|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2010 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004549
Design and characterization of microwave assisted plasma spray deposition system :
b application to eu doped y2o3 nano-particle coatings
h [electronic resource] /
by Marek Merlak.
[Tampa, Fla] :
University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains X pages.
Thesis (MS)--University of South Florida, 2010.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
ABSTRACT: This thesis presents a Microwave Plasma Assisted Spray Deposition (MPASD) system design, characterization, and application to produce nano-sized particle coatings of metal oxides. A commercially available rectangular waveguide microwave power delivery system is utilized to initiate and sustain the plasma discharge within the customized plasma applicator where micron-sized droplets of a metal ion solution are heated to evaporate the solvent and thermally process the resulting nano-sized particles. The investigation of optimum conditions for oxygen, argon, and air plasma ignition in the MPASD system was presented. Measured electron temperature of the plasma was between 6000K and 40000K for the plasma conditions used in the MPASD process. Successful deposition of Y2O3:Eu nano-particles using the MPASD system was achieved. MPASD process allows control of the particle's properties, shown through XRD and photoluminescence studies of the Y2O3:Eu coatings. The MPASD process settings effect on particles activated doping concentration and, as a result, its photoluminescence was shown
Advisor: Sarath Witanachchi, Ph.D.
t USF Electronic Theses and Dissertations.
Design and Characterization of Microwave Assisted Plasma Spray Deposition System: Application to Eu Doped Y 2 O 3 Nano Particle Coatings by Marek Rados aw Merlak A thesis submitted in partial fulfillment of the requirements for the degree of M aster of Science Department of Physics College of Arts and Sciences University of South Florida Major Professor: Sarath Witanachchi, Ph.D. Pritish Mukherjee, Ph.D. Myung Kim, Ph.D. Date of Approval: May 14 2010 Keywords: Py rolysis, Microwave Plasma, Y 2 O 3 :Eu P hosphor, Nano Particle Copyright 2010 Marek Rados aw Merlak
i TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ........... iii LIST OF FIGURES ................................ ................................ ................................ ......... i v ABSTRACT ................................ ................................ ................................ .................... x i CHAPTER 1. Nanotechnology ................................ ................................ ........................ 1 1.1. History of Nano Technology ................................ ................................ ............. 2 1.2 Nano Particles ................................ ................................ ................................ .... 6 1.2 .1. Surface Effect ................................ ................................ ........................ 6 1.2 .2. Quantum Effect ................................ ................................ ..................... 8 1.2 .3. Nanoparticle Properties ................................ ................................ ....... 10 1.3 Hazards of Nanotechnology ................................ ................................ ............ 11 1.4. Thesis Outline ................................ ................................ ................................ 1 3 CHAPTER 2. Spray Pyrolysis ................................ ................................ ....................... 14 2.1. Spray Pyrolysis Overview ................................ ................................ ................ 14 2.2. Spray Pyrolysis Atomization Techniques ................................ ........................ 1 6 2 .3. Spray Pyrolysis Heat Sour ces ................................ ................................ .......... 19 2.4. Chapter Summary ................................ ................................ ............................. 22 CHAPTER 3. Microwave Assisted Spray Deposition System Design ......................... 23 3.1. Microwave Power Delivery Subsystem ................................ ............................ 25 3.1.1. Waveguide Theory ................................ ................................ ................ 27 3.1.2. Microwave Power Delivery Subsystem Components ........................... 28 3.2. Materia l Delivery Sub system ................................ ................................ ......... 39 3.2.1. Precursor Atomizer ................................ ................................ ............... 39 3.2.2. Gas Mass Flow Controller ................................ ................................ .... 43 3.3. Deposition Chamber ................................ ................................ ........................ 4 5 3.4. Vacuum Subsystem ................................ ................................ .......................... 46 3.5. Control and Monitoring Sub System ................................ ............................... 48 3.6. Chapter Summary ................................ ................................ ............................ 49 CHAPTER 4. Microwave Plasma Assisted Spray Deposition System Characterization ................................ ................................ ................................ .............. 51 4.1. Plasma Temperature Model ................................ ................................ .............. 51
ii 4.2. Emission Spectroscopy for Plasma Temperature Determination .................... 58 4.2.1. Spectroscopic Temperature Measurement Theory .............................. 58 4.2.2. Spectroscopy Exp erimental Setup ................................ ....................... 59 4.2.3. Spectroscopy Experimental Results ................................ ..................... 60 4.3. Plasma Ignition Study ................................ ................................ ...................... 64 4.3.1. Plasma Ignition Experimental Setup ................................ .................... 66 4.3.2. Plasma Ignition Study Initial Results ................................ ................... 6 9 4.3.3. Oxygen Plasma Ignition Study Results ................................ ................. 71 4.3.4. Argo n Plasma Ignition Study Results ................................ .................. 73 4.3.5. Ai r Plasma Ignition Study Results ................................ ....................... 74 4.3.6. Plasma Ignition Study Summary ................................ .......................... 7 5 4.4. Plasma Tail Temperature Study ................................ ................................ ....... 7 6 4.4.1. Plasma Tail Temperatur e Study Experimental Setup .......................... 7 6 4 .4.2. Plasma Tail Temperature Study Results ................................ .............. 7 7 4.4.3. Plasma Tail Temperature Study Summary ................................ .......... 80 4.5. Atomizer Generation Rate ................................ ................................ ............... 8 1 4.6. Chapter Summary ................................ ................................ ............................ 8 2 CHAPTER 5 Case Study: Y 2 O 3 :Eu Nanophosphor ................................ ...................... 8 3 5.1. Y 2 O 3 and Y 2 O 3 :Eu Overview ................................ ................................ ........... 8 3 5.2. Thermolysis of Y(NO 3 ) 3 powder ................................ ................................ .... 8 6 5.3. Deposition of Y 2 O 3 :Eu Y 2 O 3 :Eu Coatings using MPASD System .................. 8 8 5.4. Morphology of Y 2 O 3 :Eu Coatings ................................ ................................ ... 90 5.5. Growth Rate Study of Y 2 O 3 :Eu Coatings ................................ ........................ 9 7 5.6. Crystal Structure Study of Y 2 O 3 :Eu Coatings ................................ ................ 10 1 5.7. The energy dispersive X r ay spectrum of Y 2 O 3 :Eu coatings ........................ 10 7 5.8. Anneal Study of Y 2 O 3 :Eu Coatings ................................ ............................... 10 8 5.9 Photoluminescence of Y 2 O 3 :Eu CoatingsChapter Summary ......................... 10 9 5.10. Chapter Summary ................................ ................................ .......................... 1 20 LIST OF REFFERENCES ................................ ................................ .............................. 12 1 APPENDI CES ................................ ................................ ................................ .............. 1 30 A PPENDIX A: Electric D iagrams ................................ ................................ ........... 1 31
iii LIST OF TABLES Table 2.1. Comparison of different liquid atomization techniques (Lucas 1 994) ......... 16 Table 3.1. The commonly used ISM ba nds of radio frequency spectrum ..................... 26 Table 3.2. List of major components of microwave power delivery subsystem of MPASD system ................................ ................................ ............................. 29 Table 3.3. Major components list of the material delivery subsystem of the MP ASD system ................................ ................................ ............................. 39 Table 3.4. List of major components of v acuum subsystem of MPASD system .......... 47 Table 3.5. List of major components of control and monit oring subsystem of MPASD system ................................ ................................ ............................. 48 Table 4.1. Wavelength, transition probability, upper state statistical weight, and upper state energy for selected strong emission lines of neutral oxygen ato m. (nist.gov) ................................ ................................ ............................. 63 Table 4.2. List of major components of plasma ignition study experiment al setup ...... 67 Table 4.3. MPASD process setting for opt imum plasma ignition conditions ............... 7 5 Table 5.1. MPASD process conditions for Y 2 O 3 :Eu coating deposition ....................... 8 9 Table 5.2. X ................................ .............. 10 1 Table 5.3. The particle diameter calculation results using the broadening of XRD diffrac tion peaks for selected samples ................................ ........................ 10 3
iv LIST OF FIGURES Figure 1.1. Timeline of nanotechnology selected events. ................................ ................ 3 Figure 1.2. Example of surface atoms and surface vacancies in cubic structure. ............ 8 Figure 1.3. The de nsity of states function for bulk (3D), thin film (2D), nanowire (1D) and quantum dot (0D). ................................ ................................ ............ 9 Figure 2.1. Spray pyrolysis proc ess. The pyrolysis occurs (a) in flight of the atomized precursor droplets or (b) upon contact on a hot surface. .............. 15 Figure 2.2. Diagram of examples of atomization techniques. The (a) liquid jet, (b) internal air assisted, and (c) external air assisted nebulizer are shown. ........ 17 Figure 2.3. Diagram of examples of atomization techniques. The (a) jet instabilities and (b) surface instabilities atomizers are shown. ..................... 18 Figure 2.4. Electrostatic atomization technique. The (a) diagram of the apparatus an d (b) the principle of the drop formation. ................................ .................. 19 Figure 2.5. Selected pyrolysis processes. The (a) hot surface pyrolysis, (b) oven pyrolysis, and (c) laser assisted pyrolysis processes are shown. .................. 20 Figure 2.6. Plasma spray pyrolysis processes. The (a ) flame, (b) ICP plasma, and (c) microwave plasma assisted spray pyrolysis systems are shown. ............ 21 Figure 2.7. Material side injection technique. The material is injected into (a) the plasma and (b) the plasma tail. ................................ ................................ ..... 22 Figure 3.1. The Microwave Plasm a Assisted Spray Pyrolysis (MPASD) system component diagram. ................................ ................................ ...................... 24 Figure 3.2. Microwave power delivery subsystem diagram. ................................ ......... 25 Figure 3.3. Selection of the United States radio spectrum frequency allocations chart (year 2003). (US dept of commerce) ................................ ................... 2 5
v Figure 3.4. Wave propagating in a rectangular waveguide (TE01 mode). The relation between the free space propagation and waveguide propagation is shown. ................................ ................................ ................... 28 Figure 3.5. Magnetron head (a) location within MPASD, (b) picture of GA4002A magnetron head a nd SM745G power supply, and (c) operation diagram of magnetron launcher. ................................ ................................ ... 30 Figure 3.6. Waveguide isolator (a) location within MPASD, (b) picture of GA1107 waveguide isolator and (c) its operation diagram. ......................... 31 Figure 3.7. Waveguide directional coupl er (a) location within MPASD and (b) picture of GA3106 dual directional waveguide coupler. .............................. 32 Figure 3.8. Waveguide three stub tuner (a) location within MPASD, (b) operation diagram, and (c) picture of GA1002 precision 3 stub tuner. ........................ 32 Figure 3.9. Example circuit model of waveguide components. ................................ ..... 33 Figure 3.10. Plasma applicator (a) location within MPASD, (b) picture of GA6103 downstream plasma applicator, and (c) its detailed cross section. ................................ ................................ ................................ .......... 34 Figure 3.11. Plasma applicator redesign stages cr oss section diagram. Features marked red were new or redesigned from original. Seals are marked with solid black and coolant with light blue. ................................ ................ 35 Figure 3.12. Different flow patterns possible in plasma zone. Material flow is marked gray and gas flow is marked black. ................................ .................. 36 Figure 3.13. Plasma applicator final design. ................................ ................................ .. 37 Figure 3.14. Adjustable waveguide short (a) location within MPASD, (b) operation diagram, and (c) picture of GA1219A basic sliding short circuit. ................................ ................................ ................................ ........... 38 Figure 3.15. The diagram of u ltrasonic atomizer utilized in MPASD system. .............. 40 Figure 3.16. The effect of the precursor contaminate level on contaminate concentration within nanoparticle. The curves represent the contamination due to 0.01ppm, 0.4ppm, 1ppm, and 2ppm contaminated solv ent if used to produce 100nm Y 2 O 3 particles. .................. 42 Figure 3.17. Basic operation of gas mass flow meter (a) and the diagram of typical gas mass flow controller. ................................ ................................ .. 44 Figure 3.18. Picture and diagram of deposition chamber. ................................ ............. 45
vi Figure 3.19. Operational diagram of vacuum sub system. The different pressure conditions are shown. ................................ ................................ .................... 47 Figure 3.20. Front panel and program diagram (part) for MPASD control and monitoring program written in LabView. ................................ ..................... 49 Figure 4.1. The published data for heat capacity of ox ygen gas (engineeringtoolbox.com) as a function of temperature and the fitted curve. ................................ ................................ ................................ ............. 52 Figure 4.2. The plasma temperature as a function of absorbed microwave power. The oxygen gas temperature for 700sccm (solid line) and 5000sccm (dash line) gas flow w ere calculated. No radiation and conduction was included. ................................ ................................ ................................ ........ 53 Figure 4.3. The plasma temperature as a function of absorbed microwave power for (a) 5000 ccm and (b) 700 sccm oxygen flow rate. The thick solid, dash, and red thin lines presents data for 1 g/min, 0.1 g/min, and 0.01 g/min precursor (H 2 O) flow respectively. ................................ ..................... 54 Figure 4.4. Emissivity data (a) and the calculated plasma temperature as a function of absorbed microwave power for (b) 700 sccm and (c) 5000 sccm oxygen flow rate. The thick soli d, dash, and red thin lines present data for 1 g/min, 0.1 g/min, and 0.01 g/min precursor ................................ 55 Figure 4.5. Plasma temperature for selected plasma discharge size. The plasma calculation results for (a) 2 cm diameter sphere, (b) 3 cm diameter sphere, and (c) 3 x 3 x 5 cm ellipsoid are presented by dash, solid thick, and solid thin line respectfully. Calculated for precursor flow rate of 0.01g/hr of water. ................................ ................................ .......................... 56 Figure 4.6. Experimental setup diagram for plasma emission spectroscopy. ................ 59 Figure 4.7. Mic rowave plasma emission spectra. Plasma absorbed microwave power was 200 W, process pressure was 40 Torr, nebulizer duty cycle was ................................ ................................ ................................ ............... 60 Figure 4.8. Calculated electron temperature as a function of process pressure from the measured emission 777 nm and 84 4 nm triplet ratio. The data was collected at 400W absorbed microwave power, 700+18 sccm O 2 flow, and 3% nebulizer duty cycle. ................................ ............................... 6 2 Figure 4.9. Calculated electron temperature as a function of process absorbed microwave power from the measured emission 7 77 nm and 844 nm triplet ratio. The data was collected at 40 torr process Table 4.1. Wavelength, transition probability, upper state statistical weight, and upper state energy for selected strong emission lines of neutral oxygen atom. (nist.gov) ................................ ................................ ............................. 62
vii Fig ure 4.10. High frequency ionization coefficient of air as a function ...................... 6 5 Figure 4.11. Breakdown electric field for air as a function of pressure for parallel plate configuration at three different distances (Herlin 1948) ...................... 6 6 Figure 4.12. Diagram o f the experimental setup for the electrical discharge of oxygen and argon gas in MPASD system. ................................ .................... 6 7 Figure 4.13. Experimentally determined minimum microwave forward power level during electrical breakdown of oxygen gas in MPASD system as a function of pr essure at selected positions of waveguide short. .................. 69 Figure 4.14. Microwave power flow diagram for (a) ignition study experimental setup and (b) the setup modified to reduce reverse power. ........................... 70 Figure 4.15. The breakdown electric field strength of dif ferent gasses as a function of pressure. Data collected using system shown in figure 4.14b. ................................ ................................ ................................ ............. 71 Figure 4.16. Average microwave power level needed to cause electrical breakdown of oxygen gas as a function of (a, b, c) pressure and (d) waveguide short pos ition. The arrows show the direction of increasing waveguide short position (away from the plasma). ................................ ...... 7 2 Figure 4.17. 3D (a) and (b) contour plot of experimentally obtained average microwave power level needed for oxygen plasma ignition as a function of position and pressure. ................................ ................................ 7 3 Figure 4.18. 3D (a) and (b) contour plot of experimentally obtained average microwave power level needed for argon plasma ignition as a function of position and pressure. ................................ ................................ ............... 7 4 Figure 4.19. 3D (a) and (b) contour plot of expe rimentally obtained average microwave power level needed for air plasma ignition as a function of position and pressure. ................................ ................................ .................... 74 Figure 4.20. Plasma tail temperature study experimental setup. ................................ ... 76 Figure 4.21. Plasma tail temperature as a function of prob e position for selected material flow rates. The process pressure was 60 Torr and process power was 1000 W of absorbed power and nebulizer duty cycle was 3%. ................................ ................................ ................................ ................ 77 Figure 4.22. Plasma tail temperature as a function of process pressure for selected probe p ositions. The process power was 600 W, the process gas flow was 700 + 18 sccm O2 and nebulizer duty cycle was 3%. ........................... 78
viii Figure 4.23. The plasma tail temperature as a function of nebulizer duty cycle for probe positions of (a) +4 mm, (b) +44 mm, and (c) 84 mm away from plasma zone. The process power was 600 W, process pressure was 60 Torr and the O2 gas flow was 700 + 18 sccm. ................................ .............. 7 9 Figure 4.24. Plasma tail temperature as a function of absorbed microwave power for selected probe positions. The process pres sure was 60 Torr, the process gas flow was 700 + 18 sccm O2 and nebulizer duty cycle was 3%. ................................ ................................ ................................ ................ 80 Figure 4.25. The atomizer generation rate as a function of power duty cycle. .............. 81 Figure 5.1. High pressure transition of (a, b) Y 2 O 3 and (c, d) Eu:Y 2 O 3 Reproduced from Wang (2009). X . ................ 8 4 Figure 5.2. The Crystal structure of (a) cubic, (b) hexagonal, and (c) monoclinic Gd2O3 crystal. Reproduced from Zhang (2008). ................................ ......... 8 4 Figure 5.3. The cation site coordination for (a) cubic Y 2 O 3 and (b, c, d) monoclinic Eu2O3. Reproduced from Maslen (1996) and Yakel (1978). ................................ ................................ ................................ ........................ 8 5 Figure 5.4. Powder XRD patterns of body centered cubic and base centered monoclinic Y 2 O 3 X . ................................ ........ 8 6 Figure 5.5. The powder XRD patterns of a nnealed Y(NO 3 )3xH 2 O reagent. Some of the XRD peaks of BCC Y 2 O 3 (solid diamond), YO(NO 3 ) (open diamond), Y(NO 3 ) 3 3H 2 O (circle), Y(NO 3 ) 3 5H 2 O (triangle), and Y(NO 3 ) 3 6H 2 O (sqare) are marked. ................................ .............................. 8 7 Figure 5.6. The processed nanoparticle size as a function o f precursor concentration for MPASD system for Y 2 O 3 process using Y(NO 3 ) 3 solution precursor. ................................ ................................ ......................... 8 8 Figure 5.7. Picture of selected coatings. Samples S15 (a) and S16 (b) are shown. ...... 90 Figure 5.8. SEM images of coatings deposited at (a) 20 Torr and (b c, d) 60 Torr. The coating deposited at 60Torr is shown at three magnification levels. Samples S15 (a) and S16 (b, c, d) are shown. ................................ ............... 9 1 Figure 5.9. SEM images of a coating and the bald spot on the coating. The images were taken at same magnificatio n and 1mm apart. Sample S16 is shown. The images were taken at same magnification and 1mm apart. ................................ ................................ ................................ .............. 9 2 Figure 5.10. The droplet and resulting calculated particle size distribution. Data for droplet size distribution supplied by manufacturer. ................................ 9 3
