Your SlideShare is downloading. ×
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Improvement of ZnO and SnO2 Hydrogen GAs Sensors
Upcoming SlideShare
Loading in...5
×

Thanks for flagging this SlideShare!

Oops! An error has occurred.

×
Saving this for later? Get the SlideShare app to save on your phone or tablet. Read anywhere, anytime – even offline.
Text the download link to your phone
Standard text messaging rates apply

Improvement of ZnO and SnO2 Hydrogen GAs Sensors

3,332

Published on

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

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

Published in: Education, Technology, Business
0 Comments
2 Likes
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total Views
3,332
On Slideshare
0
From Embeds
0
Number of Embeds
1
Actions
Shares
0
Downloads
226
Comments
0
Likes
2
Embeds 0
No embeds

Report content
Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
No notes for slide

Transcript

  • 1. Republic of IraqMinistry of Higher Education& Scientific ResearchUniversity of BaghdadCollege 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 ByDr. Abdulla M. Suhail Dr. Wasan R. Saleh2011 A D 1432A H I
  • 2. Supervisor certification We certify that this thesis was prepared by Mr. Qahtan Ghatih Hialunder our supervision at the Physics Department, College of Science,University of Baghdad as a partial requirement for the degree of doctor ofphilosophy in Physics.Signature: Suhail Signature: WasanName: Abdulla M. Suhail Name: Wasan R. SalehTitle: Assist. Professor Title: Assist. ProfessorAddress: College of Science, Address: College of Science,University of Baghdad University of BaghdadDate: November , 2011 Date: November , 2011 In view of the available recommendation, I forward this thesis fordebate by the Examining Committee.Signature: RaadName: Dr. Raad M. S. Al-HaddadTitle: ProfessorAddress: Collage of Science, University of BaghdadDate: November 29, 2011 II
  • 3. Examination Committee Certification We certify that we have read the thesis entitled “Improvement of ZnOand SnO2 Hydrogen Gas Sensors” as an examining committee, examined theStudent “Qahtan Ghatih Hial” in its contents, and that in our opinion it meetsthe standard of a thesis for the degree of Doctor of Philosophy in Phys-ics/Optoelectronics. Signature: Raad Title: Professor Name: Dr. Raad M. S. Al-Haddad Date: November 29, 2011 ChairmanSignature:Title: Professor Signature: Emad Title: ProfessorName: Dr. Izzat M. AL-Essa Name: Dr. Emad Kh. Al-ShakarchiDate: November , 2011 Date: November 27, 2011Member MemberSignature:M. B. Q. Signature:BassamTitle: Assist. Professor Title: Assist. ProfessorName: Dr. Mayada Bedry Al-Quzweny Name: Dr. Bassam Ghalib RasheedDate: November 28, 2011 Date: November 29, 2011Member Member Signature: Suhail Signature:WasanTitle: Assist. Professor Title: Assist. ProfessorName: Abdulla M. Suhail Name: Wasan R. SalehDate: November , 2011 Date: November , 2011Supervisor Co-Supervisor Digitally signedApproved by the Dean of college of Science Qahta by Qahtan IliyaSignature:Salih M. Ali DN: cn=Qahtan IliyaTitle: ProfessorName: Dr. Saleh Mahdi AliThe Dean of the College of Science n Iliya Date: 2012.05.19 15:07:50 -0700Date: December 5, 2011 III
  • 4. ABSTRACT Spray – pyrolyzed palladium – doped metal oxides (zinc oxideZnO 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 temperaturewith 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 techniquehas proven its simplicity and reliability in realizing polycrystalline innature ZnO films which crystallized along the (002) phase with preferen-tial orientation along the c – axis of the ZnO hexagonal wurtzite structureas 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 ultravioletcut – off at approximately 380 nm. The transmission but not the estimateddirect band gap Eg increased with decreasing film thickness. ScanningElectron Microscope SEM and Atomic Force Microscope AFM surfacemorphology studies of the ZnO films reveal a uniform distribution ofporous spherical – shaped nanostructure grains of 20 nm diameter. Theelectrical characterization of the sprayed thin films shows that they arehighly resistive, but that their properties vary considerably when themeasurements are conducted in vacuum or in air. For both ZnO and SnO2 metal oxides, the doped sensor exhibit anincrease of the conductance upon exposure to hydrogen gas of variousconcentrations and at different operating temperatures, showing excellentsensitivity. IV
  • 5. It was found that the sensing mechanism of hydrogen gas in thepresent metal oxide sensors is mostly related to the enhancement of ad-sorption of atmospheric oxygen. The excellent selectivity and the highsensitivity for hydrogen gas can be achieved by surface promotion ofZnO/SnO2 metal oxide films. The observed conductance change in Pd –doped ZnO sensors after exposure to H2 gas (3%) is about two times aslarge as that in the undoped ZnO sensors. The variation of the operating temperature of the film has led to asignificant change in the sensitivity of the sensor with an ideal operatingtemperature 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 gasis characterized to be relatively extremely short. ZnO thin films of 20 –time dipping in palladium chloride solution have the highest sensitivity of97% and extremely short response time of 3 s, which fit for practice sinceit is crucial to get fast and sensitive gas sensor capable of detecting toxicand flammable gases well below the lower explosion limit (4% by vol-ume for H2 gas). For SnO2 sensing elements, the optimum operating temperature isaround 210 0C and 95.744 % sensitivity to 4.5% H2: air mixing ratio. V
  • 6. Dedicated toAll Those Who Care… Including… Her VI
  • 7. Acknowledgments It would be impossible to express my thanks on this page to all those whohave 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 methe opportunity to work on a challenging and interesting project over the past threeyears 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 utmostvaluable 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 forthe 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 materialsupport provided. Also, the great help of the XRD, AFM at the Material Physics &Chemistry Research Establishment labs at the Ministry of Science and Technology areacknowledged. This thesis would not have been possible without their willingness towork 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 skinnedknees and shaken confidence. They read me my first book, and never failed to callwhen my studies overwhelmed me. They gave me the ground I could stand on when-ever the path ahead seemed dim. No son (or brother) could ask for better. VII
  • 8. Curriculum VitaeJuly 1, 1972 ...................................................................................................Born – Iraq.1994 ..................................................... B.Sc., Physics/Physics – Baghdad University1997........................................ M.Sc., Physics/Laser Technology – Baghdad University2002 – 2007.........................................................Assist. Lecturer – Physics Department2007 – October 31, 2011 .............................. Ph. D. Postgraduate – Physics DepartmentPUBLICATIONSJournal Articles [1] Q. G. Al-zaidi, Abdulla. M. Suhail, Wasan R. Al-azawi, Palladium – doped ZnO thin film hydrogen gas sensor, Applied Physics Research Vol. 3, No. 1, pp. 89 – 99, (2011). [2] H. A. Thjeel, A. M. Suhail, A. N. Naji, Q. G. Al-zaidi , G. S. Muhammed, and F. A. Naum, Fabrication and characteristics of fast photo response high responsivity ZnO UV detector, Sensors and Actuators A: Physicsl, revised manuscript submitted for publication. [3] A. M. Suhail, O. A. Ibrahim, H. I. Murad, A. M. Kadim and Q. G. Al-zaidi, Enhancement of white light generation from CdSe/ZnS core – shell system by adding organic pyrene molecules, Journal of Luminescence, revised manuscript submitted for publication. :‫إِنِّي رأَيت أَنَّهُ الَ يَكتُب أَحد كتَابا ً فِي يَومه إِالّ قَال فِي غده‬ َِِ َ ِ ِ ِ ٌ َ ْ ُ َ ".‫"لَو غيِّر هَذا لَكانَ أَحْ سن، ولَو زيد هَذا لَكانَ يُستَحسن، ولَو قُدم هَذا لَكان أَفضل، ولَو تُرك هَذا لَكانَ أَجْ مل‬ َ َ َ َ َ َ ْ َ َ َ ِّ َ َ ْ َ َ َ ِ َ َ َ َ َ ُ .‫وهَذا من أَعظَم العبَر، وهُو دلِيل علَى اِستِيالَء النُقص علَى جُملَة البَشَر‬ ِ ِ َ ِ ِ َ ٌ َ َ ِ ِ ِ ْ ِ َ َ ‫العمـــاد األصفهانـــي‬ Assuming that he’s not dead, Qahtan can best be reached at his “lifetime” email address of: qahtaniliya@yahoo.co.uk Mobile No.: +009647702981421 VIII
  • 9. Contents PageAbstract ........................................................................................................................ IVDedication .................................................................................................................... VIAcknowledgments...................................................................................................... VIICurriculum Vitae ...................................................................................................... VIIIList of Tables ............................................................................................................. XIIList of Figures ........................................................................................................... XIIIList of Symbols .......................................................................................................... XXChapter 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...................................................................................................... 6Chapter 2 Working Principles of Semiconductor Metal Oxide Gas Sensors ................ 8 2.1 Adsorption Mechanisms ...................................................................................... 8 2.2 Non-Stoichiometry in Semiconductors ............................................................. 11 2.3 Gas Sensor Operation: Catalysis and Adsorption ............................................. 13 2.4 Semiconductor Metal Oxide Gas Sensors ......................................................... 21 2.5 Gas Sensor Metrics ............................................................................................ 27 2.5.1 Sensitivity ............................................................................................... 27 2.5.2 Selectivity .............................................................................................. 28 2.5.3 Stability .................................................................................................. 29 2.5.4 Response and Recovery Times ............................................................... 30 2.6 Sensing Mechanism ........................................................................................... 31 2.7 Factors Influencing the Performance ................................................................ 33 2.7.1 Long term effects / Baseline Drift .......................................................... 33 2.7.2 Sensor surface poisoning ........................................................................ 34 2.8 Optimization of Sensor Performance ................................................................ 34 2.8.1 Use of Catalyst ........................................................................................ 34 IX
  • 10. 2.8.1.1 Spill over Mechanism ................................................................ 36 2.8.1.2 Fermi Energy Control ................................................................ 37 2.8.2 Grain size effects .................................................................................... 39 2.8.3 Thickness dependence ............................................................................ 40 2.8.4 Temperature Modulation ........................................................................ 41 2.8.5 Filters for Selectivity............................................................................... 42 2.8.6 AC and DC measurements ...................................................................... 43 2.9 Zinc Oxide ......................................................................................................... 45 2.9.1 Properties of Zinc Oxide ......................................................................... 45 2.9.2 Defects chemistry.................................................................................... 49 2.9.3 Spray pyrolysis deposition technique ..................................................... 55 2.9.3.1 The deposition process and atomization models ........................ 59 2.9.3.2 Deposition parameters ................................................................ 63 I. Substrate temperature............................................................... 63 II. Influence of Precursors ............................................................ 63 III. Spray Rate ................................................................................ 64 IV. Other Parameters ...................................................................... 65 2.9.4 Metal Oxide Gas Sensors ........................................................................ 65Chapter 3 Experimental Procedure .............................................................................. 78 3.1 Gas Sensor Fabrication ...................................................................................... 78 3.2 Spray pyrolysis experimental set up .................................................................. 80 3.3 Precursor solution .............................................................................................. 81 3.4 The determination of film thickness .................................................................. 83 3.5 Surface modification of ZnO by palladium noble metal ................................... 84 3.6 Al Interdigitated Elecrtodes (IDE) .................................................................... 85 3.7 Gas sensor testing system .................................................................................. 87 3.8 Sensor testing protocol ...................................................................................... 89 3.9 Crystalline structure of the prepared ZnO thin films ........................................ 91 3.10 Thin film surface topography............................................................................ 92 3.11 Optical properties ............................................................................................. 92 3.12 Tin oxide (SnO2) hydrogen gas sensor.............................................................. 94 X
  • 11. Chapter 4 Results and Discussion ................................................................................ 95 4.1 ZnO thin film deposition ................................................................................... 95 4.2 Crystalline structural properties of the ZnO thin film ....................................... 98 4.3 Surface topography and morphology studies ................................................. 100 4.4 Optical Properties ............................................................................................ 105 4.5 Electrical Properties ........................................................................................ 108 4.5.1 Resistance – Temperature Characteristic .............................................. 108 4.5.2 I – V characteristic of the zinc oxide films ........................................... 110 4.5.3 AC Impedance Spectroscopy ................................................................ 112 4.6 Gas Sensing Measurements ............................................................................. 114 4.6.1 Sensing Characteristics of Pure ZnO towards hydrogen gas ................ 114 4.6.2 Sensing characteristics of Pd – doped ZnO towards hydrogen gas ...... 118 4.7 Operation temperature of the sensor ............................................................... 119 4.8 Tin oxide (SnO2) hydrogen gas sensor ............................................................ 122 4.8.1 Crystalline structure and morphology of undoped SnO2 thin film ....... 122 4.8.2 Optical properties of the undoped tin oxide SnO2 thin films ................ 125 4.8.3 Sensing characteristics of pure SnO2 towards hydrogen gas ................ 126 4.8.4 Sensing characteristics of Pd – doped SnO2 towards hydrogen gas ..... 128 4.9 Conclusions and Future work Proposals ......................................................... 132 References ....................................................................................................... 135 XI
  • 12. List of TablesTable page1. Table 1.1 ----------------------------------------------------------------------- 5 Examples of application for gas sensors and electronic noses2. Table 2.1 ----------------------------------------------------------------------- 9 Comparison of physisorption and chemisorption3. 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 films8. Table 4.1 --------------------------------------------------------------------- 96 Spray pyrolysis deposition optimum parameters9. Table 4.2 -------------------------------------------------------------------- 100 Crystalline structure, Miller indices and d spacings of the as – deposit- ed ZnO crystal planes10.Table 4.3 -------------------------------------------------------------------- 100 Crystalline structure, Miller indices and d spacings of the Pd – doped ZnO crystal planes. XII
  • 13. LIST OF FIGURES Figure Page1. Figure 2.1 -------------------------------------------------------------------- 15 Microstructure and energy band model of a gas sensitive SnO2 thick film. The potential barriers form as a result of oxygen adsorption.2. Figure 2.2 -------------------------------------------------------------------- 17 The nature of oxygen species adsorbed on ZnO as reported by several researchers.3. Figure 2.3 -------------------------------------------------------------------- 18 The energy barriers in the transformation from reactants (A + B) to products (C + D). The uncatalyzed reaction is characterized by a large activation energy (Eg), while the barrier to product formation is low- ered (Ec) when a catalyst is used.4. Figure 2.4 -------------------------------------------------------------------- 22 Schematic view of gas sensing reaction in (a) Compact layer and (b) Porous layer. a: grain boundary model. b: open neck model. c: closed neck model.5. Figure 2.5 ------------------------------------------------------------------- 23 Schematic of a compact layer with geometry and energy band repre- sentation; Z0 is the thickness of the depleted surface layer; Zg is the thickness of the surface and eVS the band bending. (a) A partly deplet- ed compact layer (“thicker”) and (b) A completely depleted layer (“thinner”).6. Figure 2.6 ------------------------------------------------------------------- 24 Schematic representation of a porous sensing layer with geometry and energy band for small and large grains. λD Debye length, Xg grain size.7. Figure 2.7 -------------------------------------------------------------------- 25 Schematic of a porous layer with geometry and surface energy band with necks between grains; Zn is the neck diameter; Z0 is the thickness of the depletion layer and eVS the band bending. (a) a partly depleted necks and (b) a completely depleted necks.8. Figure 2.8 -------------------------------------------------------------------- 26 Influence of particle size and contacts on resistances and capacitances in thin films are shown schematically for a current flow I from left to right. XIII
  • 14. 9. Figure 2.9 -------------------------------------------------------------------- 30 Drawing showing how response and recovery times are calculated from a plot of sensor conductance versus time.10.Figure 2.10 ------------------------------------------------------------------- 35 Illustration of the catalyst effect. Nano – particles, having higher sur- face area, act as catalysts. Here, R stands for reducing gas.11.Figure 2.11 ------------------------------------------------------------------- 36 Mechanism of sensitization by metal or metal oxide additive.12.Figure 2.12 ------------------------------------------------------------------- 37 Illustration of Spill Over caused by catalyst particles on the surface of the grain of the polycrystalline particle.13.Figure 2.13 ------------------------------------------------------------------- 38 An adequate dispersion of the catalysts is required in order to effec- tively affect the grains of the semi-conducting material to serve the implied purpose of increase in sensitivity.14.Figure 2.14 ------------------------------------------------------------------- 39 Schematic models for grain – size effects.15.Figure 2.15 ------------------------------------------------------------------- 44 Equivalent circuit for the different contributions in a thin film gas sen- sor; intergranular contact, bulk and electrode contact.16.Figure 2.16 ----------------------------------------------------------------- 46 T – X diagram for condensed Zn- O system at 0.1 MPa.17.Figure 2.17 ------------------------------------------------------------------- 47 Many properties of zinc oxide are dependent upon the wurtzite hexag- onal, close-packed arrangement of the Zn and O atoms, their cohe- siveness and void space [59].18.Figure 2.18 ------------------------------------------------------------------- 49 Ellingham diagram of oxides.19.Figure 2.19 ------------------------------------------------------------------- 50 Various types of point defects in crystalline materials.20.Figure 2.20 ------------------------------------------------------------------- 60 Schematic diagram of chemical spray pyrolysis unit21.Figure 2.21 ------------------------------------------------------------------- 62 XIV
