<%BANNER%>

Matrix-Assisted Resonant Infrared Pulsed Laser Ablation of Electroluminescent Dendrimers

Permanent Link: http://ufdc.ufl.edu/UFE0024390/00001

Material Information

Title: Matrix-Assisted Resonant Infrared Pulsed Laser Ablation of Electroluminescent Dendrimers
Physical Description: 1 online resource (236 p.)
Language: english
Creator: Torres, Ricardo
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: ablation, deposition, explosion, fel, frozen, infrared, laser, matrix, nucleation, phase, pulsed, resonant, solvent, vapor
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The deposition techniques for polymer thin films in organic light emitting diodes are limited to wet methods since molecular pyrolysis prevents the use of dry vacuum thermal evaporation methods. Wet methods have critical limitations such as poor thickness control, drying patterns and re-dissolution of previous layers. In this work, a novel approach, Matrix-Assisted Resonant Infrared Pulsed Laser Ablation (RIM-PLA) has been studied as an alternative deposition method for electroluminescent polymer films. RIM-PLA was successfully used for the deposition of two model dendrimers: fluorescent and phosphorescent Ir-cored. A free-electron laser was tuned to resonance frequencies for the vibrational modes of two solid matrix solvents: chloroform (C-H stretch; C-H bending) and toluene (C-H stretch; C=C stretch). The temperature-dependent absorption coefficients for each resonance mode were measured. Targets made from flash-frozen, low-concentration solutions of the dendrimers were irradiated at each frequency while varying fluence and exposure times. The molecular structure integrity of the targets was characterized. The deposited films were characterized to assess structure fidelity, roughness and topography, and luminance. All RIM-PLA deposited films were compared with spin-coated films. The ablation characteristics for each mode were found to be dependent on the solvent and not the dendrimer. Calculations from a temperature-rise model show that FEL pulsed- irradiation results in heating rates on the order of 10^8 ? 10^9 K/s, resulting in metastable condensed targets. Thermodynamic and kinetic relations were used to calculate the relevance of three ablation mechanisms: normal vaporization, normal boiling and phase explosion. The latter mechanism has a critical threshold ( > 0.8 Tc) for each solvent, and proceeds through spinodal decay followed by rapid homogeneous nucleation of vapor bubbles within the focal volume. For both chloroform modes, the primary ablation mechanism was phase explosion from the solid state, occurring fast enough (~100 ns) as to prevent significant dendrimer degradation. For both toluene modes, the primary mechanism was normal vaporization, which resulted in inefficient dendrimer deposition with high degradation. The results from RIM-PLA depositions indicate that films with minimal molecular degradation can be deposited. Further studies are necessary to understand ways to reduce the high film roughness from phase explosion in the target.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Ricardo Torres.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Holloway, Paul H.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024390:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024390/00001

Material Information

Title: Matrix-Assisted Resonant Infrared Pulsed Laser Ablation of Electroluminescent Dendrimers
Physical Description: 1 online resource (236 p.)
Language: english
Creator: Torres, Ricardo
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: ablation, deposition, explosion, fel, frozen, infrared, laser, matrix, nucleation, phase, pulsed, resonant, solvent, vapor
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The deposition techniques for polymer thin films in organic light emitting diodes are limited to wet methods since molecular pyrolysis prevents the use of dry vacuum thermal evaporation methods. Wet methods have critical limitations such as poor thickness control, drying patterns and re-dissolution of previous layers. In this work, a novel approach, Matrix-Assisted Resonant Infrared Pulsed Laser Ablation (RIM-PLA) has been studied as an alternative deposition method for electroluminescent polymer films. RIM-PLA was successfully used for the deposition of two model dendrimers: fluorescent and phosphorescent Ir-cored. A free-electron laser was tuned to resonance frequencies for the vibrational modes of two solid matrix solvents: chloroform (C-H stretch; C-H bending) and toluene (C-H stretch; C=C stretch). The temperature-dependent absorption coefficients for each resonance mode were measured. Targets made from flash-frozen, low-concentration solutions of the dendrimers were irradiated at each frequency while varying fluence and exposure times. The molecular structure integrity of the targets was characterized. The deposited films were characterized to assess structure fidelity, roughness and topography, and luminance. All RIM-PLA deposited films were compared with spin-coated films. The ablation characteristics for each mode were found to be dependent on the solvent and not the dendrimer. Calculations from a temperature-rise model show that FEL pulsed- irradiation results in heating rates on the order of 10^8 ? 10^9 K/s, resulting in metastable condensed targets. Thermodynamic and kinetic relations were used to calculate the relevance of three ablation mechanisms: normal vaporization, normal boiling and phase explosion. The latter mechanism has a critical threshold ( > 0.8 Tc) for each solvent, and proceeds through spinodal decay followed by rapid homogeneous nucleation of vapor bubbles within the focal volume. For both chloroform modes, the primary ablation mechanism was phase explosion from the solid state, occurring fast enough (~100 ns) as to prevent significant dendrimer degradation. For both toluene modes, the primary mechanism was normal vaporization, which resulted in inefficient dendrimer deposition with high degradation. The results from RIM-PLA depositions indicate that films with minimal molecular degradation can be deposited. Further studies are necessary to understand ways to reduce the high film roughness from phase explosion in the target.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Ricardo Torres.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Holloway, Paul H.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024390:00001


This item has the following downloads:


Full Text

PAGE 1

1 MATRIX -ASSISTED RESONANT INFRARED PUL SED LASER ABLATION OF ELECTROLUMINESCENT DENDRIMERS By RICARDO DANIEL TORRES -PAG N A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT O F THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Ricardo Daniel Torres Pag n

PAGE 3

3 To the many people who have supported my education, in particular, to my parents and my adviso r.

PAGE 4

4 ACKNOWLEDGMENTS The work presented in t his dissertation would not have been possible without the help and collaboration of so many people. I show my deepest gratitude to those mentioned and everyone e lse who contributed in any way to suc cessfully completing this work First and foremost, I thank Dr. Paul H. Holloway, my advisor and mentor God knows that it i s hard to say that this document would exist, if it had not been for his insurmountable patience and unconditional support. His pr ofessional character is one of a kind and admirable. He allowed me to explore my research interests and learn from continued and multiple mistakes without the fear of doing the wrong thing. His complete confidence in my abilities, even when I doubted t hem myself, made this work possible. It has been a pleasure and a privilege working for him. I would also like to thank my supervisory committee: Dr. David P. Norton, Dr. Brent P. Gila Dr. Christopher Batich and Dr. John R. Reynolds for their comments a nd insights into my work. I express deep gratitude to Dr. Paul L. Burn and his synthetic chemistry group at the University of Oxford (now at the University of Queensland) for supplying the dendrimer materials. None of the data presented in this dissertati on would have been possible without the collaboration of Dr. Richard F. Haglund and Dr. Stephen L. Johnson from Vanderbilt University. Stephens work and help for all depositions was invaluable and essential to say the least I sincerely appreciate his willingness to assist during those long night shifts and all the helpful conversations we had I would li ke to thank the FEL staff and director, Dr. J ohn Kozub for their assistance and gettin g the laser back up running the multiple times it broke down. M y deepest gratitude goes to Dr. Jungseek Hwang at the Department of Physics, now an Assistant Professor at Pusan National University (Korea), for his assistance with the tedious liquid -cell FTIR transmission measurements.

PAGE 5

5 A very important person for our re search groups subsistence wa s Ludie A. Harmon. She always made sure we had what we needed to do our work, and best of all, she did it with a smile on her f ace. I would like to thank previous and current members of Dr. Holloways research group for all t heir assistance and useful discussions in particular Debasis, Martin, Narada, Evan Bryan, Lei, Phil Teng -Kuan and Sergey I would also like to thank Dr. Mark Davidson (for substituting in my committee) and Chuck Rowland at MICROFABRITECH, for their ass istance and support with building BOB and fixing other vacuum equipment. I also need to thank Monta R. Holzworth for assisting with tasks of multiple projects, some tedious and repetitive. I worked on multiple projects while at UF, the latest materializin g in this dissertation. For my second study, I would like to thank Dr. Markus Albrecht and Dr. Olga Osetska at the RWTH Aachen (Germany) for supplying the templated Alq3 derivatives. I am very grateful to Dr. Nisha Ananthakrishnan for teaching me how to make and test PLED devices, Dr. Katsu Ogawa for his useful insights, and Dr. John R. Reynolds and Dr. Kirk S. Schanze for allowing me to use their laboratories. I would also like to thank Dr. David Witker for his assistance with the polymer synthesis work of my initial project. I express sincere appreciation to the staff of MAIC and PERC for their assistance with characterization equipment. Also, special thanks go to Dr. Chris Dancel for her help with the MALDI TOF MS data as well as the Mass Spectrometry Laboratory at the Department of Chemistry for their analysis. It is very important to acknowledge the funding support I have received, because without it this research would not have been possible. I give my sincere thanks to the UF Foundation in partic ular to Dr. Cammy Abernathy for recommending me for th e Alumni F ellowship that

PAGE 6

6 supported me for my first four years. Additional support came from the Army Research Laboratory Grant #W911NF 04 200023. On a personal note, I want to thank my very good friends Robert and Stephanie. Stephanie abandoned us half -way our journey, but I will a lways remember all the stressful and fun times with classes and all our outings. We made it to several continents together and hope to make it to a few more in the future. It was a great pleasure. I would also like to thank Vasana for her support and insights for so many things personal and professional Finally, I express my gratitude to Guillermo for being a ver y important part of my life and adding extra challenges to it. I leave the most important people in my life for last: my family. The uncondi tional love and support of my parents Gladys and V ctor have given me strength throughout the times that I wanted to quit this process It gives me joy that at least one of my grandparents gets to hear about this dissertati on, although I m sure abuela T ina is equally proud of me. I thank my brothers, Leo and Vctor, especially the oldest, for making fun of my life and research situations making me laugh along the way; and to my beautiful nieces, for giving me immense happiness and making me look forwar d to future joys in life. And to the One that has guided me through all the good and the bad: Gracias Papa Dios!

PAGE 7

7 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 4 L IST OF TABLES .............................................................................................................................. 11 LIST OF FIGURES ............................................................................................................................ 15 ABSTRACT ........................................................................................................................................ 21 CHAPTER 1 INTROD UCTION ....................................................................................................................... 23 1.1 Motivation and Primary Goal ........................................................................................... 23 1.2 Objectives .......................................................................................................................... 24 1.3 Structure of the Dissertation ............................................................................................. 25 2 LITERATURE REVIEW ........................................................................................................... 26 2.1 Introduction ....................................................................................................................... 26 2.2 Electroactive Polymers ..................................................................................................... 26 2.2.1 Conducting (Conjugated) Polymers .................................................................... 26 2.2.2 Photoluminescence and Electr oluminescence in Conjugated Molecules ......... 28 2.3 Technology of Interest: Organic and Polymer Light Emitting Diodes .......................... 29 2.4 Pulsed L aser Deposition of Organic Materials ................................................................ 32 2.4.1 Ultraviolet Pulsed Laser Deposition (UV -PLD) ................................................. 32 2.4.2 Resonant Infrared Pulsed L aser Deposition (RIR PLD) .................................... 35 2.5 Electroactive Dendrimers ................................................................................................. 37 2.5.1 Motivation for Materials Selection ...................................................................... 37 2.5.2 Studied Dendrimer Families ................................................................................ 38 2.6 Conclusion ......................................................................................................................... 39 3 EXPERIMENTAL METHODS ................................................................................................. 46 3.1 Materials ............................................................................................................................ 46 3.1.1 Dendrimers ............................................................................................................ 46 3.1.2 Solvents ................................................................................................................. 46 3.2 Measurements of Optical Constants of Solvents ............................................................. 46 3.2.1 Transmission Liquid Cell ..................................................................................... 47 3.2.2 Fourier Transform Infrared (FTIR) Spectrometer .............................................. 47 3.3 Control Films ..................................................................................................................... 47 3.4 Laser Processing ................................................................................................................ 48 3.4.1 Laser Source .......................................................................................................... 48 3.4.1.1 Free -electron laser (FEL) ...................................................................... 48 3.4.1.2 W. M. Keck FEL at Vanderbilt University .......................................... 49 3.4.2 Preparation for Ablation ....................................................................................... 51

PAGE 8

8 3.4.2.1 FEL beam control .................................................................................. 51 3.4.2.2 Substrate preparation ............................................................................. 52 3.4.2.3 Target preparation .................................................................................. 53 3.4.3 Thin -Film Depos ition ........................................................................................... 53 3.4.4 Constant and Variable Deposition Parameters ................................................... 54 3.5 Characterization ................................................................................................................. 54 3.5.1 Target Characterization ........................................................................................ 54 3.5.2 Thin -Film Characterization .................................................................................. 55 3.5.3 Techniques and Apparatus ................................................................................... 55 3.5.3.1 Nuclear Magnetic Resonance (NMR) spectroscopy ........................... 55 3.5.3.2 Fourier Transform Infrared (FTIR) spectroscopy ................................ 55 3.5.3.3 Matrix Assisted Laser Desorption/Ionization Time -of Flight (MALDI TOF) mass spectrometry ....................................................... 56 3.5.3.4 Optical microscopy ................................................................................ 56 3.5.3.5 Fluorescence microscopy ...................................................................... 57 3.5.3.6 Atomic Force Microscopy (AFM) ........................................................ 57 3.5.3.7 Stylus profilometry ................................................................................ 58 3.5.3.8 Thin -film photoluminescence quantum efficiency .............................. 59 4 OPTICAL CONSTANTS OF FROZE N SOLVENT MATRICES ......................................... 69 4.1 Introduction ....................................................................................................................... 69 4.2 Light Absorption by Dendrimer Solutions ...................................................................... 69 4.3 Absorption Coefficients, Optical Penetration Depths and Lifetimes of Resonance Modes ................................................................................................................................. 71 4.3.1 Literature Survey .................................................................................................. 71 4.3.2 Experimental Method ........................................................................................... 72 4.3.3 Absorption by Water / Atmospheric Moisture Condensation............................ 74 4.3.4 Spect ral Data Fittings ........................................................................................... 74 4.3.5 Lifetimes of Resonance Modes ............................................................................ 74 4.3.6 Results ................................................................................................................... 75 4.3.6.1 Liquid chloroform (298 K) .................................................................... 75 4.3.6.2 Solid chloroform .................................................................................... 76 4.3.6.3 Liquid toluene (298 K) .......................................................................... 77 4.3.6.4 Solid toluene .......................................................................................... 77 4.4 Light Coupling Efficiency ................................................................................................ 78 4.5 Affected Fo cal Volumes ................................................................................................... 79 4.6 Conclusions ....................................................................................................................... 79 5 TARGET CHARACTERIZATION .......................................................................................... 89 5.1 Introduction ....................................................................................................................... 89 5.2 Introduction to NMR Spectroscopy ................................................................................. 90 5.3 Peak Assignments for NMR Spectra of Dendrimers ...................................................... 91 5.3.1 Phosphorescent Dendrimer .................................................................................. 91 5.3.2 Fluorescent Dendrimer ......................................................................................... 92 5.4 Res onance Assignments for FTIR Spectra of Dendrimers ............................................. 92

PAGE 9

9 5.4.1 Phosphorescent Dendrimer .................................................................................. 92 5.4.2 Fluorescent Dendrimer ......................................................................................... 93 5.5 Temporal Stability of Dendrimers ................................................................................... 93 5.5.1 Phosphorescent Dendrimer .................................................................................. 93 5.5.2 Fluorescent Dendrimer ......................................................................................... 95 5.6 Target Characterization of Phosphorescent Dendrimer Fluence: 1 J/cm2 .................. 9 5 5.6.1 Chloro form ............................................................................................................ 96 5.6.1.1 CH alkyl stretch mode: 3.32 m ......................................................... 96 5.6.1.2 CH bending mode: 8.18 / 8.28 m ...................................................... 96 5.6.2 Toluene .................................................................................................................. 97 5.6.2.1 C................................................................... 97 5.6.2.2 ................................................... 97 5.6.3 Conclusions ........................................................................................................... 98 5.7 Target Characterization of Phosphorescent Dendrimer: Changes in Fluence ............... 99 5.7.1 Chloroform C H alkyl stretch mode: 3.32 m .................................................... 99 5.7.2 Toluene C H stretch mode: 3.31 m ................................................................. 100 5.7.3 Conclusions ......................................................................................................... 101 5.8 MALDI TOF MS Res ults for RIM PLA Targets ......................................................... 101 5.9 Conclusions ..................................................................................................................... 102 6 FILM CHARACTERIZATION ............................................................................................... 126 6.1 Introduction ..................................................................................................................... 126 6.2 Film Surface Topography and Thickness of Control Films ......................................... 126 6.3 Characterization of Den drimer Films Fluence: 1 J/cm2 ............................................ 127 6.3.1 Chloroform .......................................................................................................... 128 6.3.1.1 CH alkyl stretch mode: 3.32 m ....................................................... 128 6.3.1.2 CH bending mode: 8.18 / 8.28 m .................................................... 130 6.3.1.3 Comparison between modes ............................................................... 131 6.3.2 Toluene ................................................................................................................ 132 6.3.2.1 CH stretch mode: 3.31 m ................................................................. 132 6.3.2.2 ................................................. 134 6.3.2.3 Comparison between modes ............................................................... 136 6.3.3 Conclusions on Low Fluence (1 J/cm2) Films .................................................. 138 6.4 Characterization of Phosphorescent Dendrimer Films Increasing Fluence ............. 139 6.4.1 Chloroform C H alkyl stretch mode: 3.32 m .................................................. 139 6.4.2 Toluene C H stretch mode: 3.31 m ................................................................. 140 6.4.3 Conclusions on Film Deposition with Increasing Fluence .............................. 140 6.5 Mole cular Structure of the Deposited Films .................................................................. 141 6.5.1 MALDI TOF Mass Spectrometry ..................................................................... 141 6.5.2 FTIR Spectroscopy ............................................................................................. 142 6.6 Photoluminescence Quantum Yields of Films .............................................................. 143 6.7 Reduced Film Roughness by Annealing ........................................................................ 144 6.8 Conclusions ..................................................................................................................... 144 7 THERMALLY INDUCED PHASE TRANSITIONS AND ABLATION MECHANISMS FROM FEL IRRADIATION OF FROZEN SOLVENT MATRICES ..... 177

PAGE 10

10 7.1 Introduction ..................................................................................................................... 177 7.2 Mechanisms for High Power Laser Absorption in Materials ....................................... 177 7.3 Confinement Co nditions ................................................................................................. 178 7.3.1 Thermal Confinement ......................................................................................... 178 7.3.2 Stress Confinement ............................................................................................. 179 7.4 Assumptions .................................................................................................................... 180 7.4.1 Negligible Plume Shielding ............................................................................... 180 7.4.2 Steady State Ablation Model ............................................................................. 180 7.5 Temperature Rise in Solvent .......................................................................................... 182 7.6 Thermally Induced Phase Transitions in Solvent Matrices .......................................... 184 7.6.1 Melting ................................................................................................................ 184 7.6.2 Evaporation ......................................................................................................... 185 7.7 Phase Diagrams of Solvents Used as Matrices ............................................................. 186 7.8 Ablation Mechanisms from FEL Irradiation of Solvent Matrices ............................... 189 7.8.1 Surface Evaporation (Normal Vaporization) .................................................... 189 7.8.2 Spontaneous Nucleation ..................................................................................... 192 7.8.2.1 Volume evaporation (normal boiling) ................................................ 192 7 .8.2.2 Macroscopic sputtering via phase explosion (explosive boiling) ..... 195 7.9 Correlation to Ablation Results ...................................................................................... 203 7.9.1 Chlor oform Results ............................................................................................. 205 7.9.2 Toluene Results ................................................................................................... 208 7.10 Conclusion ....................................................................................................................... 211 8 CONCLUSIONS ....................................................................................................................... 218 REFERENCE LIST .......................................................................................................................... 224 BIOGRAPHICAL SKETCH ........................................................................................................... 236

PAGE 11

11 LIST OF TABLES Table page 2 1 Solution and thin -film PLQY of all fluorescent dendrimer generations. ........................... 45 2 2 Solution and thin -film PLQY of all phosphor escent dendrimer generations. .................... 45 3 1 Fundamental properties of phosphorescent dendrimer. ....................................................... 66 3 2 Fundamental properties of fluore scent dendrimer. .............................................................. 66 3 3 Fundamenta l properties of matrix solvents. ......................................................................... 67 3 4 Radii of spot sizes for the FEL beam tuned to the wave lengths of interest. ...................... 68 4 1 Selected resonances for each solvent used as a matrix. ....................................................... 86 4 2 Averaged infrared optical constant s fo r toluene from Bertie, et al. .................................... 86 4 3 Mid infrared absorption coefficients for toluene from Anderson et al. .............................. 86 4 4 Absorption coe fficients for modes of liquid chloroform and dendrimer matrices. ............ 86 4 5 Optical penetration depths for modes of liquid chloroform and dendrimer matrices. ....... 86 4 6 Absorption coefficients for modes of frozen chloroform and dendrimer matrices. ........... 86 4 7 Optical penetration depths for modes of frozen chloroform and dendrimer m atrices. ...... 87 4 8 Results from Lorentzian fits of chloroform spectra. ............................................................ 87 4 9 Absorption coefficients for modes of liquid toluene a nd dendrimer matrices. .................. 87 4 10 Optical penetration depths for modes of liquid toluene and dendrimer matrices. ............. 87 4 11 Absorption coefficients for modes of frozen toluene and dendrimer matrices. ................. 87 4 12 Optical penetration depths for modes of frozen toluene and dendrimer matrices. ........... 88 4 13 Results from Lorentzian fits of toluene spectra. ................................................................... 88 4 14 FWHM results from Gaussian fits of FEL wavelength distribution. .................................. 88 4 15 Laser -mode coupling efficiency for selected solvent resonance peaks. ............................. 88 4 16 Average FEL irradiated focal volumes for the selected resonance modes. ........................ 88

PAGE 12

12 4 17 Estimated volume percentages of the ablated matrices affected by FEL irradiation at resonant modes. ...................................................................................................................... 88 5 1 Proton assignments for the phosphorescent dendrimer and appropriate signal legend for NMR spectra. .................................................................................................................. 118 5 2 Proton assignments for the fluorescent dendrimer and appropriate signal legend for NMR spectra. ........................................................................................................................ 119 5 3 Characteristic mid IR resonances observed in the phosphorescent dendrimer. ............... 120 5 4 Characteristic mid IR resonances observ ed in the fluorescent dendrimer. ...................... 121 5 5 Relative proton counts from 1H NMR spectra of control phosphorescent dendrimer. .... 122 5 6 Relative proton counts from 1H NMR spectra of control fluorescent dendrimer. ........... 122 5 7 Relative proton counts from 1H NMR spectra presented in Figure 5 13: C H alkyl stretch mode of chloroform at the sa me fluence. ............................................................... 122 5 8 Relative proton counts from 1H NMR spectra presented in Figure 5 15: C H bending mode of chloroform at the same fluence. ........................................................................... 123 5 9 Relative proton counts from 1H NMR spectra presented in Figure 5 16: C H stretch mode of toluene at the same fluence. .................................................................................. 123 5 10 Relative proton counts from 1H NMR spectra p resented in Figure 5 17: C=C aromatic stretch mode of toluene at the same fluence. ...................................................... 124 5 11 Relative proton counts from NMR spectra presented in Figure 5 18: C H alkyl stretch mode of chloroform at increasing fluences. ........................................................... 124 5 12 Relative proton counts from NMR spectra presented in Figure 5 19: C H stretch mode of toluene at increasing fluences. .............................................................................. 125 5 13 Comparison of ratios of intensities of C H stretches of target shown in Figure 5 20 with the control spectra (Figure 5 7). .................................................................................. 125 5 14 Values and error for the monoisotopic mass for M+ from the MALDI TOF MS data for the phosphorescent dendrimer targets. .......................................................................... 125 6 1 Thickness of spin-coated (control) dendrimer films using AFM or stylus profilometry. ......................................................................................................................... 172 6 2 AFM Z ranges and RMS roughness of spin -coated (control) dendrimer films. .............. 172 6 3 AFM RMS roughness and stylus profilometry th ickness of phosphorescent dendrimer films: C H alkyl stretch mode of chloroform (1 J/cm2) .................................. 173

PAGE 13

13 6 4 AFM RMS roughness and stylus profilometry thickness of fluorescent dendrimer films: C H alkyl stretch mode of chloroform (1 J/cm2). .................................................... 173 6 5 AFM RMS roughness and stylus profilometry thickness of phosphorescent dendrimer films: C H bending mode of chloroform (1 J/cm2). ......................................... 173 6 6 AFM RMS roughness and stylus profilometry thickness of fluorescent dendrimer films: C H bending mode of chloroform (1 J/cm2). ........................................................... 173 6 7 AF M RMS roughness and stylus profilometry thickness of phosphorescent dendrimer films: C H stretch mode of toluene (1 J/cm2) ................................................. 173 6 8 AFM RMS roughness and stylus profilometry thickness of fluor escent dendrimer films: C H stretch mode of toluene (1 J/cm2). .................................................................... 174 6 9 AFM RMS roughness and stylus profilometry thickness of phosphorescent dendrimer films: C=C aromatic stretch mode of tolue ne (1 J/cm2) ................................. 174 6 10 AFM RMS roughness and stylus profilometry thickness of fluorescent dendrimer films: C=C aromatic stretch mode of toluene (1 J/cm2). ................................................... 174 6 11 Average deposition rates for all resonance modes at a fluence of 1 J/cm2. ...................... 174 6 12 AFM RMS roughness and stylus profilometry thickness of phosphorescent den drimer films: C H alkyl stretch mode of chloroform (20 min) .................................. 175 6 13 AFM RMS roughness and stylus profilometry thickness of phosphorescent dendrimer films: C H stretch mode of toluene (20 min ). .................................................. 175 6 14 Values and error for the monoisotopic mass for M+ from the MALDI TOF MS data for the phosphorescent dendrimer films. ............................................................................ 175 6 15 Values and error for the monoisotopic mass for M+ from the MALDI TOF MS data for the fluorescent dendrimer films. .................................................................................... 175 6 16 Comparison of ratios of intensities of C H stretches of tar get shown in Figure 6 37 with the control spectra. ....................................................................................................... 176 6 17 Phosphorescent dendrimer thin -film PLQYs for depositions at low fluence (1 J/cm2). .. 176 6 18 PL quantum yields of phosphorescent dendrimer films deposited at increasing fluence. .................................................................................................................................. 176 6 19 Measured RMS roughness of AFM images shown in Figure 6 40. .................................. 176 7 1 T) for selected solvent resonance modes. .......................... 215 7 2 Peak temperatures in the frozen and melted solvent matrice s for selected resonance modes in chloroform and toluene (1 J/cm2). ....................................................................... 216