ix Figu re 5.11. The TEM image of several particle clusters from Eu:Y 2 O 3 nanoparticle coating. ................................ ................................ ..................... 9 4 Figure 5.12. The STEM image of several particle clusters from Y 2 O 3 :Eu nanoparticle coating. ................................ ................................ ..................... 9 5 Figure 5.13. The histogram of particle size distribution of Y 2 O 3 : Eu ............................ 9 6 Figure 5.14. The normal (a) and high contrast (b) image of Y 2 O 3 coating deposited for 16 minutes. The observed area coverage was 41%. ............... 9 8 Figure 5.15. SEM images of Y 2 O 3 coating deposited for (a) 30s, (b) 1min, (c) 2min, (c) 4min, (d), 8m in, and (f) 16min. ................................ ..................... 9 9 Figure 5.16. Observed coating percentage coverage as a function of time for deposition of Y 2 O 3 coating (solid line). The dash line represents the calculated 6.4%/min growth. ................................ ................................ ...... 100 Figure 5.17. The powder XRD pattern of (a) Si 10 0 single crystal substrate and Y 2 O 3 :Eu sample deposited on Si 100 substrate under pressure of (b) 20 Torr and (c) 40 Torr. The monoclinic phase peaks are marked. ................. 10 2 Figure 5.18. The HRTEM image of Y 2 O 3 :Eu nanoparticle. The observed interplanar spacin g is 0.69 nm. ................................ ................................ ... 10 4 Figure 5.19. The HRTEM image of Y 2 O 3 :Eu nanoparticle. The observed interplanar spacing is 0.69 nm. ................................ ................................ ... 10 5 Figure 5.20. The HRTEM image of Y 2 O 3 :Eu nanoparticle. The observed interplanar spacing is 0.68 nm. ................................ ................................ ... 10 6 Figure 5.21. The H RTEM image of Y 2 O 3 :Eu nanoparticle. The observed interplanar spacing is 0.37 nm. The insert shows the SAED diffractogram. ................................ ................................ .............................. 10 7 Figure 5.22. The energy dispersive X ray spectrum of small cluster of Y 2 O 3 :Eu particles. ................................ ................................ ................................ ...... 10 8 Figure 5.23. The XRD pattern s for sample of Y 2 O 3 :Eu annealed at selected annealing temperatures. The diamond markers point to the characteristic XRD peak locations of base centered monoclinic phase of Y 2 O 3 :Eu. ................................ ................................ ................................ 10 9 Figure 5.24. Photoluminescence experiment setup diagram. ................................ ...... 1 10 Fi gure 5.25. Transmission spectrum of filters utilized in the luminescence experiment ................................ ................................ .............................. 11 1
x Figure 5.26. The power and pressure conditions for the thirty samples tested. The red bubble represents the photoluminescence detected. The varing red area re presents the relative intensity of photoluminescence. ...................... 11 2 Figure 5.27. Optical spectrum of the photoluminescence of Y 2 O 3 :Eu particles deposited at 600W absorbed power. The excitation spectrum was filtered using Hoya U340 glass. ................................ ................................ .. 11 3 Figure 5.28. The spectral intensity photoluminescence of Y 2 O 3 :Eu particles deposited at 40 Ttorr and 60 Torr process chamber pressure. The excitation spectrum was filtered using Hoya U340 glass. .......................... 11 3 Figure 5.29. Luminescence intensity as a function of sample depositio n Absorbed power. The excitation spectrum was filtered using Hoya U340 glass. ....... 11 5 Figure 5.30. Optical spectrum of the photoluminescence of Y 2 O 3 :Eu particles deposited at 600 W absorbed power. The excitation spectrum was filtered using Hoya U330 glass. ................................ ................................ .. 11 5 Figure 5.31. The photoluminescence of Y 2 O 3 :Eu particles deposited at 40 Torr and 60 Torr process chamber pressure. The excitation spectrum was filtered using Hoya U330 glass. ................................ ................................ .. 11 6 Figure 5.32. Luminescence intensity as a function of sample depo sition Absorbed power. ................................ ................................ ................................ .......... 11 7 Figure 5.33. Comparison between the Y 2 O 3 :Eu absorption spectrum and the excitation spectrum of experiment with U330 and U340 excitation light filters. ................................ ................................ ................................ .. 11 8 Figure 5.34. The spectrum of absorbed excitation light in experime nt with U330 and U340 excitation path filter. ................................ ................................ .. 11 9 Figure A.1. Electric diagram of nebulizer power control adapter. ............................... 131
xi Design and Characterization of Microwave Assisted Plasma Spray Deposition System: Application to Eu Doped Y 2 O 3 Nano Particle Coatings Marek Rados aw Merlak ABSTRACT This thesis presents a M icrowave P lasma Assisted S pray Deposition (MPAS D ) system design, characterization and application to produce nano sized particle coatings of metal oxides. A c ommercially available rectangular waveguide microwave power delivery system is utilized to initiate and s ustain the plasma discharge within the customized plasma applicator where micron sized droplets of a metal ion solution are heated to evaporate the solvent and thermally process the resulting nano sized particles. The investigation of optimum conditions for oxygen argon, and air plasma ignition in the MPAS D system was presented. Measured electron temperature of the plasma was between 6000K and 40000K for the plasma conditions used in the MPAS D process. Successful deposition of Y 2 O 3 :Eu nano particles using the MPAS D system was achieved MPAS D process allows control of the particles properties shown t h rough XRD and photoluminescence studies of the Y 2 O 3 :Eu coatings The MPASD process settings effect on particles activated doping concentration and, as a resu lt, its photoluminescence was shown
1 CHAPTER 1. Nanotechnology The topic of this manuscript is the design, characterization, and application of a system to produce nano particle coatings This chapter starts with a brief intr oduction to nanotechnology. Nanotechnology has been hailed to be the technology of the 21 st century. The rapid growth of our knowledge in this field and its many existing and predicted applications are critical to science, technology, economy, and most imp ortantly to the well being of people. There are multitudes of effort s from scientific communities, commercial entities, governments and public to envision the predicted advancements nanotechnology promises to bring When trying to define what is nano scien ce often the dimension is the criteria of choice. The m ost often utilized criteria is that at least one dimension of an object is between 1nm and 100nm (nano.gov). However, a more insightful definition of nano science uses system properties as criteria an d it states that : Nano science is dealing with functional systems based on the use of sub units with specific size dependent properties or of individual or combined functionalized sub units. (Schmid 2006)
2 In essence, this definition does not specify the size limits but that the size is such that size change varies the properties of material or combination of such materials. This definition better explains what nano science is about: it is about controllable or variable properties of materials. 1.1 H istory of Nano T ech nology When nano science and technology start is difficult to say. There are some early examples of nanotechnology. Nearly 2000 years ago Greeks and Romans used dyeing methods ( Walter 2006 ) described later by Arabian authors of medieval times to dye their gray hair black. This method worked because dyeing process created nano crystals of PbS within the hair volume. The Lycurgus cup made by Romans in 4 th century A.D. is an example of ancient nanotechnology as its glass optical properties are due to the 60nm nano particles of gold silver alloy dispersed within it (Freestone 2007). A s imilar technique was in use since medieval times for creating stained glass (Minko 2008). Si lver and copper nano clusters found in the luster decorations of Italian pottery were made by the method described by Piccolpasso in 1557 ( Borgia 2001 Padovani 2004). Photography developed as early as 1790 (Lichfield 1973) is an example of nanotechnology since it relies on decomposition of silver halides into silver na no particles under exposure of light. These early applications show that nanotechnology is not a new topic. However, the understanding of the nanotechnology is a work of modern science. It is worthy to mention Michael Faraday in his 1857 lecture titled Ex perimental Relations of Gold to Light presented his work in gold particle suspension that we know
3 now to be nano p articles. This lecture is not only the beginning of colloidal chemistry but could be the earliest scientific work in nano science. In his wo rk Faraday investigated the light interaction with particles very minute in its dimensions, and he concluded that phenomena appeared to indicate that a mere variation in the size of its particles gave rise to a variety of resultant colours (Faraday 185 7). Later work of Gustav Mie quantified the light interaction with the small particles that Faraday observed (Mie 1908). The first instruments to probe into matter beyond what the optical microscope allowed was the e lectron m icroscope invented by Max Knott and Ern st Ruska in 1931 (Ruska 1987). 1965 1960 1951 1931 1908 1857 1790 1557 350 50 B.C. 1970 1990 1980 2000 2005 Word Nanotechnology Pople's Software Zeolite Moore's Law Ferrofluid PbS Hair Dye The Lycurgus Cup Decorations of Pottery Photography Faraday's Gold Suspension Mie Light Scattering Electron Microscope Invention of STEM Field Ion Microscope Feynman's Talk DNA as Nano Assembler Carbon Nanotubes Manipulation of Atoms Quantum Dots AFM Bucky Ball STM Drexler's Surface Enchanced Raman Morrison's Molecular Computer National Nanotech Initative First Nanotechnology college in USA First molecular-scale circuit over 350 consumer nanoproducts over 1000 consumer nanoproducts Figure 1.1. Timeline of nanotechnology selected events. This invention which was improved in later years allowed observation of atoms using the field ion electron microscope developed by Erwin Mueler in 1951 ( Melmed
4 1995 ). The invention of new microscopy was important for the field of nano science. T he ancients did not know about nano sized particles of PbS in their hair or that the nano particles of gold and silver within the glass that absorb the blue light r esult ed in the deep red color of the light coming through the Lycurgus Cup. They could not know because one cannot see the nano particles with the bare eye. Faraday t h rough his systematic study suspected particles too small to see were responsible for the red color of the gold suspension but the verification did not come until necessary instrumentation was developed nearly a century later. N ano science and technology as we know it today is often said to start with Richard Feynman s talk titled There is Plenty of Room at the Bottom (Fenman 1959). This is the beginning of the so called Feynman STM IBM NNI history of nanotechnology: introduction of the field by Feynman followed by the invention of the scanning tunneling microscope (STM) (Golovchenko 1986) manipulation of atoms by scientists at IBM (Eigler 1990) and the foundation of the national nanotechnology institute (NNI) (nano.gov). The timeline presented in F igure 1.1 outlines some of the more important discoveries and inventions relating to nanote ch nology However, history, goals, market forces, and even politics have complicate d the nanotech nology story For example, the beginning of modern nanotech nology is unclear. Feynmans talk is often introduced in nano science and engineering classrooms as the bir th of nano science and his talk is said to inspire the research in nanotech nology However, Feynmans role was more that of the Nostradamus of nanotech nology Many scientists such as the inventors of the STM and AFM stated years after F e ynmans ta lk that There is Plenty of Room at the Bottom lecture did not influence their work or they
5 were not aware of it at the time of their discoveries. This suggests that his talk was not as influential or important as it is often credit ed In addition, Feynm an did not do any work in the nano science field, and his talk at the time of its introduction sparked little debate and went mostly unnoticed (Toumey 2005, 2008). It is interesting to notice that it is not until 1974 that Norio Taniguchi uses the term n anotechnology Deciding the birth date of nanotechnology sparks controversy, but not as much as deciding what is its direction: the nano science road is littered with conflicting views and directions, and sometimes with the ideas of the impossible. Richard Smalley and Eric Drexler are both highly regarded scientists yet their conflicting ideas sparked the Drexler Smalley open letters exchange (Baum 2003). Drexler in his work strongly emphasized his nano assembler direction for nanotechnology (Drexler 1986, 2003). Drexler approaches nanotechnology in an engineering manner: the nano assembler perform s task s and can self replicate in turn a large amount of them can quickly build devices in a nano factory manner. Smalley opposes such direction pointing to pro blems in this concept. His fat fingers and sticky fingers arguments show that mechanical nano assembler would not be able to manipulate atoms or molecules in the precise or reliable manner required In addition, he counters the Drexler argument of util izing enzymes or ribosome as manipulators by noticing that such method requires water chemistry. This would greatly limit the application s and would not be able to create the devices that Drexler envisioned. The history and direction of the nanotechnology is unclear but its growth is real, as can be seen by the growth in research and applications (Hullmann 2007, Salerno 2008).
6 1.2 Nano P articles The f ocus of this manuscript is on nano particles and a technique for their synthesis. Nano particles are of g reat interest for science and industry for their new or enhanced properties. The changing properties of materials in nan o scale are due to the quantum and surface effects occurring in small dimensioned obj ects. The se two phenomena are discussed separately. The t hermal, m echanical, m agnetic, c hem ical, c onductivity, and o ptical properties of nano particles are of great interest for scientific as well as commercial use. Some examples of chemical and physical methods for producing nano crystalline materials inc lude sol gel synthesis (Bui 2009) laser vaporization (Geohegan, Abdelsayed 2007 ), gas phase condensation (Champion, Guo 2007) plasma processin g (Jerby 2009), mechano chemical synthesis (Yang 2003, Tsuzuki 2004 ) combustion synthesis (Sun 2008), and sputt ering techniques (Dreesen 2009, Chen 2008). The optimal synthesis process should produce particles with small diameter (less than 100nm), narrow size distribution, and crystalline structure Small diameters are preferred because above 100 nm, the propertie s tend to approach those of bulk materials. Narrow particle distributions are preferred for potential application purposes Crystalline phase particles often have the desired properties over the amorphous phase. 1. 2 .1. Surface Effect Often the ratio of s urface atoms to the bulk atoms, or dispersion, is the source of the changing properties of the material on the nano scale. The atoms at the surface do not have the ability to form the same bonds as the bulk atoms do. This creates dangling bonds
7 and changes in crystal structure of the surface. Therefore, the ratio of an objects surface area to its volume does affect the properties of the object For an object with spherical geometry the surface area ratio to its volume is R R R V A 3 3 4 4 3 2 (1.1) Sin ce the ratio is inversely proportional to the objects radius then the ratio is small and does not change significantly for large objects. However, for small objects the ratio increases significantly with decreasing radius. This is the main reason why the surface effects are largely unnoticeable for objects with macroscopic dimensions, while the surface effects are significant on the nano scale. However, for real objects with nano scale dimensions the ratio of surface atoms to the total number of atoms is a more appropriate estimate of the surface effect. For a cubic crystal the ratio is R N N N N n n n F 1 6 6 8 2 1 6 8 12 6 3 1 3 2 2 1 3 1 3 2 (1.2) Because the ratio varies as 3 1 N then the surface effect on the nano scale can change greatly with the dimension s of the pa rticle while the change is unnoticed with objects of larger dimensions This is because for the smallest nano particles a ll atoms can be surface atom s, while larger particles, like 100nm lithium nano particle s ~5% of atoms are surface atoms (N~10 6 ).
8 Fig ure 1.2. Example of surface atoms and surface vacancies in cubic structure. Surface atoms behave different ly than bulk atoms, and not all surface atoms necessary behave alike. For example in the cubic crystal in F igure 1.2, the lack of some of the neare st neighbors for the surface atoms creates dangling bonds. The n umber of nearest neighbor atoms affects the number of dangling bonds for each atom and it depends on the location on the surface. Similarly, a vacancy at a surface creates dangling bonds from the atoms surrounding the vacancy. Dangling bonds could be terminated to lower the surface energy by creating bonds with other surface atoms, by absorbing atoms from the surrounding atmosphere that will terminate the dangling bonds, or by changing the chem istry of the surface in multi element compounds (Kawarada 1996, Shiraishi 1996, Camillone 2002) 1.2 .2. Quantum Effects On the other hand, the quantum effects are due to the small dimensions of the object and decreasing number of atoms within the particl e. As the size of the particle decreases the interaction of the particle with light changes. The magnetic particles can become magnetic single domains. The number of energy levels within the energy band of
9 the crystal decrease as the number of atoms decre ase. The density of states function changes depending on dimensionality and is summarized in F igure 1.3. The quantum effect can be explained using model of single electron being confined within infinite square well potential. Time independent Schrdinger equation is r E r r V r m ` 2 (1.3) For square well of width L, zero potential inside and infinite potential outside the solution is 2 2 2 2 2 mL n E n (1.4 ) The discrete energy levels depend on the square inverse of the dimension of the obje ct. Although, nano sized objects are multi electron this simple example can be extended to illustrate the size effect on small dimensioned objects Figure 1.3. The density of states function for bulk (3D), thin film (2D), nanowire (1D) and quantum dot (0D). If only one dimension is in the nano size range then the surface is often referred to as a thin film. Due to the thickness of such surface quantum, effects can be observed
10 that can be explained by modeling the surface as a two dimensional object. S imilarly, if two dimensions are small enough quantum effects occur that can be explained using a one dimensional model. E xample s of such 1D object s are nanowire and nanotube s If the size of the object in all three dimensions is small enough quantum effect s occur that can be explained by the zero dimensional model. Such 0D object s are often referred to as quantum dot s (QD). 1.2 .3. Nano P article Properties The study of nano structures is driven by the ability of nano structures to have different proper ties from the bulk materials. For nano particles, t he changes observed in thermal, mechanical, magnetic, chemical, or optical properties create great interest in scientific and commercial worlds The thermal properties of particles on the nano scale are diffe rent from the bulk such as t he m elting point of particles was observed to change (Goldstein 1992, Sun 2007). This allow s for lower temperature process ing of materials such as in sintering. The thermal conductivity of fluids is enhanced in the presence of suspended nano particles (Keblinski 2001, Das 2003). The thermal effects are often due to the high hydrostatic pressure due to particle surface curvature. Super elasticity has been observed in nano structured ceramics (Gao 2006) and thin films (Rumpf 200 5). Rumpf also has shown enhanced tensile strength in thin films. Strengthening of nano structured materials has been observed and modeled (Scattergood 2007).
11 A p articles size as well as its shape can greatly affect the magnetic properties of nano partic les Of most interest is the single domain and super paramagnetic effects observed in nanoparticles (Lu 2007). Magnetic nano particles are of great interest in biology and medicine for applications in therapy, drug delivery and diagnostic (Pankhurst 2003). The large surface are a to volume ratio of nano particles can greatly increase the chemical activity of a material. This is utilize d to increase the sensitivity of chemical sensors and the effectiveness of the catalysts (Astruc 2005). It was found that the se materials have higher organic compound adsorbity than that of activated carbon. The s mall size of nano particle s can create quantum confinement of an electron often called quantum dot (Meijerink). Quantum confinement can change the overlap of the wave function of the doping atoms wi th those of host atoms result ing in more efficient interaction between the doping and host atoms. The decreas e in size causes a blue shift in absorption that reflects the band gap increase of the host material, increased pho toluminescence, and decreased luminescent decay as shown in published studies (Bhargava 1994) 1. 3 Hazards of Nanotech nology Nano technology has allow ed us to discover new and exciting applications for materials, but there is a great risk that the se new ly discovered materials can cause harm to health and the environment. Just like Cur i e, todays scientists are looking for discoveries but sometimes not realizing the potential harm to themselves. In
12 addition, the discoveries can harm society t h rough their effects on health and the environment. For example, a carbon na notube with its very small diameter yet comparatively very large length and many interesting properties is proposed to solve many technological issues Many applications of carbon nanotube are studied from e lectronic devices (Javey 2003) to mechanical act uators (Baughman 1999). However, studies show that carbon nanotube s can be as dangerous as the now infamous asbestos (Poland 2008). Another example of problematic nanotechnology application is nano particle silver. Nano particle silver is one of the most common materials used in nanotechnology based consumer products. Because of its antimicrobial properties the nano particles of silver are used in bedding, washers, water purification, toothpaste, nursing nipples and bottles, toys, and kitchen utensils. How ever, the EPA decided to regulate washing machines in similar way as they regulate pesticides Their decision was made because in contrast to manufacturers claims nano particles of silver were released with wastewater (EPA 2008). This poses question abo ut other applications of nano particle silver: is it safe? In light of the fast growing nanotechnology field governmental regulatory bodies struggles to protect society from the potential harm. Th is is often difficult since the materials once known to be s afe, can become dangerous just by being nano sized Possible dangers can appear in the research, production, use, and disposal of products that utilize nanotechnology.
13 1.4. Thesis Outline The thesis topic is the design and characterization of a microwa ve plasma assisted spray deposition sys tem, its application to grow Y 2 O 3 :Eu nanoparticle coatings, and the characterization of these coatings. A brief description of the spray pyrolysis process is presented in C hapter 2 of this manuscript. The design of t he Microwave Plasma Assisted Spray Deposition (MPAS D ) system and its characterization are presented in C hapters 3 and 4 respectively. The depos ition and characterization of Y 2 O 3 :Eu is presented in C hapter 5.