  • 15. Spray processes (A, B, C, and D) occurring with increase in substrate temperature.22.Figure 3.1 -------------------------------------------------------------------- 78 Schematic of a typical gas sensor structure.23.Figure 3.2 -------------------------------------------------------------------- 81 Spray pyrolysis experimental set up.24.Photo plate 3.1 -------------------------------------------------------------- 82 A: experimental set up of the spray pyrolysis deposition SPD. B: Air atomizer. C: Gemo DT109 temperature controller, and D: Digital bal- ance with the magnetic stirrer.25.Figure 3.3 -------------------------------------------------------------------- 85 Vacuum system for the vaporization from resistance – heated sources. When replacing the transformer and heater with an electron gun, va- porization by means of an electron beam occurs.26.Figure 3.4 -------------------------------------------------------------------- 86 A schematic diagram of the IDE masks utilized in this work.27.Figure 3.5 -------------------------------------------------------------------- 87 Gas sensor testing system.28.Photo plate 3.2 -------------------------------------------------------------- 88 A photo of the sensor testing system.29.Figure 3.6 -------------------------------------------------------------------- 91 A schematic diagram of the gas sensor basic measurement electrical circuit.30.Photo plate 3.3 -------------------------------------------------------------- 92 LabX XRD – 6000 Shimadzu diffractometer unit.31.Photo plate 3.4 -------------------------------------------------------------- 93 Ultra 55 SEM unit from ZEISS.32.Photo plate 3.5 -------------------------------------------------------------- 93 AA3000 Scanning Probe Microscope SPM, tip NSC35/AIBS, from Angstrom Advance Inc.33.Photo plate 3.6 -------------------------------------------------------------- 94 Optima sp-3000 plus UV-Vis-NIR spectrophotometer.34.Figure 4.1 -------------------------------------------------------------------- 96 XV
  • 16. A photo of spray pyrolyzed ZnO thin film on glass samples.35.Figure 4.2 -------------------------------------------------------------------- 97 Scanning Electron Micrograph photo of spray pyrolyzed ZnO thin film on glass.36.Figure 4.3 -------------------------------------------------------------------- 97 Enlarged photos of Al interdigitated electrodes IDE evaporated on ZnO thin film sample. A: 1 – mm finger spacing IDE on glass, and B: 0.4 – mm finger spacing IDE on silicon.37.Figure 4.4 -------------------------------------------------------------------- 99 XRD crystal structure of as deposited ZnO thin film prepared from 0.1 M Zinc Chloride aqueous precursor.38.Figure 4.5 -------------------------------------------------------------------- 99 XRD crystal structure of Pd – doped ZnO thin film prepared from 0.1 M Zinc Chloride aqueous precursor.39.Figure 4.6 ------------------------------------------------------------------- 101 Scanning Electron Micrograph of ZnO film prepared at a) 400 0C and the inset b) 200 0C.40.Figure 4.7 ------------------------------------------------------------------- 102 Scanning Probe Microscope images of zinc oxide thin film spray py- rolysed on glass substrate at 450 0C spraying temperature with the precursor of 0.2 M zinc acetate dissolved in 100 mL distilled water.41.Figure 4.8 ------------------------------------------------------------------- 103 Scanning Probe Microscope images of zinc oxide thin film spray py- rolysed on glass substrate at 450 0C spraying temperature with the precursor of 0.2 M zinc acetate dissolved in 100 mL isopropyl alcohol.42.Figure 4.9 ------------------------------------------------------------------- 104 Granularity cumulation distribution report of ZnO thin film deposited at 450 0C on glass substrate using 0.2 M zinc acetate in distilled water precursor solution.43.Figure 4.10 ------------------------------------------------------------------ 105 Transmission spectra of ZnO thin films of different thicknesses sprayed on – glass at 400 0C temperature. The precursor was 0.1 M dissolved in distilled wa-ter except the 189.34 – nm thick sample which was a 0.2 M zinc acetate dissolved in 3:1 volume ratio isopro- pyl alcohol and distilled water. XVI
  • 17. 44.Figure 4.11 ------------------------------------------------------------------ 106 Absorption spectra of ZnO thin films of different thicknesses sprayed on – glass at 400 0C temperature. The precursor was 0.1 M zinc ace- tate dissolved in distilled water.45.Figure 4.12 ------------------------------------------------------------------ 107 Plots of (αhν)2 vs. photon energy hν for ZnO thin films of different energy gaps and thicknesses.46.Figure 4.13 ------------------------------------------------------------------ 108 Relationship of energy gap Eg of sprayed ZnO thin films with film thickness.47.Figure 4.14 ------------------------------------------------------------------ 109 The variation of resistance of the spray – pyrolyzed deposited zinc ox- ide film of 668 nm film thickness with temperature.48.Figure 4.15 ------------------------------------------------------------------ 110 The I – V characteristic in dark and under UV illumination.49.Figure 4.16 ------------------------------------------------------------------ 111 The effect of vacuum on base line current of a ZnO thin film at 200 0C and 10 v bias voltage.50.Figure 4.17 ------------------------------------------------------------------ 112 The I – V characterization of sprayed ZnO film in the temperature range from RT to 300 0C.51.Figure 4.18 ------------------------------------------------------------------ 113 The Cole – Cole plot for the impedance spectrum of the films at room temperature. The inset is the R-C equivalent circuit of the simulation of the impedance spectrum.52.Figure 4.19 ------------------------------------------------------------------ 114 Sensing behavior of ZnO thin film at 6 v bias voltage and 210 degrees temperature to traces of hydrogen reducing gas mixing ratio in air of 3%, 2%, and 1% respectively.53.Figure 4.20 ------------------------------------------------------------------ 115 The sensitivity dependence of as – deposited ZnO sensor on hydrogen gas mixing ratio.54.Figure 4.21 ------------------------------------------------------------------ 116 XVII
  • 18. Transient responses of ZnO thin film (245 nm thick) at 210 0C testing temperature upon exposure to hydrogen gas of mixing ratios of 1%, 2%, and 3% respectively.55.Figure 4.22 ------------------------------------------------------------------ 117 Response and recovery time of the sensor as a function of testing gas mixing ratio at a testing temperature of 210 0C and bias voltage of 6 v.56.Figure 4.23 ------------------------------------------------------------------ 117 I - V characteristics of undoped ZnO gas sensor to 5%, 3%, and 1% Hydrogen gas mixture in air and at 200 degrees temperature.57.Figure 4.24 ------------------------------------------------------------------ 118 The switching behavior of the Pd – sensitized ZnO thin film maximum con-ductance to hydrogen of 3% H2:air mixing ratio at 200 0C and bias voltage of 10 v.58.Figure 4.25 ------------------------------------------------------------------ 119 Effect of the testing temperature on the Pd – sensitized ZnO thin film maximum conductance to hydrogen of 3% H2:air mixing ratio and bias voltage of 10 v.59.Figure 4.26 ------------------------------------------------------------------ 121 The variation of sensitivity with the operating temperature of the Pd – doped ZnO gas sensor.60.Figure 4.27 ------------------------------------------------------------------ 122 Transient responses of Pd – sensitized ZnO thin film (668 nm thick) as exposed to hydrogen gas of mixing ratio of 3% and at three different testing temperatures of (1) 250, (2) 300, and (3) 350 0C successively.61.Figure 4.28 ------------------------------------------------------------------ 123 X – ray diffraction (XRD) pattern of SnO2 thin film spray pyrolyzed on glass substrate at temperature of 450 0C.62.Figure 4.29 ------------------------------------------------------------------ 124 AFM image of undoped SnO2 thin film deposited at 450 0C on glass substrate with the precursor being tin dichloride dehydrate dissolved in isopropyl alcohol.63.Figure 4.30 ------------------------------------------------------------------ 125 Transmission spectra of undoped SnO2 thin films of different thick- nesses deposited at 450 0C on glass substrates.64.Figure 4.31 ------------------------------------------------------------------ 126 XVIII
  • 19. Absorption coefficient versus the photon energy for energy gap esti- mation of undoped SnO2 thin films of different thicknesses deposited at 450 0C on glass substrates.65.Figure 4.32 ------------------------------------------------------------------ 127 Sensitivity behavior of undoped tin oxide SnO2 thin film to different hydrogen concentrations. The bias voltage was 5.1 v with the tempera- ture set to 210 0C.66.Figure 4.33 ------------------------------------------------------------------ 127 Sensitivity versus H2 gas concentration of undoped tin oxide SnO2 thin film. The bias voltage was 5.1 v with the temperature set to 210 0C.67.Figure 4.34 ------------------------------------------------------------------ 128 Sensing behavior of Pd – doped SnO2 gas sensor to different H2 : air mixing ratios. The tests were performed at 210 0C temperature and 10 v bias.68.Figure 4.35 ------------------------------------------------------------------ 129 Response transient of Pd – doped SnO2 gas sensor to different H2 : air mixing ratios. The tests were performed at 210 0C temperature and 10 v bias.69.Figure 4.36 ------------------------------------------------------------------ 130 Sensitivity and Response time as a function of the H2 test gas mixing ratio. The test was performed at 210 0C and 10 v bias on SnO2 sample sprayed over the IDE and surface coated with 20 PdCl2 layers sprayed at 400 0C over the film.70.Figure 4.37 ------------------------------------------------------------------ 130 Transient responses of SnO2 thin film of 248 nm thick at 150, 175, and 210 0C testing temperature upon exposure to1% H2:air gas mixing ra- tio.71.Figure 4.38 ------------------------------------------------------------------ 131 Variation of sensor response current with temperature of Pd - doped SnO2 thin film exposed to 4.5% hydrogen gas mixing ratio in air and at 10 v bias voltage. XIX
  • 20. LIST OF SYMBOLS ε Static dielectric constant λ Wavelength nm μ (electron) mobility cm2.s-1.v-1 ρ Density kg/m3 σ Conductivity Ω.cm τrec Recovery time s τres Response time s Ω Ohm qVS Surface Potential Barrier eVAFM Atomic Force MicroscopeCVD Chemical Vapor Deposition CSP Chemical Spray PyrolysisDMM Digital Multi MeterENC 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 dioxideNTCR Negative Temperature Coefficient of Resistance PID Proportional–Integral–Derivative Controller ppm Parts Per MillionPTCR Positive Temperature Coefficient of Resistance R Resistance (electrical) Ω S Siemenssccm Standard Cubic Centimeter per MinuteSEM Scanning Electron MicroscopeSMO Semiconductor Metal Oxide t90 Time to accomplish 90% of sensor response change sTEM Transmission Electron MicroscopeVOC Volatile Organic CompoundXRD X-Ray Diffraction Z0 Depletion Region nm XX
  • 21. Chapter 1Motivation and Project ObjectivesIntroduction The purpose of this chapter is to provide a general framework andintroduction for the work presented in the current Ph.D. project. Thischapter 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 inresponse to a specified input stimulus [1]. This stimulus can be a physicalstimulus like temperature and pressure or a concentration of a specificchemical or biochemical material. The output signal is typically an elec-trical signal proportional to the input variable, which is also called themeasurand. Sensors can be used in all three phases of matter although gasand liquid sensors are the most common. The presence of a reducing/oxidizing gas at the surface of certainmetal oxide semiconductors changes its electrical resistance R. It is thisphenomenon that has spurred the use of these materials in the detection ofa gaseous ambient. The theoretical basis for semiconductor gas sensorsarose in 1950, when Carl Wagner proposed a concept to explain the de-composition of nitrous oxide (N2O) on zinc oxide (ZnO) [2]. He made thenovel assumption that an exchange of electrons was taking place betweenthe gaseous N2O and the solid ZnO, which possessed a layer of adsorbedoxygen. A few years later, Brattain et al. found that ambient gas producedchanges in potential between an electrode and a germanium surface [3].These findings were explained in a theory outlining the existence of do- 1
  • 22. nor and acceptor traps that lead to the generation of a space charge layeron the surface of the germanium. A working gas sensor was realized in1962, 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]. Itwas in that same year that Naoyoshi Taguchi issued a patent for a gassensor based on tin oxide (SnO2) [5]. As such, gas sensors based on SnO2are 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 devicewhose resistance in air is very high and drops when exposed to reducinggases such as combustibles (H2, CO, CH4, C3H8) or volatile organic va-pors and it has enjoyed a substantial popularity because of its ease offabrication, low cost, robustness, and their sensitivity to a large range ofreductive and oxidative gases [7]. In addition to research on understand-ing the fundamentals of the sensing mechanism, the studies on ZnO andSnO2 sensors have been directed on enhancing the sensor performancethrough the addition of noble metals (Pt, Pd, etc.), synthesis of thick andthin film sensors, and doping with other semiconductors [7-10]. Othermetal oxides such as Fe2O3, TiO2, WO3 and Co3O4 have also been used asgas sensors. Despite these broad studies in the semiconductor sensor area,problems such as insufficient gas selectivity, slow response and recoverytimes, inability to detect very low gas concentrations, and degradation ofthe sensor performance by surface contamination still persist. Thus, thereis 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 adepletion layer (for n-type semiconductors), and hence reduces conduc-tivity [11]. Here, ionosorption refers to the process where a species is 2
  • 23. adsorbed and undergoes a delocalized charge transfer with the metal ox-ide. This can then be used to measure reducing and oxidizing gases, asthey will change the amount of ionosorbed oxygen, and therefore theconductivity 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 lowerresponse time for the gas sensor. However, the physical properties of themetal oxides place an upper limit on the temperatures that can be used. Ifthe temperature is too high, the stability and reliability of the sensorsdiminishes because of possible coalescence and structural changes [12].Furthermore, as temperature increases, the charge – carrier concentrationwill increase and the Debye length, LD, will decrease, resulting in lesssensitivity [13]. In most cases, the optimal temperature for metal oxidegas sensors is between 200 0C and 500 0C [17]. There are two well – known ways for improving the gas sensingproperties 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 tobetter sensitivity and stability, e.g. [14, 15]. The second is to reduce grainsize, which has been shown to increase sensitivity [16]. This is becausethe depletion layer caused by ionosorption has a greater effect on theconduction channel of the grain as the grain size decreases. Consequently,there is great interest in using nanoparticles in gas sensors, since they canbe used to make films with very small grain sizes. These two approacheshave 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 ZnOsensing element with palladium has greatly enhanced the sensor sensitivi- 3
  • 24. ty and response time of minute – grain size ZnO thin films made possiblethrough 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 theenvironment has grown. Gas sensors find applications in numerous fields[17, 18]. Two important groups of applications are the detection of singlegases (as NOx, NH3, O3, CO, CH4, H2, SO2, etc.) and the discriminationof odours or generally the monitoring of changes in the ambient. Singlegas sensors can, for examples, be used as fire detectors, leakage detectors,controllers of ventilation in cars and planes, alarm devices warning theovercoming of threshold concentration values of hazardous gases in thework places. The detection of volatile organic compounds (VOCs) orsmells 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 gassensing devices designed to analyze such complex environmental mix-tures [19]. In Table 1.1 [17] examples of application for gas sensors andelectronic noses are reported. Industry currently employs many varieties of gas sensing systemsfor monitoring and controlling emissions from their processes. Applica-tions exist in the steel, aluminum, mineral, automotive, medical, agricul-tural, aroma and food industries. Analytical instruments, based on optical spectroscopy and electro-chemistry, are widely used in the scientific community. These instru-ments give precise analytical data, however are costly, slow, and cumber- 4
  • 25. some and require highly qualified personnel to operate. Current trends areto improve low cost, solid state gas sensor performance in order to obtainhigh linearity, sensitivity, selectivity and long term stability [19]. Table 1.1: Examples of applications for gas sensors and electronic noses [17]. Applications Automobiles  Car ventilation control  Filter control  Gasoline vapour detection  Alcohol breath tests Safety  Fire detection  Leak detection  Toxic/flammable/explosive gas detectors  Boiler control  Personal gas monitor Indoor air quality  Air purifiers  Ventilation control  Cooking control Environmental control  Weather stations  Pollution monitoring Food  Food quality control  Process control  Packaging quality control (off-odours) Industrial production  Fermentation control  Process control Medicine  Breath analysis  Disease detection The only practical way to monitor air quality or provide a mean toalert a human of potential danger is by direct gas sensing. A gas sensorcan form part of an early warning system, notifying the appropriate au-thorities or provide the feedback signals to a process control system. Toachieve this, a gas sensor system must be capable of accurate and stablein-situ real time measurements. Environmental factors such as operatingtemperature, vibration, mechanical shock, chemical poisoning, as well as 5
  • 26. various device characteristics (accuracy, resolution, physical size andcost) must also be taken into consideration.1.3 Focus of current research The main focus of the present thesis is on the improvement of semiconducting metal oxide (SMO) thin film based gas sensors (with specialemphasis on SnO2 and ZnO) and their characterization. An objectiveanalysis of the various substrates used by several investigators is per-formed as a part of this work. The gas sensitive zinc oxide and tin oxidefilms are deposited by chemical spray pyrolysis deposition technique withair 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 microscopeAFM. Too much effort has been spent to maximize the sensitivity S andreduces the response τres and recovery τrec times of the sensing elementupon exposing to hydrogen reducing gas H2 of various concentrations Cand at different operating temperatures T. The catalytic effect of the pal-ladium noble metal and grain size effect are exploited to accomplish thesevital objectives. The thesis also describes the development of the gassensor 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 chapteris a general introduction where the overview of the field of study andscope of the work carried out is outlined. The second chapter entitled“Working principles of semiconductor metal oxide gas sensors” brieflydescribes the working principle of the kind of sensors developed andsurveys the various methods used currently to improve the sensor charac- 6
