USF Libraries
USF Digital Collections

Growth and characterization of thermoelectric ba8ga16ge30 type-i clathrate thin-films d...

MISSING IMAGE

Material Information

Title:
Growth and characterization of thermoelectric ba<sub>8</sub>ga<sub>16</sub>ge<sub>30</sub> type-i clathrate thin-films deposited by pulsed dual-laser ablation
Physical Description:
Book
Language:
English
Creator:
Hyde, Robert Harry
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Hydrodynamic Modeling
Iccd Imaging
In-situ Growth
Pld
Polycrystalline
Power Factor
Dissertations, Academic -- Physics Materials Science -- Doctoral -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: The on-going interest in thermoelectric (TE) materials, in the form of bulk and films, motivates investigation of materials that exhibit low thermal conductivity and good electrical conductivity. Such materials are phonon-glass electron-crystals (PGEC), and the multi-component type-I clathrate Ba<sub>8</sub>Ga<sub>16</sub>Ge<sub>30</sub> is in this category. This work reports the first investigation of Ba<sub>8</sub>Ga<sub>16</sub>Ge<sub>30</sub> films grown by pulsed laser deposition (PLD). This dissertation details the in-situ growth of polycrystalline type-I clathrate Ba<sub>8</sub>Ga<sub>16</sub>Ge<sub>30</sub> thin-films by pulsed laser ablation. Films deposited using conventional laser ablation produced films that contained a high density of particulates and exhibited weak crystallinity. In order to produce high quality, polycrystalline, particulate-free films, a dual-laser ablation process was used that combines the pulses of (UV) KrF excimer and (IR) CO<sub>2</sub> lasers that are temporally synchronized and spatially overlapped on the target surface. The effect of the laser energy on stoichiometric removal of material and morphology of the target has been investigated. In addition, in-situ time-gated emission spectroscopy and imaging techniques were used to monitor expansion of components in the ablated plumes. Through these investigations, the growth parameters were optimized not only to significantly reduce the particulate density but also to produce large area stoichiometric films. Structure and electrical transport properties of the resultant films were also evaluated. This work provides new insight toward the in-situ growth of complex multi-component structures in thin-film form for potential TE applications.
Thesis:
Disseration (Ph.D.)--University of South Florida, 2011.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Robert Harry Hyde.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 207 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - E14-SFE0004896
usfldc handle - e14.4896
System ID:
SFS0028151:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Growth and Characterization of Thermoelectric Ba8Ga16Ge30 Type I Clathrate Thin Films Deposited by Pulsed Dual Laser Ablation by Robert H. Hyde A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of P hilosophy Department of Physics College of Arts and Sciences University of South Florida Co Major Professor: Pritish Mukherjee, Ph.D. Co Major Professor: Sarath Witanachchi, Ph.D. Xiaomei Jiang, Ph.D. Dale Johnson, Ph.D. George S. Nolas, Ph.D. Date of Approval: April 20, 2011 Keywords: hydrodynamic modeling, ICCD imaging, in situ growth, optical emission spectroscopy, PLD, polycrystalline, power factor, Seebeck coefficient, stoichiometric Copyright 2011 Robert H. Hyde

PAGE 2

DEDICATION I de dicate this work to my parents, Mrs. Kay A. Hyde and the late Mr. Harry W. Hyde who inspired and encouraged the completion of this doctorial degree.

PAGE 3

ACKNOWLEDGEMENTS I would like to express my gratitude and appreciation to my advisors, Dr. Sarath Witanachchi and Dr. Pritish Mukherjee, for their support guidance, encouragement, and interest in my academic and personal life throughout the years o f my graduate work. I would like to acknowledge and thank the faculty members serving on my dissertation committee for their time and participation ; Dr. Xiaomei Jiang, Dr. Dale Johnson, and Dr. George S. Nolas. I am thank ful to my colleagues Devajyoti M ukherjee and Marek Merlak for their valuable assistance and perspective during our time in the laboratory. I would also like to acknowledge the other current and past me mbers of LAMSAT; Ted Wangensteen Jason Rejman, Dr. Tara Dhakal, Dr. Gayan Dedigamuwa, Dr. James Winslow, and Dr. Houssam Abou Mourad for the friendship and comradery we have shared, both in and out of the lab. I also thank my friend Diana Nesbitt for her unique contribution to my life and efforts toward fabrication of the new multi techniq ue deposition system I would also like to thank all the past and present Physics Department staff for their invaluable assistance through the years; Mary Ann Prowant, Daisy Matos, Bob Harrington, Phil Bergeron, Kimberly Carter, Evelyn Keeton Williams, and Sue Wolfe. This project was partially supported by the US Department of Energy under grant number DE FG02 04ER46145 and the National Science Foundatio n under grant number DMI 0217939.

PAGE 4

i TABLE OF CONTENTS LIST OF TABLES iii LIST OF FIGURES v LIST OF ABBREVIATIONS xiv ABSTRACT xvii CHAPTER 1: INTRODUCTION 1 1.1. Thermoelectrics 2 1.1.1 Basic Review 2 1.1.2 Current Thermoelectric Materials Research 9 1.1.3 Clathrates 13 1.1.4 ThinFilm Thermoelectric Structures 16 1.1.5 ThinFilm Deposition Techniques 17 1.2. Pulsed Laser Vaporization and Deposition 18 1.3. Summary 25 1.4. Outline of Dissertation 25 CHAPTER 2: DUAL LASER ABLATION 27 2.1. Description of the Pu lsed Dual Laser Ablation System 32 2.1. 1. Dual Laser System Parameters 33 2.1.2 CO2 Laser Parameters 37 2.1.3 Excimer Laser Parameters 38 2.2. Summary 39 CHAPTER 3: INVESTIGATION OF TARGET SURFACE MODIFICATION BY PULSED LASER ABLATION 40 3. 1. B a8Ga16Ge30 Target Properties 44 3. 2. Experimental Details 47 3. 3. Investigation of the Excimer Laser Target Inte ractio n 50 3.4. Investigation of the CO2 Laser Target Interaction 54 3.5. Investigation of the D ual Laser Target Interaction 61 3. 6. Summary 64 C HAPTER 4 : IN SITU OPTICAL DI AGNOSTICS 66 4. 1. Experimental Setup 68 4. 2. ICCD Imaging 75 4.2.1. Dual Laser Synchronization by InSitu Optical Diagnostics 75

PAGE 5

ii 4.2.2. Elemental Expansion Profiles 80 4. 3. Optical Emission Spectroscopy 84 4. 4. Summary 91 CHAPTER 5 : HYDRODYNAMIC MODELING OF PLASMA EXPANSION 93 5.1. Isothermal and Adiabatic Expansion Model 95 5.2. Application of the Theoretical Model 100 5.3. Summary 104 CH APT ER 6: CHARACTERIZATION OF THE Ba8Ga16Ge30 THIN FILM S 106 6. 1. Film Deposition Profile 106 6.2 Thin Film Surface Morphology 110 6.3 Thin Film Crystallinity 116 6. 4. Thin Film Composition 120 6.5 Electric Transport Properties 122 6.5.1. Temperature Dependent Resistivity 123 6.5.2. Van der Pauw Hall Measurement 129 6.5.3 Seebeck Coefficient 131 6.5 .3 .1. Seebeck Appar atus 137 6.5.3 .2. Seebeck Apparatus Characterization 141 6.6 Summary 146 CHAPTER 7 : CONCLUSI ONS 148 REFERENCES 152 APPENDICIES 165 A ppendix A: Wyckoff Positions 166 App endix B: Grit Designation Table 167 Appendix C: Pro perties of Ba8Ga16Ge30 Target Components 168 Appendix D: ICCD Camera System Specifications 169 Appendix E: Spectrometer Specifications 170 Appendix F: Fiber Optic Bundles 172 Appendix G : O ES and ICCD Procedure 174 Appendix H : Hydrodynamic Plasma Model Maple Program 176 Appendix I : PLD Procedure 179 Appendix J : Electron Binding Energies 182 Appendix K : CO2 Output Couplers 183 Appendix L : Publications and Pr esentations 184 ABOUT THE AUTHOR End Page

PAGE 6

iii LIST OF TABLES Table 1. 1. Properties of pulsed laser evaporation and deposition. 19 Table 2.1. Properties of the excimer and CO2 lasers and resulting effect on laser output. Laser properties include the highvoltage of the discharge capacitor (HV), gas mixture ratio, reflection of the CO2 laser o utput coupler (OC), and cavity length of the CO2 laser. 36 Table 2.2. The CO2 laser gas premix used during dual laser ablation. 38 Table 3.1. Crystallographic and transport properties for single crystal type I clathrate Ba8Ga16Ge30 at room temperature. 41 Table 3. 2. Parameters for the KrF excimer and CO2 lasers. 42 Table 3.3. The powder XRD indices for the cold pressed Ba8Ga16Ge30 type I clathrate target. The XRD source is Cu K 46 Table 3. 4. Composition and Ga/Ge atomic ratio of the Ba8Ga16Ge30 type I clathrate cold pressed target determined by EDS analysis. 46 Table 3.5. Laser target interaction site composition by EDS for excimer laser fluences 1 to 5 J/ cm2 after 1000 pulses. The intrinsic and un abla ted target values are listed also. The Ga/Ge ratio is calculated for each. 53 Table 3. 6. Laser target interaction site composition by EDS for CO2 laser fluences 0.6 to 2.4 J/ cm2 after 1000 pulses. The intrinsic balanced and unablated target values are li sted also. The Ga/Ge ratio is calculated for each. 56 Table 3.7. The actual composition and Ga/Ge ratio of the un ablated target and laser target interaction sites after 1000 pulses per site for dual laser 1 J/cm2 UV coupled with 2 J/cm2 IR with 100 110 ns pp delay. The results for excimer only single laser 3 J/cm2 (S3) and 1 J/cm2 (S1) are included for comparison 64

PAGE 7

iv Table 4 1. The full width at half maximum (FWHM) of the OES cross section profile of ablated plumes 5 mm from the target for (a) 1 J/cm2 and (b) 3 J/cm2 single laser, and (c) dual laser ablation. ( 1.5 degree) 83 Table 5.1. Velocity, coefficient gamma, plasma thickness, and plasma temperature. for 1 J/ cm2 and (b) 3 J/ cm2 single laser and (c) dual laser. 105 Table 6.1. The n factor of th e cosn fit for the y and z axis thickness profiles for single and dual laser depositions at a target substrate distance of 4 cm 110 Table 6.2. Stoichiometry of thin films deposited by single laser (SL) and dual laser (DL) ablation. 121 Table 6.3 Barium, ga llium, and germanium composition of thin films deposited a various dual laser peak to peak delays of 175 ns, 100 ns and 25 ns. The substrate temperature during deposition was room temperature. The thickness of the resultant films is given. 122 Table 6.4. The room temperature electric transport properties of various Ag doped Ba8Ga16Ge30 thin films deposited on quartz substrates deposited by single laser (SL) and dual laser (DL) ablation. 128 Table A.1. The Wyckoff positions for the c rystallographic cubic space group Pm 3n applicable to the type I clathrate Ba8Ga16Ge30 [102]. 166 Table B.1. ISO grit designatio n and average grit diameter. 167 Table C.1. Properties of the target components; bar ium, gallium, and germanium. 168 Table J .1. The electron bindi ng energy for oxygen (O), silicon (SI), gallium (Ga), germanium (Ge), stro ntium (Sr), and barium (Ba). 182 Table K.1. The output energy and delay for the CO2 laser germanium output couplers. 183

PAGE 8

v LIST OF FIGURES Figure 1. 1. The Seebeck effect for an isolated conductor in a uniform temperature gradient ( ). The Fermi Dirac diagrams indicate the energy states for a hot and a cold temperature. The higher energy carriers diffuse from the hot sid e to the cold side generating a voltage potential ( ). 2 Figure 1. 2. Schematic diagram of the open circuit Seebeck effect for two dissimilar conductors a and b with junctions a positions 1 and 2, held at temperatures T1 > T2, respectively. 3 Figure 1.3. Arrangement of thermoelectric modules for (a) power generation by the Seebeck effect and (b) refrigeration by the Peltier effect. 4 Figure 1.4. The energy band diagrams for ideal ohmic metal semiconductor junctions at thermal equilibrium for a TE refrige ration device. The thermal energy absorption and emission as the carriers (electrons and holes) cross the junctions between the semiconductor material and metal contacts. 5 Figure 1.5. The f igure of merit ( ZT) of state of the art commercial materials f or thermoelectric power generation, ( a ) ntype and ( b ) ptype. 10 Figure 1.6. A c omparison of the Seebeck coefficient ( ), power factor ( 2 ), and conductivity ( ) as a function of the carrier concentration (10x) 11 Figure 1.7. The type I clathrate composed of the (a) facesharing pentagonal dodecahedron and tetrakaidecahedron, and the (b) cubic unit cell with the polyhedra outlined in red. The Wyckoff positions of the guest and framework atoms are indicated. 14 Figure 1. 8. Components of the type I clathra te structure with (a) guest atoms on the corners, center, and faces of the cubic unit cell, and the host atom (b) dodecahedron cages (9), (c) the tetrakaidecahedron cages (12), and (d) the complete host guest schematic. 15

PAGE 9

vi Figure 1.9. Diagram of the pulse d laser deposition process depicting the pulsed beam (i) vaporization of the target surface, (ii) transport of the vapor plume, and (iii) growth of the film on the substrate. 18 Figure 1. 10. A diagram of the excimer laser energy (i) absorption by the targ et material and initiation of ablation and (ii) continued target ablation and fractional plasma absorption of the laser pulse. 21 Figure 1.11. The evolution of the laser target interaction. The (a) The unablated target, (b) laser pulse arrival, rapid targ et surface heating, and plasma initiation, (c) pulse continuation, pulse absorption into the expanding plasma, and increased surface heating and propagation into the target, (d) pulse termination, hydrodynamic plasma expansion, and rapid target cooling, an d (e) ideal return of the target to the initial state. 22 Figure 2.1 A diagram of the pulsed dual laser ablation system. The system can be operated in either excimer only sin gle laser or dual laser mode. 28 Figure 2. 2. Oscilloscope traces of the tempora l synchronization of the (UV) excimer laser pulse and the (IR) CO2 laser pulse. The pp delay (peak to peak) is shown relative to the peak of the excimer laser pulse. 29 Figure 2.3. Evolution of the laser target interaction for dual laser ablation. (a) CO2 laser pulse arrival and on set of surface melt, (b) synchronized arrival of the excimer laser pulse to interact with the premelt zone, initiate evaporation and plasma formation, and dual laser plasma absorption, and (c) termination of the excimer pulse, and continued plasma absorption of the CO2 laser pulse, increasing plas ma temperature and expansion. 30 Figure 2.4. The time line of component synchronization for the dual laser ablation system. The t0 trigger to initiate the dual laser pulse cycle is th e (a) CO2 laser internal trigger, followed by the (b) CO2 laser capacitor discharge, which triggers the (c) digital delay generator (DDG) t0 signal to trigger the excimer laser, followed by the (d) excimer laser output pulse, and (e) the CO2 laser output pulse. 34 Figure 2.5. Detail of the cumulative oscilloscope traces of the excimer and CO2 laser pulses to illustrate the 16 ns jitter of the CO2 laser pulses relative to the excimer laser pulses. The jitter is measured between the minimum and maximum occur rence of the full width at half maximu m (FWHM) of the laser pulses. 35

PAGE 10

vii Figure 2.6. Schematic diagram of the CO2 laser with the ZnSe Brew ster window extension tube and Ge output coupler. 37 Figure 2.7. Output characteristics of the excimer laser after a n ew fill, the (a) output delay and (b) pulse jitter relative to the trigger pulse, as a function of the output energy. 39 Figure 3.1. The normalized, indexed, powder XRD pattern for the cold pressed Ba8Ga16Ge30 type I clathrate target. 45 Figure 3. 2. SEM image of the coarse polished (529 nm rms) Ba8Ga16Ge30 type I clathrate cold pressed target. Approximately 315315 m viewing area. 47 Figure 3.3. Diagram of the laser spot size at the laser target interaction site. 48 Figure 3. 4. Photograph of the Ba8Ga1 6Ge30 target with various laser interaction sites at fluences of 1 to 5 J/cm2 and repetitions of 1, 10, 100, and 1000 pulses. 49 Figure 3. 5. Diagram of SEM and EDS analysis locations for the laser target interaction study on the Ba8Ga16Ge30 target. 50 Fi gure 3.6a. SEM images of the UV KrF excimer only, laser target interaction sites for fluences of 1, 2, and 3 J/ cm2, and cumulative laser pulses per site of 10, 100, and 1000. 51 Figure 3.6b. SEM images of the UV KrF excimer only, laser target interaction sites for fluences of 4 and 5 J/ cm2, and cumulative laser pulses p er site of 10, 100, and 1000. 52 Figure 3.7. SEM images of the IR CO2 laser only, laser target interaction sites at 1000 pulses per site for fluences of 0.6 J/ cm2, 1.3 J/ cm2, 1.8 J/ cm2, 2.0 J/ cm2, and 2.4 J/ cm2. 55 Figure 3.8. The schematic diagram of the experiment setup for the single beam transie nt reflectivity measurements. 57 Figure 3. 9. The monitor pulse before target interaction and reflection pulse after target interaction. The 0.6 J/ cm2 CO2 laser fluence was used for alignment and calibration. 58 Figure 3. 10. Representation of IR reflection pulse showing features: on set of melt, on set of ablation, and peak of the monitor pulse. 58

PAGE 11

viii Figure 3.11. Single beam reflectivity for CO2 l aser fluences of (a) 1.8 J/ cm2, (b) 2.0 J/ cm2, and (c) 2.4 J/ cm2, before (Monitor) and after Reflection from the Ba8Ga16Ge30 targe t, displaying on set of melt. 60 Figure 3.12. The monitor oscilloscope traces of the UV excimer laser pulse and the IR CO2 la ser pulse with a 100 110 ns pp delay. The on set of the melt is indicated at m. 61 Figure 3.13. Images of the laser target interaction site s for 10, 100, and 1000 pulses/site using dual laser f luences of 1 J/cm2 UV and 2 J/cm2 IR with 100 110 ns pp delay 62 Figure 3.14. SEM i mages of dual laser ablation of an unpolished target. A region of un ablated target surface is shown and the laser target interaction site s for 10, 100, and 1000 pulses/site using dual laser f luences of 1 J/cm2 UV and 2 J/cm2 I R with 100 ns pp delay 63 Figure 4.1. Diagram of the imaging system identifying the coordinate system, object plane, image plane, and an example light collecting device (optical fiber). The magnification is for this configuration. 68 Figure 4.2. The schematic diagram s of the dual laser deposition system with the (a) optical emission spectroscopy (OES), and (b) intensified chargecoupled device (ICCD) imaging system included. 69 Figure 4.3. Specification of the wavelength sensitivity of the ICCD came ra. The wavelength cut off due to the imaging window is shown at 350 nm. 70 Figure 4. 4. Photographs of the alignment and calibration procedure using a 4 cm long helium lamp as the object (not shown) placed along the plume axis of propagation and imaged at the light collection plane. Photograph (a) is an example of proper object/lens alignment and (b) is misaligned displaying distortion of the image. 72 Figure 4.5. Schematic diagram of the imaging system alignment and calibration layout using a helium lamp masked for 4 cm length. The diagram is drawn to scale. 73 Figure 4. 6. Intensity of the He lamp (587.6 nm wavelength) as the optical fiber position was varied along the x axis (horizontal) and y axis (vertical) of the image plane. 74 Figure 4.7. The (a) back lit scale card used to determine image magnification and orientation for the optical diagnostics system(s), and (b) a photograph of the imaged card. 74

PAGE 12

ix Figure 4.8. Time integrated ICCD images of CO2 laser target interactions with laser fluences of (a) 1.8 J/ cm2, (b) 2.0 J/ cm2, and (c) 2.4 J/ cm2. Images are normalized with respect to image (c). 76 Figure 4.9. The normalized time resolved total emission intensity of the CO2 laser target interactions for 1.8 J/ cm2, 2.0 J/ cm2, and 2.4 J/ cm2. The emission was imaged at the target surface (0 mm position). 76 Figure 4. 10. ICCD imaging of total emission of ablated plumes from the target for (a) 1 J/cm2 and (b) dual laser ablation. 77 Figure 4.11. Overlapped oscilloscope traces for the excimer laser pulse (UV ) and CO2 laser pulses (IR) for peak to peak delays of 25 ns, +100 ns, and +175 ns, relative to the UV pulse at 0 ns. 78 Figure 4.12. Dual laser synchronization parameters as a function of the dual laser pulse peak to peak delays obtained by total emissi on ICCD imaging of the laser induced plasma expansion. The (a) normalized peak intensity and (b) normalized FWHM are shown. 79 Figure 4.13. ICCD imaging of the time integrated laser generated plume emission for excimer only 1 J/ cm2 (S1), high fluence sing le laser 3 J/ cm2 (S3), and dual laser ablation. For each laser configuration the unfiltered total emission, Ba+ 455 nm, Ga+ 426 nm, and Ge+ 481 nm are shown. 81 Figure 4.14. OES cross section of ablated plumes 5 mm from the target for (a) 1 J/cm2 and (b) 3 J/cm2 single laser, and (c) dual laser ablation. The emission shown is for Ba+ 455 nm, Ga+ 426 nm and Ge+ 481 nm. 82 Figure 4.15. The time integrated spectra for (a) singlelaser 1 J/ cm2 fluence and (b) dual lasers ablated plumes. The neutral (I) and si ngly ionized (II) species of Ba, Ga, and Ge are indicated. 85 Figure 4.16. The total intensity of the time integrated optical emission obtained 2 cm from the target surface for cumulative laser pulse for single and dual laser ablation. 86 Figure 4.17. An example of a time of flight (TOF) profile positioned at 2 mm from the target surface. 88

PAGE 13

x Figure 4.18. The normalized intensity, time resolved, total emission time of flight profiles for positions 0 mm, 1 mm, 2 mm, 4 mm, and 6 mm from the target surface. The profiles were imaged along the plume axis for single laser (a) 1 J/ cm2, and (b) 3 J/ cm2 and (c) dual laser ablation configurations. The dual laser pp delay was 100 ns. 89 Figure 4.19. The distance time plot of the TOF peak arrival time at various d istances from the target surface, for singlelaser 1 J/ cm2 and 3 J/ cm2, and dual laser 1 J/ cm2 UV + 2 J/ cm2 IR. The slopes were velocity determination. 90 Figure 4.20. The temporal FWHM of the TOF profiles for single laser 1 J/ cm2 and 3 J/ cm2 and for th e dual laser 1 J/ cm2 UV + 2 J/ cm2 IR. 90 Figure 4.21. The time resolved elemental TOF profiles for the ions Ba+ 455 nm, Ga+ 426 nm, and Ge+ 481 nm at 4 mm from the target surface. 91 Figure 5.1. Schematic diagrams showing the three regimes during the pul sed laser irradiation of a bulk target: the (a) evaporation regime (0 < t < 100s ps), (b) isothermal regime (100s ps < t ), and (c) the adiabatic regime ( t > ). 94 Figure 5. 2. Schematic diagram of the (a) laser spot size at the target surface and (b) the plasma volume with the orthogonal edges (Xo,Yo, Zo) indicated. 97 Figure 5. 3. TOF profiles and theoretical model fit for single laser ablation (a) 1 J/ cm2, (b) 3 J/ cm2, and (c) dual laser 1 J/ cm2 UV + 2 J/ cm2 IR. 103 Figure 6. 1. The dependence of th e on axis film deposition rate per laser pulse on the excimer laser fluence. The substrate temperature was room temperature. 107 Figure 6. 2. Profile film thickness along the (a) y axis and (b) z axis for Ba8Ga16Ge30 deposited on quartz substrates. The tar get to substrate distance was 4 cm. The solid lines indicate the theoretical fit cosn 108 Figure 6.3. A SEM at 3000 times magnification which shows examples of the various types of particulates found on the thin films. 109 Figure 6. 4. SEM at 300 times magnification at (a) 1 J/ cm2 (b) 2 J/ cm2 (c) 3 J/ cm2 (d) 4 J/ cm2 and (e) 5 J/ cm2. 111 Figure 6.5. SEM images of thin films deposited by single laser PLD with fluences of (a) 1 J/cm2 and (b) 3 J/cm2, and dual laser fluences of (c) 1 J/cm2 UV and 2 J/cm2 IR 112

PAGE 14

xi Figure 6. 6. Overlapped oscilloscope traces for the excimer laser pulse (UV) and CO2 laser pulses (IR) for peak to peak delays of 25 ns, +100 ns, and +175 ns, relative to the UV pulse at 0 ns. 113 Figure 6.7. Scanning electron micrographs of films d eposited at various dual laser peak to peak delays with 1 J/cm2 UV and 2 J/cm2 IR laser fluences. The CO2 laser pulse peak is delayed by (a) 25 ns, (b) 100 ns, and (c) 175 ns, relative to the excimer laser pulse peak. The film deposited at 100 ns delay re sulted in reduced particulate generation. 115 Figure 6.8. The XRD dependence on the substrate temperature. The excimer laser fluence was 1 J/ cm2. 116 Figure 6. 9. XRD intensity of the peak 31.2 (321) as a function of temperature. 117 Figure 6. 10. The XR D patterns dependence on the excimer laser fluence. The substrate temperature was 400 C. 118 Figure 6.11. XRD intensity of the peak 31.2 (321) as a function o f the excimer laser fluence. 118 Figure 6.12. X ray diffraction patterns for single laser 1 J/ cm2 UV (SL1) and 3 J/cm2 UV (SL3), and dual laser (DL1+2) 1J/cm2 UV + 2 J/cm2 IR. The film thickness was 267 30 nm. The type I clathrate XRD peaks are identified to a cubic unit cell structure with space group Pm 3n 119 Figure 6.13. EDS quantification of the films deposited as a function o f the excimer laser fluence. 120 Figure 6.14. The temperature dependent resistivity of a Ba8Ga16Ge30 thin film sample at various applied injection currents, demonstrating the semiconducting behavior. 123 Figure 6.15. Temperature dependent resistivity of the Ba8Ga16Ge30 thin films deposited by single and dual laser ablation for various thicknesses. 125 Figure 6.16. A scanning electron micrograph image of a film deposited by single laser PLD with a fluence of 3 J/cm2. The average film thickness is 623 nm. The micro cracks are visible on the film. 125

PAGE 15

xii Figure 6.17a. Schematic diagram of progressive film nucleation for single laser deposition, (a) sparse nucleation, (b) increased coalescence growth, (c) final, discontinuo us crystal growth within an amorphous fill and or channels and holes, and (d) an SEM image of a Ba8Ga16Ge30 film grown with 3 J/ cm2 single laser ablation. Note the 50 150 nm crystallites encased within the film growth. 127 Figure 6.17b. Schematic diagra m of progressive film nucleation for dual laser deposition, (a) high energy, ion bombardment promoting increased nucleation, (b) increased, dense crystal growth, (c) final, uniform continuous growth, and (d) an SEM image of a Ba8Ga16Ge30 film grown from 1 J/ cm2 UV and 2 J/ cm2 IR dual laser ablation. Note the smooth film growth as compared to the image in Figure 6.20a. 127 Figure 6.18. The schematic diagram of the Van der Pauw technique for determination of resistivity and Hall values for thin film samples. 125 Figure 6.19. A s chematic diagram of a basic thermocouple arrangement ( ab) between hot (TH) and cold (TL) sources 132 Figure 6.20. Schematic arrangement of thermocouples on the sample surface for determination of the thermopower by the differentia l method. 135 Figure 6.21. A schematic wiring diagram of the apparatus used to measure the temperature gradient and Seebeck voltage, used to determine the Seebeck coefficient. 140 Figure 6.22. Diagram of the Seebeck coefficie nt measurement apparatus. 141 Figure 6.23. Photographs of the (a) Seebeck apparatus and DAQ board and (b) detail of the sample and thermocouple placement. 141 Figure 6.24. Arrangement of the T type thermocouples on the surface of the hater block for char acterization of the thermocouples. 142 Figure 6.25. Time required for the heater block temperature to stabilize with 15 V AC applied to the heater. 142 Figure 6.26. The stabilized temperature recorded by the thermocouples at various heater voltages. 143 Figure 6.27. The voltage potential as a function of the temperature difference at various average temperatures. 145 Figure 6.28. The thermopower ( b) of the Cu lead of the thermocouples used in the Seebeck apparatus. 145

PAGE 16

xiii Figure E.1. Spectrograph optical layout. 171 Figure F.1. 10 mm diameter ferrule on spectrograph end for (a) two leg fiber bundle and (b) four leg fiber bundle. 172 Figure F.2. The three fiber configuration on the source end of the four leg fiber bundle. An approximate 528 m diameter fi ber bundle zone (including cladding), an approximate 483 m diameter light collection zone (not includ ing cladding at outer edge). 173 Figure I .1. Substrate heater cool down time from 300 C to room temperature with 1 V AC reduction every 2.5 minutes until 0 V AC is reached. 181

PAGE 17

xiv LIST OF ABBREVIATIONS angstrom ArF argon fluorine B a8G a16G e30 Barium Gallium Germanium C heat capacity CA clear aperture CCD charge coupled device CJC cold junction compensation CO2 carbon dioxide CWL center wavelength DC direct current T change in temperature V change in voltage e electron charge EF Fermi energy f Ladenurgs oscillator strength v1 laser frequency vp plasma resonant frequency g upper level statistical weight Hz hertz IR infrared

PAGE 18

xv ICCD intensified cha rge coupled device I current J Joule K Kelvin KrF krypton fluorine kB Boltzmann constant thermal conductivity wavelength MOCVD metal organic chemical vapor deposition micrometer microvolt mobility mJ millijoules ne e lectron number density nm nanometer ns nanosecond OES Optical Emission Spectroscopy OMA optical multi channel analyzer n(E) density of states PGEC phonon glass electron crystal 2 Power Factor PLD Pulsed Laser Deposition q carrier charge

PAGE 19

xvi s second S Seebeck coefficient SiO2 silicon dioxide electrical conductivity T t emperature TH high temperature TL low temperature T/C thermal couple TE thermoelectric TISR LIPS time integrate spaceresolved laser induced plasma spectro scopy TEA transversely excited atmospheric pressure UV ultra violet V voltage VH high voltage VL low voltage XeCl xenon chloride YSZ yttrium stabilized zirconium ZT figure of merit

PAGE 20

xvii ABSTRACT The ongoing interest in thermoelectric ( TE ) materials in the form of bulk and films, motivates investigation of materials that exhibit low thermal conductivity and good electrical conductivity. Such materials are phononglass electron crys tals (PGEC) and the multicomponent type I clathrate Ba8Ga16Ge30 is in this category. This work reports the first investigation of Ba8Ga16Ge30 films grown by pulsed l aser deposition (PLD) This dissertation details the in situ growth of poly crystalline ty pe I clathrate Ba8Ga16Ge30 thin films by pulsed laser ablation Films deposited using conventional laser ablati on produced films that contained a high density of particulates and exhibited weak crystallinity In order to produce high quality, polycrystalli ne, particulatefree films a dual laser ablation process was used that combin es the pulses of (UV) KrF excimer and (IR) CO2 laser s that are temporally synchronized and spatially overlapped on the target surface. The effect of the laser energy on stoichiom etric removal of material and morphology of the target has been investigated. In addition, in situ time gated em ission spectroscopy and imaging techniques were used to monitor expansion o f components in the ablated plume s Through these investigations the growth parameters were optimized not only to significantly reduce the particulate density but also to produce large area stoichiometric films S tructure and electrical transport properties of the resultant films were also evaluated This work provides new insight toward the in situ growth of com plex multi component structures in thin film form for potential TE applications.

PAGE 21

1 CHAPTER 1: INTRODUCTION The increasing need for energy production and conservation has motivated interest in technologies that address energy harvesting [ 1]. Waste heat recovery through the application of thermoelectric materials ( TE ) can play a majo r role in this area [ 25]. The impact that TE materials will have on the energy crisis is a matter of debate [ 6], however, niche military, space [ 7] and commercial products have show n promise over the past decade. Small scale thermo generators for low pow er consumption in high volume commercial applications are the most promising [ 8] Vehicle exhaust heat recovery has considerable potential benefit and along with a U.S. Department of Energy (DOE) goal of ~10% improved fuel economy is stimulating on going r esearch [ 79] T E materials have found applications as self powered monitors, transponders, and temperature warning systems [ 9]. Other viable applications include localized active cooling or heating of devices such as micro processors, laser diodes, detect ors, and biological specimens [ 1012]. However, t he main shortcoming with current thermoelectric materials is their poor efficiency [ 3 ]. From the experimental point of view, one of the motivations is to have the ability to grow TE materials in thin film fo rm for devices with improved efficiencies.

PAGE 22

2 This chapter describes the physical properties that contribute to the TE conversion efficiency, and material systems an d structures used in TE devices. C omm only used film fabrication techniques will be discussed, as well as the PLD that was utilized 1.1. Thermoelectrics A basic review of thermoelectrics ( TE ) is presented as well as the current focus for improving TE materials being researched. The novel materials known as clathrates will be discussed, and thin film TE structures and their deposition techniques will be reviewed. 1.1.1 Basic Review In a conductor or semiconductor the free carriers (electrons and holes) carry both charge and heat and can be induced to diffuse through the material (Figure 1.1) [ 13, 14 ]. The diffusing carriers are scattered by impurities, imperfections, and lattice vibrations (phonons) and these hot and cold carriers diffuse at different rates The Fermi Dirac distribution for a hot and cold temperature is depicted in Figure 1.1 [ 15, 16 ]. Within the F(0) f ( ) 0 1 T=0 F(0) f ( ) 0 1 T=0 Hot Cold + + + + V Figure 1.1. The Seebeck effect for an isolated conductor in a uniform temperature gradient ( ). The Fermi Dirac diagrams indicate the energy states for a hot and a cold temperature. The higher energy carriers diffuse from the hot side to the cold side generating a voltage potential ( ).