14 CHAPTER 2. Spray Pyrolysis There are mu ltiple methods of producing nano scale particles. However, for many materials, these methods require a post annealing step to create the desired crystalline phase. The annealing step often leads to particle agglomeration and prevents depositions on temper ature sensitive surfaces such as polymers. In addition, the high temperature materials often are difficult to make using standard techniques because of the temperature limitations of the process. To eliminate agglomeration of the particles many techniques utilize a surfactant that terminates the surface dangling bonds. Although the surfactant reduces agglomeration, often the surfactant causes changes in the surface effect of the nano particle and obscure quantum effects within the particle. There is a need to overcome these limitations, as ability to produce size controlled, surfactant free, and crystalline nano particles of high temperature materials can lead to new discoveries and applications. The spray pyrolysis techniques can achieve such results. 2.1. Spray Pyrolysis Overview Pyrolysis is a process by which a solid (or liquid) undergoes degradation of its chemicals into smaller volitale molecules under heat, without interaction with
15 oxygen or any other oxidants, that is necessary for almost all solids (or liquids) to burn (Stauffer 2003) Pyrolysis is one form of the more general thermolysis process. The MPAS D system described in C hapters 3 and 4 is a pyrolysis system. However, the application of the MPASD system presented in C hapter 5 is strictly sp eaking not a pyrolysis process but technically a thermolysis process since the carrier gas used was oxygen. Nevertheless, the term pyrolysis will be used throughout the manuscript for all of the MPAS D processes due to its similarity to other spray pyrolys is techniques Other authors had allowed for this discrepancy, one example being the flame spray pyrolysis (Madler 2002). Figure 2.1. Spray pyrolysis process. The pyrolysis occurs (a) in flight of the atomized precursor droplets or (b) upon contact on a hot surface. Spray pyrolysis is a process of pyrolysis of a sprayed material. This can be achieved by thermally processing the spray after its atomization but before reaching a surface, or the pyrolysis can occur on a hot surface ( F igure 2.1). The produ ct of spray pyrolysis can be gas, liquid, or a solid. This manuscript describes a processs of spray
16 pyrolysis where the pyrolysis occurs before reaching surface and focuses on a final product that is a powder coating of solid particles. The topic most rele vant to this manuscript is the application of spray pyrolysis to produce particles. In such an application a precursor is atomized and the pyrolysis is achieved by heating the atomized droplets. Upon heating the solvent is removed and the reaction occurs where the solid product of pyrolysis forms a particles. F igure 2.1b presents the process where the precursor is a liquid solution and the particles are captured on the substrate surface. The process of pyrolysis can be as simple as solid precipitation wit hin droplet s or more complicated as in metal citric chelate pyrolysis. 2 .2. Spray Pyrolysis Atomization Techniques The precursor can be atomized using several techniques. The most common means of atomization are the liquid jet atomizer, air assisted atom izer, jet instabilities atomizer, surface instabilities atomizer, and the electrostatic spray atomizer techniques. The basic characteristic of each method are summarized in T able 2.1. Table 2.1. Comparison of different liquid atomization techniques (Luca s 1994). Liquid jet atomizer Air assisted atomizer Jet instabilities atomizer Surface instabilities Electrostatic spray Droplet size dispersion Broad Broad Very narrow Narrow Very narrow Independent gas and liquid flow rates Independent Gas dependent I ndependent Independent Liquid dependent Possible liquid flow rate Small to large Small to large Very small Small Very small Initial kinetic energy of the droplet Very high High High Very low Small Design and fabrication technology Intermediate Easy Diff icult Intermediate Difficult
17 The liquid jet atomizer is a hydraulic method where a liquid is pumped t h rough a narrow opening. The liquid turbulence in the jet, jet surface instabilities, and jet velocity profile are just few methods of atomization of the liquid jet (Reitz 1982, Ibrahim 1999). This method allows for large material flow, but suffers from broad droplet size distribution. The air assisted atomizer allows atomizing the liquid stream by pheumatic means. The liquid jet or droplets are mixed (int ernal mix method) before reach ing a nozzle, the turbulence of the mixture causes atomization of the stream. In the external mix method the liquid jet is ejected similarly, but not at a large pressure, as in the liquid jet atomizer the atomization is achie ved by the air stream impacting on the jet. The air assisted atomization methods are widely used in industrial application s especially in paint systems and fuel injectors for turbine engines (pnr nozzles.com, Watanawanyoo 2009 ). Although the air assisted atomizers are easy in application their broad droplet size distribution and the high bas e flow rates limit its applications. Figure 2.2. Diagram of examples of atomization techniques. The (a) liquid jet, (b) internal air assisted, and (c) external air assisted nebulizer are shown.
18 The jet instabilities atomizer is a combination of liquid jet atomizer with sonic or ultrasonic technology. The liquid jet is again atomized by the instabilities in the jet. However the instabilities are due to the sonic vib rations induced in the jet. This allow for lower jet velocities b ecause only the induced sonic vibrations cause atomization and the resulting droplet size is very narrow ( Avvaru 2005 ). The s urface instabilities atomizer also uses sonic or ultrasonic method s of atomization. The sonic vibration disturbs the liquid surface Given large enough amplitude of vibration the surface disturbance can result in liquid atomization. This method uses gas flow to transport the particles and the gas flow is independent from the material flow. The surface instabilities atomizer has a narrow droplet size distribution compared to those previously presented Figure 2.3. Diagram of examples of atomization techniques. The (a) jet instabilities and (b) surface instabilities atom izers are shown. Electrostatic atomization uses intense electric field between the charged atomizer and surface to be coated. The charge transfers to the fluid and repulsive forces between the atomizer and the fluid tear the droplets from the atomizer and send them toward the work surface. The energy source for electrostatic atomization is the electric charge that
19 the fluid receives. The particle size with electrostatic atomization is a function of three main factors electric ; field strength, liquid flow r ate, and fluid properties (including its electrical properties). This method results in a narrow droplet size distribution. Figure 2.4. Electrostatic atomization technique. The (a) diagram of the apparatus and (b) the principle of the drop formation. 2 3. Spray Pyrolysis Heat Sources Spray pyrolysis depends on heating to achieve its chemical decomposition. There are many techniques of heating that can be utilized in spray pyrolysis. Some of the common techniques include the hot surface (Mooney 1982), column oven (Messing 1993), laser induced (Dedigamuwa 2009), flame heating (Madler 2002, Sheen 2009), and plasma heating (Hoder 2005, Gell 2007, Jia 2009) as shown in F igure 2.5 and 2.6. The choice of the technique utilized strongly depends on the particl es desired physical properties, chemistry and desired deposition rate. The choice of the heating technique will dictate the resulting process temperature and the time that the droplet/particle is being heated.
20 In hot surface pyrolysis process es the atom ized precursor is sprayed onto a hot surface. The pyrolysis occurs upon contact with the hot surface ( F igure 2.5a). Because the liquid precursor is allowed to contact the surface, the coating is usually a film made up of platelets/splats. This process doe s not result in nano particles but allows for easy process temperature adjustment The oven spray pyrolysis process uses one or more heating columns. The atomized droplets are moved t h rough a hot area created by the oven where the pyrolysis occurs ( F igur e 2.5b). The number and length of the ovens depends on desired process temperature and required reaction time. In this method both temperature and pyrolysis time can be adjusted. Figure 2.5. Selected pyrolysis processes. The (a) hot surface pyrolysis, (b) oven pyrolysis, and (c) laser assisted pyrolysis processes are shown. In laser assisted spray pyrolysis a focused laser beam accomplished the heating The atomized precursor is directed t h rough the beam waist of focused laser beam ( F igure 2.5c). Of ten the carrier gas is chosen to absorb the laser light creating a hot zone. The reaction occurs within the hot zone and the reaction temperature can be adjusted by regulating the power of the laser beam or change in material flow rate
21 Flame spray pyroly sis uses the heat generated by burning a fuel to promote pyrolysis of the precursor droplets. A d iagram of one arrangement of this technique is presented in F igure 2.6a. The precursor carried by the fuel (gas) flow mixes with an oxi d izer at the nozzle tip. The ignited flame supplies heat to promote reaction. The processed particles are collected on the samples surface. The process temperature can be changed by using different fuels and oxidizers, changing fuel to oxidizer ratio, and their flow rates. In a ddition, contact time with precursor droplet can be adjusted in this technique. Figure 2.6. Plasma spray pyrolysis processes. The (a) flame, (b) ICP plasma, and (c) microwave plasma assisted spray pyrolysis systems are shown. The flame in the previous technique can be replaced with an induction coupled plasma (ICP) or microwave induced plasma. Plasma heating can reach very high temperatures without the need for a fuel. Plasma gasses can be reactive o r inert depending on the needs of the particular depo sition process In these techniques, the temperature of the hot zone can be adjus ted by changing the plasma operation parameters. The droplet/particle contact time with the plasma can be adjusted in the ICP arrangement
22 by changing the RF coil and reaction tube length. The waveguide dimensions limit the microwave plasma size Figure 2.7. Material side inje ction technique. The material can be injected into (a) the plasma or (b) the plasma tail. The material flow presented in F igures 2.5 and 2.6 is the s ame as fuel or plasma gas. However, the material and fuel for plasma gas flow can be separated. The side flow pattern shown in F igure 2.7a is often utilized (Gell 2007). This technique allow s injecting the material at different parts of the plasma: materi al can be injected to the plasma itself or into the plasma tail. By injecting the material into different parts of the plasma the temperature and precursor/particle contact time with the plasma can be adjusted. 2 .4. Chapter Summary Spray pyrolysis is often utilized for synthesis of fine and ultrafine particles. A l arge selection of atomization, heating, and material injection techniques allow s one to tailor spray pyrolysis to a specific application. Proper choice of the spray pyrolysis technique utiliz ed one must consider the desired temperature, time needed to complete reactions, and the chemistry of the deposited material.
23 C HAPTER 3. Microwave Assisted Spray Deposition System Design The Microwave Plasma Assisted Spray Deposition (MPAS D ) system w as designed at the Laboratory for Advanced Materials Science and Technology (LAMSAT). The main advantage of the MPASD was its capability for depositing nano sized particles that were surfactant free, needed no further annealing, and could be deposited in the required final crystal form on surfaces that are sensitive to high temperatures The MPASD system was composed of five sub systems: microwave power delivery, material delivery, deposition chamber, vacuum pump and control sub system s F igure 3.1 prese nts a diagram showing the key components and important electrical, gas, v acuum, and water connections The role of the microwave power delivery sub system was to produce a high electric field localized within a plasma tube. The material delivery sub syste m supplied metered amounts of process gas and the atomized precursor. The deposition chamber was used to create a controlled deposition environment and to enclose the sample mounting assembly. The vacuum sub system created a stable and precise process pres sure. The automated control helps to control and monitor the process variables and record experimental data and process paramete rs. The control software was customized to accommodate many experiment scenarios
24 Figure 3.1. The Microwave Plasma Assisted Spray Pyrolysis ( MPASD ) system co mponent diagram.
25 The MPASD system presented was successfully utilized to produce Eu doped Y 2 O 3 particles with a diameter of 100nm. The results of this application are presented in C hapter 5 of this manuscript. 3.1. Microwave Power Delivery Subsystem. Figu re 3.2. Microwave power delivery subsystem diagram. The MPASD system uses a microwave power delivery sub system to deliver microwave power to the plasma discharge zone. This subsystem was constructed using standard WR340 size waveguide components designe d to work at a 2.45 GHz frequency in TE 0 1 mode. Figure 3.3. Selection of the United States radio spectrum f requency allocations chart (year 2003 ). (US dept of commerce)
26 Component availability and governmental regulations limit the choice of system oper ating frequency to the 2.45 GHz. The radio frequency spectrum is strictly regulated to avoid interference between users of the busy radio frequency (RF) spectrum ( F igure 3.3). The RF spectrum allocated for industrial, scientific, and medical (ISM) applica tions use frequency bands ranging from 6.78 MHz to 245 GHz. The common ISM bands are listed in T able 3.1. Table 3.1. The commonly used ISM bands of radio frequency spectrum. Frequency range [Hz] Center frequency [Hz] Availability 6.765 6.795 MHz 6.7 80 MHz Subject to local acceptance 13.553 13.567 MHz 13.560 MHz 26.957 27.283 MHz 27.120 MHz 40.66 40.70 MHz 40.68 MHz 433.05 434.79 MHz 433.92 MHz 902 928 MHz 915 MHz Region 2 only (includes US) 2.400 2.500 GHz 2.450 GHz 5.725 5. 875 GHz 5.800 GHz 24 24.25 GHz 24.125 GHz 61 61.5 GHz 61.25 GHz Subject to local acceptance 122 123 GHz 122.5 GHz Subject to local acceptance 244 246 GHz 245 GHz Subject to local acceptance The most common microwave heating frequencies used commercially are the 915 MHz, 2.45 GHz, and the 5.8 GHz bands. Operation at these frequencies was desired since the availability off the shelf components speeds up the designing and prototyping of the system. The choice of the 2.45 GHz over the 915 M Hz and 5.8 GHz was dictated by the overall system size and cost The waveguide system operating at 915 MHz would require bulky WR975 (9.75 in by 4.875 in rectang ular cross section) or WR770 (7.7 in by 3.85 in cross section). On the other hand the syst em operating at 5.8 GHz would require small WR159 (1.590 in by 0.795 in rectang ular cross section) waveguide, with
27 higher components cost than for systems operating at 2.45 GHz. The 2.45 GHz system can use convenient sized WR340 waveguide (3.4 in by 1.7 in cross section ) and the component cost is lower than the other systems. The 2.45 GHz band is used in common household microwave oven s In fact, very early designs leading up to MPASD system utilized common household microwave oven components. 3.11. W aveguide Theory For a rectangular waveguide propagating a TE wave in the z direction the propagation is: t i z i y e a x B c a i E sin 0 (3.1) e space and critical wavelength, and 2 0 1 2 c c (3.2) For oper ation wavelengths longer than the critical wavelength, the propagation coefficient is a purely imaginary number and the resulting wave is attenuated as it propagat es ( evanescent wave ). On the other hand, for wavelengths shorter than the critical wavelengt h, the propagation coefficient is a real number and as a result, the wave freely propagates t h rough the waveguide. For the TE 01 mode of propagation the electric field within a rectangular waveguide that depends on the power level and the time averaged el ectric field strength is given by (Metaxas 1988) a a g TE b a P E 1 4 0 01 (3.3)
28 Due to the propagation path of electromagnetic wave within a rectangular waveguide ( F g is different from the free space 0 T he waveguide wavelength depends on free space wavelength and waveguide critical wavelength (Metaxas 1988) 2 0 0 1 c g (3.4) Figure 3.4. Wave propagating in a rectangular waveguide (TE 01 mode). The relation between the free space propagatio n and waveguide propagation is shown. c for a rectangular waveguide depends on its dimensions. For the TE m,n mode the critical wavelength is 2 2 2 b n a m c ( b a and n m TE mode) (3.5) The above equations allow calculati on of the electric field strength of a free running wave in a rectangular waveguide. 3.1.2. Microwave Power Delivery Subsystem Components The microwave power delivery sub system utilized a waveguide terminated magnetron head with power supply, waveguide iso lator, waveguide directional coupler,
29 waveguide three stub tuner, waveguide plasma applicator, and waveguide adjustable short to create a standing EM wave of controllable position and strength. The m agnetron head utilized a waveguide with one end closed an d an additional opening in the broad wall. This additional opening allow ed the magnetron antenna to be placed inside the waveguide cavity. The location of the antenna is usually 4 g from the back wall of the waveguide. This location allo ws for the best coupling of the energy between the antenna and the waveguide. Table 3.2. List of major components of microwave power delivery subsystem of MPASD system PN: Manufacturer Description Comments GA4005 GAE Magnetron Head 1.8 kW GA1107 GA E Waveguide Isolator GA3106 GAE Waveguide Directional Coupler GA1002 GAE Waveguide Tuner GA6103 GAE Waveguide Plasma Applicator GA1219 GAE Waveguide Adjustable Short SM745G.1001 Alter Magnetron Power Supply 2 kW The magnetron head us ed in MPASD was model GA4005 from Gerling Applied Engineering utilizing the water cooled magnetron (Panasonic model 2M137) The microwave head was capable of generating microwave radiation a frequency of 2.45 GHz at up to 1.8 kW of power. The microwave out put used the WR340 waveguide with UG554 flange. The model GA4006 magnetron head required a minimum of 1 GPM of coolant flow.
30 Figure 3.5. Magnetron head (a) location within MPASD (b) picture of GA4002A magnetron head and SM745G power supply, and (c) ope ration diagram of magnetron launcher. The magnetron head was powered using the high voltage power supply model SM745G made by Alter. This power supply was capable of generating a voltage of 4.6 kV and a current of up to 750 mA. The controls built into th e magnetron head w ere used during deposition to turn on and off the microwave generation and to adjust the microwave power level. The built in safety interlock system prevented magnetron operation in case of coolant flow failure, magnetron over temperature and mechanical waveguide mounting failure. The main role of the waveguide isolator in the MPASD was to protect the magnetron from damage due to the reverse microwave signal. The waveguide isolator is a device similar in design to the waveguide circulator in that both utilize the Faraday rotation of the EM wave within the ferromagnetic medium. In contrast from the circulator the isolator used in the MPASD had port #3 terminated with an absorbing media. In this configuration the forward wave entering port # 1 (from magnetron head) was directed to port #2 (to plasma applicator). As the reflected wave entered the isolator at port #2 (from plasma applicator) it was redirected to port #3 where it was absorbed.
31 Figure 3.6. Waveguide isolator (a) location within MPASD (b) picture of GA1107 waveguide isolator and (c) its operation diagram. The waveguide isolator used in MPASD was the model GA1107 made by Gerling Applied Engineering. It is capable of operating at a frequency of 2.45 GHz and 6 kW of microwave powe r. The waveguide isolator utilized the WR340 waveguide with UG554 flanges. This waveguide isolator required a minimum of 0.9 GPM of coolant flow during operation. The waveguide coupler is a waveguide device where a small portion of the microwave power was redirected from the waveguide into an external output. This task could be achieved by the use of an aperture array coupler, resistive loop coupler, and multi probe reflectometer. Because of the small size and low cost of application, the loop coupler was i mplemented. Other authors (Kulinski 1998, Mazur 2004) have treated the theory of operation of the waveguide loop coupler The output signal of the loop coupler is rectified and the rectified voltage relate d to the power flow within the waveguide. The dire ctivity of the loop coupler enabled separate measurement of the forward and revers ed power.
32 Figure 3.7. Waveguide directional coupler (a) location within MPASD and (b) picture of GA3106 dual directional waveguide coupler. The waveguide coupler used in the MPASD was the model GA3106 made by Gerling Applied Engineering capable to operate at a frequency of 2.45 GHz. The directional coupler was matched with two detectors and the set was calibrated for operation up to 2 kW of microwave power. The waveguide coupler utilized the WR340 waveguide with UG554 flanges. The directivity of the coupler was measured by the manufacturer to be 56 dB. Figure 3.8. Waveguide three stub tuner (a) location within MPASD (b) operation diagram, and (c) picture of GA1002 prec ision 3 stub tuner.
33 The waveguide system can be modeled using circuit theory. In this approach, each waveguide component is viewed as an equivalent R L G C circuit (Figure 3.9) which the values depend among other parameters, on geometry and dimensions o f the waveguide. Figure 3.9. Example circuit model of waveguide components. The role of three stub waveguide tuner was to match the impendence of the plasma discharge to the impendence of the microwave source. The three metal stubs protruding through t he broad wall of the waveguide had the protruding depth s controlled. The metal stubs changed the impedance of the waveguide section and helped to reduce the reverse power within the waveguide by matching the load to the source impedance Other authors (Gri ffin 1976, Muehe 1968) explain the theory and show experimental results for waveguide tuners The waveguide tuner used in MPASD was the model GA1002 made by Gerling Applied Engineering capable to operate at frequency of 2.45 GHz and up to 6 kW of microwave power. The waveguide tuner utilized the WR340 waveguide with UG554 flanges. The p lasma applicator is a section of waveguide with openings in the wide side of the waveguide that allow placing the plasma discharge tube inside the waveguide.