  • 27. teristics. The fabrication and characterization of the sensing element aswell as testing the sensors towards hydrogen reducing gas is dealt with inthe third chapter. Moreover, the development of the gas sensor testingchamber and the protocol to use it are also detailed in this chapter. Thesubsequent chapter (chapter four), a discussion of the experiments carriedout and the results obtained in the development of gas sensors is submit-ted. At the end of the chapter four, the conclusions and the scope for fu-ture work are summarized. 7
  • 28. Chapter 2Working Principles of Semiconductor Metal Oxide Gas SensorsBackground 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 sensorsbuilt using metal oxide materials. Particular emphasis is placed on theadsorption of oxygen because most sensors operate in air and oxygen isthe dominant adsorbed species in this case. The mechanism where anoxide transforms gas – surface interactions into a measurable electricalsignal is reviewed with a focus on the effects of particle size on this phe-nomenon. The current understanding of the gas sensing mechanism and abrief discussion of theoretical and empirical models proposed for semi-conductor metal oxide (SMO) gas sensors are discussed. The metrics bywhich gas sensor performance is judged are defined in this chapter and anintroduction to SMO gas sensors is presented. Background information isconcluded with a discussion of reported spray pyrolysis deposition tech-nique for oxide semiconductors.2.1 Adsorption Mechanisms Physical adsorption (physisorption) is defined as an adsorptionevent where no geometric change occurs to the adsorbed molecule andvan der Waals forces are involved in the bonding between the surface andadsorbate [11]. Chemical adsorption (chemisorption) is the formation of achemical 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 strongerbond than physisorption and the characteristics of each are summarized in 8
  • 29. Table 2.1. Two types of chemisorption occur on the surface of metal ox-ides: (1) molecular or associative chemisorption, in which all the atomicbonds are preserved in the adsorbed molecule; and (2) dissociative chem-isorption, where bonding within the adsorbed molecule decomposes andmolecular fragments or ions are bound to oxide surface. Molecular chem-isorption is the most probable type of adsorption for molecules that pos-sess free electrons or multiple bonds. Gas molecules with single bonds Table 2.1: Comparison of physisorption and chemisorption [20]. Physisorption Chemisorption Covalent Bonding Intermolecular (van der Waals) Force (adsorbate & surface) Low Temperature High Temperature Low Activation Energy High Activation Energy (<< 0.5 eV) (> 0.5 eV) Low Enthalpy Change High Enthalpy Change (ΔH < 20kJ/mol) (50kJ/mol < ΔH < 800kJ/mol) Reversible Reversible at High Temperature Adsorbate energy state unaltered Electron density increases at interface Multilayer formation possible Monolayer surface coveragetend to react via dissociative chemisorption; however; there is an activa-tion energy associated with dissociation. The type of chemisorbed oxygenon the surface of a metal oxide is dependent on the temperature of thesystem. Barsan and Weimar compiled results from a survey of the litera-ture concerning oxygen adsorption on SnO2 and correlated the adsorbedoxygen species to temperature where techniques such as infrared spec-troscopy, temperature programmed desorption and electron paramagneticresonance were used [21]. Table 2.2 summarizes the temperature rangesassociated with each species of oxygen adsorption. 9
  • 30. Table 2.2: Temperature ranges associated with molecular and dissociative oxygenadsorption reactions [20]. Temperature Range (°C) Adsorption Reaction(s) O2 (g) + e− → O− (ads) 2 Room Temp.< T < 175°C O2 (g) + 2e− → O2− (ads) 2 175°C < T < 500°C O2 (g) + 2e− → 2O− (ads) O− (ads) + e− → O2− (ads) T > 500°C O2 (g) + 4e− → 2O2− (ads) In the reaction shown in Table 2.2, (g) indicates the gaseous form,(ads) indicates the molecule or ions that are adsorbed on a surface and e-is an electron initially in the metal oxide. Oxygen interactions with thesurface of an oxide are of utmost importance in gas sensing. Oxygen is astrong electron acceptor on the surface of a metal oxide. A majority ofsensors operate in an air ambient; therefore, the concentration of oxygenon the surface is directly related to the sensor electrical properties. Theconversion to O- or O2- at elevated temperatures are useful in gas sensingsince only a monolayer of oxygen ions are present with these stronglychemisorbed 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 achievedby thermal stimulation up to a specific temperature or by reactions withother gaseous species. A desorption process that is isothermal occurswhen, for instance, a reducing gas such as carbon monoxide (CO) is in-troduced into the surrounding atmosphere. Oxygen is consumed in a reac-tion with the CO to form carbon dioxide (CO2) as written in equation 2.1. CO(g) + O− (ads) → CO2 (g) + e− (2.1) 10
  • 31. where the extra electron generated is injected back into the metal oxide.This desorption reaction results in a lower surface coverage of oxygenadsorbates which influences the electrical properties of the oxide.2.2 Non – Stoichiometry in Semiconductors The relevance of non – stoichiometry to the transport properties ofmetal oxide semiconductors will now be explored using ZnO as an exam-ple. It is well – known that ZnO is stoichiometrically deficient in oxygentraced to either zinc interstitials or oxygen vacancies. To begin with, thenotation 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 relativecharge, 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 vacancyoccupying an oxygen lattice site. Electrons and holes may be signified bye and h, respectively. The equilibrium constant, K(T), for the general chemical reactionof reactants A, B and products C, D [19]: aA + bB → cC + dD (2.2)can be written as: [C]c [D]d ∆G K(T) = = e−KT (2.3) [A]a [B]b 11
  • 32. where ΔG represents the standard change in free energy for the reaction.An oxygen vacancy, known as a type of Schottky defect, can be generatedin ZnO through the following reaction: ∙∙ 1 ZnZn + OO → ZnZn + VO + 2e′ + O2 (g) (2.4) 2It is conventional practice to denote the left side of (2.4) as “nil”. As seenin (2.4), a positively charged oxygen vacancy is compensated by the gen-eration of electrons, thus leading to n-type conductivity of ZnO. The mass action constant for (2.4) can be written as: 1 ∙∙ K R = [VO ](pO2 )2 ⋅ n2 (2.5) ∙∙where pO2 denotes the partial pressure of oxygen, [VO ] and n represent theoxygen vacancy and electron concentrations, respectively. To solve for ∙∙[VO ] or n, one must evoke the electroneutrality condition (ENC), whichstates that the concentration of positive defects present in the materialmust equal the concentration of negative defects. The ENC for (2.4) is: ∙∙ 2[VO ] = 𝑛 (2.6) ∙∙using (2.6) and solving for [VO ] or n in (2.5) yields: 1 KR 3 1 [VO ] = ( ) (pO2 )−6 ∙∙ (2.7) 4and 1 1 1 n = (2)3 (K R )3 (pO2 )−6 (2.8)Thus, the logarithmic concentration of oxygen vacancies and electrons inZnO, plotted against log (pO2 ), is shown to have what is termed a -1/6pO2 dependence. The electronic conductivity, 𝜎, is given by: 12
  • 33. 1 1 1 σ = q(2)3 (K R )3 (pO2 )−6 μe (2.9)where q is the electronic charge and 𝜇 𝑒 is the electronic mobility. An oxygen deficiency in ZnO may also be realized through theformation of a Zn interstitial, known as a type of Frenkel defect. This canbe formed through the following reaction: 1 nil = Zn⋅⋅ + 2e′ + O2 (g) i (2.10) 2using 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 largepart by the concentration of electrons or holes present in the material. Incertain 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 thebasis for semiconductor gas sensor operation. This behavior is unlikemetals, where the adsorption of a gas may cause small conductancechanges due to a modification of charge carrier mobility [6]. As an exam-ple, recall the experiments of Wagner on the decomposition of N 2O onZnO. If the generation of two electrons proceeds as in (2.4), the decom-position reaction was proposed as follows [2]: 2e− + N2 O → N2 + O2− (ads. ) (2.11) O2− (ads. ) + N2 O → N2 + O2 + 2e− (2.12) The adsorption of oxygen in (2.11) would result in an increase inZnO resistivity due to the capture of majority charge carriers. The subse- 13
  • 34. quent reaction between the adsorbed oxygen and N2O in (2.12) acts torestore the supply of conduction electrons and thus, an increase in con-ductivity may be observed. It is this simple and reversible change incharge concentration that drives the use of metal oxide gas sensors. For a visual perspective, a schematic of an n – type semiconductortin dioxide SnO2 thick film with an accompanying band structure modelis shown in figure 2.1 [19]. For conduction to occur, an electron mustpass 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 theadsorbed oxygen leave behind positively charged donor ions. An electricfield develops, between these positive donor ions and the negativelycharged adsorbed oxygen ions, which serves to impede the flow of elec-trons between neighboring grains. The barrier generated by the electricfield has a magnitude of eVS, where e is the electronic charge and VS isthe 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 barrierand reach a neighboring grain is given by: eVs ns = Nd exp (− ) (2.13) kTwhere Nd is the concentration of donors, k is Boltzmann’s constant, and Tis the temperature. Since conductance (or resistivity) is proportional to n s,an increase in the adsorbed oxygen content will raise eVS and thus, fewerelectrons will cross the potential barrier. This may be empirically moni-tored as an increase in resistivity. The introduction of a reducing gas willreverse this effect, lowering the potential barrier and decreasing resistivi-ty. It is this reducing gas, often termed the analyte, whose presence is ofinterest. 14
  • 35. 1 − Electron – depleted O region 2 2 O- O- O- O- O- O- - O O- Grain Grain (a) In air O- O- O- O- O- Thermionic emission 𝑒𝑉 𝑆1 𝑒𝑉 𝑆1 ; 𝐺1 = 𝐺0 exp(− ) 𝐾𝑇 EC EF CO2 CO - O O- O O- O- Grain Grain (b) In the presence of reducing gas O- O- O O O- O- CO CO2 CO2 CO 𝑒𝑉 𝑆2 𝑒𝑉 𝑆2 ; 𝐺2 = 𝐺0 exp(− ) 𝐾𝑇 𝐺2 𝑒 𝑉 𝑆1 − 𝑉 𝑆2 𝑠𝑒𝑛𝑠𝑖𝑡𝑖𝑣𝑖𝑡𝑦 ⇉ 𝑆 = = exp ( ) 𝐺1 𝐾𝑇Figure 2.1: Microstructure and energy band model of a gas sensitive SnO2 thick film.The potential barriers form as a result of oxygen adsorption [19]. SnO2 15
  • 36. Using an example of ZnO in the detection of hydrogen, the followingreactions may occur [24]: 2e− + O2 → 2O− (ads. ) (2.14) 2O− (ads. ) + 2H2 → 2H2 O + 2e− (2.15) The reaction of H2 with adsorbed oxygen on the surface of ZnO(2.15) will result in a measurable reduction in the resistivity. It should benoted that adsorbed oxygen may exist in multiple forms. Takata et al.proposed that oxygen adsorbed on ZnO is transformed with increasingtemperature in the following manner [25]: O2 → O− → 2O− → 2O2− 2 (2.16)of these forms, O2 is considered fairly inactive due to its non – dissocia-tive state. With regards to O− and O−, electron spin resonance (ESR) 2studies have shown that O− is far more reactive than O− [26]. The nature 2of adsorbed oxygen on ZnO as reported by several researchers is shownin 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), ZnOacts as a heterogeneous catalyst, as it is a phase distinct from the reactantsand products. To illustrate the phenomenon of catalysis, consider thefollowing reaction: A+B→ C+D (2.17) Two possible paths in which this reaction may proceed are shownin figure 2.3 [27]. In the absence of a catalyst, the reaction of (2.17) ischaracterized by a large activation energy, Ea. When a catalyst such asZnO or SnO2 is used, the gaseous products adsorb onto the metal oxide 16
  • 37. Figure 2.2: The nature of oxygen species adsorbed on ZnO as reported by severalresearchers [25].surface with an exothermic heat of adsorption ΔH (State I). The reactionto form adsorbed products then proceeds with a lower activation energyEc (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 thereaction to proceed. An undesirably low activation energy will cause thereaction 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 subjectto an attractive potential [28]. This potential is the origin of adsorptionand arises from the multitude of unsatisfied bonds that exist at the surfaceof the solid. The adsorbed species is often called the adsorbent and thesolid surface is termed the adsorbate [11]. Physical adsorption, or phy-sisorption, occurs as a result of electrostatic and van der Waals forces thatexist between the adsorbent and adsorbate. Heats of adsorption for phy-sisorption tend to be low, with ΔH values typically in the range of 2 – 15kcal/mole [29]. In the case of chemical adsorption, (termed chemisorp- 17
  • 38. EaEnergy A+B ΔH Ec II I C+D Reaction coordinateFigure 2.3.The energy barriers in the transformation from reactants (A + B) to products(C + D). The uncatalyzed reaction is characterized by a large activation energy (E a),while the barrier to product formation is lowered (Ec) when a catalyst is used [27].tion), the adsorbent forms a chemical bond with the solid surface. Valuesfor ΔH tend to be higher for chemisorption and are often in the range of15 – 200 kcal/mole [29]. As chemisorption tends to provide the necessarycatalysis 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 forgas detection, then changes in conductivity must be proportional to theconcentration of the gaseous analyte. To understand this relationship,adsorption kinetics must be discussed. The residence time, of an ad-sorbed atom is given by [28]: ∆H = exp ( ) (2.18) T 18
  • 39. where is related to surface vibration time and R is the universal gasconstant. The surface coverage, S, of a gaseous species is dependent onboth 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 tothe pressure through the kinetic theory of gases will yield: N ∆H =( ) exp ( ) (2.20) √2 T Twhere NA is Avogrado’s number, P is the partial pressure of the gas, andM is the average molar weight of the gaseous species. Experimentalcurves of S plotted as a function of P at a given temperature are known asadsorption isotherms [28]. One particular isotherm derived by IrvingLangmuir 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 realsurfaces, modified isotherms have been developed [29]. Regardless, thederivations of Langmuir provide a sound qualitative relationship betweensurface 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 completelycovered surface, then a concentration of S adsorbed molecules will resultin So – S available sites [28]. The fluxes for both reflected molecules, FR,and adsorbed molecules, FA, are given by: R =( ) = (1 − ) (2.21)substitution of FA into (2.19) and subsequent rearrangement will yield: 19
  • 40. a = = (2.22) + +a The constant a is comprised of the grouping of terms, with the ex-ception of P, from (2.20). If θ = (S/So), where θ is defined as the degreeof coverage, then (2.22) takes the following form: b = (2.23) 1+bwhere b = (a/So). When the degree of surface coverage is proportional to the partialpressure, changes in the electrical conductivity may be related to gasconcentration. 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 ofproportionality makes the gas sensor insensitive to coverage. If a compe-tition ensues for surface sites between two gas species, A and B, then(2.23) becomes [27]: b b = = (2.24) 1+b +b 1+b +b If bBPB >> bAPA, then the equations of (2.24) become: b b = = (2.25) 1+b 1+bin the case that bBPB is >> 1, the equations of (2.25) reduce to: b 1 (2.26) 1+bThus, 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 willbecome independent of the gas concentration as approaches 1. The rate of a reaction between A and B may be given by [27]: ae=k (2.27) 20
  • 41. If the rate in (2.27) is much higher than the rate of adsorption of say, A,then will fall to zero and will increase. As an example, note that thecoverage 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 quitehigh. As the reaction rate between the oxygen and the reducing speciesincreases, falls to zero, enabling the reducing gas to be detected with ahigh 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 theirgas 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 layeris 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 usedmetal – oxides are n – or p – type semiconductors, due to the presence ofoxygen – vacancies in the bulk. Generally the conductance or the re-sistance of the sensor is monitored as a function of the concentration ofthe target gases. Additionally the performance of the sensor depends onthe 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. Thesensing itself can take place at different sites of the structure dependingon the morphology. They will play different roles, according to the sens-ing layer morphology. An overview is given in figure 2.4. 21
  • 42. Gas Product Gas Product desorption adsorp- tion (a) Compact layerGas Product B C A Sensitive layer (b) Porous layer Electrodes Substrate Figure 2.4: Schematic view of gas sensing reaction in (a) Compact layer and (b) Po- rous layer. A: grain boundary model. B: open neck model. C: closed neck model [32] A simple distinction can be made between:  Compact layers; the interaction with gases takes place only at the geometric surface (Figure 2.5, such layers are obtained with most of the techniques used for thin film deposition such as pulsed laser deposition, sputtering etc.) and  Porous layers; the volume of the layer is also accessible to the gas- es and in this case the active surface is much higher than the geo- metric one (Figure 2.6, such layers are characteristic to thick film techniques and RGTO (Rheotaxial Growth and Thermal Oxida- tion). For compact layers, gases only interact with metal oxides at geo- metrical surfaces, resulting in a surface depletion layer through the film either partly (a) or completely (b) (Figure 2.5). Whether the sensor oper- ates under a partly or a completely depleted, the condition is determined 22
  • 43. by the ratio between layer thickness Zg and Debye length λD (or LD). For partly depleted layers, when surface reactions do not influence the con- duction in the entire layer (Zg >Z0 see Figure 2.5), the conduction process takes place in the bulk region (of thickness Zg −Z0, much more conduc- tive than the surface depleted layer). Formally two resistances occur in parallel, one influenced by sur- Gas Product Surface band bending Volume not accessible (a) Z eVS to gases Z0 Current flow Zg Conducting channel Z Energy X eVS (b) ZFigure 2.5: schematic of a compact layer with geom- e∆VS etry and energy band representation; Z0 is the thickness of the depleted surface lay- Zg er; Zg is the thickness of the surface and eVS the band bending. (a) A partly deplet- ed compact layer (“thicker”) and (b) A Eb Energy completely depleted layer (“thinner”) [32]. face reactions and the other not; the conduction is parallel to the surface, and this explains the limited sensitivity. Such a case is generally treated as a conductive layer with a reaction-dependent thickness. For the case of completely depleted layers in the absence of reduc- ing gases, it is possible that exposure to reducing gases acts as a switch to the partly depleted layer case (due to the injection of additional free charge carriers) [32]. 23
  • 44. Large grains Product Gas Z X Energy Current flow qVS Eb Xg> λD Xg X 2X0 Small grains Current flow q∆VS <KB T Energy Flat band condition q∆VS Eb Xg< λD X Extended surface influenceFigure 2.6: Schematic representation of a porous sensing layer with geometry andenergy band for small and large grains. λD Debye length, Xg grain size [33]. 24
  • 45. It is also possible that exposure to oxidizing gases acts as a switch between partly depleted and completely depleted layer cases depending on the initial state of the sensing film. Figure 2.6 illustrates the conduction model of a porous sensing lay- er with geometry and surface energy band for small and large grains. For large grains, conduction can be hindered by the formation of depletion layers at surface/bulk regions and grain boundaries; the presence of ener- gy barriers blocks the motion of charge carriers. In contrast, flat band conditions dominate in case of small grains, allowing fast conduction for this case [34]. For porous layers the situation may be complicated further by the presence of necks between grains (Figure 2.7). It may be possible to have all three types of contribution presented in figure 2.8 in a porous Z (b) eVS Zn X eVS Z Z (a) eVSCurrent Zo flow Zn Conduction Zn- 2Z0 X Channel Energy Figure 2.7: schematic of a porous layer with geometry and surface energy band with necks between grains; Zn is the neck diameter; Z0 is the thickness of the depletion layer and eVS the band bending. (a) a partly depleted necks and (b) a completely depleted necks [32]. 25