PAGE 14

14 7 3 Required volumetric energy density for a melting transition in the frozen targets. ......... 216 7 4 Calculated recession at the target surface from FEL induced surface evaporation. ........ 216 7 5 Volume diffusion coefficients and diffusion lengths for FEL puls e duration. ................. 216 7 6 Critical temperature thresholds ( Tpe) and maximum rates ( Jcr max) for homogeneous nucleation in the superheated solvents. ............................................................................... 217 7 7 Time of stratification ( sd) of superheated liquid solvents. ................................................ 217 7 8 Time lag (n) for homogeneous nucleation of superheated matrix solvents. .................... 217 7 9 FE L effective volumetric energy densities for solid -state chloroform (1 J cm2). ............ 217 7 10 FEL effective volumetric energy densities for solid -state toluene (1 J cm2). .................. 217

PAGE 15

15 LIST OF FIGURES Figure page 2 1 Formation of a bond by weak overlapping of the pz orbitals of two successive carbon atoms and resulting electron delocalization. ............................................................ 41 2 2 Radiative and non -radiative decay processes in a typical pol yatomic m olecule. .............. 41 2 3 Electroluminescence in a mon olayer organic or polymer LED. ......................................... 42 2 4 Evolution of OLED device structures ................................................................................... 42 2 5 Schematic of basic components of a PLD system. .............................................................. 43 2 6 Energy diagram comparison between ultraviolet (electronic) and infrared (vibr ational) laser excitation ................................................................................................. 43 2 7 Dendrim er families studied in this work. ............................................................................. 44 2 8 UV-visible and photoluminescence spectra of fl uorescent dendrimer fam ily.. ................. 44 2 9 Solution UV -visible and film photoluminescence spectra of phosph orescent dendrimer family. ................................................................................................................... 45 3 1 Dendrimers used in this research: a) Phosphorescent fac tris(2 -phenylpyridine) iridium(III) -co red; b) Fluorescent tris(distyrylbenzenyl)amine cored dendrimers used in this research. ....................................................................................................................... 60 3 2 Molecular structures of matrix solvents. ............................................................................... 60 3 3 Demountable liquid cell used for transmission spectroscopy. ............................................ 60 3 4 Schematic of the principal components of t he free -electron laser. ..................................... 61 3 5 Pulse structure of th e Vanderbilt FEL. ................................................................................. 61 3 6 Calibration curves for radius of a circular spot si ze at different FEL wavelengths (CaF2 focusing lens). .............................................................................................................. 62 3 7 FEL laboratory: a) FEL beam is guided through (blue) pipes under low vacuum towards the exit window; b) monitor displaying pulse intensity/bandwi dth. ..................... 62 3 8 Setup for beam guiding to the vacuum deposition chamber. .............................................. 63 3 9 Frozen target of phosphorescent dendrimer (G1) in ch loroform. ....................................... 63 3 10 RIM -PLA thin -film deposition chamber. ............................................................................. 64

PAGE 16

16 3.11 Frozen dendrimer target during ablation: (a) under normal room l ight and (b) under a UV lamp. ................................................................................................................................. 64 3 12 Dendrimer covered substrate after RIM PLA deposition. ................................................... 65 3 13 Schemati c of an atomic force microscope. ........................................................................... 65 3 14 Schematic of integrating sphere setup used for thin -film quantum yield measurements. ........................................................................................................................ 65 4 1 Absorption, ref lection and transmission of light through a liquid solution. ....................... 81 4 2 Optical penetration depth relationshi p to absorption coefficient. ....................................... 81 4 3 Mid IR atmospheric absorption spectrum and wavelengths of interest. ............................ 81 4 4 Absorption coefficient spectra for liquid chloroform and dendrimer matrices. ................. 82 4 5 Optical penetration depths for liquid chloroform and dendrimer matrices. ....................... 82 4 6 Absorption coefficient spectra for frozen chloroform and dendrimer m atrices. ................ 82 4 7 Optical penetration depths for frozen chloroform and dendrimer matrices. ...................... 83 4 8 Crystal structure ( phase) of solid chloroform. .................................................................. 83 4 9 Chloroform spectra: a) Evolution of absorption coefficient spectra (Lorentzian fits) of chloroform as the temperature of the transmission cell rose. b) Comparison of integr ated areas under fits for both chloroform modes. ....................................................... 83 4 10 Absorption coefficient spectra for liquid toluene and dendrimer matrices. ....................... 84 4 11 Optical penetration depths for liquid toluene and dendrimer matrices. .............................. 84 4 12 Absorption coefficient spectra for frozen toluene and dendrimer matrices. ...................... 84 4 13 Optical penetration depths for frozen toluene and dendrimer matrices. ............................. 85 4 14 Lorentzian fits for frozen toluene absorption coefficient data. ........................................... 85 4 15 FEL wavelength distributions at the selected wavelengths. ................................................ 85 5 1 1H NMR chemical shifts of different organic moieties. .................................................... 103 5 2 Molecular structure of the phosphorescent dendrimer with proton assignments. ............ 103 5 3 Molecular structure of the fluorescent dend rimer with proton assignments. ................... 104

PAGE 17

17 5 4 Expansion of the downfield region of the fluorescent dendrimer control 1H NMR spectra. .................................................................................................................................. 104 5 5 FTIR spectra of phosphorescent dendrimer control. .......................................................... 105 5 6 FTIR spectra of fluorescent dendrimer control. ................................................................. 105 5 7 1H NMR sp ectra of phosphorescent dendrimer control: (a) as received; (b) 15 months after receipt. .......................................................................................................................... 106 5 8 MALDI TOF MS (large scale) of phosphorescent dendrimer control 15 months after receipt. ................................................................................................................................... 107 5 9 MALDI TOF MS (M+ area) of phosphorescent dendrimer 15 months after receipt. ...... 107 5 10 1H NMR spectra of fluorescent dendrimer co ntrol: (a) as received; (b) 15 months after receipt. .......................................................................................................................... 108 5 11 MALDI TOF MS (large scale) of fluorescent dendrimer control 15 months after receipt. ................................................................................................................................... 108 5 12 MALDI TOF MS (M+ scale) of fluorescent dendrimer control 15 months after receipt. ................................................................................................................................... 109 5 13 1H NMR spectra comparison for phosphorescent dendrimer targets ablated fo r different times at 1 J/cm2: C H alkyl stretch resonance mode of chloroform. ................. 110 5 14 FTIR spectra comparison for phosphorescent dendrimer targets ablated for different times at 1 J/cm2: C H alkyl stretch resonance mode of chloroform. ................................ 111 5 15 1H NMR spectra comparison for phosphorescent dendrimer targets ablated for different times at 1 J/cm2: C H bending resonance mode of chloroform. ........................ 112 5 16 1H NMR spectra comparison for phosphorescent dendrimer targets ablated for different times at 1 J/cm2: C H stretch resonance mode of toluene. ................................. 113 5 17 1H NMR spectra comparison for phosphorescent dendrimer targets ablated for different times at 1 J/cm2: C=C aromatic stretch resonance mode of toluene. ................ 114 5 18 1H NMR spectra comparison for phosphorescent dendrimer targets ablated at increasing fluences: C H alkyl stretch resonance mode of chloroform. ........................... 115 5 19 1H NMR spectra comparison for phos phorescent dendrimer targets ablated at increasing fluences: C H stretch resonance mode of toluene. ........................................... 116 5 20 FTIR spectra of phosphorescent dendrimer target ablated at 55 J/cm2: C H stretch re sonance mode of toluene. ................................................................................................. 117

PAGE 18

18 6 1 AFM images of razor -blade cut control dendrimer films and average thickness analysis. ................................................................................................................................. 145 6 2 A FM height images and optical micrographs (transmission-mode) of spin -coated control dendrimer films. ...................................................................................................... 146 6 3 Optical micrograph (transmission -mode) of a bare glass substrate. ................................. 146 6 4 Optical micrographs (reflection -mode) and AFM height images of phosphorescent dendrimer films: C H alkyl stretch resonance mode of chloroform (1 J/cm2). ................ 147 6 5 AFM phase images of phosphorescent dendrimer films: C H alkyl stretch resonance mode of chloroform (1 J/cm2). ............................................................................................ 147 6 6 Fluorescence and bright -field mode images of phosp horescent dendrimer films: C H alkyl stretch resonance mode of chloroform (1 J/cm2). ..................................................... 148 6 7 Optical micrographs (reflection -mode) and AFM height images of fluorescent dendrimer films: C H alk yl stretch resonance mode of chloroform (1 J/cm2). ................ 149 6 8 AFM phase images of fluorescent dendrimer films: C H alkyl stretch resonance mode of chloroform (1 J/cm2). ............................................................................................ 149 6 9 Fluorescence and bright -field mode images of fluorescent dendrimer films: C H alkyl stretch resonance mode of chloroform (1 J/cm2). ..................................................... 150 6 10 Compa rison of AFM RMS roughness and stylus profilometry thickness for both dendrimer films: C H alkyl stretch resonance mode of chloroform (1 J/cm2). ................ 150 6 11 Optical micrographs (reflection -mode) and AFM height images of phosphorescent dendrimer films: C H bending resonance mode of chloroform (1 J/cm2). ....................... 151 6 12 Fluorescence and bright -field mode images of phosphorescent dendrimer film s: C H bending resonance mode of chloroform (1 J/cm2). ............................................................ 152 6 1 3 Optical micrographs (reflection -mode) and AFM height images of fluorescent dendrimer films: C H bending resonance mode of chloroform (1 J/cm2). ....................... 153 6 14 Fluorescence and bright -field mode images of fluorescent dendrimer films: C H bending resonance mode of chloroform (1 J/cm2). ............................................................ 154 6 15 Comparison of AFM RMS roughness and stylus profilometry thickness for both dendrimer films: C H bending resonance mode of chloroform (1 J/cm2). ....................... 154 6 16 Compari son of AFM RMS roughness and stylus profilometry thickness for all dendrimer films at all chloroform resonance modes studied (1 J/cm2). ........................... 155 6 17 Optical micrographs (reflection -mode) and AFM h eight images of phosphorescent dendrimer films: C H stretch resonance mode of toluene (1 J/cm2). ................................ 156

PAGE 19

19 6 18 Fluorescence and bright -field mode images of phosphorescent dendrimer films: C H stret ch resonance mode of toluene (1 J/cm2). ..................................................................... 157 6 19 Optical micrographs (reflection -mode) and AFM height images of fluorescent dendrimer films: C H stretch resonance mode of toluene (1 J/cm2). ................................ 158 6 20 Fluorescence and bright -field mode images of fluorescent dendrimer films: C H stretch resonance mode of toluene (1 J/cm2). ..................................................................... 159 6 21 Comparison of AFM RMS roughness and stylus profilometry thickness for both dendrimer films: C H stretch resonance mode of toluene (1 J/cm2). ................................ 159 6 22 Optical micrographs (reflection -mode) and AFM height images of phosphorescent dendrimer films: C=C aromatic stretch resonance mode of toluene (1 J/cm2). ............... 160 6 23. Fluorescence and bright -field mode images of phosphorescent dendrimer films: C=C aromatic stretch resonance mode of toluene (1 J/cm2). ..................................................... 161 6 24 Optical micrographs (reflection -mode) and AFM height images of fluorescent dendrimer films: C=C aromatic st retch resonance mode of toluene (1 J/cm2). ............... 162 6 25 Fluorescence and bright -field mode images of fluorescent dendrimer films: C=C aromatic stretch resonance mode of toluene (1 J/cm2). ..................................................... 163 6 26 Comparison of AFM RMS roughness and stylus profilometry measurements for both dendrimer films: C=C aromatic stretch resonance mode of toluene (1 J/cm2). ............... 163 6 27 Comparison of AFM RMS roughness and stylus profilometry measurements for all dendrimer films at all toluene resonance modes studied (1 J/cm2). .................................. 164 6 2 8 Optic al micrographs (transmission-mode) and AFM height images of phosphorescent dendrimer films: C H alkyl stretch resonance mode of chloroform at increasing fluences. .............................................................................................................. 165 6 29 AFM phase images of phosphorescent dendrimer films: C H alkyl stretch resonance mode of chloroform at increasing fluences. ....................................................................... 165 6 30. Low magnification optical micrographs (reflection-mode) of phosphorescent dendrimer films: C H alkyl stretch resonance mode of chloroform at increasing fluences. ................................................................................................................................ 166 6 31 Optical micrographs (transmission -mode) and AFM height images of phosphorescent dendrimer films: C H stretch resonance mode of toluene at increasing fluences. .............................................................................................................. 167 6 32 Large area optical micrographs (reflection -mode) of phosphorescent dendrimer films: C H stretch resonance mode of toluene a t increasing fluences. ............................. 168

PAGE 20

20 6 33 MALDI TOF MS (large scale) of phosphorescent dendrimer: C H alkyl stretch resonance mode of chloroform (1 J/cm2, 20 min). ............................................................. 168 6 34 MALDI TOF MS (M+ area) of phosphorescent dendrimer: C H alkyl stretch resonance mode of chloroform (1 J/cm2, 20 min). ............................................................. 169 6 35 MALDI TOF MS (large scale) of fluorescent dendrimer: C -H alkyl stretch resonance mode of chloroform (1 J/cm2, 20 min). ............................................................................... 169 6 36 MALDI TOF MS (M+ area) of fluorescent dendrimer: C H alkyl stretch resonance mode of chlorofor m (1 J/cm2, 20 min). ............................................................................... 170 6 37 FTIR spectra of control film and phosphorescent dendrimer film ablated at 55 J/cm2: CH stretch resonance mode of toluene. ............................................................................. 170 6 38 PL spectra comparison of phosphorescent dendrimer films deposited at low fluence (1 J/cm2exc = 390 nm). .................................................................. 171 6 39 PL spectra comparison of phosphorescent dendrimer films deposited at increasing fluence for the C exc = 390 nm). ............................... 171 6 40 AFM height images and phase images of a phosphorescent dendrimer film annealed above its Tg: C H control spin -coated dendrimer films (1 J/cm2, 20 min). ...................... 172 7 1 Possible vibrational energy relaxation mechanisms in solvent molecules ...................... 212 7 2 Typical phase diagram for a pure substance. ...................................................................... 212 7 3 Liquid -gas phase diagram of pure chloroform. .................................................................. 213 7 4 Liquid -gas phase diagram of pure toluene. ........................................................................ 213 7 5 Temperature dependence of the rate of homogeneous nucleation for metastable liquid chloroform at two different pressures. ..................................................................... 214 7 6 Temperature dependence of the rate of homogeneous nucleation for metastable liquid toluene at two different pressures. ............................................................................ 214 7 7 Temperature -entropy phase diagram of a single component material with three different isochores.172 ........................................................................................................... 215 7 8 Ablation scenarios at different temperature ranges for RIM PLA targets. ...................... 215

PAGE 21

21 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MATRIX -ASSISTED RESONANT INFRARED PULSED LASER ABLATION OF ELECTROLUMINESCENT DENDRIMERS By Ricardo Daniel Torres -Pagn August 2009 Chair: Paul H. Holloway Major: Materials Science and Engineering The deposition techniques for polymer thin films in organic light emitting diodes are limited to wet methods since molecular pyrolysis prevents the use of dry vacuum thermal evaporation methods. Wet me thods have critical limitations such as poor thi ckness control, drying patterns and re dissolution of previous layers In this work, a novel approach Matrix Assisted Resonant Infrared Pulsed Laser Ablation (RIM -PLA) has been studied as a n alternative dep osition method for electroluminescent polymer films RIM PLA was successfully used for the deposition of two model dendrimers: fluorescent and phosphorescent Ir -cored. A free electr on laser was tuned to resonance frequencies for the vibr ational modes of two solid matrix solvent s: chloroform (C -H stretch; C H bending) and toluene (C H stretch; C=C stretch) The temperature-dependent absorption coefficients for each resonance mode were measured Targets made from flash -frozen, low -concentration solutions of the dendrimers were irradiated at each frequency while varying fluence and ex posure times. The molecular structure integrity of t he targets was characterized. The deposited film s were characterized to assess structure fidelity, roughness and topography and luminance All RIM -PLA deposited films were compared with spin -coated films T he ablation characteristics f or each mode were found to be dependent on the solvent and not the dendrim er. Calculations from a temperaturerise model

PAGE 22

22 show that FEL pulse d irradiation results in heating rates on the order of 108 109 K/s, resulting in metastable condensed targets Thermodynamic and kinetic relations were used to calculate the relevance of three ablation mechanisms: normal vaporization, normal boiling an d phase explosion. The latter mechanism has a critical threshold (> 0.8 Tc) for each solvent, and proceeds through spinodal decay followed by rapid homogeneous nucleation of vapor bubbles within the focal volume. For both chloroform modes, the primary abla tion mechanism was phase explosion from the solid state, occurring fast enough (~100 ns) as to prevent significant dendrimer degradation. For both toluene modes, the primary mechanism was normal vaporization, which r esult ed in i nefficien t dendrimer depos ition with high degradation. The results from RIM -PLA depositions indicate that films with minimal molecular degradation can be deposited. Further studies are necessary to understand ways to reduce the high film roughness from phase explosion in the target.

PAGE 23

23 CHAPTER 1 INTRODUCTION 1.1 Motivation and Primary Goal The discovery of efficient electroluminescence in organic and conjugated polymer thin films has resulted in a tremendous r esearch effort on the structure property-processing relationships of thes e materials.1 2 Evidence of the considerable progress made in the field is that flat panels based on organic and polymeric light emitting diodes (OLEDs and PLEDs, respectively) are rapidly emerging in commercial products replacing current liquid crystal display (LCD) technology. The low cost of materials and their simple device fabrication processes, including ink -jet printing, m ean that larger device s can be produced economically.3 The processing techniques for the deposition of polymer thin films for PLEDs has been limited to wet methods since molecular pyrolysis prevents the use of dry v acuum thermal evaporation methods.4 Wet methods have critical limitations such as poor thickness c ontrol, drying patterns, re -dissolution of previous layers and substrate limitations. These limitations have resulted in increased attention to Pulsed Laser Deposition (PLD) as an alternative method for the deposition of electrolumine scent and conducting polymers.511 The use of ultraviolet lasers (UVPLD) for polymer ablation (material removal) has result ed in significant degradation to the molecular structure of these organic macromolecules.58 Consequently, infrared (IR) lasers have been studied as alternative light sources, particularly the tunable mid -infrared free -electron laser (FEL).911 The intense FEL irradiation (~0.25 eV) gives rise to multi -photon excitation over the vibrational manifolds of g round electronic states; the lack of electronic excitation (~3 4 eV) as in the case of UV -PLD results in preservation of the complex organic structures.12 In order to further minimize direct FEL -polymer interaction, a matrix assisted method has been used where the organic sample is dissolved at a low -

PAGE 24

24 c oncentration in a volatile solvent. The solutions can be flash-frozen, resul ting in solid targets that are ablated with an FEL tuned to high absorption resonance mode s of the solvent. This processing technique is known as Matrix -Assisted Resonant Infrared Pulsed Laser Ablation (RIM -PLA). The ablation mechanisms of this technique have never been fully described, as it has been assumed that the polymer molecules attain sufficient kinetic energy through collective coll isions with evaporating solvent molecules eventually reaching the deposition substrate The primary goal of t his di ssertation is to provide descriptions of the ablation mechanisms involved in RIM -PLA with the use of two solvents and two model luminescent dendrimer molecules. A temperaturerise model, previously applied to solid polymer systems,1314 along with thermodynamic and kinetic relations was used to explain the target and film characteristics for each comb ination. A secondary goal of this dissertation is to identify if RIM PLA is a suitable method for the deposition of electroluminescent polymers based on the characterization results. 1.2 Objective s In order to achieve the goals previously stated, the obje ctives for this dissertation are: 1 Measure the optical constants (absorption coefficients, optical penetration depths, mode lifetimes) of the frozen matrices (solvents) and their dependence on polymer loading and temperature. 2 Determine the degree of molecul ar degradation, if any, in the post ablated targets. 3 Characterize the deposited film qualities by assessing their morphologies, molecular structure fidelity and luminescence. 4 Assess the possible thermally -induced phase transitions in the matrix and the rel evant heat dissipation mechanisms that lead to dendrimer ablation. 5 Correlate the experimental results to the previous assessment in order to establish the predominant FEL ablation mechanism s for each studied resonance mode.

PAGE 25

25 1.3 Structure of the Dissertatio n Chapter 2 gives a brief review of the technology which motivated this work (PLEDs), the relevant materials and the processing limitations which PLD attempts to overcome. Chapter 3 follows with a thorough description of the dendrimers, FEL facility and c haracterization equipment used. Chapter 4 details the results of the measurements of the optical constants of the resonance modes in the fr ozen matrices and their polymer and /or temperature dependence. Chapters 5 and 6 describe the characterization resul ts for the ablated targets and deposited films, respectively. Chapter 7 follows with a thorough discussion of the relevant ablation mechanisms and their correlation to each studied resonance mode. Chapter 8 concludes by addressing the main goals describe d in this introduction.

PAGE 26

26 CHAPTER 2 LITERATURE REVIEW 2.1 Introduction The deposition process for conducting and ele ctroluminescent organic materials is dictated primarily by the size of the molecules. For small organic molecules vacuum sublimation off ers great advantages and fewer limitations than the polymers wet processing methods. Several variations of Pulsed Laser Deposition (PLD) have been studied as possible alternatives to vacuum sub limation. The progress of this technique for electroactive p olymer deposition is discussed in this chapte r as a prelude to the studies in this dissertation. 2.2 Electroactive Polymers 2.2.1 Conducting (Conjugated) Polymers Although the initial studies of conducting polymers date back to 1862 when polyaniline was fi rst synthesized,15 the field would not attract considerable attention until 1977 when Heeger, MacDiarmid and Shirakawa discovered and developed the c onducting properties of polyacetylene.1619 Their findings led to a massive interest for an exciting and promising new class of materials, i.e. polymers that exhibited the optical and electrical properties of metal s and semiconductors, while retaining the attractive mechanical properties and processing advantages of polymers. Non -conjugated polymers are large macromolecules of repeating chemical units (typically carbon (C) containing) with excellent mechanical prope rties but electrically insulating behavior The backbone carbon atoms in non -conjugated polymers are fully covalent; each with an sp3 hybridization resulting in four energy -equivalent bonds.20 From molecular orbital theory s uch bonds have a energy gap in the range of ~ 8 10 eV, rendering them electrically insulating and vi sibly transparent.