34 Additional side view holes allow for optical diagnostic of the plasma. The size of the openings and their length are chosen to reduce the microwave leakage by taking advantage of the attenuation of the wave while operating below the cut off frequency for the opening. F igure 3.10. Plasma applicator (a) location within MPASD (b) picture of GA6103 downstream plasma applicator, and (c) its detailed cross section. The waveguide plasma applicator used in the MPASD was the model GA6103 made by Gerling Applied Engineering ca pable of operating at radio frequency of 2.45 GHz and up to 6 kW of microwave power. This waveguide plasma applicator utilized the WR340 waveguide with UG554 flanges. The plasma applicator was modified to improve the MPASD process. The original plasma appl icator design ed for generation of downstream plasma for semiconductor applications was found to be inadequate for use in MPASD process. The discharge chamber of the original applicator had a shape th at allowed for better localization of the discharge for d ownstream plasma applications. However, in this design the discharge was difficult to maintain at larger process pressures of MPASD and the plasma had routinely melted the narrow discharge tube. The redesigned applicator had the plasma tube assembly chan ged to increase the electric field
35 strength and to use a lar g er diameter discharge tube (Figure 3.11a) This resulted in ease of plasma ignition and plasma and maintenance during deposition a nd the tube melting incidents w ere reduced. Figure 3.11. Plas ma applicator redesign stages cross section diagram. Features marked red were new or redesigned from original. Seals are marked with solid black and coolant with light blue. Plasma zone vacuum seals needed to be re designed as the original seal design wer e inadequate The original design utilized high temperature kalvare z o ring seals, and the seals w ere indirectly cooled by water cooling of the metal assembly. The exotic and expensive kalvarez orings are capable of operating at higher temperatures than th e common buna of viton elastomeric o rings H owever, the original design did not efficien tly cool the seals and seal failure occurred within few minutes of operation. To reduce the seal failure rate an improved seal cooling scheme was introduced. The cool ant was allowed to be in contact with the seal and the plasma tube. This
36 approach reduced the seal failure rate and did allow the use of less expensive buna and viton seals. Although this redesign ( F igure 3.11b) increased the lifetime of the seal, it did n ot protect the seals sufficiently as damage due to overheating and UV radiation occurred The second redesign of the seal improved the reliability of the seal by allowin g for direct contact of the coolant with the seals and the plasma tube as well as shi elding them from the plasma radiation ( F igure 3.11d and e). Figure 3.12. Different flow patterns possible in plasma zone. Material flow is marked gray and gas flow is marked black. One reason for plasma tube and seal failure was overheating. This prob lem was reduced by changing the gas flow pattern within the plasma tube. The original flow of the plasma applicator was inappropriate for the MPASD process and required redesign. The wide straight t h rough flow pattern of the GA6103 created significant plas ma tube coating, material flow within the tube was unpredictable, and the pla sma tube habitually overheated due to coating of the tube These shortcomings of the plasma applicator were addressed in the plasma applicator redesigns. Several methods that in clude d the m aterial flow nozzle, sheath gas flow, reverse vortex flow techniques were experimented with to determine the optimal design ( F igure 3.12). Although the material flow nozzle ( F igure 3.11 c and d) improved the results the tube coating was still a significant issue and the
37 flow pattern, although better, was unsatisfactory. The plasma tube overheating due to material buildup persisted. The forward vortex ( F igure 3.12b) and reverse vortex flow pattern ( F igure 3.12c) both resulted in satisfactory f low pattern control: overheating of the plasma tube was reduced significantly and the tube coating was reduced. However, high flow of the gas was needed for the vortex to develop. This resulted in short contact time between the plasma and the material to b e processed. The final design utilized the material flow nozzle with gas forward sheath flow ( F igure 3.12d). This design allowed for lower gas flow rates and still maintained good plasma control and predictable material flow pattern. Figure 3.13. Plasma applicator final design. The m aterial flow pattern was improved by introduction of the material nozzle. The f irst and simplest material nozzle was made of a glass tube ( F igure 3.11c) was found to be inappropriate for the MPASD The high temperatures of t he plasma melted the nozzle tip. The improved flat top metal nozzle ( F igure 3.11e) helped to remove the heat from the nozzle tip, but also created instability within the plasma. Because this nozzle was made of metal it also most likely increased the electr ic field within the plasma zone,
38 but its flat top design created a situation where the plasma would dance around the center of mater ial flow. These instabilities w ere eliminated using a dome shaped plasma nozzle. The dome design ( F igure 3.11f and 3.13) l ocalized the plasma discharge to the center of the plasma tube by increasing the electric field in the center of the nozzle (the top of the dome) as opposed to the edges.The final design of the plasma applicator detail is presented in F igure 3.13. The expe riments described in C hapter 4 and 5 utilized this design. An a djustable waveguide short was required to localize the standing wave in order to create an anti node within the plasma cavity. T he electric field at the movable wall was zero this locates the node of standing wave was wall. Therefore, the anti nodes were located at 2 1 4 1 n g away from the shorting wall. As th e waveguide short was moved, one of the anti nodes could be located within the plasma tube. Figure 3.14. Adjustable waveg uide short (a) location within MPASD (b) operation diagram, and (c) picture of GA1219A basic sliding short circuit. The adjustable waveguide short used in MPASD was the model GA1219 made by Gerling Applied Engineering capable to operate at frequency of 2 .45GHz and up to 6kW
39 of microwave power. This waveguide short utilized the WR340 waveguide with UG554 flanges. 3. 2 Material D elivery Sub system The role of the material delivery sub system of the MPASD system was to supply the atomized precursor and process g ases to the plasma applicator. The main parts of the sub system were the precursor atomizer and gas metering controllers Table 3.3. Major components list of the material delivery subsystem of the MPASD system. PN: Manufacturer Description Comments UF C 1000 MKS Mass Flow Controller 50sccm N 2 UFC 81000 MKS Mass Flow Controller 1slm NH 3 S tec Mass Flow Controller 241T Sonaer Ultrasonic Nebulizer 3 2 .1 Precursor Atomizer The size and size distribution of the precursor droplet used in MPASD directly affected the size and size distribution of the particles created. Proper choice of atomization technique of precursor solution was critical. The choice was made after considering the different atomizer techniques listed in T able 2.1. A v ery narrow droplet size distribution of jet instability or electrostatic spray is desired for MPASD technique. However, because of the ease of application the ultrasonic nebulizer was chosen. The ultrasonic nebulizer that was used is an example of the surface inst ability atomizer. The narrow distribution of the ultrasonic nebulizer, low
40 fluid flow rate, and low initial kinetic energy of the droplet is appropriate for application in MPASD The dynamic instability of surface submitted to oscillations at high frequen cy was used to obtain atomization. One method to create surface vibrations utilizes a vibrating piezo crystal driven by alternating current ( F igure 3.5). The vibrating crystal plate created disturbances on the surface of the liquid that resulted in liquid atomization. Figure 3.15. The diagram of ultrasonic atomizer utilized in MPASD system. Given sufficient energy of the oscillations, the surface instabilities can result in formation of droplets. The droplet size distribution is narrow and the size depe nds on the oscillation frequency. According to Kelvins expression (Lacas 1994) the disturbance wavelength relates to surface tension liquid density and oscillation frequency f by 2 3 2 2 f (3.6) Mizutani (1972) has shown that the droplet diameter was directly proportional to disturbance wavelength. Then
41 2 3 3 3 3 2 2 f k k d (3.7) Mizutani showed that 53 0 3 k for Sauters mean diameter then 3 2 3 2 55 1 8 53 0 f f d (3.8) For the nebulizer used in MPASD (Soaner model 241) the manufacturer indicated that the volume mean diameter of atomized droplets was ( sonozap.com ) 3 2 73 0 f d (3.9) The difference between equations 3.8 and 3.9 was due to di fferent statistic s used to calculate the average diameter. Now suppose a solution droplet created by the nebulizer have its solvent removed and undergoes pyrolysis. The remaining solute will create a particle. The final diameter of such a particle relates to the diameter of the initial droplet by 3 3 solution particle d k d (3.10) H ere the constant k directly depends on molar concentration of the precursor solution sol c the molar mass of particle particle M density of particle particle and molar ratio The factor is the molar ratio of product to reagent in pyrolysis reaction. Then 3 3 solution particle particle solution particle d M c d (3.11) and combining equations 3.9 and 3.11 results in 3 2 73 0 particle particle solution particle M c f d (3.12)
42 for a particle mean volume diameter. The model 241T nebulizer utilized in MPASD system was operated at a frequency of 2.4 MHz. The precursor used during the deposition process used water as a solvent. Assum ing that the small concentrations of the material in the solvent did not change the properties of the precursor from the solvent, surface tension and density, then for water ( m N 0728 0 and 3 1000 m kg ) the atomized precursor mean volume diameter is 1.7 d solution Then the mean volume diameter of particle is 3 3 0120 0 solution particle particle particle c M d (3.13) The cube root dependence of particle diameter on concentration requires t hat the solvent contamination needs to be considered when res ulting particles are of nano size The contamination of precursor solvent is important issue for the small contamination of the precursor become much larger contamination of the resulting nano particle. 0.001% 0.010% 0.100% 1.000% 10.000% 0 50 100 150 200 250 Size [nm] Contamination Figure 3.16. The effect of the precursor contaminate level on contaminate concentration within nanoparticle. The curves represent the contamination due to 0.01ppm, 0.4ppm, 1ppm, and 2ppm contaminated solvent if used to produce 100nm Y 2 O 3 particles. If depositions of nano scale particles are desired then solvent quality plays a significant role as it is shown in F igure 3.16. Some of contaminates of the solvent might
43 incorporate into the particles resulting in unexpected or even undesired results. For example the reagent grade dis tilled water (Alpha Aesar PN:36645) can contain 0.01ppm of silicate, 0.01ppm of heavy metals, 0.04ppm chlorine, 0.4ppm NO, 1ppm PO, and 1ppm SO. The ultrapure spectrometric quality water (Alpha Aesar PN:19391) contains as much as 2ppm evaporation residue t hat includes 0.01ppm of heavy metals. This might result in over 1% contamination of the resulting 100 nm nano particles of Y 2 O 3 used in deposition s presented in S ection 5.2. In order to reduce the solvent contamination effect on the resulting product, asid e from using extremely pure solvents, the atomized droplet size would need to be decreased. This would allow using higher precursor concentrations and thus reducing the contaminant to solute ratio within each droplet. The practical method of reducing the d roplet size requires nebulizer redesign. Changing the nebulizer operation frequency will allow reduc tion of the atomized droplet size. Atomized droplet size depends on the frequency as 3 2 ~ f d In order to reduce the particle volume by a fa ctor of 10 the frequency of the nebulizer operation needs to be increased by a factor of 3.16. This would decrease the contamination of the resulting coating by a factor of 10. Although high purity solvents are recommended for the MPASD process, the practi cal method of the contamination levels control in the nano particle appears to be reduction of atomized droplet size by higher frequency nebulizer design. 3.2.2. Gas Mass Flow Controller The precise control of the process gas mass flow is achieved by th e use of mass flow controllers (MFC). The MFCs used in MPASD system utilize the commercially
44 available devices (UNIT models UFC 1000 and UFC 8100). The MFCs used have built in thermal flow sensor and electromagnetic (UNIT) or piezo electric (STEC) flow con trolling valves. The t hermal flow sensor detects change in the thermal profile along the gas flow tube. Under a zero flow condition the temperature profile will be symmetric on both sides of the heater ( Figure 3.17a). During gas flow through the tube dis turbance of the temperature profile occurs. The temperature distribution is sensed by the thermistors mounted on each side of heating element. Thermisotors are connected as part of a Wheatstone bridge allow ing m easurement of the small voltage corresponding to a temperature difference. Therefore the voltage difference relates to gas flow through the tube. Figure 3.17. Basic operation of gas mass flow meter (a) and the diagram of typical gas mass flow controller. The flow signal from the thermal flow se nsor is compared to the flow control signal supplied from outside of the control ler. Internal circuitry readjusts the gas valve position according to the difference signal between the gas flow measured and set value.
45 3.3. Deposition Chamber As the proces sed material exited the plasma zone, it entered the deposition chamber where the material was deposited onto a substrate. The deposition chamber consisted of an 8 inch diameter aluminum base plate, cylindrical glass wall the choice of various aluminum top plates, sample holder, deflection shield, and optional plasma tail directing tube. Figure 3.18. Picture and diagram of deposition chamber. To vacuum pump Sample holder Sample Electrical connections Plasma appl icator Flow directing tube Deposition shield
46 The base plate acted as a mechanical base for the va cuum chamber as well as the connection for gas, and vacuum lines The base plate was mounted onto the plasma applicators top water jacket that also acted as a heat sink for the chamber. The center opening of the base plate allowed for the gas flow directing tube to be installed. This improved the coating growth rate. The cylindrical sidewall of the deposition chamber was made of Pyrex glass, and allowed for easy alignment of the sample before operation and visual diagnostic s during operation of MPASD system. The vacuum seal was accomplished using the flat rubber gasket that makes the seal between the bottom plate and the chamber wall and the o ring seal that makes the seal between the cylindrical wall and top plate. The top plate is equipped with two mechanical feed t h rough that allow ed manipulat i on of the sample holder by adjusting height and position of the depositing sample and operati on of the deflection shield. The electric motor mounted on the top plate was utilized to rotate the multiple position sample holder which allowed deposition on se veral samples during one deposition rum In addition the electric feed t h rough in the top plate was used to connect probes located inside the chamber. 3.4. Vacuum Sub System As the material flows into the system the role of the vacuum sub system was to remove the gasses from the system at a rate that will result in a controlled pressure within the deposition chamber. The vacuum subsystem utilized the dual stage rotary vacuum pump with a pumping speed of 70 cfm and capable of creating a vacuum level of 10 4 Torr.
47 The pressure control was achieved by utilizing a throttle valve located between the vacuum pump and the deposition chamber. The t hrottle valve was actuated by an electric motor controlled by an electronic controller. Figure 3.19. Operational d iagram of vacuum sub system. The different pressure conditions are shown. The p rocess pressure was measured by a silicon strain gauge pressure sensor. The throttle controller utilized a circuit that compares the set voltage representing the desired set pr essure and the pressure sensor voltage representing the actual pressure. If the two were not at the desired balance, the internal circuit energized the throttle valve to open when the pressure is too high, or to close if the pressure was too low as shown in F igure 3.19. The throttle control utilized a simple PID circuit that prevents control oscillations and improves the response of the syst em. The pressure set voltage was generated by the control and monitoring subsystem described later. Table 3.4. List of major components of vacuum subsystem of MPASD system. Model Manufacturer Description Comments 2008A Alcatel Vacuum Pump 253B 1 2CF 1 MKS Throttle Valve 152H P0 MKS Throttle Valve Control 54 760 eng Sensor Systems Pressure Gauge 0 25 P SI a in, 1 5VDC out
48 3.5. Control and Monitoring Sub System The control and monitoring sub system simplifies the MPASD deposition process by automatically monitoring and controlling some of the variables of the process. The deposition variables include micro wave power, process pressure, precursor atomization rate, and process gas flow. In addition, this sub system allows monitoring of the diagnostic temperature probe. This sub system utilizes a computer based data acquisition card (DAQ) together with pressure temperature, microwave power, and gas flow sensors and controllers. Table 3.5. List of major components of control and monitoring subsystem of MPASD system. PN Manufacturer Description Comments PCI 6221 National Instruments DAQ card SCB 68 Nationa l Instruments DAQ terminal b lock Computer Generic National Instruments LabView 8.5 Software Power One Power supply m odule 12V DC, Two needed PWM c ircuit See Appendix A Diff a mp c ircuit See Appendix A The heart of the control sub syste m is the DAQ card controlled by the LabView software. The PCI 6221 DAQ card used is capable of measuring up to 16 voltage signals, generating 2 voltage signals and 2 PWM digital pulse trains. The DAQ card measured the pressure sensor, temperature probe, microwave crystal rectifier voltage, and gas mass flow output voltage
49 Figure 3.20 Front pa nel and program diagram (part ) for MPASD control and monitoring program written in LabView. The software interface recalculates the voltages into meaningful values representing the respect ive measurements that were accessed through the virtual control panel. The voltage output was utilized in the plasma ignition study ( S ection 4.3) to automatically vary the microwave power during the study. The PWM pulse train generated by the DAQ card was used to control the power of the nebulizer and in turn the rate of precursor atomization. A differential amplifier circuit was used to eliminate common mode voltage of the temperature probes The isolating circuit was used be tween the DAQ and the nebulizer allowing computer control of the nebulizer material atomization rate The electrical diagram of the control and monitoring sub system and schematics of the differential amplifier and the PWM cir cuit are presented in A ppendix A 3.6. Chapter Summary The MPASD system utilizes microwave power to produce a thermal plasma. MPASD system utilizes rectangular waveguides to create a standing wave within the plasma applicator as described in S ection 3.1.2. The process precursor was a tomized using the nebulizer described in S ection 3.2.1 and the material was heated in the plasma
50 applicator descr ibed in S ection 3.1.2. The process gas was metered using the mass flow controller described in S ection 3.2 .2. The processed particles were coll ected onto a substrate located in deposition chamber described in S ection 3.3. The vacuum system p resented in S ection 3.4 removed the process gas from vacuum chamber and controled the process pressure. The a utomatization sub system utilized computer contro ls to ease the deposition process and monitor the multitude of process parameters ( S ection 3.5).
51 C HAPTER 4. Microwave Plasma Assisted Spray Deposition System Characterization. The characterization of the MPASD system is presented in this chapter. Th e MPASD process relies on the plasma thermal heating of the material. The initial estimate of the plasma temperature, plasma emission spectroscopy study, as well as the plasma ignition study is presented in this chapter. In addition, the results of the pre cursor atomization rate study are presented. 4.1. Plasma Temperature Model During the steady state of the MPASD operation, the microwave energy absorbed by the plasma gases could be estimated by adding the energy lost due to radiation, conduction, and ma ss transfer. Assuming thermal equilibrium within the plasma zone of MPASD process then the plasma temperature can be estimated using simple calorimetric and radiative relations. Since most of the materials grown by the MPASD system are oxides, the plasma temperature estimate was calculated for oxygen as carrier gas and water as a precursor solvent. In the first approximation, it was assumed that the radiation and the conduction are not significant. Therefore, during steady state operation of the MPASD the power
52 absorbed by the plasma is equal to the thermal power flowing out due to the hot gas and material exiting the plasma zone. The power balance can be written as: precursor transfer mass gas transfer mass transfer mass absorbed P P P P (4.1) The power needed to increase the temperature for a gas flow rate mass f is T c f P gas mass gas transfer mass (4.2) The heat capacity of gas c can be treated as constant for Argon gas ( R c Ar p 2 3 ) but it is a function of temperature for molecular oxygen. The heat capacity for o xygen has four main contributions: translational, rotational, vibrational and electronic modes. In the most general case, the heat capacity for molecular oxygen can be then written as sum of heat capacities due to translational, rotational, vibrational and electr onic modes R e T e T e T e T c i T i T T T o v i 2 2 3 2 2 2 1 2 1 2 1 2 2 2 3 3 2 1 2 (4.3) 3 4 5 6 0 2000 4000 6000 8000 10000 T [K] Cp/R [-] Figure 4.1. The experimental data f or heat capacity of oxygen gas as a function of temperature and the fitted curve. The 1 2 3 and i are the characteristic translational, rotational, vibrational, and electronic temperatures respectively. For the temperature range of interest (T>300K) the translational and ro tational modes are fully available as it can be seen from the experimental data presented in Figure 4.1.
53 Using the published data and equation 4.3 the heat capacity function was determined as R e T e T e T c T T T o p 10354 2 2483 2 1917 2 10354 878 0 2483 2 1 1917 2 1 2 7 2 (4.4) for temperature T in Kelvin. Now, taking equation 4.2 with 4.4 the relation between process temperature and absorbed microwave power for oxygen plasma can be summarized as T T T O absorbed e T e T e T T f R P 10354 2 2483 2 1917 2 10354 878 0 2483 2 1 1917 2 1 2 7 2 (4.5) Using equation 4.5 the oxygen plasma temperature dependence on absorbed power for gas flow rates of 5000 sccm and 700 sccm was calculated and the results are presented in Figure 4.2. 0 2000 4000 6000 0 200 400 600 800 1000 1200 Power [W] Temperature [K] Figure 4.2. The plasma temperature as a function of absorbed microwave power. The oxygen gas temperature for 700sccm (solid line ) and 5 000sccm (dash line) gas flow w ere calculated. No radiation and conduction was included.
54 Since radiative losses increase as T 4 for temperatures higher than 1000K radiative losses can be significant. Therefore, the temperature prediction for argon or oxygen plasmas for the power range given should not ignore the radiative component. After including the heat absorbed by water vapor, through similar analysis to equation 4.3, the plasma temperature was computed. Adding the precursor flow decreases the plasma temperature for the same absorbed power level as compared to the pure gas case. This was summarized in Figure 4.3. It can be seen that the plasma temperature decreases for larger precursor flow rates. Again, the high plasma temperatures suggest that the radiative energy losses are significant and should not be ignored. Figure 4.3. The plasma temperature as a function of absorbed microwave power for (a) 5000 ccm and (b) 700 sccm oxygen flow rat e. The thick solid, dash, and red thin lines presents data for 1 g/min, 0.1 g/min, and 0.01 g/min precursor (H 2 O) flow respectively. To account for plasma radiation on the plasma temperature the gray body model was used for the radiation of oxygen gas and water vapor mixture. Plancks law describes the radiative power of black body. 0 2000 4000 6000 0 500 1000 Power [W] Temperature [K] a) 0 2000 4000 6000 0 500 1000 Power [W] b)
55 1 1 5 T k c h B e T u (4.6) Integrating the above over all frequencies, the power output of black body radiation is described by Stefan Boltzmann law 4 T A P (4.7) The gray body model assumes that the object radiates similarly to the black body except that the total radiated power of the gray body is less than that of black body. The proportionality factor (emissivity) relates the gra y body radiation power to black body radiation power. 4 T A P P black gray (4.8) b) 0 2000 4000 6000 0 500 1000 Power [W] c) 0 2000 4000 6000 0 500 1000 Power [W] Temperature [K] Figure 4.4. Emissivity data (a) and the calculated plasma temperature as a function of absorbed microwave power for (b) 700 sccm and (c) 5000 sccm oxygen flow rate. The thick solid, dash, and red thin lines present data for 1 g/min, 0.1 g/min, and 0.01 g/min precursor (H 2 O) flow respectively. Now, emissivity of oxygen and water vapor mixture strongly depend on its temperature. Using the emissivity data (Hottel 1967) the plasma temperature was
56 calculated for oxygen and water mixture accounting for the radiative transfer. Figure 4.4 shows a summary of the results. The plasma temperature changes significantly after the emissivity was accounted for. The material trough output of nebulizer utilized in the MPASD system is about 0.05 g/min. In this regime, the plasma temperature depends little on material flow rate. The resulting temperature changes with gas flow as well as material flow. However, the plasma dimension affects the radiative transfer rate, thus affecting the plasma temperature. Figure 4.5 shows the effect of plasma size on its temperature. As the plasma size increases, the plasma effective surface area incr eases resulting in larger radiative transfer. As a result, the plasma temperature decreases. 0 2000 4000 6000 0 500 1000 Power [W] Temperature [K] Figure 4.5. Plasma temperature for selected plasma discharge size. The plasma calculation results for (a) 2 cm diameter sphere, (b) 3 cm diameter sphere, and (c) 3 x 3 x 5 cm ellipsoid are presented by dash, solid thick, and solid thin line respectfully. Calculated for precursor flow rate of 0.01g/hr of water. It is important to notice that the model presented does not predict the size of the plasma discharge. The calculations presented are very rudimentary and can only be used as an estimate of the plasma temperature. Correct estimation of radiative energy transfer is important, as its effect on temperature prediction is significant: s imple gray body modeling is not appropriate for reliable plasma temperature predictions. The spectro
57 radiative analysis of oxygen plasma emission shows a spectrum rich in features (Figure 4.X) that change with plasma conditions. The plasma size varies with process conditions, changing the effective gray body surface area. The plasma dimensions cannot be predicted from the simple model presented. In addition, real oxygen plasma is composed not only of molecular oxygen, but also of atomic oxygen, oxygen ions, and electrons. This complicates the calorimetric analysis shown earlier. Plasma modeling is treated by many authors, and the complexity of these models requires computational techniques to be used. Experimental measurement of the plasma temperatures prese nts less of a challenge. Experimental plasma characterization methods include the use of probes (Kissovski 2006, 2007), s pectroscopic methods (Tatarova 2007, Kitamura, Rabat 2004), and use of waves (Irish 1964, Stott 1969). Electrostatic probes, such as t he Langumuir probe allows the determination of the density and temperature of ions and electrons in the plasma. Wave methods include RF cavity perturbation methods and the microwave propagation techniques. These methods allow the determination of electron and ion electrostatic, cyclotron resonance frequency, RF absorption, diffusion, reflection, and phase lag among other parameters. The refraction index found by these methods has direct relationship with plasma density. The spectroscopic methods allow for p lasma species identification as well as electron, ion, atom, molecule vibrational, and molecule rotational temperature determination. Plasma emission spectroscopy is non invasive and experimentally straightforward An introduction to plasma temperature det ermination by emission spectroscopy is described in the next section.