  • 46. layer: I. Surface/bulk (for large enough necks Zn >Z0, figure 2.5). II. Grain boundary (for large grains not sintered together), in which conduction can be hindered by the formation of depletion layers at surface/bulk regions and grain boundaries; the presence of energy barriers blocks the motion of charge carriers. III. Flat bands conditions which dominate in case of small grains and Surface Grain Model bulk nano crystal Schottky contact − boundary O2 Geometric Z X and Metal I 𝐿 𝐷>1 2 CO effect EC Electronic EC EF EF EF EF Band LD − LD LD O2 EV EV bulk Surface Metal Z X X X SurfaceElectrical equiva- lent circuit (Low current) Bulk Figure 2.8 : Influence of particle size and contacts on resistances and capacitances in thin films are shown schematically for a current flow I from left to right [35]. small necks, allowing fast conduction for this case. Of course, what was mentioned for compact layers, i.e. the possible switching role of reducing gases, is valid also for porous layers. For small grains and narrow necks, when the mean free path of free charge carriers becomes comparable with the dimension of the grains, a surface influence 26
  • 47. on mobility should be taken into consideration. This happens because thenumber of collisions experienced by the free charge carriers in the bulk ofthe grain becomes comparable with the number of surface collisions; thelatter 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 theseparameters 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 alsobeen extensively examined in the literature since this is a crucial factorwhen creating a commercially viable device [37, 38]. Other issues such asresponse 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 ofthem will be applied in this research.2.5.1 Sensitivity The response of a sensor upon the introduction of a particular gasspecies is called the sensitivity (S The most general definition of sensitiv-ity applied to solid – state chemi – resistive gas sensors is a change in theelectrical resistance (or conductance) relative to the initial state uponexposure to a reducing or oxidizing gas component). The sensitivity de-pends on many factors including the background gas composition, rela-tive humidity level, sensor temperature, oxide microstructure, film thick-ness and gas exposure time. One of the most common methods is to re- 27
  • 48. port the ratio of the electrical resistance (R) in air to the resistance meas-ured when a gas is introduced as shown in equation 2.28 [39]: R G = (2.28) RGwhere R is the electrical resistance and G is the electrical conductanceand the subscript “AIR” indicates that background is the initial dry airstate and the subscript “GAS” indicates the analyte gas has been intro-duced. Another common approach to report S is shown in equations 2.29and 2.30 [40]: |∆R| RG −R = 100 = 100 (2.29) R R |∆ | 100 100 (2.30) The values calculated using the above equations scale from zerowhile 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 issimply to add 1 to the reported value and they are equivalent. Both valuesare acceptable and useful metrics for gas sensor response testing. In thecurrent research, the percentage conductance change has been selected asthe 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 gasspecies from the background atmosphere. This is a task where metal ox-ide based sensors face significant challenges and show poor discrimina-tion between gas species. Selectivity has been defined as the ratio of sen-sitivities to a particular gas as shown in equation 2.31. SG ⁄ = (2.31) SG 28
  • 49. In this work only a single gas was introduced in a background of dry airso 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 alsobeen applied to describe the ability of a sensor (or array) to detect anddistinguish 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 theselectivity 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 fromn – type and p – type materials in a sensor array are valuable for dataanalysis algorithms to enhance selectivity. Selectivity is not a focus ofthis study. It will become a more important issue once other reducingand/or oxidizing gases become available.2.5.3 Stability Stability measures the capability of a sensor to maintain sensitivityover durations of time for a particular gas species. Stability is measuredin terms of baseline “drift” which is the change in baseline conductanceover some duration of time at a particular temperature. Here we definedrift as the change in baseline conductance relative to the initial conduct-ance as written in equation 2.32 below: 1 − D=| | (2.32) 1− 0 Drift is reported in units of Siemens per hour (S/hr). Elvin R.Beach III [20] reported the drift over durations of 48 hours by individual-ly introduceing reducing and oxidizing gases during testing to simulatemore realistic conditions the sensor would operate under. 29
  • 50. 2.5.4 Response and Recovery Times The response time (τres) of a gas sensor is defined as the time ittakes the sensor to reach 90% of maximum/minimum value of conduct-ance upon introduction of the reducing/oxidizing gas [42]. Similarly, therecovery time (τrec) is defined as the time required to recover to within10% of the original baseline when the flow of reducing or oxidizing gasis removed. Figure 2.9 shows how this is measured from sensor dataplotting the conductance as a function of time. 90% of maximum Maximum conductance conductance Conductance Recovered to within 10% of original baseline τres τrec Time Figure 2.9: drawing showing how response and recovery times are calculated from a plot of sensor conductance versus time [20].Xu et al., [43] prepared ultrathin Pd nanocluster film capable of detecting2% 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. Fouryears earlier that time, in 2001, Favier et al., [44] obtained a comparableultrafast response time of less than 75 ms towards H2 : N2 gas mixturefrom 2 to 10% at room temperature using palladium mesowire arrays. 30
  • 51. 2.6 Sensing Mechanism Both thin – film and bulk semiconducting metal oxide materialshave been widely used for the detection of a wide range of chemicalssuch as H2, CO, NO2, NH3, H2S, ethanol, acetone, human breath, andhumidity. The sensing mechanism of metal oxide gas sensors mainlyrelies on the change of electrical conductivity contributed by interactionsbetween metal oxides and surrounding environment. The exchange ofelectrons between the bulk of a metal oxide nanostructure grain and thesurface states takes place within a surface layer (charge depletion layer),thus, contributing to the decrease of the net charge carrier density in thenanomaterial conductance channel. This will also lead to band bendingnear the surface of both conduction and valence bands. The thickness ofthe 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 inthe Schottky approximation [32]: 1 eVS 2 ω = λD ( ) (2.33) kT 1 εε0 kT 2 λD = ( 2 ) (2.34) e n0where ω is the width of the surface charge region that is related to theDebye length λD of the nanomaterial, ε0 is the absolute dielectric con-stant, ε is the relative dielectric permittivity of the structure, k is theBoltzmann’s constant, T is the temperature, e is the elementary charge, n0is the charge carrier concentration and VS is the adsorbate – induced bandbending. The Debye length λD is a quantum value for the distribution ofthe space charge region. It is defined as the distance to the surface atwhich the band bending is decreased to the 1/e – th part of the surfacevalue [11]. 31
  • 52. The conductance of 1-D metal oxide nanomaterial can be ex-pressed as [45] 1 A π(D − 2ω)2 = = = n0 eμ (2.35) ρl 4lwhere R is the electrical resistance, ρ is the resistivity, no is the ini-tial/nominal charge carriers concentration, e is the electron charge, μ isthe 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 regionthat is related to the Debye length of the nanomaterial. Likewise, theelectrical conductance of ZnO nanofilms can be expressed as dependent[24] upon the charge carriers’ concentration: 1 A no |e|μA = = = (2.36) ρl lhere A, l are the area and length of the nanofilm channel, respectively.Therefore, the change in electrical conductance of the nanofilm exposedto gas atmosphere is determined by the change in electrical charge carri-ers’ concentration ∆n [12]: ∆no |e|μA ∆ = (2.37) lThe gas sensitivity S is given by [24]: |∆ | ∆nS 𝑆= 100 = (2.38) noAccording to this expression, higher gas sensitivity could be obtained bya larger modulation in the depletion region of the ZnO metal oxide nano-film. The width of the depletion region is inversely proportional to thesquare root of the free charge carrier concentration. When the radius of ZnO nanostructure (grain) is of the order of orless than Debye length/radius, the conductive channel is reduced substan-tially. The modulation of the depletion region width can also be produced 32
  • 53. 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 ambientenvironment can be defined as [13]: g − a 4 4 𝜀𝜀0 1 2 1 = 100 = ωa − ωg = √ (VSa − VSg2 ) (2.39) g D D 𝑒𝑛0where Gg and Ga are the conductance of ZnO nanostructure in H2 gas andin air ambient, respectively, n0 is carrier concentration in air. VSa and VSgare the adsorbance – induced band bending in air and in H2 gas, respec-tively. According to this equation, enhancement of H2 gas sensitivity canbe realized by controlling the geometric factor (4/D), electronic character- 1 1istics (εε0/en0), and adsorption induced band bending (VSa2 − VSg2 ) dueto adsorption on the ZnO nanostructure surface. This can be done bydoping, or by using modulation of operation temperature which is notdesirable for H2 gas sensors on single ZnO nanowire. Another way is tomake 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 performance2.7.1 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 allimportant in utilization of the sensors. These determine the frequency atwhich the calibration checks should be carried out and the frequency atwhich the sensors may have to be replaced. They can only be determinedover long periods of time and no method by which the process can beaccelerated is valid [18]. 33
  • 54. 2.7.2 Sensor surface poisoning The surface of ZnO and other oxides may become unstable becauseof “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 tothe presence of H2S in commercial SnO2 based sensors. Another domi-nant poison is chlorine gas. Thus it is important in the development ofsensors to be aware of the other reactive gases in the measurand environ-ment [18].2.8 Optimization of Sensor Performance2.8.1 Use of Catalyst Metal oxide gas sensors need a catalyst deposited on the surface ofthe film to accelerate the reaction and to increase the sensitivity. A cata-lyst is a material that increases the rate of chemical reactions withoutitself getting changed. It does not change the free energy of the reactionbut 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 fieldclose to that of heterogeneous catalysis, with the only difference that incatalysis one is mainly interested in the products of the reaction whereasin gas sensing one is interested in the reactants as shown in the figure2.10. This approach is considered relatively standard in fields such asheterogeneous catalysis but so far it has rarely been applied to solid-stategas sensors. The chosen catalyst influences the selectivity of sensor. Ideally, ifone wants to detect a particular gas in a mixture of gases, one will like acatalyst combination that catalyzes the oxidation of the gas of interest and 34
  • 55. does not catalyze the oxidation of any other gas. Unfortunately, suchideal combinations are not easily found [46]. The widespread applicability of semi-conducting oxides such asSnO2 or ZnO, as gas sensors is related both to the range of conductancevariability and to the fact that it responds to both oxidizing and reducinggases. Small amounts of noble metal additives, such as Pd or Pt are com-monly dispersed on the semiconductor as activators or sensitizers to im- RH2+O2 RH2+H2O O- O- O- O- O- O- O- O- O- O- O- - O- O- O - O- O- - O O Figure 2.10: Illustration of catalyst effect. Nano – particles, having higher surface area, act as catalysts. Here, R stands for reducing gas [46].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 oneis 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 orexplains sensor behavior in different environments [48]. In spite of all thework reported, a deep analysis of the material – gas interaction and itsinfluence on the sensor electrical response is still lacking to completely 35
  • 56. understand the role played by the additives on the gas sensing mecha-nism. A model for increase in sensitivity using nanoparticles has beenexplained 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 interactionis a chemical reaction by which additives assist the redox process of met-al oxides. The term spillover refers to the process, illustrated in figure2.11, namely the process where the metal catalysts dissociates the mole-cule, then the atom can ‘spillover’ onto the surface of the semiconductorsupport. At appropriate temperatures, reactants are first adsorbed on tothe surface of additive particles and then migrate to the oxide surface toreact 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 asshown in the figure 2.12. H2 H2O H2O H H H2O O- O- - H M H O M O - O- O- e O- O- Activation of gas followed Acceptor of electrons by spilling over Change in redox state of additive Change in surface oxygen concentration Chemical sensitization Electronic sensitization Figure 2.11: Mechanism of sensitization by metal or metal oxide additive [48]. 36
  • 57. 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]. Thus, for a catalyst to be effective there must be a good dispersionof the catalysts, as shown in the figure 2.13, so that catalyst particles areavailable near all inter – granular contacts. Only then can the catalystsaffect the important inter-granular contact resistance. The inverse effectmay also occur [49] when a nascent oxygen or gas atom is newly formedfrom a reaction on a metal oxide site. The nascent atom may migrate to ametal site and desorbs into gas molecule from there. This is called reversespillover or the porthole effect.2.8.1.2 Fermi Energy Control The second interaction is the electronic sensitization (figure2.11); in which additives interact electronically with the metal – oxide asa sort of electron donor or acceptor. For example, changes in the workfunction of the additive due to the presence of a gas will cause a changein the Schottky barrier between the metal and the oxide and thus, achange in conductivity. This simply means that oxygen adsorption on the 37
  • 58. 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.O 2 CatalystO- Electron flow (a) Poor catalyst dispersion (b) Need adequate catalyst dispesion (Particle separation < 500 A) 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]. 38
  • 59. 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 polycrystallineelement [48]. Each crystallite of semiconductor oxide in the element hasan electron depleted surface to a depth of L in air, where L is determinedthe by Debye length LD (or λD ) and the strength of chemisorptions. Thegrain size effects are pictorially depicted in figure 2.14. If the diameter Dof the crystallite is comparable to 2L, the whole crystallite will be deplet- 𝐷 ≫ 2𝐿 (𝑔𝑟𝑎𝑖𝑛 𝑏𝑜𝑢𝑛𝑑𝑎𝑟𝑦 𝑐𝑜𝑛𝑡𝑟𝑜𝑙) Core region 𝐷 ≥ 2𝐿 (𝑛𝑒𝑐𝑘 𝑐𝑜𝑛𝑡𝑟𝑜𝑙) (Low re- sistance) Space – charge region (High re- 𝐷 < 2𝐿 (𝑔𝑟𝑎𝑖𝑛 𝑐𝑜𝑛𝑡𝑟𝑜𝑙) sistance) Figure 2.14: Schematic models for grain – size effects [31]ed of electrons and this will cause the gas sensitivity of the element to thereducing 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 changeswith 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- 39
  • 60. ture of the channel is attenuated by the surface space charge layer. Thismodel 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 sensingelement is dominated by conduction through grain boundary contacts(grain boundary control). For D≥2L, neck control forms the primarymechanism of conductivity modulation (neck control). For D<2L, theelectrical 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 thicknessbut also in their microstructures and can thus lead to rather differenttransducer functions [48, 51]. The sensitivity of the layers dependsstrongly 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 thefilm surface relative to the thickness of the bulk semiconductor. In gen-eral, when the depletion width equals the film thickness, more sensitivityis expected [51]. Adsorption of the atmospheric oxygen on the surface of sensingfilm, results in an increase of the resistance of sensing thin film. Uponexposing to reducing gas, reduction in depletion layer depth occurs thusdecreasing the resistance of the film. When the depletion depth is more or less equal to the thickness ofsensing film, the resistance will be high and hence contributing for thehigher sensitivity. However, it has been pointed out that the columnar 40
  • 61. 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 doesplay a role in determining the sensitivity of the sensor for different gases[50]. The thin SnO2 layer, (thickness 50-300 nm) mainly responds to theoxidizing gases such as Ozone and NO2 whereas thick films (thickness15-80 μm) respond to reducing gases like CO and CH4. However, uponreducing the temperature of the sensor, the thick film showed a signifi-cant response to oxidizing gases. This behavior can be explained with thediffusion 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 alsotemperature dependent, resulting in different static characteristics at dif-ferent temperatures. On the other hand, temperature has an effect on thephysical properties of the semiconductor sensor material such as chargecarrier concentration, Debye length, work function etc. The optimum range of temperature for an effective sensor responsecorresponds to that where the material is able to catalytically reduce oroxidize the target gas, simultaneously changing the electrical propertiesof the sensor material. The rate of reaction depends on the exact reducingagent under study. It is found that, with a given reducing agent, there ispeak in the sensitivity: If the temperature is too low, the rate of reaction istoo slow to give a high sensitivity, whereas if the temperature is too high, 41
  • 62. the overall oxidation reaction proceeds so rapidly that the concentrationof 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 bereduced/oxidized without producing a perceptible electrical change on themetal-oxide material. The sensitivity again is low. However, on the onehand, temperature should be high enough to allow gas reaction on thematerial 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 propertiesand the sensor electrical response is indispensable to understand thewhole gas sensing mechanism. For each sensor-gas combination, an optimum temperature be-tween these limits must be used. When higher degrees of selectivity areneeded, sensor arrays are used (sometimes termed “electronic noses”),where the different response of different sensors is used for identifyingthe gaseous species by pattern matching [52]. With such sensor array, thelack 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 atdifferent operating temperatures or of the device using different materialsat the same temperature [53, 54].2.8.5 Filters for selectivity The use of filters forms another approach to improve the selectivityof gas sensors. These filters either consume gases that one does not wishto pass to the gas sensor or to permit the passage of selected gases to thesensor. Their use is to a great extent empirical. For example, Ogawa et alclaim that ultra-fine SnO2 rejects methanol. Carbon cloth and low porosi-ty materials are used to prevent highly reactive or large molecules from 42