PAGE 27

27 Conjugated polymers have a fundamental difference in their carbon -carbon bonding scheme, as their sp2pz hybridization results in three bonds and an additional higher -energy bond.20 The overlap of two adjacent pz orbitals lead s to the -bonding, resulting in substantial electron delocalization along the polymer backbone (Figure 2 1) The energy gap in a conjugated polymer is ~14 eV granting their semiconducting behavior Polyacetylene and similar systems can be treated as unifor m quasi one -dimensional atomic chains, as intra -molecular interactions are weak.21 However, treating such molecules as metallic systems proved to be inaccurate as the ground state of such a one -dimensional metal is unstable with respect to a structural distortion This is a consequence of the so -called Peierls distortion, a process in which a conjugated chain lowers its energy by doubling its period, forming a chain of buckled dimers.22 Such distortion is consistent with conjugated polymers having alternating single and do uble bonds, i.e. different average bond lengths (1.446 and 1.346 respectively)23 along their backbones. The lowering of the symmetry lowe rs the energy of the occupied states and stabilizes the distortion. An energy gap now separates the occupied ( bonding) orbitals from the unoccupied ( antibonding) band of states as those states near the Fermi level are eliminated by the Peierls transition .22 The width of the band gap near the Fermi level is defined by the difference between the Highest Occupied Molecular Orbital (HOMO) and the Lowest Unoccupie d Molecular Orbital (LUMO), which depends prim arily on the effective conjugation length.21, 23 These energy levels are often referred to as a valence and a conduction band in analogy with inorganic semiconductor nomenclature. The doping mechanism s in conducting polymers ha ve been discussed in terms of self localized excitations, such as solitons, polarons and bipolarons .2122, 2427 The doping process in conjugated polymers is quite different from inorganic semiconductors. The introduction of a

PAGE 28

28 charges in these polymer systems leads to a structural distortion of the lattice in the vicinity of the c harge, which lowers the energy of the system and stabilizes the charge. Solitons are topological excitations in degenerate ground state polymer s that result in a distortion in the latti ce pattern of the polymer chain; such distortions crea te single localized electronic states in the middle of the energy band gap which result in higher conductivity.22, 28 Polarons and bipolarons are other type of excitations or charge -storage species that occur in non -degenerate ground state polymers; most refer to p ositively charged species, i.e. p type doping or oxidation o f the conjugated polymers, because most polymers cannot form stable negative species.27 An increa sed conductivity can also be achieved by photoexcitation or by applying an external bias, in the case of electroluminescence; such excitation result s in neutral polaron-excitons (bound excited states) with weak to intermediate binding energies (at most a f ew tenths of an electron volt) that may or may not radiatively decay.21, 29 2.2.2 Photoluminescence and Electroluminescence in Conjugated Molecules Upon excitation t he spin wavefunction of the formed exciton may be described as either a singlet ( S = 0) or triplet ( S = 1).21 A singlet is a molecular electronic state in which all electron spins are paired, i.e. the spin of the excited electron is still paired wit h that of the ground state resulting in a net spin of zero. A triplet state results when the excited electron is no l onger paired with that of the ground state, i.e. they are parallel and have the same spin resulting in a net spin of one. Photoluminescence (i.e. radiative exciton recombination) processes in polyatomic molecules can be divided into fluorescent and phosph orescent, depending on the spin state transitions involved, i.e. an excitation to a singlet or a triplet state (Figure 2 -2).21 Fluorescen ce is s pin allowed radi ative emission process, which occur s from a singlet excited state (Si) to the singlet ground state (S0). Phosphorescence occurs when the absorbed photon energy leads to intersystem crossing from a singlet excited state (Si) to a triplet excited state (Tj) (Figure 2 2)

PAGE 29

29 The energy transition fro m a triplet excited state back to a singlet ground state, although classically forbidden is not limited by quantum mechanics.21 However, the rate of crossing is kinetically limited which result s in a time delayed relaxation (photon emission).21 In conjugated organic systems, absorption of a photon can give rise to a n or the more likely transition, which promotes an electron to the LUMO. The energy levels are broadened by vibration. R apid vibronic relaxation to the bottom of the LUMO (S1) is followed by radiative decay to the ground state (S0). Due to the con finement of the excitation, usually to a single polymer chain, the energy difference between singlets and triplets (the exchange energy) may be large. Consequently, the rate of intersystem crossing will be low but will eventually result in phosphorescence or the alternate indirect process of triplet triplet annihilation .21 Several reports have been written on fluorescence efficiencies, singlet triplet probabilities, triplet triplet annihilation and pho sphorescence of electroactive polymers.3 In particular, phosphorescent polymers are technologically important since they theoretically allow for electroluminescent devices with internal efficiencies of 100% due to their ability to harness both triplet and singlet excited states.3033 2.3 Technology of Interest: Organic and Polymer Light Emitting Diodes The first breakthrough in organic electroluminescence was achieved by Tang and Van Slyke in 1987 from Kodak when they reported efficient and low -voltage organic LEDs from p n heterostructure devices using thin films of vapor -deposited small organic materials.1 A few years later, researchers from Cambridge reported the first electroluminescence results from the polymer poly( p phenylene vinylene) (PPV) .2 Since then OLEDs and PLEDs, as they are commonly referred to, have garnered considerable interest owing to their promising applications in flat panel displays and potential replacement of c urrent liquid crystal display (LCD) technology. Some of the key advantages of OLEDs for flat -panel display applications are the

PAGE 30

30 molecules self -emitting property, high luminous efficiency, full color tuning by molecular design, wide viewing angles, high co ntrast, lo w power consumption, low weight and potentially large area flexible color displays.3 Electr oluminescence in organic solids proceeds differently than in their inorganic analogues as most polymeric and low -molecular weight materials form disordered amorphous films without a macroscopic crystal lattice.29 Because of the absence of extended delocalized states, charge transport is usually not a coherent motion in well -defined bands but rather a hopping process betwe en localized energy states at different distances resulting in low carrier 1 cm2 V1 s1).3 The basic electroluminescence steps in an organic solid include: (i) injection, (ii) transport, (iii) capture and (i v) radiative recombination of positive and neg ative charge carriers inside the layer. A highly simplified energy diagram for a single l ayer O LED is shown in Figure 2 3 .34 For simplicity the spatial variation of the molecular energy levels is drawn in a band like fashion c) metal cathode (e.g. Ca, Mg, Al) into the LUMO of the polymer, while holes are injected from the high work function a) anode ( e.g. indium tin oxide) into the HOMO (Step 1 in Figure 2 3) Both injection steps are associated with energy barriers (offsets)e h, for electrons and holes, respectively. Under an applied bias (V) the injected carriers will drift towards the counter electrode by hopping between different energy state s at varying distances (Step 2) Finally, the charge carriers will form an exciton with intermediate binding energy (few tenths of eV) (Step 3), which upon recombination leads to photon emission (Step 4). The efficiency and stability of the example device in Figure 2 3 is considerably limited by the energy barriers f or carrier injection and inefficient exciton confinement within the

PAGE 31

31 luminescent layer. Consequently OLED and PLED structures have increased significantly in complexity (Figure 2 4). The use o f specialized layers (hole injecting, hole transporting, hole blocking, emitting, electron transporting) have increased significantly the ir efficiency ;3, 35 E lectroluminescence in OLEDs can be significantly increas ed by carrier or exciton confinement within a multilayer device.36 37 Although some conducting and electroluminescent polymers dissolve in aqueous solutions (e.g. poly(3,4 -ethylenedioxythiophene) poly(styrene) : PED OT PSS), most can only be deposited from organic solvents.3, 35 The similar solubilities of materials in adjacent OLED layers limit significantly the processing, as re -dissolution of previously deposited layers bec omes an issue. Present organic LED thin film deposition procedures fall into two major categories: (i) vacuum thermal evaporation for small organics (< ~ 1 kDa) and (ii) wet coating techniques for polymer layers.29 Thermal vacuum evaporation and sublimation allows for tight tolerances on deposition rates, film thickness and areas of coverage. Since polymers generally crosslink or decompose upon heating at high temperatures, they cannot be thermally evaporated in a vacuum chamber.4 Hence, they are generally deposited by wet -coating a thin film from solution. Drop casting is used for ink jet print processing, while spin -coating is used generally for monochromatic area displays. The fabrication of PLEDs by processing from solution has promised to be less costly as the de position of the active layers does not require the use of vacuum technology. A lthough the thickness of spin-coated films may be controlled by the concentration of the polymer in the solution, the spinning rate, and the spincoating temperature, it is difficult to fabr icate thick films and thickness cannot be monitored during deposit ion. Additionally, inherent to these deposition processes is film non uniformity from resulting drying pattern s, particularly in ink jet printing Some reports have demonstrated an efficient

PAGE 32

32 combinatorial approach for the processing of solely spin coated multilayer PLEDs ; the technique utilizes orthogonal solvents for the individual layers, i.e. alternated deposition from hydrophilic and hydrophobic solvents. The layers must be free of pinholes which could allow for re dissolution. Most importantly, the success of this deposition process is limited to polymers with alternating solubilities, which is unrealistic for most multiple layer devices. Therefore, there continues to be a need for a deposition process for electroluminescent and conjugated polymers that removes the inherent limitations of wet -coating processes. Pulsed laser deposition has been and continues to be studied as a possible alternative. 2.4 Pul sed Laser Deposition of Organic Materials 2.4.1 Ultraviolet Pulsed Laser Deposition (UV -PLD) On e of the most technologically important applications of the interaction of laser light with a solid material is known as pulsed laser deposition (PLD) .12, 38 The fact that a pulsed laser beam is able to readily va porize almost any material suggested that it could be used to deposit thin films. The interactions of laser pulses with organic materials were initially investigated a few years after the first laser (ruby, 694.3 nm ) was successfully demonstrated in 1960.39 However, the fie ld did not gain much traction until decades later when further studies on ultraviolet (UV) ablation of polymers were done.40 41 These initial reports show ed promise for the precise control over the depth of ablation (etching) by controlling the number of laser pulses and fluence (energy/area) and what at that time seemed little to no photothermal damage. The results led to an increasing interest in the field that permeated to a wide range of precise -etching and thin -film applications in the pharmaceutical, electronics, microsensi ng and bioengineering industries.12, 38, 42 Figure 2 5 shows a typical PLD system. A target and a substrate in close proximity are placed inside the vacuum chamber.12, 38 The intense pulse laser beam is focused onto the target

PAGE 33

33 surface. Above a threshold energy densi ty, significant material removal will occur in the form of a substrate directed plume, which may contain neutral and ionized species from the target. The thres hold energy density will be dependent on the optical response of the target material, its morphology, and the laser pulse duration and wavelength.12 Nearly all orga nic polymers h ave moderate to intense absorption coefficients in the UV region.38 These absorptions are usually associated to electronic transitions from ground electronic states to excited singlet states as described in Section 2.2.2 Consequently, t he use of a UV laser, as in conventional PLD typically leads to fragmentation of the polymer chains due to the hi gh photon energies (~3 8 eV) which tend to exceed the binding energies of most common types of present bonds (~3 4 eV ). The fragmentation may also lead to a variety of photochemical reactions. For instance, p revious reports have successfully utilized th e absorbed radiation to photothermally depolymerize the target material, with subsequent repolymerization at the substrate.43 45 However, if the degree of re -polymerization was incomplete, such as when a photochemi cal reaction causes the loss of a pendant fu nctional group, the polymer would be subject to substantial structural modification in comparison with the bulk native material. UV-PLD has been used for the deposition of conducting and electroluminescent organi c materials Although the deposition of a lmost intact polymers is possible near the ablation threshold fluence, the process is complicated by molecular structures that cannot be subjected to the same bondbreaking and repolymerization that was used for a d dition polymers. Some of the reports include the use of a KrF excimer laser (248 nm) to deposit amorphous copper phthalocyanine (CuPc), aluminum tris 8 -hydroxyquinline (Alq3) and 4 -dialkylamino 4 nitrostilbensen (DANS)56, and a Nd:YAG laser (355 nm) for poly( p -phenylene vinylene) (PPV) films.7 Although the se films demonstrated electroluminescen ce, their efficiencies were not

PAGE 34

34 comparable to those of spin-coated or sublimated controls as a result of poor molecular structur e fidelity from target to su bstrate. The lack of excellent structural fidelity of the organic material from target to substrate in UV-PLD led to the development of a new variation of the technique where the composition of the target material was changed to provid e a softer desorption process. Matrix Assisted Pulsed Laser Evaporation (MAPLE), originally developed at the Naval Research Laboratory, minimized the photochemical damage that resulted from direct interaction of the UV laser light with the organic target.46 Th e MAPLE technique has been successfully used to deposit uniform thin layers of small organic mo lecules and polymers (thickness from 10 nm to over 1 accurate thickness control (<0.05 nm/laser shot) and good to excellent structural fidelity.9, 46 50 In MAPLE, a flash -frozen matrix consisting of a dilute solution (1 5%) of the polymer in a relatively vol atile and highly UV absorbing solvent is used as the laser target. The light material interaction in MAPLE can be described as a photothermal process.12 The photon energy is pr eferentially absorbed by the solvent caus ing a temperature rise in the focal volume; the solvent v aporizes and the polymer is heated. As the surface solvent molecules are evaporated into the gas phase, polymer molecules are exposed to the gas -target matr ix interface. If surface evaporation is the predominant ablation (material removal) mechanism, t he polymer molecules attain sufficient kinetic energy through collective collisions with the evaporating solv ent molecules, eventually reaching the deposition substrate. This deposition process can occur without any significant polymer decomposition by careful optimization of the processing conditions, i.e. laser wavel ength, repetition rate, solvent properties concentration, temperat ure, and background gas and pressure.12

PAGE 35

35 In light of the wet processing limitations for organic and polymer LE Ds, MAPLE was used for the deposition of light emitting polymer films. Fitz Gerald, et al. use d a pulsed excimer laser (KrF, 248 nm, 10 Hz ) at a very low fluence 0.04 J/ cm2 to grown thin films of a ruthenium tris(bipyridine) -centered star -s haped poly(meth yl methacrylate) from a solution of dimethoxy ethane.8 However, comparison of 1H NMR (nuclear magnetic resonance) and GPC (gel permeation chromatography) spectra between the native an d deposited material revealed distinguishable d egradation. These results confirmed that irradiation at high -energy ultraviolet wavelengths, irrespective of the target composition, would never be a suitable method for the deposition of intact luminescent organic materials. 2.4.2 Resonant Infrared Pulse d Laser Deposition (RIRPLD) A less explored technique, Resonant Infrared Pulsed Laser Deposition (RIR -PLD) is a novel variant of conventional PLD in which the laser is tuned to vibrational modes in the target material (solid polymer).51 53 The intense laser irradiation (~0.25 eV) gives rise to multi -photon excitation over the vibrational manifolds of g round electronic states which may later be followed by thermal decomposition and deposition (Figure 2 6) In the a bsence of electronic excitation, the complex chemical and physical s tructure of the polymer may be preserved. The term resonance in RIR -PLD implies that excitation occurs at the mode maxima when the optical (light) penetration depth will be minimal (see Se ction 4.2). A consequence of resonance is that the volumetric energy density will be high enough to result in ablation. For non resonant irradiation, characterization of the deposited films has shown substantia l modification in the polymers mid IR finge rprint r egion of the absorption spectrum.54 Since every polymer has different a different vibrational mode fingerprint, a tunable laser in the mid infrared region allows for model studies (Section 3.4.1) of RIR PLD ablation How ever, the

PAGE 36

36 availability of commercial table top lasers close to resonance wavelengths (e.g. Er:YAG; 2.94 -scale processing if the technique proves valuable .11, 55 A recent comprehensive study has shed more light onto the ablation mechanisms of RIR PLD.13 14 The study used plume shadowgraphy1 and temperature rise calculations at the target surface to identify the predominant mechanisms of ablation. The conclusions showed that phase explosion (explosive boiling; see Section 7.8.2.2) of the superheated surface layers is responsible for the initial stages of polymer removal, followed by a recoil induced ejection of lower molecular -weight species The se results confirmed the mechanisms previously suggested for tissue ablation.56 T he matrix assisted analogue of RIR PLD (also known as RIR MAPLE) has been studied for a variety of systems particularly on the deposition of the electroactive polymer poly[2 methox y 5 -(2 -ethylhexyloxy) 1,4 -phenylenevinylene] (MEH -PPV) .911 In one of the studies, PLED devices were fabricated; similar electroluminescence spectra and current volage (I -V) characteristics were observed when compa red to conventional spin -coated devices.10 The peak emission wavelengths were all within 10 nm of electroluminescence spectra of spin coated devices with only slight spectral broadening. FTIR spectra of the deposited films also showed little changes from the native material. Most recently, Pate et al. lo oked at ablation of MEH PPV at a variety of processing conditions (different target to -substrate distances, ambient base pressures, laser fluences, etc.) and target compositions with a t able -top laser (2.9 m).11 This study varied deposition parameters to improv e the film properties of the deposited films However it failed to describe the underlying mechanisms of ab lation involved in RIR MAPLE. 1 The technique setup consists of a nitrogenpumped dye laser that irradiates parallel to the target surface at a height of several millimeters, and a telephoto lens that images the silhouette of the ablation plume onto a color charge c oupled device.

PAGE 37

37 This dissertation attempts to provide those details through modeling matched to the characterization of deposited luminescent polymer films at different laser processing conditions for different frozen matrix compositions. F or the first time, the molecular structure of post ablated targets is thoroughly characterized by nuclear magnetic resonance spectroscop y and other techniques in order to support the ablation mechanisms proposed. The experimental data are correlated to thermodynamic and kinetic relatio ns in order to understand the FEL -frozen target interactions. As will be discussed RIR MAPLE thin -film deposition can occur through different target ablation mechanisms; consequently, the acronym RIR -MAPLE is inappropriate, as it suggest s only surface evaporation of the solvent ( and polymer ) molecules. In this dissertation the technique will rather be called RIM PLA, a more accurate description of the technique: resonant infrared (matrix assisted) pulsed laser ablation. 2.5 Electroactive Dendrimers 2.5.1 Motivation for Materials Selection D egradation in organic and polymer L EDs essentially appears primarily in the form of a decrease in device luminance over time, which can proceed through three indepen dent and visually distin ct mechanisms.5758 These mechanism s are referred to as (i) dark -spot degradation,5964 (ii) catastrophic failure, and (iii) operational/ intrinsic degradation.6567 The first mode, associated primarily with degradation at the device electrodes, occurs through the formation of non -emissive regions.58 64 The second mode, mainly associated with defects in the organic layers, occurs through the development of electrical shorts that result in a sudden decrease or total loss of luminance as a result of large leakage currents.68 The third degradation mechanism is reflected in a long -term intrinsic decrease in the brightness of the emissive area (fade ), without any obvious change in device appearance, and mainly occurs during de vice operation.65 67 This mechanism has been extensively studied and associated with the

PAGE 38

38 accumulation of nonradiative recombination centers and luminescence quenchers in the emissive zone.65 67, 6980 Recrystallization has been an extensively studied intrinsic degradation mechanism of light emitting diodes based on amorphous organic and polymer layers.29, 8182 Singlet excitons are e fficiently quenched by defects and charge -dipole induced effects at the surface of grain boundaries of polycrystalline layers primarily induced by Joule heating upon device operation.3, 29, 83 Structure modificati on by the introduction of bulky molecules resistant to crystallization has offered an appropriate solution to this intrinsic problem.58, 84 85 Electroactive dendrimers are highly non -planar macromolecules exhibitin g high glass transition temperatures ( Tg). The existence of different conformers in addition to the nonplanar structures result in amorphous glasses, as determined by x ray diffraction and differential scanning calorimetry.84 Dendrimers contain well defined chromophore core s, whose absorption and solution emission are unaffected by the attachment and successive branching of dendrons.86 As has been shown by electrochemical st udies on dendrimers, charge injection takes place into the core and not into the dendrons which have a wider energy gap.87 The dendrons hence act as a spatial separator and insulate the cores. Thus the barrier for injection of charge into a PLED remains unaffected by dendrimer generation but a change in the chromophore separation manifests itself in the charge transport properties.8788 2.5.2 Studied Dendrimer Families Two families of dendrimer structures were use d for RIM -PLA in this work ( shown in Figure 2 7 ).89 90 For the zeroth generation (G0) the fluorescent dendrimer possesses no dendrons but merely the emissive distyrylbenzene chromophores. The number of stilbene un its per dendrimer increases as 6 x (2G 1) with generation, where G denotes the generation number. IrG1 and Ir G2 contain a fac -tris(2 phenylpyridine) iridium cored, phenylene based dendrons and

PAGE 39

39 2 -ethylhexyloxy surface groups. The use of two different dendrimers (one of which has a heavy center ion) help ed understand if the ablation characteristics are polymer -dependent. The UV -visible a bsorption and photoluminescence (PL) spectra of the fluorescent dendrimer family are shown in Figure 2 8.90 For all generations, the photoluminescence quantum yields (PLQY PL) in solution following excitation at 325 nm (in the dendron band) and 442 nm (in the core) are the same within the experimental error (Table 2 1).90 In the solid state, the steady -state PL spectra and the PLQY are ve ry di fferent from the solution results. The steady -state PL spectra (Figure 2 8) in the zeroeth (G0) and first generations show ed a substantial red -tail. Mo st importantly, the film PLQY s are up to an order of magnitude smaller than in solution ; the values inc rease with generation (Table 2 1).90 The UV -visible absorption and PL spectra of the phosphorescent dendrimer family is shown in Figure 2 9.89 The PL spectra for both Ir -cored dendrimers contain the same features, a peak at ~520 nm and a shoulder at ~550 nm. The decrease in red shift of the spectrum relative to Irppy3 core i s due to the dendrons on Ir G1 decreasing the intermolecular interactions of the emissive cores within the film.89 Attachment of the second -generation dendron enhances this effect. The thin -film PLQY of the first generation of the phosphorescent dendrimer is ~3 times that of the fluorescent analogue. Consequently this dendrimer was more effective at testing any photoluminescence degradation from RIM -PLA deposition. 2.6 Conclusion Pulsed laser deposition (PLD) of organic and polymeric materials has been an area of intense study for the over 25 years. The advent of electroluminescent polymer technologies has attracted more attention to PLD due to the inherent limitations of wet coating me thods. Although, RIM PLA has effectively been demonstrated for the deposition a luminescent polymer followed by device fabrication ,10 no study has yet led to an unders tanding of the underlying

PAGE 40

40 mechanisms of ablation for this matrix assisted technique; that is the emphasis of this dissertation.

PAGE 41

41 Figure 2 1. F ormation of a bond by weak overlapping of the pz orbitals of two successive carbon atom s and resulting electron delocalization. Figure 2 2. Radiative and non radiative decay processes in a typical polyatomic molecule assuming that the radiative transition in a given spin manifold always occur from t he lowest excited state ( knr: nonradiative decay rate to the ground state ; kr: radiative decay rate to the ground state ; kis: intersystem crossing rate to the triplet manifold).21

PAGE 42

42 Figure 2 3. Electroluminescence in a m onolayer organic or polymer LED: ( 1) charge carrier injection, (2) charge carrier transport, (3) exciton formation, (4) radiative exciton ach: hole injection ebi: builtin potential, V : appl ied voltage, Ve: effective voltage across the organic layer, q : elementary charge (adapted from Brtting ).34 Figure 2 4. Evolution of OLED device structures (HIL: hole injecting layer; HTL: hole transporting layer; EML: emitting layer; HBL: hole blocking layer ; ETL: electron transporting layer).35

PAGE 43

43 Figure 2 5. Schematic of basic components of a PLD system. Figure 2 6. Energy diagram comparison between ultraviolet (electronic) and infrared (vibrational) laser excitation (adapted from Chrisey, et al.) .12

PAGE 44

44 Fi gure 2 7 Dendrimer families studied in this work: f luorescent tris(d istyrylbenzenyl)amine cored (G0 G3) and phosphorescent fac -tris(2 phenylpyridine ) i ridium (III) -cored (Ir G1 IrG2). Figure 2 8 UV-visible and photoluminescence spectra of fluorescen t dendrimer family (G0 G3).90

PAGE 45

45 Figure 2 9 Solution UV -visible and film photoluminescence spectra of phosphorescent dendrimer family (Irpp y3 core, IrG1 and Ir -G2).89 Table 2 1. Solution and thin-film PLQY of all fluorescent dendrimer generations.90 Fluorescent Dendrimer Generation PL, %) Thin -PL, %) G0 61 5 G1 63 7 G2 63 8 G3 62 12 Table 2 2. Solution and thin-film PLQY of all phosphorescent dendrimer generations.89, 9192 Fluorescent Dendrimer Generation So PL, %) Thin -PL, %) IrG1 70 22 IrG2 69 31

PAGE 46

46 CHAPTER 3 EXPERIMENT AL METHODS 3.1 Materials 3.1.1 Dendrimers Two types of families of electroluminescent dendrimers we re studied in this research: fluorescent t ris(distyrylbenzenyl) amine -cored and phosphorescent fac tris(2 -phenylpyridine) iridium(III) -cored macro molecules. Due to the difficulty in the synthesis of these dendrimers, only the first gener ation of each family was used for the depositions the structure s of which are sho wn in Figure 3 1 Table s 3 1 and 3 2 list fundamental properties of the materials. A collaboration was established with Dr. Paul L. Burn who was then at the University of Oxford and is now at the University of Queensland, to o btain ~750 mg of the fluores cent and ~1 g of the phosphorescent macromolecule The purity of the as received materials was characterized by Nuclear Magn etic Resonance (NMR) spectroscopy, M atrix -Assisted Laser Desorption/ Ionization Time -of -Flight (MALDI TOF) spect r ometry and Fourier Transform Infrared (FTIR) spectroscopy. The dendrimer powders were wrapped in aluminium foil and kept in a desiccator at all times. 3.1.2 Solvents Figure 3 2 shows the molecular structure of the solvents used in this work. The solvents used were certified ACS grade (purity assay: 99.5% including prese rvative, Fisher Scientific ): Table 3 3 list s the relevant fundamental properties for these solvents.93 3.2 Measurements of Optical Constants of Solvents Inf rared transmission measurements were done on the solvents used in this work, in order to determine the liquid and solid -state absorption coefficients and optical penetration depths at the resonance frequencies of interest.

PAGE 47

47 3.2.1 Transmission Liquid Cell A variable path length demountable liquid cell (DLC 2 Harrick Scientific ) was used for transmission spectroscopy (Figure 3 3 ). The cell has a chemically resistant PTFE body with luer lok fittings. ZnSe windows ( index of refraction n = 2.2 2.4, 13 mm 2 mm disks, International Crystal Laboratories) were used due to their minimal water absorption (1 mg / 100 g H2O) and high transmission (~70 %) in the mid infrared spectrum (0.6 21 m ) The aperture in the cell is 8 mm; 15 500 m PTFE spacers are plac ed between the windows. 3.2.2 Fourier Transform Infrared (FTIR ) Spectrometer A Bruker 113v fast -scan Fourier transform i nterferometric spectrometer or FTIR was used for all optical constant measurements. The main difference between an FTIR and a monochrom atic spectrometer is that the system has an interferometer instead of a prism or grating monochromators. The theoretical basis of the Michelson interferometer and its application to the FTIR have been discussed elsewhere.9496 The system uses a Globar (silicon carbide) light source, a KBr beam splitter and a room temperature pyroelectric deuterated triglycine sulfate (DTGS) detector The double -sided movable mirror in the chamber is air bearing, which limits the ultimate chamber pressure to ~1 x 103 bar (~ 1 Torr). The mirror moved cont inuously at a scan speed of 12.5 kHz and its position was calibrated by a HeNe laser and a white light reference interferometer. Each spectrum was the result of 32 coadded scan s ranging from 400 to 4000 cm with a spectral resolution of 2 cm. 3.3 Control Films Control dendrimer films were deposited via spin -coating for comparison purposes. Solutions in chloroform (12 mg/mL) wer e spin -coated (Specialty Coating Systems, P6700) at 3000 rpm for 30 s. The films were dried for 4 hours at 25 C in low vacuum.