58 4.2. Emission Spectroscopy for Plasma Temperature Determination Preliminary spectroscopic measurements of microwave plasma emissions are presented here and the possibility of plasma characte rization using emission spectroscopy was discussed. The spectroscopic methods allow for making a qualitative d etermination of the chemical composition by identifying plasma emission or absorption lines. In addition, spe ctroscopic methods allow for measurem ent of electron, ion or atom temperatures especially in hot gasses and plasmas The electron temperature was calculated using the emission spectroscopy for several plasma conditions. 4.2.1. Spectroscopic Temperature Measurement Theory Electronic temperat ure can be determined by measuring the intensity of characteristic emission lines. Assuming local thermal equilibrium ( LTE ) condition of an optically thin plasma the ratio of two emission lines for a same atom or ion is (Griem 1997) T k E E m m n m n m m n m n m n m n b m m e g A g A i i R 2 1 1 2 2 2 2 1 1 1 1 1 2 2 1 1 (4.9) For the elemental neutral and ionized species the statistical weight, the transitional probability, and frequency can be found from references (nist.gov). The temperature can be determined by direct calculation or by a graphical slope method. Th e Doppler Effect due to thermal energy of the atom or ion can be quantified by measuring the emission line broadening and the broadening relates to temperature as (Kettani 1973) M T k c B D D 2 2 ln (4.10)
59 This method needs to be applied carefully as o ther effects, such as plasma turbulence and pressure, can cause significant broadening of the emission lines. To reduce contribution of other effects this method is mainly used for high temperature plasmas. The molecule rotational temperature in a plasma has been found to be the same as the gas temperature under certain conditions (Touzeau 1991). The rotational mode emission spectra creates emission bands with closely spaced emission peaks. The rotational temperature of a molecule can be determined from t he relation rot B nm J e T k c h J F A S I ln (Kylian 2002) (4.11) where A nm is the spontaneous transition probability, S j is line strength, F(J) energy of J level and I e is the intensity of the spectral line. Using the graphical slope method the temperature T rot can be determined if several spectral lines intensities are known. 4.2.2. Spectroscopy Experimental Setup Figure 4.6. Experimental setup diagram for plasma emission spectroscopy.
60 The spectroscopy measurements were conducted utilizing the USB2000 spectrometer capable of detecting in the 200 900 nm range with optical resolution of 5 nm FWHM. The instrument sensitivity was sufficient to measure the spectra using an optical fiber and without the need of light collecting optics. The experimental setup is presented in Figure 6.4. 4.2.3. Spectroscopy Experimental Results An example of the collected plasma emission spectrum is presented in Figure 4.7. 200 300 400 500 600 700 800 900 Wavelength [nm] Intensity [arb] possible OHrot bands 200 300 400 500 600 700 800 900 Wavelength [nm] Intensity [arb] 750 770 790 810 830 850 Wavelength [nm] Intensity [arb] 777nm O triplet 844nm O triplet possible O2 rot bands Figure 4.7. Microwave plasma emission spectra. Plasma absorbed microwave power was 200 W, proc es s pressure was 40 Torr, nebulizer duty cycle was 3%, and oxygen gas flow was 700 +18 sccm.
61 During a short spectrometer detector integration time (low sensitivity), three emission bands were observed at 285 nm, 311 nm, and 350 nm. In addition, there were two peaks observed at 590 nm and 777 nm. During longer a integration time (higher sensitivity), the plasma emission spectrum shows a rich spectrum of atomic emission lines as well as several emission bands. The emission lines at 777 nm and 844 nm are of g reat interest as these lines are characteristic to neutral oxygen and the electron temperature can be determined from these emission lines. The emission band at around 311 nm is possibly due to a OH rotational mode and can be useful to determine the rotat ional temperature of the OH ion (Pellerin 1996). In addition, the emission band around 766 nm is possibly due to O 2 rotational mode, and could be used to calculate the O 2 rotational temperature. A closer look at the emission lines at 777 nm and 844 nm al lows for determination of the electron temperature. The intensity of the emission lines was measured from the collected spectra. The raw data collected by the spectrometer was rescaled to account for nonlinear intensity spectral response of the spectromete r. The reference used for recalibration was emission from a tungsten lamp (2900K). The plasma emission spectral data collected for plasma under selected pressures was processed and the intensity of 777 nm triplet to 844 nm triplet was calculated. Then the plasma electron temperature was calculated using equation 4.9 from the measured triplet ratio. The resulting calculations were compared in Figure 4.8. According to these calculations, the electron temperature was as low as 6000 K for 120 Torr plasma and as high as 40000 K for a 20 Torr plasma. The large uncertainty
62 of the triplets ratio is due to radiometric calibration uncertainty as well as the uncertainty of the measured intensity. 1000 10000 100000 0 50 100 150 Power [W] Temperature [K] a) 2 3 4 0 50 100 150 Power [W] Ratio [-] b) Figure 4.8. Calculated electron temperature as a function of process pressure from the measured emission 777 nm and 844 nm triplet ratio. The data was collected at 400W absorbed microwave power, 700+18 sccm O 2 flow, and 3% nebulizer duty cycle. Similar calculations of electron temperature conducted on plasma emissions data give absorbed microwave power effect on electron temperature. These results are presented in Figure 4.9. 8000 12000 16000 20000 100 500 900 Power [W] Temperature [K] a) 2 3 100 500 900 Power [W] Ratio [-] b) Figure 4.9. Calculated electron temperature as a function of process absorbed microwave power from the measured emission 777 nm and 844 nm triplet ratio. The data was collected at 40 torr process pressure, 700+18 sccm O 2 flow, and 3% nebulizer duty cycle.
63 The temperature change is smaller than the change observ ed in Figure 4.8. The temperature was determined to be between 1100 K for the 200 W generated plasma and 1500 K for 1000 W generated plasma. Again, the uncertainty of the observed temperature is due to radiometric calibration uncertainty and the uncertain ty of the measured intensity. The electron temperature measurement results might improve if additional spectral lines are measured. Looking at Table 4.1, the triplet emission at 926 nm is a good candidate for such calculations. Table 4.1. Wavelength, tra nsition probability, upper state statistical weight, and upper state energy for selected strong emission lines of neutral oxygen atom. (nist.gov) A nm [s 1 ] g m [ ] E m [eV] 777.194 3.69 10 7 7 10.740931 777.417 3.6810 7 5 10.740475 777.539 3.7010 7 3 10.740224 844.625 3.2210 7 1 10.988880 844.636 3.2210 7 5 10.988861 844.676 3.2210 7 3 10.988792 926.267 2.6010 7 5 12.078644 926.277 2.97 10 7 7 12.078629 926.601 4.4410 7 9 12.078618 It is worth notice that the electron temperatures measured at higher pressures (80, 100, and 120 Torr) are around 6000 K a number similar to the calculated values presented in Figure 4.5. The emission band s observed around 310 nm and 766 nm suggest that the rotational temperature of OH ion and O 2 molecule can be calculated. However, the resolution of the experimental setup used to collect data prevents measurement of
64 necessary data for calculations. In ord er to perform the calculation from equation 4.11, spectroscopic data with a resolution of 0.025 nm or better is recommended. The temperature broadening for 777 nm line changes as T and can be calculated using equation 4.10 This broa dening varies from 0.0025 nm FWHM at 300 K to less than 0.05 nm FWHM for temperature of 40000 K. This suggests that high resolution spectrometers are necessary for the determination of temperature. However, the pressure broadening and Stark broadening due to RF electric field can make it difficult to measure the plasma temperature. The experimental setup utilized for this initial study had optical resolution of only 9 nm. This prevents observation of the thermal broadening. In theory, this method allows for calculation of ion or atom temp erature, however, the required instrumentation and possible difficulty of removal of effects of peak broadening resulting from strong electric field within the plasma and broadening due to plasma pressure might render this m ethod impractical. 4.3. Plasma Ignition Study The ignition of gas plasma discharge occurs when the rate of gas ionization exceeds the rate of electron recombination. In high frequency discharge, the rate of gas ionization depends on pressure, the strength of e lectric field and its orientation as well as the discharge chamber dimensions and geometry and the properties of the gas.
65 1.E-05 1.E-04 1.E-03 1.E-02 1000 10000 100000 1000000 E/p [V/( m Torr)] Ionization coefficient [1/V] L=0.0318cm L=0.0318cm L=0.0635cm Figure 4.10. High frequency ionization coefficient of air as a function of p E / (Herlin 1948). In a parallel plate configuration with plate separation L and for a gas with ionization coefficient the electric breakdown condition is given by (Herlin 1948) 2 2 1 E L (4.12) The ion ization coefficient varies with strength of the electric field, pressure, and depends on experimental system dimensions as presented in Figure 4.10. The resulting break down electric field dependence is shown in Figure 4.11. It is clear that there exists a pressure at which the electrical breakdown occurs its lowest electric field. As presented in Figure 4.11, the plate separation L affects the minimum electric field needed to breakdown air and the pressure at which the minimum occ urs. Similar studies for other gases such as oxygen, air, and argon show a similar characteristic.
66 10000 100000 1000000 0.1 1 10 100 1000 Pressure [Torr] Electric Field [V/m] L=0.0635cm L=0.318cm L=0.157cm Figure 4.11. Breakdown electric field for air as a function of pressure for parallel plate configuration at three different di stances (Herlin 1948) The geometry of MPASD plasma zone creates non uniform electric field unlike parallel plate configuration used to derive equation 4.12, the n the electrical breakdown of gas cannot be readily pr edicted for there is no ionization coe fficient data available for MPASD plasma applicator geometry. However, it is reasonable to expect that the electric field strength needed for electric breakdown of gas in MPASD will change with process pressure. This change is expected to have similar feat ures to the data shown in Figure 4.11: there exists a pressure at which the electric field strength needed for electric breakdown of gas is at a minimum. 4.3.1. Plasma Ignition Experimental Setup The system condition at which the electric field strength needed f or plasma ignition is at its minimum is the optimum condition for plasma ignition. In MPASD this condition depends on the process pressure and the waveguide short position. In order to determine the optimum ignition configuration of MPASD system a gas elec tric breakdown (or plasma ignition) study was conducted. In this study the output power of
67 the microwave source was varied to determine the strength of electric field needed to cause electric breakdown of the process gas. The study measured the forward mic rowave power level needed for gas electric breakdown as a function of pressure and position of waveguide short. Figure 4.12. Diagram of the experimental setup for the electrical discharge of oxygen and argon gas in MPASD system. Figure 4.12 presents th e diagram of the experimental setup used for this study. The power level of the microwave magnetron was computer controlled. Table 4.2. List of major components of plasma ignition study experimental setup. PN: Manufacturer Description Comments MP ASD system See table 3.2 EP200 Verity Monochromator GA1219 GAE Waveguide Isolator Optional P400 5 UV VIS Ocean Optics Optical Fiber The plasma ignition was detected using optical monochromator with build in PMT detector (Variety EP200) tuned to a selected atomic line of the process gas emission.
68 The electrical signal from the monochromator was connected to the computer through the DAQ card (National Instruments PCI 6251) and the signal was processed with the help of the custom written ignition.v i a LabView virtual instrument program. The ignition.vi program also allows controlling and monitoring of the process pressure and microwave forward power. After manually adjusting the variable position short at the desired location the ignition.vi progr am sweeps through the desired pressure range. The waveguide short position is measured from an arbitrarily chosen 0 cm location, and this location is kept constant throughout the experiment. At each pressure step, the program increases the microwave forwar d power from 0 W until plasma is detected by the monochromator. The power level at which the plasma begins to occur together with the corresponding pressures is then stored into a file for further data analysis. However, after the plasma is extinguished a nd after the microwave power is reduced to 0 W there is some ionized gas remaining in plasma tube. This ionized gas reduces the electric field needed to ignite the plasma in next experiment cycle. This in turn gives lower values for electric field strength s. The goal of the ignition study was to determine the electric field strength needed for plasma ignition in a situation where the gas is not already ionized. To reduce the remaining ions effect on ignition during the experiment the process gas was flown t hrough the plasma zone at a rate of 18 sccm. This helps to remove ionized gas after each gas electrical breakdown and before the microwave power is cycled again.
69 4.3.2. Plasma Ignition Study Initial Results Figure 4.13 presents two data steps of the ignition stu dy for oxygen gas. As it is expected there is a pressure at witch the microwave power needed for ignition is at its lowest. From the graph it appears that the minimum occurs in the pressure range of 1.5 to 4 Torr. The electric field within a rect angular w aveguide can be calculated using the microwave power level and waveguide dimensions. The peak electric field strength in the rectangular waveguide with dimensions a and b for a wave with free space wavelength 0 operating at TE m,n mode is given by equation 3.3. In the case of WR340 (a=86.4 mm, b=43.2 mm) waveguide operating at 2.45 GHz in TE 1,0 mode and power level P the strength of electric field within the waveguide is P E RMS 535 (4.13) However, the strength of electric field in the discharge tube during the ignition experiment can not be estimated using equation 4.13. 0 500 1000 1500 0 2 4 6 8 10 Pressure [Torr] Power [W] +1.0cm +5.7cm Figure 4.13. Experimentally determined minimum microwave forward power level during electrical breakdown of oxygen gas in MP ASD system as a function of pressure at selected positions of waveguide short.
70 The dependence of the results in Figure 4.13 on the waveguide short position suggests that the electric field strength within the discharge tube varies depending on the wavegui de short position. This effect is due to interference of the wave reflected by the waveguide short with the forward wave (see Figure 4.14a). This interference of the two waves creates locations of high electric field and locations of low electric field. This effect is verified by rearranging the experimental setup to eliminate the reverse power. The new setup is presented in Figure 4.14b. An additional waveguide isolator was installed after the plasma applicator. In this arrangement, all of the microwave energy that passes through the plasma applicator is absorbed in the waveguide isolator and not reflected back as was the case in the experimental setup from Figure 4.14a. Figure 4.14. Microwave power flow diagram for (a) ignition study experimental setup and (b) the setup modified to reduce reverse power. Utilizing the modified experimental se tup shown in Figure 4.14b, values of the minimu m microwave power needed to create breakdown of gas were collected for argon, oxygen, and air. The electric field strength from the collected data was calculated using
71 equation 4.13 and presented in Figure 4.15. The results in this configuration do not dep end on the waveguide short position. The results are sim ilar to the result in Figure 4.11 Both have its minimum in the 2 to 4 T orr range The minimum breakdown electric field strength found for air t h rough experiment (16kV/m) is lower from the one shown in Figure 4.11 (40kV/m 120kV/m). The difference is due to different geometry of MPASD plasma discharge zone from the idealized parallel plate capacitor arrangement used to compute Figure 4.11. 0 5 10 15 20 25 0 2 4 6 8 10 12 14 Pressure [Torr] Electric field [kV/m] argon oxygen air Figure 4.15. The breakdown electric field strength of different gasses as a function of pressure. Data collected using system shown in figure 4.14 b. 4.3.3. Oxygen Plasma Ignition Study Results The results of the ignition of oxygen plasma study performed using the experimental setup shown in Figure 4.12 are summarized in Figure 4.16. Each data set was processed to show the average microwave power at plasma ignition as a function of pressur e
72 Figure 4.16. Average microwave power level needed to cause electrical breakdown of oxygen gas as a function of (a, b, c) pressure and (d) waveguide short position. The arrows show the direction of increasing wave guide short position (away from the plasma). The dependence of ignition condition on the waveguide short position is summarized in Figures 4.16(a,b,c). Initially, as the waveguide short is close to plasma zone, the microwave power level needed for plasma igni tion is high. However, as the waveguide short is moved away from the plasma zone the microwave power needed decreases (Figure 4.16a). While the waveguide short is moved from the 0 cm to 3.5 cm location the minimum microwave power needed remains constant, b ut the pressure range of the minimum microwave power range decreases (Figure 4.16b). Further movement of the waveguide short results in an increase of the plasma ignition minimum microwave power and reduces the pressure range (Figure 4.16c).
73 Figure 4.1 7. 3D (a) and (b) contour plot of experimentally obtained average microwave power level needed for oxygen plasma ignition as a function of position and pressure. The Figure 4.17 allows to determination of the minima of the average microwave power level needed to ignite the oxygen plasma. Figure 4.17a shows that such a minima exists, and Figure 4.17b locates the minima in the pressure power plane. The minima (the optimum ignition condition for MPASD system) for oxygen plasma ignition setting was determine d to be 2 Torr, 2 cm waveguide short position, and microwave forward power of more than 225 W. 4.3.4. Argon Plasma Ignition Study Results The ignition experiment was repeated for argon as the plasma medium. Similar to the results for oxygen there was significant dependence of microwave power needed to ignite plasma as a function of position and gas pressure. Figure 4.18 summarizes the collected data and allowed determination of the minima of the average microwave power level needed to ignite the argon plasma. Figure 4.18a shows that such a minima exist as well, and Figure 4.18b locates the m inima in the pressure power plane. The minima (the optimum ignition condition for MPASD system) for argon plasma ignition setting was
74 determined to be 3.5 Torr, 3 cm waveguide short position, and microwave forward power of more than 50 W Figure 4.18. 3D (a) and (b) contour plot of experimentally obtained average microwave power level needed for argon plasma ignition as a function of position and pressure. 4.3.5. Air Plasma Ignit ion Study Results Figure 4.19. 3D (a) and (b) contour plot of experimentally obtained average microwave power level needed for air plasma ignition as a function of position and pressure. The plasma ignition study was again conducted but using air as the plasma discharge gas. This study again showed the strong position and pressure dependence of
75 the power necessary to ignite the plasma. Figure 4.19 summarizes the microwave power level dependence on position and pressure during plasma ignition. Similar to the study for oxygen and argon there is a condition where the air plasma ignition occurs at the lowest power level. From Figure 4.19b this occurs at a pressure of 8 Torr, 3 cm position of the waveguide short, and with minimum power of 180 W. 4.3.6. Plasma Ig nition Study Summary The plasma ignition study showed that there was a strong relationship between the microwave power level necessary to ignite the plasma, the plasma gas pressure, and the MPASD system waveguide short position. The optimum conditions for the plasma ignition for oxygen, argon, and air were determined and the values were summarized in Table 4.3. Table 4.3. MPASD process setting for optimum plasma ignition conditions. Plasma gas Waveguide position Process gas pressure Minimum power Oxygen 2cm 2.0Torr 225W Argon 3cm 3.5Torr 50W Air 3cm 8.0Torr 180W In addition, it was verified that the ignition microwave power level dependence on position is due to the reverse microwave interfering with the forward wave to create areas of high and low e lectric field strength.
76 4.4. Plasma Tail Temperature Study As the gas exits the plasma zone, the MW source supplied energy is no longer available to maintain the temperature and the temperature decreases due to radiation, gas expansion, and mixing with cooler gases. The temperature of the plasma tail in relation to the distance from the waveguide at given gas flow, process power, and pressure was studied. The tail temperature decreases with distance from the waveguide 4.4.1. Plasma Tail Temperature Study Experiment al Setup The experimental setup for this study is shown in Figure 4.20. The k type thermocouple was used to probe the temperature. The location of the thermocouple was adjusted using the sample holder assembly and the plasma temperature was measured in the center of the gas flow. Figure 4.20. Plasma tail temperature study experimental setup. Plasma tail temperatures were measured after the process pressure, microwave power, waveguide position and waveguide tuner setting w ere adjusted until microwave r eflected power was minimized and the desired absorbed power was reached.