  • 63. reaching the sensor. Silica can be used to increase hydrogen sensitivity,as hydrogen passes more freely through a silica surface layer. SimilarlyTeflon is helpful in stopping H2O reaching the sensor and Zirconia can beused 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 andby DC measurement. The inherently noisy behavior of the resistor, 1/fnoise also known as flicker noise in the DC resistance measurements canoften approach the desired sensitivity threshold of the sensor. AC re-sistance measurements are one way to overcome prohibitive 1/f noise, butthey 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 Göpelhave reported [55] that sensitivity and selectivity of the gas sensors canbe improved by applying different conduction measurement methods viz.DC and AC conduction measurement methods. They have shown that theuse 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; intergranularcontact, bulk and electrode contact is illustrated in figure 2.15. Intergran-ular contact: the ionosorption of oxygen at the grain surface results in thecreation of a potential barrier and the corresponding depletion layers atthe intergranular contacts. An intergranular contact can be representedelectronically by a resistor Rgb (due to the high resistive depletion layers)and a capacitor Cgb (due to the sandwiching of high resistive depletionlayers between two high conductive ‘plates’ of bulk material) in parallel.The electrode contact can also be represented by a (RC) element. The 43
  • 64. Large grains Product Xg> λD Gas Z X Energy Current flow qVS ∆Φ Eb Xg X 2X0 𝑞𝑉 𝑆 ∆Φ 𝑅 𝑔𝑏 ~𝑛 𝑏 exp( ) 𝑅 𝐶 ~𝑛 𝑏 exp( ) 𝑘𝐵 𝑇 𝑘𝐵 𝑇 𝑅 𝑏 ~𝑛 𝑏 𝑅 𝑏 ~𝑛 𝑏 … … 𝜀 0.5 𝜀 0.5 𝐶 𝑔𝑏 ~( ) 𝐶 ~( ) 𝑞𝑉 𝑆 𝐶 ∆ΦFigure 2.15: Equivalent circuit for the different contributions in a thin film gas sen-sor; intergranular contact, bulk and electrode contact. 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. 44
  • 65. 2.9 Zinc oxide2.9.1 Properties of Zinc oxide Zinc oxide, ZnO, is an interesting II – VI compound semiconductorwith a wide direct band gap of 3.4 eV at room temperature [56]. It is awidely used material in various applications such as gas sensors, UVresistive coatings, piezoelectric devices, varistors, surface acoustic wave(SAW) devices and transparent conductive oxide electrodes [57]. In theearly 2000 ‘s, ZnO also attracted attention for its possible application inshort – 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 0.1MPa hydrostatic pressure are the hexagonal closed packed (hcp) Zn witha very narrow composition range, the hexagonal compound, ZnO (49.9 to50.0 at % O), with a narrow but significant composition range, and acubic 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, aface centered cubic (fcc) modification of ZnO is stable. Also, it has beenreported that ZnO can exist metastably at room temperature in either oftwo cubic modifications with structure of ZnO (sphalerite) and NaCl(rock salt) types [58]. Table 2.3 summarizes data related to Zn – O crystalstructures. ZnO crystals are composed of alternate layers of zinc and oxygenatoms disposed in a wurtzite hexagonal close – packed structure with alongitudinal axis (c – axis) as shown in figure 2.17 [59]. The oxygenatoms (ions) are arranged in close hexagonal packing, with zinc ions 45
  • 66. Zn – Rich Boundary O – Rich 49.999 Boundary Liq. ̴~0.005 Unknown 800 Liquid 0 T C 600 ZnO (Zn)~O Liq. ̴~ 7 10−7 419.58 0 ~419.80 ~50.00 M.P. 400 (Zn) ZnO2 0 10 20 30 40 50 60 70 Z At. % O n T – X diagram for condensed Zn- O system at 0.1 MPa [58]. Figure 2.16: Table 2.3: Zn – O crystal structure data [58] Stable phases at 0.1 MPa Other phases Zn ZnO (I) ZnO2 ZnO (II)(a) ZnO (III)Composition, ~0 49.9 to 50.0 ~66.7 ~50 ~50 at. % O Pearson hP2 hP4 cP12 cF8 cF8 SymbolSpace group P63/mmc P63mc Pa3 Fm3(-)m F4(-)3m ZnO FeS2 ZnS Prototype Mg NaCl (wurtzite) (pyrite) (sphalerite) 46
  • 67. O Zn Zn Zn Zn O O O O Zn 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].occupying half the tetrahedral interstitial positions with the same relativearrangement as the oxygen ions. In the crystal structure, both zinc andoxygen ions are coordinated with four ions of the opposite charge, andthe 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 somerelatively large open spaces (0.095 nm). Typical properties of ZnO arelisted in Table 2.4 [58, 60] and Ellingham diagram including ZnO isshown 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 byspecial heat treatments and by the introduction of specific impurities intothe crystal lattice. ZnO can even be made to exhibit metallic conductivityas for transparent electrodes similar to ITO. In general, 0.5 – 1% addi- 47
  • 68. tions of trivalent cations (e.g. Al and Cr) decrease the resistivity of ZnOby about 10 orders of magnitude. [58]. Table 2.4: Typical properties of zinc oxide [58, 60]. Property Value Crystal structure Hexagonal, wurtzite Molecular weight Zn:65.38, O:16 and ZnO:81.38 a0: 0.32495 Lattice parameters at 300 K (nm) c0: 0.52069 5.606 or Density (g cm-3) 4.21 x 10 ZnO molecules/mm3 19 Stable phase at 300 K Wurtzite Melting point (ºC) 1975 Thermal conductivity 0.6, 1-1.2 a0: 6.5  10-6 Linear thermal expansion coefficient 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 <106 Intrinsic carrier concentration (cm-3) 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 1010 Piezoelectric coefficient D33 = 12 pC/N Electron effective mass 𝑚∗𝑒 0.24Electron 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 48
  • 69. Figure 2.18: Ellingham diagram of oxides [61].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 ionsand their associated valence electrons as shown in figure 2.19. A principaldifference between point defects in ionic solids and those in metals is thatin the former, all such defects can be electrically charged. Ionic defects 49
  • 70. 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 energylevels. Such an excitation may create an electron in the conduction band A A A ASubstitutional B A A impurity Vacancy A A A A Vacancy A A Interstitial impurity A A A A Figure 2.19: various types of point defects in crystalline materials [62].and/or an electron hole in the valence band of the crystal. In terms ofspatial positioning, these defects may be localized near atom sites, inwhich case they represent changes in the ionization state of an atom, ormay 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 andcharge 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 bepreserved, although the total number of sites can be increased or de-creased. 50
  • 71. For example, the Schottky disorder for NaCl and Frenkel disorderfor ZnO, respectively, can be written using Kröger – Vink notation as: ′ ⋅ null = VNa + VCl (2.40)and ZnX = Zn⋅⋅ + VZn Zn i ′′ (2.41)where null indicates the creation of defects from a perfect lattice. Therespective mass – action equilibrium constants are: ′ ⋅ K Scho ky = [VNa ] ⋅ [VCl ] (2.42)and KF enkel = [Zn⋅⋅ ] ⋅ [VZn ] i ′′ (2.43)The brackets denote concentration, usually given in mole fraction. Writ-ing the equilibrium constant as the product of concentrations implies thatthe thermodynamic activity of each defect is equal to its concentration.The free energies for these quasichemical reactions are simply theSchottky or Frenkel formation energy, and the equilibrium constant isgiven by: ∆HS K Scho ky = K ∘ exp (− S ) (2.44) kTand ∆HF KF enkel = K ∘ exp (− F ) (2.45) kTThe equilibrium constant is a function of temperature only and the prod-uct of the cation and anion vacancy concentrations is a constant at fixed 51
  • 72. temperature. Furthermore, when only intrinsic defects are present, theconcentration of anion and cation vacancies must be equal for chargeneutrality considerations, ∆HS [VNa ] = [VCl ] = K ∘ 1⁄2 exp (− ′ ⋅ S ) (2.46) 2kTand ∆HF [Zn⋅⋅ ] = [VZn ] = K ∘ 1⁄2 exp (− i ′′ F ) (2.47) 2kTIn Kröger – Vink notation, free electrons and holes do not themselvesoccupy lattice sites. The process of forming intrinsic electron – hole pairsis excitation across the band gap, which can be written as the intrinsicelectronic reaction: null = e′ + h⋅ (2.48)The equilibrium constant for this reaction is: g K e = n. p = [e′ ]. [h⋅ ] = NC . NV exp (− ) (2.49) kTwhere NC and NV are the density of state of conduction band and valenceband, respectively, and Eg is the energy band gap of the material. When electrons and holes are tightly bound to an ion, or otherwiselocalized at a lattice site, the whole is considered to be one ionic defect.Thus, the valence state of defects such as vacancies and interstitials canvary. For instance, an oxygen vacancy can in principle take on differentvalence states (VO VO and VO ), as can cation interstitials, e.g., Zn⋅⋅ , Zn⋅ ⋅⋅ ⋅ X i iand ZnX in the wurtzite structure compound ZnO. i 52
  • 73. 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 ofoxygen to the gas phase leaving behind doubly charged oxygen vacanciesor cation interstitials: 1 1 OX = O2 (g) + VO + 2e′ O ⋅⋅ o OX = O2 (g) + Zn⋅⋅ + 2e′ O i (2.50) 2 2The equilibrium constant for the creation of double ionized oxygen va-cancies is: 1 R ⋅⋅ K R = n2 . [VO ]. O2 2 = K 0 . exp (− R ) (2.51) kTIn ZnO, the electron is a major electronic charge carrier. Thus, the con-ductivity of ZnO is: 1 R −1σ ∝ n = 2 ⋅ [VO ] = 2 ⋅ (2K 0 ) ⋅⋅ R 3 ⋅ exp (− ) ⋅ (O2 ) 6 (2.52) 3kTWith background acceptor, ⋅⋅ [A′ ] = 2 ⋅ [VO ] (2.53)The conductivity of ZnO is: 1 R −1 −1 σ ∝ n = (2K 0 ) R 2 ⋅ exp (− ) ⋅ (O2 ) 4 ⋅ [A′ ] 2 (2.54) 2kTAlso, the reduction of ZnO can be expressed by creating single chargedoxygen vacancies or cation interstitials: 1 1 OX = O2 (g) + VO + e′ O ⋅ o OX = O2 (g) + Zn⋅ + e′ O i (2.55) 2 2 53
  • 74. The equilibrium constant for the creation of single ionized oxygen vacan-cies is: 1 R ⋅ K R = n. [VO ]. O2 2 = K 0 . exp (− R ) (2.56) kTThus, the conductivity of semiconducting ZnO in which single ionizedoxygen vacancies are dominant is: 1 R −1 σ ∝ n = [VO ] = (K 0 ) ⋅ R 2 ⋅ exp (− ) ⋅ (O2 ) 4 (2.57) 2kTWith background acceptor, ⋅ [A′ ] = [VO ] (2.58)The conductivity of ZnO is: R −1 σ ∝ n = (K 0 ) ⋅ exp (− R ) ⋅ (O2 ) 2 ⋅ [A′ ] −1 (2.59) kTThus, investigating the conductivity of ZnO in reducing environments canassist in determining the valence state of defects and the activation energyfor releasing electrons. Substituted foreign atoms can also enhance the semiconductingproperties of ZnO. In the presence of selected metallic vapors at elevatedtemperatures, the foreign metallic atom replaces a portion of the Zn at-oms. The zinc atoms, on release from their lattice positions, diffuse to thecrystal surface where they are vaporized. This substitution process cansubstantially 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. 54
  • 75. 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, solidoxide fuel cells etc. It has evolved into an important thin film depositiontechnique 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 – effectivenesswith 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 offilm 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 followingatomizers, table 2.5, are usually used in spray pyrolysis technique: airblast (the liquid is exposed to a stream of air) [66], ultrasonic (ultrasonicfrequencies produce the short wavelengths necessary for fine atomiza-tion) [67, 68] and electrostatic (the liquid is exposed to a high electricfield) [69, 70]. Table 2.5: Characteristics of atomizers commonly used for SPD [63]. Atomization rate Droplet velocity Atomizer Droplet size µm cm3/min. 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 Key parameters of this process are the atomization technique, aero-sol transport (carrier gas, pressure, distance, and reactor geometry), sub- 55
  • 76. strate temperature and material, the relative expansion coefficients of thefilm and the substrate upon which it is deposited, and most importantly,the chemical composition of the solution (both solvent and precursor salttypes). [71]. Many studies were made on CSP process since the pioneering workby Hill and Chamberlin in 1964 on CdS films for solar cells [65]. Severalreviews on this technique have also been published.Siu and Kwok made a detailed study of the properties of the Cu 2S/CdSthin film solar cells formed on chemically sprayed CdS films. Good andreproducible films could be obtained using a spray-rate of 2.8 ml min-1and a substrate temperature of around 340 0C. They demonstrated thatcells 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 filmsof oxide superconductors in relation to deposition parameters and theirdevice 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 depositiontemperature [74].Song et al., presented a preparation procedure of spray pyrolyzed un-doped and aluminium doped zinc oxide thin films for solar cell. Theyinvestigated 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]. 56
  • 77. Next, in 1995, Roh et al., [76] did employ ultrasonic nozzle to depositCdS thin films on SAW devices intended for SO2 gas sensing.Polycrystalline tin oxide SnO2 films with nano – size crystallites (8 – 20nm) were prepared by Korotcenkov et al., [77]. The crystallites with ori-entation (110) or (200) plane parallel to substrate, forming the surface ofthe film, were predominant in the sprayed SnO2 films. The latter factor isimportant in influencing the gas sensitivity characteristics similar to thegrain size effect.Different atomization techniques and properties of metal oxide, chalco-genide and superconducting films prepared using CSP were discussed byPatil [78]. The results proved that film properties depended on the prepa-ration conditions and could be easily tailored via optimizing sprayingparameters viz. substrate temperature, spray rate, precursor concentrationetc.After that, Ebothe and El Hichou [79] examined the role of differentspraying flow rates deposition parameter f, between 1 and 8 mil min-1, onthe surface irregularities evolution of a sprayed ZnO thin film of the samethickness e evaluated at 1 mm by always adjusting the deposition time, t,to the f value. This thickness has been confirmed from cross-sectionalimages of the samples examined by scanning electron microscopy (SEM)using a LEO 982 set. The substrate – nozzle distance d=44.5 cm is keptconstant and the optimal spraying temperature used is 450 0C. The XRDresults revealed that the variation of f has no effect on the material’sstructure as it remains hexagonal and has (002) preferred growth orienta-tion which is normal to the substrate plane. 57
  • 78. In 2005, Perednis and Gaukler gave an extensive review on the effect ofspray parameters on films as well as models for thin film deposition byCSP [63].Gümüs et al. [66] reported highly transparent ZnO thin films that hadsuccessfully been prepared by pyrolytically spraying zinc acetate solutionon 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 exhibitsexcellent 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 ofover 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 eVoptical 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 propertiesof a ZnO thin film obtained by spraying a 0.1 M zinc acetate precursor onglass substrate held at 370 oC temperature. The optical energy gap for thefilm of different thicknesses is estimated to be in the range 2.98 – 3.09eV. The film exhibits thermally activated electronic conduction and theactivation 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 filmsat substrate temperatures around 475 0C. Zinc acetate dehydrate (Zn(CH3COO)2.2H2O) and Aluminum acetylacetonate (C15H21AlO6) were 58
  • 79. used as precursors and the solvent was a mixture of de – ionized water,methanol and acetic acid. The obtained films are polycrystalline with ahexagonal wurtzite structure and are preferentially oriented in the (002)crystallographic direction. Grain sizes varied from 21.3 to 25.3 nm basedon substrate temperature. An average transmission of 75% is observedand the optical band gap of AZO films is varied from 3.26 to 3.29 eVwith 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 anaerosol 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 than1020 cm−3. The carrier density and mobility are measured by both Hallprobe and near infrared spectroscopy. Film growth and grain size, mor-phology and orientation have been altered using an increased percentageof ZnCl2 in the precursor, which resulted in a 10 – fold increase in chargecarrier 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 aheated substrate. Typical CSP equipment consists of an atomizer, a sub-strate heater, temperature controller and a solution container. Additionalfeatures like solution flow rate control, improvement of atomization byelectrostatic spray or ultrasonic nebulization can be incorporated into thisbasic system to improve the quality of the films. To achieve uniform large 59
  • 80. area deposition, moving arrangements are used where either nozzle orsubstrate or both are moved. The schematic diagram of a typical sprayunit is depicted in figure 2.20. Only crude models about the mechanism of spray deposition andfilm formation have been developed. There are too many processes thatoccur sequentially or simultaneously during the film formation by CSP. Mechanical Power Supply Gas Regulator System Valve Electronic Compressed Controller Air Spray nozzle Spray Solution Substrate Hot Plate Thermocouple Heater Temperature Controller Power Supply Figure 2.20: Schematic diagram of chemical spray pyrolysis unit [83].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 ofprecursor solutions, transportation of the resultant aerosol and decompo-sition of the precursor on the substrate. Atomization of liquids has been 60
  • 81. investigated for years. It is important to know which type of atomizer isbest suited for each application and how the performance of the atomizeris affected by variations in liquid properties and operation conditions. Airblast, ultrasonic and electrostatic atomizers are normally used. Amongthem, air blast atomization is the simplest. However this technique haslimitation in obtaining reproducible droplets of micrometer or submicronsize and in controlling their distribution [63, 83]. In ultrasonic nebulized atomization, precursor solutions are foggedusing an ultrasonic nebulizer [68]. The vapour generated is transported bycarrier gas to the heated substrate. Precursor solution is converted tosmall droplets by ultrasonic waves and such droplets are very small withnarrow size distribution and have no inertia in their movement. Pyrolysisof 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 significanceonly in recent years. Electrostatic atomization of liquids was first reportedby Zeleny [84]. Jaworek et al. published an article on this type of atomi-zation [70]. A positive high voltage applied to the spray nozzle generateda positively charged spray. Stainless steel discs acted as cathode and thedroplets under electrostatic force moved towards the hot substrate wherepyrolysis took place. In electrostatic spray, depending on the spray pa-rameters, various spraying modes were obtained. They were classified ascone – jet mode and multi – jet mode. Cone – jet mode split into multi –jet mode with increase in electric field, where number of jets increasedwith applied voltage. 61
  • 82. The reaction process taking place in CSP is interesting. Manymodels exist for the decomposition of precursor. Many simultaneousprocesses occur when a droplet hits the substrate surface: evaporation ofresidual solvent, spreading of droplet and salt decomposition. Vigue andSpitz 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 dryprecipitate in which decomposition occurs.In process B, solvent evaporates before the droplet reaches the surfaceand precipitate impinges on the surface where decomposition occurs.In process C, solvent vaporizes as droplet approaches the substrate, thensolid melts and sublimes and vapour diffuses to substrate to undergo het-erogeneous reaction there.In process D, at highest temperature, the metallic compound vaporizesbefore it reaches the substrate and chemical reaction takes place in vapourphase. Substrate Finely divided solid product Vapour Precipitate A B C D Figure 2.21: spray processes (A, B, C, and D) occurring with increase in substrate temperature [85, 63]. 62
  • 83. 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]. 63
  • 84. For ZnO thin films grown by spray pyrolysis, it was found that organicsalts (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 thesample surface by their burning. It was observed that transparency of asdeposited ZnO films increased when ethanol was used instead of water assolvent for zinc acetate [71].III Spray rate Spray rate is yet another parameter influencing the properties offilms formed. It has been reported that properties like crystallinity, sur-face morphology, resistivity and even thickness are affected by changes inspray rate [88]. It is generally observed that smaller spray rate favours formation ofbetter crystalline films. Smaller spray rate requires higher deposition timefor obtaining films of the same thickness prepared at higher spray rate.Also, the surface temperature of substrate may deviate to a lower value athigh spray rate. These two factors may contribute to the higher crystal-linity at small spray rates. Decrease in crystallinity at higher spray rates isobserved in sprayed CuInS2 thin films [88]. Decrease in crystallinityusually results in increased resistivity of the films. Surface morphology of the films varies with spray rate. Higherspray rate results in rough films. Also, it is reported that films depositedat smaller spray rates are thinner due to the higher re-evaporation rate. 64