PAGE 23

3 temperature gradient ( ) the average energy per electron in an isolated conductor is greater in the hot end. Consequently, the more energetic electrons diffuse toward the cold region until a potential difference ( ) is built up which prevents further diffusion. The basis of this process is called the Seebeck or thermoelectric effect, discovered in 1821. Thomas Seebeck observed when two dissimilar conductors, a and b, are joined (Figure 1.2) and the junctions are held at two different temperatures, T1 > T2, a voltage differenc e ( ) develops that is proportional to the temperature difference ( ). The and are related by an intrinsic property of the materials called the Seebeck coefficient or thermopower ( ) where ab is defined as the differential Seebeck coefficient between materials a and b such that: T Vab (1.1) for small as 0 T A reversal of this process is known as the Peltier effect, discovered in 1834. It was observed when an electric current flows through the junction of two dissimilar metals, heat is exchanged [ 17, 14]. If a current source is applied across the points 3 and 4 (Figure 1.2), then heat is either absorbed or rejected depending on the direction of the a b b V 3 4 Cold (T 2 ) 2 Hot (T 1 ) 1 Figure 1.2. Sch ematic diagram of the open circuit Seebeck effect for two dissimilar conductors a and b with junctions a t positions 1 and 2, held at temperatures T1 > T2 respectively.

PAGE 24

4 current. The Seebeck coefficient ( ) and the Peltier ef fect are related to each other in the definition of the Peltier coefficient ( ) where: Tab (1.2) where T is the temperature of the junction. The rate of heating ( Q ) or cooling is given by: IT Qab (1.3) where I is the current through the junction. The Seebeck effect and Peltier effect are the foundation for solid state TE devices that can be used to directly convert heat into electricity for power generation, or for refrigeration, depending the configuration (Figure 1.3) [2, 3 ]. A TE couple, is made up of two types of semiconductor materials, an ntype (electron majority carriers) and a ptype (hole majority carriers). The two branches of the device are connected by a metallic contact to form an electrical series, but thermally parallel circuit [ 13, 14]. The Seebeck effect is temperaturegradient driven; a heat source applied to one side will drive the Figure 1.3 Arrangement of thermoelectric modules for (a) power generation by the Seebeck effect and (b) refrigeration by the Peltier effect. (a) (b) I p type n type Heat Rejection (T H ) Cooling (T C ) + + R L I Heat Sink (T C ) p type n type H eat Input (T H ) V +

PAGE 25

5 majority carriers toward the cooler region (Figure 1.3a). This type of TE configuration will produce a voltage that can be utilized drive a current through a load resistance ( RL) for power generation. The Peltier effect is current driven; an electric current applied to the material system diffuses the hot carriers through the circuit (Figure 1.3b). The electrons in ntyp e material will flow against the direction of the current and holes in ptype material will flow with the current direction. Since ideal ohmic metal/semiconductor contacts at thermal equilibrium share Fermi levels ( Ef), heat will be absorbed or emitted as carriers (electrons and holes) cross the junctions [ 18, 19]. For example, as current passes through the metal/ ntype semiconductor junction an electron acquires energy (cooling) as it enters the conduction band. Energy is released (heating) as an electron passes through an ntype semiconductor/metal junction (Figure 1.4). Thus, this TE configuration provides Figure 1.4 The energy band diagrams for ideal ohmic metal semiconductor junctions at thermal equilibrium for a TE refrigeration device. The thermal energy absorption and emission as the carriers (electrons and holes) cross the junctions between the semiconductor material and metal contacts. p-type n type Heat Re leased (T H ) Cooling (T C ) + E v Ec E f Metal p type p type Metal Ev E c E f E v E c E f Metal n type E f E v E c Metal n -type + Heat Re leased Heat Re leased Heat Absorbed Heat Absorbed Heat Absorbed

PAGE 26

6 refrigeration capability with no moving parts. A TE couple is consists of TE material segments a and b, which are ptype and ntype respectively. The efficiency ( ) of a TE power generation couple is measured as the ratio of electrical power ( PL) delivered through the load resistance ( RL), relative to the heat flow ( QH) into the hot side of the TE couple. The electrical power is given by PL = I2RL where I is the cu rrent through RL. H LQ P (1.4) The QH input is comprised of three terms. The heat flow through the TE couple due to the thermal conductance ( ); the heat absorbed at the hot junction due to the Peltier coefficient ( Q = abI TH), equatio n (1.3); and the heat due to Joule heating ( I2R ) of the TE couple with the assumption that it is divided between the hot and cold sides. K is the total thermal conductance of materials a and b in parallel, and R is the series resistance of a and b. Theref ore, QH is given by: R I I T T K QH ab H 2 2 1 (1.5) then the power generation efficiency ( ) is given by: R I IT T K IRH ab L 2 2 1 (1.6) where H TC, and ab p n where p and n are the Seebeck coefficients of the ptype and ntype materials, respectively. The efficiency () of TE refrigeration is measured by the coefficient of performance (COP), defined as the ratio of the heat absorbed ( QC) to the electrical power input. The rate of the heat absorbed is given by:

PAGE 27

7 T K R I I T QC ab C 22 1 (1.7) The power input ( P ) is given by: R I T I Pab 2 (1.8) where the applied volt age to the TE couple is balanced between the resistance ( R ) of the TE segments and the Seebeck voltage due to the between the junctions. The efficiency (C) of refrigeration is then given by: R I T I T K R I I T P Qn p C n p C C C 2 22 1 (1.9) The efficiency (H) of the heat released by the hot side for TE heating is obtained by substitution of TH for TC in equation (1.9). The perform ance of a TE couple can be evaluated at a specified temperature by the figure of merit ( Z ) where: RK Zab 2 (1.10) If the geometries of the segments a and b are matched to minimize heat absorption, then the segment cross sectional area ( A ) a nd length ( L ), can optimize the Z of a TE couple. An optimization of Z occurs if the product RK is minimized, such that when: 2 1 p n n p n p p nA L A L (1.11) The figure of merit for the couple can then be rewritten in terms of the two materials, an ntype and a ptype where:

PAGE 28

8 2 1 2 1 2 n n p p n pZ (1.12) or as the dimensionless figure of merit ( ZT): 2 / 1 2 / 1 2) (n n p p n pT ZT (1.13) The performance of an individual TE material segment relies directly on the ZT [ 14]. The parameter ZT is then given by: T ZT2 (1.14) where is the Seebeck coefficient, is the electrical conductivity, T is the temperature of interest, and is the total thermal conductivity. The is approximately described as: e L (1.15) where L is the lattice thermal conductivity and is the heat conducted by phonons (lattice vibrations). The electronic thermal conductivity ( e) is the heat conducted by mobile electrons. Since the e for TE materials is proportional to via the Wiedemann Franz law [ 5] given by: LTe (1.16) where L is the Lorentz number (2.45 108 V2K2 for free electrons). The product 2 is referred to as the electrical power factor and is a value commonly reported as an indicator of TE performance. The early application of the TE effect was met al thermocouples which are still widely used today. During the 1950s and 1960s extensive research was conducted on

PAGE 29

9 materials such as alloys of Bi2Te3 and Si1 -xGex for thermoelectric applications. These materials remain state of the art for their specific t emperature use. In the 1990s, resurgence in the field began with advances in novel processes for producing bulk materials and progress in thin film growth techniques [ 11]. 1.1.2 Current Thermoelectric Materials Research At room temperature, T = 300 K, de sired values for the thermoelectric parameters are = 225 V/K [ 20], = 105 1m1, and = 1.5 W/m K, which results in a ZT 1. These values are typical for the best TE materials such as Bi2Te3 and Sb2Te3 alloys, which are presently used by industry in devices that operate near room temperature and are well investigated [ 2, 21]. Current TE devices operate at an efficiency of about 5 6 %. By increasing ZT by a factor of 4 predicted efficiencies can increase to 30 % [3]. A comparison of various ntype and ptype TE materials is shown in Figure 1.5 where the ZT peaks for each material such that each material has an optimal operating temperature range [ 5]. CsBi4Te6 ( ptype) has shown promise f or low temperature applications ( 100 200 K ) [ 22, 23]. Mid temper ature range (500 900 K) power generation materials have been based on tellurides like LAST (Lead Antimony Silver Telluride, ntype ) and TAGS (Te Ag Ge Sb, ptype ) such as (AgSbTe2)0.15(GeTe)0.85. TAGS has been shown to be more efficient than PbTe but is ve ry expensive. High temperature TE generators (>900K) have used Si Ge alloys ( ntype and ptypes) as well as La2Te3 ( ntype ) and Yb14MnSb11 ( ptype ). Lower thermal conductivity has been achieved wit Zn4Sb3 (0.9 W/m K) [24], ntype PbTe (0.15 W/m K) [24], and inhomogeneous nano structured materials such as SALT (Sodium Antimony

PAGE 30

10 Lead Telluride) and LASTT (Lead Antimony Sodium Tin Telluride) [3, 5]. Current research i s aimed at increasing the ZT by maximizing the power factor ( 2 ) and/or minimizing the thermal conductivity ( ), or finding materials capable of operating in new or broader temperature ranges [ 20 ]. A maximized power factor implies that a high voltage and current are generated, and a low allows a larger to be created across the material [ 25]. Increasing the power factor ( 2 ) involves increasing and together. In reality, it is difficult to enhance without lowering Since is the energy tran sported per carrier q C (1.17) where C is the specific heat and q is the charge of the carrier, a low gives rise to high Figure 1.5 The f igure of merit ( ZT ) of state of the art commercial materials f or thermoelectric power generation, ( a ) n type and ( b) ptype [3, 5, 2225] (a) (b) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 100 300 500 700 900 1100 1300 Temperature (K) ZT Bi2Te3 SiGe CoSb3 LAST PbTe La2Te3 n -type 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 100 300 500 700 900 1100 1300 Temperature (K) ZT Yb14MnSb11 SiGe Zn4Sb3 Sb2Te3 PbTe TAGS CsBi4Te6 SALT p -type

PAGE 31

11 The and can be adjusted through doping that changes the free carrier concentration. Since these two parameters typically change in opposite directions, a compromised set of values needs to be achieved such that there is a peak power factor through optimization of the carrier concentration ( n) [5]. Depending on the material system, the optimum carrier conce ntration typically occurs at 1019 to 1021 cm3, a concentration found in heavily doped semiconductors (Figure 1.6). To effectively minimize minority carrier contributions, leading high mobility carriers, the direct band gap needs to be large enough, on the order of kBT where kB is the Boltzmann constant (1.381 1023 J/K) [ 25]. For semiconductors, the carriers must be thermally excited across the band gap for conduction to occur, and the temperaturedependence of the electrical conductivity is appro ximated by; T k E ob ge (1.18) 0 200 400 600 800 1000 1200 1400 17 18 19 20 21 22 Carrier Concentration (log) Seebeck Coefficient (V/K) 0 2000 4000 6000 8000 10000 Seebeck Power Factor 22 Conductivity 1/ Semiconductor Heavily Doped Semiconductor Metal Figure 1.6 A c omparison of the Seebeck coefficient ( ), p ower factor ( 2 ), and conductivity ( ) as a function of the carrier concentration (10x) [5]

PAGE 32

12 There are two primary ways to achieve a high conductivity, by having a very small band gap to excite across ( Eg kbT 0.25 eV at 300 K), or by having very high mobility carriers ( cm2/V s) [ 20]. Optimizing the power factor of a material or system would have to be based on band gap size, shape and width of the bands near the Fermi level ( EF), and the carrier effective masses and mobilities [ 25]. A common characteristic of recently developed t hermoelectrics wi th Z T > 1 is that most have lattice thermal conductivities ( L) that are lower than the current commer cial materials. The lattice thermal conductivity ( L) component of the total thermal conductivity is the only parameter not determined by the electronic structure [ 2, 3 ]. Lowering e would come at the cost of lower ing which is not desirable in a TE material. A reduced lattice thermal conductivity directly improves the TE efficiency, and allows re optimization of the carrier concentration for additional Z T improvement. Productive s trategies to improve the ZT relat ed to minimizing the thermal conductivity are the use of materials with intrinsically low L, and to scatter phonons within the unit cell or structure. Materials with intrinsically low L follow the phonon glass electron crystal (PGEC) approach. The PGEC approach is a concept introduced by Slack (1995) as a guide for new TE materials with improved properties [ 26]. An ideal PGEC material would have the thermal conductivity associated with amorphous materials (phonon glass), and the electronic properties as sociated with good semiconductor single crystals (electron crystal). Scattering phonons c an be accomplished by (i) creating rattling structures or point defects such as intersti tials, vacancies or by alloying, (ii) the use of complex crystal structures (iii) increasing the lattice period, and (iv) by scatter ing phonons at interfaces utilizing boundary scattering.

PAGE 33

13 1.1. 3 Clathrates PGEC materials that can satisfy these strategies include the class of novel materials known as clathrates [ 27, 28]. Inorgan ic clathrates were first systematically investigated in the 1960s with Na8Si46 (Kasper et al ., 1965), Si and Ge (Cros et al ., 1968), and Sn clathrates (Gallmeier et al ., 1969) [ 29]. Type I inorganic clathrates are characterized by a framework of host atoms which occupy sites at the vertices of facesharing polyhedral cages which enclose loosely bond guest atoms. A general chemical formula for type I clathrates can be written as A8ByY46y, where A represents the guest atom (Group IIA, alkaline earth metal, Sr, Ba, or Eu), and B framework substituting atom (Group IIIB, Ga or In), and Y (Group IVB, Si, Ge, or Sn) represent the framework atoms. The host atom cages form two pentagonal dodecahedrons and six tetrakaidecahedrons ( hexagonal truncated trapezohed ron ) per cubic unit cell [ 29, 30]. The 12 sided ( 12 pentagons), 20 atom dodecahedron and the 14 s i ded (12 pentagons and 2 hexagons), 24 atom tetrakaidecahedron cages with the enclosed guest atom are shown in Figure 1.7a. The unit cell crystal structure hig hlighting the dodecahedra outlined at the center of F igure 1.7b, a nd the tetrakaidecahedron is outlined at the right of the figure The green shaded spheres represent atoms forming the host framework and the guest atoms inside the polyhedra are represented in shades of blue T ype I clathrate belongs to the c rystallographic cubic space group Pm 3n and in the Wyckoff position notation the host atoms are found at three different sites 6c 16i and 24k (see Appendix A) and the centers of the cages are denoted 2a and 6 d for the dodecahedr on and tetrakaidecahedr on cages, respectively.

PAGE 34

14 The fully occupied ( A8ByY46y) host atom framework of the t ype I c lathrate contains 46 tetrahedrally coordinated host atoms in the unit cell. The geometrical arrangement of the 4 6 host atoms leaves 8 voids per unit cell for the guest atoms to occupy. An illustrative descri ption of the three dimensional structure is shown in Figure 1.8a d, where depth perspective has been added for clarity The one guest atom is located at each cor ner ( 2a ) and two on each face ( 6d) of the cubic unit cell ( Figure 1.8a ) Part of the structure can be envisioned where a dodecahedron cage encloses the guest atom at each 2a site ( Figure 1.8b ) Each tetrakaidecahedr on cage enclose s a 6d site guest atom whe re these cages are joined by their hexagonal faces (Figure 1.8c) The complete structure with the unit cell outlined is illustrated in Figure 1.8d which can be visualized as the dodecahedron cages join ed through the 16 i site s with vacancies at the 6 c site s. The dodecahedron cage is a subunit present in all inorganic clathrates and f illing the 6 c sites form the outline of the tetrakaidecahedron cages. Describing the type I clathrate with this Figure 1.7. The type I clathrate composed of the (a) face sh aring pentagonal dodecahedr on and tetrakaidecahedron, and the (b) cubic unit cell with the polyhedra outlined in red. The Wyckoff positions of the guest and framework atoms are indicated. x y z 24k 16i 6c 6d 2a (a) (b)

PAGE 35

15 manner has been used to help explain the preference for vacancies and lower valence host elements on the 6 c site [ 124] The loosely bound guests, or rattlers, in these crystalline materials can undergo (a) Figure 1.8 Components of the type I clathrate structure with (a) eight guest atoms on the corn ers one on center, and twelve on the faces of the cubic unit cell, an d the (b) nine dodecahedron cages, (c) the twelve tetrakaidecahedron cages, and (d) the complete host guest schematic. (b) (c) (d)

PAGE 36

16 large, localized low frequency vibrations, which in turn can scatter the heat carrying acoustic phonons resulting in thermal conductivities with magnitudes similar to amorphous materials. The guest host interaction in clathrates continues to be of scientific interest. It has also been demonstrated that the transport properties can be optimized through Si substitution wi thin the Ga Ge framework [ 3133 ] and manipulation of the carrier concentration [ 34]. The type I class of clathrates meet desired features for possible ZT improvement with their intrinsically low complex crystal structure with the phonon scattering rattl er As a result of the interesting properties and technological promise of typeI clathrates, the bulk properties of these materials have been studied extensively using a wide range of experimental and theoretical techniques [34]. 1.1.4 Thin Film Thermo electric Structures Phonon scattering at interfaces lead s to the use of multiphase composites mixed on the nanomet er scale. These nanostructured materials can be formed as thin film hetero structures and superlattices, or as intimately mixed composite stru ctures TE materials in thin film form offer promise for ZT enhancement [ 35] Three thin film approaches have been suggested. The first approach use s quantum confinement effects to obtain an enhanced density of states near the Fermi level The second appro ach employs phononblocking/electron transmitting superlattices. These structures utilize the acoustic mismatch between the superlattice components to reduce L, rather than using the conventional alloying approach, thereby potentially eliminating alloy scattering of carriers. The third thin film approach is based on thermionic effects in heterostructures

PAGE 37

17 [ 36] Thin film heterostructures, or s uperlattices cons ist of alternating thin layers (2 10 nm each) of different materials stacked periodically (~100 m total thickness) The lattice mismatch and electronic potential differences at the interfaces and resulting phonon and electron interface scattering and ban d structure modifications can be exploited to reduce phonon heat conduction, i.e. L, while maintaining or enhancing the electron transport This in turn, improv es the thermoelectric figure of merit [ 35] 1.1.5 Thin Film Deposition Techniques Films of TE materials have been deposited by a variety of fabrication techniques. Research based thin film deposition techniques include molecular beam epitaxy (MBE) [37] and low pressure chemical vapor deposition (CVD) [38] of Si Ge alloys ; low temperature metal organic chemical vapor deposition (MOCVD) [39, 40], flash evaporation [ 41, 42], MBE [43], sputtering [ 44], and pulsed laser ablation (PLA) [45] of Bi2Te3 and Sb2Te3 alloys The method of PL A profitably differs from the thermal method s of continuous deposition of thin semiconductor films. The presence of a large fraction of excited atoms and ions allows lower growth temperatures and a high nucleation rate allows the deposit ion of thin continuous films on the order of several nanometers. In addition, since an insignificant mass of the target material evaporates for one pulse, the film th ickness can be precisely controlled by the number of laser pulses. In some cases, the quality of the films obtained by laser ablation is comparable with the quality of films grown by MBE [ 46]. One of the main advantages of laser ablation for film growth is the ability of this process to closely reproduce the target stoichiometry in the deposited film [ 4749].

PAGE 38

18 For a target that contains elements with a wide variation of physical properties, pulsed laser ablation is uniquely suited for the growth of complex s tructured, multicomponent films from a single composite target such as Ba8Ga16Ge30. 1.2. Pulsed Laser Vaporization and Deposition Pulsed laser deposition (PLD) is often described as a threestep process consisting of (i) vaporization of a target materi al, (ii) transport of the vapor plume, and (iii) film growth on a substrate (Figure 1. 9) [50, 51]. During the first step, a pulsed laser beam strikes the surface of the source material, a rotating target, and the energy from the laser rapidly evaporates th e targets surface. In most materials, the radiation is absorbed by only the outermost layers of the target to a depth on the order of fractions of a micrometer The short laser pulses lasting less than 50 ns cause the temperature of the surface to rise rapidly to thousands of degrees C elsius, but the temperature of the target (iii) Film Growth (i) Vaporization (ii) Transport of Vapor Plume Rotating Target Pulsed Beam Substrate Figure 1. 9 Diagram of the pulsed laser deposition process depictin g the pulsed beam (i) vaporization of the target surface, (ii) transport of the vapor plume, and (iii) growth of the film on the substrate.

PAGE 39

19 bulk remains unchanged The rapid, nonequilibrium heating produces evaporants along with the characteristic luminous plasma that expand and propagate, as the second step. This vaporized material then deposit s on the substrate, the third step, producing a film with composition typically identical to that of the target surface. The ability to reproduce the composition of the target in the film may be the highest benefit of PLD [ 52] This technique has many unique advantages for thinfilm deposition (Table 1.1). An energy source (laser) outside of the deposition chamber produces a highly forwarddirected and confined plume of materials, which can be deposited with less contamination than the unconfined plasma in a sputter process. In complex multi component material deposition with conventional evaporation methods, the various cations come from different sources. To produce the right mixture in the deposited film, the rate of arrival of each species must be monitored and controlled. PLD does not require such monitoring, because the composition of the film typically replicates the composition of the target referred to as congruent evaporation and stoichiometric deposition Producing mu ltilayer materials can also be accomplished rather easily with PLD, becaus e different targets can be alternately posi tioned in the path of the laser beam which is done with a controlled target carousel. PLD is a good technique for depositing Property Effect on film properties Congruent evaporation StoichiometryHigh kinetic energy of species Crystalline structure. Sputtering of light atoms High instantaeous evaporation rates Low background gas incorporation. Crystalline structure Low power input Low thermal radiation Local evaporation Preservation of source purity Particulate ejection Degradation of optical, electrical properties Table 1.1. Properties of pulsed laser evaporation and deposition.

PAGE 40

20 extremely pure films. In most proc esses the film includes contaminants. For example, in thermal and electron beam evapo ration a container is used to hold the source material, and the high temperatures needed to evaporate the source materials can also evaporate parts of the container thereby contaminating the deposition process. In metal organic chemical vapor deposition (MOCVD), volatile organic molecules transport the desired cations from the source to the fi lm, and fragments of the organics can end up in the film [ 52] It is desirable to have the photon energy of the laser greater than the band gap of the material which is being ablated, thus giving lower densities required for ablation. The photon energy, E, is given by; hc E ( 1. 19) where h is Plan ks constant (6.626 1034 J s), c is the speed of light (2.998 108 m/s), and is the laser wavelength. The laser power density at the target surface can be divided as low (< 106 W/ cm2), intermediate (> 106 W/ cm2 and < 108 W/ cm2), and high (>108 W/ cm2). The intermediate to high power density regimes correspond to typical PLD laser fluences of 1 to 6 J/ cm2 with laser pulse widths of 20 to 60 ns. An advantage of laser evaporation is the small heat content of the evaporation target due to the small size o f the evaporating volume and shorter radiation time of the pulsed laser. The small volume of heated material is defined by the area of the laser spot size at the target surface and its absorption heated depth ( dH). This volume is quickly brought to boiling temperature which then rapidly cools, by conduction into the bulk of the target, upon termination of the laser pulse. These cycle times are on the order of milliseconds and are conducive for congruent evaporation [ 53].

PAGE 41

21 This process is depicted in Figures 1. 10 and 1.11. The dominant mechanisms involved in this process are found to depend on laser parameters such as the energy density (fluence), pulse duration ( p) wavelength, polarization, laser repetition rate [ 54] The material properties that affect the surface temperature and nature of evaporation are the thermal diffusivity ( Dt) and optical absorption coefficient ( l) at the laser wavelength. During initia l stages of the laser target interaction thermal energy is absorbed into a thin layer on the surface of the target and then diffuses into the target during the laser pulse duration. The depth of the heated volume of material is given by the thermal diffu sion length ( ld) [ 55]: p T dD l2 ( 1.20) where the thermal diffusivity ( DT) of the material is given by: p TC D (1.21) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 -100 -50 0 50 100 150 200 Time (ns) Normalized Voltage Energy absorption and initiation of ablation 40 ns Continued ablation and plasma absorption Figure 1.10 A diagram of the excimer laser energy (i) absorption by the target material and initiation of ablation and (ii) continued target ablation and fractional plasma absorption of the las er pulse.

PAGE 42

22 where is the thermal conductivity, is the material density, and Cp is the heat capacity. If the thermal diffusion length ld << l 1 then the heated depth dH l 1, or if ld >> l 1 then dH ld [ 55] (Figure 1.10b). Low diffusivity and high absorption, resu lts in high peak temperatures and concentration of thermal energy at the surface, which are good for congruent evaporation. The transfer of energy occurs within a few picoseconds, then heating of the absorption layer will begin. Vaporization of the absorpt ion depth will begin with a timeframe on the order of 100 ps from the arrival of a 20 ns laser pulse duration [ 54]. Figure 1.1 1 The evolution of the laser target interaction. The (a) The unablated target, (b) laser pulse arrival, rapid target surface heating, and plasma initiation, (c) pulse continuation, puls e absorption into the expanding plasma, and increased surface heating and propagation into the target, (d) pulse termination, hydrodynamic plasma expansion, and rapid target cooling, and (e) ideal return of the target to the initial state. ( d ) (a) (c ) ( e ) (b ) Plasma Laser beam d T

PAGE 43

23 At sufficiently high laser fluence, a plasma is formed in the vapor over the target area irradiated by the incident laser pulse. The interac tion of this plasma with the remainder of the laser pulse and evaporation source, strongly affects the nature of the evaporation and energy coupling to the target surface. A general feature of pulsed laser ablation plasmas, or plumes, is the high ion and e lectron temperatures on the order of several thousand Kelvin, and the high degree of ionization [ 54]. When this plasma is dense enough it will absorb laser radiation directly by inverse Bremsstrahlung process [ 56] and the plasma absorption coefficient ( ) is given by: T k h i cBe T n Z 1 10 69 33 2 1 2 3 8 (1.22) where Zc is the average charge in the plasma of ion density ni, at temperature T and laser frequency Also, ne, q, me, are the electron density, electronic charge, and electron mass, respectively. This absorption increases the plasma temperature and shields the evaporating surface from laser radiation so that the evaporation enters a self regulating regime. Radiant heating from the plasma, and plasma impingement further heats the surface, increasing the energy coupling efficiency to the material (Figure 1.10c). In instances where a plasma is produced, the wavelength of irradiation determines the fractional absorption of the beam by the plasma. The plasma resonant frequency ( vp) is given by: 2 / 1 3 2 / 1 210 9 8 4e p e e pn v m q n v (1. 23)

PAGE 44

24 A ratio of v1/ vp is necessary i n order for substantial absorption to occur in the plasma, w here v1 is the laser frequency [ 56, 57]. A laser pulse optical frequency greater than the plasma frequency, l > p, results in the laser ene rgy predominately penetrating the plasma and continuing to interact with the target surface. I f p > l then proportionally more of the laser energy will be reflected from the plasma. The number and density of the excited and ionized atoms is a strong function of laser fluence for evaporation. Typical values of the charge density are ne = 1019 1020 cm-3 for plasmas generated by PLA. This will lead to critical plasma frequency values in the range of 3 1013 to 9 1013 s1. Formation of the plasma and it s subsequent hydrodynamic expansion are the origin of the high kinetic energy, and ionized and excited neutral species, and their impingment on the film growth surface, which play important roles in the nucleation, structure, and growth of thin films [ 58]. The plasma will produce species with high kinetic energies ranging between 1 eV and more than 100 eV [ 59, 60]. Particle energies in the 10 to 1000 eV range are sufficient to break atomic bonds, cause thermal spikes, generate subsurface vacancies and displ acement of atoms, and enhance surface diffusivity of the adsorbed atoms. These mechanisms generally improve the film microstructure and morphology in a manner similar to ion assisted deposition [ 61]. They can also cause damage to the substrate and film as with ion etching [62]. Lower substrate temperatures can be used in pulsed laser deposition due to enhanced surface mobility caused by energetic bombardment during deposition [ 54]. For example, epitaxial germanium film growth was possible at 300 C using PL D while for molecular beam epitaxy (MBE), temperature in excess of 700 C are required [ 51].

PAGE 45

25 PLD stands out as perhaps the most promising deposition choice for a broad range of materials. Reliable congruent transfer of material from bulk to film is possib le, the energy range of the impinging particles promotes surface diffusion while avoiding bulk damage, the species are often activated either as ions or excited neutrals which facilitate associative chemistry on the growing film, and virtually all materials can be ablated [ 54]. 1.3. Summary The pursuit of improved performance of TE materials is motivated by energy production and conservation issues. TE m aterials with complex structures and intrinsically low thermal conductivity such as type I clathrates ar e promising candidates for investigation. In addition, the enhanced performance of TE structures composed of thin films is of interest. TE device s composed of thinfilms may be appropriate for small scale electronic and optoelectronic applications where sm all heat loads or low level of power generation are more appropriate [ 71]. The advantages of PLD are ideal for in situ growth of such complex multi component materials such as type I clathrate. Based on the results obtained, this will help establish a basic understanding of film growth of complex structures, and to date the quantity of reports on the production and characterization of any type I clathrate thin films are extremely limited [ 72]. 1.4. Outline of Dissertation The thinfilm growth and character ization of the novel thermoelectric material type I clathrate Ba8Ga16Ge30 deposited by pulsed dual laser ablation was studied through the following investigations. The surface modification of the polycrystalline target by

PAGE 46

26 pulsed laser ablation techniques w as investigated through surface morphology and composition studies (Chapter 3). The optical emission and expansion of the vaporized material plume in transit to the substrate was studied by in sit u optical diagnostics (Chapter 4 ), and a hydrodynamic model of the plasma expan sion was investigated (Chapter 5 ) to determine plasma temperatures of the ablated material. The resultant films were studied for the structure and electric transport properties (Chapter 6). A comprehensive summary of these results and a brief discussion of future directions will be presented (Chapter 7).

PAGE 47

27 CHAPTER 2: DUAL LASER ABLATION For certain applications, there remains an obstacle with PLD despite its advantages. Some materials when used with PLD pro duces fine particulates rang ing in size from microns down to submicron dimen sions. Particulates generated by pulsed laser ablation can be distinguished by whether the original material when ejected from the target is in the solid, liquid, or vapor state [ 55]. The size of particulates formed from the vapor state tends to be in the nanometer range and may be spherical or polyhedral in shape. The shape of particulates formed from liquid ejecta tend to be spherical and in the sub micron to micron size range. Particulates formed from solid ejecta are also in the micron size range and tend to be irregularly shaped. There are multiple sources of particulate generation: (1) dislodged fragments from pits and micro cracks due to laser induced thermal and mechanical shock; (2) rapid expansion of trapped gas beneath the target surface during laser irradiation caus ing forcible ejection of material; (3) splashing of the molten layer caused by the presence of a subsurface superheated layer or from laser induced rapid surface evaporation; and (4) condensation from vapor species due to supersaturation observed with high ambient gas pressure [ 55, 53]. Particulates incorporated into the film degrade its properties and the tolerance for particulates is application specific as well as dependent on their size and density. Considering an electronic thin film component with a t ypical width of 10 m, a particulate density of less

PAGE 48

28 that 1 per 100 m2 may be desirable [ 55]. A number of techniques exist to eliminate particulates. A high quality high density target can minimizes the problem for some materials and other active and passive techniques have been developed to filter out the heavy particulates [ 52] A simple approach to reduce the number of particulates is to reduce the laser power density to below the threshold level that causes the splashing of the molten layer, also b y increasing the spot size [ 63, 55]. Rotating vane velocity filters can be used to stop the slow moving massive particles (~1 10 cm/ms), but let the fast atomic and molecular evaporants through (~1 cm/s) [ 64, 55]. Other techniques involve gas dynamic separation [ 62], off axis deposition [ 65], and substrate biasing [ 66] to separate massive particles from the vapor plume [ 55]. UV Detector Excimer Laser CO2 Laser Delay Generator IR Detector Oscilloscope Rotating Target Vacuum Chamber Energy Monitor Substrate Heater Figure 2 1 A diagram of the pulsed dual laser ablation system. The system can be operated in either excimer only single laser or dual laser mode.