77 4.4.2. Plasma Tail Temperature Study Results The relation of plasma temperature and probe position is summarized in Figure 4.21. The tail temperature decreases with probe distance from the plasma zone. The distances of 44 mm and higher are possible sample locations and the te mperatures in these locations w ere as high as 615 K. The lowest measured temperature was 370 K at 84 mm and 218 sccm of O 2 flow. The plasma tail can be utilized to hea t the substrate by placing it close to the plasma zone opening. On the other hand samples that are sensitive to temperature can be placed farther away from the plasma opening where the tail temperature is lower and the process parameters can be changed to further decrease the tail temperature at the sample location. 0 500 1000 1500 2000 -20 0 20 40 60 80 100 Height [mm] Temperature [K] 700+18sccm O2 600+18sccm O2 500+18sccm O2 400+18sccm O2 300+18sccm O2 200+18sccm O2 Figure 4.21. Plasma tail temperature as a function of probe position for selected material flow rates. The process pressure was 60 Torr and process power was 1000 W of absorbed power and nebulizer duty cycle was 3%. It was found that the plasma temperature decreases with gas flow rate. This was surprising since the previous models predict that the temperature of the plasma increases with decreasing flow. The decreas e of the temperature of the plasma tail with gas flow can be explained by looking at the time a gas molecule takes to move from plasma to the
78 location probed in the experiment. At lower gas flows this time is longer. This will result in more energy radiate d away from the gas thus lower temperature of the gas in plasma tail. The plasma tail temperature dependence on process pressure is presented in Figure 4.22. The process pressure was varied between 20 and 16 Torr. The tail temperature was measured at posit ions of 4 mm, 44 mm, and 84 mm away from the plasma zone. There was some variation (125 K) at position of 4 mm away from plasma zone, however, the variation is less significant at other distances. 500 1000 1500 0 50 100 150 200 Pressure [Torr] Temperature [K] +4mm +44m m +84m m Figure 4.22. Plasma tail temp erature as a function of process pressure for selected probe positions. The process power was 600 W, the process gas flow was 700 + 18 sccm O 2 and nebulizer duty cycle was 3%. The plasma tail temperature dependence on nebulizer duty cycle is summarized in Figure 4.23. The nebulizer output mass flow dependence on its power duty cycle was presented in Section 4.5.
79 1400 1420 1440 1460 1480 1500 0% 50% 100% Duty Cycle [%] Temperature [K] 800 820 840 860 880 900 0% 50% 100% Duty Cycle [%] Temperature [K] 600 620 640 660 680 700 0% 50% 100% Duty Cycle [%] Temperature [K] Figure 4.23. The plasma tail temperature as a function of ne bulizer duty cycle for probe positions of (a) +4 mm, (b) +44 mm, and (c) 84 mm away from plasma zone. The process power was 600 W, process pressure was 60 Torr and the O 2 gas flow was 700 + 18 sccm. The temperature of the plasma tail at a location of 4 mm away from the plasma zone decreased with increased nebulizer duty cycle. This is expected as the plasma temperature is expected to decrease with increased material flow. However, the temperature of the plasma tail at locations of 44 mm and 84 mm away from the plasma zone increase with increasing duty cycle of the nebulizer. The increase of the temperatures at the 44 mm and 84 mm is consistent with the results for plasma temperature increase with gas flow as presented in Figure 4.21. Higher gas flow is exp ected for experiments with higher generated droplet flow rate due to higher water vapor content.
80 0 500 1000 1500 2000 0 200 400 600 800 1000 1200 Power [W] Temperature [K] +4mm +44mm +84mm Figure 4.24. Plasma tail temperature as a function of absorbed microwave power for selected probe positions The process pressure was 60 Torr, the process gas flow was 700 + 18 sccm O 2 and nebulizer duty cycle was 3%. Plasma tail temperature also depends on absorbed microwave power. This relation is shown in Figure 4.24. Since increased absorbed microwave power is expected to incr ease the plasma temperature then the plasma tail temperature is also expected to increase. Indeed the plasma tail temperature increases with increasing absorbed power. However, the increase is small. Increase of the absorbed power from 200 W to 1000 W rais es the tail temperature only 400 K at a position of 4 mm away from plasma zone. The plasma temperature at 200 W of absorbed power is 870 K higher than the temperature of the process gas as it enters the plasma zone. 4.4.3. Plasma Tail Temperature Study Summary T he plasma tail temperature decreases as the distance from the plasma increases. The plasma tail temperature pressure dependence is significant only at close distances from the plasma. Power changes do change the tail temperature especially at close
81 distan ces and the effect decreases at large distances from plasma. The material flow effect on the plasma tail temperature is small for the material flows utilized. The plasma tail can be utilized as a mechanism of substrate heating or the plasma temperature at the sample location can be reduced by changing location of the sample. If the heating of the substrate by the plasma tail is insufficient for a given sample then the active heating can be implemented by utilizing an electric sample heater. For some samples the plasma tail heating might be excessive then active cooling of the sample would be required. 4. 5 Atomizer Generation Rate The generation rate study was conducted to determine the amount of material being generated under selected duty cycle operation of th e nebulizer. The atomization rate is too small to measure below a 2% duty cycle and it remains constant above 10%. Between 2% and 10% of duty cycle the generation rate of the atomizer can be adjusted from 0 to 13 g/hr. 0 10 0 5 10 15 20 25 30 35 Duty cycle [%] Flow [g/hr] Figure 4.25. The atomizer generat ion rate as a function of power duty cycle.
82 4. 6 Chapter Summary MPASD system was shown to be capable of creating plasmas in argon, oxygen, and air, and optimal ignition settings w ere determined. The plasma was sustained upon introduction of atomized water. T he electron temperature for the plasma was measured to be between 6000 K and 40000 K. The pressure and microwave power effect on electron temperature was observed. The plasma tail was measured and the position, pressure, power, and flow relation to tempera ture was measured. In addition, the atomizer generation rate was characterized.
83 C HAPTER 5. Case Study: Y 2 O 3 :Eu Nanophosphor The synthesis of Y 2 O 3 :Eu nano particles using the MPASD system described in Chapter 3 and Chapter 4 is presented here. The ability to successfully synthesize nanoarticles with varying properties that depend on process settings is shown. The coatings of Y 2 O 3 :Eu w ere analyzed and their crystal structure, crystal lattice size, luminescence was determined. In addition coating grow th rate was measured and coating quality was observed. 5.1. Y 2 O 3 and Y 2 O 3 :Eu Overview Yttrium oxide, Y 2 O 3 is an air stable, white solid substance. It is the most important yttrium compound and is widely used to make YVO 4 europium and Y 2 O 3 europium phos phors that give the red color in color TV picture tubes and it is a common starting material for b oth materials science and inorganic compounds Yttrium oxide is also used to make yttrium iron garnets, which are v ery effective microwave filters and yttrium aluminum garnet (YAG) for use as a lasing media. The Y 2 O 3 crystal is excellent host for lanthanides. Lanthanide doped Y 2 O 3 crystals are widely used for their luminescence in visible and infrared range. The luminescence ranging from blue (Tm), trough green (Tb), yellow (Dy), pink (Sm), red (Eu), and infrared (Er, Tm, Yb, Nd) is possible for lanthanide doped Y 2 O 3 (Guyot 1996,
84 Yoda 2004, Anh 2007). High thermal conductivity of Y 2 O 3 makes it a promising material for high power laser media, being a possible rep lacement for the YAG crystal. Figure 5.1. High pressure transition of (a, b) Y 2 O 3 and (c, d) Eu:Y 2 O 3 Reproduced from Wang (2009). X The Y 2 O 3 crystal has been observed to have cubic, monoclinic and hexagonal structures. The most common body centered cubic structure is widely utilized and synthesized commercially. The phase change from cubic to hexagonal upon compression and from hexagonal to monoclinic during decompression has been observed in high pressure studies of Y 2 O 3 phase transitions (Wang 2009). Figure 5.2. The Crystal structure of (a) cubic, (b) hexagonal, and (c) monoclinic Gd 2 O 3 crystal. Reproduced from Zhang (2008)
85 Similar studies for Y 2 O 3 :Eu showed that the crystal cubic structure changes to monoclinic un der high pressure and then to hexagonal at even higher pressures. Upon decompression the hexagonal crystal reverts to monoclinic phase. The monoclinic phase is stable at atmospheric pressure. The structure of Y 2 O 3 is similar to the Gd 2 O 3 crystal presented in Figure 5.2 where the cubic, hexagonal, and monoclinic phase crystal structure is shown. The differences between crystal structures are important for Eu doped Y 2 O 3 crystal as the crystal effect on the doped site is important for luminescence. Cubic Y 2 O 3 has two distinct cation sites each six way coordinated. The oxygen atoms are located about the cation site on a distorted cube structure with two oxygen sites empty (Figure 5.3a), a quasi octahedral arrangement (Frantisek 1984, Jollet 1990, Maslen 1996). Figure 5.3. The cation site coordination for (a) cubic Y 2 O 3 and (b, c, d) monoclinic Eu 2 O 3 Reproduced from Maslen (1996) and Yakel (1978). For the monoclinic phase of Y 2 O 3 there are three distinct cation sites each seven fold coordinated. Two sites can be described as having oxygen located at the apexes of a trigonal prism with seventh oxygen located on a normal to the face. The third site is a distorted octahedron with seventh oxygen laying at large distance. This is similar to the monoclinic Eu 2 O 3 pre sented in Figure 5.3b, c, and d.
86 The XRD patterns of most often observed body centered cubic (BCC) or the base centered monoclinic (monoclinic) crystal structures of Y 2 O 3 are presented in Figure 5.4. 20 30 40 50 60 70 2Theta [deg] Count [arb] BCC Y2O3 BCM Y2O3 Figure 5.4. Powder XRD patterns of body centered cubi c and base centered monoclinic Y 2 O 3 X . The distinct XRD patterns help to identify the phase quickly. Eu doped Y 2 O 3 is known for its red luminescence. The luminescence observed is usually due to transition from 5 D 2 5 D 1 5 D 0 to 7 F J ground level, 5 D 0 to 7 F J being the most dominant resulting in emission at 611nm. Since the emission of the Eu cation is affected by the host lattice, the emission was found to change with size and crystal phase for Eu:Y 2 O 3 nano particle. 5.2. Thermolysis o f Y(NO 3 ) 3 powder The study of thermal breakdown of Y(NO 3 ) 3 xH 2 O in air was conducted to determine the dynamics of formation of Y 2 O 3 The reagent quality Yttrium (III) nitrate hydrate (Alfa Aesar #11187) was heated in an oven (Jelrus Temp Master model 18000 ) at selected temperatures for the duration of 1 hour for each step. After each heating step the sample was ground and the powder XRD pattern was recorded. The measured XRD patterns are summarized in Figure 5.5.
87 The XRD pattern of Y(NO 3 ) 3 xH 2 O powder befor e heating showed that it contains Y(NO 3 ) 3 6H 2 O. Upon heating the sample dehydration occurred and the phases of Y(NO 3 ) 3 5H 2 O, Y 2 (NO 3 ) 6 7H 2 O, Y(NO 3 ) 3 3H 2 O, and Y(NO 3 ) 3 H 2 O were the main phases detected for samples heated up to 300C. 10 20 30 40 50 60 2Theta [deg.] Count [arb.] Reagent Figure 5.5. The powde r XRD patterns of annealed Y(NO 3 ) 3 xH 2 O reagent. Some of the XRD peaks of BCC Y 2 O 3 (solid diamond), YO(NO 3 ) (open diamond), Y(NO 3 ) 3 3H 2 O (circle), Y(NO 3 ) 3 5H 2 O (triangle), and Y(NO 3 ) 3 6H 2 O (sqare) are marked. The sample processed at 400 C and 450 C pres ented different picture. XRD pattern for this sample showed strong presence of tetragonal YO(NO 3 ) suggesting that dehydration ended and thermal breakdown of Y(NO 3 ) 3 began. Upon further heating the Y(NO 3 ) 3 XRD pattern disappeared at 500 C and broad peaks
88 were observed. These peaks upon further heating had grown sharper and several other peaks evolved into the pattern representing the BCC phase of Y 2 O 3 This study suggest that the solution of Y(NO 3 ) 3 is appropriate for MPASD proce ss where it is expected that the temperatures reach above 500 C needed to synthesize Y 2 O 3 5.3. Deposition of Y 2 O 3 :Eu Coatings using MPASD System The samples of Y 2 O 3 :Eu w ere deposited using the MPASD system. The precursor used during deposition was a s olution of Y(NO 3 ) 3 xH 2 O (Alfa Aesar #11187) and Eu(NO 3 ) 3 xH 2 O (Alfa Aesar #15290) in water with the concentration of 0.0081 mol/L and 0.0009 mol/L respectively. Using equation 3.13, or easy to use Figure 5.6, the resulting particle size was expected to be 1 00 nm, and the resulting Y 2 O 3 :Eu particle to have 10% Yttrium replaced with Europium (Eu 0.2 Y 1. 8O 3 )). 0 100 200 0.00 0.05 0.10 Concentration [mol/L] Size [nm] 0.009 [mol/L] = 100 [nm] Figure 5.6. The processed nanoparticle size as a function of precursor concentration for MPASD system for Y 2 O 3 process using Y(NO 3 ) 3 solution precursor.
89 Table 5.1. MPASD process conditions for Y 2 O 3 :Eu coating deposition. Sample ID Deposition pressure Deposition power Deposition O2 flow Nebulizer duty cycle Sample distance from plasma Precursor Y concentration Precursor E u concentration Comment Torr W S ccm % mm mol/L mol/L S1 20 200 700+18 2.5 80 0.0081 0.0009 Luminescence S2 40 200 700+18 2.5 80 0.0081 0.0009 Luminescence S3 60 200 700+18 2.5 80 0.0081 0.0009 Luminescence S4 80 200 700+18 2.5 80 0.0081 0.0009 Lumin escence S5 100 200 700+18 2.5 80 0.0081 0.0009 Luminescence S6 120 200 700+18 2.5 80 0.0081 0.0009 Luminescence S7 140 200 700+18 2.5 80 0.0081 0.0009 Luminescence S8 20 400 700+18 2.5 80 0.0081 0.0009 Luminescence S9 40 400 700+18 2.5 80 0.0081 0.000 9 Luminescence S10 60 400 700+18 2.5 80 0.0081 0.0009 Luminescence S11 80 400 700+18 2.5 80 0.0081 0.0009 Luminescence S12 100 400 700+18 2.5 80 0.0081 0.0009 Luminescence S13 120 400 700+18 2.5 80 0.0081 0.0009 Luminescence S14 140 400 700+18 2.5 80 0.0081 0.0009 Luminescence S15 20 600 700+18 2.5 80 0.0081 0.0009 Luminescence S16 40 600 700+18 2.5 80 0.0081 0.0009 Luminescence S17 60 600 700+18 2.5 80 0.0081 0.0009 Luminescence S18 80 600 700+18 2.5 80 0.0081 0.0009 Luminescence S19 100 600 700+ 18 2.5 80 0.0081 0.0009 Luminescence S20 120 600 700+18 2.5 80 0.0081 0.0009 Luminescence S21 140 600 700+18 2.5 80 0.0081 0.0009 Luminescence S22 160 600 700+18 2.5 80 0.0081 0.0009 Luminescence S23 40 800 700+18 2.5 80 0.0081 0.0009 Luminescence S24 60 800 700+18 2.5 80 0.0081 0.0009 Luminescence S25 80 800 700+18 2.5 80 0.0081 0.0009 Luminescence S26 100 800 700+18 2.5 80 0.0081 0.0009 Luminescence S27 120 800 700+18 2.5 80 0.0081 0.0009 Luminescence S28 140 800 700+18 2.5 80 0.0081 0.0009 Lumin escence S29 60 1000 700+18 2.5 80 0.0081 0.0009 Luminescence S30 80 1000 700+18 2.5 80 0.0081 0.0009 Luminescence REF 70 700 700+18 2.5 80 0.0090 0 Reference Y 2 O 3 G1 70 700 700+18 2.5 80 0.0090 0 Growth study G2 70 700 700+18 2.5 80 0.0090 0 Growth st udy G3 70 700 700+18 2.5 80 0.0090 0 Growth study G4 70 700 700+18 2.5 80 0.0090 0 Growth study G5 70 700 700+18 2.5 80 0.0090 0 Growth study G6 70 700 700+18 2.5 80 0.0090 0 Growth study
90 The MPASD deposition condition is summarized in Table 5.1. Du ring deposition oxygen was used for the main and nebulizer flow. The gas flow is noted as sh i eld gas flow + nebulizer gas flow. For example, if shield gas flow is 700sccm and nebulizer gas flow is 18sccm then the gas flow will be written as 700+18sccm. The generation power of the nebulizer was controlled using the circuit described in Appendix A. All samples in th e power versus pressure study w ere deposited at the same operating conditions of MPASD except that the microwave power and process gas pressure w ere varied and the location of the waveguide short position was adjusted for each power and pressure setting in order to localize the plasma discharge within the discharge tube. 5.4. Morphology of Y 2 O 3 :Eu Coatings Figure 5.7. Picture of selected coa tings. Samples S15 (a) and S16 (b) are shown. The as deposited samples w ere inspected visually for the coating quality. The coatin gs deposited at pressures of 20 Torr had poor surface quality with uneven coating
91 and many visual blemishes. On the other han d, the coatings deposited at pressures of 40 Torr and above appeared to have smooth surface with only single spot blemish as it is presented in Figure 5.7. These coatings w ere easily disturbed and could be removed from the substrate by a gentle wipe down. In addition, the samples deposited at pressures of 120 Torr and above had either a very thin or no coating. Further investigation of the coating using SEM showed that the samples deposited under 20 Torr of pressure did not have the desired nanoparticles. T hese samples had a solid thick film with an uneven surface (Figure 5.8a). On the other hand the samples deposited at pressures of 40 Torr and above had a smooth coating that was composed of nano particles and some sub micron sized particles (Figure 5.8b, c and d). Figure 5.8. SEM images of coatings deposited at (a) 20 Torr and (b, c, d) 60 Torr. The coating deposited at 60Torr is shown at three magnification levels. Samples S15 (a) and S16 (b, c, d) are shown
92 One feature worth notice was a small patc h, usually in the center of deposition, where the deposition was thinner. The analysis of the thinning, or bald spot, of the coating is presented in Figure 5.9. Each image was taken in 1 mm intervals sweeping across the bald spot, which clearly shows a spa rser deposition in the center of the bald spot. The bald spot can be explained by looking at the possible flow patterns of the gas as it exits the plasma zone. The bald spot is more apparent at low pressure depositions than at higher pressure depositions. For the low pressure depositions the velocity of the gas, under same mass gas flow, is higher than for high pressure depositions. Assuming the same temperature and same mass gas flow the velocity of the gas will be twice as high at 40 Torr as at 80 Torr. A s a result, the fast moving gas blows off the coating in the center of the sample. Figure 5.9. SEM images of a coating and the bald spot on the coating. The images were taken at same magnification and 1mm apart. Sample S16 is sho wn. The images were taken at same magnification and 1mm apart.
93 The presence of large particles can be contributed to size distribution of the nebulizer droplet generation and to atomized droplet collision before pyrolysis. 0 100 200 300 400 500 600 700 800 900 1000 1um 2um 3um 4um 5um 6um 7um 8um 9um 10um 11um 12um 13um 14um 15um 16um 17um 18um 19um 20um 21um Diameter [nm] Count Droplet 0 100 200 300 400 500 600 700 800 900 1000 58nm 120nm 180nm 230nm 290nm 350nm 410nm 470nm 530nm 580nm 640nm 700nm 760nm 820nm 880nm 940nm 990nm 1.05um 1.11um 1.17um 1.23um Diameter [nm] Count Particle Figure 5.10. The droplet and resulting calculated particle size distribution. Data for droplet size distribution supplied by manufacturer. The nebulizer data shows that the droplet size distribution will result in some large particles (Figure 5.10). The data shows that most part icles fall in the 60 nm, and 120 nm statistical bins, but particles with sizes larger than that account for 10% of the distribution of the nebulizer since droplets of 2 The TEM and STEM images (Figure 5.11. and 5.12) show that the deposited coatings are composed of spherically shaped particles. Many particles are in contact with each other, but do not appear to be bonded to each other
94 Figure 5.11. The TEM image of several particle clusters from Eu:Y 2 O 3 nanoparticle coating.
95 Figure 5.12. The STEM image of several particle clusters from Y 2 O 3 :Eu nanoparticle coating. The TEM and STEM images presented in Figure 5.11 and 5.12 show that the particles in the deposited coating have a size distribution. An analysis of diameters of 120 particles is summarized in Figure 5.13. The largest particle found in the data set was 329nm and the smallest was 44nm. The average, median, and mean volume diam eter are 151nm, 143nm, and 169nm respectfully. Although the mean particle diameter calculated from the precursor chemistry and the nebulizer is 100nm the distribution shown in Figure
96 5.13 is skewed toward larger diameters and peaks for particles in the 12 5nm bin. This could be due to the assumptions made about the precursor solution surface tension, density, and viscosity during calculation (Equation 3.13). In this calculation the surface tension and viscosity was assumed to be the same as of pure water and the actual the values for the precursor could account for the discrepancy between the calculated and the observed diameters. 0 20 40 60 80 100 120 140 0 5 10 15 20 25 30 35 0 50 100 150 200 250 300 350 Cumulative count Count Diameter [nm] Distribution Cumulative Mean Diameter = 151nm, S.D. = 53nm Figure 5.13. The histogram of particle size distribution of Eu:Y 2 O 3 The nebulizer operating frequ ency affects the droplet size. The actual operating frequency was not measured and its value could be different from the value advertised by nebulizer manufacturer, affecting the deposited coatings mean particle diameter. In addition, the collisions of p recursor droplets increase the final particle size. As the atomized droplet is moved toward the plasma zone, each collision between two droplets creates one larger droplet with volume that is a sum of the initial droplets. This will skew the statistical distribution toward larger sizes for the deposited particles. If two droplets of 1.7um diameter collide, this would result in two 100nm particles collide, then the resulting particle size increases to 125nm. The particle size
97 after eight 1.7 um droplets c ollide will result in processed particle diameter of 200nm. These collisions could play large role in the distribution peak shift toward larger diameters in the statistical distribution shown The statistical distribution of generated droplets is inherent to the atomization technique s and any improvements in this distribution would require redesign of the atomizer. However, the particle size distribution shift toward larger particles due to droplet collisions could can be addressed. The number of the col lisions mainly depends on droplet concentration in the material flow and the time that it takes the droplet to reach plasma zone. Both parameters can be regulated by changing the nebulizer gas flow, changing the process pressure, and changing the nebulize r generation rate. Increase in gas flow or decrease in process pressure decreases both concentration and transit time. The decrease in nebulizer generation rate decreases the concentration of particles in nebulizer flow. 5.5. Growth Rate Study of Y 2 O 3 : Eu Coatings In order to determine the growth rate of the MPASD process several depositions of Y 2 O 3 w ere performed and the coatings were analyzed using SEM. Each deposition was performed under same process conditions but with different deposition times. Si x depositions with deposition times varying from 30 seconds to 16 minutes w ere analyzed using SEM to make qualitative and quantitat ive observations. All samples w ere deposited at an 80 torr process pressure, 700 W absorbed microwave power, 500 sccm and 18 sccm of oxygen flow through the sheath and nebulizer nozzle, respectfully, and 2.5% nebulizer duty cycle.