PAGE 48

48 3.4 Laser Processing 3.4.1 Laser Source 3.4.1.1 Free-electron l aser (FEL) A variety of reviews about the physics of Free -Electron Lasers ( FEL s) are available.97101 In essence, FELs convert the kinetic energy of electrons into light .102 103 Figure 3 4 shows a schematic of the typical FEL. An accelerator produces free electrons at relativistic speeds in a vacuum. The electrons then pass through a wiggler or an undulator a periodic magnetostatic field in an optical cavity. The essential feature of FELs is that the kinetic energy of the electrons is decremented when the undulator deflects the electrons into a serpentine path while giving off photons of lig ht which propagat e and are amplified along the optical cavity axis By properly phasing the maximum amplitudes of the electromagnetic and electron waves, it is possible to create a resonance condition in which the radiative process is enhanced or amplified by classical electromagnetic -field effects .97 Experimental realizations of FELs are based on a series of accelerator technologies, operating with electron energies from as low as keV and up to GeV. Collectively, FELs operate from the terahertz through the ultraviolet range and via intracavity Compton backscattering104 105 into the X ray and gamma ray regimes ; however, each accelerator technology corresponds to a restricted wave length range of operation. FELs are versatile and in some cases as in this research, are unique light sources Because the wavelength of the emitted light depends on the energy of the electrons and the properties (spacing, strength) of the magnetic field, a n FEL can be tuned within its wavelength range of operation with relative ease. Whereas the lasing medium for FELs is the free beam of electrons, the lasing medium for conventional lasers is bound electrons and the accompanying atomic nuclei, w hich res tricts laser output. Consequent ly, FELs can be tuned to wavelength ranges in

PAGE 49

49 which conventional lasers cannot operate and can have optical power, pulse and polarization characteristics that conventional lasers have not achieved. The FEL optical pulse stru cture correlates with the bunching of the electron beam whereas the polarization is determi ned by the magnetic field.97 98 FELs can continuously produce optical pulses at a set repetition rate or at multiples of it Other FELs have a complex optical pulse structure with a burst (macropulse) of micropulses such as the one used for this work The micropulse lifetime can be picoseconds or shorter and the micropulse repetition rate within a macropulse can be as high as gigahertz. The repetition rate of the macropulse can be as high as many tens of Hertz. 3.4.1.2 W. M. Keck FEL at Vanderbilt University The Mark -III FEL located at Vanderbilt University was used for all the depositions carried out in this research. The temporal pulse structure in a free -electron laser is determined by the choice of electron accelerator and rf generation system. In the Mark -III FEL electrons are extracted from a thermionic cathode by a microwave field in a resonant cavity (S -band klystron, 2.865 GHz), producing a train of micropulses, which after comp ression, are ~ 0.7 1 ps long and separated by approximately 350 ps.106 This train of ~3 1 04 micropulses extends for about a 4 s macropulse at a repetition rate of 30 Hz. Figure 3 5 shows a breakdown of the pulse structure of the Vanderbilt FEL. The electrons then enter the radio -frequency linear accelerator into the wiggler with an energy of ~43 MeV. The energy of the electr on beam and the spacing/strength of the FEL wiggler allow for tuning of the laser in the midinfrared region from 2 10 m. There is a frequency chirp during the macropulse of approximately 2 3% of the central frequency Tuning the FEL from one freque ncy to another in a narrow range (10%) around a central

PAGE 50

50 frequency is generally accomplished by adjusting the wiggler magnetic field; larger adjustments of frequency require returning the electron-beam energy and sometimes inserting a new mirror set in the optical cavity to optimize the laser gain. The energy in each FEL micropulse is of the 7 W).52 The macropulse energy is dependent on the tuned wavelength, but c ould reach up to 22 mJ in our experiments. Previous studies have shown no difference in the dynamics of ablation (i.e. ablation threshold radiant exposure and the ablated crater depth for a defined radiant exposure) due to the micropulse (ps) structure of the FEL.107 108 Therefore, throughout this dissertation a continuous FEL macropulse ( s) is assumed The laser operates at the lowest order (fundamental) transverse electromagnetic mode (TEM00) resulting in a Gaussian spatial beam profile. The laser spot size (radius) of the Gaussian laser beam is conventionally defined as the point at which the original intensity I0 falls to 1 / e2 of its value.14 All spot size measurements in this study were measured by Stephen L. Johnson, a collaborator at Vand erbilt University.14 The measurements were done via the knife edge method. The method uses a razorblade to block the transmission of the laser beam while its intensity is recorded as a function of the r azors po sition. Figure 3 6 shows the calibration curves for several wavelengths of the FEL beam with accuracy to within 10 % based on standard deviation The measurements of spot size are critical to our work, as t hey define the appropriate pulse energy for the d esired fluence. The knife -edge method calibration was used for determining the appropriate spot sizes for all the fluences used in this research. For the longer wavelengths, a minimum spot size was interpolated or extrapolated from the available data, an d the macropulse energy was adjusted as necessary. Specifically, for the 6.23 and 8.22 m wavelengths, a minimum spot size was assumed at

PAGE 51

51 distance s of 53 and 56 cm, respectively. Note that this has proved to be a more appropriate method than the alternat ive measurement from the burnt ellipse pattern on a thermal paper target. Being heat sensitive, thermal paper turn s dark wi th laser irradiation. However, a measurement from thermal paper implies a large measurement error, as the size of the burnt ellipse can dramaticall y change from a small variation of energy, or by increasing the number of macro pulses. Therefore, we believe th e thermal paper method was not appropriate for fluence calculations as it can easily over -estimate the real spot size. Table 3 4 tabulates the values (and estimated maximum errors) of the spot size assumed for the FEL wavelengths used in this work, as determin ed from the calibrations curves Since the laser beam enters the window at a 45 angle, as it rasters the target (1 cm radi us) the actual calibration distance will vary up to 0.5 cm (from calculations with the law of cosines) As the spot size gets smaller, the error increas es due to the steep curvature of the calibration curves at low FEL wavelengths. T herefore, the spot s ize for ablation at 3.3 m c an change up to 6 8 % as it moves radially across the target This in turn means that the average fluence that will be cited in this work could change as much as 300% as a result of the chamber setup 3.4.2 Preparation for Ablati on 3.4.2.1 FEL beam c ontrol Once the laser is tuned to the desired wavelength, the beam is sent through vacuum pipes acros s the facility rooms (Figure 3 7 a). The beam spectral profile is simu ltaneously monitored (Figure 3 7 b). When it exits the delivery pipe, t he beam is guided by gold -coated mirrors while its intensity is adjusted by polarization sensitive attenu ation. The macropulse energy is measured with a pyroelectric joulemeter (EPM 2000, Molectron Detector, Inc. ). Two galvometric motors control m irrors that set the rastering profile. A LabVIEW program was written by Stephen L. Johnson that controls the laser beam position.14 The emphasis of the code was in guaranteeing

PAGE 52

52 that the rastering speed would be fast enough as to prevent a second macropulse from striking a previously irradiated area. In that sense, a previously melted section of the target would have enough time to resolidify assuming good insulat ion of the target well. The program scans the beam in a spiral pattern, with its angular frequency increasing as the radius decreases (~2 Hz in periphery, ~6 8 Hz at center). The radial frequency was ~0.2 Hz. Figure 3 -8 shows all previously described components, and their location relative to the deposition chamber. 3.4.2.2 Substrate preparation Soda lime silicate plain microslide s (2947, Corning Glassworks ; 73% SiO2, 14% Na2O, 7% CaO, 4% MgO, 2% Al2O3) were used as substrates for all fi l ms, as these off ered the most versatility for all the required characterization techniques. Substrates were prepared by cutting microscope slides with a diamond cutter into 2.5 cm 2.5 cm 1 mm squares. The following cleaning method was used for all substrates of both spin -coated and laser ablated films, in order to minimize changes in sticking coefficients. The substrates were secur ed in an inert customized PTFE holder and then sonicated for 15 minutes each in the following s olvents: a cleaning solution of ~10 0 mg of Alconox detergent per 500mL of deionized water, deionized water, acetone, and isopropyl alcohol. The organic solvents used were all either ACS or HPLC grade (Fisher Scientific). After the final wash, the glass substrates were placed in a plasma cleaner (Harrick Scientific PDC 32G) and treated with UHP oxygen plasma for 20 minutes. Each clean substrate was then placed in an individual plastic container, which had been previously dusted. In the case of the spin coated control films, depositions followed immediately after cleaning. For laser ablated films, deposition occurred 23 days after cleaning, but the substrates were additionally dusted with a triple -filtered ( 0.3m ) inert gas (Fisherbrand Aerosol Duster) prior to placing inside the chamber.

PAGE 53

53 3.4.2.3 Target p reparation The dendrimer was dissolved in each solvent at a concentration of 0.5 % (by weight) The assumed densities for chloroform and toluene are li sted in Table 3 3 A stainless steel well (2.5 cm diameter, 2 mm depth) was fitted with a PTFE screw and supported inside cup filled with liquid nitrogen. The dendrimer solution (~1.5 2 mL) was slowly added to the well and allowed to solidify. The tar g ets would typically be frozen after 5 minutes, but would be left on top of the liquid nitrogen for at least 5 additional minutes. Figure 3 9 show s the typical resulting target. Minimal water condensation occurred at the target surface while being frozen due to the exerted nitrogen pressure. As the target was moved and settled inside the chamber, water condensation became evident in all targets, even at the hydrophobic surface. All targets discussed in this work remained visibly f rozen during all deposition times except some ablations at high fluence (particularly for toluene) I f target melting was not iced those depositions were stopped when the target was still solid The target would then be removed from the deposition chamber and refrozen. The sa me target would then be placed back in the chamber and ablation would be restarted and completed 3.4.3 Thin-Film Deposition The deposition system used for all ablations was assembled by students from Dr. Richard F. Haglunds group and staff at Vanderbilt University. The custom -built stainless steel chamber used for all depositions is shown in Figure 3 10. Once the laser energy has been adjusted, the FEL beam is sent into the vacuum deposition chamber at a 45 angle through a BaF2 window and the mirrors are adjusted to center the beam at the target surface. A coaxial HeNe laser is used to guide the infrared laser to a thermal paper target used to test the beam position. In a typical deposition process, a clean substrate is attached to the chamber holder with a small piece of carbon tape. A freshly frozen dendrimer target is placed atop an insulating PTFE

PAGE 54

54 ring inside the chamber. The chamber is then pumped down to high vacuum (8 105 P0 1 104 T orr ) in ~ 5 minutes. The laser rastering program w as then started and timed for the desir e d irradiation time. Figure 3 11 shows a target being ablated under normal room lighting and under UV lamp irradiation. The maximum pressure rise d uring ablation was to 1 103 Torr. The chamber pressure was monitored by a Pirani gauge from 2 103 Torr to 760 T orr, and a standard hot -filament ion gauge for lower pressures. A turbomolecular system (Leybold BMH 70 DRY) was used for high vacuum pum ping. Once the deposition was finished, the chamber wa s vented with dry nitrogen in order to prevent water vapor from ambient condensing on the chamber walls. The substrate wa s rem o ved from the holder (Figure 3 12), placed in a case and covered in alumini um foil. All substrates were stored in a dry desiccator for as long as possible. 3.4.4 Constant and Variable Deposition Parameters For all depositions, the FEL was operated at a constant macropulse repetition rate (30 Hz) at the same rastering speed (spir al profile described in S ection 3.4.2.1) The working distance between the target and substrate was constant (2.38 cm ). The concentration of the dend rimer solutions was never varied The processing variable parameters were: (i) FEL wavelength (determine d by the resonant mode of the solvent), (ii) the irradiation time, and (iii) the laser fluence (energy per unit area). 3.5 Characterization 3.5.1 Target Characterization Following a deposition, each re -melted target was collected in an individual vial, lab eled appropriately and allowed to dry. The molecular and structural integrity of these targets was examined b y FTIR spectroscopy, NMR spectroscopy and M ALDI TO F spectrometry

PAGE 55

55 3.5.2 T hin-Film Characterization The molecular and structural fidelity of the la ser -deposited thin films was characterized by FTIR spectroscopy and MALDI TOF spectrometry. NMR spectroscopy could not be done on any of the films as depositio ns result in nanogram quantities, which are insufficient for the technique. Film morphology was studied with optical microscopy, fluorescence microscopy, Atomic Force Microscopy (AFM). Thickness and roughness measurements were do ne with a stylus profilometer. 3.5.3 Techniques and Apparatus 3.5.3.1 Nuclear Magnetic Resonance (NMR) s pectroscopy Targe t molecules were characterized by 1H NMR using a Varian VXR 3 00 MHz or a Gemini 300 MHz FT -NMR spectrometer. All measurements were done at 298 K in CDCl3 solution with tetramethylsilane as the re ference. In order to increase signal to noise, 256 transien ts were added in each spectra. 3.5.3.2 Fou rier Transform Infrared (FTIR) s pectroscopy Laser abl ated films were deposited on polished 13 mm 2 mm NaCl disks (International Crystal Laboratories). Targets were re -dissolved in their previous solvent and drop -casted on top of clean NaCl disks and allowed to dry under vacuum (no heat). Transmission measurements were carried out in a nitrogen -purged Nicolet Magna 760 spectr o meter ( Thermo Electron Corp.) and data was processed with the OMNIC software suite. Th e system uses a Globar (silicon carbide) light source, a KBr beam splitter, and a room temperature pyroelectric deuterated triglycine sulfate (DTGS) detector. Each spect rum was the result of 64 coadded scans ranging from 650 to 4000 cm with a spectral resolution of 2 cm.

PAGE 56

56 3.5.3.3 Matrix Assisted Laser Desor ption/ Ionization Time -of -Flight (MALDI TOF) mass spectrometry High resolution mass spectrometry was performed on a n Agilent 6210 TOF -MS with a MALDI attachment The matrix used was 2,5 dihydroxybenz oic acid Isotopic d istribution: The calculated (theoretical) isotopic distribution for the phosphorescent dendrimer (C135H156IrN3O6) is 2106. 1595 (30% monoisotope mass for M+ ), 2107.1628 (48%), 2108.1636 (89%), 2109.1660 (100%), 2110.1689 (70%), 2111.172 0 (34%), 2112.1752 (15%), 2113.1784 (5%).89 The calculated isotopic distribution for the fluorescent dendrimer (C162H183N) is 2142.4345 (55% monoisotope mass for M+ ), 2143.4379 (100%), 2144.4413 (89%), 2145.4446 (54%), 2146.4480 (24%), 2147.4514 (9%), 2148.4547 (3%).90 Note that M+ stands for the singly ionized molecule ; that is, it was not pr otonated during the ionization process. The distributions were calculated with the Agi lent MassHunter software suite. Error : The error fr om the MALDI TOF spectra result s was calculated with Equation 3 1. 610 ) ( ) ( l theoretica actual l theoretica ppm Error (3 1) 3.5.3.4 Optical m icros copy An Olympus BX60 optical microscope fitted with a charge coupled device (CCD) digital color camera ( Spot RT ; Diagnostic Instruments ) was used. P ho tographs of depositions on glass substrates were processed with the SPOT Advanced software The microsco pe is equipped with dual light sources (Olympus TH3 power supply ) for both transmitted and reflectance imaging. The contrast in transmitted/reflected light is based on the variations of optical density and color within the film. All measurements were don e with bright -field illumination at both transmission

PAGE 57

57 and reflection modes with 10X and 50X objectives. Reflection mode resulted in better contrast for rough films 3.5.3.5 Fluorescence m icroscopy An Olympus IX70 inverted microscope fitted with a side -por t CCD camera (Princeton, RTE 1300 1030) was used for fluorescence image captures Bright -field mode images from a tungsten lamp excitation were also acquired at the same locations for comparison In fluorescent microscopy filtered high energy light (1 00 W Hg light source) illum inates the sample focal spot but the lower energy emission (fluorescence) fro m the material film results in the acquired image. A dichroic mirror is used to separate the excitation and emission light paths. Two filters are used a long with the dichroic mirror: (i) a bandpass filter for the e xcitation from the Hg lamp (360 nm peak / 40 nm bandwidth; Chroma Technology), and (ii) a bandpass filter for the emission from the dendrimer films (525 nm peak / 5 0 nm bandwidth; Chroma Technol ogy). The d ichroic mirror is mounted on a modular optical block along with the excitation and the emission bandpass filters All images were collected through 10x and 40x objective lenses (Olympus U Plan Fl, 0.30NA and SLC Plan Fl, 0.55NA, respectively ), and processed with the TSView data acquisition software. A camera calibration for a grid of 1030 1030 pixels was used (10X: 1.32 m/pixel; 40X: 0.345 m/pixel). 3.5.3.6 Atomic Force Microscopy (AFM) A Dimension 3100 (Veeco Digital Instruments) scanning probe microscope operated with a Nanoscope V controller was used. The m icroscope was operated in tapping mode with cantilever tips with a nominal radius of 8 nm (Model RTESP Veeco Instruments). AFM images were obtained in height and phase modes. In both modes, the substrate surface is scanned with a sharp tip located at one end of a cantilever (P -doped Si, Model RTESP Veeco Instru ments )

PAGE 58

58 (Figure 3 13) Piezoelectric scanners are used to raster the tip over the surface, while the cantilever is oscillating at resonance (~ 300 kHz) A constant resonance amplitude is set for the cantilever. A He Ne laser beam reflects off the cantilev er onto a position-sensitive photodiode in order to monitor tip -surface interactions. Feedback between the scanners and the photodiode through an electronics interfa ce is used to produce an image. As the tip interacts with the surface, the amplitude, reso nance frequency and the phase angle of the oscillating cantilever vary. As it oscillates over the surface, the amplitude will change depending on the artifacts the tip encounters resulting in a topographic height image. Add itionally, these interactions will cause a phase off -set in the resonance frequency, resulting in the phase image. Typically, the phase image will result in higher contrast than the phase image but is also more s ensitive to material dislodges as polymer buildup (dragging) in the tip leads to changes in the oscillating phase of the cantilever. Material dislodges were a high occurrence issue with RIM PLA films Phase images have been used in the past to study changes in stiffness, viscoelasticity and chemical composition.109 111 All measurements were done at room temperature and 1 atm. 3.5.3.7 S tylus profilometry A KLA Tencor P 2 long scan p rofilometer was used for thickness measurements. The e quipment can measure thickness from ~ 5 nm to ~ 65 210 mm. The profilometer scan data also gives results on the surface roughness of the film. Dendrimer films were deposited on glass substrates partially masked with pressure sens itive tape, which was removed post abla tion Thickness measurements were made by scanning the tip radius) along different lengths orthogonal to the film step created by the mask. A force of 20 mg was applied to the stylus at a scan length of 30

PAGE 59

59 mm for all measurements. The vertical deflection measures the change in step height which is t he thickness of the sample. 3.5.3.8 Thin-film p hotoluminescence q ua n tum e fficiency Room -temperature thin -film photoluminescence was m easured with a spectrophotometer (F P 6500, Jasco, Inc. ) equipped with a DC -powered 150W Xenon l amp source and a photomultiplier tube (PMT) detector A quantum yield measurement system (Figure 3 14) for solid film samples acquires the light with an integrating sphere (60 mm diameter; BaSO4 coating; Spectralon reflect ance standards). Equations 3 2 and 3 3 were used for the qua n tum yield calculations:112 ) ( ) ( ) 1 ( ) (0 e i PLL E A E ) ( ) ( ) (0 0 L L L Ai (3 2 ), (3 3) where Ei( ) is the integrated luminescence from direct excitation of the substrate, E0( ) is the integrated luminescence from secondary excitation of the substrate, Li( ) is the integrated excitation from direct excitation of the film L0( ) is the inte grated excitation from indirect excitation of the film Le( ) is the in tegrated excitation without a film.

PAGE 60

60 Figure 3 1 Dendrimers used in this research: a) P hosphorescent fac tris(2 -phenylpyridine) iridium(III) -cored ; b) F luorescent tris( distyrylbenzenyl)amine cored Figure 3 2. Molecular structures of matrix solvents. Figure 3 3. Demountable liquid cell used for transmission spectroscopy.

PAGE 61

61 Figure 3 4. Schematic of the principal components of the free -electron laser. Figure 3 5. Pulse structure of the Vanderbilt FEL (not to scale).

PAGE 62

62 Figure 3 6. C alibration curves for radius of a circular spot size at different FEL wavelengths (CaF2 focusing lens). Figure 3 7. FEL laboratory: a) FEL beam is guided through (blue) pipes under low v acuum towards the exit window; b) monitor displaying pulse intensity/bandwidt h.

PAGE 63

63 Figure 3 8. Setup for beam guiding to the vacuum deposition chamber. Figure 3 9. Frozen target of phosphorescent dendrimer (G1) in chloroform.

PAGE 64

64 Figure 3 10. RIM -PLA thin -film deposition chamber. Figure 3.11. Frozen dendrimer targ et during ablation: (a) under normal room light and (b) under a UV lamp.

PAGE 65

65 Figure 3 12. Dendrimer covered substrate after RIM -PLA deposition. Figure 3 13. Schematic of an atomic force microscope.113 Figure 3 14. Schematic of integrating sphere setup used for thin-film quantum yield measur ements (adapted from FP 6500 manual).

PAGE 66

66 Table 3 1. Fundamental properties of phosphorescent dendrimer. Molecular Formula Molecular Weight TGA(5 %) Tg C135H156IrN3O6 2108.9 g mol 1 400 C 132 C Table 3 2. Fundamental properties of fluorescent dendrime r. Molecular Formula Molecular Weight Melting Point C162H183N 2144.2 g mol1 238 239 C

PAGE 67

67 Table 3 3 Fundamental properties of matrix solvents.93 Property Chloroform Tolue ne Molecular formula CHCl3 C7H8 Molecular weight (M) (g mol-1) 119.4 92.1 -3) (l.s.) 1.479 (298 K) 0.867 (293 K) -3) (s.s.) 1.978 (295 K, 0.6 GPa)114 2.960 (301 K, 0.9 GPa)115-116 NV (molecules cm-3) 7.46 x 1021 (298 K) 5.67 x 1021 (293 K) Molar volume (Vm = M -1) (cm3 mol-1) 80.7 (298 K) 106.3 (293 K) Molar concentration (mol L-1) 12.4 (298 K) 9.41 (293 K) Melting point ( Tsl) (K) 209.7 178.2 Normal boiling point ( Tlv) (K) (at 1 atm) 334.3 383.8 Vapor pressure (Torr) (at 298 K) 196.5 28.43 Critical tempe rature (K) 536.4 591.8 Critical pressure (Torr) 41.0 x 103 (5.47 G Pa) 30.8 x 103 (4.11 G Pa) Critical volume (cm3 mol-1) 239 316 Specific heat capacity ( cp) (J g-1 K-1) (l.s.) 0.885 1.707 Specific heat capacity ( cp) (J g-1 K-1) (s.s.) 0.920 (240 K, 760 Torr)117 0.640 (78 K)115 fus ( Tsl) (J mol-1) 9.5 x 103 6.64 x 103 vap ( Tlv) (J mol-1) 29.24 x 103 33.18 x 103 vap (298 K) (J mol-1) 31.28 x 103 38.01 x 103 sub ( Tsv) (J mol-1) 31.28 x 103 38.01 x 103 Thermal conductivity ( kT) (W cm-1 K-1) (l.s.) 1.17 x 10-3 1.31 x 10-3 Thermal conductivity ( kT) (W cm-1 K-1) (s.s.) 7 x 10-3 (77 K)118-120 2.51 x 10-3(301 K, 0.9 GPa)116 T = kT -1 cp -1) (cm2 s-1) (l.s) 8.93 x 10-4 8.85 x 10-4 T = kT -1 cp -1) (cm2 s-1) (s.s) 3.82 x 10-3 1.32 x 10-3 Cubic expansion coefficient (K-1) (293 K) 1.21 x 10-3 1.05 x 10-3 Speed of sound ( US) (cm s-1) (l.s.) 9.87 x 105 1.31 x 106, 121-123 Speed of sound ( US) (cm s-1) (s.s.) 2.23 x 106, 2.62 x 106, 124* Surface tension ( ) (N m-1) (293 K) 26.5 x 10-3, 125-127 28.4 x 10-3, 128 Flash point (K) None 277.2 Autoignition temperature (K) None 753.2 Di 1.04 0.37 l.s.: liquid state; s.s.: solid state; *Using Raos rule (described in Section 7. 3.2 ).

PAGE 68

68 Table 3 4. Radii of spot sizes for the FEL beam tuned to the wavelengths of interest. CHCl3 C6H5CH3 FEL wavelength 3.3 2 m 8.22 m 3.3 1 m 6.23 m Spot Size Radius (m) 7 0 30 225 25 70 3 0 175 25

PAGE 69

69 CHAPTER 4 OPTICAL CONSTANTS OF FROZEN SOLVENT MATRI CES 4.1 Introduction Three main material properties should be considered relevant to the depositions rates from RIM -PLA and the r esulting film and target c haracteristics: (i) the absorption coefficient spectra of the resonance mode; (ii) its lifetime and (iii) the efficiency of FEL light coupling to the mode. In this chapter we describe the results of measurements of these propert ies 4.2 Light Absorption by Dendrimer Solutions Whe n a beam of light irradiates a homogeneous dendrimer solution in a cuvette, as shown in Figure 4 1 a part of the incident light is absorbed by the sample, while a small portion is reflected, and the rest is transmitted, that is, r t a oI I I I (4 1) where Io is the intensity of the incident light and Ia, It and Ir denote the intensities of the absorbed, transmitted, and reflected light respectively (Figure 4 1) .20, 129 In practice, the materials chosen for the nonabsorbing windows may absorb and reflect a considerable amount of light. For instance the ZnSe windows used in the measurements to be discussed have ~70% transmission in the mid infrared spectrum (0.6 21 m ). The absorption of light is defined on the basis of the Beer -Lambert law. The law considers the light a bsorbed by a homogeneous medium when a parallel beam of monochromatic radiation is incident on it, and it states that the rate of change with the thickness of the sample medium is directly proportional to the intensity That is, x xI dx dI (4 2)

PAGE 70

70 where Ix is the intensity of the light at a point in the medium and x is the path length of the light medium up to that poi nt. By integrating Equation 4 2 for the entire path length ( l ) from x = 0 to x = l we obtain: l I Io ln (4 3) where is the characteristic absorption coefficient of the medium at that wavelength of light, Io is the intensity of the incident light at x = 0 and I is the intensity of the transmitted light at x = l Experimentally, the ratio I / Io is known as the transmittance ( T ) of the sample. In terms of transmittance, Equation 4 3 may be rewritten as: l I I To ln ln (4 4) According to Beer -Lamberts law, the amount of light absorbed is directly proportional to the number of molecules of the absorbing species, that is, the concentration of the sample in the path of the light. A similar expression as Equation 4 4 may be written for t he dependence of transmittance on the concentration C of the solution: l C I I To ln ln (4 5) In Equation 4 5 is the molar absorptivity, which in the older literature is referred to as the molar extinction coeffi cient of the sample. Now, if we co nsider the irradiation of the surface of a solvent target with a broad laser beam, the effective irradiance ( I, W/cm2) (or conversely fluence, ) varies as a function of depth, i.e. opticalz z k I z Iexp ) ( ) (0 (4 6 )

PAGE 71

71 which follows from the Beer -Lambert law.12, 130 In Equation 4 6 I is the irradiance at z and the parameter k (z ) accounts for non linear effect s (e.g. reflection losses at the surface, changes in index of refraction through the sample leading to back -scatter ed irradiance etc. ). The quanti t y optical (cm) is the optical penetration depth, and is defined as the depth at which the light has been attenuated by 1/ e i.e. to about 37% of it s initial value at the surface of the target. The optica l penetration depth is defined by convention as the i nverse of the absorption coefficient ,42, 131 e.g. obtained from transmittance data (Figure 4 2). We assume that optical scattering, i.e. spatial variations of the refractive index within the target, is negligible (i.e. k (z ) = 1, in Equation 4 6) In the case that optical scattering was significant, optical 42, 131 The relevance of the optical penetration depth is that it can be used to estimate the irradiated (affected) focal volumes at the selected resonance modes The f ocal volume in turn, defines the number of dendrimer molecules exposed to the energy of the laser beam. With respect to target characterization (Chapter 5 ), the affected focal volume will be used to determine if the number of exposed molecules is large enough to de te ct molecular alterations by Nuclear Magnetic Resonance spectroscopy. With respect to film characterization (Chapter 6) the quantity will be used to compare the deposited volumetric energy densities to the ablation mechanisms seen in the film. More imp ortant, optical is needed to calculate two import ant parameters, the thermal confinement time (t) and the stress confinement time (s) as wi ll be defined further in Chapter 7 (Section 7.3) 4.3 Absorption Coefficients, Optical Penetration Depths and Life times of Resonance Modes 4.3.1 Literature Survey Table 4 1 lists the selected resonance modes for the solvents used in this study. Several groups have measured some of the infrared optical constants of the solvents used in this work,132 136 mainly refractive index and optical dispersion. However, t he mid infrared absorption

PAGE 72

72 coefficient or molar absorptivity data have seldom been reported for toluene and none reported for chloroform at the wavelengths of interest fo r this work .137 139 Bertie, et al. compiled infrared transmission measurements for toluene that were recorded on four different instruments ( three different manufacturers ) by six separate spectroscopists .140 They reported averages of the peak intensities of the molar extinction coefficients at 25 C, citing estimated errors of 2.5%, with baseline values accurate to between 3 a nd 10% (Table 4 2). We converted these values to the absorption coefficients with the molar concentration of the pure solvent (Table 3 3) In a separate work, Anderson measured completely different values for the same solvent toluene (Table 4 3 ).141 In this work we ablate solid solvent matric es which are frozen at cryogenic temperatures (~77 K). Considering that no absorption or molar extinction coefficient data h ave been reported at these temperatures for either solvent (nor for chloroform at room temperature) we decided to measure the abs orption coeffic ients of both solvents Transmittance spectra were measured at room temperature (298 K) and at cryogenic temperatures (~77 K, boiling temperature of nitrogen at 1 atm ) in order to establish any possible changes to the absorption of the sele cted resonance modes. 4.3.2 Experimental Me thod The demountable liquid cell and the FTIR spectrometer used for the transmission measurements have been previously described in S ection 3.2 In a typical procedure a PTFE spacer ( dspacer = 15, 25 m) was plac ed between Zn Se crystals, while o rings created a tight seal upon closing the cell. A reference spectrum was obtained with an empty cell. T he fringe pattern in the spectrum resulting from the different indexes of refraction was used to calculate the exact path length (Equation 4 7 ).