PAGE 49

29 A modified approach for particulate reduction is employing a second pulsed laser to control par ticulate production (Figure 2.1). The plume absorption of the second laser pulse, to heat the plume formed by the first laser, evaporates the ejected matter in the plume. This pulsed dual laser ablation technique utilizes spatial overlap of two laser pulses of different wavelengths on the target surface. The ejection of particulates is essentially eliminated with use of a suitable temporal delay between the two laser pulses (Figure 2.2). Additionally, this technique enhances the evaporated species kinetic energy and ionization in the laser ablated plume by enhanced plume excitation. This allows the reduction of the substrate temperature for epitaxial film growth as a result of the increased mobility on the substrate due to enhanced plume excitation. Figure 2.2. Oscilloscope traces of the temporal synchronization of the (UV) excimer laser pulse and the (IR) CO2 laser pulse. The pp delay (peak to peak) is shown relative to the peak of the excimer laser pulse. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 0 100 200 300 400 500 600 700 800 900 Time (ns) Normalized Voltage UV IR p p delay

PAGE 50

30 Figure 2.3 Evolution of the laser target interaction for dual laser ablation. (a) CO 2 laser pulse arrival and on set of surface melt, (b) synchronized arrival of the excimer laser pulse to interact with the pre melt zone, initiate evaporation and plasma formation, and dual laser plasma absorption, and (c) termination of the excimer pulse, and continued plasma absorption of the CO2 laser pulse, increasing plasma temperature and expansion. ( a ) Surface pre-melt 0.0 0.2 0.4 0.6 0.8 1.0 -100 0 100 200 300 400 500 600 700 800 Time (ns) Normalized Voltage IR Pre-melt (b ) 0.0 0.2 0.4 0.6 0.8 1.0 -100 0 100 200 300 400 500 600 700 800 Time (ns) Normalized Voltage UV IR Evaporation and plasma absorption ( c ) 0.0 0.2 0.4 0.6 0.8 1.0 -100 0 100 200 300 400 500 600 700 800 Time (ns) Normalized Voltage IR UV Plasma absorption

PAGE 51

31 During ablation (Figure 2.3) using the dual laser process the IR CO2 laser pulse arrives at the target first and begins to heat up the material (Figure 2.3a). When the UV excimer laser (KrF) pulse arrives at the same spot on the target that has been heated just above the melting point by the CO2 laser pulse, it interacts with a smooth molten layer of the material (Figure 2 .3b). The molten material lacks any cracks and loose particles that could produce particulate ejection [ 53]. The plasma plume formed at the target surface will interact with th e remaining part of the CO2 laser pulse to heat the plume to a higher temperature, through the inverse Bremsstrahlung process (eq. 1.22) that would re evaporate any submicron particles remaining in the plume (Figure 2.3c). This dual laser deposition proces s reduces the particulates without losing any of the basic advantages that makes pulsed laser deposition appealing for thin film deposition [ 4749, 6769]. Material ablation by synchronized dual lasers is a promising technique to simultaneously minimize bo th particulates and the narrow angular distribution problem of standard single laser PLD [ 70]. This dual laser ablation technique has been successfully employed in the control of Pb depletion in PZT (PbZr1 xTixO3) films [125], and film growth of Y2O3 [126, 132], Zn [127], ZnO [126, 128, 130, 131], diamondlike carbon [129], multicomponent films of CuIn0.75Ga0.25Se2 [131], and Er doped Y2O3 [133] The implementation of the dual laser system ultimately involves the relationship between the two laser pulses at the target surface and the qualities of the material. The effect of laser target interactions of each laser individually will guide selection of promising fluences for each laser when used in combination. Generally, the applicable fluence of the excimer l aser is greater than the threshold fluence but less than the optimum fluence required during the standard single laser ablation method. The choice of

PAGE 52

32 the CO2 laser energy for a particular material is a fluence that produces surface melting without causing significant material removal through ablation. The two lasers then require temporal synchronization such that the CO2 laser pulse arrives at the target first to initiate surface melting, followed by the arrival of the excimer laser pulse at on set of the m elt, see Figure 2.2. The proper coupling of the infrared (IR) CO2 laser energy with the ultraviolet (UV) excimer laser generated plasma results in ionic enhancement and increased kinetic energy of the evaporated material may allow even lower substrate temp eratures to be used as compared to standard singlelaser PLD due to the enhanced surface mobility caused by energetic bombardment during deposition [ 68]. 2.1. Description of the Pulsed Dual Laser Ablation System The concept of singlelaser PLA, or PLD, w as introduced Chapter 1.2. The dual laser ablation system, shown in Figure 2.1, can be operated in either single laser or dual laser modes of operation. The evacuation of deposition chamber is performed by a roughing pump and turbo molecular pump in series to the upper micro torr range, then by a closed cycle cryo pump, to the 107 torr range. The substrate heater is a stainless steel block internally heated by a 600 watt halogen bulb, with a temperature range of room temperature (26 C) to 700 C. The substrate temperature is monitored by a K type thermocouple with the tip in thermal contact by conductive silver paste to the face of the heater block. The substrates are located adjacent to the thermocouple tip and fixed to the heater block with conductive s ilver paint to improve thermal contact. The substrate temperature was referred to, and recorded as, the surface temperature of the heater block. The target to substrate distance was typically maintained at 4 cm on axis to the laser -

PAGE 53

33 target interaction site. The laser utilized during single laser ablation is a UV KrF excimer laser with a pulse full width half maximum (FWHM) duration of 28 56 ns, emitting a wavelength of 248 nm. During experiments using dual laser ablation the second laser is a far IR CO2 la ser with a pulse FWHM of 250 ns, emitting a wavelength of 10.6 pulses from the two lasers are synchronized through the use of delayed triggering from an inductive pickup of the CO2 laser discharge. The reported laser fluences are measured at the target surface, accounting for losses through the optical path(s) of the laser pulse. Losses occur at optical elements such as the mirrors, lenses, and chamber windows; there was 70.3 % transmission of the excimer laser pulse, and 86.9 % transmission of the CO2 laser pulse, to the target. 2.1. 1. Dual Laser System Para meters The dual laser system described in Figure 2.1 is initiated by the internal trigger of the CO2 laser (Figure 2. 4a). A n inductive pickup of the CO2 laser capacitor discharge (Figure 2. 4 b) is used to trigger the digital delay generator (DDG) t0 signal (Figure 2. 4c) to trigger the excimer laser. The temporal delay between the excimer laser output pulse (Figure 2. 4d) and the CO2 laser output pulse (Figure 2. 4e) is detected by the UV and IR detectors (Figure 2.1) and monitored on a fast oscilloscope, 1 gigi samples per second, and regulated by the excimer laser triggering delay set by the DDG. There are innate properties of the lasers that affect the internal delay between their trigger and output. These properties include the level of the high voltage (HV ) applied to the discharge capacitors, the ratio of the gas mixtures, and for the CO2 laser specifically, the output coupler reflection (OC) and laser cavity geometry. Typically, it is desired to

PAGE 54

34 have the excimer laser output occur as early as possible, an d the CO2 laser pulse to occur as late as possible, while still maintaining the quality of the laser pulse. This allows the pp delay to be controlled over a large range by the DDG. The temporal delay is referenced relative to the peak of the excimer laser pulse (0 ns), called the peak to peak delay ( pp 0 500 1000 1500 2000 2500 3000 Time (ns) (a) (e) (d) (c) (b) Normalized Voltage Figure 2.4 The time line of component synchronization for the dual laser ablation system. The t0 trigger to initiate the dual laser pulse cycle is the (a) CO2 laser internal trigger, followed by the (b) CO2 laser capacitor discharge, which triggers the (c) digital delay generator (DDG) t0 signal to trigger the excimer laser, followed by the (d) excimer laser output pulse, and (e) the CO 2 laser output pul se.

PAGE 55

35 delay). If the peak of the CO2 laser pulse occurs after the peak of the excimer laser pulse, the pp delay is positive (+), and a CO2 laser peak before the excimer pulse would be negative ( ). The typic al pp delay range is limited to approximately +250 ns by the CO2 laser gain build up time, and is essentially unlimited in the negative pp direction. For practical purposes, a negative pp delay greater than 50 ns would cause undesired ablation by the C O2 laser pulse energy, and not allow enough remaining energy in the CO2 laser pulse to interact with the excimer laser generated plasma. Laser pulse quality is evaluated by uniformity of the energy distribution across the laser spot area, and the jitter. E nergy distribution becomes less uniform at low laser energy. Jitter is a measure of the fluctuations of the temporal positions of pulses, as illustrated in Figure 2.5 Previous experiment reports have demonstrated that precise synchronization of the two la ser pulses 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 0 25 50 75 100 125 150 175 200 225 250 Time (ns) Normalized Voltage 16 ns jitter Figure 2 5 Detail of the cumulative oscilloscope traces of the excimer and CO 2 laser pulses to illustrate the 16 ns jitter of the CO2 laser pulses relative to the excimer laser pulses. The jitter is measured between the minimum and maximum occurrence of the fullwidth at half maximum (FWHM) of the laser pulses.

PAGE 56

36 to within tens of nanoseconds is crucially important for the particulatefree deposition of thin films [ 67, 69] and excessive jitter will adversely affect pulseto pulse synchronization. Laser pulse j itter is affected by the HV, gas mixture, OC, and CO2 laser cavity geometry. Greater levels of HV will reduce jitter, depleted gas mixture will increase jitter, high OC reflection decreases jitter, and longer laser cavity length increases jitter. A summary of the laser parameters effect on the laser output is given in Table 2. 1. The listed parameters are the high voltage applied to the discharge capacitors (HV), the gas mixture ratio, the reflection of the CO2 laser output coupler (OC), and the cavity length of the CO2 laser. The effects on the laser output by changing these parameters are evaluated by the pulse energy, delay, jitter, and duration ( p). Adjusting one parameter to achieve a desired output effect generally requires compensation by other parameters. Therefore, a compromise of the parameters is established to have control over the range of values desired. Energy Delay Jitter Duration High Increase Decrease Decrease Increase Low Decrease Increase Increase Decrease Optimum Increase Decrease Decrease Negligible Depleted Decrease Increase Increase Slight Increase High Increase Decrease Decrease Negligible Low Decrease Increase Increase Negligible High Negligible Increase Increase Negligible Low Negligible Decrease Decrease Negligible Effect on laser output Property Value HV OC Reflection Cavity Length Gas Mixture Table 2 1 Properties of the excimer and CO 2 lasers and resulting effect on laser output. Laser properties include the high voltage of the discharge capacitor (HV), gas mixture ratio, reflection of the CO2 laser output coupler (OC), and cavity length of the CO2 laser.

PAGE 57

37 2.1.2 CO2 Laser Parameters I t was found that configuring the CO2 laser with a ZeSe Brewster window attached to an extension tube yielded the best results for achieving a consistent and uniform laser pulse with a practical energy density range required for experimental purposes Plac ing the window as far or farther (20.75 inches, on center) from the end of the electrodes as the gold back mirror s distance (12 inches), kept the ZnSe window from becoming damaged due to sputtering from ionized gas within the laser gas cavity (Figure 2.6) The final cavity length was 161.8 cm which was the longest length which could be accommodated to maximize the delay and also allow ample working space for adjustment s and maintenance. A n 80% reflectivity and 20 meter curvature germanium output coupler was chosen which produced approximately 550 mJ of energy at a HV of 36.5 kV, while providing a delay of 9751310 ns (Appendix K ) The spatial filter was completely opened for maximum power output and the output coupler was adjusted for 6 2 multi mode. A st andard gas premix was used for the CO2 laser, see Table 2.2, and was Cavity Length Back Mirror ZnSe Brewster Window Output Coupler Spatial Filter Electrodes Cu Mirror Figure 2.6. Schematic diagram of the CO 2 laser with the ZnSe Brewster window extension tube and Ge output coupler.

PAGE 58

38 flowed constantly in order to maintain consistent energy output. The laser applied HV was maintained at 30 kV or greater to minimize the jitter. 2.1.3 Excimer Laser Parameters The cav ity length and output coupler of the excimer laser were not modified from the original manufactures configuration. As shown in Figure 2.7a b, as the HV is increased the output energy increases, and as the energy is increased, the delay and jitter decrease. These are desirable characteristics; however, the output energy is too high for the required dual laser conditions. The gas mixture was then intentionally depleted by manually injecting the neon buffer gas into the laser cavity while maintaining required 3000 mbar cavity pressure. This buffer injection results in decreased energy output at elevated HV settings without significantly increasing the jitter or delay. Table 2.2. The CO 2 laser gas premix used during dual laser ablation. Gas Mix Percent (%) CO 1 N 4 CO212 He Balance

PAGE 59

39 2.2. Summary Through proper configuration of the dual laser system it was possible to achieve the desired output conditions CO2 and excimer laser fluences as high as 3 J/ cm2 for each laser could be achieved, with a pp delay range of 50 to 200 ns, and typical dual laser jitter of 16 ns with 8 ns capability. This degree of flex ibility provides a versatile system for experimental investigations of pulsed laser ablation. 1050 1100 1150 1200 1250 1300 25 75 125 175 225 275 325 375 Energy (mJ) Delay (ns) 16 18 20 22 24 26 28 30 32 34 25 75 125 175 225 275 325 375 Energy (mJ) Jitter (ns) ( a ) ( b ) Figure. 2.7. Output ch aracteristics of the excimer laser after a new fill, the (a) output delay and (b) pulse jitter relative to the trigger pulse, as a function of the output energy.

PAGE 60

40 CHAPTER 3: INVESTIGATION OF TARGET SURFACE MODIFICATION BY PULSED LASER ABLATION In order to produce high quality thin films by PLD with basic properties such as broad area, smooth topography, uniform distribution, and stoichiometric deposition, it is essential to gain an understanding of the interaction between the focused laser pulse and the target surface. Insight into this laser target interaction can be gained through investigations of the target surface modification by the high power, pulsed la ser ablation of the material of interest. Stoichiometric deposition is dependent on congruent evaporation of the laser irradiated target, and to achieve this, the entire heated volume (the laser irradiated spot area and heated depth) must be vaporized duri ng each laser pulse. However, if a high percentage of the molten material is left behind, phase segregation and recrystallization will take place [ 50]. Factors relevant to surface modification follows: (i) the maximum temperature reached in the solid and the melt depth, increase with laser fluence, (ii) the thermal pulse penetrates deeper into solids with low absorption coefficients and high thermal conductivities, and (iii) shorter pulse lengths produce higher melting and solidification velocities [ 50, 73]. The more intrinsic aspects of the laser/solid interaction are dependent on the important laser parameters; pulse width, wavelength, and fluence (energy density). It is typically desirable to have the pulse width on the order of nanoseconds or shorte r, f or longer pulse widths, the energy

PAGE 61

41 delivered by the laser is carried away by thermal conduction to a significant extent. Use of a KrF excimer laser (248 nm) as the primary laser for ablation, results in pulse widths in the range of 20 to 60 ns. There are f ew restrictions placed on the targets used in a pulsed laser deposition ( PLD ) system. Successful deposition can be made from pressed powders, sintered pellets, cast material, single crystals, and metal foils. The main differences between these different target morphologies are in the nature of the target erosion, and the generation of particulates. A general guide is that high density and highly homogenous targets yield the best films [ 74 ]. Parameter Symbol Units Value Ref. Formula B8Ga16Ge30 Temperature T K 295 Formula Weight (g/mol) 4391.94 15 Space Group Pm-3n 15 Lattice Parameter a 10.767(10) 15 Volume V 31248.2(2) 15 Atoms/Unit Cell Z 1 15 Density calcg/cm35.843 15 Linear Absorption Coefficient lmm-132.383 15 Melting Point C 963 5 18 Seebeck CoefficientV/K 50 18 Carrier Concentration n cm-3 ~102118 Mobility cm2/Vs 10-15 18 Band gap EgeV 0.89 19 Heat Capacity CpJ/gK 0.3065 4 Lattice Thermal Conductivity LW/mK 1.9-2.1 4 Table 3 .1. Crystallographic and transport properties for single crystal type I clathrate Ba8Ga16Ge30 at room temperature.

PAGE 62

42 Reported values for properties of single crystal type I clathrate Ba8Ga16Ge30 are given in Table 3 .1. The band gap of Ba8Ga16Ge30 is 0.89 eV, and s ince it is desirable to have the photon energy ( E) of the laser greater than the band gap of the material being ablated (equation 1.19) then the KrF excimer laser (5.00 eV) will satisfy this parameter (Table 3 .2) This will allow lower energy densities for ablation. Whereas, the CO2 laser with 10.6 m wavelength ( E = 0.1 17 eV) would not be a n appropriate ablation source. Based on the material properties values (Table 3 .1) for t he thermal conductivity ( L) densit y ( ), and heat capacity ( Cp), the calculated thermal diffusivity ( DT) is 1.06 1.17 mm2/s. From previously reported excimer laser ablation of Ba8Ga16Ge30 using fluences of 1 3 J/ cm2 [6], the laser pulse duration ( p) is 30 43 ns T he calculated thermal diffusion length ( ld) is then 0.25 0.32 m. The linear absorption coefficient ( l) o f Ba8Ga16Ge30 is 32.383 mm-1 [ 76] and since ld << l 1 where l 1 = 30.88 m then the Excimer CO2Type KrF CO2Spectrum Range Ultraviolet (UV) Far Infrared (IR) Wavelength nm 248 10.6 103Photon Energy EeV 5.000 0.117 Optical Frequency ls-11.2 10152.8 1013High-Voltage HV kV 16 -24 30 40 Delay ns 1076 1264 975-1395 Fluence FlJ/cm20.5 6 0.3 3.5 Duration (FWHM) pns 25-56 261 6 Rise-Time ns 6.4 0.4 98 6 Jitter ns 6.6 (S.D.) Laser Parameter Symbol Units Table 3 .2. Parameters for the KrF excimer and CO 2 lasers

PAGE 63

43 heated depth is dH l 1 30.88 m [ 55] Since ad >> ld, the complete volume is vaporized during a high en ergy laser pulse, preventing any phase segregation. This leads t o congruent evaporation; however, at low laser pulse energies the entire volume does not reach the boiling point of the material and this leaves the low volatile material behind leading to non stoichiometric evaporation [ 50] During dual laser ablation, t he majority of the IR CO2 laser pulse is absorbed into the excimer laser generated plasma since the absorption coefficient ( ) equation 1.22, is proportional to 3 or 3, where the absorption is much stronger for the longer wavelength CO2 laser than the excimer laser radiation. Then, by the inverse Bremsstrahlung process [ 56], the target is screened from further ablation by the relatively long CO2 laser pulse. In addition, t he absorbed CO2 laser energy causes intense heating of the plasma and can evaporate the submicron particulates passing through the plasma and at the same time enhance the kinetic energies of the plume spec ies [ 47]. Also, since the wavelength of irradiation determines the fractional absorption of the beam by the plasma, as compared to the plasma frequency ( vp), equation 1.23. When v1/ vp substantial absorption to the plasma occurs and given t ypical values of the charge density are n = 1019 1020 cm3 for plasmas generated by PLA then the critical plasma frequency values will be in the range of 31013 to 91013 s1. As compared to t he optical frequency of the KrF excimer laser UV = 1.21015 s1, and the CO2 laser IR = 2.81013 s1, the CO2 laser frequency is near ideal for absorption into the plasma. Due to the low thermal conductivity and thermal diffusion length, the heated v olume of Ba8Ga16Ge30 will be small. These results, combined with the short KrF excimer laser pulse duration, are good conditions for congruent evaporation at relatively

PAGE 64

44 moderate ablation fluences. Under dual laser ablation, absorption of CO2 laser pulse en ergy into the excimer laser generated plasma will contribute to increased plasma temperatures, leading to enhanced kinetic energy and ionization. Investigation of the Ba8Ga16Ge30 target surface modification by pulsed laser ablation will lead to understandi ng of conditions which are conducive to congruent evaporation and contribute to growth of high quality thin films. 3.1 Ba8Ga16Ge30 Target Properties The Ba8Ga16Ge30 type I clathrate material used for the packed powder laser ablation target was synthesiz ed at the Novel Materials Laboratory at the University of South Florida, Physics Department [ 77]. The material was prepared by mixing stoichiometric quantities of the high purity elements Ba (99%, Aldrich), Ga (99.9999%, Chameleon), and Ge (9 9.99%, Alfa Ae sar). The powder mixture was placed in a p yrolytic boron nitride crucible and sealed in a fused quartz ampoule under high purity nitrogen gas at a pressure of 2/3 atm osphere T o form Ba8Ga16Ge30, t he mixtu re was heated at 1 C / min, held at 1000 C for 24 hours, and then cooled at a rate of 2 C/min ; the total synthesis time was approximately 48 hours The resulting product consisted of ingots possessing a metallic luster, which was then ground to 325 mesh, approximately 45 m. The density of the material w as higher than 95 % of the theoretical X ray density (5.843 g/ cm3 [78]), a s analyzed by powder X ray diffraction (XRD). The clathrate powder was uni axial cold die pressed in a 1.25 inch diameter die with a load of 27.4 103 lb/in2 and maintained for six hours. The compacted bulk density of the final target was 4.19 0.07 g/ cm3, which is 71.8 % 1.1 % of the theoretical density.

PAGE 65

45 The indexed powder XRD pattern of the cold pre ssed target is shown in Figure 3.1, and the correspondin g indices are given in T able 3.3 The XRD pattern was indexed to a cubic unit cell structure with space group Pm 3n. The diffraction pattern displays peaks which are characteristic of typeI clathrates and are in good agreement with those reported for Ba8Ga16Ge30 [ 78 79]. The XR D spectrum indicates trace amounts of the impurity phase diamond structured germanium observed in other work [ 34, 80, 81]. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 20 25 30 35 40 45 50 55 60 2 (degree) Normalized Intensity (arb.) Ge (222) (320) (321) (400) (410) (411) (420) Ge + (520) (440) (530) (531) (600) (611) Ge + (620) (540) Figure 3 .1. The normalized, indexed, powder XRD pattern for the cold pressed Ba8Ga16Ge30 type I clathrate target.

PAGE 66

46 The samples of the cold pressed target were flat polished to a surface roughness of 10.1 9.6 nm (RMS 13.9 nm) and the chemical composition was analyzed by energy dispersive x ra y spectroscopy (EDS) (JEOL) at 20 keV, as given in Table 3.4. The results Table 3.4 Composition and Ga/Ge atomic ratio of the Ba 8 Ga 16 Ge 30 type I clathrate cold pressed target determined by EDS analysis. Ba Ga Ge Ga/Ge Stoichiometic 8.0 16.0 30.0 0.533 Target 8.5 0.2 15.6 0.6 30.0 0.6 0.53 0.03 Atomic % Table 3.3 The powder XRD indices for the cold pressed Ba 8 Ga 16 Ge 30 type I clathrate target. The XRD source is Cu K Intensity (%) (hkl) d () Note 27.27 13.1 3.268 Ge 28.66 26.0 (222) 3.112 29.85 35.8 (320) 2.991 30.99 100.0 (321) 2.883 33.21 15.6 (400) 2.696 34.27 15.4 (410) 2.615 35.25 7.5 (411) 2.544 37.26 8.9 (420) 2.411 45.28 15.4 (520) 2.001 Ge + Ba8Ga16Ge30 47.66 7.3 (440) 1.907 49.17 29.3 (530) 1.851 49.99 13.4 (531) 1.823 50.73 17.9 (600) 1.798 52.24 38.0 (611) 1.750 53.68 10.1 (620) 1.706 Ge + Ba8Ga16Ge30 54.45 6.4 (540) 1.684

PAGE 67

47 were normalized to the average Ge30 atomic percent (at.%) composition. The target material was found to appear slightly Ba rich, as compared to the stoichiometric 8:16:30 ratio. Howe ver, the Ga/Ge ratio of 0.53 0.03 is consistent with the intrinsic material. 3.2 Experimental Details In order to study the effect of the surface modification at the laser target interaction site by the pulsed laser ablation processes, the Ba8Ga16Ge3 0 packed powder target was resurfaced and polished with super fine ISO P1200 grit (15.3 m average particle diameter) sand paper (see Appendix B ) to provide a smooth target surface which has not been processed by laser ablation (Figure 3.2). The target was then sprayed with pressurized dry nitrogen gas to remove loose particles. The target was mounted in the appropriate target holder and installed in the vacuum chamber. The target was kept stationary during each trial so the laser pulse could interact wit h the same region for a given laser fluence and number of pulses. This reduced the Figure 3 2 SEM image of the coarse polished (529 nm rms) Ba 8 Ga 16 Ge 30 type I clathrate cold pressed target. Approximately 315315 m viewi ng area. 50 m

PAGE 68

48 experiment duration and more importantly, the overall erosion of the target. For subsequent trials, the next laser impingement site was changed to a clean, un ablated locati on either by moving the laser spot location by vertical translation of the focusing, or rotating the target. The process was repeated for the next number of repetitions or laser fluence. The focused laser spot size at the target surface (Figure 3.3 ), was m aintained at 2.0 mm wide by 3.0 mm high for each laser fluence, by adjusting the laser focusing lens(es). The vacuum chamber was evacuated to 106 Torr for this investigation. The focused laser pulse was impinged onto target surface locations for 1, 10, 100, and 1000 counts. The UV KrF excimer laser fluence was varied to 1, 2, 3, 4, and 5 J/ cm2. The IR CO2 laser fluence was varied to 0.3, 0.6, 1.3, 1.8, 2.0, and 2.4 J/ cm2. The surface morphology and composition of the laser target interaction sites were t hen examined for 2 mm 3 mm Target Laser Target Interaction Site Figure 3.3 Diagram of the laser spot size at the laser target interaction site.

PAGE 69

49 morphological and compositional changes by scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS), respectively. A photograph of the Ba8Ga16Ge30 target after various UV laser fl uences and repetitions (Figure 3 4), dis plays the result of the accumul ated laser impingement. Figure 3. 5 shows the typical layout of the SEM and EDS analysis sites. SEM observations were conducted at 30X to 30000X magnifications to determine appropriate representation of the morphology of the s urface. It was concluded that 300X magnification (approximate scan area of 315 315 m) is representative of the laser target interaction area. EDS was performed and at four adjacent locations at each site, the average and standard deviation these locatio ns were reported Each interaction site was compared to a bare, un ablated region of the target surface for surface morphology and stoichiometry. The results of the single laser, excimer only and CO2 only, laser target interactions will determine the las er fluence combination to be used for the dual laser ablation investigations. 1 J/cm 2 2 J/cm 2 3 J/cm 2 4 J/cm 2 5 J/cm 2 1000 pulses 100 pulses 10 pulses 1 pulse Figure 3.4 Photograph of the Ba 8 Ga 16 Ge 30 target with various laser interaction sites at fluences of 1 to 5 J/cm2 and repetitions of 1, 10, 100, and 1000 pulses.

PAGE 70

50 3.3 Investigation of the Excimer Laser Target Interaction SEM images are shown for UV KrF excimer only, laser target interaction sites for fluences of 1, 2, 3, 4, and 5 J/ c m2 at cumulative pulses or 10, 100, and 1000 pulses per site, respectively (Figure 3 6a and b). The surface modification can be compared to the un ablated target region shown in Figure 3. 2. It was observed that practical ablation did not occur with an exci mer laser fluence of 0.5 J/ cm2. This establishes that the ablation threshold ( FT) is approximately 0.5 J/ cm2. Ablation at a low fluence of 1 J/ cm2 eventually produces defined cones as the laser pulses progress from 10 to 100 pulses/site, and become more defined as the pulses reach 1000 pulses/site. The cones are directed toward Target Laser Target Interaction Site Laser Target Interaction Site 30X SEM EDS Site (4) SEM EDS Site 300X Figure 3.5 Diagram of SEM a nd EDS analysis locations for the laser target interaction study on the Ba8Ga16Ge30 target.

PAGE 71

51 the incident laser beam, and are the result of cumulative shadowing caused by ripples in the incomplete melt zone of the interaction site. Typically, the formation of cones is and indication of non congruent evaporation. At 2 J/ cm2 the result was eventual partial cone and molten zone formation as 1000 pulses/site is achieved. At fluences of 3, 4, and 5 J/ cm2, a smooth melt zone is produced quickly, and produced increasing degrees o f a re Figure 3.6a SEM images of the UV KrF excimer only, laser target interaction sites for fluences of 1, 2, and 3 J/ cm2, and cumulative laser pulses per site of 10, 100, and 1000. 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 1 J/cm 2 10 pulse 2 J/cm 2 10 pulse 3 J/cm 2 10 pulse 1 J/cm 2 100 pulse 2 J/cm 2 10 0 pulse 3 J/cm 2 100 pulse 1 J/cm 2 1000 pulse 2 J/cm 2 1000 pulse 3 J/cm 2 1 000 pulse

PAGE 72

52 solidified molten zone. It can be seen that 1000 pulses/site long order ripples or waves have started to form. 50 m 50 m 50 m 50 m 50 m 50 m 4 J/cm 2 1 0 pul se 5 J/cm 2 1 0 pulse 4 J/cm 2 1 00 pulse 5 J/cm 2 1 00 pulse 4 J/cm 2 1 00 0 pulse 5 J/cm 2 1 00 0 pulse Figure 3.6b SEM images of the UV KrF excimer only, laser target interaction sites for fluences of 4 and 5 J/ cm2, and cumulative laser pulses per site of 10, 100, and 1000.

PAGE 73

53 The effect of the cumulative laser ablation on the target composition was investigated by EDS after 1000 pulses for the different laser fluences. The results were normalized to germanium with an atomic percentage of 30% (Ge30), to be compared to the intrinsic Ba8Ga16Ge30 clathrate (Table 3 .5). The results for the un ablated target are given as reference. It can be seen that at low fluences, 1 and 2 J/ cm2, there is a significant deficiency of barium and slight excess gallium, relative to germanium and as compared to the original target and higher fluences. This is due to the incomplete melt of the target at low fluences, allowing the lower boiling point barium (2171 K) to be evaporated at a higher rate than the other components (Ga 2478 K, Ge 3107 K). Properties of the target components, barium, gallium, and germanium are listed in Appendix C This nonuniformity in evaporation can cause noncongruent expansion in the ablated plume and/or nonstoichiometric deposition on the film. The presence of the molten zone produced with higher laser fluences is more conducive for stoichiometric evaporation and this conclusion is supported by the energy dispersive spectroscopy (EDS) scans of the ablated spots on the targe t, which is presented Table 3.5 Laser target interaction site composition by EDS for excimer laser fluences 1 to 5 J/ cm2 after 1000 pulses. The intrinsic and un ablated target values are listed also. The Ga/Ge ratio is calculated for each. Ba Ga Ge Intrinsic 8.0 16.0 30.0 0.533 Target 8.8 0.4 15.8 0.7 30.0 0.9 0.53 0.04 1 J/cm26.4 0.4 17.5 0.4 30.0 0.7 0.58 0.01 2 J/cm27.0 0.2 16.6 0.6 30.0 0.6 0.55 0.01 3 J/cm27.3 0.1 16.2 0.5 30.0 0.7 0.54 0.01 4 J/cm27.3 0.2 16.1 0.5 30.0 0.8 0.54 0.01 5 J/cm27.7 0.5 16.0 0.6 30.0 0.7 0.53 0.01 Atomic % Sample Ga/Ge

PAGE 74

54 in Table 3 .5. At a 1 J/cm2 excimer laser fluence, as the ablation continues, the target had greater deficiency of the more volatile barium, signifying germanium enri chment. 3.4 Investigation of the CO2 Laser Target Interaction The laser target interaction was further studied by the CO2 laser target interaction and the single beam transient reflectivity The results presented here will be combined with the results of the excimer laser target interaction investigation (Chapter 3.3) for the investigation of the dual laser target interaction. CO2 laser pulses usually induce a thermal evaporation process [ 82], which can lead to no n congruent evaporation. Evaporation of the target material is not the goal of this investigation, whereas the purpose is to determine conditions to provide surface melting with minimal ablation. The SEM images (Figure 3.7) of the CO2 laser target interact ion site for an un ablated target, and at 1000 pulses each for fluences of 0.6, 1.3, 1.8, 2.0 and 2.4 J/ cm2, are shown in Figure 3.7. A fluence of 0.6 J/ cm2 produced slightly detectable melt of the smallest particles on the target surface, even after 1000 pulses. Increasing fluence through 2.4 J/ cm2 resulted in an increasingly partial melt of the target surface. The higher fluences produced a partial melt that re solidified into larger feature than the original target granule sizes. The laser fluences from 0.6 to 1.8 J/ cm2 did not produce detectable plasma generation nor ablation of the target. The higher fluence of 2.4 J/ cm2 was observed to produce strong plume production and emission, which is undesirable for use during dual laser ablation. The effect on t he target composition was investigated by EDS after 1000 pulses for the different IR laser fluences, as with the UV irradiation, shown in Table 3.6. It can

PAGE 75

55 be seen that there is no significant change in the percentage of the elemental composition of the t arget. There is a slight non significant increase in the barium content as the fluences increases, as compared to the bare target, possibly due to a higher absorption of far IR irradiation by the germanium. The appropriate CO2 laser energy to be utilized for dual laser ablation is determined by the amount of surface melt and ablation. The surface morphology investigation allows for the observation of a cumulative effect. However, the effect is not conveyed as a temporal evolution. Temporal synchronization is determined by the on set of the target surface melt measured by singlebeam transient reflectivity observations. The single beam transient target reflectivity method [ 49 ] is used to obtain information about target surface modifications by observing the temporal profile of the Figure 3.7 SEM images of the IR CO 2 laser only, laser target interaction sites at 1000 pulses per site for fluences of 0.6 J/ cm2, 1.3 J/ cm2, 1.8 J/ cm2, 2.0 J/ cm2, and 2.4 J/ cm2.

PAGE 76

56 reflected laser pulse as it interacts with the target. As the target melts, it become smoother and the surface is expected to result in the enhancement of the reflection, followed by a decrease due to absorption of the remainder of the pulse into the self ablated plume. This single beam configuration provides a simple way of monitoring the plume cutoff so that the time for the melting ( m) can be observed directly. Figure 3. 8 illustrate s the CO2 laser single beam transient reflectivity configuration. The target surface was polished to aid in the detector alignment since the intensity of the reflected portion of the beam is significantl y weaker than the monitor pulse. IR detector 1 was used to monitor the initial laser pulse shape directly, so that the scattered pulse collected by IR Detector 2 could be compared to the actual laser pulse directly for each individual laser pulse. Both of the detectors were high sensitivity HgCdTe infrared detectors to permit single shot detection for very weak probe levels. Table 3.6 Laser target interaction site composition by EDS fo r CO 2 laser fluences 0.6 to 2.4 J/ cm2 after 1000 pulses. The intrinsic balanced and un ablated target values are listed also. The Ga/Ge ratio is calculated for each. Ba Ga Ge Balanced 8.0 16.0 30.0 0.533 Target 8.80.4 15.80.7 30.00.9 0.530.04 0.6 J/cm28.80.4 15.70.9 30.00.9 0.520.04 1.3 J/cm28.80.5 15.70.7 30.01.1 0.520.04 1.8 J/cm28.80.2 15.70.6 30.00.9 0.520.04 2.0 J/cm28.80.6 15.70.9 30.00.8 0.520.04 2.4 J/cm29.10.1 160.9 30.01.1 0.530.05 Sample Atomic % Ga/Ge

PAGE 77

57 Consistent with results of the CO2 laser target interaction investigation, a CO2 laser fluence of 0.6 J/ cm2 did not result in appreci able melting of the target surface. This fluence was then chosen for alignment and calibration of the single beam transient measurements, which involves a degree of difficulty due to the low light level that reaches the reflection detector. The properly al igned and calibrated monitor and reflection pulses (Figure 3. 9) will over lap one another well with proper alignment, with only a Oscilloscope IR Detector 2 IR Detector 1 Target Vacuum Chamber CO 2 Laser B.S. Mirror ZnSe Lens ZnSe Lens Figure 3 8 The schematic diag ram of the experiment setup for the single beam transient target reflectivity measurements.

PAGE 78

58 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 100 200 300 400 500 600 700 800 900 1000 Time (ns) Normalized Voltage Monitor Reflection Figure 3.9 The monitor pulse before target interaction and reflection pulse after target interaction. The 0.6 J/ cm2 CO2 la ser fluence was used for alignment and calibration. Figu re 3 10 Representation of IR reflection pulse showing features: on set of melt, on set of ablation, and peak of the monitor pulse. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 100 200 300 400 500 600 700 800 900 1000 Time (ns) Normalized Voltage Monitor Reflection on-set of melt on-set of ablation peak of monitor pulse m

PAGE 79

59 reduced change in magnitude for the reflected pulse. Figure 3.1 0 schematically demonstrates the features of the reflected pulse that are noted; (i) on set of the target surface melt ( m), (ii) onset of ablation/evaporation of the molten material, in reference to (iii) the peak of the monitor pulse. The target was impinged by CO2 laser pulses of 1.8, 2.0 and 2.4 J/ cm2 and the reflection pulses were recorded concurrently with the monitor pulse (Figure 3 .1 1a c) respectively. Each recorded pulse is for a single pulse per target location. Each subsequent pulse was on an un ablated region of the target as to not record a cumulative effect. The results of 1.8 J/ cm2 (Figure 3.11a) show a late arrival and low onset of melt increase 115 ns before the peak, with little to no ablation contribution. The late onset of melt may not allow sufficient CO2 laser energy rem aining to interact with the excimer laser generated plasma. The 2.0 J/ cm2 fluence (Figure 3.1 1b) had a 145 ns time span before the pulse peak and weak ablation. The 2.4 J/ cm2 fluence (Figure 3.1 1c) had a 180 ns time difference, causing melting at the onset of the pulse and significant ablation, not desired in this process. The transient reflectivity results combined with the surface morphology and compositional results lead to 2.0 J/ cm2 CO2 laser fluence being the most promising candidate for use with dual laser ablation. When the excimer laser pulse is synchronized with the CO2 laser pulse during dual laser ablation, based on the above results the on set of the excimer laser pulse should occur at the same time as the on set of the melt ( m).