  • 85. IV Other parameters Parameters like height and angle of spray head, angle or span ofspray, type of scanning, pressure and nature of carrier gas etc., influencethe properties of deposited films. Different types of spray heads whichproduce different spray patterns are commercially available. Relativemotion of the substrate holder and spray head should ensure maximumuniformity and large area coverage.2.9.4 Metal oxide gas sensors The idea of using semiconductors as gas sensitive devices leadsback to 1952 when Brattain and Bardeen first reported that gas adsorptionon germanium semiconductor surface caused a variation in its electricalconductivity [3]. The first realization of a working gas sensor was in1962, when Seiyama et al. detailed the use of zinc oxide ZnO thin filmsin the detection of gases as ethanol (C2H6O) and carbon dioxide (CO2)[4]. Naoyoshi Taguchi, in that same year, published a patent for a tinoxide (SnO2) – based gas sensor [5]. From then, the detection of hydro-gen (H2), oxygen (O2) and hydrocarbon by means of surface conductivitychanges on various metal oxide crystals and thin films have been pro-posed and demonstrated [89]. Although, many metal oxides have beensuccessfully 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 easyfabrication 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- 65
  • 86. pends on the electronic and structural properties of the sensor material[36]. The sensing parameters can be promoted by the addition of metaladditives such as Al, Sn, Sb, Cu, Pt and Pd. These additives greatly en-hance the gas sensor sensitivity, shorten the response time, and shift thevolcano – shaped correlations between gas response and temperaturetoward 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 tomany other reducing inflammable and organic gases (H2, methane CH4,butane C4H10, acetone C3H6O, ethanol C2H5OH). The resistance changeand the selectivity of the sensor were enhanced by doping group III metalimpurities 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 undopedZnO sensors. The lower limit of the detection for ammonia gas was about1 ppm at a working temperature of 350 0C. The sensing mechanism ofammonia 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 thegas sensitivity of RF and DC magnetron – sputtered aluminum dopedzinc oxide (ZnO:Al) thin film on corning 17059 glass substrate with afilm thickness of about 300 nm. The sensitivity measurements were car-ried out at an operating temperature of 200 – 350 °C in air. A testing gasof 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- 66
  • 87. ten sea foods such as oyster, squid, sardine or fragrance from wine andcoffee. The high sensitivity of the sensor for the odor from rotten seafoods 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 gasesin the 0.1 – 100 ppm concentration range by a gas recognition systemusing a thin film oxide semiconductor micro gas sensor array and theneural – network pattern recognition technique. The sensing materials of1 wt. % Pd – doped SnO2, 6 wt. % A1203 – doped ZnO, WO3 and ZnOwere used for the gas sensor array whose power consumption was only 65mW at 300 0C, and the back-propagation algorithm was applied as thesupervised learning rule. The recognition probability of the neural-network was 100% for the discrimination of the gases and concentrationsused 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 bythis method were highly resistive, signifying the presence of a large den-sity of oxygen adsorbed acceptor – like trap states (O− , O-, etc.). Prelim- 2inary studies of gas sensing characteristics, performed at 150 – 375 0Ctemperature range, indicated that the Pd – sensitized ZnO films respondstrongly 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 exhibitedexcellent sensitivity (more than 99%) at 200 0C and the response timewas 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 67
  • 88. upon being exposed to methane (CH4), hydrogen (H2) and ethane (C2H6)reductive gases. The sensitivity of the sensor increased with operatingtemperature and its highest values were obtained at 200 ºC for methaneand 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 theselectivity could be promoted by the use of an appropriated catalyticmetal 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 goodquality undoped ZnO films deposited on glass and quartz by a novel CVDtechnique using a 0.5 M zinc acetate as the starting solution. They ob-tained faster response and higher sensitivity. The operating temperatureplayed a key role in the selectivity of such sensors, with the optimumoperating 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 toxicgases and vapours. Sensitivity, selectivity and stability of sensors workingin different temperature depend on the way the tin dioxide and additiveswere prepared. A construction also plays an important role. He presentedan attitude towards the evaluation of transport of electrical charges insemiconductive grain layer of SnO2, when dangerous gases appeared inthe surrounding atmosphere.Bârlea et al., [8] characterized SnO2 metal oxide gas sensors deposited ona ceramic cylinder heated to the functioning temperature (100 – 400 °C),and exposed it to an atmosphere containing a reducing gas: carbon mon- 68
  • 89. oxide (CO), liquid petroleum gas (LPG, mainly butane) and methane gas.They investigated the influence of the supplied electric current to thesensor sensitivity and the response time and recovery time. The electricresistance of the semiconducting material dramatically modified, even atvery low gas concentrations. The response time was longer for lowersupply 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 ofgas sensing characteristic parameters particularly when it comes toproviding 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 gassensor based on self – assembled monolayer promoted 2 – dimensionalpalladium film on glass substrate. Their enhanced sensor was sensitiveenough to detect hydrogen levels as low as 25 ppm with a fast responsetime 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 ofZnO:Al films as an ethanol vapor gas sensor obtained by RF magnetronsputtering on Si substrate using Pt as interdigitated electrodes. The struc-tural characteristics revealed that flat and well – defined columnar filmswith c – axis textured were formed. The film exhibited good sensitivity tothe ethanol vapors with quick response – recovery characteristics, and itwas found that the sensitivity for detecting 400 ppm ethanol vapor was~20 at an operating temperature of 250 °C. The high sensitivity, fast re- 69
  • 90. covery, and reliability implied that ZnO:Al seemed to be a promisingsemiconducting material for the detection of ethanol vapor.Patil et al., [41] explored the characterization and ethanol gas sensingproperties of pure and doped ZnO thick films prepared by the screenprinting technique. Pure zinc oxide was almost insensitive to ethanol.Thick films of (1 wt%) Al2O3 – doped ZnO were observed to be highlysensitive to ethanol vapours at 300 °C. Doping Al2O3 into zinc oxidecreated 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 showedvery rapid response and recovery to ethanol vapours. Moreover, it hadgood 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 thinfilm semiconducting layer and a Pd – catalyst layer coated on the surfaceof the semiconducting ZnO. The sensor exhibited reasonable sensitivityof 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 byLiewhiran and Phanichphant [42]. They fabricated ZnO thick films usingflame spray pyrolysis (FSP) on Al2O3 substrate interdigitated with Auelectrodes with various thicknesses (5, 10, 15 μm). The gas sensing char-acteristics to ethanol (25 – 250 ppm) evaluated as a function of filmthickness at 400 °C in dry air, displayed the tendency of the sensitivity todecrease with increasing film thickness and response time; with the thin-nest sensing film (5 μm) showed the highest sensitivity and the fastestresponse time (to 250 ppm, S=801, τres =5 s). They discussed the behavior 70
  • 91. on the basis of diffusively and reactivity of the gases inside the oxidefilms. The sensing characteristics were deteriorated evidently with in-creasing film thicknesses. The recovery times were quite long withinminutes.Aroutiounian et al., [97] reported hydrogen sensors working at and closeto room temperature and made of porous silicon covered by the TiO 2-x orZnO<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 explosionlimits of 4 – 75% by volume. The sensor had a relatively short time ofresponse and recovery (~20 s). Such sensors could also be a part of asilicon 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 temperaturerange 40 −100 0C to different concentrations of hydrogen (1000 − 5000ppm) in air was investigated. The glass ceramic/ZnO<Al>/Pt structureshowed 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 mostimportant factors affecting sensing properties, especially sensitivity. Thegas response to different gases is related to a great extent to the surfacestate and morphology of the material [13]. The use of micro andnanostructured films is advantageous due to the high surface to volumeratio the nanostructure exhibits, offering faster response and higher gassensitivities to low concentration of the tested gases and eliminating theneed of operation at elevated temperatures. Some groups have explored 71
  • 92. nano grain – sized ZnO structures and tested it for gas sensitivity. Mat-thew Szeto [98] fabricated ZnO chemical gas sensors from prepared ZnOnanoplatelets and their sensitivities to H2 gas were investigated underconditions of varying concentration, sensor temperature, and intensity ofUV light. It was found that at room temperature and a source voltage of5V that the ZnO sensor had the best sensitivity of greater than 85% to H2gas at 50 ppm and was most sensitive in the absence of UV radiation. At130 0C temperature, the ZnO sensor showed sensitivities near 100% to~50 ppm of H2, where it both responded and recovered faster. Memoryeffect previously observed was also non – existent at temperatures nearand above 100°C. The sensor was also observed to be both slower inresponse and lower in sensitivity in the absence of UV light in a darkenedchamber.Savu et al., [16] deposited a Porous nano and micro crystalline tin oxidefilms 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 todetermine the influence of the microstructure and working conditionsover the electrical response of the sensors. The conductivity of all sam-ples increases with the temperature and decreases in oxygen, as expectedfor an n-type semiconducting material. The improved sensitivity andresponse times at the 200 °C working temperature are due to the higherrate 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 tookplace in the grain boundary region, with the contribution of the grainsbeing indicated by the formation of a second semicircle at high frequen-cies. The better sensing performance of the doctor bladed samples can be 72
  • 93. explained by their lower initial resistance values, bigger grain sizes andhigher 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 werefound to be the most important factors for controlling response parame-ters of resistive type gas sensors. Thus, it was shown that the decreasingof thickness, grain size and degree of texture was the best way to decreasetime constants of metal oxide sensors. However, it was concluded thatthere is no universal decision for simultaneous optimization of all gas-sensing characteristics. One have to search for a compromise betweenvarious engineering approaches because adjusting one design feature mayimprove 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 ahome-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 thepotential of applying these films as sensing elements in future metal ox-ide gas sensing devices. 73
  • 94. Al-hardan et al., [101] prepared ZnO thin films by thermal oxidation ofZn metal at 400 0C for 30 and 60 min. The XRD results showed that theZn metal was completely converted to ZnO with a polycrystalline struc-ture. The sensors had a maximum response to H2 at 400 0C and showedstable behavior for detecting H2 gas in the range of 40 to 160 ppm. Filmoxidized for 60 min in oxygen flow exhibited higher response than that ofthe 30 min oxidation which was approximately 4000 for 160 ppm H2 gasconcentration. The sensor with higher resistance yields higher response tothe gas under test. The sensing mechanism was modeled according to theoxygen – vacancy model.Tamaekong et al [102] investigated the gas sensing properties towardhydrogen (H2) of ZnO nanoparticles doped with 0.2 – 2.0 at. % Pt whichwere successfully produced in a single step by flame spray pyrolysis(FSP) technique. ZnO nanoparticles paste was coated on Al2O3 substrateinterdigitated with gold electrodes to form thin films by spin coatingtechnique. The gas sensing properties toward hydrogen gas revealed thatthe 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 lowhydrogen 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 ZnOsensor has been shown to have a lower operating temperature of around250 0C and enhanced sensitivity than previously reported. Good stabilityand repeatability of the sensor were demonstrated when tested underdifferent concentration of oxygen atmosphere. The enhancement waslikely attributed to the higher oxidation state of the chromium. 74
  • 95. Also, Al-hardan et al., [104] investigated the mechanism of hydrogen(H2) gas sensing in the range of 200 – 1000 ppm of RF-sputtered ZnOfilms. The I – V characteristics as a function of operating temperatureproved the ohmic behavior of the contacts to the sensor. The compleximpedance spectrum (IS) of the ZnO films exhibited a single semicirclewith shrinkage in the diameter as the temperature increased and as thehydrogen concentration was increased in the range from 200 ppm to 1000ppm.One month later, Al-hardan et al. [12] studied the gas sensing propertiesof 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 isopropanoland ethanol. An optimum operating temperature for maximum sensitivityof 400 0C for the above vapors was obtained. The sensor showed a stable,reversible and repeatable behavior in the acetone concentration of 15 upto 1000 ppm. They explained the sensing mechanism in accordance withthe ionosoption model. The same ZnO based sensor also exhibited goodsensitivity for vinegar test in the concentration range of 4% to 9% and themaximum sensitivity to vinegar test application was obtained at 400 0C.The work revealed the validity of using ZnO gas sensor in estimating theacid concentrations of the vinegars for food requirements [106].Hussain et al., [105], grew ultra – fine thin films of pure and SnO dopedZnO nanosensor on gold interdigitated ceramic substrate by ultrasonicaerosol assisted chemical vapor deposition technique (UAACVD) ataround 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 atemperature range of 60 – 150 °C. At room temperature (25 °C), the re- 75
  • 96. sponse and recovery time of the sensor increased many orders of magni-tude compared to 60 °C. Sensitivity of the ZnO sensor demonstratedlinear 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 hydrogengas sensor. The prepared ZnO films were doped by dipping in palladiumchloride PdCl2. Sensitivity dependence on the temperature and test gasconcentration was tested and the optimum operation temperature wasdetermined at around 280 oC. The response time of 2-3 s of the dopedZnO film was so fast to detect flammable H2 leaks well below the lowerexplosion limit (LEL) of 4%.Chen et al., [118] successfully prepared tin dioxide SnO2 thin films withinteresting fractal features by pulsed laser deposition techniques underdifferent 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 tindioxide thin films formed at different substrate temperatures are of fun-damental importance in the development of microdevices, such as gassensors for the detection of environmental pollutants. Fractal methodwas applied to the evaluation of this material. The measurements of car-bon monoxide gas sensitivity confirmed that the gas sensing behaviorwas sensitively dependent on fractal dimensions, fractal densities, andaverage sizes of the fractal clusters. The random tunneling junction net-work mechanism was proposed to provide a rational explanation for thisgas sensing behavior. The formation process of tin dioxide nanocrystalsand fractal clusters could be reasonably described by a novel model. 76
  • 97. In spite of the many researches on the as – deposited and dopedmetal oxide ZnO and SnO2 – based gas sensors prepared via differentdepositions 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 metaloxides by chemical spray pyrolysis deposition are presented. The effectof Pd doping on the sensitivity S, operating temperature and the responsetime of both metal oxides gas sensing elements to hydrogen (H2) gas ofdifferent concentrations is examined. 77
  • 98. Chapter 3Experimental ProcedureIntroduction The present chapter gives a detailed account of the work carriedout for the development of Zinc Oxide and tin oxide thin film based gassensors. The different steps followed in this regard are described. Chemi-cal spray pyrolysis deposition on glass and silicon substrates, which isused in the present work, is discussed in details. Following it is a discus-sion on the characterization of the deposited gas sensitive ZnO and SnO2materials. Details of the experimental set up made for testing and study-ing the performance of the developed gas sensors is also presented in thischapter. The results obtained and analyses of data on the performance ofthe 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 Electrical Measurement ZnO film (150 nm) Electrode: Pt (200 nm)/ Ta (25 nm) H2 Pt electrode film H2 H2 Insulation layer: SiO2 layer (1 μm) SiO2 layer H2 ZnO film Substrate: Si wafer Si wafer Figure 3.1: Schematic of a typical gas sensor structure. Thicknesses are not to scale.semiconductor metal oxide, ZnO or SnO2, gas sensor are outlined below:  Selection of substrate 78
  • 99.  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 ofthe gas sensitive material [36]. The grain size of a given material on asubstrate is known to depend on the wettability of the substrate. Sub-strates with lower surface tension are believed to result in smaller grainsize. Glass, having the lowest surface tension, resulted in the smallest 79
  • 100. 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 topographyexhibited by the substrate itself.3.2 Spray pyrolysis experimental set up Chemical techniques for the preparation of thin films have beenstudied extensively because such processes facilitate the designing ofmaterials on a molecular level. Spray pyrolysis, one of the chemical tech-niques applied to form a variety of thin films, results in good productivityfrom a simple apparatus. In the current research, zinc oxide thin films aredeposited on glass substrates employing locally – made spray pyrolysisdeposition chamber whose main components set up is illustrated in theschematic diagram of figure 3.2. It is essentially made up of a precursorsolution, 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 adjustablecopper capillary tube nozzle of 0 - 0.8 mm inner diameter clamped to aholder and supported by a metal tripod. The nozzle is driven by a com-pressed atmospheric air. The prepared precursor solution is pumpedthrough the metal nozzle with a solution flow rate ranging from 1 to 2mL/min. Due to the air pressure of the carrier gas; a vacuum is created atthe tip of the nozzle to suck the solution from the tube after which thespray starts [63]. To regulate spraying time, a 16 – Bar Tork solenoidvalve controlled by an adjustable timer has been incorporated. The atom-izer and the 1500 Watts hot plate heater are enclosed in a 1 11 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). 80
  • 101. Ventilation Fan Compressed Air Tube Measuring Air Nozzle Cylinder Capillary Tube Precursor Sprayer Spray Solution 30 cm Holder with cone stand Solenoid Valve And Timer 0 Substrate Temperature 0450° C Controller Substrate heater Thermocouple Air in Figure 3.2: Spray pyrolysis experimental set up3.3 Precursor solution A 0.2 M concentration precursor solution of zinc acetate dihydrateZn(CH3COO)2.2H2O (molecular weight 219.4954 g/mole) has been pre-pared by dissolving a solute quantity of 4.389908 g ofZn(CH3COO)2.2H2O (as weighed by a 10−4 g - precision balance) in 100mL isopropyl alcohol C3H9O (the solvent). A magnetic stirrer is incorpo-rated for this purpose for about 10 – 15 minutes to facilitate the completedissolution of the solute in the solvent. Furthermore, aqueous precursor ofZinc chloride ZnCl2 (molecular weight 136.3146 g/mole) dissolved indistilled water has also been employed in getting ZnO thin films. Organic 81