PAGE 73

73 ) ( 21 cm n m dfringes vac real (4 7 ) In Equation 4 7 dreal is the real separation between the ZnSe crystal disks, m is the number of fringes, nvac is the index of refraction of vacuum (i.e. 1) and fringes is the wavenumber differe nce between the selected peaks of the fringe pattern. In order to obtain a smooth reference, a software package ( Fourier Transform Smooth (FTS) Charles Porter) was used to remove the fringe patterns. The program performs a Fourier transformat ion on the spectra. The fringes (noise) have a sinusoidal pattern in the spectrum, and can be easily identified as a high amplitude aberration at the high frequency part of the transform. The transform can be expanded and the fringe pattern can be removed without affecting the rest of the spectrum. The equations described in Section 4.2 were used to determine the absorption coefficient and optical penetration depth from the data For room temperature m easurements, the cell was filled with the appropriate solvent or dendrimer solution (0.5 wt. %) and placed inside a dry nitrogen purged chamber. A spectrum was acquired. If there was no intensity (i.e. full absorption) at the absorption mode of the solvent the cell was removed, cleaned and a thinner spacer w as use d For measurements at cryogenic temperatures, the clean transmission cell was filled with the sol ution and sealed The entire cell was then rapidly submerged into liquid nitrogen and allowed to equilibrate for a minimum of five minutes. The cell would t hen be quickly removed from the dewar and placed inside the FTIR chamber. The chamber was evacuated to ~1 Torr a nd the data were collected All data used to determine the absorption coefficients are the results of a minimum scan time, typically ~2 minute s after the removal of the cell from the dewar. Since the absorption coefficient is a property dependent on temperature it changes as the cell warms Therefore, any inclusion of post 2 minute spectra would lead to measurement error as the cell is

PAGE 74

74 not a perfect insulator To establish the amount of experimental error, a t least three spectra were obtained for each condition i.e. no dendrimer, phosphoresc ent and fluorescent dendrimer. Each sample intensity spectrum was divided (as per the B eer L ambert law ) by the data of a reference (empty cell) collected under the same conditions, i.e. the earliest scan collected about 2 minutes after removal from the dewar. All the a verage spectra shown in this chapter is intact of any baseline manipulation with the acq uisition software 4.3.3 Absorption by Water / Atmospheric Mo i sture Condensation None of the solvent modes studied in this work overlapped with a ny water vapor absorption spectral line (Figure 4 3) The 1605 cm1 (6.23 m) mode of toluene (aromatic C=C st retch) is the cl osest to a high water absorption region, but occurs in -between two large groups of H2O spectral lines. U pon cell freezing however, the transmittance through an empty cell (reference) decreased about 50%, most likely from scattering of the light by ice crystals frosting the surface of the ZnSe crystals Since the reference used for each transmission calculation had about the same amount of frost (constant room humidity, data acquired at same time post removal from dewar), the error associa ted with ice scattering is incorporated in the standard deviation. 4.3.4 Spectral Data Fittings T he Cauchy Lorentz (Lorentzian ) distribution was used for fitting the absorption data, as it is the solution for the differential equation describing forced res onance and describes the homogeneous broadening seen in the spectral absorption lines.142 4.3.5 Lifetime s of Resonance Modes T he redistribution of the vibrational energy of excited polyatomic molecules in condensed states occurs typically through the loss of energy from the specific vibrational mode to some or all the other mechanical degrees of free dom (the phonon bath) in the molecule.143 144 This

PAGE 75

75 process ordinarily occurs on a 1012 109 s time -scale, also known as the modes lifetime (T1).143, 145 The lifetime of a vibrational mode is inversely related to the full width at half maximum (FWHM) of the spectral peak (Equation 4 8).142, 146147 FWHM c T 2 11 (4 8) Previous work by Bubb, et al.51 suggested a lifetime -dependence on the deposition rates from resonant ablation of solid polystyrene targets. It was proposed that for modes with similar ab sorption coefficients, the longe r li fetimes (lower spectral widths) resulted in higher yields due to a higher probability of disrupting the relatively weak Van der Waals bonds. However, the authors also stated that because of the relatively large spectral bandwidth of the FEL other non lin ear processes may be occurring that could affect the depo sition rates The widths of vibrational modes will be further used in the discussion to determine any correlation to the deposition rates in RIM PLA. 4.3.6 Results 4.3.6.1 Liquid c hloroform (298 K ) Figure 4 4 shows the absorption coefficient spectra for liquid chloroform and the frozen matrices of dendrimer solution Fig ure 4 5 shows the optical penetration depth data T able s 4 4 and 4 5 tabulate the average va lues and standard deviation s for the t wo resonant modes T he maxima for the resonance s are single peaks at 3.31 and 8.22 m for the C H alkyl stretch and the C H bending modes, respectively. It is evident that the absorbance does not change significantly with the addition of dendrimer (0.5 w t. %) to the matrices. T he average change of the coefficient is < 10%. In the liquid solvent, the absorption of the C H bending mode is ~6.5 times that of the C H stretch mode. Correspondingly, the optical penetration depth for the C -H stretch mode is m uch larger than the C -H bending mode

PAGE 76

76 4.3.6.2 Solid c hloroform Figure s 4 6 and 4 7 show the absorption coefficient spectra and optical penetration depths for frozen chloroform and the frozen dendrimer matrices The average values and standard deviations a re tabulated in Tables 4 6 and 4 7 With respect to the frozen chloroform absorption coefficient spectra: 1 Similar to the liquid solvent results, the absorption coefficients do not vary from pure solvent to frozen matrices of dendrimer solution All av era ge values are within 10% of one an other. 2 T he C -H st retch resonance shifted to lower energy upon freezing. This change has been previously observed138, is repeatable and its magnitude is four times the scan resolution (2 cm1) The absorption coefficient for this mode doubled from the room temperature average value, which has been attributed to weaker hydrogen bonding in the solid.137 3 The C -H bending resonance splits into two sepa rate peaks upon solidification. The splitting has previously been observed138, 148 and is a result of crystallization upon freezing. Chloroform may crystallize as two different phases, and .114, 148 149 The phase crystallizes at 185 K into an orthorhombic unit cell with a space group Pnma (i.e. a p rimitive lattice with a n glide perpendicular to the a axis, a mirror plane perpendicular to the b axis and a a -glide perpendicular to the c axis ) with four molecules per unit cell (Figure 4 8 ).138, 148 The (hexagonal lattice, 2 atoms/unit cell, P63 space group) phase has only been observed at high pressures, therefo re is not considere d relevant to our study.114 4 The splitting is a consequence of a loss of degeneracy from lower site symmetry in the crystal ( E species ).2, 138139, 148, 150 That is, in a crystal, surrounding atoms exert additional deformational force s compared to the liquid state resulting in vibrations with different phases between molecules. The magnitude of the splitting is depend ent on the temperature of the solvent. Note that chloroform may also form a glass; its glass transition temperature (Tg) has been reported as 114 and 109.5 K.151152 The glassy state is not considered in this work s ince the crystal forms prior to reaching the Tg and the measured spectra is representative of this phase .138, 148 Considering that the spectra does not change from pure solvent to dendrimer matrix, it i s reasonable to conclude that the polymer molecules do not prevent the formation of the crystal. 2 The chloroform molecule is represented by the C3v point group, i.e. a cyclic symmetry group with 3fold rotation and a mirror plane parallel to the axis of rotation. Every point group has different irreducible representations with differe nt symmetry operations. Chloroform has three of these symmetry representations: A1, A2 and E ; while only the A1 and A2 species result in mid IR resonance. The E species are responsible for the C H bending resonance. See: F. A. Cotton, Chemical applicati ons of group theory 3rd ed. (Wiley, New York, 1990).

PAGE 77

77 Figure 4 9 shows the evolution of the C -H bending mode from spectra acquired at different times after removal fro m the dewar i.e. at different temperatures T he f igure also includes a plot of the relative integrated areas of the Lorentzian fits for each resonance mode of chloroform at different temperatures Table 4 8 includes the results for all Lorentzian fits (with baseline corrections) for the solvents average measurements. Although the C -H stretch resonance area stays almost constant upon freezing the C H bending mode displays an almost exponential dependence on temperature. Based on the results from Table 4 8, the absorption coefficient of the C H bending mode decreased significantly as a result of freezing, while its lifetime increases by 30%. Conversely, the C -H stretch absorption coefficient for the frozen solvent doubled while its lifetime tripled. 4. 3.6.3 Liquid toluene (298 K ) Figure 4 10 shows the absorption coefficient spectra for pure liquid toluene and frozen matrices of dendrimer solutions Figure 4 1 1 shows the optical penetration depth data Tables 4 9 and 4 10 tabulate the average values an d standard deviations f or both resonance modes T he maxima for the resonance s are single peaks at 3.30 and 6.23 m for the C H stretch and the C=C aromatic stretch modes, respectively. In the liquid state, the absorption coefficients of the C -H stretch m ode of toluene is ~12% higher than that of the C=C aromatic stretch. T he absorption coefficients change less than 4% as a result of the addition of dendrimer (0.5 wt. %) to toluene. Note that all the spectral lines surrounding the C=C aromatic stretch mo de arise from water vapor absorption and not from toluene 4.3.6.4 Solid toluene Figure s 4 1 2 and 4 13 show the absorption coefficient spectra and optical penetration depth data, respectively, for frozen toluene and the dendrimer matric es. Tables 4 11 and 4 12

PAGE 78

78 tabulate the average values and st andard deviations for both modes With respect to the frozen toluene absorption coefficient spectra: 1 T he absorption coefficients do not vary from pure solvent to dendrimer m atrices after considering the standard dev iations 2 The maximum absorption coefficient for the C H stretch mode decreased by ~40% and peak shifted to lower energy by 2 cm1 (same as the scan resolution). Although the shift was repeatable, no reports exist in the literature for this behavior. Sinc e there is n o hydrogen bonding in toluene, the change must be related to other interactions in the glassy or crystal line structure of toluene. Toluene is one of the most common organic liquids known to supercool easily to the point of vitrification ; its g lass transition temperature (Tg) from calorimetry measurements has been previously reported to be 117 K.115, 153 The symmetry of the -toluene cryst al has been previously reported (165 K) to have a space group of P 21/c i.e. a simple monoclinic lattice with parameters a = 7.666 b = 5.832 c = 28.980 and = 105.73.154 There are 8 molecules per unit cell. Altho ugh the crystal is stable prior to reaching the Tg, the high degree of supercooling inherent in the measurement technique is expected to lead to a glassy state 3 The maximum absorption coefficient for the C=C aromatic stretch mode decreased by ~20% and the center peak did not shift upon freezing. Table 4 13 lists the results from the Lorentzian fits to the toluene data. Figure 4 1 4 shows the fits for the frozen solvent. 4.4 Light Coupling Efficiency Figure 4 1 5 shows the normalized spectral intensity for the FEL when tuned to the selected resonance wavelengths. The spectra was fitted to Gaussian distributions (not shown) as the laser was operated in its fundamental (TEM00) mode ; Table 4 1 4 tabulates the spectral bandwidth (FWHM) from these curve fits F rom Table 4 14 it is evident that the FEL spectral distribution widens as the laser wa s tuned to higher wavelengths. The results are important as the efficiency of light absorption is dependent on the overlap (coupling) of the absorption peak with the FEL waveleng th distribution i.e. the efficiency to achieve resonance. Table 4 1 5 lists a comparison of the ratios of the FWH M of the resonance peaks to those of the FEL spectral distributions i.e. laser -mode

PAGE 79

79 coupling efficiency Note that both peaks of the split C H bending resonance of chloroform in the solid state were considered in the respective calculation. These results will be used to determine effective volumetric energy densities when assessing the ablations characteristics for each mode (Section 7.9). 4. 5 Affected Focal Volumes Based on the average FEL spot sizes (Section 3.4.1.2) and optical penetration depths at the modes maxima, the affec ted focal volumes at the surface of irradiated targets were calculated Table 4 1 6 tabulates the results of the ca lculations (beam focal spot area X optical penetration depth). Table 4 17 lists the estimated material amounts (volume %) in the targets affected by FEL irradiation. The calculations assumed that: (i) the spiral length from the FEL rastering profile (Sect ion 3.4.2.1) covers 50% of the target surface area; (ii) the beam is rastered to a maximum radii of 1 cm; and (iii) the dimensions of the target well described in Section 3.4.2.3 (1.25 cm radius 2 mm depth ). As the target was ablated, the affected vol ume percentages increased The results from Table 4 17 indicate that the amount of target material affected by the FEL beam is minimal ( less than 1% for all modes ) H owever, these calculation s assume a n FEL beam with a line spectral pro file tuned to each mode maximum, which is not the case (S ection 4.4 ). T he optical penetration depth s for non resonant wavelengths (experimentally ~ 40 80% of the FEL distribution based on Table 415) may be as high as 2 00 300 m Consequently, the FEL affected volumes at n on resonant wavelengths could be 1530 times those tabulated in Table 4 17. Therefore, the affected volumes from FEL irradiation could be as high as 5 7%. 4.6 Conclusions T he absorption coefficients of selected vibratio nal modes in room -temperature and fr ozen (solid at ~77 K ) sol vents have been measured T he absorption coefficients were found to not

PAGE 80

80 change upon addition of small amount of dendrimer (0.5 % by weight). With respect to the solid -state chloroform resonance modes: 1 The peak absorption c oefficie nt for the C H alkyl stretch mode was 15% higher than that of the C -H bending mode. 2 The C -H stretch mode has a ~11 times longer lifetime of the C -H bending mode based upon the mode width. 3 The C -H bending mode was 200% more efficient than the C -H stretch mo de at coupling with the FEL laser beam. With respect to the solid -state toluene resonance modes: 1 The maximum coefficient for C=C aromatic stretch mode wa s 150% higher than that of the C H alkyl stretch mode. 2 The C=C aromat ic stretch mode has a 42% longer l ifetime than the C H stretch mode. 3 The C -H stretch mode was 300% more efficient at coupling with the FEL laser beam.

PAGE 81

81 Figure 4 1. Absorption, reflection and transmission of light through a liquid solution. Figure 4 2. Optical penetration depth ( tical Figure 4 3. Mid -IR atmospheric absorption spectrum and wavelengths of interest.

PAGE 82

82 Figure 4 4. Absorption coefficient spectra for liquid chloroform and dendrimer matrices. Figure 4 5. Optical pene tration depths for liquid chloroform and dendrimer matrices. Figure 4 6. Absorption coefficient spectra for frozen chloroform and dendrimer matrices.

PAGE 83

83 Figure 4 7. Optical penetration depths for frozen chloroform and dendrimer matrices. Figure 4 8 Crystal structure ( phase) of solid chloroform (x is the polarity axis).148 Figure 4 9. Chloroform spectra: a) Evolution of absorption coefficient spectra (Lorentzian fits) of chloroform as the temperature of the transmission cell rose. b) Comparison of integrated areas under fits for both chloroform modes.

PAGE 84

84 Figure 4 10. Absorption coefficient spectra for liquid toluene and dendrimer matrices. Figure 4 11. Optical penetration depths for liquid toluene and dendrimer matrices. Figure 4 12. Absorption coefficient spectra for frozen toluene and dendrimer matrices.

PAGE 85

85 Figure 4 13. Optical penetration depths for frozen toluene and dendrimer m atrices. Figure 4 14. Lorentzian fits for frozen toluene absorption coefficient data. Figure 4 15. FEL wavelength distributions at the selected wavelengths.

PAGE 86

86 Table 4 1. Selected resonances for each solvent used as a matrix. Chloroform Toluene Resonance mode 1 CH alkyl stretch (3.3 m) CH stretch (3.3 m) Resonance mode 2 CH bending (8.2 m) C=C aromatic stretch (6.2 m) Table 4 2. Averaged infrared optical constants for toluene from Bertie, et al.140 (C = 9.41 mol L1 at 25 C). Frequency of interest (cm1) Molar extinction coefficient (L mol1 cm1) Absorption coefficient = C (cm1) 3026.9 54.3 511.0 1604.5 28.9 271.9 Table 4 3. Mid -infrared absorption coefficients for toluene from And er son et al.141 Frequency of interest (cm1) Absorption coefficient, (cm 1) 3026.6 364.9 1606.7 100.7 Table 4 4. Absorption coefficients for modes of liquid chloroform and dendrimer matrices. Wavelength (m) Wavenumber (cm1) Solvent 1) Solvent Std. Dev. Phosph. 1) Phosph. Std. Dev. Fluo. 1) Fluo. S td. Dev 3.31 3021 485 7 445 5 459 4 8.22 1217 3148 263 2999 273 3035 450 Table 4 5. Optical penetration depths for modes of liquid chloroform and dendrimer matrices. Wavelength (m) Wavenumber (cm1) Solvent opt. (m) Solvent Std. Dev. Phosph. opt. (m) Phosph. Std. Dev. Fluo. opt (m) Fluo. Std. Dev 3.31 3021 20.6 0.3 22.5 0.2 21.8 0.2 8.22 1217 3.2 0.3 3.4 0.3 3.3 0.5 Table 4 6. Absorption coefficients for modes of frozen chloroform and dendrimer matrices. Wavelength (m) Wavenumber (cm1) So lvent 1) Solvent Std. Dev. Phosph. 1) Phosph. Std. Dev. Fluo. 1) Fluo. Std. Dev 3.32 3013 980 182 945 102 901 85 8.18 1223 554 53 539 42 491 16 8.28 1208 842 98 771 72 814 72

PAGE 87

87 Table 4 7. Optical penetration depths for modes of frozen chloroform and dendrimer matrices. Wavelength (m) Wavenumber (cm1) Solvent opt. (m) Solvent Std. Dev. Phosph. opt. (m) Phosph. Std. Dev. Fluo. opt (m) Fluo. Std. Dev 3.32 3013 10.4 1.9 10.7 1.2 11.1 1.2 8.18 1223 18.1 1.7 18.6 1.4 20.4 0.7 8.28 1208 12.0 1.4 13 1.3 12.3 1.1 Table 4 8. Results from Lorentzian fits of chloroform spectra. Mode Time Post Freeze (min) Peak Center (m) Intensity (cm1) FWHM (m) Integrated Area (a.u.) Relative Area (%) C H stretch 2 min 3.32 896 0.005 7 100 298 K 3.31 438 0.015 10 100 C H bending 2 min 8.18 499 0.023 18 33.6 8.28 777 0.030 36 66.4 3 min 8.19 993 0.020 31 30.5 8.28 1152 0.040 70 69.5 4 min 8.21 1914 0.047 135 69.7 8.26 1844 0.021 59 30.3 298 K 8.22 3374 0.069 346 100 Table 4 9. Absorption coefficients for modes of liquid toluene and dendrimer matrices. Wavelength (m) Wavenumber (cm1) Solvent 1) Solvent Std. Dev. Phosph. 1) Phosph. Std. Dev. Fluo. 1) Fluo. Std. Dev 3.3 0 3027 811 4 783 4 780 3 6.23 1604 698 1 712 1 709 3 Table 4 10. Optical penetration depths for modes of liquid toluene and dendrimer matrices. Wavelength (m) Wavenumber (cm1) Solvent opt. (m) Solvent Std. Dev. Phosph. opt. (m) Phosph. Std. Dev. Fluo. opt (m) Fluo. Std. Dev 3.3 0 3027 12.3 0.1 12.8 0.1 12.8 0.1 6.23 1604 14.3 0.1 14 0.1 14.1 0.1 Table 4 11. Absorption coefficients for modes of frozen toluene and dendrimer matrices. Wavelength (m) Wavenumber (cm1) Solvent 1) Solvent Std. Dev. Phosph. 1) Phosph. Std. Dev Fluo. 1) Fluo. Std. Dev 3.31 3025 488 40 476 39 491 70 6.23 1604 965 38 846 118 890 145

PAGE 88

88 Table 4 12. Optical penetration depths for modes of frozen toluene and dendrimer matrices. Wavelength (m) Wavenumber (cm1) Solvent opt. (m) Solvent St d. Dev. Phosph. opt. (m) Phosph. Std. Dev. Fluo. opt (m) Fluo. Std. Dev 3.31 3025 20.6 1.7 21.1 1.7 20.7 2.6 6.23 1604 10.4 0.4 12 1.7 11.5 1.8 Table 4 13. Results from Lorentzian fits of toluene spectra. Mode Time Post Freeze (min) Peak Center (m) Intensity (cm1) FWHM (m) Integrated Area (a.u.) CH stretch 2 min 3.31 390 0.026 15 298 K 3.30 701 0.024 27 C=C stretch 2 min 6.23 976 0.015 23 298 K 6.23 652 0.023 23 Table 4 14. FWHM results from Gaussian fits of FEL wavelength distribution. center 3.31 m 6.23 m 8.22 m FWHM (m) 0.044 0.101 0.150 Table 4 15. Laser -mode coupling efficiency for selected solvent resonance peaks. Chloroform Toluene CH stretch CH bending CH stretch C=C stretch Liquid state 0.341 0.153 0.545 0.223 So lid state 0.114 0.353 0.590 0.149 Table 4 16. Average FEL irradiated focal volumes for the selected resonance modes. Focal volume (m m3) Chloroform Toluene CH stretch CH bending CH stretch C=C stretch Liquid state 3.24 0.50 1.93 2.25 Solid state 1 .63 2.84 3.24 1.63 Table 4 17. Estimated volume percentages of the ablated matrices affected by FEL irradiation at resonant modes. Affected volume (%) Chloroform Toluene CH stretch CH bending CH stretch C=C stretch Liquid state 0. 52 0.08 0.31 0.36 Solid state 0.26 0.45 0.51 0.26

PAGE 89

89 CHAPTER 5 TARGET CHARACTERIZAT ION 5.1 Introduction Previous research on resonant infrared pulsed laser ablation has focused on the characterization of the deposited films.9 11 However, no study has thoroughly examined the molecular structure changes at the ablated target. The dendrimers used in this work are ideal for looking at this issue as they are large enough to be considered polymers, but small enough to offer good r esolu tion with practical organic characterization techniques. Th e following chapter details the characterization of the molecular structure of the dendrimer retained in the ablated target s. It contains two major sections the first of which discusses targets a blated at the same fluence (1 J/cm2) for all the studied resonance modes. The second section discusses targets ablated at increasing fluence (10, 20 and 30 J/cm2) for one of the resonance modes in each solvent as the FEL is tuned to the same wavelength (3.31 m). Upon melting, the target solutions were collected and characterized via 1H NMR spectroscopy, FTIR spectroscopy and MALDI TOF spectrometry. These three techniques are used to determine if any degradation results from laser ablation. In general: ( i) NMR data will give information about relative changes in t he number of protons and if protons with different electronic environment are created ; (ii) FTIR data will tell if functional groups have been retained and/or created in the molecule ; and (iii) M ALDI TOF will confirm the presence of intact dendrimer and possib ly other predominant fragments. Due to the nature of the RIM -PLA process, it is not possible to characterize only the targets surface thus the presented data are for the bulk of the target s As determined in Section 4.5 the affected volumes from FEL irradiation could be as high as 5 7% (for nonresonant excitation) All of these techniques, particularly NMR, will detect molecular alterations for these concentrations.