PAGE 80

60 Figure 3 11 Single beam reflectivity for CO 2 laser fluences of (a) 1.8 J/ cm 2 (b) 2.0 J/ cm2, and (c) 2.4 J/ cm2, before (Monitor) and after Reflection from the Ba8Ga16Ge30 target, displaying on set of melt. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 100 200 300 400 500 600 700 800 900 Time (ns) Normalized Voltage Monitor Reflection 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 100 200 300 400 500 600 700 800 900 Time (ns) Normalized Voltage Monitor Reflection 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 100 200 300 400 500 600 700 800 900 Time (ns) Normalized Voltage Monitor Reflection (a) ( b ) ( c ) m m m

PAGE 81

61 3. 5. Investigation of the Dual Laser Target Interaction Based upon the results of the 3 investigation of excimer laser target interaction (Section 3) and the investigation of CO2 laser target interaction (Section 3.4) moderate laser fluences were combined for the dual laser ablation technique. The low excimer laser fluence of 1 J/ cm2 was chosen as the excimer laser energy density. This investigation is to determine if the cumulative surface features and non congruent evaporation can be contro lled when coupled with the appropriate CO2 lase r energy. Referring to Section 3.3 Figure 3. 6a, after 10 laser pulses with 1 J/ cm2, the laser target interaction site had a smooth melt zone. The goal is to maintain this when coupled with the CO2 laser pulse. The results of the CO2 laser target interaction and singlebeam transient reflectivity studies lead to the selection of 2.0 J/ cm2 CO2 laser fluence, and a excimer laser pulse on set of approximately 6080 ns (100120 ns pp delay) (Figure 3.1 2). Figure 3 .1 2 The monitor oscilloscope traces of the UV excimer laser pulse and the IR CO2 laser pulse with a 100 110 ns pp delay. The on set of the melt is indicated at m. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 -100 0 100 200 300 400 500 600 700 800 Time (ns) Normalized Voltage UV IR 100 -110 ns p-p delay m

PAGE 82

62 SEM image s of the dual laser target interaction sites for 10, 100, and 1000 pulses/site for laser fluences of 1 J/ cm2 UV coupled with 2 J/ cm2 IR and a 100110 ns pp delay are shown in Figure 3.1 3. These parameters produced and maintained a smooth melt zone through all laser pulse/site conditions. This is similar to the high fluence single laser ablation results, but without the long order surface waves (see Figure 3. 6a and b). These dual l a ser target surface modification results are conducive for reduced particulate generation. It was observed that proper evaporation using the dual laser technique with these parameters is dependent on the target surface preparation. Polishing the target surface prior to ablation controlled the formation of conical features on the surface. The dual laser ablation was able to maintain a smooth melt feature when starting from a smooth target, but could not produce a smooth melt from a target surface that began with an un polished, granulated target surface (Figure 3.1 4). Rastering the laser spot location across the rotating target during long deposition runs can be used to also minimize feature formation. Spot rastering was performed to minimize the average laser pulses per site to Figure 3.13 Ima ges of the laser target interaction site s for 10, 100, and 1000 pulses/site using dual laser f luences of 1 J/cm2 UV and 2 J/cm2 IR with 100 110 ns pp delay 50 m 50 m 50 m Dual 100 pulse Dual 1000 pulse Dual 10 pulse

PAGE 83

63 have a quantity of pulses/site which was comparable to other single laser fluences during film deposition investigations. The effect on the target composition was investigated by EDS after 1000 pulses per site for the different laser fluences, as shown in Table 3.7. The dual laser results are shown compared to the excimer only laser fluence of 1 J/ cm2. The dual laser was also compared to the excimer only single laser fluence of 3 J/ cm2 since the melt zones had a similar appearance. The results were normalized to germanium with an atomic percentage of 30% (Ge3 0), to be standardized to the Ba8Ga16Ge30 clathrate. The result for the un ablated target was not irradiated by the pulsed laser. The dual laser ablation resulted in Figure 3 .1 4 SEM i mages of dual laser ablation of an unpolished target. A region of un ablated target surface is shown and the laser target interaction site s for 10, 100, and 1000 pulses/site using dual laser f luen ces of 1 J/cm2 UV and 2 J/cm2 IR with 100 ns pp delay 50 m Un-ablated Target Dual 10 pulse 50 m Dual 10 0 pulse 50 m Dual 1 000 pulse 50 m

PAGE 84

64 near stoichiometric evaporation and a significant improvement as compared to the excimer o nly laser fluence of 1 J/ cm2. 3. 6. Summary Based upon general qualities of type I clathrate Ba8Ga16Ge30, high quality thin film deposition by PLD may incur some challenges. These qualities include the complex clathrate crystal structure, multi component material with widely varied physical properties, and coldpressed powder target as the source. Investigation of the target surface modification by pulsed laser ablation was undertaken to gain understanding of the parameters that can be controlled to provide ablation conditions that may be conducive for quality thin film deposition. The target surface irradiation using KrF excimer only single laser ablation revealed that use of low fluences produced conical surface modifications and non congruent evaporation of the material. High fluence ablation produced near congruent Table 3 7 The actual composition and Ga/Ge ratio of the un ablated target and laser target i nteraction sites after 1000 pulses per site for dual laser 1 J/cm2 UV coupled with 2 J/cm2 IR with 100 110 ns pp delay. The results for excimer only single laser 3 J/cm2 (S3) and 1 J/cm2 (S1) are included for comparison Ba Ga Ge Balanced 8.0 16.0 30.0 0.533 Target 8.60.4 15.60.2 30.00.2 0.520.01 Dual 7.60.1 15.70.2 30.00.1 0.520.01 S37.30.1 15.90.5 30.00.7 0.530.01 S16.50.4 17.60.7 30.00.9 0.590.03 Atomic % Sample Ga/Ge

PAGE 85

65 evaporation with smooth laser target melt zone, which is conducive for good film deposition. However, fluences greater than 3 J/ cm2 do not appear to result in an increased benefit. The laser target interaction was also investigated using the dual laser ablation technique. Based on the excimer only investigations and CO2 laser only results, it was found that conducive evaporation conditions could be achieved using low to moderate lase r fluences. Coupling 2 J/ cm2 CO2 laser fluence with the 1 J/ cm2 excimer laser generated plasma produced results comparable to those of high fluence excimer only results. The outcome of this surface modification investigation have provided the ablation cond itions that can now be used during insitu optical emission studies of the laser generated plasma. Optical diagnostic studies can further the understanding of the evaporation processes that occur during pulsed laser ablation of this particular material.

PAGE 86

66 CHAPTER 4: IN SITU OPTICAL DIAGOI STIC S During the laser target interaction in pulsed laser ablation the evaporated material forms a plasma whose ion density is determined by the laser parameters. The plasma contains both multiply ionized species and highly excited species. Subsequent expansion of the plasma and collisional excitation of the ionized and excited species extend the visible emission in a plume of material traveling from the target to the substrate. In situ optical diagnostics provides a n on invasive probe to investigate atoms, ions, and molecules within the laser generated plasma without interfering with the ablation process. It can yield information about properties such as excited state species densities, collisional effects, and distrib ution of species to name a few. These details are utilized to develop a physical understanding about the plasma dynamics that affect the growth of stoichiometric Ba8Ga16Ge30 film s A number of diagnostic techniques have been developed to assist in the char acterization of plume dynamics such as spectroscopic analysis and timegated imaging. Optical emission spectroscopy (OES) is well suited to yield information about PLD ablation plumes. Emission spectroscopy is used to identify the species in the laser abl ated plume from referenced atomic lines [ 83]. The emission intensity is very high during the initial stages of plume expansion corresponding to a continuum of radiation through the visible spectrum. The plasma appears white within the first millimeter of

PAGE 87

67 e xpansion. After expansion to distances of a few millimeters, PLD plumes exhibit an assortment of atomic and ionic lines. The spectral intensities change dramatically with time as the atoms cool and deexcite. Imaging techniques are extremely useful for studies of ablation plumes in vacuum by providing two dimensional snapshots of the three dimensional plume propagation and dynamics. This capability becomes essential for understanding of the plume propagation. Intensified charge coupled device array (ICCD) c ameras allow high speed photography with the use of electronic gates that can be used to dynamically image the plume emission [ 84, 83]. The light emitted by the excited sample can be spectrally resolved through the use of notch filters to yield qualitative and quantitative information on the elemental constituents of the sample. In a typical optical emission diagnostics system the plasma plume is imaged outside of the vacuum chamber using rel evant optical elements (Figure 4 .1). A specific object plane alon g the axis of the plume is image d onto a plane which is incident to optical fiber s or directly onto the detector of the camera. This arrangement allows collection of light from a specific position a plane or points on that plane with in the plasma while c ontributions from other positions are minimized, as shown in Figure 4 1. O ptical plasma emission measurements may be time gated dependent upon the information desired. Time integrated data collection from the entire laser generated plume volume is adequat e for species identification or expansion profiles. This involve s gating on the order of micro seconds or longer However, spatially and temporally resolved measurements are necessary to infer the timeof flight (TOF) and local populations. T ime resolved m easurements require gating on the order of nano seconds.

PAGE 88

68 Key parameters for determining the quality of a growing film are the kinetic energy of the species and the stoichiometry of the components [ 85] Optical determination of the time of flight (TOF), a nd in turn the velocity, is an attractive method for estimating the energy of the species [86] Insight into stoichiometric deposition is examined by comparative species expansion profiles. The OES and ICCD imaging techniques were utilized for characterization of the pulsed laser generated plasma. 4.1 Experimental Details The experimental set up consists of the dual laser ablation system described in Chapter 2.3 with the integration of the optical diagnostics system (Figure 4.2a b). The center piece of th e in situ optical diagnostics is the high speed ICCD camera. The ICCD Figure 4 .1. Diagram of the imaging system identifying the coordinate system, object plane, image plane, and an example light collecting device (opt ical fiber). The magnification is for this configuration. x z y x' z' y' Spectrometer O bject P lane Vacuum Chamber Imaging Window Lens System Image P lane S 3 J/cm 2 G Target 2 cm

PAGE 89

69 Figure 4 2 The schematic diagram s of the dual laser deposition system with the (a) optical emission spectroscopy (OES), and (b) intensified chargecoupled device (ICCD) imaging system included. UV Detector Excimer Laser CO 2 Laser IR Detector Oscilloscope Rotating Target Vacuum Chamber Substrate Heater Lens System Digital Delay Generator ICCD 10.00ns Sync Out Trig In Trig In ICCD Sync Control PC (b) (a) UV Detector Excimer Laser CO2 Laser IR Detector Oscilloscope Rotating Target Vacuum Chamber S ubstrate Heater 10.00ns OMA System Lens System Digital Delay Generator 0 500 1000 1500 2000 2500 3000 400 450 500 550 600 650 700 750 800 Wavelength (nm) Intensity (arb) Optical Fiber Spectrograph ICCD Sync Out Trig In Trig In

PAGE 90

70 camera was interchangeable between the two configurations, OES (Figure 4.2a) and ICCD imaging (Figure 4.2b). The camera was part of a PI Acton PI MAX:512 UNIGEN Digital ICCD Camera System, with 512 512 pixels (19 19 m pixel active area), 12.4 12.4 mm image area, and < 5 ns gating capable (see Appendix D ) The gating capability allows for high speed, short time frame exposure to be used for laser generated p lume analysis. The spectral sensitivity of the camera ranges from 150 nm (UV) to 925 nm (Figure 4.3). The timing of the camera is controllable by the Programmable Timing Generator (PTG). The spectrometer (Appendix E ) used in the OES experiments, was a 50 0 mm focal length, Czerny Turner type, with 1:1 magnification, and t riple i ndexable gratings ( 150, 600, and 1200 grooves/mm ), adjustable entrance slit set to 10 m, and resolution to Figure 4 .3. Specification of the wavel ength sensitivity of the ICCD camera. The wavelength cut off due to the imaging window is shown at 350 nm. 0 5 10 15 20 25 30 35 40 45 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 Wavelength (nm) Quantum Effiency (%) Imaging window w avelength cut -off

PAGE 91

71 0.05 nm. The optical fiber bundle (Appendix F ) was a three fiber bundle w ith 200 m diameter fibers (~245 m dia. with cladding). The spectrometer was joined with the ICCD camera to produce broad range wavelength spectrums of the emission, as in an optical multichannel analyzer (OMA) system, also referred to as time integrate spaceresolved laser induced plasma spectroscopy (TISR LIPS) [ 87]. The camera, spectrometer, and PTG were controlled by the WinSpec/WinView softwa re (version 2.5.21.0, ). The software is for real time acquisition, display, and data processing operated unde r Microsoft Windows 2000/XP. The OES and ICCD imagi ng standard operating procedure for data collection and analysis is given in Appendix G The imaging system was a 10.5 cm focal length lens set, which was positioned perpendicular to the plume axis. The op tic axis of the lens system was centered 2 cm from the face of the target, see Figure 4.1. The axis of the luminous laser generated plume (object) was imaged onto a plane outside the deposition chamber. A magnification of was produced for use during OES The fiber was able to be translated in the x y plane of the image plane. During ICCD imaging, the lens position was adjusted to produce a 1/4 magnification and the camera was positioned so that the image plane formed on the front surface of the cameras ICCD detector. The camera was mounted on a stage with y z axis adjustment to facilitate proper focus of the image and positioning of the image on the detector of the camera. The entire target substrate distance of 4 cm was detectable without requiring mov ement of the camera. The system is aligned and calibrated by placing a standard source (helium lamp) on axis in place of the target and substrate. The lamp was masked to simulate the 4 cm target substrate distance. This alignment was performed to minimize effects due to

PAGE 92

72 distortion and aberration caused by the optical elements of the system such as windows, lenses, filters, and fiber optics. The position of the lens was adjusted to minimize image distortion losses, and position of the optical fiber or camer a detector was defined relative to the aligned image plane. Figure 4.4a shows an example of the properly aligned image system and Figure 4.4b is an example of the distortion that can result due to misalignment. The He lamp was translated vertically along t he object plane (Figure 4. 5) and He spectrums were recorded by OES at various x y positions. The optical fiber position was translated horizontally (0 to 40 mm) and also vertically ( 30 mm to 30 mm) as the lamp was moved. The normalized intensity of the 58 7.6 nm wavelength line of He, as a function of the x y plane position of the op tical fiber is shown in Figure 4 6. This gives a representation of the field of view of the imaging system without resulting in losses due to distortion. The alignment allows qu ality image rendition within the needed Figure 4.4 Photographs of the alignment and calibration procedure using a 4 cm long helium lamp as the object (not shown) placed along the plume axis of propagation and imaged at the light collection plane. Photograph (a) is an example of proper object/lens alignment and (b) is misaligned displaying distortion of the image. Optical Fiber He Lamp Image Distortion (a) (b)

PAGE 93

73 0 40 cm horizontal range from 20 to 20 cm vertically. The spectrometer and data collection software were also calibrated in this fashion to account for contributions due to the optical system components. The image magnification and orientation were determined by placing a scale card between the target and substrate holder (Figure 4. 7). The scale card is aligned along the plume axis then back lit and focused onto the image Figure 4 5 Schematic diagram of the imaging system alignment and c alibration layout using a helium lamp masked for 4 cm length. The diagram is drawn to scale. Substrate Holder He Lamp Position(s) Target Optic Fiber Position(s) ( 0 0 ) Lens Dia. Chamber Window x y

PAGE 94

74 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 0 5 10 15 20 25 30 35 40 45 Horizontal Position (mm) Intensity (arb.) -30mm -20mm -10mm 0mm 10mm 20mm 30mm Figure 4.6 Intensity of the He lamp (587.6 nm wavelength) as the optical fiber position was varied along the x axis (horizontal) and y axis (vertical) of the image plane. Figure 4 .7. The (a) back lit scale card used to determine image magnification and orientation for the optical diagnostics system(s), and (b) a photograph of the imaged card. y x 1 cm 0 1 2 3 4 Target Holder Image 1cm x y (a) (b )

PAGE 95

75 plane. The timing of the in situ optica l diagnostics system was triggered ( t0) simultaneously with the excimer laser via the digital delay generator. The camera delay and gatewidth were controlled by the t0 triggered PTG. 4. 2. ICCD Imaging 4.2.1 Dual Laser Synchronization by In Situ Optical Diagnostics The laser generated plume emission and expansion is observed to increase dramatically when the excimer laser generated plasma is efficiently coupled with the CO2 laser energy. When the synchronization is optimized for given laser parameters th e plume ionization, plume expansion, and particle reduction are enhanced [ 48, 69]. The proper temporal synchronization of the dual laser inter pulse delay can be determined by the maximum expansion and emission of the plume when optimum energy coupling has occurred. To illustrate the CO2 laser only interaction with the target, timeintegrated ICCD images of laser ablation at various CO2 laser fluences were recorded (Figure 4 .8). A fluence of 1.8 J/ cm2 (Figure 4.8a) results in visible thermionic emission; h owever there is no visible emission from a plasma or plume formation. 2.0 J/ cm2 (Figure 4.8b) created weak visible plume emission, while 2.4 J/ cm2 (Figure 4.8c) produced strong plasma and plume formation. The dramatic increase in the ablation with the 2.4 J/ cm2 as compared to the low er fluences is shown in Figure 4.9 by time resolved emission imaged at the target surface. These results support the on set of the melt studies by single beam tra nsient reflectivity in Chapter 3 Section 4. Although the CO2 laser fluence of 2.4 J/ cm2 independently produces strong ablation and emission, but when coupled with the excimer

PAGE 96

76 laser generated plasma it would result in excessive target surface temperatures causing increasing evaporation and particulate production and would be counter productive. The ICCD imaging configuration, Figure 4.2b, of the insitu optical diagnostics system was utilized for the following results, as well as the object plane of the plume 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 0 50 100 150 200 250 300 Time (ns) Normalized Intensity (arb) 2.4J 2.0J 1.8J Figure 4.9 The normalized time resolved total emission intensity of the CO 2 laser target interactions for 1.8 J/ cm2, 2.0 J/ cm2, and 2.4 J/ cm2 The emission was imaged at the target surface (0 mm position). (b ) 5 mm Figure 4.8 Time integrated ICCD images of CO 2 laser target interactions with laser fluences of (a) 1.8 J/ cm2, (b) 2.0 J/ cm2, and (c) 2.4 J/ cm2. Images are norma lized with respect to image (c). (a) 5 mm (c ) 5 mm

PAGE 97

77 described in Figure 4.1. The length of the viewable area imaged in Figure 4.10 is approximately 48 mm. The following results presented in this section utilized time integrated gating to capture all the light emitted during the ablation process. The total intensity of the plume emission was recorded for the singlela ser low fluence of 1 J/cm2 and again when coupled with the addition of the 2 J/cm2 IR CO2 laser pulse at various peak to peak ( pp ) delays. Examples of single laser and dual laser generated plasmas are shown is Figure 4.10a and b, respectively. Also, examp les of the in situ monitored laser pulse delays for 25 ns, 100 ns, and 175 ns pp delays are shown in Figure 4 .11. The ICCD camera system was synchronized to the triggering of the excimer laser and recorded each event for 20 s to ensure that the entire p lume emission was captured (timeintegrated). The pp delay was varied from 50 ns to 200 ns, relative Figure 4.10 ICCD imaging of total emission of ablated plumes from the target for (a) 1 J/cm2 and (b) dual laser ablation. (a) Dual Ge + 481 nm

PAGE 98

78 to the peak of the excimer laser pulse at 0 ns. The on axis intensity (Figure 4.12.a), and crosssectional FWHM (Figure 4.12.b) of the plume was measured 250 m from the target surface. This observation position is beyond the edge of the laser induced plasma, within the 2 3 mm laser spot size, and before significant plume expansion and ion recombination has occurred. The cross sectional expansion profi les were compared (Figure 4.12) and the on axis intensity of the dual laser induced plume was observed to minimize as the FWHM maximized. Maximum plume expansion is indicative of optimum energy coupling of the CO2 laser pulse into the excimer laser pulse g enerated plasma as the plume becomes less forward directed and broader expansion occurs due to the higher temperatures of the plasma which can result in greater species ionization [48]. The minimum on axis intensity occurred when the pp inter pulse delay was between 105125 ns, and the maximum cross sectional FWHM occurred between 80 105 ns. It is 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 -200 -100 0 100 200 300 400 500 600 Time (ns) Normalized Voltage UV 0ns IR -25ns IR 100ns IR 175ns 0 -25 100 175 Figure 4.11 Overlapped oscilloscope traces for the excimer laser pulse (UV) and CO 2 laser pulse s (IR) for peak to peak delays of 25 ns, +100 ns, and +175 ns, relative to the UV pulse at 0 ns.

PAGE 99

79 evident from Figure 4.10b, with the 100 ns delay, that the plume expansion is dramatically greater for the dual laser emission as compared to the excimer only s ingle laser expansion. 0.6 0.7 0.8 0.9 1.0 1.1 -75 -25 25 75 125 175 225 Peak-To-Peak Delay (ns) Normalized Intensity Dual Single (b) 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 -75 -25 25 75 125 175 225 Peak-to-Peak Delay (ns) Normalized FWHM Dual Single (a) Figure 4.12. Dual laser synchronization parameters as a function of the dual laser pulse peak to peak delays obtained by total emission ICCD imaging of the laser induced plasma expansion. The (a) normalized peak intensity and (b) normalized FWHM are shown.

PAGE 100

80 4.2.2 Elemental Expansion Profiles To investigate the expansion and species ionization, the laser generated plumes were imaged with the addition of notch filtering. Wavelength filters were individually placed i n the optical path before the detector of the ICCD camera. The optical filters were chosen based on NIST emission spectra tables as strong lines and not overlapping or near other emission lines. The filters for singly ionized barium (Ba+ 455 nm), gallium ( Ga+ 426 nm), and germanium (Ge+ 481 nm) were implemented in this study. The ICCD images of the laser generated plumes for singlelaser ablation low fluence (1 J/ cm2 UV), single laser ablation high fluence (3 J/ cm2 UV), and dual laser ablation (1 J/ cm2 UV + 2 J/ cm2 IR), res pectively, are shown in Figure 4 .13. The results displayed also include unfiltered emission (total emission of the plume), allowing the contribution from the entire detectable spectrum to be recorded (350 nm 925 nm). The pseudo color int ensity of the images is normalized relative to each other. The emission from the excimer only low fluence configuration was very weak for all three element wavelengths, as compared to the singlelaser high fluence and dual laser ablation results. It resu lts from coupling the CO2 laser energy into the low fluence excimer laser generated plasma can be clearly seen. The emission is greatly increased, as is the expansion. As a comparison, the dual laser expansion is significantly greater than the high fluence single laser ablation. The angular distribution of the OES cross section of ablated plumes 5 mm from the target was derived from the wave length filtered images (Figure 4.14). To compare all three laser configurations, the 5 mm on axis distance was chosen because beyond this distance the emission from the low fluence single laser ablation was depleted below

PAGE 101

81 detectable values for the parameters utilized here. The emission shown is for Ba+ 455 nm, Ga+ 426 nm and Ge+ 481 nm. The species dependent intensity pr ofiles for (a) 1 J/cm2 and (b) 3 J/cm2 single laser, and (c) dual laser ablation, are shown in Figure 4 .1 4a c The results of the full width at half maximum (FWHM) of the OES profiles are summarized Figure 4 13 ICCD imaging of the time integrated laser generated plume emission for excime r only 1 J/ cm2 (S1), high fluence single laser 3 J/ cm2 (S3), and dual laser ablation. For each laser configuration the unfiltered total emission, Ba+ 455 nm, Ga+ 426 nm, and Ge + 481 nm are shown. S 1J/cm 2 Total S 3J/cm 2 Total Dual Total S 1 J/cm 2 Ba + 455 nm S 1 J/cm 2 Ga + 426 nm S 1 J/cm 2 Ge + 481 nm S 3 J/cm 2 Ga + 426 nm S 3 J/cm 2 Ba + 455 nm Dual Ba + 455 nm Dual Ga + 426 nm Dual Ge + 481 nm S 3 J/cm 2 Ge + 481 n m

PAGE 102

82 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 -80 -60 -40 -20 0 20 40 60 80 Angle (deg) Norm. Intensity (arb.) Ba Ga Ge 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 -80 -60 -40 -20 0 20 40 60 80 Angle (deg) Norm. Intensity (arb.) Ba Ga Ge 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 -80 -60 -40 -20 0 20 40 60 80 Angle (deg) Norm. Intensity (arb.) Ba Ga Ge Figure 4.14 OES cross section of ablated plumes 5 mm from the target for (a) 1 J/cm 2 and (b) 3 J/cm2 single laser, and (c) dual laser ablation. The emission shown is for Ba+ 455 nm Ga+ 426 nm and Ge+ 481 nm. (a) (b) (c)

PAGE 103

83 in Table 4.1. The expansion f or the dual laser ablation generated plume is significantly broader and more congruent, as compared to either singlelaser configuration. The average FWHM expansion of the dual laser ablated plume is 1.7 to 3.1 times greater than the single laser plumes. The optical emission from the Ba8Ga16Ge30 plume demonstrates that the coupling of the CO2 laser into the excimer laser generated plume causes both significant ionic excitation in addition to lateral plume expansion, as previously demonstrated with other ma terials [ 48, 68]. The highly ionized material and congruent expansion profiles of the dual laser process are desirable for quality film depositions, leading to stoichiometric deposition of the material. Congruent evaporation is one of the important advanta ges of pulsed laser evaporation (pulse widths < 1 s) for deposition of multi component materials [ 53]. There are exceptions, however, where evaporation is incongruent even under pulsed laser heating [ 90, 91], such as a volatile element component of the co mpound causing preferential evaporation leading to non stoichiometric depositions. The lower boiling point of the barium has been found to be problematic for a laser energy density (1 J/ cm2) near the excimer laser ablation threshold for Ba8Ga16Ge30, howeve r, this is circumvented with the utilization of dual laser ablation. Ba Ga Ge S129.4 29.4 22.3 S333.4 41.2 37.3 DL67.4 66.5 69.6 FWHM (deg) Table 4 .1. The full width at half maximum (FWHM) of the OES cross section profile of ablated plumes 5 mm from the target for (a) 1 J/cm2 and (b) 3 J/cm2 single l aser, and (c) dual laser ablation. ( 1.5 degree)

PAGE 104

84 4.3 Optical Emission Spectroscopy OES is primarily used for identification of the species in the laser generated plume from tabulated atomic lines and molecular bands [ 84]. Time integr ated spectra from the entire plume volume are adequate for this task. However, spatially and temporally resolved measurements are necessary to infer the timeof flight (TOF) and local populations. After expansion to distances of a few millimeters, PLD plum es generally exhibit a rich assortment of atomic and ionic lines. Because the interpretation of OES observations is dependent on the source of emission and the understanding of the physical processes occurring within the source, care must be taken in the i nterpretation of these observations [92 ]. The laser generated plume was characterized by viewing the optical emission of the plume perpendicular to t he axis of propagation (Figure 4.1). The spectra of the plume were obtained using the OES configuration of the in situ opti cal diagnostics system, Figure 4.2a. The optical fiber was positioned to collect light which was imaged 2 cm away from the target surface, along the plume/optic axis. The spectrum was started at 400 nm due to the complete transmission cut o ff at 350 nm from the deposition chamber imaging window (Borofloat). The time integrated spectra were collected for the excimer only portion (1 J/ cm2), and dual laser 100 ns pp delay (1 J/ cm2 UV + 2 J/ cm2 IR) ablation, Figures 4 .15a and b, respectively. T he spectra have shown on the same scale for comparison and show the relative emission of the excited species. The neutral (I) and singly (II) ionized elements are identified fo r barium, gallium, and germanium [ 93] The increased singly ionized intensities of the dual laser configuration can be clearly seen, demonstrating the effect of coupling the energy of the CO2 laser pulse with the excimer

PAGE 105

85 laser generated plasma. U nder the optimum conditions highly excited and ionized plumes can be produced while usin g moderate laser fluences. The higher ionization leads 0 5000 10000 15000 20000 25000 30000 35000 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 Wavelength (nm) Intensity (arb.) Ba I Ba II Ga I Ga II Ge II 0 5000 10000 15000 20000 25000 30000 35000 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 Wavelength (nm) Intensity (arb.) Ba I Ba II Ga I Ga II Ge II Figure 4.15 The time integrated spectra for (a) single laser 1 J/ cm 2 fluence and (b) dual lasers abl ated plumes. The neutral (I) and singly ionized (II) species of Ba, Ga, and Ge are indicated.

PAGE 106

86 to generation of energetic species and their impingement on the film growth surface plays an important role in the nucleation structure of high quality thin film growth [53 ]. While ablating the rotat ing target with low excimer laser fluence, it was observed that the optical emission intensity decreased as the total laser pulses accumulated. The time integrated intensity for the visible spectrum (400 nm to 715 nm) was recorded on axis, 2 cm from the ta rget, as a function of cumulative laser pulses. The normalized intensity for single and dual laser PLD is shown in Figure 4.16. As the laser pulses accumulate, the intensity decreases for the singlelaser PLD process but remains constant for the dual laser PLD. This effect may be due to the changing target surface topography from the impinging excimer laser pulses for low fluence ablation, as seen in Chapter 2, causing reduced absorption of the laser energy into the target surface. As for the dual laser pro cess, the target surface topography is smoother as a result of the more complete melt zone, providing a more consistent plume as the ablation of the target progresses. 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 0 1000 2000 3000 4000 5000 6000 Pulses Norm. Intensity Single Dual Figure 4 1 6. The total intensity of the time integrated optical emission obtained 2 cm from the target surface for cumulative laser pulse for single and dual laser ablation.

PAGE 107

87 Spatially and temporally resolved measurements were utilized to measure timeof flight ( TOF) of the transient plume for determination of the velocity of the ablated material. This velocity is proportional to the kinetic energy of the emitting material. The emission temporal profiles were recorded for the laser generated transient plumes for 1 J/ cm2 and 3 J/ cm2 excimer single laser ablation and dual laser ablation. The optical fiber was translated along the optical axis of the image plane a distances of 0 mm, 1, mm, 2 mm, 4 mm, and 6 mm, for each laser condition. The peak of the 0 mm position T OF data is utilized as a reference point to calibrate the distance and time to equal to zero. The smallest feasible camera time gate was used for each condition in order to measure the emission for only a small increment of time at a given distance from th e target, while still being able to measure appropriate emission intensity. This will aid increasing the signal to noise ratio at larger distances from the target and lower ablation energies. Due to the low intensity emission of the low fluence single lase r ablation, time gates of 10 ns (at 0 mm) to 250 ns (at 6 mm) were required. The high fluence singlelaser was gated at 10 ns each, while the dual laser was gated at 20 ns each. At each position, the delay of the timegate is increased from 0 to 1.5 s in steps equal to the gatewidth, in order to record the complete time evolution of the population passing that particular point. An example of a time gated TOF profile for a plume pulse arriving at a point 2 mm on axis from the target surface with a 10 ns gate is shown in Figure 4 .17. The peak of the intensity profile corresponds to the bulk of the light emitting material passing the imaged position. In the given example the peak on material passes the 2 mm point of interest, approximately 235 ns after the arrival of the laser pulse. This data is used to determine the velocity of the bulk of material and is referred as the TOF. The TOF

PAGE 108

88 profiles for the thre e laser configurations (Figure 4 .18a c) are intensity normalized for each individual profile. The low f luence sin gle laser TOF profiles (Figure 4 .18a) occur later in time and are significantly broader than the profiles for high fluence single la ser and the dual laser, Figure 4.18b and 4.18c, respectively. The time at which the peak intensity occurs is used to determine the TOF of the emitting material. The distance as a function of the peak time was plotted (Figure 4.19) and the slope was used to determine the velocity of the emitted material for each laser configuration. The material accelerates quickly in the initial stages of the plasma to a constant velocity. The low fluence singlelaser ablation resulted in the lowest velocity, 0.63 cm/s, and the broadest temporal FWHM (Figure 4.20). The high fluence single laser and dual laser ablation resulted in high er velocities, 0.90 and 1.56 cm/s, respectively. However, the temporal FWHM for these two conditions were equal. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 0 100 200 300 400 500 600 700 800 900 1000 Delay (ns) Normalized Intensity (arb) Figure 4.17 An example of a time of flight (TOF) p rofile positioned at 2 mm from the target surface.

PAGE 109

89 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 0 250 500 750 1000 1250 1500 Time (ns) Normalized Intensity (arb) 0mm 1mm 2mm 4mm 6mm 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 0 250 500 750 1000 1250 1500 Time (ns) Normalized Intensity (arb) 0mm 1mm 2mm 4mm 6mm 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 0 250 500 750 1000 1250 1500 Time (ns) Normalized Intensity (arb) 0mm 1mm 2mm 4mm 6mm Figure 4 .1 8 The normalized intensity, time resolved, total emissi on time of flight profiles for positions 0 mm, 1 mm, 2 mm, 4 mm, and 6 mm from the target surface. The profiles were imaged along the plume axis for single laser (a) 1 J/ cm2, and (b) 3 J/ cm2 and (c) dual laser ablation configurations. The dual laser pp delay was 100 ns.

PAGE 110

90 Figure 4.19 The distance time plot of the TOF peak arrival time at various distances from the target surface, for singlelaser 1 J/ cm2 and 3 J/ cm2, and dual laser 1 J/ cm2 UV + 2 J/ cm2 IR. The slopes were velocity determination. y = 0.6335x 0.0236 R2 = 0.9896 y = 0.8958x 0.0025 R2 = 0.9946 y = 1.563x 0.0276 R2 = 0.989 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Distance (cm) S 1J S 3J Dual Figure 4.20 The temporal FWHM of the TOF profiles for single laser 1 J/ cm 2 and 3 J/ cm2, and for the dual laser 1 J/ cm2 UV + 2 J/ cm2 IR. 100 200 300 400 500 600 700 800 900 1000 1100 1200 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Distance (cm) FWHM (ns) S 1J S 3J Dual

PAGE 111

91 The velocities of the individual elements were investigated to determine the difference b etween the velocities of the different species. Due to the low emission of Ga and Ge, relative to that of Ba, these velocities could not be resolved with certain accuracy (Figure 4.21). This has been observed elsewhere with other materials using laser ablation in vacuum, where individual species velocities have been found to not to be as expected [ 86]. 4.4. Summary Through the use of insitu optical diagnostic techniques the time gated and timeintegrated plasma emission was investigated as a technique to facilitate the synchronization of the two lasers for dual laser ablation The optimum temporal pp 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 200 300 400 500 600 700 800 900 Time (ns) Normalized Intensity (arb) Ba+ 455nm Ga+ 426nm Ge+ 481nm Figure 4.21 The time resolved elemental TOF profiles for the ions Ba + 455 nm, Ga + 426 nm, and Ge+ 481 nm at 4 mm from the target surface.