  • 102. BA Air exit Needle in nozzleC DPhoto 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. 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]. 82
  • 103. The spray rate is usually in the range 2 – 3 mL min-1. The optimumcarrier gas pressure for this rate of solution flow is around 5 kg cm-2. Atlower pressures, the size of the solution droplets becomes large, whichresults in the presence of recognized spots on the films and then reductionof transparency. This situation increases the scattering of light from thesurface 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 theprecursor 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 substratewas fixed at 30 cm. Table 3.1 summarizes the optimized thermal pyroly-Table 3.1.: Optimum thermal spray pyrolysis deposition conditions for the preparation of ZnO thin films. 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 cmsis deposition conditions for the preparation of ZnO thin films that wereemployed 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 hadpreviously been determined. After the deposition, each substrate itself isweighted again to determine the quantity of deposited ZnO. Measuring 83
  • 104. the surface area of the deposited film, taking account of ZnO specificweight of the film, the thickness is determined using the relation: ∆mZnO = (3.1) A∙ρwhere A is the actual area of the film in cm2, ∆mZnO is the quantity ofdeposited zinc oxide, and ρ is the specific weight of ZnO. Film thicknesswas 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 ofthe film to accelerate the reaction and to increase the sensitivity, impartspeed 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 toimprove 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, spraypyrolysis deposition, thermal evaporation, CVD, laser ablation, magne-tron sputtering, impregnation by salt solution. With the help of thesemethods, it was possible to form on a surface of metal oxides surfaceclusters of various components with sizes from 0.1 to 8 nm [16]. For the above reasons, the surface of the deposited ZnO thin filmswere catalyzed using successive multiple dipping (or spraying) of theprepared samples with a 50 ccm (0.0564 Molar) solution made up ofdissolving 1% by weight palladium chloride PdCl 2 (Mwt.= 177.3256g/mole) in ethanol alcohol C2H5OH. Each sample was successivelysprayed 10, 15, 20, 25, and 30 times of 4s spray interval and at 400 de-grees hot plate heater temperature. Eventually, the sensitized samples 84
  • 105. were heat – treated at the same temperature for a period of one hour inatmospheric air. About 20 – time palladium spray or dipping was found tobe 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 Bell jar Substrate holder Thickness Glow ring monitor Ring shield Shutter Cylindrical shield Boat Bus bars Valve Variable power High vacuum supply transformer Valve Rotary pump Penning gauge Pirani gauge Valve Diffusion pump 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. 85
  • 106. 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 = (3.2) 2πR2 ρ 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 15 mm 10 mm 0.4 mm 0.4 mm 2 mm 0.4 mm 13.6 mm 25 mm 2 mm 2 mm 3 mm 3 mm 3 mm 22 mm3 mm 19 mm1 mm 14 mm 2 mm 2 mm Figure 3.4.: A schematic diagram of the IDE masks utilized in this work. 86
  • 107. 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. Vacuum gage Test gas in Auxiliary inlet USB 3 mm Cable 16.3 cm 20 cm PC – interfaced ZnO DMM Sensor O –ring seal Temp. 436 Controller 2 cm V A Gas Manifold 450 Ω Gas Output to Flow meter Air vacuum 8 – pin feed through Flow pump meter Needle ValveHydrogen Air Relief valve Exhaust Vacuum Pump Figure 3.5: Gas sensor testing system 87
  • 108. Pressure Gauge Testing DC Power Chamber Supply H2 gas Gas Needle Valve Temp Controller UNI-T81 DMM Gas Flow Meter Vacuum Pump Air Supply Photo plate 3.2.: A photo of the sensor testing systemAnother third port is provided for vacuum gauge connection. A multi – pin feed through at the base of the chamber allows forthe electrical connections to be established to the heater assembly as wellas 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 ofthe sensor. The thermocouple senses the temperature at the surface of thefilm exposed to the analyte gas. The PC – interfaced multi meter, of typeUNI-T UT81B, is used to register the variation of the sensor conductance(reciprocal of resistance) exposed to predetermined air – hydrogen gasmixing ratio. The chamber can be evacuated using a rotary pump to arough vacuum of 1 10−3 ba . A gas mixing manifold is incorporated to 88
  • 109. 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 andtest 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. 89
  • 110.  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 formeasurement is described below:The flow meter connected to zero air cylinder is set to a known value (say1000 sccm) using the needle valve. Then, the flow meter connected to thetest 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 oftest gas is set to 1 sccm while keeping the flow rate of zero air at 1000sccm. A schematic diagram of the electrical circuit used for gas sensormeasurements is illustrated in figure 3.6. When the sensor is connected asshown in the basic circuit, output across the load resistor (VRL) increas-es/decreases as the sensors 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 biasvoltage Vb from 0 to 15 volts across the sensor resistance R S and the cor-responding current of the circuit is measured via the digital multimeterDMM 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 voltagecontrolled by the GEMO DT109 PID temperature controller (not shownin the figure) together with its k – type (nickel-chromium) thermocouple.The same above electrical circuit was exploited to investigate the I – Vcharacteristic of the sensitive sensing material at various temperaturesboth in pure air and in gas – containing atmosphere. 90
  • 111. PC – interfaced DMM A 220 V AC RH RS Vb DC Power Gas Supply 0 -15 V RL Figure 3.6: A schematic diagram of the gas sensor basic measurement elec- trical circuit.3.9 Crystalline structure of the prepared ZnO thin films The crystalline structure is analyzed by a SHIMADZU 6000 X-raydiffractometer (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 arerecorded at a scanning rate of 0.08333° s-1 in the 2𝜃 ranges from 20° to60°. 91
  • 112. Photo plate 3.3.: LabX XRD – 6000 Shimadzu diffractometer unit.3.10 Thin film surface topography The surface topography is analyzed with Ultra 55 scanning electronmicroscope 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 anatomic force microscope, AFM, (AA3000 Scanning Probe MicroscopeSPM, 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 equippedwith a xenon lamp. Photo plate 3.6 illustrates this spectrophotometer. 92
  • 113. 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. 93
  • 114. Photo plate 3.6.: Optima sp-3000 plus UV-Vis-NIR spectrophotometer.3.12 Tin oxide (SnO2) hydrogen gas sensors In addition to the zinc oxide – based hydrogen gas sensor, and forcomparison purposes, undoped and palladium doped tin oxide thin filmshave been prepared on glass substrates using the same deposition/dopingprocedures described previously for ZnO thin films. The 0.2 – M concen-tration spraying precursor is obtained by dissolving 4.5126 g stannouschloride dihydrate SnCl2.2H2O (molecular weight 225.63 g/mole) solutein 100 mL isopropyl alcohol C3H9O solvent. Likewise, the dissolutionprocess 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 H2gas mixing ratios. 94
  • 115. Chapter 4Results and discussionIntroduction In the preceding chapter, experimental setup and methods are de-scribed. In this chapter, the detailed experimental results and sensingperformance characteristics of the ZnO and SnO2 thin films to hydrogengas will be presented. Some important factors of sensor characteristics will be investi-gated here. These include sensitivity, response and recovery time, theoptimum operating temperature, and gas concentration. Other characteris-tics that we didnt 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 bothmetal 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)2.2H2Odissolved in water, organic solvent, and mixture of both, are also used inthe realization of zinc oxide thin films. Spraying temperature, the crucialparameter, is varied between 370 and 500 oC with the optimum sprayingtemperature being around 450 oC. Table 4.1 outlines the optimum deposi-tion conditions employed in the current research. 95
  • 116. Table 4.1: spray pyrolysis deposition optimum parameters. 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 cmFigure 4.1 illustrates a zinc oxide thin film photos of 0.2 M zinc chlorideand zinc acetate dehydrate aqueous precursors sprayed on glass substratesat 450 0C temperature. Zinc chloride aqueous precursor Zinc acetate aqueous precursor Figure 4.1: A photo of spray pyrolyzed ZnO thin film on glass samples 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 4.2with 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 96
  • 117. Figure 4.2: Scanning Electron Micrograph photo of spray pyrolyzed ZnO thin film on glassdeveloped for calculating the thickness of thin films is given elsewhere[108]. Two masks of different fingers spacing (1 mm and 0.4 mm) wereused to evaporate the Al interdigitated electrodes IDE and figure 4.3 A B 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 97
  • 118. shows two of these electrodes after being evaporated over the ZnO thinfilm 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 – dopedZnO 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 andthe effect of the Pd dopant on the structure of the film is displayed infigure 4.5. Diffraction pattern spectra are obtained with 2𝜃 starting from20 ° to 50 ° at 6 ° glancing angle. In both the as – deposited and Pd –doped ZnO thin films, the X – ray diffraction spectra possess one sharpand three small peaks. It means that the film is polycrystalline with crys-tal planes (100), (002), (101) and (102). The film is crystallized in thehexagonal wurtzite phase and presents a preferential orientation along thec – axis indicated by the plane (002). The result is in a good agreementwith data mentioned in the literature (JCPDF card no 36-1451) [109]. Thestrongest 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 orientationslike (100) and (102) at 2𝜃 = 31.7013 ° and 47.4424 °, respectively arealso seen with comparatively lower intensities. Therefore, the crystallitesare highly oriented with their c – axes perpendicular to the plane of thesubstrate. It is worth to mention here that a small intensity peak appearsin the Pd – doped film at 2𝜃 = 40.32 o which belongs to the plane (111) ofthe 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 dataas 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. 98
  • 119. 0.1958 °., while 2𝜃and d values are given in Tables 4.2 and 4.3, respec- 2500 XRD 6000 SHIMADZU XR-Diffractometer tively. (002) 2000 1500I [CPS] 1000 500 (101) (100) (102) 0 20 25 30 35 Theta - 2Theta [Degree] 40 45 50 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. 1800 (002) XRD 6000 SHIMADZU XR-Diffractometer 1600 1400 1200 1000 I [CPS] 800 600 400 (101) (100) Pd (102) 200 (111) 0 20 25 30 Theta 35 - 2Theta [Degree]40 45 50 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. 99
  • 120. Table 4.2: Crystalline structure, Miller indices and d spacings of the as – deposited ZnO crystal planes. Integrated 2Theta FWHM IntensityPeak No. dExp. dTheo I/I1 Int. deg. deg. counts 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 578Table 4.3: Crystalline structure, Miller indices and d spacings of the Pd – doped ZnO crystal planes. Integrated Peak 2Theta FWHM Intensity dTheo Å dExp. Å I/I1 Int. No. deg. Deg. counts 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 6304.3 Surface topography and morphology studies Figure 4.6 (a) shows the surface micrograph of zinc oxide filmprepared 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 otherindicating good adhesiveness of film with the substrate. The grains sizeseen is comparable with the value calculated from x-ray diffraction stud-ies. Al-Hardan et al., had a uniform distribution of RF – sputtered ZnOnanostructure grains with a similar average grain size diameter [12]. Filmprepared at 200 0C spraying temperature, displayed in the inset (b) of 100
  • 121. a bFigure 4.6: Scanning Electron Micrograph of ZnO film prepared at a) 400 0C and the inset b) 200 0Cfigure 4.6, demonstrates a discontinuous nature. E. Arca et al. believe thatat low temperature the droplet splashes onto the substrate with lesserdecomposition which leads to porous and less adhesive film which couldsometimes be observed visually [71]. The surface morphology of the undoped ZnO films as observedfrom the AFM micrograph (figures 4.7 and 4.8) confirms that the grainsare uniformly distributed within the scanning area (5 μm Χ 5 μm), withindividual columnar grains extending upwards. This surface characteristicis important for applications such as gas sensors and catalysts [101]. Itwas found that using isopropyl alcohol organic solvent other than water ispreferred. This is due to a better droplet size distribution and, also, due toadditional heat transfer toward the sample surface resulted from alcoholburning [71]. The root mean square (rms) of the film surface roughnessdeposited at 450 0C using precursor of zinc acetate dissolved in distilled 101
  • 122. 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 onglass substrate at 450 oC spraying temperature with the precursor of 0.2 M zinc acetatedissolved in 100 mL distilled water. 102
  • 123. 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 onglass substrate at 450 oC spraying temperature with the precursor of 0.2 M zinc acetatedissolved in 100 mL isopropyl alcohol. 103
  • 124. water is about 17 nm, indicating that the surface of the spray depositedZnO thin film is very smooth. This value increases to 25.8 nm and thegrain size decreases from 250 to 65 nm as zinc acetate is dissolved inisopropyl alcohol organic solvent. The higher nucleation with a lower growth rate, results in a finegrains of the films. Further, the grain size of the film can also be deducedfrom 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 130nm as illustrated in figure 4.9. The statistical mean grain size of the film 100 120 Granularity Cumulation Distribution Chart 100 80 80 Cumulation %Percentage % 60 60 40 40 20 20 0 0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 Diameter nm 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 nmFigure 4.9: Granularity cumulation distribution report of ZnO thin film deposited at 4500 C on glass substrate using 0.2 M zinc acetate in distilled water precursor solution.deposited at 450 °C is about 57.76 nm, which is a little bit higher than thecrystallite size calculated from the XRD profile (49.15211 nm). 104
  • 125. 4.4 Optical properties Figure 4.10 shows the optical transmittance spectra of the ZnO thinfilms. Approximately, all the films demonstrate more than 60% transmit-tance at wavelengths longer than 400 nm, which is comparable with thevalues for the ZnO thin films deposited by Ju-Hyun Jeong [69] usingelectrostatic spray deposition ESD method, P. P. Sahay [80] using SPDmethod, B. J. Babu [81] by ultrasonic spraying USP scheme. Below 400nm there is a sharp fall in the T% of the films, which is due to the strongabsorbance of the films in this region. It has been observed that the over-all T% increases with the decrease in the film thickness. This happensdue to the overall decrease in the absorbance with the decrease in filmthickness [80]. The relationship between the morphology of the sampleand the solvent composition is straightforward. A smooth, homogeneous,good quality layer can be obtained by just using organic solvents (isopro- 1 189 nm 0.9 0.8 279 nm 523 nm 0.7 Transmission 0.6 613 nm 0.5 0.4 0.3 0.2 0.1 0 200 300 400 500 600 700 800 900 Wavelength nmFigure 4.10: Transmission spectra of ZnO thin films of different thicknesses sprayedon – glass at 400 0C temperature. The precursor was 0.2 M zinc acetate dissolved indistilled 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. 105
  • 126. 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 2.5 613 nm 2 523 nm 1.5 Absorbance 279 nm 1 189 nm 0.5 0 200 300 400 500 600 700 800 900 Wavelength nmFigure 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 distilledwater. 106
  • 127. follows [111]: 0 log = d log e → 2.30258 A = d (4.1)where I0 and I are the intensities of the incident and transmitted lightrespectively, 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)2versus 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 portionof the plot (αhυ)2 versus (hυ) for zero absorption coefficient value givesthe 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 between3.21 eV to 3.224 eV as the film thickness is changed from 189 nm to 613 16 Eg= 3.220 eV, t=279 nm 14 12 Χ1010 Eg= 3.216 eV, t=523 nm 10 Eg= 3.224 eV, t=189 nm (αhν)2 cm-2 . eV2 8 Eg= 3.210 eV, t=613 nm 6 4 2 0 2 2.5 3 3.5 4 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. 107
  • 128. 3.25 Energy gap Eg eV 3.2 3.15 100 200 300 400 500 600 700 Film thickness t nm Figure 4.13: Relationship of the extrapolated energy gap Eg of sprayed ZnO thin films at different film thicknesses.nm. As it is obvious from the plot above, the variation of thickness t forthe sprayed ZnO thin films has no effect on the estimated values of Eg.4.5 Electrical properties4.5.1 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 thetemperature 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 – pyrolyzeddeposited zinc oxide films of 668 nm film thickness with temperature.The variation of the resistance with the temperature reveals that resistanceof the film decreases as the temperature increases from room temperature 108
  • 129. 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 maybe due to the saturation of the conduction band with electrons promotedfrom shallow donor levels caused by oxygen vacancies. At this point anincrease in temperature leads to a decrease in electron mobility and asubsequent increase in resistance. Similar observations are made by other 0.02 1000 0.018 900 0.016 800 0.014 700 Conductance S Resistance kΩ 0.012 600 0.01 500 0.008 400 0.006 300 0.004 200 0.002 100 0 0 0 50 100 150 200 250 300 350 400 Temperature ͦC Figure 4.14: The variation of resistance of the spray – pyrolyzed deposited zinc oxide film of 668 nm film thickness with temperature.research groups [104, 18]. According to Al-Hardan et. al., [12] below 1500 C temperature, oxygen adsorption at the surface is mainly in the form ofO− , while above 150 0C, chemisorbed oxygen is present in the form of 2O− or O−2. Due to the conversion of O− into O− or O−2, oxygen adsorbs 2the additional electron from the zinc oxide, which is attributed to increasein 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 greatlyaffected by the temperature variation, probably due to the equilibrium 109
  • 130. obtained between the two competing processes: thermal excitation ofelectrons and the oxygen adsorption. Finally, at temperature higher than300 0C the resistance decreases again, probably because of the dominantexcitation of electrons and desorption of electron species [12]. The tem-perature range 200 – 240 0C is suitable for sensor operation due to thesmall 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 10 UV - illuminated 8 Current μA 6 Dark 4 2 0 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 -2 -4 -6 -8 -10 Bias Voltage V Figure 4.15: The I–V characteristic at room temperature of undoped ZnO film of 1178 - nm thickness in dark and under UV illumination.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. 110
  • 131. 40 maximum Vacuum vacuum pump OFF 35 30 Current μA Vacuum 25 pump ON Atmospheric 20 air Atmospheric air 15 10 0 50 100 150 200 250 300 350 400 Time s Figure 4.16: The effect of vacuum on base line current of a ZnO thin film at 200 0C and 10 v bias voltage. It is generally accepted that oxygen is chemiadsorbed at a surfacesite 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 darkcurrent [112, 113]. The effect of temperature on the I – V characteristics is depicted infigure 4.17. It confirms the enhancement of the current with temperaturefrom 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 currentas the temperature is further raised above 200 0C indicating an increase inthe film’s resistance. This effect is observed in the chemisorption regionat 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 inthe resistance of the ZnO. This reaction can be expressed as follows: 111