PAGE 90

90 5.2 I ntroduction to NMR Spectroscopy Similar to an electrons spin quantum number (ms = from Paulis exclusion principle), atomic nuclei have an associated magnetic spin quantum number ( I ).155 156 Nuclear magnetic resonance (NMR) spectroscopy allows for the detection of nuclei with I > 0, such as 1 1H, 19 9F 13 6C, and 17 8O In a typical proton (1H ) NMR experiment, a solution of the organic sample in a uniform 5 mm glass tube is oriented between the poles of a powerful magnet while it is spun to average any magnetic field variations and tube imperfections. A radio frequency (rf) coil irradiates the sample at the spectrometers frequency while an applied external magnetic field (Bo) is varied. The electronic acquisition system of th e spectrometer observes the absorbed rf signal from the sample, while the external magnetic field is swept over a small range; an NMR spectrum is generated. Further details on the spin -state population s their corresponding energy differences and the fund amental NMR equation can be found elsewhere.155 156 An NMR spectrum is the combination of broadened lines (singlets, doublets, triplets, etc) which give important information for the orga nic molecule, including: (i) n umber of protons present, (ii) the presence of aromatic moieties and their substitutions ( ortho-, meta -, para-), (iii) t ypes of alkyl substituents and respective locations (iv) configurational isomers present and ( v) dynamic processes in the molecule by varying the sample temperature. The response of a proton and consequently its appearance in a spectra is dependent on the surrounding electronic environment of the nuclei.155156 Since electrons are charged part icles, they move in response to the external magnetic field and generate a secondary field that opposes the much stronger applied field. This secondary field shields the nucleus from the applied field, so Bo must be increased in order to achieve resonance (absorption of rf energy). Since both this frequency shift and the fundamental resonant frequency are directly proportional to the strength of the magnetic field, the shift is converted into a field -independent dimensionless value known as the chemical shift

PAGE 91

91 The chemical shift in 1H NMR spectra is reported relative to a reference compound (TMS, tetramethylsilane) with a chemical shift of 0 ppm (parts per million). The chemical shift is calculated by subtracting the frequency of the reference from the frequency of the signal and dividing by the resonance rf fr equency in the spectrometer (e.g. 300 MHz). The relative location of several organic moieties in the chemical shift s pectrum is shown in Figure 5 1.155 156 As the figure shows deshielded (unsaturated) nuclei appear downfield (l ower chemical shift ) while shielded (saturated) ones appear upfield (higher chemical shift) The splitt ing of a particular proton signal is related to J coupling, which arises from the interaction of different spin states through the chemical bonds in the molecule.155156 The spacing between the lines of a doublet, triplet or quartet is called the J coupling constant. It is given the symbol J (Hz) The magnitude of the coupling constant can be calculated by multiplying the separation of the lines in ppm units by the resonance frequency of the spectrometer (e.g. 300 MHz) 5.3 Peak Assignments for NMR Spectra of Dendrimers Due to the quantity and complexity of the NMR signals from these macromolecules, the rea der may use the discussion in the following sections to understand the location of the protons in the molecule relative to the spectras chemical shift. J coupling constants are not included but are available in the original publications detailing the synthesis89 90, or may be calculated from the listed data 5.3.1 Phospho rescent Dendrimer Figure 5 1 shows the molecular structure of the first generation phosphorescent dendrimer. All the protons have been assigned a letter, which is preceded by the number of equivalent hydrogens in that location. Due to the three dendrons symmetry and equivale nce, the letters for

PAGE 92

92 the protons repeat. The signals in the obtained NM R spectra were numbered. Their correlation to the labeled structure follows in Table 5 1 5.3.2 Fluorescent Dendrimer Similar to the previous analysis, Figure 5 3 shows the molecular structure of the first generation fluore scent dendrimer. All the protons were assigned a letter preceded by the number of equivalent hydrogens in that location. The signals in the obtained NMR spectra were numbered and the correlatio n to the labeled structure follows in Table 5 2. Table 5 2 de tails the expected chemical shift for each proton signal. However, a s the NMR spectra will show, there is little resolution of the downfield peaks (7.027.72 ppm). T herefore, that region in all the fluorescent dendrimer NMR spectra could only b e compared as a whole. Figure 5 4 shows an expansion of thi s region in the control spectra. 5.4 Resonance Assignments for FTIR Spectra of Dendrimers Similar to NMR, the mid -infrared spectra of the dendrim ers is busy and complex. In the following sections, the FTIR spectra of the control molecules are presented followed by tables listing the most predominant resonance assignments.157158 The se s pectra will be used as a comparison basis for those of the ablated targets. Note that the target FTIR data is qualitative (absorbance scale) as it only shows the retention and creation of functional groups within the molecule, but little i nformation can be obtained about the level of degradation post ablation. 5.4.1 Phosphorescent Dendrimer Figure 5 5 shows the mid infrared absorption spectrum of the first generation phos phorescent dendrimer. Table 5 3 lists the most relevant absorption frequencies with their corresponding r esonance assignment.

PAGE 93

93 5.4.2 Fluorescent Dendrimer Figure 5 6 shows the mid infrared absorption spectrum of the first generation flu orescent dendrimer. Table 5 4 lists the most relevant absorption frequencies with their corresponding resonance assignment. Note that most tert -butyl group compounds have three moderate to strong absorption bands in the re gion 29902930 cm1 (3.343.41 m ) due to asymmetric stretching vibrations. The symmetric stretching vibrations occur in the region 29502850 cm1 (3.393.51 m ) wi th aromatic t -butyl absorbing in the region 2915 2860 cm1 (3.443.50 m ).157 5.5 Temporal Stability of Dendrimers The depositions discussed in this work were carried out at different dates within the time frame of two years. Therefore, it was neces sary to test the temporal stability of the molecu les to properly isolate any degradation effects from laser ablation. 5.5.1 Phosphorescent Dendrimer Figure 5 7 shows a comparison of the 1H NMR spectra of the phosphorescent dendrimer control after it was re ceived and the same sample 15 months later. The integrated areas under the peaks are compared in order to establish any possible degradation of the molecule over time. These areas represent the number of protons for t he respective signal. Table 5 5 show s the theoretically expected number of protons for the labeled peaks and their relative ratios to an arbitrary normal peak. In the case of the phosphorescent dendrimer, the normal corresponds to th e multiplet for the methylene ( -CH2) groups from the 2 -et hylhexyloxy groups at the dendrimer s periphery. This signal (1.25 1.62 ppm) contains 48 protons. Table 5 5 shows the calculated ratios for the measured control samples at different times throughout the research. Significant differences are seen between the theoretically expected and the measured proton counts, particularly for the shielded protons. For instance, there is over a 2 0% difference in the measured proton count for signal 7 (protons labeled c in Figure 5 2 )

PAGE 94

94 relative to the expected count Thi s should not be thought of as less t -butyl groups present. Initial consideration of this difference suggested a limitation of macromolecular NMR where larger molecules have slower tumbling rates and shorter signal relaxation times.155 This could cause the peaks to become broader and/or weaker. However, upon further examination, the variations from the expected are most likely the result of poor shimming prior to NMR data acquisition .155 Shimming refers to the adjustment of the currents in the shim coils surrounding the NMR sample, so as to cancel as much as possible any magnetic field g radients in the liquid Poor shimming results in lack of symmetry in the peaks, usually altering the integrated areas under the peaks, i.e. the proton counts. Shimming issues can arise from glass tube inhomogeneities. T he differences from the expected counts were repeatable for all measured samples, as is evidenced by comparison of the as recei ved and the 15 -month old sample. Therefore, the comparison of the RIM -PLA targets will always be made to the experimental and not to the theoretical counts in the control. Both of the control spectra indicate no changes in peak positions. Moreover, t here are no new peaks appearing after 15 months indicating no temporal degradation of the molecule. From Table 5 5 it is also clear that there are small variations of proton counts for the dendrimer over the period of 15 months. The maximum change in pr oton counts was 20% for signal 1. This maximum change represents the assumed error for using the NMR proton ratios to assess RIM PLA induced degradation of the phosphorescent dendrimer targets Figure s 5 8 and 5 9 are different scales of the MALDI TOF ma ss spectra of the 15-month old phosphorescent dendrimer sample used for the NMR experiment, which confirms intact dendrimer are still present. In conclusion, t he excellent temporal stability of the phosphorescent dendrimer allows t he detection of RIM -PLA induced damages to its molecular structure

PAGE 95

95 5.5.2 Fluorescent Dendrimer Figure 5 10 shows a comparison of the 1H NMR spectra of the fluorescent dendrimer after it was received and the same sample 15 months later. The integrated areas under the peaks have b een tabulated in order to establish any possible degradation of the molecule over time. Table 5 6 shows the expected number of protons for the labeled peaks and their relative ratios to an arbitrary normal peak. In the case of the fluorescent dendrimer, the normal was set t o the upfield singlet ( 1.3 9 ppm) corresponding to the 108 protons from the tert -butyl groups at the dendrimers periphery. From Table 5 6, it is evident there are significant changes between the expected and the measured proton counts f or both samples due to the same reason stated before. There are no changes in peak positions by comparing both spectra. However, relative to the as received spectra, the 15 month old sample shows a significantly decreased proton count for signal 1 and se veral new peaks appeared upfield ( spectra b in Figure 5 9 ). These significant changes in the NMR spectra indicate degradation of the molecule s structure over time. Figure s 5 1 1 and 5 12 are different scales of t he MALDI TOF mass spectra of the 15-month old fluorescent dendrimer sample used for the NMR experiment, which confirms intact dendrimer still present. As opposed to the phosphorescent dendrimer, the lack of good temporal stability of the fluorescent dendrimer did not allow isolation of RIM -PLA in duced damages to its molecular structure Therefore the target data for the flu orescent dendrimer will not be presented in this dissertation. 5.6 Target Characterization of Phosphorescent Dendrimer Fluence: 1 J/cm2 The following discussion relates to t argets ablated at a fluence of 1 J/cm2 at each of the four resonance modes selected. Targets were ablated for 5, 10 and 20 minutes. For brevity, n o data on the shortest time scale has been included in this manuscript.

PAGE 96

96 5.6.1 Chloroform 5.6.1.1 C -H alkyl s tretch m ode : 3.32 m Figure 5 1 3 shows the 1H NMR spectra comparison of the phosphorescent dendrimer control and targets ablated with th e FEL tuned to the chloroform C H alkyl stretch mode. The targets were ablated for 10 and 20 minutes, respectively. Ta ble 5 7 lists the relative proton counts of the signals compared to the as received control. From the NMR spectra and the proton counts, it is evident that there are changes to the molecular structure from ablation at the C H alkyl stretch resonance mode of chloroform. The variations in proton counts are higher (>30%) than the maximum 20% error established in Section 5.5.1, indicating slight alterations to the dendrimer structure However, there are no new peaks and the FTIR spectra (Figure 5 14) of the t argets retain the same functional groups as the control. There is a minimal difference between the relative proton counts of the targets ablated at 10 and 20 minutes, respectively. The MAL DI TOF mass spectra of the target ablated for 20 minutes (not show n) showed intact dendrimer still predominantly present; t he result is listed in Table 5 14. 5.6.1.2 C -H bending mode : 8.18 / 8.28 m Figure 5 1 5 shows the comparison of the control and two targets ablated at the chloroform CH bending mode for 10 and 20 mi nutes, respectively From the NMR spectra and the proton counts in Table 5 8 it is evident that most downfield peaks decreased considerably (over 30 %) from ablation at the C -H bending resonance mode of chloroform. The vari ations in proton counts are sli ghtly higher than those seen for ablation at the C H alkyl stre tch resonance mode (Table 5 7) indicating similar structural damage to the phosphorescent dendrimer as ablation at the C H alkyl stretch There is no significant difference between the relati ve proton counts of the targets ablated at 10 and 20 minutes respectively The FTIR spectra of these targets (not shown

PAGE 97

97 similar to Figure 5 14) did not show any differences from the controls functional groups. The MALDI TOF mass spectra of the target ablated for 20 minutes (not shown) also s how ed intact dendrimer still present (result listed in Table 5 14). 5.6.2 Toluene 5.6.2.1 C -H stretch mode : 3.31 Figure 5 1 6 shows the 1H NMR spectra comparison of the control and two targets irradiated at the toluene C H stretch resonance mode. The targets were ablated at a fluence of 1 J/cm2 for 10 and 20 minutes, respectively Table 5 9 lists the relative pro ton counts of the signals compared to the as received control. The relative proton counts of the targets in Table 5 9 show very significant changes relative to the control. All areas of the downfield signals decreased considerably, up to 40% Notably the shape of signal 5 changed ; the part of this signal at higher chemical shift disappeared. Additionally, the triplet from signal 4 broadened and weakened relative to the other peaks These changes indicate more severe damage than irradiation at either of the chloroform resona nce modes. A small difference between the relative proton counts of the two targets indicates continued degradation over time. The FTIR spectra of these targets show ed ret ention of all functional groups of the control with no additio nal new bands The MALDI TOF mass spectra of the target ablated for 20 minutes showed i ntact dendrimer still present (Table 5 14). 5.6.2.2 C=C aromatic stretch mode : 6.23 Figure 5 1 7 shows the NMR spectra comparison of the phosphorescent dendrimer control and targets ablated for 10 and 20 minutes while the FEL was tuned to the toluene C=C aromatic stretch mode. Table 5 10 lists the relative proton counts of the signals compared to the control. The NMR spectra and the proton counts from Table 5 10 indicate significant changes to the molecular structure from ablation at the C=C stretch r esonance mode of toluene The

PAGE 98

98 variations in proton counts are comparable to the target s ablated at the C H stretch mode (Table 5 9). Similarly to this mode, signal 4 (a triplet) weakened and broadened. Moreover, signal 5 als o changed shape considerably; the part at higher chemical shift fully disappeared for both irradiation times Howev er, there are no new peak s and the FTIR spectra (not shown similar to Figure 5 14) of the targets retain the same functional groups as the control. There is no significant difference between the relative proton counts of the targets ablated at 10 and 20 minutes, respectively. The MALDI TOF mass spectra of the target ablated for 20 minutes showed intact dendrimer still predominantly present (Table 5 14). T his mode resulted in significant damage to the structure, comparable to that at the C H stretch mod e of toluene. 5.6.3 Conclusions After careful comparison of the relative proton counts for all the collected 1H NMR spectra, we conclude that all modes resulted in some degradation to the molecul ar structure of the phosphorescent dendrimer; the changes in proton counts were higher than the estimated error of 20%. Notably the chloroform resonance modes showed the least RIM PL A induced degradation, with slightly better molecular fidelity from the C H alkyl stretch mode Ablation at both toluene modes resul t ed in high dendrimer degradation evidenced by substantial changes to the spectral peaks. All targets ablated at the chloroform modes showed similar counts as a function of irradiation time. This behavior indicates that most of the irradiated material at the surface is eff iciently removed by FEL irradiation Consequently, the effective amount of degraded dendrimer remains constant over time. By contrast, the targets ablate d at the toluene modes showed slightly decreased proton counts (higher degradation) as a function of irradiation time This trend indicates that dendrimer removal at the surface of these targets is kinetically limited,

PAGE 99

99 therefore the relative amount of degraded molecule increases over time. These trends support the proposed ablation mech anisms for each of the solvents ( see Section 7.9). The c ollected FTIR data for the targets ablated at 1 J/cm2 showed no significant differences from the control spectra, irrespective of the resonance mode or solvent. MALDI TOF mass spectra showed intact d endrimers still present in all cases (Table 5 14). A lthough FTIR has been the most widely used method for testing the molecular structure of the deposited films, it is cl ear that the technique may be misleading. While this technique did not indicate degr adation, 1H NMR spectroscopy did. The high sensitivity of NMR spectroscopy proved to be a more valuable characterization technique to assess the degradation of the molecular structure of the targets. 5.7 Target Characterization of Phosphorescent Dendrimer: Changes in Fluence T he following section details the characterization of phosphorescent dendrimer targets ablated at increasing fluences: 10, 20 and 30 J/cm2, respectively. Due to limited luminescent material only the C H stretch resonance modes from b oth solvents were studied 5.7.1 Chloroform C -H alkyl stretch mode: 3.32 m Figure 5 1 8 shows the 1H NMR spectra comparison of the phosphorescent dendrimer control and targets ablated with th e FEL tuned to the chloroform C H alkyl stretch mode. All t arget s were ablated for 20 minutes. Table 5 11 lists the relative proton counts of the signals compared to the as received control. T he proton counts for all the spectra are comparable even as the fluence was increased Minimal variations are seen relative t o the control (only ~10% higher than the estimated error), and no additional peaks formed. In fact, the values are very close to those targets ablated at 1 J/cm2 for 10 and 20 minutes, respectively ( Section 5.6.1.1). It is evident that the minimal degradation occurring at the target is independent of ablation time ( Table 5 -7) and fluence (Table 5 11) This behavior indicates that most of the irradiated material at the surface is efficiently

PAGE 100

100 removed by FEL irradiation, irrespective of fluence. Consequent ly, the effective amount of degraded dendrimer in the target remains constant over time. The efficient ablation at this chloroform mode is supported by the high film yields (to be discussed in the next chapter). These trends support the proposed ablation mechanisms for this solvent (Section 7.9). 5.7.2 Toluene C -H stretch mode: 3.31 m Figure 5 1 9 shows the 1H NMR spectra comparison of the phosphorescent dendrimer control and targets ablated with th e FEL tuned to the toluene C -H alkyl stretch mo de. All t argets were ablated for 20 minutes. Table 5 1 2 lists the relative proton counts of the signals compared to the as received control. Contrary to the chloroform mode, ablation at higher fluence for the C -H stretch mode of toluene resulted in additional degr adation of t he molecular structure. C omparison of the relative proton counts of Table 5 12 to those of fluence 1 J/cm2 ( Section 5.6.2.1) indicates that a fluence increase leads to further deviation from the control counts (degradation) T he N MR spectra c hanged considerably; particularly evidenced by the normal signal which has substantially changed at 3 0 J/cm2. This upfield signal widened and changed shape considerably, suggesting new alkyl fragments are formed as the molecule breaks down. The trend of increased degradation indicates that dendrimer removal at the surface of these targets is kinetically -limited, irrespective of fluence; consequently, the relative amount of degraded molecule in the target increases over time and fluence. This conclusion i s consistent with th e low film yields for these films, as will be discussed in the next chapter. Figure 5 20 is the FTIR spectra of a target which was ablated at the highest fluence experimentally possible for that wavelength: 55 J/cm2. N ote that an additi onal band centered at 3286 cm1 formed when compared to the control spectra (Figure 5 7 ). This band corresponds to the O H stretch from the alcohol formed after

PAGE 101

101 cleavage of the 2 -ethylhexyloxy groups from the periphery.157 A dangling oxygen bond resulting from pyrolysis or photolysis cou ld easily be protonated. This band could also be the seco ndary amine resulting from cleavage of a dendron from the core.157 However, it is more likely to be the O H stretch as one examines the relative intensities of the C H stretches (Table 5 13) From Table 5 13 it is evident that the aromatic C H stretch intensity increased significantly while that of asymmetric C H alkyl stretch decreased If the infrared absorption is proportional to the functionals group concentrati on (Beer Lamberts law) it is evident that the FTIR spectra shows a decreased signal for the 2 -ethylhexyloxy groups from the dendrimers periphery. These results indicate that degradation of the molecule is likely to initially proceed by cleavage of the alkyl groups. 5.7.3 Conclusions The 1H NMR and FTIR data presented for the targets ablated at increasing fluences show that the C H alkyl stretch mode of chloroform results in minimal damage to the dendrimers molecular structure. The NMR spectra indicate little effect of increasing fluence (up to 30 J/cm2) or increasing time. That is, the amount of degraded dendrimer in the target stays constant, which is consistent with an efficient ablation mechanism (see Section 7.9). By contrast, ablation at the CH stretch mode of toluene resulted in increasing damage as the fluence increased. That is, the ablation mechanism for this mode in toluene is highly inefficient, resulting in higher amounts of degraded dendrimer (see Section 7.9). Considering that all the se targets were ablated at the same FEL wavelength, it is clear that the degradation of the molecular structure is dependent on the selected resonance mode and its absorption coefficient. 5.8 MALDI -TOF MS Results for RIM -PLA Targets MALDI TOF MS (previousl y defined in Section 3.5.3.3) was used to determine the presence of intact dendrimer in the targets ablated for 20 minutes. Representative large and M+-

PAGE 102

102 scale spectra for each den drimer are shown in Figures 5 8, 59, 511 and 512 Table 5 14 list s the re sults from the MALDI TOF mass spectra for the phosphorescent dendrimer The table lists the center value for the peak corresponding to the monoisotope mass for M+ (the singly ionized molecule) followed by the error in parenthesis. The theoretical isotopi c distribution was presented in Section 3.5.3.3.1; the value corresponding to M+ of the phosphorescent dendrimer is 2106.1595. The error was calculated with Equation 3 1 ; a positive value represents the theoretical value is higher than the signal value, w hile a negative stands for the opposite. The MALDI TOF data shows that intact dendrimer was retained for all targets. However, the MALDI TOF MS analysis of these samples is only qualitative. That is, the presence of the intact dendrimer can be determined but the relative amounts of degraded polymer, if any, cannot. The error in all o f the measurements is below 15 ppm, the standard acceptable value. 5.9 Conclusions This chapter has focused on understanding the amount of degradation in the frozen matrix t argets after FEL irradiation. The high degradation from irradiation at the toluene modes is consistent with a highly inefficient ablation mechanism that subjects the dendrimer to high temperatures long enough to induce pyrolysis. T he slight degradation f rom irradiation at the chloroform modes is consistent with a much more efficient material removal mechanism, which results in constant measured degradation with time and fluence. These results will be used in Section 7.9 to support the proposed ablation m echanisms for these two solvent matrices.

PAGE 103

103 Figure 5 1. 1H NMR chemical shifts of different organic moieties. Figure 5 2. Molecular structure of the phosphorescent dendrimer with proton assignments.

PAGE 104

104 Figure 5 3. M olecular structure of the fluor escent dendrimer with proton assignments Figure 5 4. Expansion of the downfield region of the fluorescent dendrimer control 1H NMR spectra.

PAGE 105

105 Figure 5 5. FTIR spectra of phosphorescent dendrimer control. Figure 5 6. FTIR spectra of fluorescent d endrimer control.

PAGE 106

106 Figure 5 7. 1H NMR spectra of phosphorescent dendrimer control: (a) as received; (b) 15 months after receipt.

PAGE 107

107 Figure 5 8. MALDI TOF MS (large scale) of phosphorescent dendrimer control 15 months after receipt. Figure 5 9. MA LDI TOF MS (M+ area) of phosphorescent dendrimer 15 months after receipt.

PAGE 108

108 Figure 5 10. 1H NMR spectra of fluorescent dendrimer control : (a) as received; (b) 15 months after receipt Figure 5 11. MALDI TOF MS (large scale) of flu orescent dendrimer c ontrol 15 months after receipt.

PAGE 109

109 Figure 5 12. MALDI TOF MS (M+ scale) of flu orescent dendrimer control 15 months after receipt

PAGE 110

110 Figure 5 13. 1H NMR spectra comparison for phosphorescent dendrime r targets ablated for different times at 1 J/cm2: C H a lkyl stretch resonance mode of chloroform

PAGE 111

111 Figure 5 14. FTIR spectra comparison for phosphorescent dendrime r targets ablated for different times at 1 J/cm2: C H alkyl stretch resonance mode of chloroform

PAGE 112

112 Figure 5 15. 1H NMR spectra comparison for phosphorescent dendrimer targets ablated for different times at 1 J/cm2: C H bending resonance mode of chloroform

PAGE 113

113 Figure 5 16. 1H NMR spectra comparison for phosphorescent dendrimer targets ablated fo r different times at 1 J/cm2: C H stretch resonance mode of toluene.

PAGE 114

114 Figure 5 17. 1H NMR spectra comparison for phosphorescent dendrimer targets ablated for different times at 1 J/cm2: C=C aromatic stre tch resonance mode of toluene

PAGE 115

115 Figure 5 18. 1H NMR spectra comparison for phosphorescent dendrime r targets ablated at increasing fluences : C H alkyl stretch resonance mode of chloroform.

PAGE 116

116 Figure 5 19. 1H NMR spectra comparison for phosphorescent dendrimer targets a blated at increasing fluences: C H stretch resonance mode of toluene

PAGE 117

117 Figure 5 20. FTIR spectra of phosphorescent dendrimer target a blated at 55 J/cm2: C H stretch resonance mode of toluene

PAGE 118

118 Table 5 1 Proton assignments for the phosphorescent dendrimer and appropriate signal legend for NMR spectra. Structure identification Total equiv. protons M ultiplicity C hemical shift (ppm) Assigned number in acquired NMR spectra* a 36 multiplet (2 overlapping triplets) 0.85 1.05 8 b 48 multiplet 1.25 1.62 normal c 6 multiplet 1.65 1.84 7 d 12 doublet 3.8 8, 3.89 6 e 12 doublet* 6.90 7.18 5 f 12 doublet* 7.58 7.72 3 g 9 singlet* 7.58 7.72 3 h 3 doublet* 7.72 7.83 2 i 3 singlet* 7.97 8.1 0 1 j 3 doublet* 6.90 7.18 5 k 3 doublet* 7.72 7.83 2 l 3 doublet* 7.97 8.10 1 m 3 triplet* 6.90 7.18 5 n 3 triplet 7.24 7.34 4 *Note that some mea sured signals overlap, resulting in multiplets which cannot be resolved (also seen in reference).89

PAGE 119

119 Table 5 2. Proton assignme nts for the fluo rescent dendrimer and appropriate signal legend fo r NMR spectra. Structure identification Total equiv. protons Expected m ultiplicity1 Chemical shift (ppm)1 Assigned number in acquired NMR spectra2 a 108 singlet 1.39 normal b 6 double doub let 7.39 1 c 12 doublet 7.43 1 d 24 three doublet s 7.06, 7.14 (doublet, 6H core ) 7.17, 7.28 (doublet, 12H G1 ) 7.19, 7.25 (doublet, 6H core ) 1 e 9 two broad singlets 7.61 (6H), 7.64 (3H) 1 f 12 doublet 7.54, 7.57 1 g 6 doublet (overlapping) 7.15, 7.46 1 h 6 doublet (overlapping) 7.15, 7.46 1 1Based on reference .90 2Note that all the signals at higher chemical shift overlap, resulting in a large multiplet which cannot be resolved.