PAGE 112

92 delay was able to be determined for the coupling of the 1 J/ cm2 excimer laser generated plasma with the 2 J/ cm2 CO2 laser energy. Similar resul ts were obtained when investigated by the laser target interaction; however, the use of optical diagnostic techniques is a desirable in situ process. Results of the optical emission studies, thought the use of both ICCD imaging and OES, yielded insight in to the nature of the ablated target material. The plasma expansion profiles found that the dual laser technique produced broader and congruent expansion of the species which is conducive to improved film distribution and stoichiometric deposition The OES studies demonstrate d the higher species ionization and TOF also for the dual laser technique This is advantageous for producing high quality films with good adhesion and crystal structure.

PAGE 113

93 CHAPTER 5: HYDRODYNAMIC MODELING OF PLASMA EXPANSION Results of hydrodynamic modeling of the laser generated plasma expansion due to single laser and dual laser ablation of the Ba8Ga16Ge30 target were evaluated. The interaction of the high power la ser pu lses with bulk targets results in evaporation, plasma formation, and deposition of films. The mechanisms involved in the plasma expansion process were investigated by utilizing a theoretical model developed by Singh et al. [ 94] for simulating laser p lasma solid interact ions. T he model is based on gas dynamic equations where the laser generated plasma is treated as an ideal gas at high pressure and temperature. This plasma volume is initially confined to the small dimensions of the laser spot size and plasma thickness and then is suddenly allowed to expand in vacuum. T h e process can be classified into three regimes (Figure 5.1) [ 95] : (i) interaction of the laser beam with the bulk material resulting in evaporation of the surface layers (evaporation regime) (ii) interaction of the laser beam with the evaporated material leading to formation of a high temperature isothermal expanding p lasma consisting of positively cha rged and neutral particles, mole cules, and atoms (isothermal regime), and (iii) the an isotropic threedimensional adiabatic expansion of the laser generated plasma and deposition of films (adiabatic regime) The first two regimes occur during the laser pulse duration ( t ) while the last initiates after the laser pulse terminates ( t > ) The evaporation of the target is assumed to be thermal

PAGE 114

94 in nature u nder pulsed laser evaporation conditions, while the plasma expansion dynamics is nonthermal as a result of interacti ons of the laser beam with the evaporated material. The gas dynamic equations simulate the expansion of the plasma in the last two regimes. The solution of the equations shows the expa nsion velocities of the plasma are related to its initial dimensions and temperature, and the atomic weight of the species. The model has been shown to predict characteristic features of pulse laser evaporation and deposition of films [ 95] The characteristics include (a) the pulse energy density effect on atomic velocities, (b) the forward directed nature of the plasma and its dependence on energy density, (c) spatial compositional variations in multi component thin films as a function of energy density, (d) dependence of the atomic velocities with atomic weights, (e) veloci ty distribution of the atomic and molecular species, and (f) thickness and compositional variations as a function of substratetarget distance and irradiated spot size. Figure 5.1 Schematic diagrams showing the three regimes during the pulsed laser irradiation of a bulk target: the (a) evaporation regime (0 < t < 100s ps), (b) isothermal regime (100s ps < t ), and (c) the adiabatic regime ( t > ). (a) ( c ) ( b )

PAGE 115

95 The model was initially developed for simulation of single excimer laser evaporation [ 94] but was modified for the dual laser ablation technique [ 68]. The modified plasma model has been successfully used to simulate the film thickness profiles plasma thickness [ 97], and species expansion profiles of multi component materials [ 98] for singl e laser and dual laser techniques. Under dual laser ablation, the addition of the synchronized but longer pulse duration, CO2 laser required the consideration of another isothermal regime during the initial stages of the plasma generation. For the results presented here, plasma model was applied to the time of flight ( TOF ) profiles of the single and dual laser ablated plasmas presented in Chapter 4. The model was fitted to the experimental data through computation determination of plasma expansion parameters such as initial plasma thickness, temperature, and the expansion velocities 5.1 Isothermal and Adiabatic Expansion Model The rapid expansion of the laser generated plasma in vacuum results from large density gradients. The plasma which is absorbing the laser energy can be simulated as a high temperature high pressure gas. In the initial stages of the plasma expansion the particle density is on the order of 10191020 cm3, so the mean free path of the particles is short and the plasma behaves as a flu id. The equations of gas dynamics can be applied to simulate its expansion. In the theoretical model, the density and pressure profiles of the plasma are assumed to show an exponential decrease with distance from the target surface. The density ( n) of the plasma at any given point ( x,y,z ) at time t can be expressed by [ 94];

PAGE 116

96 2 2 2 2 2 2) ( 2 ) ( 2 ) ( 2) ( ) ( ) ( ) , (t Z z t Y y t X x oe t Z t Y t X n t t z y x n ( 5.1) Where no is the total number of evaporated particles at the end of the laser pulse ( t = ). The pressure ( P ) at any point in the plasma is related to its d ensity by the ideal gas equ a tion ( P V = nkBTo) and can be expressed by [ 94]; 2 2 2 2 2 2) ( 2 ) ( 2 ) ( 2) ( ) ( ) ( ) , (t Zz t Y y t X x o oe t Z t Y t X T n t tz y x P ( 5.2) where T0 is the isothermal temperature of the plasma. The expression for the position and time dependent velocity ( v ) of the plasma is given by [ 94]; ^ ^ ^) ( ) ( ) ( ) ( ) ( ) ( ) , ( k dt t dZ t Z x j dt t dY t Y x i dt t dX t X x t z y x v ( 5.3) where dX/dt dY/dt and dZ/dt refer to the expansion velocities of the plasma edges X Y and Z respectively. The equation of gas dynamics governing the expansion of the plasma is the equation of continuity expressed as [ 94]; v S ot mn t A d N v dV t ( 5.4) where there is an injection of particles into the plasma, and V denotes the volume, S is the surface enclosing that volume, and N fluid and the mass m of th e atomic species. This equation simple states the conservation of mass for each atomic species. The last term shows the injection of atomic species into the plasma. The equation of the conservation of linear momentum or the equation of motion can be exp ressed as [ 94];

PAGE 117

97 0 dV P v v P v v v v t v t vV ( 5.5) If the relations for the velocity, density, and pressure are substituted into the equation of continuity and the equation of motion, the final solution is [ 94]; M T k dt t Z d dt t dZ t t Z dt t Y d dt t dY t t Y dt t X d dt t dX t t Xo B 2 2 2 2 2 2) ( ) ( 1 ) ( ) ( ) ( 1 ) ( ) ( ) ( 1 ) ( where t (5 .6) where To is the initial plasma temperature, M is the atomic mass of the plume species, and kB is the Boltsmann constant. This equation determines the initial expansion of the three ortho gonal plasma edges, see Figure 5 .2. The initial dimens ions of the plasma in the transverse direction are determined by the irradiated area of the laser spot size and are on the order of millimeters, whereas in the perpendicular direction the dimension is on the order of microns. Since the velocities are contr olled by t he pressure gradients, the Figure 5 .2. Schematic diagram of the (a) laser spot size at the target surface and (b) the plasma volume with the orthogonal edges (Xo,Yo, Zo) indicated. Z o y z x Yo X o 10 100 m (b ) Z o y z Yo 2.0 mm 3.0 mm (a)

PAGE 118

98 expansion is anisotropic in the direction perpendicular to the target. The second phase of the process begins after the termination of the laser pulse ( t > ), no particles are evaporated or injected into the plasma, also adiabatic expansion of the plasma occurs. The temperature of the plasma can be related to the dimensions of the plasma by the adiabatic thermodynamic equation given by [ 94]; const t Z t Y t X T 1) ( ) ( ) ( (5 .7) cities at constant pressure and volume where; V PC C (5 8) The thermal energy is rapidly converted into kinetic energy with the plasma attaining extremely high expansion velocities. Since there is no injection of particles into the plasma in this regime, the density and pressure gradients can be expressed in a form similar to equations (5.1) and (5.2) by neglecting the term ( ). The plasma density can be then expressed as [ 94]; 2 2 2 2 2 2) ( 2 ) ( 2 ) ( 2) ( ) ( ) ( ) , (t Z z t Y y t X x oe t Z t Y t X n t z y x n (5 .9) and the pressure as [ 94]; 2 2 2 2 2 2) ( 2 ) ( 2 ) ( 2) ( ) ( ) ( ) , (t Z z t Y y t X x o oe t Z t Y t X T n t z y x P (5 .10) for t > The velocity expressio n remains similar to equation (5 .3) as in the isothermal regime. The adiabatic equation of state is given as [ 94]; 0 1 n v t n n P v t P P (5 .11) And the equati on of temperature is given by [94];

PAGE 119

99 v T T v t T 1 (5 .12) The assumption is that there are no spatial variations in the plasma temperature, ( 0 T ). The equations for the velocity, density, and pressure are substituted into the diff er ential equations, Equations (5.4), (5.5), (5.11), and (5 .12). The solution which controls the expansion of the plasma in this regime is given by following set of similarity equations [94]; ) 1 ( 2 2 2 2 2 2) ( ) ( ) ( 6 ) ( ) ( ) ( t Z t Y t X Z Y X M T k dt Z d t Z dt Y d t Y dt X d t Xo o o o B (5 .13) These equations facilitate the threedimensional expansion profile of the laser ablated plume. Xo, Yo, and Zo are the three initial o rthogonal plasma edges (Figure 5.2) after the termination of the las er pulse ( t = ) ~ 30 ns FWHM at 1 J/ cm2 with the center of the laser irradiated spot at spatial coordinates (0, 0, 0) at time t The initial perpendicular dimensions ( x ) are less than 10 are determined by the di mensions of the plasma. The x coordinate is directed perpendicular to the target surface. X(t) Y(t) Z(t) refer to the spatial coordinates of the expanding plasma in the three orthogonal directions. To is the isothermal temperature of the plasma, kB is th e Boltsmann constant, M denotes the atomic mass of the plasma species, and is the ratio of the specific heat capacity at constant pressure and volume. This theoretical model for pulsed laser evaporation was applied to experimental conditions by simultaneously solving the similarity Equations ( 5.6) and (5.7) in a Maple 9.50 pr ogram to compute the plume density profile as a function of time which is TOF of the evaporated material. The program, Plasma Model .mw is given in Appendix H For single laser ablation, which this model was originally designed for

PAGE 120

100 there is an isothermal regime, which lasts for the duration of the laser pulse (20 60 ns), followed by the adiabatic regime. However, m odeling the dual laser ablated plume is complicated by the arrival of the CO2 laser pulse [ 68] There are two distinct stages to the initial i sothermal expansion in the dual laser case. The first stage, during the excimer laser pulse, is modeled similarly to the single laser expansion with a higher temperature corresponding to increased absorption of CO2 laser energy. After the conclusion of the excimer laser pulse, the continuing absorption of the CO2 laser by the plasma for approximately 264 ns due to the longer laser pulse width, without including additional incorporation of material from the t arget into the plume 5.2 Application of the T heoretical Model The previously described model was used to fit the theoretical time and position dependent density relation to the experimental temporal time of flight (TOF) profiles obtained during in situ optical diagnostics w hich were presented in Chap ter 4 The analysis was conducted on data for dual laser ablation with the combined laser fluences of 1 J/ cm2 UV and 2 J/ cm2 IR. The UV irradiation component of the dual laser ablation, low fluence excimer only laser fluence of 1 J/ cm2 was analyzed to det ermine the enhancement provided by the addition of the CO2 laser energy. The high fluence excimer only laser ablation (3 J/ cm2) was also calculated and compared due to its laser target interaction similarity to that of the dual laser effect. The input par ameters used in the simulation were the; laser spot size, at omic mass of the plasma species, delay time, time gate, distance from the target surface, plasma velocity components specific heat ratio (gamma), plasma thickness, and plasma

PAGE 121

101 temperature. The ar ea irradiated by the incident laser pulse into the target surface kept consistent and was measured to be 3 mm high 2 mm wide, and since the orthogonal egdes of the plasma volume are determined from the center of the laser spot then Y = 1.5 mm and Z = 1.0 mm. The atomic masses of the elements of the Ba8Ga16Ge30 compound are Ba8 = 137.33, Ga16 = 69.72, Ge30 = 72.59, with a total atomic mass equal to 279.64, so the average of the total atomic mass was used ( M = 93.21) for the plasma species. The laser pulse duration was used as the delay time. The timegate was the same as the experimental time gate condition utilized during in situ OES. The distance from the target surface was the same on axis position of the OES optical fiber used during data collection, 0 6 mm. The perpendicular plasma velocity ( vx) was determined for the TOF values; however the were determined by successive iterations of the model to fit the data. The remaining simulation parameters; gamma, plasma thickness, and plasma temperature, were also determined by iterations of the computational model. For those parameters which were not determined by direct experimental measurements, were varied based on practical values reported for PLD and trends similar to physical expectations. These values w ere in the range of 0.1 2.0 cm/s for transverse velocities vy and vz, 1.1 1.67 for gamma, plasma thickness of 10 100 m, and plasma temperatures of 2000 40000 K [84] The transverse velocities are on the low end of the range for singlelaser ablat ion due to the highly forward directed nature, and on the high end for dual laser ablation due to the much broader expansion profiles observed during in sit u optical diagnostics, Chapter 4 Gamma is high for low energy, low emission PLD and decreases for h igher energy conditions. Low plasma thickness and temperature occur with low energy PLE, and is counter for high energy PLE.

PAGE 122

102 The model calculated the number density of a species passing a certain point in space over a set time interval. This includes the e mitting and nonemitting atoms or ions of the species. Thus, the model predicts the total amount of material passing that point. This is ideal for fitting film thickness profiles as previously done by Singh et al. and Mukerjee et al. This makes the model a useful tool in predicting thickness profiles and the degree of expansion, however, the calculated number density does not directly correspond to the light emitted by the material at that point for the same time interval. The agreement between the model an d the data falls off as the angle from the optic axis and distance from the source increases. The plasma expansion is thought of as a high pressure gas expanding into the vacuum. The sustained emission from the plume is a result of collision between the co nstituents of the plume and the lower pressure gradient of the plasma at the outer edges of the plume will result in fewer collisions resulting in lower emission intensities. Due to this effect, the computational results and experimental data were only com pared on axis and close to the target surface (0 6 mm) where the emission is still intense and would have less inherent error as compared to longer distances and far off axis. The results were also computed for multiple points (2, 4, and 6 mm) for each e xperimental condition to further improve the output. The corresponding best theoretical fit curves are shown in Fig ure 5.3 as the solid black lines and overlaid with the experimental results for TOF from Chapter 4 The very high intensity emission of the h igh fluence single laser ablation aided in obtaining good agreement between the experimental and computational results (Figure 5 .3b). This helped with determining the affect of adjusting parameters in the model which could be applied when analyzing the res ults of lower emission data such as for the single laser 1 J/ cm2

PAGE 123

103 Figure 5 3 TOF profiles and theoretical model fit for single laser ablation (a) 1 J/ cm 2 (b) 3 J/ cm2, and (c) dual laser 1 J/ cm2 UV + 2 J/ cm2 IR. The black line is the mo del fit. 0.0 0.2 0.4 0.6 0.8 1.0 0 250 500 750 1000 1250 1500 Time (ns) Normalized Intensity (arb) 2mm 4mm 6mm (a) 0.0 0.2 0.4 0.6 0.8 1.0 0 250 500 750 1000 1250 1500 Time (ns) Normalized Intensity (arb) 2mm 4mm 6mm ( b ) 0.0 0.2 0.4 0.6 0.8 1.0 0 250 500 750 1000 1250 1500 Time (ns) Normalized Intensity (arb) 2mm 4mm 6mm (c )

PAGE 124

104 fluence. Adjustment of the parameters to have the peaks of the profiles occur at the proper times determined accurate simulation of the on axis velocity ( vx), and the broadening of the profile was found to be a factor of the initial transverse velocities ( vy and vz) The results for the velocity, plasma thickness, and plasma temperature, are given in Table 5.1. The single laser high fluence (3 J/ cm2) had the greatest plasma thickness which results in the highly forward directed plume, as discussed in the ICCD imaging section. The dual laser had a thinner plasma thickness, but a much higher plasma temperature, the combination of which resulted in the highly ion ized elemental species in the plume and the broader plume expansion. The ionization of the laser generated plasma is strongly dependent on the surface and plasma temperature. 5.3. Summary Application of the hydrodynamic modeling of the plasma expansion has successful matched the experimentally obtained TOF profiles. This has allowed determination of the laser generated plasma thickness and temperature. The high initial plasma temperature of the dual laser technique was obtained while using modest laser e nergy densities This comparatively high plasma temperature is the origin of the higher Sample T (K) S10.63 1.53 18 7600 S30.90 1.31 49 14700 DL1.56 1.17 39 28400 Table 5.1 Velocity, coefficient gamma, plasma thickness, and plasma temperature. for 1 J/ cm2 and (b) 3 J/ cm2 single laser and (c) dual laser.

PAGE 125

105 ionization which was observed during optical diagnostics, and the higher velocity which leads to higher kinetic energy. These effects are conducive for deposition of hi gh quality thin films.

PAGE 126

106 CHAPTER 6: CHARACTERIZATION OF THE Ba8Ga16Ge30 THIN FILMS This is a properties characterization report of Ba8Ga16Ge30 type I clathrate thin films deposited by pulsed laser ablation techniques Electronic devices, coatings, displays, sensors, optical equipment, and numerous other technologies all depend on the deposition high quality thin films The t hin film growth was evaluated by deposition rate and distribution, film structure which includ ed crystallinity surface morphology, and composition, and t he electric transport properties of the films The electric transport properties included temperature dependent resistivity, and room temperature Seebeck coefficient, carrier concentration, and mobility. The concept parameters and procedure (Appendix I) of the pulsed dual laser deposition system ( s ) utilized in the experiments was described in Chapter 1.2 and Chapter 2 with a system diagram shown in Figure 2.1. 6. 1. Film Deposition Profile The dependence of the on axis film deposition rate on the laser fluence was investigated. The deposited film thickness was obtained using a surface texture profilometer (Dektak 3030 ST). The step height produced by the removable stainless steel mask is measured by a stylus in contact with, and gently dragging along, th e surface of the substrate. For each trial, the target substrate distance was maintained at 4 cm, and the substrate heater block was unheated, referred to as room temperature. The on axis

PAGE 127

107 deposition rate for the laser ablation fluence was then calculated by dividing the on axis film thickness by the total number of laser pulses during the deposition. The single laser deposition rate increases linearly for fluences from 0.6 to 3 J/ cm2 at a rate of 0.55 /pulse per J/ cm2 (Figure 6.1). The ablation threshold occurred at fluences lower than 0.6 J/ cm2 where no detectable deposition occurred. The ablation threshold occurs at the value where the energy density is absorbed into the material but is not sufficient to raise the temperature of the material enough to cause evaporation. Whereas on the other end of the linear increase of the deposition rate ( 3 J/ cm2), the laser ablation causes the plasma density at the surface of the target to reach a point where higher laser fluences can not penetrate the plasma, thereb y shielding the target from further evaporation. At this point the deposition rate saturates. At this high energy region, changes in the plasma may have occurred such as plasma losses and reflectivity of the laser pulse. The on axis deposition Figure 6 1 The dependence of the on axis film deposition rate per laser pulse on the excimer laser fluence. The substrate temperature was room temperature. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0 1 2 3 4 5 6 Fluence (J/cm2) Rate (/Pulse)

PAGE 128

108 rate of 0.14 0.03 /pulse for preliminary dual laser ablation conditions (1 J/ cm2 UV and 2 J/ cm2 IR) was found to be significantly lower than the excimer only depositions. The relatively low deposition rate is a result of the combination of low laser fluences, and other factors unique to dual laser ablation such as the broader expansion profile which was presented in Chapter 4 Large area films depositions were performed for dual laser (1 J/ cm2 UV + 2 J/ cm2 IR) (DL) and the excimer only conditions of 1 J/ cm2 (S1) an d 3 J/ cm2 (S3) The xyz axis orientation of the irradiated laser spot on the target surface relative to the resulting film deposition on the substrate is illustrated in the schematic diagram shown in Figure 6.2. The x axis refers to the central axis of the laser spot, the laser generated plume of material, and the peak of the film distribution profile. This x axis is perpendicular to the surface of the target and the substrate (Figure 6.2). The film thickness profiles of these films (Figure 6. 3) along the yaxis (vertical) and z axis (horizontal) of the various laser fluences illustrate the broader deposition profile achieved by the dual laser configuration. The solid line is the cosn( corresponds to the angle between Figure 6.2 A schematic diagram illustrating the xyz axis orientation of the irradiated laser spot on the target surface relative to the resulting film deposition on the substrate. z y y z x 4 cm Rotating T arget 32 mm Laser Spot in. Substrate

PAGE 129

109 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 0 5 10 15 20 25 30 35 40 Angle (deg) Normalized Thickness S1 S3 DL (a) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 0 5 10 15 20 25 30 35 40 Angle (deg) Normalized Thickness S1 S3 DL (a) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 0 5 10 15 20 25 30 35 40 Angle (deg.) Normalized Thickness S1 S3 DL (b) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 0 5 10 15 20 25 30 35 40 Angle (deg.) Normalized Thickness S1 S3 DL (b) Figure 6.3 Profil e film thickness along the (a) z axis and (b) y axis for Ba 8 Ga 16 Ge 30 deposited on q uartz substrates. The target to substrate distance was 4 cm. The solid lines indicate the theoretical fit cosn

PAGE 130

110 the radial vector and the target normal. Depending on the pulse energy density and other parameters, the thickness variation across the films typically show a relationship where n = 8 12 [ 94]. The cosn( 6. 1. The y / z ratio for the cosn( laser indicating a broader and more uniform distribution profile. This ratio p rogresses to a value of 2 as the single laser fluen ce increases which is indicative a narrower, more forward directed deposition The lower on axis deposition rate discussed earlier for the dual laser deposition is a result of its desirable broader deposition as compared to the singlelaser conditions. 6.2 Thin Film Surface Morphology The surface topography of the excimer laser deposited films was examined by SEM. An example of the distribution of undesired particulates found on the films is shown in the 300 0X magnified SEM image ( Figure 6 .4 ) The films were deposited at excimer only laser fluences of 1 and 3 J/ cm2 and dual laser fluences of 1 J/ cm2 UV + 2J/ cm2 IR. The films deposited by excimer only single laser ablation have an undesirable range of particulates ranging from sub micron droplets to macro molten droplets, with Table 6. 1. The n factor of the cos n fit for the y and z axis thickness profiles for single a nd dual laser depositions at a target substrate distance of 4 cm Sample z y y/z Single 1 J/cm29.0 14.0 1.56 Single 3 J/cm28.5 17.0 2.00 Dual-Laser 7.0 8.5 1.21 n factor for cosn

PAGE 131

111 sizes greater than 5 m, and irregular s hards, Figure 6.5a and b, as seen previously [75, 99, 100] and with other TE materials [45]. The macro droplets originate from splashing of the irradiated target molten layer which results in particles which were molten in transit but too massive to vaporize in the laser generated plasma, these features are typical of high fluence ablation of materials composed of metallic elements. A common and simple approach to reduce the quantity of particles is to decrease the laser fluence, so films deposited with a fluence of 1 J/cm2 (Figure 6.5a) do not contain macro droplets but the micron to submicron sized particles are still present. However, the film deposited by dual laser ablation resulted in a sign ificant reduction in particulates (Figure 6.5c) Figure 6.4 A SEM at 3000 times magnification which shows examples of the various types of particulates found on the thin films.

PAGE 132

112 (a) (b) (c) Figure 6.5 SEM images of thin films deposited by single laser PLD with fluences of (a) 1 J/cm2 and (b) 3 J/cm2, and dual laser fluences of (c) 1 J/cm2 UV and 2 J/cm2 IR.

PAGE 133

113 Thin films were deposited at three distinct pp delays of 25 ns, 100 ns, and 175 ns, respectively, to investigate the effect on the morphology of the films. At 25 ns pp delay, a significant portion of the CO2 laser pulse arrives at the target surface before the arrival of t he excimer laser pulse (Figure 6 6) resulting in a condition referred to as premelt of the laser target interaction site. Based on the results presented in Chapter 4, 100 ns pp de lay was a delay at which expansion of the plasma and plume are increased. When the pp delay is 175 ns, plasma and/or plume pumping occurs. Upon inspection of the deposited films by SEM (Figure 6.7a and 6 7c) the films produced under the conditions with t he 25 ns and 175 ns pp delays, respectively, have a large amount of molten droplets, an undesirable quality for the deposition of thin films. However, the 100 ns pp delay film ( Figure 6. 7b) is smooth with few submicron sized particl es. The particulates in Figure 6. 7a have been ejected from the target surface and only partially melted while 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 -200 -100 0 100 200 300 400 500 600 Time (ns) Normalized Voltage UV 0ns IR -25ns IR 100ns IR 175ns 0 -25 100 175 Figure 6.6 Overlapped oscilloscope traces for the excimer laser pulse (UV) and CO 2 laser pulses (IR) for peak to peak delays of 25 ns, +100 ns, and +175 ns, relative to the UV pulse at 0 ns.

PAGE 134

114 passing through the laser generated plasma. The particulates in Figure 6.7 c are characteristic of ejected material that has not reached a temperature to vaporize the material but enough to liquefy the material resulting in molten droplets that splash when deposited on the substrate. Implementing the dual laser technique has greatly reduced the quantity of particulates on the film, as can be seen in Figure 6 .7 b. Dual l aser ablation has eliminated the large macro droplets through laser target interaction site pre melt from the long wavelength IR CO2 laser followed by UV laser ablation from pre smoothed target zone, then submicron sized particles are evaporated in the pla sma which has been increased in temperature due to absorption of the remainder of the long CO2 laser pulse.

PAGE 135

115 (c) (b) (a) Figure 6.7. Scanning electron micrographs of films deposited at various dual laser peak to peak delays with 1 J/cm2 UV and 2 J/cm2 IR laser fluences. The CO2 laser pulse peak is delayed b y (a) 25 ns, (b) 100 ns, and (c) 175 ns, relative to the excimer laser pulse peak. The film deposited at 100 ns delay resulted in reduced particulate generation.

PAGE 136

116 6.3 Thin Film Crystallinity The crystallinity of the deposited thinfilms was examined by x ray diffraction (XRD). The XRD peak intensity dependence on the substrate temperature was investigated over a range from room temperatur e to 500 C as shown in Figure 6.8 The excimer laser fluence was maintained at 1 J/ cm2. The germanium impurity peak at 27.27 only appears at 500 C when the film quality appears to be affected by the deposition temperature, and the deposition rate is reduced as the material is reevaporated from the heated substrate. The XRD intensity of the prominent peak at 31.2 (321) as a function of temperatu re is shown in Figure 6. 9 and a maximum occurs at a substrate temperature of 400 C, and may be near an optimal deposition temperature for this laser fluence. 0 100 200 300 400 500 600 20 25 30 35 40 45 50 55 60 Intensity (arb.) Target RT 100C 200C 300C 400C 500C Figure 6.8 The XRD dependence on the substrate temperature. The excimer laser fluence was 1 J/ cm2.

PAGE 137

117 The XRD peak intensity dependence on the excimer laser fluence was investigated over a range of 1 to 6 J/ cm2, as shown in Figure 6 .10. The substrate temperature was maintained at 400 C. The germanium impurity peak at 27.27 does not appear in any of the films. The XRD pattern was indexed to a cubic unit cell structure with space group Pm 3n. The di ffraction pattern displays peaks which are characteristic of type I clathrates and are in good agreement those reported for Ba8Ga16Ge30 [ 76 79] and for the target as reported in Chapter 3 For each of the fluences, the peak that occurs at 29.85 (320) is proportionately lower than that of the target. The XRD intensity of the prominent peak at 31.2 (321) as a function of excimer la ser fluence is shown in Figure 6.1 1. The peak intensity increases linearly from 1 to 3 J/ cm2 and a maximum occurs at a fluence of 4 J/ cm2. 40 50 60 70 80 90 100 110 120 130 0 100 200 300 400 500 600 Temperature (C) Figure 6.9 XRD intensity of the peak 31.2 (321) as a function of temperature.

PAGE 138

118 0 1000 2000 3000 4000 5000 6000 25 30 35 40 45 50 55 Scatering Angle 2 (degree) Intensity (arb.) Target 6 J/cm2 5 J/cm2 4 J/cm2 3 J/cm2 2 J/cm2 1 J/cm2 400 C Figure 6.10 The XRD patterns dependence on the excimer laser fluence. The substrate temperature was 400 C. 0 100 200 300 400 500 600 700 800 900 0 1 2 3 4 5 6 7 Fluence (J/cm2) Figure 6.11 XRD intensity of the peak 31.2 (321) as a function of the excimer laser fluence.

PAGE 139

119 A set of films were deposited by dual laser ablation with matching thickness of 264 nm for single laser films, Figure 6 .1 2. It is evident that the low fluence deposited film has a weak XRD peak intensity as compared to the other configurati ons (29% max.). The XRD peak patterns of the films deposited by the high fluence singlelaser and the dual laser configurations agree with that produced by the Ba8Ga16Ge30 target [ 79 ]. These two films have similar XRD peak heights even though the films gro wn by the dual laser deposition technique only required a substrate temperature of 300 C as compared to 400 C for the single laser deposited films. This is achieved by the increased ionization and kinetic energy of the plume material provided by the dual laser ablation leading to higher mobility of the material arriving at the heated substrate. 25 30 35 40 45 50 55 Intensity (arb.) 87% 29% 100% DL1+2 SL3 SL1 Figure 6.12 X ray diffraction patterns for single laser 1 J/ cm 2 UV (S L1 ) and 3 J/cm 2 UV (SL3), and dual laser (DL1+2) 1J/cm2 UV + 2 J/cm2 IR. The film thickness was 267 30 nm. The type I clathrate XRD peaks are identified to a cubic unit cell structure with space group Pm 3n

PAGE 140

120 6.4 Thin Film Composition The composition of the deposited thin films was analyzed by energy dispersive X ray spectroscopy ( EDS ), a standard procedure for ident ifying and quantifying the elemental composition of sample volumes as small as a few cubic micrometers (Appendix J) The dependence of the film stoichiometry on the excimer laser fluence was examined. The quartz substrate temperature was maintained at 400 C. The laser fluence was varied from 1 to 6 J/ cm2. The average film thickness was 5780 780 The atomic percentage ratios normalized to Ba8 ar e shown in Figure 6 .1 3. The deposition over the fluence range of 1 to 6 J/ cm2 produced the desired stoichiomet ry of Ba8Ga16Ge30. 5 10 15 20 25 30 35 0 1 2 3 4 5 6 7 Fluence (J/cm2) Atomic % Ba Ga Ge Ba8 Ga16 Ge30 Figure 6.13 EDS quantification of t he films deposited as a function of the excimer laser fluence.

PAGE 141

121 Films were deposited by the dual laser ablation technique along with comparative singlelaser deposited films. The films presented had an average thickness of 267 30 nm. The film deposited by the single laser 1 J/cm2 fluence has slight Ba enrichment as compared to the Ga and Ge content. This is a result of the preferential evaporation of the more volatile Ba from the target, T able 6 2, leading to a greater percentage of Ba arriving at the substrate. It is noted that th e deviation in the low fluence deposited film was much greater than the other two configurations. The on axis stoichiometry of films deposited by the single laser 3 J/cm2 fluence and dual laser are with the proper range as the outcome of phase segregation during ablation from the target. The Ga/Ge ratio of these two depositions resulted in an ntype composition ratios (Ga/Ge < 0.533) with the high fluence singlelaser closer to intrinsic [6]. The composition of the duallaser deposited films with various pp delays was investigated and the elemental composition summary of these deposited films (Table 6. 3) reveals that the 25 ns and 175 ns films are far off the Ba:Ga:Ge 8:16:30 intrinsic ratio. These films are rich in barium and gallium, the elements with t he lowest boiling points of the three. The film thicknesses are also increased as a result of increased material being Table 6.2 Stoichiometry of thin films deposited by single laser 1 J/ cm 2 (S 1 ) and 3 J/ cm 2 (S3), and dual laser (D ual ) ablation. Ba Ga Ge Balanced 8.0 16.0 30.0 0.533 Dual 8.10.4 15.30.2 30.00.2 0.510.00 S38.10.4 15.70.6 30.00.5 0.530.03 S18.50.6 16.31.2 30.00.8 0.540.04 Atomic % Sample Ga/Ge

PAGE 142

122 ejected from the target surface, in vapor and liquefied form. The slightly more forward directed plume propagation, as indicated by the n arrower and more intense plasma expansion will also contribute to higher onaxis material deposition. Although this experiment exaggerated non optimum laser pulse synchronization by 75 ns and grater, it demonstrates the need proper synchronization the dua l laser pulses for optimum energy coupling of the CO2 laser pulse into the excimer laser generated plume. It also leads to possibility of tailoring the elemental ratios through proper management laser energies and the pp delay. This could be a future inve stigation of this complex material for control of the Ga:Ge ratio for ptype and ntype film growths. 6. 5. Electric Transport Properties The thin films were further characterized by their electric transport properties. Samples deposited by single and dua l laser to various thicknesses of 200 to 650 nm were evaluated. The electric transport properties of the films included temperature dependent resistivity ( ), carrier concentration ( n), thermopower or Seebeck coefficient ( S), and thermoelectric power factor ( S2/ ). Thickness Ba Ga Ge (nm) 175 ns 17.51.4 22.24.7 30.08.3 932 100 ns 8.70.1 16.00.1 30.00.6 816 -25 ns 47.54.5 26.716.8 30.08.3 1505 Atomic % Sample Table 6 3. Barium, gallium, and germanium composition of thin films deposited a various dual laser peak to peak delays of 175 ns, 100 ns and 25 ns. The substrate tem perature during deposition was room temperature. The thickness of the resultant films is given.