  • 132. 1 O + e− → O− (4.3) 2 2 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. 15 36 0C 0C 36 50 0C 0C 50 10 100 0C 0C 100 200 0C 0C 200 5 300 0C 0C 300Current μA 0 -5 -10 -15 -12 -8 -4 0 4 8 12 Bias voltage v Figure 4.17: The I–V characterization of sprayed ZnO film in the temperature range from RT to 300 0C. 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 112
  • 133. 10000 RP RS CP Z Ω 0 0 10000 20000 30000 Z ΩFigure 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. 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 τ m = 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 + (4.4) (1 + ω2 C 2 2) ′′ ωC 2 2 Z =− (4.5) (1 + ω2 C 2 2) 113
  • 134. 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 90 3% H 2% 1% 80 H H 70 Current μA 60 50 40 30 0 100 200 300 400 500 600 Time sFigure 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. 114
  • 135. 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 equation2.30 with test gas mixing ratio C is illustrated in figure 4.20. The figuredisplays 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 60 55 Sensitivity % 50 45 40 0 0.5 1 1.5 2 2.5 3 3.5 Hydrogen : air mixing ratio %Figure 4.20: The sensitivity dependence of as – deposited ZnO sensor on hydrogen gas mixing ratio[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 Alelectrode/ZnO nanofilm interface and lack of adsorbed oxygen ions at thenanofilm surface to react with gas molecules [114]. [101] obtained a con-sistent behavior on ZnO thin film prepared by thermal oxidation exposedto H2 gas up to 120 ppm at 400 0C. 115
  • 136. Figure 4.21 exhibits the transient response as a function of H2 gasconcentration for the as deposited, 668 – nm thick, ZnO sensing elementat 210 0C. The sensitivity of the ZnO gas sensor increases as the H2 gasconcentration is increased from 1% (10000 ppm) to 3% (30000 ppm) andit drops relatively rapidly when the H2 gas is removed , indicating that the 60 3% 2% 50 1% 40 Sensitivity % 30 20 10 0 0 50 100 150 200 Time sFigure 4.21: Transient responses of ZnO thin film (668 nm thick) at 210 0C testingtemperature upon exposure to hydrogen gas of mixing ratios of 1%, 2%, and 3%respectively.gas sensor has a good response for different H2 concentrations. Besides, ittakes 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 responseand recovery of the sensor have the same monotonicity behavior as thehydrogen target gas concentration increases. They both decrease withincreasing hydrogen concentration up to 2% at which the lowest responseand recovery times of 21s and 15 s are observed. Figure 4.23 illustrates a 116
  • 137. 35 140 30 120 25 100 Response time s Recovery time s 20 80 15 60 10 40 5 20 0 0 0 0.5 1 1.5 2 2.5 3 3.5 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.comparison of the I – V characteristic curves of the undoped sensor from0 up to 10 v of 2–v increment both in atmospheric air and in three H2 gascontaining air ambients. A linear dependence is dominant for the measured maximum cur- 9 5% H2 8 3%H 1% H2 7 Maximum current Imax mA 6 5 Air 4 3 2 1 0 0 2 4 6 8 10 12 Bias Voltage vFigure 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 117
  • 138. 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 PdCl 2 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 1800 H2 OFF H2 OFF 1600 1400 1200 Rise time = 3 sec Conductance μS 1000 Recovery time = 116 s 800 600 400 200 H2 ON H2 ON 0 0 100 200 300 400 500 600 700 800 Time sec.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. 118
  • 139. 94.3787% respectively which are comparatively twice as that for theundoped ZnO sensor. These drastic performance enhancements in theZnO 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 reactionactivation energy and the low grain size ZnO crystallites of the preparedsensing 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 thehigh temperature required for the sensor operation (200 – 500 ºC). Forthat reason, the effect of the operation temperature on the thin films sensi-tivity was studied with the aim of optimizing the operation temperature tothe lowest possible value. The gas sensitivity tests performed at room temperature show novariation 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 ofhow the maximum conductance Gmax depends on the temperature T for 3500 Maximum Conductance Gmax μS 3000 2500 2000 1500 1000 500 0 0 50 100 150 200 250 300 350 400 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. 119
  • 140. hydrogen sensor based on surface – promoted, with Palladium, ZnO sens-ing layer of about 1178 nm thickness. It is seen that the film maximumconductance Gmax goes through a maximum on changing T, with the bestoperating temperature at around 280 ºC. Roughly speaking, the increaseof Gmax (the left side of the maximum) results from an increase in the rateof surface reaction of the target gas, while the decrease of Gmax (the rightside) results from a decrease in the utility of the gas sensing layer. At thetemperature of the maximum conductance (response), the target gas mol-ecules have optimum penetration depth into the gas sensing grains (largeutility) i.e., optimum reactivity for the diffusion in the whole sensinglayer, 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 metaloxide gas sensors [48]. It is evident from the figure that the ZnO film shows a negativetemperature coefficient of resistance (NTCR) up to ~ 270 oC, whereasabove 285 oC it shows a positive temperature coefficient (PTCR). Similarbehavior was obtained by other researchers [18, 101]. The sensitivity of the sensor was calculated after the response hadreached steady state condition as a function of the operating temperaturein the range of 150 to 350 oC with temperature increment of 50 oC. Thetest 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. Thevariation of sensitivity with the operating temperature is shown in figure4.26. The sensitivity increased as the operating temperature increased,reaching a maximum value (~ 97%) at 250 oC, and decreased thereafterwith further increase in the operating temperature. It is suggested that 250 120
  • 141. 100 90 Sensitivity % 80 70 60 0 50 100 150 200 250 300 350 400 Temperature 0C Figure 4.26: The variation of sensitivity with the operating temperature of the Pd – doped ZnO gas sensor.o C is the optimum operating temperature for high sensitivity of the sen-sor. The sensitivity as well as response time is temperature dependentsince the chemical kinetics governing the solid-gas interface reaction istemperature 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 mixingratio and at three different testing temperatures of 250, 350, and 300 0Csuccessively. 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 atmosphericair. As it is obvious from the figure, maximum sensitivity S of 93.01228% is obtained at a temperature around 300 0C with a comparatively fastresponse time of ~ 4 s and a baseline recovery time of 72 s. 121
  • 142. 100 3 1 80 2Sensitivity % 60 40 20 0 0 50 100 Time sFigure 4.27: Transient responses of Pd – sensitized ZnO thin film (245 nm thick) asexposed 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. There seems to be no noticeable difference in sensing behavior ofthe ZnO hydrogen gas sensor as its surface is promoted by either dippingor spraying. However, it is worth to mention here that the palladium layerover the ZnO surface, resulted by spraying, looks more homogeneousthan that obtained by dipping.4.8 Tin oxide (SnO2) hydrogen gas sensor4.8.1 Crystalline structure and morphology of undoped SnO2 thin film It is known that tin dioxide SnO2 has a tetragonal rutile crystallinestructure (known in its mineral form as cassiterite) [118]. The unit cellconsists of two metal atoms and four oxygen atoms. Each metal atom issituated amidst six oxygen atoms which approximately form the cornersof a regular octahedron. Oxygen atoms are surrounded by three tin atomswhich 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-raydiffraction (XRD) pattern of the SnO2 thin film prepared on glass sub- 122
  • 143. strate at 450 °C spraying temperature. The major diffraction peaks ofsome lattice planes can be indexed to the tetragonal unit cell structure ofSnO2 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 crystalplanes peaks of (110), (101), (200), (211), (220), and (002) respectively. 160 140 (101) 120 (110) 100 Intensity I CPS 80 (211) 60 (200) 40 (220) (002) 20 SnO2 0 15 20 25 30 35 40 45 50 55 60 65 Theta 2 -Theta degrees Figure 4.28: X-ray diffraction (XRD) pattern of SnO2 thin film spray pyrolyzed on glass substrate at temperature of 450 oC.No characteristic peaks belonging to other tin oxide crystals or impuritieswere detected. In our films, the XRD spectrum showed predomination ofthe peaks, corresponding to reflection from the crystallographic (110),(101) planes, parallel to the substrate. The intensity of other peaks issmall. It indicates that the current films are textured. At that, the degree ofthe texturing depends on kind of sprayed solution we used, and increaseswhile using water solution instead of alcohol solution of SnCl2. 123
  • 144. The high intensity of these peaks suggests that these thin filmsmainly 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 of2000 nm x 2000 nm. The average roughness, Ra of the sample is of theorder of 1.26 nm, whereas the peak – to – valley roughness, RPV takesvalue 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. 124
  • 145. phology of SnO2 thin films is almost perfectly smooth with nanosizegrains. 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 thefilm 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 %. 100.00% t=145 nm 80.00% t=240 nm Transmission % 60.00% t=466 nm 40.00% 20.00% 0.00% 200 300 400 500 600 700 800 900 Wavelength nm Figure 4.30: Transmission spectra of undoped SnO2 thin films of different thicknesses deposited at 450 oC on glass substrates. 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 thicknessas 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 thicknessfrom 145 nm to 466 nm. The decrease of band gap with the increase offilm thickness implies that SnO2 is an n-type semiconductor [80]. This 125
  • 146. 25 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 20 Χ1010 15 (αhν)2 eV2 cm-2 10 5 0 1.5 2 2.5 3 3.5 4 4.5 hν 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.decrease of band gap may be attributed to the presence of unstructureddefects, which increase the density of localized states in the band gap andconsequently 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 hydrogengas of different mixing ratios has been explored. The successive testswere 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 hydrogengas increases linearly with H2 test gas mixing ratio up to 3% H2 (S~83%)after which the sensitivity tends to saturate with increasing the analytegas. A maximum sensitivity value of 88% has been registered when thesensor is exposed to 4% H2 gas in air (figure 4.33). Moreover, it’s worthto indicate here that both the response and recovery times of the undopedSnO2 gas sensor decrease with increasing the hydrogen gas mixing ratio, 126
  • 147. 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. 100 4% H2 90 3% H2 80 70 2% H2 Sensitivity S % 60 50 1% H2 40 30 20 10 0 0 500 1000 1500 Time t sFigure 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. 100 90 80 Sensitivity S % 70 60 50 40 30 0% 1% 2% 3% 4% H2:air mixing ratio C %Figure 4.33: Sensitivity versus H2 gas concentration of undoped tin oxide SnO2 thinfilm. The bias voltage was 5.1 v with the temperature set to 210 0C. 127
  • 148. 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 SnO 2 – 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 700 4.5% H2 600 pulse due to H2 3.3% H2 remaining in the tubing of H2 when the manifold is cracked 500 open; NF is still closed 2% H2Current μA 400 1% H2 Current 300 increased upon switching ON of rotary - from atmosphere to 200 vacuum 0.5% H2 100 0 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 Time s 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. – 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- 128
  • 149. ure 4.35 exhibits the transient sensitivity S variation with hydrogen gasconcentration. The relationship of sensor response time versus hydrogen:air mix-ing ratio is plotted in figure 4.36. The maximum sensitivity obtained is95.744 % at 4.5% H2 gas concentration in air which is slightly less thanthat reported by Mitra [117]. Also, the shortest response time of about 24s was observed at 2% hydrogen gas concentration. This speed of responseis faster than that obtained by Sunita Mishra et al. [120] 100 3.3% H2 4.5% H2 80 2% H2 Sensitivity % 60 1% H2 40 0.5% H2 20 0 0 250 500 750 1000 1250 1500 1750 2000 2250 Time s 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. Figure 4.37 shows the switching behavior to 4.5% H2 gas of theSnO2 gas sensor followed film surface modification with 20 palladiumchloride PdCl2 layers. The response for three operating temperatures iscompared. 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. 129
  • 150. 80 100 70 80 60 Response time s 50 Sensitivity % 60 40 40 30 20 20 10 0 0 0 1 2 3 4 5 H2 mixing ratio %Figure 4.36: Sensitivity and Response time as a function of the H2 test gas mixingratio. The test was performed at 210 0C and 10 v bias on SnO2 sample sprayed overthe IDE and surface coated with 20 PdCl2 layers sprayed at 400 0C over the film. 100 210 0C 80 175 0C 60 Sensitivity % 40 150 0C 20 0 0 100 200 300 400 500 Time sFigure 4.37: Transient responses of SnO2 thin film of 248 nm thick at 150, 175, and210 0C testing temperature upon exposure to 4.5% H2:air gas mixing ratio. 130
  • 151. 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. 700 600 500 Maximum current Imax. μA 400 300 200 100 0 100 125 150 175 200 225 250 275 300 Temperature T 0CFigure 4.38: variation of sensor response current with temperature of Pd - doped SnO2thin film exposed to 4.5% hydrogen gas mixing ratio in air and at 10 v bias voltage. 131
  • 152. 4.7 Conclusions and future work proposals In this study, the influence of thin film processing conditions onthe 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 simpleand reliable deposition technique in acquiring high quality thin films ofdifferent types and at a wide range of deposition temperatures. Thismethod has the advantages of low cost, easy-to-use, safe and efficientroute to coat large surface areas in mass production. The spray pyrolyzed ZnO thin films obtained from non – organicsolvent (distilled water) are observed to be comparatively of less trans-parency, with thickness of about 1178 nm, as measured by weighingdifference method and confirmed via the SEM unit. The XRD diffraction pattern proves that the prepared ZnO on glasssubstrate is highly c – axis oriented, giving a peak at Bragg angle equal to34.38 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 theprepared films contain high carrier concentrations. AFM investigations reveal a porous morphology of spherical parti-cles. The electrical characterization of the sprayed thin films shows thatthey are highly resistive, but that their properties vary considerably whenthe 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 andoperating temperatures, showing excellent sensitivity. It is found that the 132
  • 153. sensing of hydrogen gas in our metal oxide sensors is related to the en-hancement of adsorption of atmospheric oxygen. The excellent selectivityand the high sensitivity for hydrogen gas can be achieved by surfacemodification of ZnO films. The observed conductance change in Pd –doped ZnO sensors after exposure to H2 gas (3%) is about two times aslarge as that in the undoped ZnO sensors. The variation of the operating temperature of the film leads to asignificant change in the sensitivity of the sensor with an ideal operatingtemperature 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 sprayingwith) in palladium chloride solution have the highest sensitivity of 97%and extremely short response time of 3 s, which fit for practice since it iscrucial to get fast and sensitive gas sensor capable of detecting toxic andflammable gases well below the lower explosion limit (4% by volume forH2 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 bulkyand requires a 220 v A.C current source, consuming much power. Thisissue can be handled by applying a thin film heating element on the backof 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 HClequivalent to 93.2% SnO2, 5.5% Sb2O3 and 1.3% ZnO. Using this ap- 133
  • 154. proach, Mochel [64] obtained such thin films of an electrical resistance of42 ohms per square. This value increased to 56 ohms per square uponpassing an alternating current, equivalent to 300 Watts at 110 v, to thedeposited film. Doing so, the sample temperature increased to 825 0C injust a few minutes. Seven heating and cooling cycles caused no substan-tial change in the resistance and other properties. 134
  • 155. 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 135
  • 156. 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. 136
  • 157. [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). 137
  • 158. [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). 138
  • 159. [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). 139
  • 160. [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). 140
  • 161. [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). 141
  • 162. [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). 142
  • 163. ‫جمهوريـــــة العــــــراق‬ ‫وزارة التعليــــم العالــــي والبحـــث العلمــــي‬ ‫جامعـــــــــة بغــــــــــداد‬ ‫كليــــــــــــة العلـــــــــوم‬ ‫تحسين متحسس أوكسيد الزنك وأوكسيد القصدير‬ ‫لغـاز الهايدروجين‬ ‫أطروحة مقدمة إلى‬ ‫مجلس كلية العلـوم – جامعـة بغـداد‬ ‫وهي جـزء من متطلبـات نيـل درجـة دكتـوراه فلسفــــة‬ ‫في الفيزيـاء‬ ‫من قبل‬ ‫قحطــان كـاطــع حيــال‬ ‫بكالوريوس علوم في الفيزياء 4119‬ ‫ماجستيـــــر علوم في الفيزياء 1119‬ ‫أشــــراف‬‫د. وسـن رشيـــد صالـح‬ ‫د.عبـدهللا محســن سهيــل‬‫9911 ميالدي‬ ‫1449هجري‬ ‫‪I‬‬
  • 164. ‫الخالصــــــه‬‫تم تحضير اغشيه نانويه ألوكسيد الزنك ‪ ZnO‬وأوكسيد القصدير المطعم بالباليديوم على قواعد‬ ‫ُ‬‫زجاجيه باستخدام طريقه التحلل الكيميائي الحراري وتم تفحصها كمتحسس سريع االستجابه لغاز‬‫الهايدروجين المختزل. أستخدمت تقنية التحلل الكيميائي الحراري بدرجه حراره حوالي 054‬‫درجه مئويه والهواء الجوي كغاز حامل لتحضير األغشيه المتحسسه ألوكسيد الزنك وأوكسيد‬‫القصدير. كانت محاليل التذريه التي طبقت هي كل من أمالح كلوريد الزنك 2‪ ZnCl‬وأسيتات‬‫الزنك ‪ Zn(CH3COO)22H2O‬بينما أستخدمت ماده كلوريد القصدير ‪SnCl2.2H2O‬‬‫للحصول على أوكسيد القصدير. أثبتت تقنيه التحلل الكيميائي الحراري بساطتها ومعوليتها في‬‫الحصول على أغشيه رقيقه متعدده التبلور وذات طور )200( وبأتجاه على امتداد المحور ‪c‬‬‫لتركيب أوكسيد الزنك السداسي كما اظهرته تحليالت التركيب لحيود األشعه السينيه. أظهرت‬‫األغشيه المحضره نفاذيه عاليه عند المدى المرئي للطيف الكهرومغناطيسي وبمعدل وصل الى‬‫حوالي %59 وعتبة قطع عند األشعه فوق البنفسجيه ذات الطول الموجي 184 نانومترتقريبا.‬‫ازدادت النفاذيه وليس فجوة الطاقه المباشره المحسوبه لألغشيه بنقصان سمك الغشاء الرقيق.‬‫أظهرت نتائج الدراسات السطحيه بالمجهر االلكتروني الماسح ومجهر القوه الذريه ألغشيه‬‫اوكسيد الزنك تشكل توزيع منتظم لحبيبات مساميه نانويه التركيب دائريه الشكل ذات اقطار‬‫بحدود 11 نانومتر. بينت الخصائص الكهربائيه لألغشيه الرقيقه المحضره بهذه التقنيه مقاوميتها‬ ‫العاليه وان هذه الخصائص تغيرت تغيراً ملحوظا ً عند أجراء القياسات في الهواء اوفي الفراغ.‬‫ابدى كل من متحسسسا اوكسيد المعدن ‪ ZnO‬و 2‪ SnO‬المطُعم تحسسيه فائقه حيث أزدادت‬ ‫المواصله له بتعرضه لغاز الهايدروجين ذا تراكيز متنوعه وعند درجات حراريه مختلفه.‬‫وجد أن تحسس متحسس أكاسيد المعادن لغاز الهيدروجين يتعلق بتحسن امتزاز االوكسجين‬‫الجوي. االنتقائيه الممتازه والتحسسيه العاليه لغاز الهيدروجين يمكن تحقيقها بالمعامله السطحيه‬‫ألغشيه اوكسيد الزنك/أوكسيد القصدير. لقد كان مقدار التغير بالمواصله ألوكسيد الزنك المطُعم‬‫بالباليديوم بتعرضه لغاز الهايدروجين ذا تركيز 4% حوالي مرتان بقدر مثيلتها للغشاء الغير‬ ‫المطعم.‬ ‫ُ‬‫أدى تغيير درجة الحرارة التي يعمل عندها متحسس أوكسيد الزنك الى تغير ملحوظ في حساسيته‬‫حيث كانت درجة الحراره المثلى لألشتغال حوالي 52 ± 052 درجه مئويه أنخفضت بعدها‬ ‫حساسية المتحسس. كان تغير التحسسيه لغشاء أوكسيد الزنك خطيا ً بزياده تركيز الغاز.‬ ‫‪II‬‬
  • 165. ‫يتميز زمن األستجابه – األفاقه لمادة أوكسيد الزنك المطعم بالباليديوم بكونه مفرط القصرنسبيا.‬ ‫ُ‬‫لقد اظهرت األغشيه المغطسه 11 مره في كلوريد الباليديوم اعلى تحسسيه قدرها 11% وزمن‬‫استجابه مفرط القصر 4 ثانيه وهذا مالئم عمليا ً طالما أنه يعتبر من االمور الحاسمه الحصول‬‫على متحسس غازي سريع االستجابه والتحسس له القدره على كشف الغازات الملتهبه والسامه‬‫وعند تراكيز قليله دون الحد االدنى ألنفجار هذه الغازات (لغاز الهيدروجين يكون هذا الحد 4‬ ‫%).‬‫ولعناصر التحسس ألوكسيد القصدير فقد كانت درجه حرارة االشتغال المثلى 191 درجه مئويه‬ ‫وبنسبه مئويه للتحسس قدرها 441.41 % لغاز هيدروجين بتركيز قدره 5.4 %.‬ ‫‪III‬‬

×