PAGE 120

120 Table 5 3. Characteristic mid IR resonances observed in the phosphore scent dendrimer.157 Wavenumber (cm1) Wavelength ( m) Intensity Resonance a ssignment 3037 3.29 m C H aromatic stretch 2956 3.38 s C H asymmetric stretch, CH 3 2928 3.42 s C H asymmetric stretch, CH 2 2871 3.48 s C H symmetric stretch, CH 3 2865 3.49 s C H symmetric stretch, CH 2 1606 6.23 s C=C arom atic stretch 1511 6.62 s C=C aromatic stretch 1474 6.78 s CH 2 scissoring in alkane ; CH 3 asymmetric (bending) deformation; aromatic C H in -plane bending 1388 7.20 w m C H symmetric deformation, CH 3 1282 7.80 m C=C asymmetric stretch (Kekul ) 1178 8 .49 w aromatic C H in plane bending, p substituted benzenes (also 1016 cm1 w) 1111 9.00 w aromatic C H in plane bending, p substituted benzenes (also 1016 cm1 w) 1028 9.73 m s C=C stretching 1016 9.84 m C=C C trigonal bending 972 10.3 w aromatic C H out of plane bending, aromatic ring breathing 872 11.5 w C H out of plane bending, m substituted benzenes 827 12 .1 s C H out of plane bending, p substituted benzenes (w=weak ; m=moderate; s=strong)

PAGE 121

121 Table 5 4. Characteristic mid IR resonanc e s observed in the flu orescent dendrimer.157 Wavenumber (cm1) Wavelength ( m) Intensity Resonance a ssignment 3025 3.31 m C H aromatic stretch C H vinyl stretch 2962 3.38 s C H asymmetric stretch, CH 3 ( t butyl) 2904 3.44 m C H symmetric stretch, CH 3 ( t butyl) 2866 3.49 s C H symmetric stretch, CH 3 (aromatic t butyl) 1593 6.28 s C=C aromatic stretch 1510 6.62 s C=C aromatic stretch 1478 6.77 s CH 3 asymmetric (bending) deformation ; aromatic C H in -plane bending 1393 7.18 w m C H symmetric deformatio n, CH 3 ( t butyl) 1362 7.34 s C H asymmetric bending deformation ( t butyl) 1286 7.78 m C=C asymmetric stretch (Kekul ) 1248 8.01 m t butyl C C vibration 1202 8.32 m t butyl C C vibration 1178 8.49 m aromatic C H in plane bending, p substituted benzene s 1108 9.03 w aromatic C H in plane bending, p substituted benzenes 958 10.4 s aromatic C H out of plane bending; C H out of plane bending trans R -CH=CH R 872 11.5 w C H out of plane bending, m substituted benzenes 847 11.8 w m C H out of plane bend ing, m substituted benzenes; CH out -of -plane bending, p -substituted benzenes 827 12 .1 s C H out of plane bending, p substituted benzenes 705 14.2 m C H out of plane bending, m substituted benzenes; CH out -of -plane bending, cis R CH=CH R (w=weak ; m=moderate ; s=strong)

PAGE 122

122 Table 5 5. Relative proton counts from 1H NMR spectra of co ntrol phosphorescent dendrimer. Signal Expected proton c ount / Ratio to n ormal As r eceived 15 months later 1 6 / 48 = 0.13 0.15 0.12 2 + 3 27 / 48 = 0.56 0.59 0.56 4* 3 / 48 = 0.06 0.13 0.11 5 18 / 48 = 0.38 0.39 0.38 6 12 / 48 = 0.25 0.25 0.25 7 6 / 48 = 0.13 0.10 0.11 8 36 / 48 = 0.75 0.64 0.68 *Signal 4 exhibits considerably higher proton count in the measured samples due to overlap with the nondeuterat ed CHCl3 singlet. Table 5 6. Relative proton counts from 1H NMR spectra of control fluorescent dendrimer. Signal Expected proton c ount / Ratio to n ormal As r eceived 15 months later 1 75 / 108 = 0.69 0.87 0.78 2 0 / 48 = 0 0.05 0.04 3 0 / 48 = 0 0.0 3 0 Table 5 7. Relative proton counts from 1H N MR spectra presented in Figure 5 13: C H alkyl stretch mode of chloroform at the same fluence Signal Expected proton c ount / Ratio to n ormal Control 1 J/cm2 10 min 1 J/cm2 20 min 1 6 / 48 = 0.13 0.15 0. 11 0.10 2 + 3 27 / 48 = 0.56 0.59 0.49 0.46 4* 3 / 48 = 0.06 0.13 0.08 0.09 5 18 / 48 = 0.38 0.39 0.34 0.33 6 12 / 48 = 0.25 0.25 0.22 0.21 7 6 / 48 = 0.13 0.10 0.10 0.09 8 36 / 48 = 0.75 0.64 0.69 0.70 *Signal 4 exhibits considerably higher proton count in the measured samples due to overlap with the nondeuterated CHCl3 singlet.

PAGE 123

123 Tab le 5 8. Relative proton counts from 1H NMR spectra presented in Figure 5 15: C H bending mode of chloroform at the same fluence Signal Expected proton c ount / Ratio to n ormal Control 1 J/cm2 10 min 1 J/cm2 20 min 1 6 / 48 = 0.13 0.15 0.10 0.10 2 + 3 27 / 48 = 0.56 0.59 0.44 0.45 4* 3 / 48 = 0.06 0.13 0.09 0.09 5 18 / 48 = 0.38 0.39 0.30 0.30 6 12 / 48 = 0.25 0.25 0.20 0.20 7 6 / 48 = 0.13 0.10 0.08 0 .08 8 36 / 48 = 0.75 0.64 0.65 0.65 *Signal 4 exhibits considerably higher proton count in the measured samples due to overlap with the nondeuterated CHCl3 singlet. Tab le 5 9. Relative proton counts from 1H NMR spectra presented in Figure 5 16: C H st retch mode of toluene at the same fluence Signal Expected proton c ount / Ratio to n ormal Control 1 J/cm2 10 min 1 J/cm2 20 min 1 6 / 48 = 0.13 0.15 0.10 0.09 2 + 3 27 / 48 = 0.56 0.59 0.46 0.42 4* 3 / 48 = 0.06 0.13 0.10 0.12 5 18 / 48 = 0.38 0.39 0.31 0.28 6 12 / 48 = 0.25 0.25 0.20 0.19 7 6 / 48 = 0.13 0.10 0.09 0.07 8 36 / 48 = 0.75 0.64 0.64 0.62 *Signal 4 exhibits considerably higher proton count in the measured samples due to overlap with the nondeuterated CHCl3 singlet.

PAGE 124

124 Table 5 10. Relative proton counts from 1H N MR spectra presented in Figure 5 17: C=C aromatic stretch mod e of toluene at the same fluence Signal Expected proton c ount / Ratio to n ormal Control 1 J/cm 2 10 min 1 J/cm 2 20 min 1 6 / 48 = 0.13 0.15 0.09 0.10 2 + 3 27 / 48 = 0.56 0.59 0.44 0.44 4* 3 / 48 = 0.06 0.13 0.16 0.13 5 18 / 48 = 0.38 0.39 0.29 0.27 6 12 / 48 = 0.25 0.25 0.20 0.20 7 6 / 48 = 0.13 0.10 0.09 0.09 8 36 / 48 = 0.75 0.64 0.60 0.60 *Signal 4 exhibits considerably higher proton count in the measured samples due to overlap with the nondeuterated CHCl3 singlet. Table 5 11. Relative proton counts from NMR s pectra presented in Figure 5 18: C H alkyl stretch mode of chloroform at increasing fluences. Signal Expected proton c ount / Ratio to n ormal Control 10 J/cm2 20 min 20 J/cm2 20 min 30 J/cm2 20 min 1 6 / 48 = 0.13 0.15 0.12 0.11 0.11 2 + 3 27 / 48 = 0.56 0.59 0.51 0.49 0.50 4* 3 / 48 = 0.06 0.13 0.10 0.09 0.08 5 18 / 48 = 0.38 0.39 0.34 0.33 0.34 6 12 / 48 = 0.25 0.25 0.22 0.21 0 .22 7 6 / 48 = 0.13 0.10 0.10 0.08 0.09 8 36 / 48 = 0.75 0.64 0.63 0.65 0.67 *Signal 4 exhibits considerably higher proton count in the measured samples due to overlap with the nondeuterated CHCl3 singlet.

PAGE 125

125 Table 5 12. Relative proton counts f rom NMR s pectra presented in Figure 5 19: C H stretch mode of toluene at increasing fluences. Signal Expected proton c ount / Ratio to n ormal Control 10 J/cm 2 20 min 20 J/cm 2 20 min 30 J/cm 2 20 min 1 6 / 48 = 0.13 0.15 0.10 0.10 0.08 2 + 3 27 / 48 = 0.56 0.59 0.43 0.42 0.36 4* 3 / 48 = 0.06 0.13 0.11 0.10 0.09 5 18 / 48 = 0.38 0.39 0.29 0.29 0.27 6 12 / 48 = 0.25 0.25 0.19 0.19 0.16 7 6 / 48 = 0.13 0.10 0.07 0.09 0.12 8 36 / 48 = 0.75 0.64 0.59 0.60 0.60 *Signal 4 exhibits considerably higher proton count in the measured samples due to overlap with the nondeuterated CHCl3 singlet. Table 5 13. Comparison of ratios of intensities of C H stretches of target shown in Figure 5 20 with the control spectra (Figure 5 7). Sample Ratio of aromatic C H stretch (~3030 cm1) to asymmetric C H alkyl stretch (~2925 cm1) Ratio of symmetric C H alkyl stretch (~2865 cm1) to asymmetric C H alkyl stretch (~2925 cm1) Control 0.27 0.81 Target 0.47 0.55 Table 5 14. Values and error for the monoisotopic mass for M+ from the MALDI TOF MS data for the phosphorescent dendrimer targets. Solvent Mode ( m) Fluence (J/cm2) Signal [amu] (Error [ppm]) Chloroform 3.32 1 2106.1567 (1.3) 8.18 / 8.28 1 2106.1634 ( 1.9) Toluene 3.31 1 2106.1779 ( 8.7) 6.23 1 2106.1696 ( 4.8)

PAGE 126

126 CHAPTER 6 FILM CHARACTERIZATIO N 6.1 Introduction The following chapter details the characterization of the films deposited from ablation at all resonance modes. Similar to the target characterization chapter, the data is divided into two majo r sections. The first section describes the characterization of the films deposited from ablation at a low fluence of 1 J/cm2 for both dendrimers. The latter section includ es data at higher fluences (10, 20 and 30 J/cm2) for only the phosphorescent dendr imer because of too little fluorescent material 6.2 Film Surface Topography and Thickness of Control Films Control dendrimer films were deposited via spin -co ating for comparison purposes. Solutions in chloroform (12 mg/mL) were spin-coated at 3000 rpm f o r 30 s. The films were dried for 4 hours at 25 C in low vacuum (0.1 atm ). The film s were then cut with a razor blade and the thicknes s was measured with AFM (Figure 6 1) The selected area in each AFM image produces an average line scan as shown in Fi gure 6 1 Sections on each line scan are enclos ed by markers; the average step height i s the film thickness. Table 6 1 tabulates the film thickness measured with AFM thickness or with stylus profilometry Although AFM offers better resolution ( nominal ti p radius ~ 8 nm ; scan rate ~1 2 Hz ) than stylus profilometry ( nominal tip radius ~ 12 m), the thickness measured by profilometry is consistent with the value obtained with AFM. However, acquisition of the AFM images in Figure 6 1 takes considerably longe r time than the s tylus profilometer measurements Due to the number of measurements needed for all the deposited films, AFM was used to quantify roughness and stylus profilometry was used for all thickness measurements.

PAGE 127

127 Figure 6 2 shows AFM height images and transmission -mode optical micrographs (50X optical objective) of the spin -coated control dendrimer films. Table 6 2 lists the v ertical scanning ranges (Z range s ) and the respective RMS (root -mean -square) roughness from multiple images. Note that the grainy appearance in the optical micrographs is a result of the underlying glass substrate. Figure 6 3 is a n optical micrograph of the bare glass substrate, which was focused on a foreign particle at the surface. 6.3 Characterization of Dendrimer Films Fluence: 1 J/cm2 The following discussion relates to films deposited from targets ablated at a fluence of 1 J/cm2 at each of the four resonance modes selected. Targets were ablated for 5, 10 and 20 minutes. Optical micrographs were obtained through 1 0X a nd 50X optical objectives; f luorescence/bright -field images were obtained through 10X and 40X opt ical objectives. In the following sections large area and zoomed-in optical micrographs are presented. The zoomed in images are representative of the featur es in the large area image The selected higher magnification sections are note d by faded squares in the large area images Note that the AFM images shown do not correspond to the zoomed in optical micrographs. These images are taken in different appara tus making it impossible to image the same area. The films used for stylus profilometry measurements were deposited on substrates masked with pressure -sensitive tape, which was removed following ablation The data presented are the average of four scans orthogonal to the film ledge created by the mask. The maximum and minimu m step heights in each scan were recorded. T he average result and respective ranges for all measurements are presented. Note that the ( ) sign in these measurements does not imply a n error in the profilometry measurement, but it gives the range of heights from the average height value In a sense, the stylus profilometer gives a maximum roughness range from the average

PAGE 128

128 film thickness. The variation in heights across the profilomete r scan length is considerable due to th e large roughness of the films, as will be shown below 6.3.1 Chloroform 6.3.1.1 C -H alkyl stretch m ode : 3.32 m Phosphorescent d endrimer: Figure 6 4 shows reflection -mode optical micrographs and representative AFM he ight images of phosphorescent dendrimer films deposited on glass substrates for target ablation at the C H alkyl stretch mode of solid chloroform (3.32 m) Table 6 3 lists the RMS roughness for the height images in Figure 6 4 as well as the profilometry thickness Ablation using this mode result ed in large film roughness and thickness which increased with irradiation time. The ablation mechanism at this mode deposits dense dendrimer beads and strings on the substrate as well as few random small circul ar features (craters) on the film (Figure 6 4, 5 min) These features are believed to result from frozen matrix particles str iking the substrate, re -dissolving previously deposited material upon melting (Section 7.9). The majority of these features range in size from 5 15 m in diameter. Although the RMS roughness for the representative AFM images are increasing, their magnitude is lower than the thickness range of stylus profilometry. As can be s een from Table 6 3, the ranges of height va riations can b e as high as 76% of the average value being the highest for the 5 minute film. Figure 6 5 shows the corresponding phase images for the height images of Figure 6 4 which s how better contrast of the film features Figure 6 6 shows fluorescence and bright -field images of the different films in Figure 6 4 Note that although the images appear dark, t he entire focal area s of all the films showed fluorescence. The CCD camera images display contrast not absolute lumin ance The high contrast indicates that th e dense deposited particles show high luminescence ; therefore they

PAGE 129

129 contain intact dendrimer core, as this is the responsible moiety in the molecule for luminescence.86 In fact, films deposited by this mode show very high luminescence relative to films deposited from toluene (Section 6.6 ). Fluorescent d endrimer : Figure 6 7 shows optical micrographs (reflection-mode) and representative AFM height images of t he fluorescent dendrimer films deposited from ablation at the same C H stretch mode (3.32 m) and fluence (1 J/cm2). T able 6 4 lists the RMS roughness f r om the se height images and the stylus profilometry thickness Figure 6 8 shows AFM phase images of so me of these films, showing better contrast. Comparison of Figure 67 with Figure 6 4 indicates no d ifference in film topographies by changing dendrimers in the same solvent. The RMS roughness and thickness increase with ablation time. The fluorescent den drimer films show the same dense b eads and strings when ablated with this mode and same frequency of random small circular patterns (Figure 6 7, 5 min). These features result in very large height variation s as shown by the stylus profilometry data. Simil ar to the phosphorescent dendrimer, the dense particles show higher fluorescence (Figure 6 9 ) than the underlying film. Note that after 20 minutes of ablation, the films are very thick; therefore the optical microscope cannot simultaneously focus all of t he features of the topography. Th e AFM height image for this irradiation time has a large range of heights sugge sting multiple layers, all with rough features Conclusions : T he images of the deposited films show that ablation with the C H alkyl stretch re sonance mode of chloroform results in very large roughness relative to the control films irrespective of whether fluorescent or phosphorescent dendrimer was in solution. The film s RMS roughness increase with irradiation time and are comparable for both dendrimers. Figure 6 10 is a plot of the roughness and thickness results for all films deposited with this mode

PAGE 130

130 T his mode resulted in a n average deposition rate of ~2 nm/s based on the thickness measurements for all films 6.3.1.2 C -H bending mode : 8.18 / 8.28 m Phosphorescent d endrimer: Figure 6 11 shows the optical micrographs (reflection-mode) and representative AFM height images of phosphorescent dendrimer films deposited from target ablation at the C -H bending mode of chloroform (1 J/cm2). Table 6 5 tabulates the RMS roughness measured for the height images, as well as thickness from profilometry. A blation at this mode produce d films with notably increa sing roughness and thickness with ablation time. Similar to ablation at the C H alkyl stretch mo de, the films exhibit much higher roughness t han the control, as they show the same de nse particle beads and strings and small circular features from solvent explosions (Figure 6 11, 5 min). These features are believed to result from frozen matrix particl es striking the substrate re dissolving previously deposited material upon melting (Section 7.9). The height variations in the films can be as high as 75% from the average value; the highest being for the shortest irradiation time (5 min) Figure 6 12 d isplays the fluorescence microscope images for all ablation times, which show higher luminescence from the dense dendrimer beads and strings (indicating intact dendrimer core) Notably, the bright -field mode images exhibit small beads of high transmission (white circles) indicating very small voids (~1 5 m) resulting from the same target explosions of frozen particles Some of these voids show high fluorescence as a result of light scattering. Fluor escent d endrimer : Figure 6 13 shows the optical microgr aphs (reflection -mode) and representative AFM height images of the fluorescent dendrimer films deposited from ablation at the same mode and fluence discussed above. T able 6 6 lists the RMS roughness measured for the heights images in Figure 6 13 with the profilometry thickness

PAGE 131

131 Comparable AFM RMS roughness and thickness values were obtained for this dendrimer rela tive to the phosphorescent molecule further supporting the conclusion that the ablation mechanism is mo de dependent but dendrimer independent. The same features of the phosphorescent dendrimer are observed in the fluorescent dendrimer films. Figure 6 14 shows fluorescence and bright -field mode images for these films. Con c lusions : The results from ablation at the C H bending mode of chloroform at 1 J/cm2 indicate increasing roughness and thicknes s with irradiation time for both dendrimers Figure 6 15 is a plot of the results for easy comparison. T his mode resulted in a n average deposition rate of ~1 nm/s based on the thickness measurements for all the dendrimer films 6.3.1.3 Comparison between modes The discussed images show that ablation at any of the studied chloroform resonance modes result ed in films with very high RMS roughness even for the short irradiation time of 5 minutes. The deposit ion rates for all films were very high, considering the low concentration of the dendrimer solutions (0.5 wt. %). Based on the film yields from profilometry, the ablation mechanism for this mode is highly efficient, resulting in films with high fluorescen ce. Section 7.9 describes this mechanism (phase explosion from the solid state ) in more detail. Figure 6 16 is a plot of the AFM and profilometry measurements for all dendrimer films at both chloroform modes By averaging both dendrimers the RMS roughne ss from ablation at the CH alkyl stretch mode resulted in ~100150% high er values relative to the C H bending mode depending on irradiation time Similarly, the av erage film thickness and consequently deposition rates, were ~ 50 150% higher, also depen ding on the ablation time.

PAGE 132

132 6.3.2 Toluene 6.3.2.1 C -H stretch m ode : 3.31 m Phosphorescent d endrimer: Figure 6 17 shows optical micrographs (reflection-mode) and representative AFM height images of the deposited phosphorescent dendrimer films from ablation with the C H stretch resonance mode of toluene (1 J/cm2). Table 6 -7 lists the RMS roughness of the shown AFM i mages, as well as the s tylus profilometry thickness The film deposition rates for this toluene mode are much lower than those from ablation at a ny of the chloroform modes. The average film thickness increase d with irradiation time T he RMS roughness of the AFM height images do not increase significantly as the irradiation time is increased. These RMS roughness are much closer to the control val ues from Table 6 2 than any of the chloroform ablated films However, the re are high variation s in heights from the profilometry measurements as this instrument has a larger scan length (30 m m). T he films have a very high thickness range r elative to the average height value, as high as 72 % for the 5 minute film These height variations are the result of liquid solvent explosions towards the target which create drying patterns in the previously deposited film (Figure 6 17) These explosions are much la rger than any of the small circular features seen for any of the chloroform -ablated films (i.e. 5 15 m in diameter) The most noticeable range in size from 50 to 150 m in diameter; however upon zoomingin on the optical micrographs, there are some as sm all as 10 m in diameter (see Figure 6 17, 50 m 50 m images). Chapter 7 will correlate these results to the proposed ablation mechanism for this toluene mode (normal vaporization with phase explosion from the liquid state ). T he micrograph of the film deposited for 5 minute s shows very dense particles (black spots on the image ; ~1 3 m in diameter ), along with higher RMS roughness relative to the 10 minute films These particles are not foreign debris ; they are in fact dendrimer, some of them retaining

PAGE 133

133 significant luminescence as shown by the fluorescence and bright -fie ld mode images (Figure 6 18). A s the irradiation time increased to 10 and 20 minutes, the contrast for these particles is not as strong as in the 5 minute micrograph since the films are getting thicker (F igure 6 17) T he density of these particles does not increase proportionally with irradiation time. Therefore, i t appears as if most of them are deposited at the initial stages of ablation, within the first five minutes. The fluorescen ce mode images for films from this mode (Figure 6 18) do not show the high contrast displayed by the chloroform images. In these toluene films, fluorescence from the entire focal area of the microscope was more uniform aside from the previously described drying patterns. However, upon comparison of the substrates under a UV lamp, these films showed considerably less luminescence than the chloroform films. In fact the quantum yields for these films we re very low relativ e to chloroform ( Section 6.6) Fluo rescent d endrimer : Figure 6 19 shows the reflection -mode optical micrographs and representative AFM height images of the fluorescent dendrimer films deposited with the C H stretch resonance mode of toluene (1 J/cm2). Ta ble 6 8 tabulates the RMS roughness measured for the heights images in Figure 6 19, as well as the profilometry thickness measurements. The characteristics of the fluorescent dendrimer films do not vary from the previously discussed phosphorescent films. The thickness height vari ations we re also very large relative to the average value, primarily due to the circular patterns from the melted target explosions. The films deposited in 5 minutes also show the same dense particle features however few show a ny fluorescence indicating degraded de ndrimer material (Figure 6 20). The density of these features did not increase with irradiation time. The similarities in film topographies and features for all

PAGE 134

134 films, irrespective of the material in solution, support the conclusion that the film p ropert ies are mode -dependent but dendrimer -independent. Conclusions : Figure 6 21is a plot of the A FM RMS roughness and stylus profilometry thic kness for all films ablated using the C -H stretch mode of toluene (1 J/cm2). On average, the fluorescent dendrimer fil ms were slightly thicker and rougher than the phosphorescent films From the average profilometry thickness this mode resulted in a deposition rate of ~0.15 0.2 nm/s. This mode produced film s with RMS roughness between 3 0 and 100 nm ; the representative measurements did not indicate dependence with irradiation time As opposed to the chloroform mode roughness, this range of AFM roughness is very close to that measured by stylus profilometry. On average, both film sets increased in thickness with irradia tion time, although the se ranges we re very large due to the drying patterns resulting from liquid target explosions 6.3.2.2 C=C aromatic stretch mode : 6.23 Phosphorescent Dendrimer : Figure 6 22 shows optical micrographs (reflection-mode) and AFM height images of the deposited films from ablation at the C=C aromatic stretch m ode of toluene. Table 6 9 tabulates the RMS roughness of the shown images, as well as the stylus profilometry thickness. Similar to ablation at the C H stretch mode, these films we r e considerably thinner than any of those deposited at the chloroform modes. The RMS roughness we re much higher than the spin coated control, but much lower than a ny of the films ablated from chloroform To the naked eye, the films appear more specular th an the rough er chloroform films. The film deposited for 5 minutes shows the same dark dense particles seen at the C H stretch mode of toluene These particles result in higher initial RMS roughness relative to the 10 minute film. A s seen in Figure 6 23, some of them we re fluorescent indicating some still have intact

PAGE 135

135 dendrimer core. Their density per unit area did not increase proportionally to irradiation time, indicating they we re deposited in the initial stages of ablation (as was the case for the C H stretch mode) The films grow with ablation time a lthough the profilometry thickness range of thickness is relatively large as a result of drying (circular) patterns. Compared to the previous mode of tol uene (C H stretch), the circular features created by the liquid explosions we re less frequent; although often they we re larger as shown by both sets of images (Figure 6 22 and 6 23). These features are consistent with a low vapor phase nucleation rate in the melted target, followed by growth of the form ed s upercritical nuclei (Chapter 7 ). The larger vapor volumes within the melted target eject larger amounts of liquid towards the substrate. Since the nucleation rate is low, the density of circular drying patterns in the film is smaller than that at the C H stretch mode of the same solvent. The m ost noticeable circular patterns range in size from 100 to 250 m in diameter; however upon zoomingin on the optical micrographs, there were some as small as 10 m in diameter (see Figure 6 22, 50 m 50 m op tical micrographs ). T hese largest circular features we re 100 m bigger, i.e. 67% larger than tho se seen at the C H stretch mode. Fluorescent d endrimer : Figure 6 24 shows the optical micrographs (reflection-mode) and AFM height images of the fluorescent de ndrimer films from ablation at the previously discussed toluene mode. The relevant AFM RMS roughness and stylus profilometry data for these fi lms is presented in Table 6 10. Relative to the phosphorescent dendrimer films, there are no significant differences by changing the material in solution. The fluorescent dendrimer films exhibit the same large particles of dendrimer for the films stopped after 5 minutes of irradiation. As shown by Figure 6 25, some of the features show sharp contrast in the fluores cence microscope. For some films

PAGE 136

136 (Figure 6 25, 10 and 20 min), the circular features from the solvent explosions show higher fluorescence from the center surrounded by a dimmer ring, suggesting a recoil of material as the solvent evaporates. Th e RMS rough ness stayed within 40 to 60 nm for this group of films although the stylus profilometer suggests a higher maximum range in topography. The film yields increase proportional to irradiation time. These films show very large circular features, some large e nough to fill the entire 50X objective focal area ( ~250 m in diameter, Figure 6 24, 20 min optical micrograph ). As previously stated, t heir size indicates nucleation and growth of large vapor bubbles within the melted solvent. These large bubbles eject significantly larger liquid droplets upon escape from the target surface. Conclusions : Figure 6 26 is a plot of the AFM RMS roughness and stylus profilometry thickness for all films ablated with the C=C aromatic stretch mode of toluene (6.23 m) The topo graphical features seen in this mode appear once again to be dendrimer -independent. The measured RMS roughness stayed between ~40 80 nm for all ablation times, irrespective of the material in solution. The liquid solvent bursts from this target wer e more violent, producing larger size droplets striking the substrate. The resulting large area circular (drying) features are responsible for the very large height ranges measured by profilometry. However, the average deposition rate for this mode was 0.2 nm/ s. 6.3.2.3 Comparison between modes Figure 6 2 7 is a plot of the AFM and profilometry results for all films ablated at both toluene modes (3.31; 6.23 m) The data and all the images show that ablation at either mode result ed in films with high RMS roughn ess relative to the dendrimer spin -coated controls, even at the short irradiation time of 5 minutes. Additionally, no particular dendrimer results in higher relative roughness supporting the conclusion that the material in solution is irrelevant to the

PAGE 137

137 resulting film topographies and deposition rates A generalized RMS r oughness range for both modes of toluene w as ~20 100 nm The high RMS roughness we re primarily a result of solvent explosions from the melted targets, leading to the very large droplets c reating the circular (drying) features seen on the films. Additionally, a ll the 5 -minute films from ablat ion with both modes showed a high density of dense dendrimer particles with average diameter of ~1 3 m The repeatability confirms that they are not foreign debri s at the substrate. Some of these particles we re fluorescent indicating that some but not all contain intact dendrimer core. The ir density did not increase with irradiation time, suggesting they originate during the first few pulses of abl ation. The average film deposition rates (~0.150.2 nm/s) did not vary considerably with resonance mode nor the dendrimer in solution. This rate is just ~ 10% of that from ablation at the CH stretch mode of chloroform (3.32 m). Considering all depositio ns used the same fluence (1 J/cm2), ablation from a toluene matrix wa s highly inefficient relative to chloroform. Chapter 7 will discuss in detail the ablation mechanism proposed for toluene (normal vaporization with some phase explosion). Even though the toluene films show better specular reflection to the naked eye than the chloroform films, upon high m agnification they all show ed circular (drying) patterns. The diameter of the se circular features can be as large as ~ 150 m for the C -H stretch mode and ~ 250 m for the C=C aromatic stretch mode. All these diameters we re significantly higher than the average diameter for the features seen with chloroform ablation (5 15 m). The large size of these features indicates that explosion from the target is less frequent. Although the vapor nucleation rate is low, the formed bubbles grow to large sizes, ejecting large amounts of liquid (droplets ) towards the target.