PAGE 143

123 6. 5.1. Temperature Dependant Resistivity The temperature dependent resistivity of the films was measured using standard a four point probe tech nique. The film was mounted in a closed cycle refrigeration system that controlled the temperature in the range starting at 320 K and was decreased to 20 K. The resistance was typically measured across a 0.20 cm 0.31 cm bridge that was scribed on the fil m. Voltage measurements were taken for injected currents of 5 to 1000 determined upon the response of the particular thin film sample. A sample temperature dependent resistivity result for a random Ba8Ga16Ge30 thin film is shown in Figure 6.14. The resistivity increases exponentially with decreasing temperature to demon strat ing t he semiconducting characteristic of the deposited material. The resistivity of this particular sample diverges as the temperature is decreased due to ohmic heating as 0 20000 40000 60000 80000 100000 120000 0 50 100 150 200 250 300 350 Temperature (K) Resistivity ( m cm ) 10nA 505nA 1000nA Figure 6.14. The temperature dependent resisti vity of a Ba 8 Ga 16 Ge 30 thin film sample at various applied injection currents, demonstrating the semiconducting behavior.

PAGE 144

124 a result of the high resistivity and applied current. This effect is avoided du ring final sample measurements. During this investigation it was found that f ilms deposited by the excimer only single laser ablation fluence of 1 J/cm2 had excessively high resistivity beyond the limits of the available instrumentation, so this configura tion was eliminated from further electric transport characterization. Thin films deposited by high fluence single laser ablation (3 J/ cm2) were also found to have high but measureable resistivity, an d were included in the following investigations as a comparison to the results of the higher quality dual laser deposited thin films Although the high fluence single laser deposited films had high particulate concentrations as compared to the dual laser deposited films, the crystallinity results were comparable and of interest, as discussed in Sections 6.3 and 6.4 of this chapter. The temperature dependent resistivity was investigated for films of various thicknesses to gain insight into the quality of the films. Films deposited to a thickness range 200 to 650 nm are presented and the results are shown in Figure 6.15. As the thickness of the film increased the resistivity decreased for the dual laser deposited samples. However the high fluence single laser film at 300 nm thick show the highest resistivity and thi cker films resulted in increased resistivity counter intuitive to typical findings These thicker single laser deposited films were imaged by SEM and were found to c ontain micro cracks (Figure 6.16) causing discontinuit y and increased resistance. These mic ro cracks are the result of stress in the film; however the film did not peel from the substrate indicating adequate adhesion. These micro cracks did not occur with the dual laser deposited thin films, again demonstrating the benefit of the technique by re ducing the stress in the film through the highly energetic and ionized plume of material

PAGE 145

125 created during ablation, as discussed in Chapters 4 and 5. In general, films are in a state of stress, compressive or tensile stress. Compressively stressed films wi ll expand parallel Figure 6.16 A scanning electron micrograph image of a film deposited by single laser PLD with a fluence of 3 J/cm2. The average film thickness is 623 nm. The micro cracks are visible on the film. 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07 0 50 100 150 200 250 300 350 Temperature (K) S 300nm D 202nm D 269nm D 435nm D 632nm Figure 6.1 5 Temperature dependent resistivit y of the Ba 8 Ga 16 Ge 30 thin films deposited by single and dual laser ablation for various thicknesses.

PAGE 146

126 to the substrate surface leading to peeling. Tensile stressed films want to contract and will crack if the elastic limit of the material is surpassed High stress in films is undesirable for device applications because they (a) tend to h ave poor adhesion, (b) are susceptible to corrosion, (c) tend to crack, and (d) have resistivity higher than relaxed annealed films [ 134]. During the SEM investigation of these films it was found that the highly resistive films grown by single laser de position display the occurrence of crystallites within the films, as seen in Figure 6.17a (d). These crystallites are 50 to 150 nm in diameter and appear to not be in contact with one another. The films deposited by dual laser ablation do not show this effect, as see in the SEM image of Figure 6.17b (d). Thin films formed by physical vapor deposition methods such as PLD grow by condensation of atoms or molecules from the vapor phase. This condensation begins with the formation of small clusters or nucle i through random agglomeration, as depicted in Figure 6.17a (a) and Figure 6.17b (a). Atoms may continue to move along the surface into other positions as a result of the kinetic energy associated with the materials arrival velocity or by thermal activati on from the heated substrate. The dual laser deposited material has a higher velocity as discussed in Chapter 4, 0.90 and 1.56 cm/s for high fluence single laser and dual laser ablation, respectively. Th e dual laser ablation results in higher surface mobi lity possibly allowing increased migration, leading to smaller and more numerous initial nucleation sites, resulting in a more uniform film growth as opposed to that of the single laser ablation results. As depicted in the progression of nucleation and gro wth, as shown in Figures 6.17a (a c) and 6.17b (a c).

PAGE 147

127 Figure 6.17 a. Schematic diagram of progress ive film nucleation for single laser deposition, (a) sparse nucleation, (b) increased coalescence growth, (c) final, discontinuous crystal growth within an amorphous fill and or channels and holes, and (d) an SEM image of a Ba8Ga16Ge30 film grown with 3 J/ cm2 single laser ablation. Note the 50150 nm crystallites encased within the film growth. (a) ( b ) ( c ) 500 nm (d) Figure 6.17b Schematic diagram of progressive film nucleation for dual laser deposition, (a) high energy, ion bombardment promoting increased nucleation, (b) increased, dense crystal growth, (c) final, uniform contin uous growth, and (d) an SEM image of a Ba8Ga16Ge30 film grown from 1 J/ cm2 UV and 2 J/ cm2 IR dual laser ablation. Note the smooth film growth as compared to the image in Figure 6.20a. (a) (b ) ( c ) 500 nm (d)

PAGE 148

128 The electrical conductivity of a material is due to the motion of charge carriers through the lattice [135]. These carriers are affected by non periodicity in the lattice, resu lting lattice vibrations, lattice defects, impurities, vacancies, interstitials, dislocations, grain boundaries, and alloying. The resistivity of thin films is generally higher than that of the same bulk material, resulting from the fact that films general ly contain more defects, dislocations, and grain boundaries [134]. The addition of even small amounts of an impurity, such as oxygen, can cause very large changes in resistivity, on the order of several orders of magnitude Vacancies, interstitials, and di slocations are considered to have a smaller effect on the resistivity of a material [136]. These issues may be factors in the resistivity values obtained for high fluence single laser deposited films as compared to the d ual laser results. They may also bec ome an increasing contribution to the dual laser deposited films as the films were made thinner. The room temperature results obtained by the temperature dependent resistivity investigation are summarized in Table 6. 4 Fluence d (J/cm2) (nm) 1 Single 3 UV 300 6.54104 2 Dual 1 UV + 2 IR 202 3.48103 3 Dual 1 UV + 2 IR 269 1.61103 4 Dual 1 UV + 2 IR 435 4.55102 5 Dual 1 UV + 2 IR 632 2.30101 PLD Sample Table 6. 4. Temperature dependent resistivity results of the Ba 8 Ga 16 Ge 30 thin films deposited by single and dual laser ablation for various thicknesses.

PAGE 149

129 6.5.2. Van der Pauw Hall Measuremen t Electrical properties of the thin films were measured using the Van der Pauw technique. Since the thickness of the films was much smaller than the length and width of the sample and grown on symmetrical ( inch) quartz substrates, the Van der Pauw method could be successfully applied. The electrical contacts were made using silver epoxy and thin, epoxy coated, copper wire. They were placed on the corners of the film as shown in the schematic diagram (Figure 6.18 ). During the Hall measurements, the m agnetic field ( B ) was applied perpendicular to the film plane as shown in the schematic diagram (Figure 6.18). The electromagnet utilized for the measurements provided a maximum field of 0.5 Telsa. The results were tabulated such that the sheet resistance ( RS) was determined by: 1 S B S AR R R Re e [6.1] The r esistivity ( ) could then be calculated using the film thickness ( d): d RS [6.2] The carrier concentration ( n) could then be determined using: B 1 4 3 2 Figure 6.18 The schematic diagram of the Van der Pauw technique for determination of resistivity and Hall values for thin film samples.

PAGE 150

130 i SV q IB n810 8 [6.3] d n nS [6.4] The m obility ( H) was then calculated using the following: S S HR qn 1 [6.5] R esults of investigating films of various thickness using the Van der Pauw method for room temperature resi stivity ( ), carrier concentration ( n), and mobility ( H) are summarized in Table 6.5. Values of 1 to 52 m for bulk polycrystalline Ba8Ga16Ge30 at room temperature have been reported [101, 33] The carrier concentration s ( n) are lower than that reported for the bulk Ba8Ga16Ge30 materi al typical for thin films. However, t he thinner films resulted in excessively low calculation of carrier concentration due to u naccounted thickness variations. A verage film thickness does not accurately correspond to calcu lations and f urther investigation is required. Fluence d n H (J/cm2) (nm) ( m cm ) (cm-3) (cm2/ Vs ) 1 Single 3 UV 300 6 7 1043 4 101610.2 2 Dual 1 UV + 2 IR 202 3 58 1039 9 101617.7 3 Dual 1 UV + 2 IR 269 1 60 1035 6 101617.1 4 Dual 1 UV + 2 IR 435 4 31 1021 4 10185.5 5 Dual 1 UV + 2 IR 632 2 10 1011 1 10197.7 Sample PLD Table 6.5 Results of the Van der Pauw measurements for room temperature resistivity ( ), carrier concentration ( n), and mobility ( H)

PAGE 151

131 6. 5. 3 Seebeck Coefficient Thermoelectric energy conversion can be discussed with reference to the schematic of a thermocouple configuration shown in Figure 6.19. It can be considered as a circuit formed from two dissimilar conductors a and b which are connected electrically in series but thermally in parallel. If junctions at 1 and 2 are maintained at different temperatures TH and TL where TH > TL, an open circuit electromotive force (emf), V is developed betw een 3 and 4 and given by V = ( TH TL) or = V / which defines the differential Seebeck coefficient ab between elements a and b. For small temperature differences the relationship is linear. The sign of is positive if the emf causes a current to flow in a clockwise direction ar ound the circuit and is measured in units of V/K, or more often in V/K [110]. The Seebeck coefficient or thermopower, is an intrinsic and bulk property of a material related to the materials electronic structure and is not geometry specific [106, 107] The thermopower yields information about the sign (+/ ) of the charge carrier and is essentially the entropy per carrier divided by the charge of the carrier [106]. When a thermoelectric material is subjected to a small and constant temperature difference, a thermoelectric potential difference is produced. The thermopower of the material is defined as = / as measurement of the ratio of the sample voltage (electric field) to the temperature difference (temperature gradient) along the sample;

PAGE 152

132 L H L H abT T V V T V / / [6.1] where ab = b a is the measured value of the therm opower which includes both the sample contribution, a, as well as the measurement lead contribution, b. The lead contribution, typically Au, Cu, or Chromel, must be known and subtracted from each data point at each temperature [106] such that the thermop ower of the sample a = b ab and equation 6.1 may be rewritten as; L H L H b aT T V V [6.2] An essential requirement for accurate thermopower measurement but also a source of error is the determination of the temperatures ( TH and TL) and voltages ( VH and VL) at the same positions, respectively. This requires the contact junction of the temperature probe to be located as close as possible to the voltage probe junction. A way to minimize this error is to use one leg of each thermocouple as the sampl e voltage lead itself, thus the thermocouples act both as voltage potential leads and for measuring the thermo voltage [106]. Another requirement is zero heat flow along the measuring leads where the leads act as heat sinks [108]. This can be minimized by utilizing leads with as small a diameter as practical. Figure 6 19 A s chematic diagram of a basic thermocouple arrangement ( a b ) between hot (T H ) and cold (T L ) sources a b b V 3 4 + p type n type I Cold (T L ) 2 Hot (T H ) 1

PAGE 153

133 The simplest technique of measuring the thermopower is the small method [106]. Typically, the temperature difference should be as small as possible by definition ( = / as a compromise has to be met in achieving sufficient measurement accuracy, thus is usually of the order of 1 to 5 K [108]. This method involves sweeping the around a set temperature and measuring the resulting voltage of the sample. The slope of as a function of is the measured thermopower ab. It is required that is linear in and goes through = 0 at = 0 [106]. For accurate and reproducible determination of the thermopower, the temperature difference should be relatively small, the sp ecimen homogeneous, and spurious voltages of thermal origin should be eliminated. When measuring at temperatures far from ambient, spurious thermally induced voltages will become more significant in the measuring circuit. It maybe also necessary to take in to account the absolute thermopower of the contact material used to thermally anchor the leads to the sample, especially if this is comparable to that of the material being measured. To minimize such effects it is recommended to measure the thermoelectric potential difference as a function of thermopower [108]. Measurement of the thermopower of thin film samples on substrates requires special consideration. The thermal conducti vity of the substrate is typically quite high and can be a source of unwanted temperature gradients. Thermally anchoring the thermocouple bead to the film surface presents sources of error since only a small fraction of the bead surface can actually be pla ced in contact with the film surface. In order to assure that the remainder of the bead is at the same temperature, solder or a

PAGE 154

134 thermally conducting epoxy can be used to thermally anchor the entire bead to the film [107]. It has been shown that there are a number of sources for error. It is quite easy to get an erroneous sign (+/ ) of the thermopower, it is essential to use consistency in defining the direction of the temperature gradient with respect to the voltage measurement, the sample will typically be opposite in sign from the measured thermopower. This is due to the fact that one has to subtract out the lead contributions ( a = b ab) which is usually small compared to the sample thermopower [107]. The voltage leads must be in excellent thermal contact to the point at which the temperature is measured [107]. High resistance contacts can exhibit ac pickup resulting in a re ctified dc offset voltage which can also lead to errors [106]. The temperature dependence of the thermopower can be quite difficult to interpret, many contributions, such as phonon drag can be involved in combination with simple diffusion thermopower (line ar in T ) which is typical of metals [106]. To check the accuracy of a Seebeck measurement both a constant offset and errors proportional to the Seebeck coefficient of the sample have to be excluded. Thus, Seebeck calibration with reference materials shoul d include a pure metal with low Seebeck coefficient (few V/K like Pt, Pd, Cu or others) as well as a reference material with higher Seebeck coefficient (preferentially several 10 to 100 V/K) this could be CuNi (constantan) or NiCr (chromel) based on stan dardized thermocouples calibration values [108]. A popular method to measure thermopower is the differential method, shown schematically in Figure 6.16. In this method, the temperature difference between two

PAGE 155

135 points of the sample and the potential difference between the same two points is measured when the net current in the sample is zero, I = 0. Therefore, the electrical field in the sample is due to the thermopower only, E = [109]. Fi gure 6.20 shows two thermocouples in contact with the surface of the solid whose Seebeck coefficient is to be measured. The temperature of junction X is raised above ambient temperature and is assumed that the sample immediately below the junction is at the same temperature, namely TH. Likewise it is assumed that TL is the temperature both of the junction Y and the solid immediately beneath it. The thermoelectric potential measured between wires 1 and 3 or 2 and 4 will depend only upon the nature of the wir es (C or A), the nature of the sample, and the temperature difference TH TL: the potential is independent of the shape of the isotherms between X and Y. If the potential difference is measured using the wire C (1 and 3), then the mean Seebeck coefficient SSC of the sample over the small temperature interval TH to TL, is given by where SC is the absolute mean Seebeck coefficient of material C between TH and TL. The sign of V is given by the sign of the potential of the cold electrode, and the V V T H TL a a b b 1 3 2 4 X Y Figure 6 20 Schematic arrangement of thermocouples on the sample surface for determination of the thermopower by the differential method.

PAGE 156

136 appropriate s ign of the material C must be used for SC. For p type material the Seebeck coefficient is positive (+), and so V will be positive, even for very low values of SSC, provided SC is chosen as positive [111]. In many applications an automatic ice bath with bui lt in reference thermocouples (T/Cs) may not be practical. Automatic ice baths are expensive and do not operate reliably in ambient temperatures below 0.0 C or above 40.0 C. Instruments available today for T/C measurement provide and electronic circuit for determining the temperature of the terminals to which the T/Cs are attached. An appropriate reference voltage is added by the system to that produced by the external T/C. Such circuits are referred to as compensators, because they compensated for the fact that the terminals to which the T/Cs were connected were not at the icepoint temperature. The compensator produces a voltage that is a function of the terminal temperature [112]. One of the main challenges is the measurement of true temperature grad ient across the sample. A true temperature gradient is measured only when good thermal contact is made between the sample and the sensors. In case of poor thermal contact, the temperature drop at the interface will underestimate due to overestimation of The only way to reduce this error is to reduce the heat flow that is necessary to develop the required temperature gradient In other words, one needs to use a long thin sample. Alternatively, one can attach thin thermocouples directly on the sampl e. In this case, finite area contact introduces error to that extent. In addition, samples that exhibit low thermal conductivity tend to give more accurate values. However, it is not always possible to obtain a sample of desired dimensions or thermal condu ctivity [113]. It has

PAGE 157

137 been reported, that it is necessary to select the metal electrode fitted to the thermoelectric material to make the thermoelectric conversion efficiency as high as possible [114]. It has been reported of the existence of a universal optimal reduced Fermi potential ( opt) and the corresponding optimal Seebeck coefficient, in the range of 103 187 V/K, where the power factor is maximized at any given temperature. These findings help to dispel the notion that good thermoelectrics should have very large | S|. Instead, it would help to refocus efforts on increasing the power factor ( S2 ) through an increase in the conductivity while maintaining |S|~ Sopt. Additionally, the identification of opt will enable easy determination of the optimal carrier concentration [115]. Simple, rapid, versatile, accurate techniques for determining the Seebeck coefficient of thermoelectric materials along with experimental difficulties have been described in publications [111, 116118]. The Seebeck coefficient data for materials, especially semiconductors, are prone to variation because of their sensitivity to the exact material composition. Also, the temperature dependency can be significant, for example a 510% increase for metals over a temperature rise of 30 C [118]. A main problem in surface temperature measurement is getting the thermocouple in good thermal contact at the chosen location. Always use the smallest diameter thermocouple wire that can be handled without too much breakage [120]. 6. 5. 3 .1 Seebeck Apparatus A differential method was chosen for the determination of the Seebeck coefficient. This method depends on the temperature difference T and potential V measured between the same two points of the sample when the net

PAGE 158

138 current in the sample is zero, I = 0. When the net current is zero, the electric field in the sample is only due to the thermopower, E T [122]. In the a pparatus shown in Figure 6.21, the heater blocks are 3.82 cm long 8.82 cm wide 1.33 cm high stainless steel with a 6.14 mm diameter 5.12 cm long, 250 watt stainless steel sheath cartridge heater inserted into the block. The heating elements are secur ed into the blocks with silver ink to improve thermal contact. The power to each heater block was adjusted by 0 120 V AC Variac power supplies and the temperature is measured and monitored with thermocouples. The sample is clamped between two stainless s teel blocks with the block sample contact areas coated with silver ink to improve thermal contact. The temperature probes are bare T type, copper constantan, 0.005 inch diameter, 12 inch long thermocouples. The junctions of the thermocouples are adjusted as to maintain their position under slight compression and then a small amount of silver ink, less than 1 mm in diameter, is applied to ensure thermal contact while securing the position. A thermocouple is affixed to each heater block and two on the sample. To minimize temperature fluctuations the apparatus is screened from drafts by an enclosure. The voltage difference is measured between the copper legs of the thermocouples and read by a Keithley 182 Sensitive Digital Voltmeter and displayed on a National Instruments LabVIEW 6.1 program which averages and displays the voltage. A simplified diagram of the Seebeck Coefficient measurement apparatus is shown in Figure 6.22 and a photograph of the a pparatus is shown in Figure 6.23 with detail of the thermal coup le positioning. The possible sources of error are compensation error, linearization error, measurement error, thermocouple wire error, and noise error. To minimize errors, the

PAGE 159

139 SCB68 shielded desktop connector block has a temperature sensor for cold juncti on compensation (CJC). The temperature sensor is powered with switches S1, S2, and S3 set for single ended or differential mode. This configuration also powers the signal conditioning area and circuitry. CJC is accurate only if the temperature sensor readi ng is close to the actual temperature of the screw terminals. Therefore, when reading the thermocouples, the SCB 68 is kept away from drafts or other temperature gradients, such as those caused by heaters, radiators, fans, and warm equipment [123].

PAGE 160

140 Figure 6.21 A schematic wiring diagram of the apparatus used to measure the temperature gradient and Seebeck voltage, used to determine the Seebeck coefficient. T H T L Voltage Data Collection Computer Temperature Data Collection Computer Heater Block Heater/Heat Sink Heater Elements T Type Thermocouples Silver Ink Pads Sample Heater Power Supply Heater Power Supply Enclosure VT H B V T LB VTH VTL NI SCB 68 CJC Temperature Sensor S1 S2 S3 3 0 6 3 33 66 65 31 28 61 VS +

PAGE 161

141 6. 5. 3 .2 Seebeck Apparatus Characterization The T type thermocouples utilized in the Seebeck apparatus were characterized to determine the relative precision of the chosen sensors. A bead of silver ink was placed on heater blo ck 1 and the tips of the four thermocouples were placed in the silver bead to V Cu T H T H T L Constantan Constantan Cu Cu Sample Heat Source Heat Sink T L Figure 6.22 Diagram of the Seebeck coefficient measurement apparatus. Figure 6.23 Photographs of the (a) Seebeck apparatus and DAQ board and (b) detail of the sample and thermocouple placement. (a) (b)

PAGE 162

142 improve thermal anchoring (Figure 6. 24). The temperature data was collected using the Figure 6.24 Arrangement of the T type thermocouples on the surface of the hater block for characterization of the thermocouples. TC1 TC 2 TC 3 TC 4 Silver Bead S.S. Heater B lock Heater Element 44 46 48 50 52 54 56 58 0 10 20 30 40 50 60 Time (min.) Temperature (C) TC1 TC2 TC3 TC4 Figure 6.25 Time required for the heater block temperature to stabilize with 15 V AC applied to the heater.

PAGE 163

143 National Instruments (NI) program Temperature.vi at a rate of 4 samples per second, 100 dat a points, and the average and standard deviation were calculated. The temperature was recorded periodically over 60 minutes. The time required for the heater block temperature to stabilize was investigated and a typical response of the various voltages is shown in Figure 6.25, when 15 V AC was applied to the heater element. It can be seen that a typical time for the temperature to stabilize was approximately 30 minutes, also that TC4 falls outside the typical response of the thermocouples, and based on the standard deviation TC1 and TC2 are not significant different. The average stabilized temperature measured by the probes was 54.74 0.50 C when the heater voltage was 15 V AC. To determine a reasonable temperature achievable by the heater blocks, voltage was applied to the heater element at approximate values of 0 (room temperature), 2.5, 5, 10, 15, 20 and 25 volts AC and the resultant temperature measured by the four probes 20 30 40 50 60 70 80 90 100 110 0 5 10 15 20 25 30 Heater Voltage (V) Temperature (C) TC1 TC2 TC3 TC4 Figure 6.26 The stabilized temperature recorded by the thermocouples at various heater voltages.

PAGE 164

144 was recorded. The results are shown in Figure 6.2 6, and at a heater voltage of 25 V and average temperature of 100.55 0.71 C was achieved. Again, based on the standard deviation TC1 and TC2 are not significantly different, and exponential trend lines determined that the R2 value for TC1 and TC2 were 0.9940 and 0.9941, respectively. The efficacy of the apparatus enclosure to minimize temperature measurement errors was investigated. As an example, when the heater block was adjusted to 25 V and TC1 measured the temperature to be 99.18 0.20 C with the enclosure and 87.53 2.05 C wit hout the enclosure in place. Based on these results, using an apparatus enclosure assists in reducing and temperature fluctuations and radiative loses. This will lead to reduction in measurement errors that are reportedly problematic of Seebeck coefficient measurements. It was also determined that TC1 and TC2 were to be used to measure the temperature of the sample, TH and TL, respectively, while the lesser accurate TC3 and TC4 were used to monitor the heater block and heat sink temperatures, respectively. A random thin film sample on a quartz substrate was affixed in the apparatus and the average temperature of the sample was adjusted for trials at approximately 27, 63, and 95 C. The temperature of heater block 1 was then adjusted to vary the temperature d ifference, = TH TL, within 0 to 5 C. As shown in Figure 6.27, at temperatures near ambient the deviation in temperature is lowest, approximately 0.28 C, resulting in low deviation of the Seebeck voltage potential, As the temperature is increased much greater than ambient, approximately 95 C, the deviation of and is strongly significant. In order to minimize error, temperatures near ambient will be utilized in determination of the Seebeck coefficient utilizing this particular apparatus.

PAGE 165

145 y = 145.71x 15.992 R2 = 0.9995 y = 154.56x 10.688 R2 = 0.9895 y = 168.35x 37.143 R2 = 0.9974 0 100 200 300 400 500 600 700 800 900 0 1 2 3 4 5 6 27C 63C 95C Figure 6.27 The voltage potential as a function of the temperature difference at various average temperatures. Figure 6.28 The ther mopower ( b ) of the Cu lead of the thermocouples used in the Seebeck apparatus. 1.90 1.91 1.92 1.93 1.94 1.95 1.96 23 24 25 26 27 28 29 30 Temperature (C)

PAGE 166

146 The thermocouple lead contribution ( b) to the thermopower (see equation 6.2) was determined by affixing a sample of the Cu lead wire to the Seebeck apparatus and measuring the resultant voltage potential a s a function of the temperat ure differential and calculating the thermopower ( b). The lead contribution must be known and subtracted from each data point at each temperature. The results are shown in Figure 6.28 and are consistent with references. 6. 6. Summary In conclusion, this chapter described the topography, structure, and electrical transport properties of the thin films deposited by single and dual laser ablation. The Ba8Ga16Ge30 material was found produce undesirable particulates and molten droplets, weak crystallinity, and excessively high resistivity when deposited by KrF single laser ablation. These poor thin film characteristics were circumvent ed by the use the dual laser ablation technique. Through proper coupling of the CO2 laser energy with the excimer laser generated plasma, control of particle reduction was demonstrated as well as high Table 6. 6 The re sults of the room temperature thermopower measurements ( S ) and power factor ( S2/ ) for thin films deposited by single and dual laser ablation at various thicknesses ( d). The resistivity determined by the Van der Pauw method (Table 6.2) are included for ref erence. p Fluence d S S2/ (J/cm2) (nm) ( m cm ) ( V / K ) ( W / cmK2) 1 Single 3 UV 300 6 7 104-864 1 1 10-2 2 Dual 1 UV + 2 IR 202 3 58 103-117 3 8 10-3 3 Dual 1 UV + 2 IR 269 1 60 103-119 8 9 10-3 4 Dual 1 UV + 2 IR 435 4 31 102-112 2 9 10-2 5 Dual 1 UV + 2 IR 632 2 10 101-105 5 3 10-1 Sample PLD

PAGE 167

147 crystallinity. These characteristics lead to decreased film resistivity and desirable thermopower. Thicker film depositions were performed to determine the effect on the electric tran s port properties. However, but due to the development of micro cracks in the high fluence single laser thin films with thickness greater than 300 nm, the resistivity increased due to discontinuity of the film. This was not the case with dual laser deposited thin films. High resistivity values may be due to vacancies on the framework sites [5]. Vacancies may introduce shallow defects near the conduction band, lowering the Hall mobility and thus increasing the resistivities as compared to Ba8Ga16Ge30. The re n) were determined using the van der Pauw Hall method. The thermopower, or Seebeck coeffiecient ( S, V/K), was determined using a thermopower differential method with small These results, Table 6.4, were used to calculate the thermoelectric power factor, P = S2/ ( W/cmK2), for comparison. With increasing film thickness the resistivity decreased by approximately two orders of magnitude, assumed to be due to reduced surface interface effects. The carr ier density increased from 1016 toward 1019 cm-3, a range approaching the optimum value. The temperature gradient established for determination of the Seebeck coefficient is driven predominately by the thermal conductivity (1.3 W/m K) and mass of the amorp hous quartz substrate and film surface losses rather than the inherent characteristics of the Ba8Ga16Ge30 clathrate material. Improvement of the power factor was shown with increasing film thickness, with a maximum room temperature power factor of 0.53 W/ cmK2 achieved which is comparable to reported values for the bulk material near room temperature.

PAGE 168

148 CHAPTER 7: CONCLUSIONS The aim of this work was to grow thin films of the type I clathrate Ba8Ga16Ge30, for thermoelectric applications by pulsed laser ablation. Type I clathrates exhibit a multicomponent, complex structure composed of elements wit h a broad range of physical properties. These characteristics can be challenging parameters for thin film deposition processes. Pulsed laser ablation/deposition is a technique which offers unique advantage s for successful deposition of complex multicompone nt materials such as Ba8Ga16Ge30. Under the proper choice of laser fluence films have the same composition as the target. This sets PLD apart from incongruent transfer methods such as thermal evaporation and sputtering. The source material was polycrystall ine type I Ba8Ga16Ge30 clathrate powder synthesized by the Novel Materials Laboratory at the Department of Physics University of South Florida under the supervision of Dr. George Nolas. The source material was cold pressed into a packed powder target for pulsed laser ablation. This work was divided into four main areas of study; (i) laser target interaction to determine ablation parameters that are conducive to stoichiometric evaporation (ii) in situ optical diagnostics of the ablated material to investi gate plume expansion dynamics (iii) deposit ion of thin film s, and (iv) investigation of their properties.

PAGE 169

149 Ablation from a stoichiometric target leading to stoichiometric evaporation of the target material, minimal ejection of particulate s or droplet s in t he evaporated material plume, and congruent expansion of the ablated elemental components is desirable The laser target interaction studies involved impinging single laser pulsed and dual laser pulses on the target surface at various laser fluences and to tal laser pulses per target site. The surfaces of the ablated target region s were examined by scanning electron microscopy (SEM) to determine the morphological changes of the target. It was found that low fluence single laser ablation, 1 J/ cm2, of the targ et developed conical structures that are about 50 m wide, which is indicative non stoichiometric evaporation of the material. High fluence single laser ablation, 3 J/ cm2, resulted in a smooth laser target melt zone which is conducive for stoichiometric ev aporation Preliminary studies found that single laser ablation lead s to varying degrees of particulate deposition while laser target studies for the dual laser ablation technique showed a smooth laser target melt zone similar to 3 J/ cm2 single laser ablat ion. Compositional analysis of the laser target interaction sites by energy dispersive spectroscopy (EDS) found near stoichiometric removal of material by the properly synchronized dual laser ablation as well as for high fluence single laser ablation Abla tion was not congruent for single laser deposition at fluences near 1 J/cm2. In situ optical diagnostics was performed to investigate the ionic emission of the ablated material, expansion profiles of the barium, gallium, and germanium components, and to de termine plasma temperatures of the various laser ablation. The diagnostics involved visible wavelength optical emission spectroscopy (OES), t wo dimensional ICCD imaging time gated imaging and time of flight spectroscopy The dual laser ablation was found to ionize the elemental components in the plume to a gr eater degree,

PAGE 170

150 as compared to it s single laser component (1 J/ cm2). The dual laser ablation conditions produced broader expansion profiles and maintained congruent expansion profiles of the components On the other hand expansion profiles of components produced by the single laser ablation d eviated from each other due to the small cone angle of expansion. A mathematical model of the plasma expansion was employed to determine the plasma temperature, with a comparison between single and dual laser ablation. The low and high fluence single laser ablation pr oduced plasma temperatures of 7600 K and 14700 K, respectively. Dual laser ablation plasma temp erature was calculated to be 284 00 K. The high plasma temp erature was the dual laser ablation characteristic which contributed to the high ionization, high velocity which leads to increased kinetic energy, and broader expansion profiles. These are all factors which are necessary for high quality thinfilm growth. Thin films deposited were investigated by x ray diffraction (XRD) to determination crys tallinity, scanning electron microscopy (SEM) to analyze the surface morphology, energy dispersive spectroscopy (EDS) to determine the elemental composition. It was fo und that deposition by the dual laser technique with modest laser energy densities produced thin films with better structural properties such as increased crystallinity, minimal density of particulates, and stoichiometric depositions. High energy single la ser depositions resulted in comparable crystallinity and stoichiometry, however, the particulate density on the films was unacceptable. The e lectric transport properties of films deposited by excimer only single laser ablation had high resistivities which were beyond the instruments range. This was found to be due to micro cracks which developed in the films, most probably due to the excessive strain in the deposited films. Dual laser deposited films produced superior

PAGE 171

151 electrical properties as compared to th at if the single laser deposited films The room temperature electric transport properties such as the Seebeck coefficient and power factor ( 2 ) were comparable to polycrystalline Ba8Ga16Ge30 bulk values.

PAGE 172

152 REFERENCES [ 1] J.C. Hemminger, G. Crabtree, A. Malozemoff, Science for Energy Technology: Strengthening the Link Between Basic Research and Industry , U.S. Department of Energy (2010); http://www.er.doe.gov/bes/reports/files/SETF_rpt_print.pdf [2] B.C. Sales, Critical Overview of Recent Approaches to Improved Thermoelectric Materials, International Journal of Applied Ceramic Technology Vol. 4, No. 4, pg. 291296 (2007). [ 3] J.R. Sootsman, D.Y. Chung, and M.G. Kanatzidis, New and Old Concepts in Thermoelectric Materials, Angewandte Chemi International Edition, Vol. 48, No. 46, pg. 86168639 (2009). [ 4] J. Zheng, Recent Advances on Thermoelectric Materials, Frontiers o f Physics in China, Vol. 3 No. 3, pg. 269 279 (2008). [ 5] G.J. Snyder and E.S. Toberer, Complex Thermoelectric Materials, Nature Materials Vol. 7, pg. 105114 (2008). [ 6] C.B. Vining, An Inconvenient Truth About Thermoelectrics, Nature Materials Vo l. 8, pg. 8385 (2009). [ 7] J. Yang and T. Caillat, Thermoelectric Materials for Space and Automotive Power Generation, MRS Bulletin Vol. 31, pg. 224229 (2006). [ 8] F.R. Stabler, Commercialization of Thermoelectric Technology, Materials Research Symposium Proceedings Vol. 886, 0886 F01 04.1 (2006). [ 9] H. Bttner, J. Nurmus, and A. Schubert, Miniaturized Thermoelectric Converters, in Thermoelectrics Handbook Macro to Nano, ed. D.M. Rowe, Chp. 46, pg. 1 18, CRC Press (2006). [ 10] J. Cancheevar am, R. alley, E. Siivola, and R. Venkatasubramanian, Performance of Integrated Thin Film Thermoelectrics in Cooling Hot Spots on Microprocessors Experimental Setup and Results, Materials Research Society Symposium Proceedings Vol. 793, S8.18.1 (2004 ).