98 Figure 5.14. SEM images of Y 2 O 3 coating deposited for (a) 30s, (b) 1min, (c) 2min, (c) 4min, (d), 8min, and (f) 16min. The SEM images of the depo sited samples are shown in Figure 5.14. Clearly, the thickness of the coating increased with longer deposition times. Usual methods of measuring thickness of a coating utilize contact probe profiling, noncontact AFM technique, x ray reflectometry techniqu e, elipsometry techniques, thin film interference techniques, and SEM cross section imaging. Unfortunately, the nature of the deposited
99 coatings did not allow one to utilize any of the standard techniques. Nevertheless, a method to estimate the coating gro wth rate was needed. A method of measuring the observed coating coverage area was utilized to estimate the coating growth rate. In this technique the SEM high contrast images w ere taken resulting in high contrast, black and white images (Figure 5.15b). T he white features of the image are due to the deposited particles while the black areas represent the uncoated substrate. The coating percentage area coverage was taken to be the ratio of white pixels to number of the image pixels. Figure 5.15. The norm al (a) and high contrast (b) image of Y 2 O 3 coating deposited for 16 minutes. The observed area coverage was 41%. The coatings made during short depositions have the characteristic of being sparse with particles separated from each other and with very few agglomerations. However, for the depositions of 4 minutes and longer there is a significant agglomeration of particles resulting in part of the particles resting on top of the other. This agglomeration is undesired. If a particle was located atop of anoth er then the observed coating coverage appeared to be lower than it actually was. The calculated coverage values are presented in Figure 5.16. Assuming constant growth rate and non overlapping coating the data should have a linear trend. From the
100 graph it i s clear that the data has a nonlinear behavior. This behavior can be explained by noticing that for the samples with long deposition times some coating particles are resting on top of others. This causes some of the particles to be obstructed by other part icles on the SEM image. The coating area coverage calculations for these overlapping particle images will incorrectly under estimate the coating growth rate. 0 20 40 60 80 100 0 2 4 6 8 10 12 14 16 Deposition Time [min] Coating % Coverage Figure 5.16. Observed coating percentage coverage as a function of time for deposition of Y 2 O 3 coating (solid line). The dash line represents the calculated 6.4%/min growth. In order to reduce the effect of particle overlap on the result of the growth rate estimation, only the coatings with sho rt deposition times ( <2 min) w ere utilized in the cal culation. In addition, it was assumed that coating coverage was 0% at the start of the deposition. Under these assumptions of the deposition it was found that the coating area coverage rate is 6.4%/min. The nebulizer manufacturer specified that the area me an diameter of the precursor droplet to be 1.41 m and volume mean diameter to be 1.79 m This allows calculating the number of particles and coating volume based on area of coverage. Given the area coverage growth rate and volume mean diameters, the equi valent volume film thickness growth rate is comparable to 100 nm/min growth rate of a solid film.
101 5.6. Crystal Structure Study of Y 2 O 3 :Eu Coatings X Ray diffraction (XRD) and electron diffraction can help to determine crystal structure, chemical makeup, p hase, crystal orientation, residual stresses, and crystallite size for powder samples. The XRD and selective area electron diffraction (SAED) techniques are utilized to study the crystal structure of the deposited coatings. In addition, the high resoluti on transmission electron (HRTEM) imaging was utilized in this crystal structure study. qualitative determination of lattice constant and crystallite size. The crystal structure was determined by comparison with the published crystallographic database. Table 5.2. X XRD model Bruker AXS D8 Monochromator model K FL Cu 2k with 0.4mm slit Source aperture 0.4 mm Divergence apreture 1.0 mm G enerator setting 40 kV, 40 mA Goniometer type 200mm Brag Brentano circle Detector model LynxEye, 192 x 75um detectors Incident aperture 2.5 soller slit with Ni Kb fiter and optional 8mm slit Scan type XRD data was collected using Bruker AX S diffractometer equipped with 2 kW Copper anode X ray source, and the high speed LynxEye X ray detector, on a 200 mm Brag Brentano circle goniometer. The XRD pattern analysis was performed with the help of DiffracPlus EVA software from Bruker AXS that inc orporated the PDF2 2005 release database for XRD pattern matching.
102 Figure 5.17 shows selected XRD patterns for substrate and samples deposited at pressures of 20 Torr and 40 Torr. The marked peak locations of corresponding base centered monoclinic phase o f Y 2 O 3 The results of this study had shown that samples deposited at pressures of 40 Torr and above are consistent with base centered monoclinic phase of Y 2 O 3 and Eu doped Y 2 O 3 On the other hand the samples deposited at a pressure of 20 Torr had not show n any presence of Y 2 O 3 --( 2 0 -2 ) --( 2 0 2 ) --( 1 1 1 ) --( 4 0 1 ) --( 4 0 -2 ) --( 0 0 3 ) --( 3 1 0 ) --( 1 1 -2 ) --( 6 0 0 ) --( 1 1 -3 ) --( 5 1 -1 ) --( 3 1 -3 ) --( 3 1 3 ) --( 0 2 0 ) --( 7 1 -2 ) --( 7 1 1 ) 20 30 40 50 60 ] Ln(Intensity) [arb] ` A C B Figure 5.17. The powder XRD pattern of (a) Si 100 single crystal substrate and Y 2 O 3 :Eu sample deposited on Si ( 100 ) substrate under pre ssure of (b) 20Torr and (c) 40 Torr. The diffraction peaks of monocl inic phase of Y 2 O 3 :Eu are marked. It is known that the crystal size affects the x ray diffraction. The finite size if a crystal causes a broadening of the x ray peak and the broadening can be described using the Sreerrer equation ) cos( hkl hkl hkl L k (5.15) Using this relation and the XRD data collected the calculat ions of the particle diameter w ere conducted for samples where the XRD pattern signal was strong. The diameter measurement for each sample was based on the XRD peak broadening for the ( 111 ), ( 401 ), ( 2 40 ), ( 003 ), ( 310 ), and ( 3 31 ) peaks. The res ults for the selected planes w ere
103 averaged and the deviation calculated. The resu lts of the calculation are summarized in Table 5.4. The calculations show that the crystal size was between 88 nm and 159 nm. There is no observable power or pressure dependence of the crystallite size. The observed particle sizes are consistent with the diameter 100 nm calculated using equations 3.13 and given precursor chemistry and concentration. Table 5.3. The particle diameter calculation results using the broadening of XRD diffraction peaks for selected samples. Deposition Power [W] Deposition Pre ssure [torr] Measured Diameter [nm] Diameter Deviation [nm] 400 40 100.7 9.0 400 40 153.1 30.4 400 60 148.2 19.2 400 80 139.0 21.6 400 100 159.7 45.3 600 60 88.2 7.9 600 40 107.3 8.3 600 60 116.7 17.2 600 80 139.4 15.2 600 100 148.8 31.3 800 80 95.7 7.1 800 40 108.3 7.6 800 60 113.7 7.5 800 80 135.9 19.7 800 100 131.8 26.0 1000 60 97.8 10.1 1000 60 106.8 12.9 1000 80 137.9 23.4 400 40 100.7 9.0
104 The powder XRD measurements allow for analysis of the coating as a whole, while the HRTEM wit h SAED allows verifying the structure of single particle within the coating. The Figures 5.18, 5.19, and 5.20 present the HRTEM image of single nanoparticle of Y 2 O 3 :Eu. Figure 5.18. The HRTEM image of Y 2 O 3 :Eu nanoparticle. The observed interplanar sp acing is 0.69 nm.
105 The interplanar distance found to be 0.69 nm, a value consistent with (6 0 0) plane being in focus of the HRTEM image. This measurement was verified in three locations of the particle. Figure 5.19. The HRTEM image of Y 2 O 3 :Eu nanopar ticle. The observed interplanar spacing is 0.69 nm.
106 Figure 5.20. The HRTEM image of Y 2 O 3 :Eu nanoparticle. The observed interplanar spacing is 0.68 nm. Figure 5.21 presents the HRTEM image of another Y 2 O 3 :Eu particle. The orientation of this parti cle differs from the orientation of the particle from Figures 5.18, 5.19, and 5.20. The observed interplanar distance is found to be 0.37nm suggesting that the image shows the (1 1 2) plane of Y 2 O 3 :Eu crystal. The SAED diffractogram for this particle is shown.
107 Figure 5.21. The HRTEM image of Y 2 O 3 :Eu nanoparticle. The observed interplanar spacing is 0.37 nm. The insert shows the SAED diffractogram. 5.7 The energy dispersive X ray spectrum of Y 2 O 3 :Eu coatings. The TEM utilized allowed to perform e nergy dispersive X ray spectroscopy (EDS) measurement. The results of this measurement are shown in Figure 5.22.The Eu signal verifies that the particle is doped with Eu. The EDS
108 measurement allowed to measure the spectrum from small area of just four p articles. Figure 5.22. The e ne rgy dispersive X ray spectrum of small cluster of Y 2 O 3 :Eu particles. However, the instrument utilized did not allow to perform quantitative analysis of the sample. Namely, the doping level of the Eu in the particles was n ot determined. 5.8. Anneal Study of Y 2 O 3 :Eu Coatings The base centered monoclinic phase of Y 2 O 3 :Eu is a high pressure stable phase for bulk material while cubic phase is the energetically favorable phase. The most common phase of Y 2 O 3 :Eu is the BCC phase In an attempt to obtain the BCC phase of Y 2 O 3 :Eu an anneal study on one sample was performed and the SEM images, and the XRD profile was recorded at five annealing temperatures. Each annealing step was done in Jelrus Temp Master M model 18000 oven in air environment for the duration of 2 hours. Additionally, the annealing was repeated for the 1100 C annealing for an additional 72hrs.
109 20 25 30 35 40 45 50 55 60 65 intensity (log scale) [arb.] 600 C 700 C 500 C no anneal 900 C 1100 C 2hrs 1100 C 72hrs intensity (log scale) [arb.] Figure 5.23. The XRD patterns for sample of Y 2 O 3 :Eu annealed at selected annealing temperatures. The diamond markers po int to the characteristic XRD peak locations of base centered monoclinic phase of Y 2 O 3 :Eu. The measured XRD patterns are presented in Figure 5.23. The monoclinic phase is present in all samples annealed at temperatures of 900 C and below. However, at a t emperature of 700 C, there is an additional broad peak at 45 that is not consistent with the monoclinic, cubic, or even hexagonal phase of Y 2 O 3 :Eu. Annealing at the temperature of 1100 C has destroyed the monoclinic phase: the characteristic peaks of mo noclinic are absent in the sample annealed at 1100 C, nor was the cubic phase of Y 2 O 3 :Eu observed. 5.9. Photoluminescence of Y 2 O 3 :Eu Coatings The Eu doped Y 2 O 3 crystal has been observed to luminescent in the red part of the visible spectrum. In order to verify this phenomenon a photoluminescence experiment was conducted.
110 A preliminary study utilizing the 410 nm wavelength laser and the set of 425, 420, 415, 405, 400, and 395 nm LEDs showed no luminescence for all of the samples. This is consistent with p ublished results for the monoclinic Y 2 O 3 :Eu that have shown that the absorption occurs around 250 nm. Therefore, the study utilizing a UV source was conducted. The experimental setup utilizes a UV light source, short pass filter, sample to be tested, long pass filter, spectrometer, and computer with A/D acquisition card and custom software. The diagram below shows the basic arrangement of experimental setup. Figure 5.24. Photoluminescence experiment setup diagram The UV light source utilized was a xenon short arc flash lamp with a borosilicate window (Perkin Elmer FX 1163). Although xenon arc lamps are capable of producing deep UV light, the construction of the arc lamp used, namely the borosilicate window of the lamp, limited the usable UV spectrum to w avelengths longer than 225nm.
111 The sample tested was exposed to UV radiation with the spectral profile controlled by the short pass filter. The transmission spectrum of the short pass filter is chosen as to prevent overlapping of the excitation signal and the photoluminescence signal of the sample. 0% 20% 40% 60% 80% 100% 200 300 400 500 600 700 800 wavelengh [nm] transmission [%] long pass filter shot pass filter (u330) short pass filter (u340) ` Figure 5.25. Transmission spectrum of filters utilized in the luminescence experiment. The short pass filters used in the experi ment w ere the Hoya u330 and u340 glass. Although these filters have the desired UV characteristics, the visible and NIR transmission of the filters prevents the measurement of luminescence in the deep red range and NIR range. This is presented in Figure 5.25. The photoluminescence signal of the sample was filtered using the long pass filter with measured spectral profile shown in Figure 5.25. The long pass filter helps to reduce the stray light signal in the spectrometer due to excitation light entering the spectrometer. The light collected by the optical fiber was analyzed using a s pectrometer (Ocean Optics model S2000) capable of spectrally resolving the light in the range of 350 nm to 1100 nm with a detector resolution of 0.36 nm/pixel and effective optical FWHM resolution of 9 nm. The spectrally resolved signal from the spectromet er was recorded on
112 the computer for further analysis with the help of an analog to digital acquisition card (National Instruments PCI 6251) and custom program written utilizing LabView software 0 200 400 600 800 1000 1200 0 20 40 60 80 100 120 140 160 180 pressure [torr] power [W] Figure 5.26. The power and pressure conditions for the thi rty samples tested. The red bubble represents the photoluminescence detected. The varing red area represents the relative intensity of photoluminescence. Thirty samples w ere analyzed for luminescence. Each sample was deposited using MPASD under different pressures and power while keeping other deposition conditions constant. The power was varied from 200 W to 1000 W absorbed power and the deposition chamber pressure was varied from 20 Torr to 160 Torr. Figure 5.25 presents each sample location on the Powe r Pressure plane.
113 500 550 600 650 700 Wavelength [nm] Intensity [arb] 140torr 120 torr 100 torr 80 torr 60 torr 40 torr 20 torr Figure 5.27. Optical spectrum of the photoluminescence of Y 2 O 3 :Eu particles deposited at 600W absorbed power. The excitation spectrum was filtered using Hoya U340 glass. The first luminescence study was performed using the U340 filt er placed in the excitation light path. The measurements showed that only the samples deposited under 40 Torr and 60 Torr of pressure had measurable luminescence, this is shown in Figure 5.27 for samples deposited under 600 W of absorbed power. Measurement s for samples deposited under different microwave absorbed power conditions had similar photoluminescence properties as a function of pressure. 0 0.2 0.4 0.6 0.8 1 1.2 1.4 585 605 625 645 40torr 200W 40torr 400W 40torr 600W 40torr 800W 0 0.2 0.4 0.6 0.8 1 1.2 1.4 585 605 625 645 60torr 200W 60torr 400W 60torr 600W 60torr 800W Figure 5.28. The spectral intensity photoluminescence of Y 2 O 3 :Eu particles deposited at 40 Ttorr and 60 Tor r process chamber pressure. The excitation spectrum was filtered using Hoya U340 glass.
114 The luminescence of the samples deposited under 40 Torr and 60 Torr chamber pressure is shown in Figure 5.28. There are two observable intensity peaks at 595 nm and 62 6 nm. The spectrometer used in the experimental setup had an optical resolution of 9 nm FWHM. As a result, any features narrower than 9 nm are not distinguishable in Figure 5.28. While comparing the photoluminescence intensity between different samples it is important to notice that the intensity of the photoluminescence does depend on the amount of the luminescent material on the sample. Although the material flow was kept constant for all deposited samples and the deposition times are constant, the actual deposited coating thickness or mass is unknown for the samples. For this study, the actual luminescence of the sample, regardless of the coating thickness or mass, was compared. From Figure 5.28 it can be seen that the luminescence increases with process absorbed power and is larger for samples deposited at 40 Torr than 60 Torr. This trend is summarized in Figure 5.29 below. Because of the limited samples deposition power range of this study, it is unknown where the intensity of the luminescence reaches i ts maximum within the power pressure range tested. This indicates the need for further study of new samples deposited under higher absorption power. Because there are only two luminescent samples at each power level it was not possible to fully present the luminescence intensity as a function of pressure. However, from Figure 5.28 it is clear that the intensity of the samples deposited at 40 Torr deposition chamber pressure is higher than those deposited at 60 Torr.
115 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0 200 400 600 800 1000 1200 Power [W] Intensity [arb.] 60torr (U340) 40torr (U340) Linear (40torr (U340)) Figure 5.29. Luminescence intensity as a function of sample deposition a bsorbed power. The excitation spectrum was filtered using Hoya U340 glass. The luminescence study was repeated with the excitation filter changed to the Hoya U330 filter. Other parameters of the study w ere kept consistent with the first part of the luminescence study. 500 550 600 650 700 Wavelength [nm] Intensity [arb.] 140torr 120 torr 100 torr 80 torr 60 torr 40 torr 20 torr Figure 5.30. Optical spectrum of the photoluminescence of Y 2 O 3 :Eu particles deposited at 600 W absorbed power. The excitation spectrum was filtered using Hoya U330 glass. Similar to the first part of lumi nescence study the measurements showed that only the samples deposited at 40 Torr and 60 Torr pressure had measurable luminescence. This is shown in Figure 5.30 for samples deposited under 600 W absorbed power.
116 Measurements for samples deposited under dif ferent absorbed power conditions have similar results. The luminescence of the samples deposited under 40 Torr and 60 Torr chamber pressure is shown in Figure 5.31. The luminescence with peak intensity at 595 nm and 626 nm is similar to the results shown i n Figure 5.28. Again, features which are narrower than the 9 nm resolution limit are not visible in Figure 5.31. 0 1 2 3 4 5 6 585 605 625 645 Wavelength [nm] Intensity [arb.] 40torr 200W 40torr 400W 40torr 600W 40torr 800W 0 1 2 3 4 5 6 585 605 625 645 Wavelength [nm] 60torr 200W 60torr 400W 60torr 600W 60torr 800W Figure 5.31. The photoluminescence of Y 2 O 3 :Eu particles deposited at 40 Torr and 60 Torr process chamber pressure. The excitation spectrum was filtered using Hoya U330 glass. From Figure 5.31 can be seen that the luminescence increases with process absorbed power and is larger for samples deposited at 40 Torr than 60 Torr. This trend is summarized in Figure 5.32.