PAGE 173

153 [ 11] T.M. Tritt and V.M. Browning, Overview of Measurement and Characterization Techniques for Thermoelectric Materials, in Recent Trends in Thermoelectric Materials Research II Semiconductors and Semimetals ed T.M. Tritt, Vol. 70, pg. 2549, Acad emic Press (2001). [ 12] R. Venkatasubramanian, E. Siivola, and B OQuinn, Superlattice Thin Film Thermoelectric Material and Device Technologies, in Thermoelectrics Handbook Macro to Nano ed. D.M. Rowe, Chp. 49, pg. 115, CRC Press (2006). [13] S.O Kasap, Thermoelectric Effects in Metals: Thermocouples, Web Materials, 3rd ed., (November 6, 2001) http://materials.usask.ca/samples/ThermoelectricSeebeck.pdf [14] D.M. Ro we, General Princiles and Basic Considerations, in Thermoelectrics Handbook Macro to Nano, ed. D.M. Rowe, Chp. 1, pg. 1 14, CRC Press (2006). [15] J.P. McKelvey, Statistical Mechanics, in Solid State Physics for Engineering and Material Science, Chp 5, pg. 187229, Krieger Publishing, Malabar, FL (1993). [16] N.W. Ashcroft and N.D. Mermin, The Sommerfeld Theory of Metals, Solid State Physics Chp. 2, pg. 2956, Saunders College Publishing, Fort Worth, TX (1976). [ 17] H.J. Goldsmid, Introduction, in Recent Trends in Thermoelectric Materials Research I Semiconductors and Semimetals ed T.M. Tritt, V. 69, pg. 1 24, Academic Press (2001). [ 18] R S. Muller and T I. Kamins Device Electronics for Integrated Circuits, 3rd ed ., pg 141170, Wiley ( 2003). [ 19] S.M. Sze, Metal Semiconductor Contacts, in Physics of Semiconductor Devices 2nd Ed ., Chp. 5, pg. 245311, Wiley, New York (1981). [20] T.M. Tritt and M.A. Subramanian, Thermoelectric Materials, Phenomena, and Applications: A Birds Eye View, MRS Bulletin Vol. 31, pg. 188198 (2006). [21] O. Yamashita and S. Sugihara, "High P erformance B ismuth T elluride C ompounds with H ighly S table T hermoelectric F igure of M erit," Journal of Materials Science V ol. 40, pp. 64396444 ( 2005 ) [22] D.Y. Chung T.P. Hogan, M.Rocci Lane, P. Brazis, J.R. Ireland, C.R. Kannewurf, M. Bastea, C. Uher, and M. G. Kanatzidis, "A New Thermoelectric Material:CsBi4Te6," Journal of the American Society of Chemistry V ol. 126, pp. 64146428 ( 2004 )

PAGE 174

154 [23] D.Y. Chung, T. Hog an, P. Brazis, M. Rocci Lane, C. Kannewurf, M. Bastea, C. Uher, and M. G. Kanatzidis, "CsBi4Te6: A High Performance Thermoelectric Material for Low Temperature Applications," Science, V ol. 287, pp. 10241027 ( 2000) [24] T. Caillat, J. P. Fleurial, and A. Borshchevsky, Preparation and Thermoelectric Properties of Semiconducting Zn4Sb3, Journal of Physics and Chemistry of Solids 58, (7), pp. 11191125, (1997). [ 25] G. Mahan, B. Sales, J. Sharp, Thermoelectric Materials: New Approaches to an Old Problem , Physics Today, Vol. 50, No. 3, pg. 42 47 (1997). [ 26] G.A. Slack, New Materials and Performance Limits for Thermoelectric Cooling, in CRC Handbook of Thermoelectrics (D.M. Rowe, ed.) pg 407, CRC Press, Boca Raton, FL (1995). [27] G.A. Slack, Design Concepts for Improved Thermoelectric Materials, Materials Research Society Symposium Proceedings Vol. 478, pg. 4754 (1997). [28] G.S. Nolas, J.L. Cohn, G.A. Slack, and S.B. Schujman, Semiconducting Ge Clathrates: Promising Candidates for Thermoelectr ic Applications, Applied Physics Letters, Vol. 73, No. 2, pg. 178180 (1998). [29] G.S. Nolas, G.A. Slack, and S.B. Schujman, Semiconductor Clathrates: A Phonon Glass Electron Crystal Material with Potential for Thermoelectric Applications, in Recent T rends in Thermoelectric Materials Research I Semiconductors and Semimetals (T.M. Tritt, ed.) Vol. 69, pg. 255300, Academic Press, San Diego, CA (2001). [30] G.S. Nolas, C.A. Kendziora, Ramen Scattering Study of Ge and Sn Compounds with Type I Clathrat e Hydrate Crystal Structure, Physical Review B, Vol. 62, No. 11, pg. 7157 (2000). [31] G.S. Nolas, X. Lin, J. Martin, M. Beekman, and H. Wang, Open Structured Materials: Skutterudites and Clathrates, Journal of Electronic Materials Vol. 38, No. 7, pg. 10521055 (2009). [32] G.S. Nolas, M. Beekman, J. Martin, D. Wang, and X. Lin, Bulk Materials Research for Thermoelectric Power Generation Applications, Materials Research Society Symposium Proceedings Vol. 044 (2008). [33] J. Martin, S. Erickson, G.S. Nolas, P. Alboni, T. Tritt, and J. Yang, Structural and Transport Properties of Ba8Ga16SixGe30 x Clathrates, Journal of Applied Physics Vol. 99, No. 4, pg. 044903 (2006).

PAGE 175

155 [34] G.S. Nolas, Structure, Thermal Conductivity, and Thermoelectric Properti es of Clathrate Compounds, in Thermoelectrics Handbook Macro to Nano, ed. D.M. Rowe, Chp. 33, pg. 1 8, CRC Press (2006). [34] J. Martin, H. Wang, and G.S. Nolas, Optimization of the Thermoelectric Properties of Ba8Ga16Ge30, Applied Physics Letters V ol. 92, No. 22, pg 222110 (2008). [35] H. Bttner, G. Chen, and R. Venkatasubramanian, Aspects of Thin Film Superlattice Thermoelectric Materials, Devices, and Applications, MRS Bulletin, Vol. 31, pg. 211217 (2006). [36] R. Venkatasubramanian, E. Siiv ola, T. Colpitts, and B. OQuinn, Thin Film Thermoelctric Devices with High Room Temperature Figures of Merit, Nature Vol. 413, pg. 597602 (2001). [37] J. Nurnus, H. Bttner, C. Knzel, U. Vetter, A. Lambrecht, J. Schumann, and F. Vlklein, Thin Film Based Thermoelectric Energy Conversion Systems, [38] M. Takashiri, T. Borca Tasciuc, A. Jacquot, K. Miyazaki, and G. Chen, Structure and Thermoelectric Properties of Boron Doped Nanocrystalline Si0.8Ge0.2 Thin Films, Journal of Applied Physics Vol. 100, No. 5, pg. 054315 (2006). [39] R. Venkatasubramanian, B. OQuinn, E. Siivola, K. Coonley, P. Addepally, M. Napier, and T. Colpitts, Superlattice Thin Film Thermoelectric Materials and Devices, Materials Resea r ch Societ y Symposium Proceedings Vol. 793, pg. S2.3.1 (2004). [ 40] A. Giani, F. Pascal Delannoy, A. Boyer, A. Foucaran, M. Gschwind, P. Ancey, Thin Solid Films 303 (1997) 817. [41] J. Dheepa, R. Sathyamoorthy, and S. Velumani, Characterization of Bismuth Telluride thin Films Flash Evapora tion Method, Materials Characterization, Vol. 58, pg. 782785 (2007). [4 2] F. Volklein, V. Baier, U. Dillner, E. Kessler, Transport Properties of Flash Evaporated (Bi1 -xSbx)2Te3 Films 1. Optimization of Film Properties, Thin Solid Films Vol. 187, No. 2, pg. 253262 (1990). [ 43] A. Boyer, E. Cisse Properties of Thin Film Thermoelectric Materials Application to Sensors Using the Seebeck Effect , Materials Science and Engineering B Solid State Materials for Advanced Technology Vol. 13, No. 2, pg. 103111 (1992). [ 44] H. Noro, K. Sato, H. Kagechika, The Thermoelectric Properties and Crystallography of Bi Sb Te Se Thin Films Grown by Ion Beam Sputtering, J ournal of Applied Physics Vol. 73, No. 3, pf 12521260 (1993).

PAGE 176

156 [45] I.S. Virt, T.P. Shkumba tyuk, I.V. Kurilo, I.O. Rudyi, T.Ye. Lopatinskyi, L.F. Linnik, V.V. Tetyorkin, and A.G. Phedorov, Deposition of Thin Bi2Te3 and Sb2Te3 Films by Pulsed Laser Ablation, Semiconductors Vol. 44, No. 4, pg. 544549 (2010). [46] J.J. Dubowski, Chemotronics, Vo l 3, No. 7, pg. 66 (1988). [ 47] S. Witanachchi, K. Ahmed, P. Sakthivel and P. Mukherjee, Dual Laser Ablation for particulateFree Film Growth, Applied Physics Letters Vol. 66, No. 12, pg. 14691471 (1995). [ 48] P. Mukherjee, J.B. Cuff, and S. Witan achchi, Plume E xpansion and S toichiometry in the G rowth of M ulti C omponent T hin F ilms U sing D ual L aser A blation, Applied Surface Science Vol. 127129, pg. 620625 (1998). [ 48] P. Mukherjee, J.B. Cuff, and S. Witanachchi, Plume E xpansion and S toichiome try in the G rowth of M ulti C omponent T hin F ilms U sing D ual L aser A blation, Applied Surface Science Vol. 127129, pg. 620625 (1998). [49] P. Mukherjee, S Chen, J.B. Cuff, P. Sakthivel, S. Witanachchi, Evidence for the Physical Basis and Universality o f the Elimination of Particulates Using DualLaser Ablation. I. Dynamic Time Resolved Target Melt Studies, and Film Growth of Y2O3 and ZnO, Journal of Applied Physics Vol. 91, No. 4, pg. 18281836 (2002) [ 50] S.R. Foltyn, Surface Modification of Mater ials by Cumulative Laser Irradiation in Pulsed Laser Deposition of Thin Films (D.B. Chrisey and G.K. Hubler, eds.) Wiley, New York, pp. 89113 (1994). [ 51] H. Sankar, W.J. Gunning, J. DeNatale, and J.F. Flintoff, J. Applied Physics, Vol. 65, 2475 (1986). [ 52] T. Venkatesan and Steven M. Green, Pulsed Laser Deposition: Thin Films in a Flash The Industrial Physicist pp. 22 (1996). [53] H. Sankur and J.T. Cheung, Formation of Dielectric and Semiconductor Thin Films by Laser Assisted Evaporation, Appli ed Physics A: Solids and Surfaces Vol. 47, pg 271284 (1988). [54] P.R. Willmott and J.R. Huber, Pulsed Laser Vaporization and Deposition, Review of Modern Physics Vol. 72, No. 1, pg. 315328 (2000). [ 55] L.C. Chen, Particulates Generated by Pulsed Laser Ablation in Pulsed Laser Deposition of Thin Films (D.B. Chrisey and G.K. Hubler, eds.) Wiley, New York, pp. 167198 (1994).

PAGE 177

157 [56] J.F. Ready, Laser Induced Particle Emission, in Effect of High Power Laser Radiation, Chp. 4 (Academic, New York, 1971). [57] A.D. Sappey and N.S. Nogar, Diagnostic Studies of Laser Ablation for Chemical Analysis, in Laser Ablation Principles and Applications (J.C. Miller, ed.), Springer Verlag, Berlin, pp. 159160 (1994). [ 58] T.P. Hughes, Plasma and Laser Light Wi ley, New York, (1975). [59] L. Wiedeman and H. Helvajian Laser P hotodecomposition of S intered YBa2Cu3O6+ x: Ejected S pecies P opulation D istributions and I nitial K inetic E nergies for the L aser A blation W avelengths 351, 248, and 193 nm , Journal of Applied Physics Vol. 70, No. 8, pg. 4513 4523 (1991). [60] R.W. Dreyfus Cu0, Cu+, and Cu2 from E xcimer A blated C opper , Journal of Applied Physics Vol. 69, No.3, pg 1721 1729 (1991). [ 61] P.J. Martin, J. Material Science, Vol. 21, 1 (1983). [ 62] S.V. Gaponov, B.M. Luskin, B.A. Nesterov, N.N. Salashchenko, Sov. Phys. Solid. State, Vol. 19, 1736 (1977). [63] D.H.A. Blank, R.P.J. Ijsselsteijn, P.G. Out, H.J.H. Kuiper, J. Floksura, and H. Rogalla, Material Science Engineering, Vol. B13, 67 (1992). [64] W.P. Barr, The P roduction of L ow S cattering D ielectric M irrors U sing Rotating V ane P article F iltration , Journal Physics E: Journal Sci. Instruments Vol. 2, No. 2, pg. 11121114 (1969). [65] R.J. Kennedy, Thin Solid Films Vol. 214, 223 (1992). [66] D.S. Lubden, S.A. Barnett, K. Suzuki, S. Gorbatkin, and J.E. Greene, Laser I nduced P lasmas for P rimary I on D eposition of E pitaxial Ge and Si F ilms , Journal of Vac uum Science Technology Vol. B3, pg. 968 (1992). [67] A. Gupta, Novel Pulsed Laser Deposition Approaches in Pulsed Laser Deposition of Thin Films (D. B. Chrisey and G. K. Hubler, eds.), Wiley, New York, pp. 285 287 (1994). [68] P. Mukherjee, S. Chen, S. Witanachchi, Effect of Initial Plasma Geometry and Temperature on Dynamic Plume Expansion in Dual Laser Ablation, Applied Physics L e tters Vol. 74, No. 11, pg. 15461548 (1999) [ 69] S. Witanachchi and P. Mukherjee, Role of Temporal Delay in Dual Laser Ablated Plumes, Journal of Vacuum Science and Technology A Vol. 13, No. 3, pg. 11711174 (1995).

PAGE 178

158 [ 70] D.H. Lowndes, D.B. Geohegan, A.A. Puretzky, D.P. Norton, and C.M. Rouleau, Synthesis of Novel Thin Film Materials by Pulsed Laser Deposition, Science, Vol. 273, No. 5277, pg. 898903 (1996). [71] T.M. Tritt and M.A. Subramanian, Thermoelectric Materials, Phenomena, and Applications: A Birds Eye View, MRS Bull etin Vol. 31, pg. 188198 (2006). [72] L.Miao, S. Tanemura, T. Watanabe, M. Tanemura, S. Toh, K. Kaneko, Y. Sugahara, and T. Hirayama, Epitaxial Growth of Ba8Ga16Ge30 Clathrate Films on Si Substrate by RF Helicon Magnetron Sputtering with Evaluation on Thermoelectric Properties, Applied Surface Science Vol. 254, pg. 167172 (2007). [73] R.K. Singh, D. Bhattacharya, and J. Narayan, Laser Target Interactions and its Effect on Surface Morphology of Laser Deposited Thin Films Material Research Society Symposium Proceedings, Vol. 201, pg. 427432 (1991). [74] S.M. Green, A. Pique, K.S. Harshavardhan, and J.S. Berstein, Equipment in Pulsed Laser Deposition of Thin Films (D. B. Chrisey and G. K. Hubler, eds.), Wiley, New York, pp. 2354 (1994). [75] R. Hyde, M. Beekman, G.S. Nolas, P. Mukherjee, and S. Witanachchi, Growth and Characterization of Germanium Based Type I Clathrate Thin Films Deposited by Pulsed Laser Ablation, in Advances in Electronic Ceramics, Ceramic Engineering and Science Proceeding s (C. Randal, Hua Tay Lin, K. Koumoto, and P. Clem, eds.), 2007 Proceedings of the 31st International Conference on Advanced Ceramics and Composites, American Ceramics Society, Vol. 28, Issue 8, pg. 211, Wiley (2008). [76] H.F. Wan g, K.F. Cai, H. Li, D.H. Yu, X. Wang, C.W. Zhou, X.L. Li, Y.Y. Wang, B.J. An, and Y. Du, Synthesis and Thermoelectric Properties of Single Crystalline and Poly crystalline Ba8Ga16Ge30 Journal of Alloys and Compounds Vol. 491, pg. 684688 (2010). [77] S. Witanachchi, R. Hyde, M. Beekman, D. Mukherjee, P. Mukherjee, and G.S. Nolas, Synthesis and Characterization of Bulk and Thin Film Clathrates for Solid State Power Conversion Applications 25th International Conference on Thermoelectrics pg. 4447 (2006). [78] H.F. Wan, K.F. Cai, H. Li, D.H. Yu, X. Wang, C.W. Zhou, X.L. Li, Y.Y. Wang, B.J. An, and Y. Du, Synthesis and Thermoelectric Properties of Single Crystalline and Poly crystalline Ba8Ga16Ge30 Journal of Alloys and Compounds Vol. 491, pg. 684688 (2010). [79] J.D. Bry an, V.I Srdanov, G.D Stucky, D. Schmidt, Superconductivity in Germanium Clathrate Ba8Ga16Ge30 Physical Review B Vol. 60, Number 5, pg. 30643067 (1999).

PAGE 179

159 [80] Y. Yan, X. Tang, P. Li, and Q. Zhang, Microstructure and Thermoelectric Transport Properties o f Type I Clathrates Ba8Sb2Ga14Ge30 Prepared by Ultrarapid Solidification Process, Journal of Electronic Materials Vol. 38, No. 7, pg. 12781281 (2009). [ 81] A. Saramat, E.S. Toberer, A.F. May, and G.J. S nyder Thermal Stability and Phase Purity in Poly crystalline Ba8GaxGe46-x Journal of Electronic Materials Vol. 38, No. 7, pg 1423 1426 (2009). [82] D.J. Burgess, P.C. Stair, E. Weitz. Calculations of the S urface T emperature R ise and D esorption T emperature in L aser I nduced T hermal D esorption , Journal of Vacuum Science Technology Vol. A4 No. 3, pg. 13621366 (1986). [83] T. Hirata and Z. Miyazaki, High Speed Camera Imaging for Laser Ablation Process: For Further Reliable elemental Analysis Using Inductivey Coupled Plasma Mass Spectroscopy, Analyt ical Chemistry Vo. 79, No. 1, Pg 147 152 (2007). [84] D.B. Geohegan, Diagnostics and Characteristics of Pulsed Laser Deposition Laser Plasmas, in Pulsed Laser Deposition of Thin Films (D.B. Chrisey and G.K. Hubler, eds.) Wiley, New York, pp. 115165 (1 994). [85] G. Dinescu, M.N. Oliveira, O. Conde, Emission Characteristics of the Plasma Plume Generated During KrF Laser Ablation of a Cr2O3 Target in Vacuum, Applied Surface Science Vol. 208209, pg. 9095 (2003). [86] A. De Giacomo, Experimental Characterization of Metalic Titanium Laser Induced Plasma by Time and Space resolved Optical Emission Spectroscopy, Spectrochimica Acta Part B, Vol. 58, Pg. 71 83 (2003). [87] M A Khater, P van Kampen, J T Costello, J P Mosnier and E T Kennedy, Time integrated laser induced plasma spectroscopy in the vacuum ultraviolet for the quantitative elemental characterization of steel alloys, Journal of Physics D: Applied Physics Vol. 33, pg. 22522262 (2000). [90] D. Mukherjee, R. Hyde T. Dhakal, H. Srikan th, P. Mukherjee and S. Witanachchi. Investigation of the Pb Depletion in Single and Dual Pulsed Laser Deposited Epitaxial PZT Thin Films and Their Structural Characterization, in Multiferroic and Ferroelectric Materials (A. Gruverman, C.J. Fennie, I. Ku nishima, B. Noheda, T.W. Noh, ed s .) 2009 Materials Research Symposium Poceedings, Vol. 1199E, Vol. 1199E, pg. 1199F03 37, Warrendale, PA (2010). [ 91] P.K. Bhat, J.T. Dubowski, and D.F. Williams, Detection of O ptical E mission from InSb I nduced by P ulsed L aser E vaporation , Applied Physics Letters Vol. 47 No.10, pg. 1085 1087 (1985).

PAGE 180

160 [92] J. R. Roberts Optical Emission Spectroscopy on the Gaseous Electronics Conference RF Reference Cell , Journal of Research of the National Institute of Standards and Technology Vol 100, N o. 4, pg. 353 371 ( 1995 ). [93] NIST Physics Laboratory, Handbook for Basic Atomic Spectroscopic Data, [94] R.K. Singh and J. Narayan, Pulsed Laser Evaporation Technique for Deposition of Thin Films: Physics and Theoretical Model , Physical Review B Vol. 41, No. 13, pg. 88438859 (1990). [95] R.K. Singh, O.W. Holland, and J. Narayan, Theoretical Model for Deposition of Superconducting Thin Films Using Pulsed Laser Evaporation Technique, Journal of Applied Physics Vol. 68, No 1, gg. 233 247 (1990). [96] C.L. Liu, J.N. Leboeuf, R.F. Wood, D.B. Geobegan, J.M. Donato, K.R. Chen, A.A. Puretzky, Computational Modeling of Physical Processes During Laser Ablation, Materials Science and Engineering B, Vol. 47, pg. 7077 (1997). [ 97] S. Chen, CO2 LaserTarget Interaction and Plume Imaging in Dual Laser Ablation, Thesis, University of South Florida (1999). [ 98] J.B. Cuff, Plume Dynamics and Laser Pulse Synchronization in Dual Laser Ablation, Thesis, University of South Florida (2000). [ 99] S. Witanachchi, R. Hyde, H. S. Nagaraja, M. Beekman, G. S. Nolas, and P. Mukherjee, Growth and Characterization of Germanium Based Type I Clathrate Thin Films Deposited by Pulsed Laser Ablation, in Materials and Technologies for Direct Thermal t oElectric Energy Conversion (J. Yang, T.P. Hogan, R. Funahashi, G.S. Nolas, eds.) 2005 Materials Research Society Symposium Proceedings, Vol. 886, pg. 0886F10 03, Warrendale, PA (2006). [ 100] S. Witanachchi, R. Hyde, M. Beekman, D. Mukherjee, P. Mukherj ee, and G. S. Nolas, Synthesis and Characterization of Bulk and Thin Film Clathrates for Solid State Power Conversion Applications, IEEE Proceedings of the 25th International Conference on Thermoelectrics Vienna, Austria, pg. 4447 (2006). [ 101] A.F. M ay, E.S. Toberer, A. Saramat, and G.J. Snyder, Characterization and Analysis of Thermoelectric Transport in nType Ba8Ga16xGe30+x Physical Review B Vol 80, No. 12, pg. 125205 (2009). [102] M I. Aroyo, A. Kirov, C. Capillas, J.M. Perez Mato and H. Won dratschek. "Bilbao Crystallographic Server II: Representations of crystallographic point groups and space groups". Acta Cryst., A62, 115 128 (2006). [103] J.A. Bearden and A.F. Burr, "Reevaluation of X Ray Atomic Energy Levels," Rev. Mod. Phys. 39 pg. 125 (1967)

PAGE 181

161 [104] M. Cardona and L. Ley, Eds., Photoemission in Solids I: General Principles (Springer Verlag, Berlin, 1978) [105] J.C. Fuggle and N. Mrtensson, "Core Level Binding Energies in Metals," J. Electron Spectrosc. Relat. Phenom. 21, pg. 275 (19 80) [106] T.M. Tritt, Measurement and Characterization Techniques for Thermoelectric Materials, Materials Research Society Symposium Proceedings Vol. 478, pg. 2536 (1997). [107] T.M. Tritt, Electrical and Thermal Transport Measurement Techniques f or the Evaluation of the Figure of Merit of Bulk Thermoelectric Materials, in Thermoelectrics Handbook Macro to Nano (D. M. Rowe, ed.) CRC Press, Boca Raton, ch. 23 pg. 210 (2006). [108] E. Mller, C. Stiewe, D.M. Rowe, and S.G.K. Williams, Approaches to Thermoelectric Standardization, in Thermoelectrics Handbook Macro to Nano (D.M. Rowe, ed.) CRC Press, Boca Raton, ch. 26 pg. 4 6 (2006). [109] A.T. Burkov, Measurements of Resistivity and Thermopower: Principles and Practical Realization, in Thermoe lectrics Handbook Macro to Nano (D.M. Rowe, ed.) CRC Press, Boca Raton, ch. 22 pp. 4 6 (2006). [110] D.M. Rowe, General Principles and Basic Considerations, in Thermoelectrics Handbook Macro to Nano (D.M. Rowe, ed.) CRC Press, Boca Raton, chp. 1, pg. 114 (2006). [111] W. Gee and M. Green, An Improved Hot Probe Apparatus for the Measurement of Seebeck Coefficient, Journal of Physics E: Scientific Instruments Vol. 3, pg. 135136 (1970). [112] C. Kemper and G. Bringert, Temperature Measurements, in Validation of Pharmaceutical Processes 3rd ed (J.P. Agalloco and F.J. Carleton, eds.) Informa Healthcare, ch. 8, pg. 109 128 (2007). [113] V. Ponnambalam, S. Lindsey, N.S. Hickman, and T.M. Tritt, Sample Probe to Measure Resistivity and Thermopower in t he Temperature Range of 300 1000 K, Review of Scientific Instrunments Vol. 77, pg. 073904 (2006). [114] O. Yamashita, Effect of Metal Electrode on Seebeck Coefficient of p and nType Si Thermoelectrics, Journal of Applied Physics Vol. 95, No. 1, pg. 178183 (2004). [115] P. Pichanusakorn and P.R. Bandaru, The Optimal Seebeck Coefficient for Obtaining the Maximum Power Factor in Thermoelectrics, Applied Physics Letters Vol. 94, No. 22, pg. 223108 (2009).

PAGE 182

162 [116] H.J. Goldsmid, A Simple Technique fo r Determining the Seebeck Coefficient of Thermoelectric Materials, Journal of Physics E: Scientific Instruments Vol. 19, pg. 921922 (1986). [117] M. Fraser, Apparatus for the Rapid Measurement of Seebeck Coefficient, Journal of Scientific Instruments Vol. 39, No. 5, pg. 227228 (1962). [118] W.H. Kettler, R. Wernhardt, and M. Rosenberg, Differential AC Method of Thermopower Measurement, Review of Scientific Instruments Vol. 57, No. 12, pg. 30533058 (1986). [119] C.J.M. Lasance, The Seebeck Coe fficient, Electronics Cooling Magazine Vol. 12, No. 4 (2006). [120] R.J. Moffat, Notes on Using Thermocouples, Electronics Cooling Magazine Vol. 3, No. 1, (1997). [121] H. Werheit, U. Kuhlmann, B. Herstell, and W. Winkelbauer, Reliable Measurement of Seebeck Coefficient in Semiconductors, Journal of Physics: Conference Series, Vol. 176, pg. 012037 (2009). [122] A.T. Burkov. Measurements of Resistivity and Thermopower: Principles and Practical Realization in Thermoelectrics Handbook Macro to Nano. D. M. Rowe (ed.). CRC Press, Boca Raton, Chp. 22, pg. 4 6 (2006). [123] Temperature Sensor and Thermocouple, in SCB 68 User Manual for Advanced Functions National Instruments Corporation, Chp. 2 (2009). [124] M Christensen, S Johnsen, and B B Ive rsen Thermoelectric C lathrates of T ype I , Dalton Transactions, Vol. 30, No. 4, pg. 978992 (2010). [125] D. Mukherjee R. Hyde T. Dhakal, H. Srikanth, P. Mukherjee and S. Witanachchi. Investigation of the Pb Depletion in Single and Dual Pulsed Laser Deposited Epitaxial PZT Thin Films and Their Structural Characterization, in Multiferroic and Ferroelectric Materials (A. Gruverman, C.J. Fennie, I. Kunishima, B. Noheda, T.W. Noh, eds.) 2009 Materials Research Society Symposium P r oceedings Vol. 1199E, pg. 1199F03 37, Warrendale, PA (2010). [126] P. Mukherjee, S. Chen, J.B. Cuff, P. Sakthivel, and S. Witanachchi, "Evidence for the Physical Basis and Universality of the Elimination of Particulates using Dual Laser Ablation. I. Dynamic Time Resolved Targ et Melt Studies, and Film Growth of Y2O3 and ZnO", Journal of Applied Physics Vol. 91, pg. 1828 1836 (2002).

PAGE 183

163 [127] P. Mukherjee, S. Chen, J. B. Cuff, and S. Witanachchi, "Evidence for the Physical Basis and Universality of the Elimination of Particulates using Dual Laser Ablation. II. Dynamic Time Resolved Target Reflectivity and Film Growth of Zn", Journal of Applied Physics Vol. 91, pg. 18371844 (2002). [128] P. Mukherjee, J. B. Cuff, and S. Witanachchi, "A Novel Technique for Low Jitter Dual Lase r Synchronization in a Thin Film Deposition System", Review of Scientific Instruments Vol. 72, pg. 2380 2386 (2001). [129] S. Witanachchi, A. M. Miyawa, and P. Mukherjee, "Highly Ionized Carbon Plasma Generation by Dual Laser Ablation for Diamond Like Ca rbon Film Growth", LaserSolid Interactions for Materials Processing, Materials Research Society Symposium P r oceedings Vol. 616, pg. 235240 (2000). [130] P. Mukherjee, S. Chen, and S. Witanachchi, Effect of Initial Plasma Geometry and Temperature on Dy namic Plume Expansion in Dual Laser Ablation, Applied Physics Letters, Vol. 74, pg. 15461548 (1999). [131] P. Mukherjee, J. B. Cuff, and S. Witanachchi, Plume Expansion and Stoichiometry in the Growth of Multi Component Thin Films Using Dual Laser Abl ation, Applied Surface Science Vol. 127129, pg. 620625 (1998). [132] S. Witanachchi, K. Ahmed, P. Sakthivel, and P. Mukherjee, "Dual Laser Ablation for ParticulateFree Film Growth", Applied Physics Letters Vol. 66, pg. 1469 (1995). [133] S. Witan achchi and P. Mukherjee, "Role of Temporal Delay in Dual Laser Ablated Plumes", Journal of Vacuum Science and Technology Vol. A13 pg. 11711174 (1995). [134] S. Wolf and R.N. Tauber, Silicon Processing for the VSLI Era 2nd ed, Vol. 1, pg. 104118, Lat tice Press (2000). [135] L.B. Valdes, Procedings IRE, Vol 42, pg. 420 (1954). [136] L.I. Maissel and R. Glang, Handbok of thin Films, McGraw Hill (1970). [137] H. Anno, M. Hokazono, M. Kawamura, and K. Matsubara Effect of Transition Element Substitution on Thermoelectric Properties of Semiconductor Clathrate Compounds , 22nd International Conference on Thermoelectrics (ICT 03) La Grande Motte, France, pg. 121126 (2003). [138] H. Anno, M. Hokazono, M. Kawamura, a nd K. Matsubara Effect of Transition Element Substitution on Thermoelectric Properties of Semiconductor Clathrate Compounds , 22nd International Conference on Thermoelectrics (ICT 03) La Grande Motte, France, pg. 121126 (2003).

PAGE 184

164 [139] E. S. Toberer, M. Ch ristensen, B.B. Iversen, and G.J Snyder H igh Temperature Thermoelectric E fficiency in Ba8Ga16Ge30, Physical Review B Vol. 77, No. 7, pg. 075203 ( 2008).

PAGE 185

165 APPENDICES

PAGE 186

166 A ppendix A: Wyckoff Positions Multiplicity Wyckoff letter Site Symmetry (0,y,z) (0,-y,z) (0,y,-z) (0,-y,-z) (z,0,y) (z,0,-y) (-z,0,y) (-z,0,-y) (y,z,0) (-y,z,0) (y,-z,0) (-y,-z,0) (y+1/2,1/2,-z+1/2) (-y+1/2,1/2,-z+1/2) (y+1/2,1/2,z+1/2) (-y+1/2,1/2,z+1/2) (1/2,z+1/2,-y+1/2) (1/2,z+1/2,y+1/2) (1/2,-z+1/2,-y+1/2) (1/2,-z+1/2,y+1/2) (z+1/2,y+1/2,1/2) (z+1/2,-y+1/2,1/2) (-z+1/2,y+1/2,1/2) (-z+1/2,-y+1/2,1/2) (1/4,0,1/2) (3/4,0,1/2) (1/2,1/4,0) (0,1/2,1/4) (0,1/2,3/4) (1/2,3/4,0) (x,x,x) (-x,-x,x) (-x,x,-x) (x,-x,-x) (x+1/2,x+1/2,-x+1/2) (-x+1/2,-x+1/2,-x+1/2) (x+1/2,-x+1/2,x+1/2) (-x+1/2,x+1/2,x+1/2) (-x,-x,-x) (x,x,-x) (x,-x,x) (-x,x,x) (-x+1/2,-x+1/2,x+1/2) (x+1/2,x+1/2,x+1/2) (-x+1/2,x+1/2,-x+1/2) (x+1/2,-x+1/2,-x+1/2) 2 a m-3. (0,0,0) (1/2,1/2,1/2) (1/4,1/2,0) (3/4,1/2,0) (0,1/4,1/2) (1/2,0,1/4) (1/2,0,3/4) (0,3/4,1/2) m.. Coordinates d -4m. 2 16 i .3. 6 c -4m. 2 24 k 6 Table A 1. The Wyckoff positions for the c rystallographic cubic space group Pm 3n applicable to the typeI clathrate Ba8Ga16Ge30 [102]

PAGE 187

167 A ppendix B : Grit Designation Table ISO/FEPA Grit Designation Average Particle Diameter (m) P12 1815 P16 1324 P20 1000 P24 764 P30 642 P36 538 P40 425 P50 336 P60 269 P80 201 P100 162 P120 125 P150 100 P180 82 P220 68 P240 58.5 P280 52.2 P320 46.2 P360 40.5 P400 35 P500 30.2 P600 25.8 P800 21.8 P1000 18.3 P1200 15.3 P1500 12.6 P2000 10.3 P2500 8.4 Extra Coarse (Very fast removal of material) Coarse (Rapid removal of material) Ultra fine (final sanding and polishing of thick finishes) Grit Size Table MACROGRITS MICROGRITS Medium (preparation for finishing, for gentle removal of varnish) Extra fine (start polishing) Super fine (final sanding of finishes) Fine (preparation for finishing, not suitable for removing varnish or paint, use for cleaning) Very Fine (sanding of bare material) Very Fine (sanding finishes between coats) Table B. 1. ISO grit designation and average grit diameter.

PAGE 188

168 A ppendix C : Properties of Ba8Ga16Ge30 Target Components Table C.1. Properties of the target components; barium, gallium, and germanium. Units Barium Galium Germanium Symbol Ba Ga Ge Atomic Number 56 31 32 Atomic Weight 137.327 69.723 72.61 Density g/cm33.62 5.91 5.32 Atomic Volume cm3/mol 39.24 11.8 13.6 Covalent Radius 1.98 1.26 1.22 Atomic Radius 2.78 1.81 1.52 Oxidation States 2 3 4 Electron Configuration (Xe)6s2(Ar)3d104s2p1(Ar)3d104s2p2 Melting Point C 727 29.8 937.2 Boiling Point C 1640 2403 2830 Acid-Base Properties basic weak amphoteric weak amphoteric Thermal Conductivity (300K) [W/(cmK)] 0.184 0.406 0.599 Specific Heat (300K) J/(gK) 0.204 0.37 0.32 Heat of Fusion (melting pt.) kJ/mol 7.750 5.590 36.940 Heat of Vaporization (boiling pt.) kJ/mol 142.0 258.70 330.90 Latent Heat of Fusion k-cal/gatom 111.5 Latent Heat of Vaporization cal/g 1100 First Ionization Potential (300K) V 5.212 5.999 7.899 Electronegativity Paulings 0.9 1.6 1.8 Electrical Resistivity microOhmscm 50 53.4 47 (27C) Electrical Conductivity (293K) 106 0.030 0.0678 1.45x104 Crystal Structure bcc Orthorhombic Diamond Work Function eV 2.5 4.5 Vapor Pressure mm Hg 760 (2400C) 760 (27C) Vapor Pressure Temp. (10-6 Torr) C 350 740 945 Cubic Coefficient of Expansion K-15.4x10-6 (20C) Linear Coefficient of Expansion K-16.1x10-8 Mohs Hardness 1.5-2.5 (20C) Magnetic Susceptibility emu/mole -22x106 (20C) Debye Temperature K 362

PAGE 189

169 A ppendix D : ICCD Camera System Specifications PI Acton PI MAX:512 UNIGEN Digital ICCD Camera System: Manufacturer: Princeton Instruments Acton Model: 73610017 Serial Number: 2911060001 Thompson TH7895M scientific grade 1, front illuminated CCD, 512 x 512 pixels 1.27:1 fiber optic bonded 18 mm grade 1, Gen III extended blue image intensifier ideal for 350 nm to 900 nm range. Fiber optic input, proprietary UV coating 19 x 19 m pixels (12.4 x 12.4 mm image area ) < 5 ns gating capable ICCD USB2 Controller: with 16 bit 100 kHz and 16bit 1 Mhz digitizers ST 133 B/U PTG Dual Model 75130001 Serial Number: 2911060002 Programmable Timing Generator (PTG): Delay Range: ~0 ns to 20 ms Timing Resolution: 40 ps Timing Jitter: 40 ps rms Repetition Rate: 50 kHz su stained; 500 kHz burst Gate Width: 5 ns to 10 ms; 40 ps resolution Fast Gate: 2 ns; 5 kHz repetition rate Bracket Pulse: 5 kHz repetition rate; 500 ns lead time Insertion Delay: 15 ns

PAGE 190

170 A ppendix E : Spectrometer Specifications Manufacturer: Acton Research Corporation Model: 1237 SpectraPro 500 Serial Number: 500173S Controller Serial Number: 100240V 2A Focal Length: 500 mm Optical System: Czerny Turner type with inline (180) optic al path, and optical multi port configuration featuring 90 and 180 optical path. Wavelength Scanning System: Direct Digital Scanning with exclusive AutotrackTM electronics. Scan Linearity: Scans linear with respect to wavelength. Triple Indexable Grat ings: 150, 600, and 1200 grooves/mm. Wavelength Display: Automatically displays correct wavelength for gratings installed. Displays grating position, groove spacing and blaze wavelength for gratings. Resolution: 0.05 nm with standard 1200 g/mm grating, 1 measured at 435.8 nm. Reciprocal Linear Dispersion: 1.67 nm/mm with 1200 g/mm grating (nominal). Aperture Ratio: f/6.9 Wavelength Operating Range: Up to the far infrared with available gratings. 185 nm to mm grating. Wavelength Accuracy: 0.2 nm/500 nm with a 1200 g/mm grating. Wavelength Reproducibility: 0.5 nm with 1200 g/mm. Focal Plane Detector Compatibility: 25 mm focal plane extends 0.750 inch (19 mm) beyond housing for easy positioning of focal plane detectors. Provides nominal 280 nm coverage with 150 g/mm grating, 140 nm with 300 g/mm grating, 70 nm with 600 g/mm grating, and 35 nm with 1200 g/mm grating.

PAGE 191

171 Slits: micrometer. Standard slit height is 4 mm. Other slit heights up to 20 mm can be supplied. Computer Compatibility: RS 232 port 9600 baud, no parity, 8 data bits, 1 start bit, 1 stop bit; or optional IEEE 48 port. Entrance Slit Triple Grating Turret End Port for Detector Figure E. 1: Spectrograph optical layout.

PAGE 192

172 A ppendix F : Fiber O ptic Bundles Standard bundles come with 200 m diameter fibers in silica for UV VIS. Fibers are arranged as a slit pattern (in a 10 mm diameter ferrule) on the spectrograph end, with a round configuration (in SMA 905 connectors) on the source end. Two Leg Fiber Bundle Model BFB 4557: a 1 m long UV_VIS fiberoptic bundle for 190 to 1100 nm. It contains two groups of 200 m diameter fibers (~245 micron diameter with cladding), with seven fibers per group and ~1 mm spacing between groups. Four Leg Fiber Bundle Model QFB 4553: a 1 m long UV_VIS fiberoptic bundle for 190 to 1100 nm. It contains four groups of 200 m diameter fibers (~245 micron diameter with cladding), with three fibers per group and ~1 mm spacing between groups. ~4.4 mm ~5.9 mm Figure F.1. 10 mm diameter ferrule on spectrograph end for (a) two leg fiber bundle and (b) fourleg fiber bundle. (a) (b)

PAGE 193

173 Figure F.2. The three fiber configuration on the source end of the four leg fiber bundle. An approximate 528 m diameter fiber bundle zone (including cladding), an approximate 483 m diameter light collection zone (not including cladding at outer edge). ~528 m dia. 1/8 inch dia. connector fibers, ~245 m dia. including cladding

PAGE 194

174 A ppendix G : O ES and ICCD Procedure Imaging Data Collection Data Analysis Following is the procedure to extract the desired R egion of I nterest ( ROI ) (ex.: Fiber Bundle Leg 3) from the collected WinSpec Glue Files and then convert the resulting ROI file s to Text ASCII files for further analysis in a spreadsheet program such as Microsoft Excel Create the following folders: Step Files Original Glue Files Excel Files ROI Leg 3 with 2 sub folders of Text Files and WinSpec Files Sort the Step Files and the Original Glue Files into their respective folders. Start the WinSpec/32 program. When using the program without the spectrometer and camera, a warning window will appear Camera not found click Cancel and a warning Error Communicating With Acton SP500 on Com1 click Cancel Under the Process menu select Binning and Skipping Under the Input tab: Select the Input Image (*.SPE file) from the Original Glue Files Frame is set at 1 to 1 X Range is set at 1 to 1294 Y Range is set at 271 to 310 from Leg 3 Fol lowing are the Regions of Interest ( ROI ) for each leg of the four leg fiber bundle: 126 to 165 for Leg 1 199 to 238 for Leg 2 271 to 310 for Leg 3 342 to 381 for Leg 4 Datatype is shown as FLOAT this parameter is set later Under the Parameters tab: Binning is set at X Dimension 1 and Y Dimension Average is not selected Under Skipping, X Dimension is set at 0 and Y Dimension is set at 0 Under the Output tab: for Output Image type the filename (usually the Input Image filename with 3 added to the end 3.SP E ) in folder ROI Leg 3 Display Result is selected

PAGE 195

175 Datatype is set at FLOAT Click Apply button, the resulting region of interest ( ROI ) is displayed. Repeat the procedure for the remaining files. The previous settings remain except for Y Range To convert t he ROI WinSpec Files to Text files: In the WinSpec/32 program under the Tools menu select Convert to ASCII In the .SPE To ASCII Conversion window: Click the Choose Files button, select all the ( 3.SPE ) files from the WinSpec Files folder, all the filename s selected will be displayed in the File Name window Click the Choose Output Directory button and select the Text Files folder under the ROI Leg 3 folder Frame No. : set at 1 to 512 File Extension: set to txt Delimiter set to Tab X Axis set to nm Output Ode r set to Pixel, Intensity New Line Characters select both Carriage Return and Line Feed Output File Options select One File for all Frames and Single Column Pixel Format select Preserve CCD X/Y/Frame Dimension Value Click Convert to ASCII button. While the conversion is being performed, the Convert to ASCII button is dimmed, and depending on the number file files being converted and the destination, each file takes approximately 1 second to convert. Click the Done button when finished. Exit the WinSpec pr ogram.

PAGE 196

1 7 6 A p p e n d i x H : H y d r o d y n a m i c P l a s m a M o d e l M a p l e P r o g r a m T h e p l a s m a m o d e l w a s c o m p u t e d u s i n g M a p l e 9 5 0 D i s t a n c e s a r e e n t e r e d a s c e n t i m e t e r s ( c m ) a n d t i m e a s m i c r o s e c o n d s ( s ) T h e p r o g r a m P l a s m a M o d e l m w i s a s f o l l o w s w i t h M a p l e c o d e ( r e d p r i n t ) a n d s a m p l e i n p u t d a t a ( b l u e p r i n t ) > r e s t a r t ; > > # I n p u t P a r a m e t e r s > S h : = 1 0 ; # l a s e r s p o t h o r i z o n t a l d i m e n s i o n f r o m c e n t e r ( m m ) S h : = 1 0 > S v : = 1 5 ; # l a s e r s p o t v e r t i c a l d i m e n s i o n f r o m c e n t e r ( m m ) S v : = 1 5 > P t : = 2 2 5 ; # p l a s m a t h i c k n e s s ( m i c r o m e t e r ) P t : = 2 2 5 > T e m p : = 8 0 0 0 ; # p l a s m a t e m p e r a t u r e ( K ) T e m p : = 8 0 0 0 > a w : = 8 1 3 3 ; # a t o m i c w e i g h t a w : = 8 1 3 3 > v x : = 0 6 3 3 5 ; # i n i t i a l x v e l o c i t y ( m / s ) v x : = 0 6 3 3 5 > v y : = 0 1 9 0 1 ; # i n i t i a l y v e l o c i t y ( m / s ) v y : = 0 1 9 0 1 > v z : = 0 1 2 6 7 ; # i n i t i a l z v e l o c i t y ( m / s ) v z : = 0 1 2 6 7 > t i d : = 0 ; # i n i t i a l d e l a y t i m e ( n s ) t i d : = 0 > t s t a r t : = 0 ; # s t a r t t i m e ( n s ) t s t a r t : = 0 > t e n d : = 2 0 0 0 ; # e n d t i m e ( n s )

PAGE 197

1 7 7 t e n d : = 2 0 0 0 > t g a t e : = 2 0 ; # g a t e t i m e ( n s ) t g a t e : = 2 0 > g a m a : = 1 6 ; # e x p o n e n t f a c t o r g a m a : = 1 6 > x d : = 6 ; # o n a x i s d i s t a n c e f o r T O F ( m m ) x d : = 6 > D i g i t s : = 5 : # n u m b e r o f d i g i t s u s e d f o r c a l c u l a t i o n s > > # C o n v e r s i o n s C o n s t a n t s a n d C a l c u l a t i o n s > a : = P t / ( 1 1 0 ^ 6 ) : # p l a s m a t h i c k n e s s u n i t c o n v e r s i o n ( m ) > b : = S h / ( 1 1 0 ^ ( 3 ) ) : # l a s e r s p o t h o r i z o n t a l d i m e n s i o n u n i t c o n v e r s i o n ( m ) > c : = S v / ( 1 1 0 ^ ( 3 ) ) : # l a s e r s p o t v e r t i c a l d i m e n s i o n u n i t c o n v e r s i o n ( m ) > t i : = t i d / ( 1 1 0 ^ 9 ) : # i n i t i a l d e l a y t i m e u n i t c o n v e r s i o n ( s e c ) > t s : = t s t a r t / ( 1 1 0 ^ 9 ) : # s t a r t t i m e u n i t c o n v e r s i o n ( s e c ) > t f : = t e n d / ( 1 1 0 ^ 9 ) : # e n d t i m e u n i t c o n v e r s i o n ( s e c ) > D T : = t g a t e / ( 1 1 0 ^ 9 ) : # g a t e t i m e s t e p u n i t c o n v e r s i o n ( s e c ) > d : = x d / ( 1 1 0 ^ 3 ) : # o n a x i s d i s t a n c e u n i t c o n v e r s i o n ( m ) > k : = 1 3 8 1 0 1 0 ^ ( 2 3 ) : # B o l t z m a n n c o n s t a n t ( J / K ) > e V : = 1 6 0 2 1 0 1 0 ^ ( 1 9 ) : # J o u l e s t o e V c o n v e r s i o n ( e V ) > u : = 1 6 6 1 1 0 ^ ( 2 7 ) : # u n i f i e d a t o m i c m a s s u n i t ( k g ) > C o n s t : = ( 6 k T e m p ) / ( a w u ) ; # c o n s t a n t C o n s t : = 4 9 0 6 9 1 06 > w i t h ( D E t o o l s ) : > e q n s : = d i f f ( x ( t ) t $ 2 ) x ( t ) = C o n s t ( ( a b c ) / ( x ( t ) y ( t ) z ( t ) ) ) ^ ( g a m a 1 ) d i f f ( y ( t ) t $ 2 ) y ( t ) = C o n s t ( ( a b c ) / ( x ( t ) y ( t ) z ( t ) ) ) ^ ( g a m a 1 ) d i f f ( z ( t ) t $ 2 ) z ( t ) = C o n s t ( ( a b c ) / ( x ( t ) y ( t ) z ( t ) ) ) ^ ( g a m a 1 ) ; # p l a s m a e x p a n s i o n e q u a t i o n s e q n s : = d 2 d t 2 x t ( ) x t ( ) = 2 5 5 7 2 1 x t ( ) y t ( ) z t ( ) 0 6 d 2 d t 2 y t ( ) y t ( ) = 2 5 5 7 2 1 x t ( ) y t ( ) z t ( ) 0 6 d 2 d t 2 z t ( ) z t ( ) = 2 5 5 7 2 1 x t ( ) y t ( ) z t ( ) 0 6

PAGE 198

1 7 8 > i n i t v a l s : = x ( t i ) = a y ( t i ) = b z ( t i ) = c D ( x ) ( t i ) = v x D ( y ) ( t i ) = v y D ( z ) ( t i ) = v z : # i n i t i a l c o n d i t i o n s f o r p o s i t i o n a n d v e l o c i t y > > L : = d s o l v e ( { e q n s i n i t v a l s } { x ( t ) y ( t ) z ( t ) } t y p e = n u m e r i c o u t p u t = l i s t p r o c e d u r e ) ; # s o l v e o r d i n a r y d i f f e r e n t i a l e q u a t i o n w i t h i n i t i a l c o n d i t i o n s L : = t = p r o c t ( ) . e n d p r o c ; x t ( ) = p r o c t ( ) . e n d p r o c ; d d t x t ( ) = p r o c t ( ) . e n d p r o c ; y t ( ) = p r o c t ( ) . e n d p r o c ; d d t y t ( ) = p r o c t ( ) . e n d p r o c ; z t ( ) = p r o c t ( ) . e n d p r o c ; d d t z t ( ) = p r o c t ( ) . e n d p r o c ; n > T : = r h s ( L [ 1 ] ) : # t i m e > X : = r h s ( L [ 2 ] ) : # x p o s i t i o n > X 1 : = r h s ( L [ 3 ] ) : # x v e l o c i t y > Y : = r h s ( L [ 4 ] ) : # y p o s i t i o n > Y 1 : = r h s ( L [ 5 ] ) : # y v e l o c i t y > Z : = r h s ( L [ 6 ] ) : # z p o s i t i o n > Z 1 : = r h s ( L [ 7 ] ) : # z v e l o c i t y > F L X : = t > ( ( X ( t ) ^ 2 Y ( t ) Z ( t ) ) ^ ( 1 ) ) d X 1 ( t ) e x p ( ( d ^ 2 / ( 2 ( X ( t ) ^ 2 ) ) ) ) ; # f l u x F L X : = t rr d X 1 t ( ) e d 2 2 X t ( ) 2 X t ( ) 2 Y t ( ) Z t ( ) > T O F : = t > F L X ( t ) D T ; # t i m e o f f l i g h t T O F : = t rr F L X t ( ) D T > f o r t f r o m t s b y D T t o t f d o T O F ( t ) e n d d o ; 1 6 8 2 2 1 0 1 5 4 3 7 1 1 6 7 7 1 0 2 5 9 3 2 0 7 8 0 1 0 5 9 6 1 0 9 6 3 1 0 2 3 4 1 9 7 5 9 1 0 1 1 9 4 5 1 1 8 1 0 7 0

PAGE 199

179 A ppendix I : PLD Procedure Substrate P reparation o Standard clean Cut substrate, typical substrate size is inch square Sonic bath in acetone 15 minutes Sonic bath in methanol 15 minutes De ionized water rinse Blow dry with nitrogen gas o Mount substrate Adhere with conductive silver paint Cover edge strip with stainless steel step mask for thickness measurement Dry silver paint by pre heating heater block and substrate to approximately 100 C for 20 minutes, then allow to cool to touch o Mount substrate hol der/ heater block Secure the substrate holder at the predetermined target substrate distance Center the substrate on axis with the target laser spot center Rotate the substrate shield to block the plume from the substrate o Check bulb heater and thermocouple connections Chamber P reparation o Clean laser windows Polish laser windows with optical polishing powder Rinse with water Blow dry with nitrogen gas Wipe Viton o ring and o ring groove with methanol Reseat window Wipe chamber inside walls with acetone fol lowed with methanol Wipe chamber lid o ring with methanol and reseat o ring Laser P reparation o Turn on the laser(s) Adjust laser energy to achieve the predetermined laser fluence Adjust the laser focusing lens for proper laser spot size and position Evacu ate Chamber o Close chamber lid o Close pressure relief valve o Turn on roughing pump o Open gate valve o Open cryo pump gate valve, if the pump is to be employed

PAGE 200

180 o Pump chamber until pressure is less than 200 milli torr, typically requires 3 minutes or less o Turn on t urbo molecular pump o Turn on substrate heater to assist in degassing and baking out the chamber o Evacuate chamber to less than 105 torr o If needed, close cryo pump gate valve, turn on cryo pump o Evacuate chamber to 1106 torr or less o If needed, close the roughing/turbo pump gate valve, and open the cryo pump gate valve, the pressure should drop to low 107 torr Deposition o When the chamber pressure is in the 106 torr range, raise the substrate heater temperature to the desired temperature o Allow the tempe rature to stabilize for at least 20 minutes o Start the target rotation o Turn on laser(s) o Laser condition the target surface o Rotate the substrate shield from in front of the substrate o Reset the laser pulse counter o Start the laser ablation for the predetermined number of pulses o Stop laser pulses Substrate C ool D own o Five minutes after the deposition was completed, reduce the heater voltage by 1 volt AC every 2. 5 minutes until 0 volts is reached Sample cool down rate is shown in Figure M.1. o Allow the heater blo ck to continue to cool to approximately room temperature

PAGE 201

181 y = -2.6355x + 346.94 R2 = 0.9995 20 45 70 95 120 145 170 195 220 245 270 295 320 0 15 30 45 60 75 90 105 120 135 150 165 180 Time (min.) Temperature (C) Figure I .1. Substrate heater cool down time from 300 C to room temperature with 1 V AC reduction every 2.5 minutes until 0 V AC is reached.

PAGE 202

182 A ppendix J: Electron Binding Energies The energies are given relative to the vacuum level for rare gases and H, N, O, F, and Cl diatomic molecules; relative to the Fermi level for metals; and relative to the top of the vale nce band for semiconductors. Values are taken from [103105]. Table J .1. The electron binding energy for oxygen (O), silicon (SI), gallium (Ga), germanium (Ge), strontium (Sr), and barium (Ba). 8 14 31 32 38 56 O Si Ga Ge Sr Ba K 1s 0.543 1.839 10.367 11.103 16.105 37.441 L-I 2s 0.042 0.150 1.299 1.415 2.216 5.989 L-II 2p1/2 0.100 1.143 1.248 2.007 5.624 L-III 2p3/2 0.099 1.116 1.217 1.94 5.247 M-I 3s 0.160 0.180 0.3587 1.293 M-II 3p1/2 0.104 0.125 0.2803 1.137 M-III 3p3/2 0.100 0.121 0.27 1.063 M-IV 3d3/2 0.019 0.030 0.136 0.796 M-V 3d5/2 0.019 0.029 0.1342 0.781 N-I 4s 0.0389 0.254 N-II 4p1/2 0.0216 0.192 N-III 4p3/2 0.0201 0.179 N-IV 4d3/2 0.093 N-V 4d5/2 0.090 N-VI 4f5/2 N-VII 4f7/2 O-I 5s 0.030 O-II 5p1/2 0.017 O-III 5p3/2 0.015 O-IV 5d3/2 O-V 5d5/2 P-I 6s P-II 6p1/2 P-III 6p3/2 Electron Binding Energy (keV) Atomic Number Element Orbital

PAGE 203

183 A ppendix K : CO2 Laser Output Couplers 30kV 36.5kV 1 80 20 553 1310 975 2 50 3002250 1460 3 90 10 400 1070 865 4 70 350 1480 1110 5 99 7 140 1000 825 6 93 10 524 1020 865 7 85 40 564 1150 915 8 99 3 220 992 810 9 77 6 236 975 792 10 88 10 570 1130 888 11 40 1160 875 12 89 20 546 1125 895 13 90 2 260 1060 855 14 95 400 1100 870 15 10 0 16 95 475 1080 860 17 90 510 1150 915 18 60 180 3980 1800 19 50 260 2430 1490 20 25 0 21 50 0 22 25 0 23 100 0 24 65 254 1980 1340 25 45 151 3070 1870 26 99 130 990 820 27 90 390 1110 890 28 75 525 1380 1050 29 63 290 2390 1470 30 95 37 1860 1350 31 5 0 32 87 545 1200 950 Delay (ns) Output Coupler Reflection (%) Curvature (m) Energy @ 36.5KV Table K .1. The output energy and delay for the CO 2 laser germanium output couplers.

PAGE 204

184 A ppendix L : Publications and Presentations A ppendix L .1. Journal Publications [1] D. Mukherjee, T. Dhakal, R. Hyde P. Muhkerjee, H. Srikanth, and S. Witanachchi, Role of Epitaxy in Controlling the Magnetic and Magnetostrictive Properties of Cobalt Ferr ite PZT Bilayers, Journal of Physics D: Applied Physics Vol. 43, pg. 485001 (2010). [ 2] T. Dhakal, D. Mukherjee, R. Hyde P. Mukherjee, M.H. Phan, H. Srikanth, and S. Witanachchi. Magnetic Anisotropy and Field Switching in Cobalt Ferrite Thin Films Depo sited by Pulsed Laser Ablation, Journal of Applied Physics Vol. 107, pg., No. 5, 053914 16 (2010). Appendix L .2. Published Conference Proceeings [5] D. Mukherjee, R. Hyde T. Dhakal, H. Srikanth, P. Mukherjee and S. Witanachchi. Investigation of the Pb Depletion in Single and Dual Pulsed Laser Deposited Epitaxial PZT Thin Films and Their Structural Characterization, in Multiferroic and Ferroelectric Materials (A. Gruverman, C.J. Fennie, I. Kunishima, B. Noheda, T.W. Noh, eds.) 2009 Materials Researc h Symposium Proceedings, Vol. 1199E, pg. 1199F03 37, Warrendale, PA (2010). [4] T. Dhakal, D. Mukherjee, R. Hyde H. Srikanth, P. Mukherjee and S. Witanachchi. Enhancement in Ferroelectricity in V Doped ZnO Thin Film Grown Using Laser Ablation, in Multiferroic and Ferroelectric Materials (A. Gruverman, C.J. Fennie, I. Kunishima, B. Noheda, T.W. Noh, ed.) 2009 Materials Research Symposium Proceedings, Vol. 1199E, pg. 1199 F03 44, Warrendale, PA (2010). [3] R. Hyde M. Beekman, G.S. Nolas, P. Mukherjee, an d S. Witanachchi, Growth and Characterization of Germanium Based Type I Clathrate Thin Films Deposited by Pulsed Laser Ablation, in Advances in Electronic Ceramics, Ceramic Engineering and Science Proceedings (C. Randal, Hua Tay Lin, K. Koumoto, and P. C lem, eds.), 2007 Proceedings of the 31st International Conference on Advanced Ceramics and Composites, American Ceramics Society, Vol. 28, Issue 8, pg. 211, Wiley (2008). [2] S. Witanachchi, R. Hyde M. Beekman, D. Mukherjee, P. Mukherjee, and G. S. Nolas, Synthesis and Characterization of Bulk and Thin Film Clathrates for Solid State Power Conversion Applications, IEEE Proceedings of the 25th International Conference on Thermoelectrics Vienna, Austria, pg. 4447 (2006). [1] S. Witanachchi, R. Hyde H. S Nagaraja, M. Beekman, G. S. Nolas, and P. Mukherjee, Growth and Characterization of Germanium Based Type I Clathrate Thin Films Deposited by Pulsed Laser Ablation, in Materials and Technologies for Direct Thermal to Electric Energy Conversion (J. Yang, T.P. Hogan, R. Funahashi, G.S. Nolas, eds.) 2005 Materials Research Society Symposium Proceedings, Vol. 886, pg. 0886F10 03, Warrendale, PA (2006).

PAGE 205

185 Appendix L .3. Conference Presentations Presentations [11] D. Mukherjee T Dhakal, R. Hyde H Srikanth P Mukherjee, and S Witanachchi Investigation of the Pb D epletion in S ingle and D ual Pulsed Laser D eposited E pitaxial PZT T hin F ilms and T heir F erroelectric C haracterization, Materials Research Society Fall Meeting Boston, MA ( Dec. 2009) [10] D. Mu kherjee R. Hyde T Dhakal, H Srikanth, P Mukherjee, and S Witanachchi, Dual L aser D eposition of S toichiometric PZT/CoFe2O4 E pitaxial H eterostructures Materials Research Society Spring Meeting, San Francisco, CA, (April 2010). [9] R. Hyde P. Mukherj ee, M. Beekman, G. S. Nolas, and S. Witanachchi Growth and Characterization of Dual Laser Deposited Films of Ba8Ga16Ge30 for Thermoelectric Applications, 27th International Conference on Thermoelectrics Corvallis OR (2008). [8] R. Hyde P. Mukherjee, M Beekman, G. S. Nolas, and S. Witanachchi, Growth of Stoichiometric Ba8Ga16Ge30 Films by Dual Laser Ablation and Study of Growth Dynamics by Emission Spectroscopy, Materials Research Society Fall Meeting Boston MA (2007). [7] R. Hyde M. Beekman, G. S. Nolas, P. Mukherjee, and S. Witanachchi, Growth and Characterization of Germanium Based Type I Clathrate Thin Films Deposited by Pulsed Laser Ablation, Univ of South Florida CAS 6th Graduate Research Symposium Tampa FL (2007). [6] M. Beekman R. Hyde D. Mukherjee, S. Witanachchi, P. Mukherjee, and G. S. Nolas, Preparation and Physical Properties of Group IV Clathrates, 31st Int. Conf. on Advanced Ceramics and Composites, American Ceramic Society Daytona Beach FL (2007). [5] R. Hyde M. Beekman, G. S. Nolas, P. Mukherjee, and S. Witanachchi, Growth and Characterization of Germanium Based Type I Clathrate Thin Films Deposited by Pulsed Laser Ablation, 31st Int. Conf. on Advanced Ceramics and Composites, American Ceramic Society, Daytona Beach FL (20 07). [4] Sarath Witanachchi R. Hyde V. Vithianathan, M. Beekman, P. Mukherjee, and G. S. Nolas, Synthesis and Characterization of Bulk and Thin Film Type I and Type II Clathrate Materials for Thermoelectric and Optoelectric Applications , 25th Int. Conf on Thermoelectrics Vienna Austria (2006). [3] M. Beekman D. Wang, R. Hyde H. S. Nagaraja, P. Mukherjee, S. Witanachchi, and G. S. Nolas, Synthesis and Characterization of Bulk and Thin Film Silicon and Germanium Clathrate Materials, Materials Research Society Spring Meeting, San Francisco CA (2006). [2] Sarath Witanachchi P. Mukherjee, H. S. Nagaraja, R. Hyde M. Beekman, H. F. Rubin, and G. S. Nolas, Dual Laser Deposition of Type I Clathrate Films, Materials Research Society Fall Meeting Boston MA (2005). [1] R. Hyde P. Mukherjee, and S. Witanachchi Role of the Magnetic Field on LargeArea Carbon Film Growth on Silicon in a Hollow Anode Arc Plasma Process, Materials Research Society Spring Meeting, San Francisco CA (2002).

PAGE 206

186 Appendix L .4. Conf erence Presentaions Poster Presentations [8] D. Mukherjee R. Hyde T. Dhakal, S. Hariharan, P. Mukherjee, a nd S. Witanachchi. Investigation of the Pb Depletion in Single and Dual Pulsed Laser Deposited Epitaxial PZT Thin Films and Their Structural Cha racterization. Materials Research Society Fall Meeting, Boston MA (2009). [7] T. Dhakal D. Mukherjee, R. Hyde H. Srikanth, P. Mukherjee, and S. Witanachchi. Enhancement in Ferroelectricity in V Doped ZnO Thin Film Grown Using Laser Ablation, Materials Research Society Fall Meeting, Boston MA (2009). [6] J. Rejman T. Dhakal, R. Hyde H. Srikanth, P. Mukherjee, and S. Withanachchi. Pulsed Laser Deposition as a Novel Growth Technique for Thin Film LuFe2O4 and Related Multiferroic Nature, Materials Rese arch Society Fall Meeting, Boston MA (2009). [5] D. Mukherjee T. Dhakal, R. Hyde P. Mukherjee, S. Hariharen, and S. Witanachchi. Growth of Epitaxial CoFe2O4/PZT Heterostructures and FerroelectricFerromagnetic Characterization, Materials Research Society Fall Meeting, Boston MA (2008). [4] D. Mukherjee T. Dhakal, R. Hyde P. Mukherjee, S. Hariharen, and S. Witanachchi. Growth of Epitaxial CoFe2O4/PZT Heterostructures and Ferromagnetic Characterization University of South Florida Poster Symposium & Competition, Tampa FL (2008). [3] R. Hyde P. Mukherjee, M. Beekman, G. S. Nolas, and S. Witanachchi, Growth of Ba8Ga16Ge30 Films by Pulsed Laser Ablation and Study of Growth Dynamics by Optical Emission Spectroscopy, Univ of South Florida Poster Symposium & Competition Tampa FL (2008). [2] R. Hyde M. Beekman, G. S. Nolas, P. Mukherjee, and S. Witanachchi, Growth and Characterization of Germanium Based Type I Clathrate Thin Films Deposited by Pulsed Laser Ablation, Univ of South Florida CAS 5th Graduate Research Symposium Tampa FL (2006). [1] R. Hyde P. Mahawela, S. Witanachchi, and P. Mukherjee, A Laser Triggered, Pulsed Plasma Process for LargeArea Thin Film Growth, Joint 29th Applied Vacuum Science and Technology Symposium and 19th Meeting of t he Florida Society for Microscopy Surface Analysis

PAGE 207

ABOUT THE AUTHOR Robert H. Hyde received his Bachelors of Science degree in Physics from Lock Haven University of Pennsylvania in 1991. Prior to attending the University of South Florida, he was employed in Research & Development with Wardley Corp./ Hartz Maountain, leading to co inventor of three U.S. Patents. He completed the Dual Masters of Science program in Engineering Science and Physics from the University of South Florida in 2006 and entered the Applied Physics Ph.D. program in 2007. He has been awarded the University of South Florida Graduate Research Symposium Award, Frank E. Duckwall Graduate Fellowship, and Fred L. & Helen M. Tharp Endowment Scholarship during the course of his graduate work. The research conducted during his graduate tenu re at USF resulted in presentations at conferences and symposiums, multiple poster and oral presentations. This work also lead to five published conferences proceedings and two peer reviewed journal publication s


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 s2011 flu ob 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004896
035
(OCoLC)
040
FHM
c FHM
049
FHMM
090
XX9999 (Online)
1 100
Hyde, Robert Harry.
0 245
Growth and characterization of thermoelectric ba8ga16ge30 type-i clathrate thin-films deposited by pulsed dual-laser ablation
h [electronic resource] /
by Robert Harry Hyde.
260
[Tampa, Fla] :
b University of South Florida,
2011.
500
Title from PDF of title page.
Document formatted into pages; contains 207 pages.
Includes vita.
502
Disseration
(Ph.D.)--University of South Florida, 2011.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
520
ABSTRACT: The on-going interest in thermoelectric (TE) materials, in the form of bulk and films, motivates investigation of materials that exhibit low thermal conductivity and good electrical conductivity. Such materials are phonon-glass electron-crystals (PGEC), and the multi-component type-I clathrate Ba8Ga16Ge30 is in this category. This work reports the first investigation of Ba8Ga16Ge30 films grown by pulsed laser deposition (PLD). This dissertation details the in-situ growth of polycrystalline type-I clathrate Ba8Ga16Ge30 thin-films by pulsed laser ablation. Films deposited using conventional laser ablation produced films that contained a high density of particulates and exhibited weak crystallinity. In order to produce high quality, polycrystalline, particulate-free films, a dual-laser ablation process was used that combines the pulses of (UV) KrF excimer and (IR) CO2 lasers that are temporally synchronized and spatially overlapped on the target surface. The effect of the laser energy on stoichiometric removal of material and morphology of the target has been investigated. In addition, in-situ time-gated emission spectroscopy and imaging techniques were used to monitor expansion of components in the ablated plumes. Through these investigations, the growth parameters were optimized not only to significantly reduce the particulate density but also to produce large area stoichiometric films. Structure and electrical transport properties of the resultant films were also evaluated. This work provides new insight toward the in-situ growth of complex multi-component structures in thin-film form for potential TE applications.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
590
Advisor:
Mukherjee Witanachchi, Pritish Sarath .
Advisor:
Mukherjee Witanachchi, Pritish Sarath .
653
Hydrodynamic Modeling
Iccd Imaging
In-situ Growth
Pld
Polycrystalline
Power Factor
690
Dissertations, Academic
z USF
x Physics Materials Science
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.4896