117 0 2 4 6 0 200 400 600 800 1000 1200 Power [W] Intensity [arb.] 60torr (U330) 40torr (U330) 60torr (U340) 40torr (U340) Linear (40torr (U330)) ` Figure 5.32. Luminescenc e intensity as a function of sample deposition Absorbed power. The luminescence spectral response in this measurement is similar to the luminescence measurement wh ere the U340 excitation filter was used. However, the intensity of the luminescence is signi ficantly higher. The luminescence intensity when using the U330 filter has been increased by five times as compared to the study when the U340 filter was used. The increase in luminescence intensity can be explained by comparing the excitation spectrum in the experiment that used the U330 filter to the experiment with U340 filter (Figure 5.33). This can be explained by looking at the absorption range of the Y 2 O 3 nano particles. The published absorption studies of Y 2 O 3 :Eu (Figure 5.33) show that the maximum absorption occurs at 250 nm. However, the peak intensity of excitation source used for this luminescence study is 330 nm and 340 nm for U330 and U340 filters, respectively. However, there is an overlap of the published absorption range of Y 2 O 3 :Eu and the excitation range in the experimental setup as it is shown in Figure 5.33. The excitation spectrum during the experiment utilizing both U330 and U340 filters clearly overlaps the absorption spectra of the Y 2 O 3 :Eu. However, the experiment utilizing the
118 U330 filter has greater overlap than in case of the U340 filter, which can explain the stronger photoluminescence in the experiment utilizing the U330 filter. 150 200 250 300 350 400 450 Wavelength [nm] Intensity [arb] Y2O3 absorption excitation u330 excitation u340 Figure 5.33. Comparison between the Y 2 O 3 :Eu absorption spectrum and the excitation spectrum of expe riment with U330 and U340 excitation light filters. The effect of the absorption and excitation spectral overlap on luminescence can be estimated. The intensity of luminescence will depend on the amount of the energy absorbed by the sample. The absorption range and excitation spectral profiles allow calculation of the difference between the two experiments due to the different excitation sources. In the calculations the Xenon lamp and the filter manufacturers data was utilized together with the absorption study published in. The effective absorbed energy was calculated by multiplying the absorption and the excitation intensity at each wavelength in the 200 to 300 nm range. The calculated effective absorbed spectrum for the experiment with U330 and U340 fi lters is presented in Figure 5.34. It is clear that the sample absorbs significantly less energy in experiment with the U340 filter than with the U330 filter. Simple integration shows that the energy absorbed by the sample in the experiment with U330 filte r is about 3.5 times larger than with the U340 filter. This difference can be the most
119 significant reason for the difference in luminescence intensity between the experiment with U330 and U340 filters. 200 220 240 260 280 300 Wavelength [nm] Intensity [arb.] excitation u330 excitation u340 Figure 5.34. The spectrum of absorbed excitation li ght in experiment with U330 and U340 excitation path filter. The lack of luminescence at pressures of 60 Torr and above combined with the crystal lattice size sheds light onto the state of the Eu doping in the crystal. Comparing lattice volume size and lu minescence study shows that the samples with larger lattice volume are also the samples that luminescence. It is possible that the luminescence is due to the Eu 3+ ion replacing the Y 3+ ion of the Y 2 O 3 structure. Since the ionic radius of the Eu 3+ ion is la rger than of the Y 3+ ion then the lattice of Eu doped Y 2 O 3 it is expected to increase if the Eu 3+ ion is activated to the lattice location. The luminescence study showed that the samples of Y 2 O 3 :Eu prepared using MPASD system do exhibit luminescence under some deposition conditions. The luminescence was observed only for the samples deposited under pressures of 40 and 60 Torr. The samples deposited under 40 Torr did luminescence more intensely than the samples deposited under 60 Torr of pressure by a facto r of two. There is a luminescence dependence on power and the samples deposited at higher power levels luminescent
120 stronger. The lack of luminescence in samples deposited at 80 Torr and above is observed for samples with smaller lattice size. This suggest that Eu is not activated into lattice size in the samples deposited at pressures of 80 Torr and higher. Although the absorption study was not performed, the use of two different excitation filters during the study showed different luminescence levels cons istent with the published absorption data for Y 2 O 3 :Eu 5.10. Chapter Summary The deposition of Y 2 O 3 :Eu using MPASD system was successful. MPASD system allowed for synthesis of a crystalline nanoparticle coatings of Y 2 O 3 :Eu without the need for further po st annealing. The effects of process pressure and process microwave power on the optical and structural properties of the deposited samples were presented. As deposited the coatings had monoclinic structure and annealing did not change the structure to the more common cubic phase. Changes in properties of the coating wit h changing process parameters w ere observed. The coatings luminescence and crystal size depended on the deposition conditions. This suggests that the process parameters can be controlled to yield desired coating properties.
121 LIST OF REFERENCES Abdelsayed, V. Glaspell, G. Nguyen, M. Howe, J.M. El Shall, S. (2008) Laser synthesis of bimetallic nanoalloys in the vapor and liquid phases and the magnetic properties of PdM and PtM nanop articles (M=Fe, Co and Ni). The Royal Society of Chemistry: Faraday Discussions. 138, 163 180. Anh, T. Benalloul, P. Barthou, C. Giang, L. Vo, N. Minch, L. (2007) Luminescence, energy transfer, and upconversion mechanisms of Y 2 O 3 nanomaterials do ped with Eu3+, Tb3+, Tm3+, Er3+, and Yb3+ ions. Journal of Nanomaterials. 48247. Astruck, D. Lu, F. Aranzaes, J.R. (2005) Nanoparticles as recyclable catalysts: the frontier between homogeneous and heterogeneous catalysts. Angewandte Chemie Internation al Edition. 44, 7852 7872. Avvaru, B. Patil, M.N. Gogate, P.R. Pandit, A.B. (2006) Ultrasonic atomization: effect of liquid phase properties. Ultrasonics. 44, 146 158. Baughman, R.H. Anvar, A.Z. de Heer, W.A. (2002) Carbon nanotube the routhe t oward applications. Science. 297, 787 792. Baughman, R.H. Cui, C. Zakhidov, A.A. Iqbal, Z. Barisci, J.N. Spinks, G.M. Wallace, G.G. ... Kertesz, M. Carbon nanotube actuators. Science. 284, 1340 1344. Baum, R. (2003) Drexler and Smalley make t he case for and against molecular assemblers. Chemical & Engineering News. 81 (48), 37 42. Bhargava, R.N. Gallagher, D. (1994) Optical properties of manganese doped nanocrystals of ZnS. Physical Review Letters. 72 (3), 416 419.
122 Borgia, I. Brunetti, B Mariani, I. Sgamellotti, A. Cariati, F. Fermo, P. Mellini, M. Viti, C. Padeletti, G. (2002) Heterogenous distribution of metal nanocrystals in glazes of historical pottery. Applied Surface Science. 185, 206 216. Camillone, N. Adib, K. F itts, J. Rim, T.R. Flynn, G.W. Joyce, S.A. Osgood, R.M. (2002) Surface termination dependence of the reactivity of single crystal hematite with CCl4. Surface Science. 511, 276 282. Champion, Y. (2007) Gas phase synthesis of nanopowders. Nanomateria ls and Nanochemistry (16). Retrieved from www.springerlink.com Chen, Q. Zhou, M. Fu, Y. Weng, J. Zhang, Y. Yue, L. Xie, F. Huo, C. (2008) Magnetron sputtering synthesis silver and organic PEO nanocomposite. Surface & Coatings Technology. 202, 5576 5578. Das, S.K. Putra, N. Thiesen, P. Roetzel, W. (2003) Temperature dependance of thermal conductivity enchancement for nanofluids. Journal of Heat Transfer. 125, 567 574. Dreesen, L. Colomer, J. Limage, H. Gigure, A. Lucas S. (2009) Synthesis of titanium dioxide nanoparticles by reactive DC magnetron sputtering. Thin Solid Films. 518, 112 115. Drexler K. E. (1986) Engines of creation. Garden City, N.Y.: AnchorPress/Doubleday. Drexler K. E. (1992) Nanosystems: molecular machinery, manu facturing, and computation. New York, N.Y.: Willey. Eigler, D. M. Schweizer, E. K. (1990) Positioning single atoms with a scanning tunnelling microscope. Nature. 344, 524 526. Faraday, M. (1857) The Bakerian lecture: Experimental relations of gold (and o ther metals) to light. Phillosophical Transactions of the Royal Society of London. 147, 145 181. Frantiek, H. (1984) Real structure of undoped Y 2 O 3 single crystals. Acta Crystallographica Section B. B40, 76 82.
123 Freestone, I. Meeks, N. Sax, M. Higgit t, C. (2007) The Lycurgus cup a Roman nanotechnology. Gold Bulletin. 40 (4), 270 277. Gao, P.X. Mai, W. Wang, Z.L. (2006) Superelasticity and nanofracture mechanics of ZnO nanohelices. Nano Letters. 6 (11), 2536 2543. Gell, M. Jordan, E.H. Teicho iz, M. Cetegen, B.M. Padute, N.P. Xie, L. Chen, D. Ma, X. Roth, J. (2008) Thermal barrier coatings made by the solution precursor plasma spray process. Journal of Thermal Spray Technology. 17 (1), 124 135. Geohegan, D. B. Puretzky, A. A. Ro uleau C. Jackson J. Eres, G. Liu, Z. Styers Barnett, D. Hu, H. Zhao, B. Ivanov, I. Xiao, K. More, K. (2009) Laser interactions in nanomaterials synthesis. Laser Surface Interactions for New Materials Production (1). Retrieved from www.spr ingerlink.com. Goldstein, A.N. Echer, C.M. Alivisatos, A.P. (1992) Melting in semiconductor nanocrystals. Science. 256 (5062), 1425 1427. Golovchenko, J. A. (1986) The tunneling microscope: a new lok at the atomic world. Science. 232. 48 53. Griem, H.R (1997) Principles of plasma spectroscopy. Cambridge; New York: Cambridge University Press. Griffin, E.J. (1976) Design of 3 screw tuners. Electronics Letters. 12 (25), 657 658. Guo, B. Harvey, A. Risbud, S.H. Kennedy, I.M. (2006) The formation of c ubic and monoclinic Y 2 O 3 nanoparticles in a gas phase flame process. Philosophical Magazine Letters. 86 (7), 457 467. Guyot, Y. Moncorg, R. Merkle, L.D. Pinto, A. McIntosh, B. Verdun, H. (1996) Luminescence properties of Y 2 O 3 single crystals dop ed with Pr3+ or Tm3+ and codoped with Yb3+, Tb3+ or Ho3+ ions. Optical Materials. 5, 127 136. Hao, B. V. Huy, P. T. Khiem, T. N. Ngan, N. T. T. Duong, P.H. (2009) Synthesis of Y 2 O 3 :Eu3+ micro and nanophosphors by sol gel process. Journal of Physic s: Conferences Series. 187, 1 6.
124 Herlin, M.A. Brown, S.C. (1948) Breakdown of gas at microwave frequencies. Physical Review. 74 (3), 291 296. Hottel, H.C. (1967) Radiative transfer. New York, McGraw Hill. Hullman, A. (2007) Measuring and assessing the d evelopment of nanotechnology. Scientometrics. 70 (3), 739 758. Ibrahim, E.A. Marshall, S.O. (2000) Instability of a liquid jet of parabolic velocity profile. Chemical Engineering Journal. 76, 17 21. Irish, R.T. Bryant, G.H. (1964) Determination of ioni zation rate and electron temperature in a mercury discharge by a microwave method. Proceedings of Physical Society. 84, 975 982. Javey, A. Guo, J. Wang, Q. Lundstrom, M. Dai, H. (2003) Ballistic carbon nanotube field effect transistors. Nature. 424 (7), 654 657. Jerby, E. Golts, A. Shamir Y. Wonde, S. Mitchell, B.A. LeGarrec, J.L. Narayanan, T. Sztucki, M. Ashkenazi, D. Barkay, Z. Eliaz, N. (2009) Nanoparticle plasm elected directly from solid copper localized microwaves. Applied Physics Letters. 95, 191501. Jia, L. Gitzhofer, F. (2010) Induction plasma synthesis of nano structured SOFCs electrolyste using solution and suspension plasma spraying: a comparative study. Journal of Thermal Spray Technology. 19 (3), 566 574. Jollet, F. Noguera, C. Thormat, N. Gautier, M. Duraud, J.P. (1990) Electronic structure of yttrium oxide. Physical Review B. 42 (12), 7587 7595. Kawarada, H. (1996) Hydrogen terminated diamond surfaces and interfaces. Surface Science Reports. 26, 205 259. Keblinski, P. Phillpot, S.R. Choi, S.U.S. Eastman, J.A. (2002) Mechanisms of heat flow in suspensions of nano sized particles (nanofluids). International Journal of Heat and Mass Transfer. 45, 855 863. Kettani, M.A. Hoyaux, M.F. (1973) Plasma Engin eering. New York: Wiley.
125 Kissovski, Z. Brochhaus, A. Korzec, D. Kytzia, S. Engemann, J. (2007) Plasma parameters of an oxygen microwave discharge obtained by probe diagnostics: I. radial distributions. Plasma Sources Science and Technology. 16, 52 3 528. Kissovski, Z. Tsankov, T. Korzec, D. Kytzia, S. Engemann, J. (2006) Plasma parameters of an oxygen microwave discharge obtained by probe diagnostics: I. axial distributions. Plasma Sources Science and Technology. 15, 126 134. Kitamura, N. Sakamoto, T. Matsuura, H. Akatsuka, H. (2005) Dependence of vibrational and rotational temperatures of microwave discharge nitrogen oxygen mixed discharge on their mixture ratio. Bulletin Research Labolatory for Nuclear Reactors.29 (1), 118 120. waveguides with small dimensions. Microwaves and Radar, 1998. MIKON '98., 12th International Conference on. 2, 615 619. Lacas, F. Versaevel, P. Scouflaire, P. Co eur Joly, G. (1994) Design and performance of an ultrasonic atomization system for experimental combustion applications. Particle & Particle Systems Characterization. 11 (2), 166 171. Litchfield, R.B. (1973) Tom Wedgwood, the first photographer. New York: Arno Press. Lu, A Salabas, E.L. Schth, F. (2007) Magnetic nanoparticles: synthesis, protection, fictionalization, and application. Angewandte Chemie International Edition. 46, 1222 1244. Mdler, L. Kammler, H.K. Mueller, R. Pratsinis, S.E. (2002 ) Controlled synthesis of nanostructured particles by flame spray pyrolysis. Journal of Aerosol Science. 33, 369 389. Maslen, E.N. Streltsov, V.A. Ishizawa, N. (1996) A synchrotron x ray study of the electron density in c type rare earth oxides. Acta C rystallographica Section B. B52, 414 422. Mazur, A. Kostka, S. (2004) Loop coupler in rectangular waveguide working near to the cut off frequency. Microwaves, Radar and Wireless Communications, 2004. MIKON 2004. 15th International Conference on. 2, 751 754.
126 Meijernik, A. (2008) Exciton dynamics and energy transfer processes in semiconductor nanocrystals. Semiconductor Nanocrystal Quantum Dots. Retrieved from www.springerlink.com. Melmed, A. J. (1996) Recollection s of Erwin Mllers labolatory: the development of FIM (1951 1956). Applied Surface Science. 94, 17 25. Messing, G.L. Zhang, S. Jayanthi, G.V. (1993) Ceramic powder synthesis by spray pyrolysis. Journal of the American Ceramic Society. 76 (11) 2707 272 6. Mie, G. (1908) Beitrge zur Optik trber Medien, speziell kolloidaler Metallsungen. Annalem der Physik. 25 (3), 377 445. Minko, N. I. Nartsev, V. M. (2008) Sciencefor Glass Production : Nanotechnology in glass materials (review). Glass and Ceramics 65 (5), 148 153. Mizutani, Y. Uga, Y. (1972) An investigation on ultrasonic atomization. Bulletin of the JSME. 15 (83), 620 627. Mooney, J.B. Radding, S.B. (1982) Spray pyrolysis processing. Annual Review of Materials Science. 12, 81 101. Muehe, C.E. (1968) A high power waveguide tuner. IEEE Transactions on Microwave Theory and Techniques. 16 (10), 882 883. Padovani, S. Boriga, I. Brunetti, B. Sgamellotti, A. Giulivi, A. DAcapito, F. Mazzoldi, P. Sada, C. Battaglin, G. (2004) Silver a nd copper nanoclusters in the lustre decoration of Italian Renaissance pottery: and EXAFS study. Allied Physics A. 79, 229 233. Pankhurst, Q.A. Connolly, J. Jones, S.K. Dobson, J. (2003) Applications of magnetic nanoparticles in biomedicine. Journal of Physics D: Allied Physics. 36, R167 R181. Pellerin, S. Cormier, J.M. Richard, F. Musiol, K. Chapelle, J. (1996) A spectroscopic diagnostic method using UV OH band spectrum. Journal of Physics D: Applied Physics. 29, 726 739.
127 Pfender, E. (1999) T hermal plasma technology: where do we stand and where are we going?. Plasma Chemistry and Plasma Processing.19 (1), 1 31. Poland, C.A. Duffin, R. Kinloch, I. Maynard, A. Wallace, W.A. Seaton, A. Donaldson, A. (2008) Carbon nanotubes introduced into the abdominal cavity of mice show asbestos like pathogenicity in a pilot study. Nature: Nanotechnology. 3, 423 428. Rabat, H. Izarra, C. (2004) Check of OH rotational temperature using an interferometric metchod. Journal of Physics D: Applied Physic s. 37, 2371 2375. Reitz, R.D. Bracco, F.V. (1982) Mechanism of atomization of a liquid jet. Physics of Fluids. 25, 1730 1742. Ruska, E. (1987) The development of the electron microscope and of electron microscopy. Reviews of Modern Physics. 59 (3), 627 6 38. Salerno, M. Landoni, P. Vergantti, R. (2008) Designing foresight studies for nanoscienceand nanotechnology (NST) future developments. Technological Forecasting & Social Change.75, 1202 1223. Scattergood, R.O. Koch, C.C. Murty, K.L. Brenner, D (2008) Strenghtening mechanisms in nanocrystalline alloys. Materials Science and Engineering A. 493, 3 11. Schmid et al. (2006) Nanotechnology. Assessment and perspective. Berlin ; New York: Springer. Scott, P.E. (9169) Microwave scattering by turbulent plasma. Journal of Physics A: General Physics. 2 (2), 75 86. Sheen, S. Yang, S. Jun, K. Choi, M. (2009) One step flame method for the synthesis of coated composite nanoparticles. Journal for Nanoparticle Research. 11 (7), 1767 1775. Shiraishi, K. (19 96) First principles calculations of surface adsorption and migration on GaAs surfaces. Thin Solid Films. 272, 345 363.
128 Smalley, R. (2001) Of chemistry, love and nanobots. Scientific American. 10, 76 77. Stauffer, E. (2003) Concept of pyrolysis for fire de bris analysts. Science & Justice. 43, 29 40. Sun, J. Simon, S.L. (2007) The melting behavior of aluminum nanoparticles. Thermochimica Acta. 463, 32 40. Sun, X.L. Tok, A.I.Y. Lim, S.L. Boey, F.Y.C. Kang, C.W. Ng, H.W. (2008) Combustion aided sus pension plasma spraying of Y 2 O 3 nanoparticles: sunthesis and modeling. Journal of Applied Physics. 103, 034308. Toumey, C. (2005) Apostolic succession. Engineering & Science. 1, 16 23. Toumey, C. (2008) Reading Feynman into nanotechnology: a text for a new science. Techn. 12 (3), 133 168. Touzeau, M. Vialle, M. Zellagui, A. Goussett, G. Lefebvre, M. (1991) Spectroscopic temperature measurements in oxygen discharges. Journal of Physics D: Applied Physics. 24, 41 47. Tsuzuki, T. McCormick, P. (2004 ) Mechanochemical synthesis of nanoparticles. Journal of Materials Science. 39, 5143 5146. Walter, P. Welcomme, E. Hallgot, P. Zaluzec, N. J. Deeb C. Castaing J. Veyssire, P. Brniaux, R. Lvque, J., Tsoucaris, G. (2006) Early use of P bS nanotechnology for an ancient hair dyeing formula. Nano Letters. 6 (10), 2215 2219. Wang, L. Pan, Y. Ding, Y. Yang, W. Mao, W.L. Sinogeikin, S. ... Mao, H. (2009) High pressure induced phase transitions of Y 2 O 3 and Y 2 O 3 :Eu 3+ Applied Physic s Letters. 94, 061921. Watanawanyoo, P. Chaitep, S, Hirahara, H. (2009) Development of an air assisted fuel atomizer (liquid siphon type) for a continuous combustor. Americal Journal of Applied Sciences. 6 (3), 380 386.
129 Witanachchi, S. Dedigamuwa, G. Mukherjee, P. (2007) Laser assisted spray pyrolysis process for the growth of TiO2 and Fe2O3 nanoparticle coatings. Journal of Materials Research. 22 (3), 649 654. Yakel, H.L. (1979) A refinement of the crystal structure of monoclinic europium sesquioxide Acta Crystallographica Section B. B35, 654 659. Yang, H. Hu, Y. Zhang, X. Qiu, G. (2004) Mechanochemical synthesis of cobalt nanoparticles. Materials Letters. 58, 387 389. Yoda, T. Miyamoto, S. Tsuboya, H. (2004) Widely tunable CW Nd doped Y 2 O 3 ceramic laser. Conference on Laser and Electro Optics. Zhang, F.X. Lang, M. Wang, J.W. Becker, U. Ewing, R.C. (2008) Structural phase transitions of cubic Gd2O3 at high pressures. Physics Review B. 78, 064114. Zyryanov, S.M. Lopaev, D.V. (2006) Measurements of the gas temperature in an oxygen plasma by spectroscopy of the transition Plasma Physics Reports. 33 (6), 563 574.
131 A PPENDIX A : Electric D iagrams Figure A.1. Electric diagram of nebulizer power control adapter.