PAGE 138

138 6.3.3 Conclusions on Low Fluence (1 J/cm2) Films Table 6 11 lists the average deposition rates fo r all solvent resonance modes at a fluence of 1 J/cm2, as determined from the profilometry readings. While the ablation yields and respective roughness varied, the films deposited by each mode showed features which were characteristic of the solvent. For instance, in all chloroform ablated films dense beads and strings of dendr imer were deposited which were highly fluorescent under UV microscopy. The films in both modes for this solvent showed high ablation rates coupled with very high roughness All the toluene films showed more specular visible reflection to the naked eye However, upon high magnification (50X objective 50 m bar ), circular (drying) features could be seen as a result of a very high incidence of explosions from the melted target T he size and the frequency of the exploding droplets was dependent on the resonance mode The striking differences between the films indicate that the thermal properties of the solvent play a very crucial role in the dissipation of the energy deposited into the focal volume. The density and average size of the circular (drying) features seen with the different modes suggests that ablation occurs from different condensed states for each solvent. For all chloroform films, t he small average diameter (5 15 m) of the features ind icate that few, if any, liquid droplets reach the substrate. More likely, the high deposition rates (efficiency) for this mode suggest that ablation occurs directly from the solid state without the solvent undergoing a melting transition. As solid matrix particles strike the substrate, dendrimer and solvent will be deposited. The frozen solvent particles will eventually melt redistribute due to s urface tension effects and re -dissolve a small amount of previously deposited dendrimer. Chapter 7 will further discuss the proposed ablation mechanism of phase explosion from solid state (frozen) chloroform targets.

PAGE 139

139 In the case of the toluene films, the low deposition rates (efficiency) indicate that ablation occurs primarily through a diffe rent m echanism than that of chloroform ( normal vaporization, Section 7.9) The high density and large average diameter (50150 m) of drying patterns indicate that ablation does not occur from the solid state, but a fter the target has melted. Upon meltin g, vapor bubbles will nucleate and grow within the affected volume. The coalescing of vapor bubbles leads to a high pressure buildup in the liquid, which results in violent liquid explosions (shockwaves) limited by the speed of sound of the liquid The l arge droplets of liquid striking the substrate result ed in the drying patterns. 6.4 Characterization of Phosphorescent Dendrimer Films Increasing Fluence The following section details the characterization of deposited phosphorescent dendrimer films from targets ablated at increasing fluences of 10, 20 and 30 J/cm2, respectively. Due to the small amount of material available, only the C H stretch resonance modes from chloroform and toluene were studied. 6.4.1 Chloroform C -H alkyl stretch mode: 3.32 m Fig ure 6 2 8 shows transmission -mode optical micrographs and AFM height images of films ablated at increasing fluences with the chloroform C H alkyl stretch mode. Figure 6 29 shows the AFM phase images. Table 6 1 2 lists the RMS roughness from AFM and thickne ss from stylus profilometry. The characteristic features of this mode carry on from the previous low fluence results, i.e. small dense beads and strings of dendrimer with random craters resulting from solvent explosions. A small aver age increase in film t hickness resulted from increasing the fluence indicating that larger sections of the affected volumes are effectively ablated T he range of heights measured by the profilometer also incr eased with energy While the 50X (50 m bar) optical micrographs sh ow films with similar features, lower magnification images ( 10X 250

PAGE 140

140 m bar ) show larger circular patterns as fluence is increased (Figure 6 30). The number and the size of liquid droplets striking the substrate increase d with the energy deposited into t he focal volume. 6.4.2 Toluene C -H stretch mode: 3.31 m Figure 6 3 1 shows transmission -mode optical micrographs and AFM height images of films ablated at increasing fluences with the toluene C H stretch mode. Table 6 1 3 lists the results of roughness from AFM and thickness from stylus profilometry. T he film yields from this mode increase d significantly with fluence. The RMS roughness we re all comparable to the range of data shown in Figure 6 2 7 Similar to ablation at the chloroform mode, the size and n umber density of droplets striking the substrate also increase d with fluence However, starting from ablation at 20 J/cm2, the droplets carry significant amounts of dendrimer with them. As shown in the lower magnification (10X) micrographs (Figure 6 32), some drying patterns appear very dense indicating heavy amounts of dendrimer carried in the explosion. The diameter of these features can be as large as 600 m. Although, no fluorescenc e imag es were available, a continuous decrease in relative photolum inescence was measured with increasing fluence, indicat ing the de posited material is not intact (Section 6.6). 6.4.3 Conclusions on Film Deposition with Increasing Fluence Increasing the FEL fluence did not alter the film characteristics of e ither mode (Se ction 6.3.3) T he average deposition rates as given by the profilometry thickness increase with fluenc e. The relative increase is higher for the toluene mode, since much larger liquid solvent d roplets are driving the transfer of material to the substra te F or both modes in the different solvents, the size and number density of droplets striking the su bstrate increase d with fluence although more pronounced in the case of toluene. Th e fact that an increase is independent of sol vent suggests that the add itional energy is used to melt

PAGE 141

141 larger volumes in the frozen target. In fact, while the toluene targets were ablated, they could be observed being melted. In particular, the 20 minute ablations for toluene at 20 and 30 J/cm2 had to be split into two 10 mi nute depositions, refreezing the target in between The low ablation yields from toluene (0.090.17 nm/s) suggest that t he ablation mechanism continued t o be normal vaporization accompanied by increasing phase explosion (Section 7.9). In the case of chlor oform, the increase in exploding droplet size with fluence is less dramatic as compared to toluene, as seen by comparing Figures 6 3 0 and 6 -3 2 The high deposition rates and same features as the low fluence films indicate that the primary ablation mechani sm in chloroform did not change with fluence. However, the additional energy does not only cause ablation from the solid state, but from melted target. The higher fluence corresponds with larger drying patterns, which indicate an increased melted volume in the target (liquid droplets striking the substrate). A more detailed discussion of these mechanisms will follow in Chapter 7. 6.5 Molecular Structure of the Deposited Films 6.5.1 MALDI -TOF Mass Spectrometry MALDI TOF MS (previously defined in Section 3.5.3.3) was used to determine the presence of intact dendrimer in the films. The technique was performed only on the 20 minute films, as the material yields were high e nough to provide sufficient signal. Representative large and M+-scale spectra of the f ilms for each dendrimer are shown in Figures 6 3 3 through 6 36. Table s 6 1 4 and 6 15 list the results from the MALDI spectra for each dendrimer The table lists the center value for the peak corresponding to the monoisotope mass for M+ (the singly ionize d molecule) followed by the error in parenthesis. The theoretical isotopic distribution was presente d in Section 3.5.3.3.1; the values corresponding to M+ are 2106.1595 and 2142.4345 for the phosphorescent and the fluorescent dendrimer, respectively. The e rror wa s calculated with

PAGE 142

142 Equation 3 1 ; a positive value represent s the theoretical value is higher than the signal value while a negative stands for the opposite. The MALDI TOF data shows that i ntact dendrimer was retained for all films, e xcept those at high fluence for the toluene C H stretch mode when no dendrimer peak was detected. However, this MALDI TOF MS analysis of these samples is only qualitative. That is, the pres ence of the intact dendrimer could be determined, but the relative amounts of d egraded polymer, if any, cannot. The error in all but one of the measurements is below 15 ppm, the standard acceptable value. 6.5.2 FTIR Spectroscopy FTIR spectroscopy was done on phosphorescent dendrimer films depos ited at mid to high fluences on NaCl po lished disks No films showed any significant difference from the control, except the one ablated at the maximum average fluence of 55 J/cm2 for the toluene C -H stretch mode Figure 6 3 7 shows a comparison of the spectra for this film and the drop -casted control. A low intensity broad peak centered at 3290 cm1 appears for very high fluence ablation at this toluene mode. As mentioned in Section 5.7.2, this band could correspond t o either the O H stretch from the alcohol formed after cleavage of the 2 -eth ylhexyloxy groups from the periphery or the secondary amine resulting from cleavage of a dendron from the core .157 A dangling oxygen bond resulting from pyrolysis of the alkyl oxy side groups cou ld easily be protonated It is more likely to be the O H str etch as one examines the relative intensities of the C -H stretches (Table 6 16). From Table 6 16 it is evident that the aromatic C H stretch intensity increased significantly while that of asymmetric C H alkyl s tretch stayed constant. If the infrared abso rption is proportional to the functionals group concentration (Beer Lamberts law) it is evident that the FTIR spectra shows a decreased signal for the 2 ethylhexyloxy groups from the de ndrimers

PAGE 143

143 periphery. D egradation of the molecule is likely to initi ally proceed by cleavage of the se alkyl groups. 6.6 Photoluminescence Quantum Yields of Films The motivation of this study was to assess if RIM -PLA is an appropriate proc essing technique for the deposition of electroluminescent polymers for organic light e mitting diodes. It was unpractical in this study to build electroluminescent devices with films ablated with the FEL due to the laser location. H owever it was possible to obtain relative photoluminescence (PL) quantum yields (QY) and spectr a for some of the studied films, particularly the phosphorescent dendrimer. Most of the fluorescent dendrimer f ilms, even the control, show ed weak photoluminescence. The s e data will not be included as this dendrimer was shown to degrade over time. Figures 6 38 show th e PL spectra (exc = 390 nm) of phosphorescent dendrimer films deposited at the low fluence of 1 J/cm2. The data w ere obtained with the same acquisition parameters, in particular constant excitation and emission slit openings an d integration times. Table 6 17 lists the PL QY of these films Each listed value is an average of three measurements done by rotating the substrate next to the int egrating sphere. Note that any quantum yield below 3% (i.e. all toluene films) was not considered reliable as the sta ndard deviation from different measurements was too large relative to the average value Figure 6 39 shows the photoluminescence (PL) spectra (exc = 390 nm) comparison of the films deposited at increasing fluences for both C H stretch modes in the different solvents. Table 6 18 includes the respective PLQY results. It is evident that the films deposited from a chloroform matrix resulted in much higher PLQY than those of toluene. The yields for the C H stretch mode of chloroform did not change significantly with increasing fluence. The measured PLQY were considerably lower than that

PAGE 144

144 reported in the literature (22%); however, the films were meas ured a couple of days after deposition. No changes to the spectra shape were observed from the literature reports (Section 2.5.2 ). All of the PLQY of the toluen e ablated films were below 3%. 6.7 Reduced Film Roughness by Annealing It wa s possible to redu ce the deposited film roughness by annealing at a temperature higher than the polymers glass transition temperature (Tg). Figure 6 4 0 shows a comparison of a pre and post annealed film, in particular for the very rough 20 minute film deposited at the c hloroform C -H stretch mode. The film was a nnealed inside a low vacuum oven (0.1 atm) for 12 hours at 150 C. The Tg of this dendrimer is 132 C (Table 3 1) Table 6 19 lists the RMS roughness for the films. Although the RMS roughness decreased from 471 to 54 nm ( almost 90% ), the value is still much higher than that of the control (< 1 nm). 6.8 Conclusions Comparable AFM RMS roughness and thickness values we re obtained for both dendrimers in each selected mode This support s the conclusion that the RIM -P LA ablation mechanism s are mode -dependent but material independent, at least for this tight molecular weight range (21002150 g/mol) and solute concentration (0.5 wt. %) T he presence of the large iridium core ion did not affect the film properties deposi ted by either mode. The film characteristics have been shown to be primarily solvent -dependent; although the density and average size of the drying circular patterns is mode dependent. The predominant ablation mechanism of chloroform resulted in good ret ention of photoluminescence, although the morphology of the films makes them unsuitable for organic light emitting diodes. Substrate heating above the polymers Tg may ameliorate these effects. In the next chapter, thermodynamic and kinetic relations are used to correlate these results to appropriate ablation mechanisms.

PAGE 145

145 Figure 6 1. AFM images of razor blade cut control dendrimer films and average thickness analysis

PAGE 146

146 Figure 6 2. AFM height images and optical micrographs (trans mission -mode) of spin-coated control dendrimer films. Figure 6 3. Optical micrograph (transmission-mode) of a bare glass substrate

PAGE 147

147 Figure 6 4. Optical micrographs (reflection-mode) and AFM height images of phosphorescent dendrimer films: C H alkyl stretch resonance mode of chloroform (1 J/cm2). Figure 6 5. AFM phase images of phosphorescent dendrimer films: C H alkyl stretch resonance mode of chloroform (1 J/cm2).

PAGE 148

148 Figure 6 6. Fluorescence and bright -field mode images of phosphorescent dendr imer films: C H alkyl stretch resonance mode of chloroform (1 J/cm2).

PAGE 149

149 Figure 6 7. Optical micrographs (reflection-mode) and AFM height images of fluorescent dendrimer films: C H alkyl stretch resonance mode of chloroform (1 J/cm2). Figure 6 8. AFM phase images of fluorescent dendrimer films: C -H alkyl stretch resonance mode of chloroform (1 J/cm2).

PAGE 150

150 Figure 6 9. Fluorescence and bright -field mode images of fluorescent dendrimer films: C H alkyl stretch resonance mode of chloroform (1 J/cm2). Fi gure 6 10. Comparison of AFM RMS roughness and stylus profilometry thickness for both dendrimer films: C H alkyl stretch resonance mode of chloroform (1 J/cm2).

PAGE 151

151 Figure 6 11. Optical micrographs (reflection -mode) and AFM height images of phosphorescent dendrimer films : C H bending resonance mode of chloroform (1 J/cm2)

PAGE 152

152 Figure 6 12. Fluorescence and bright -field mode images of phosphorescent dendrimer films: C H bending resonance mode of chloroform (1 J/cm2).

PAGE 153

153 Figure 6 13. Optical micrographs (re flection -mode) and AFM height images of fluorescent dendrimer films: C H bending resonance mode of chloroform (1 J/cm2)

PAGE 154

154 Figure 6 14. Fluorescence and bright -field mode images of fluorescent dendrimer films: C H bending resonance mode of chloroform (1 J/cm2) Figure 6 15. Comparison of AFM RMS roughness and stylus profilometry thickness for both dendrimer films: C H bending resonance mode of chloroform (1 J/cm2)

PAGE 155

155 Figure 6 16. Comparison of AFM RMS roughness and stylus profilometry thickness for all dendrimer films at all chloroform resonance modes studied (1 J/cm2)

PAGE 156

156 Figure 6 17. Optical micrographs (reflection -mode) and AFM height images of phosphorescent dendrimer films: C H stretch resonance mode of toluene (1 J/cm2).

PAGE 157

157 Figure 6 18. Fluor escence and bright -field mode images of phosphorescent dendrimer films: C H stretch r esonance mode of toluene (1 J /cm2)

PAGE 158

158 Figure 6 19. Optical micrographs (reflection -mode) and AFM height images of fluorescent dendrimer films: C H stretch resonance mode of toluene (1 J/cm2).

PAGE 159

159 Figure 6 20. Fluorescence and bright -field mode images of fluorescent dendrimer films: C H stretch r esonance mode of toluene (1 J/cm2). Figure 6 21. Comparison of AFM RMS roughness and stylus profilometry thickness for both d endrimer films: C H stretch resonance mode of toluene (1 J/cm2).

PAGE 160

160 Figure 6 22. Optical micrographs (reflection -mode) and AFM height images of phosphorescent dendrimer films: C=C aromatic stretch resonance mode of toluene (1 J/cm2).

PAGE 161

161 Figure 6 23. Fluo rescence and bright -field mode images of phosphorescent dendrimer films: C=C aromatic stretch r esonance mode of toluene (1 J/cm2)

PAGE 162

162 Figure 6 24. Optical micrographs (reflection -mode) and AFM height images of fluorescent dendrimer films: C=C aromatic str etch resonance mode of toluene (1 J/cm2).

PAGE 163

163 Figure 6 25. Fluorescence and bright -field mode images of fluorescent dendrimer films: C=C aromatic stretch r esonance mode of toluene (1 J/cm2). Figure 6 26. Comparison of AFM RMS roughness and stylus profi lometry measurements for both dendrimer films: C=C aromatic stretch resonance mode of toluene (1 J/cm2).

PAGE 164

164 Figure 6 27. Comparison of AFM RMS roughness and stylus profilometry measurements for all dendrimer films at all toluene resonance modes studied (1 J/cm2)

PAGE 165

165 Figure 6 2 8 Optical micrographs (transmission -mode) and AFM height images of phosphorescent dendrimer films: C H alkyl stretch resonance mode of chloroform at increasing fluences. Figure 6 29. AFM phase images of phosphorescent dendrim er films: C H alkyl stretch resonance mode of chloroform at increasing fluences

PAGE 166

166 Figure 6 30. Low magnification optical micrographs (reflection-mode) of phosphorescent dendrimer films: C H alkyl stretch resona nce mode of chloroform at increasing fluenc es.

PAGE 167

167 Figure 6 3 1 Optical micrographs (transmission -mode) and AFM height images of phosphorescent dendrimer films: C H stretch resonance mode of toluene at increasing fluences.

PAGE 168

168 Figure 6 32. Large area optical micrographs (reflection -mode) of phosphorescent dendrimer films: C H stretch resona nce mode of toluene at increasing fluences. Figu re 6 33. MALDI TOF MS (large scale) of phosphorescent dendrimer: C H alkyl stretch resonance mode of chloroform (1 J/cm2, 20 min).

PAGE 169

169 Figure 6 3 4 MALDI TOF MS (M+ area) of phosphorescent dendrimer: C -H alkyl stretch resonance mode of chloroform (1 J/cm2, 20 min). Figure 6 3 5 MALDI TOF MS (large scale) of fluorescent dendrimer: C H alkyl stretch resonance mode of chloroform (1 J/cm2, 20 min).

PAGE 170

170 Figu re 6 36. MALDI TOF MS (M+ area) of fluorescent dendrimer: C H alkyl stretch resonance mode of chloroform (1 J/cm2, 20 min). Figure 6 37. FTIR spectra of control film and phosphorescent dendrimer film a blated at 55 J/cm2: C H stretch r esonance mode of toluene

PAGE 171

171 Figure 6 38. PL spectra comparison of phosphorescent dendrimer films deposited at low fluence (1 J/cm2) at all studied modes ( exc = 390 nm). Figure 6 39. PL spectra comparison of phosphorescent dendrimer films deposited at increasing fl uence for the C H stretch modes in both solvents ( exc = 390 nm).

PAGE 172

172 Figure 6 40. AFM height images and phase images of a phosphorescent dendrimer film annealed above its Tg: C H control spin coated dendrimer films (1 J/cm2, 20 min). Table 6 1. Thickne ss of spin -coated (control) dendrimer films using AFM or stylus profilometry. Dendrimer film AFM (nm) Stylus profilometry (nm) Phosphorescent 49.9 53.1 7.91 Fluorescent 89.9 94.7 9.75 Table 6 2. AFM Z ranges and RMS roughness of spin-coated (contr ol) dendrimer films. Control dendrimer film Z -range (nm) RMS roughness (nm) Phosphorescent 8.89 0.12 0.85 0.05 Fluorescent 10.9 0.01 1.35 0.14

PAGE 173

173 Table 6 3. AFM RMS roughness and stylus profilometry thickness of phosphorescent dendrimer films: C H alkyl stretch mode of chloroform (1 J/cm2). Target irradiation time (min) RMS roughness (nm) Stylus profilometry thickness Average and range (nm) 5 139 533 40 6 10 277 1260 449 20 471 2485 8 19 Table 6 4. AFM RMS roughness and stylus profilometry thickness of fluorescent dendrimer films: C H alkyl stretch mode of chloroform (1 J/cm2). Target irradiation time (min) RMS roughness (nm) Stylus profilometry thickness Average and range (nm) 5 187 438 4 14 10 320 1076 5 1 5 20 506 1826 9 13 T able 6 5. AFM RMS roughness and stylus profilometry thickness of phosphorescent dendrimer films: C H bending mode of chloroform (1 J/cm2). Target irradiation time (min) RMS roughness (nm) Stylus profilometry thickness Average and range (nm) 5 80 307 232 10 138 654 351 20 196 865 378 Table 6 6. AFM RMS roughness and stylus profilometry thickness of fluorescent dendrimer films: C H bending mode of chloroform (1 J/cm2). Target irradiation time (min) RMS roughness (nm) Stylus profilometry thickness Average and range (nm) 5 93 409 207 10 106 565 284 20 265 928 394 Table 6 7. AFM RMS roughness and stylus profilometry thickness of phosphorescent dendrimer films: C H stretch mode of toluene (1 J/cm2). Target irradiation time (min) RMS roughn ess (nm) Stylus profilometry thickness Average and range (nm) 5 34 60 43 10 28 88 45 20 40 166 86

PAGE 174

174 Table 6 8. AFM RMS roughness and stylus profilometry thickness of fluorescent dendrimer films: C H stretch mode of toluene (1 J/cm2). Target irrad iation time (min) RMS roughness (nm) Stylus profilometry thickness Average and range (nm) 5 46 74 55 10 87 126 73 20 69 210 108 Table 6 9. AFM RMS roughness and stylus profilometry thickness of phosphorescent dendrimer films: C=C aromatic stret ch mode of toluene (1 J/cm2). Target irradiation time (min) RMS roughness (nm) Stylus profilometry thickness Average and range (nm) 5 55 124 82 10 38 139 84 20 70 199 107 Table 6 10. AFM RMS roughness and stylus profilometry thickness of fluore scent dendrimer films: C=C aromatic stretch mode of toluene (1 J/cm2). Target irradiation time (min) RMS roughness (nm) Stylus profilometry thickness Average and range (nm) 5 43 89 59 10 61 122 63 20 55 203 162 Table 6 11. Average deposition ra tes for all resonance modes at a fluence of 1 J/cm2. Solvent Mode FEL wavelength ( m) Deposition rate (nm s1) Chloroform CH alkyl stretch 3.32 2.0 CH bending 8.28 1.0 Toluene CH stretch 3.31 0.15 0.2 C=C aromatic stretch 6.23 0.2

PAGE 175

175 Table 6 12. AFM RMS roughness and stylus profilometry thickness of phosphorescent dendrimer films: C H alkyl stretch mode of chloroform (20 min) Fluence (J/cm2) RMS roughness (nm) Stylus profilometry thickness Average and range (nm) 10 152 956 4 1 5 20 258 1044 512 30 174 1323 867 Table 6 13. AFM RMS roughness and stylus profilometry thickness of phosphorescent dendrimer films: C H stretch mode of toluene (20 min) Fluence (J/cm2) RMS roughness (nm) Stylus profilometry thickness Average and range (nm) 10 56 106 76 20 73 134 83 30 71 198 117 Table 6 14. Values and error for the monoisotopic mass for M+ from the MALDI TOF MS data for the phosphorescent dendrimer films. Solvent Mode ( m) Fluence (J/cm2) Signal [amu] (Error [ppm]) Chloroform 3. 32 1 2106.1613 ( 0.9) 10 2106.1727 ( 6.3) 20 2106.1734 ( 6.6) 30 2106.1677 ( 3.9) 8.18 / 8.28 1 2106.1788 ( 9.2) Toluene 3.31 1 2106.1668 ( 3.5) 10 2106.1653 ( 2.8) 20 N/A 30 N/A 6.23 1 2106.1696 ( 4.8) Table 6 15. Values and erro r for the monoisotopic mass for M+ from the MALDI TOF MS data for the fluorescent dendrimer films. Solvent Mode ( m) Fluence (J/cm2) Signal [amu] (Error [ppm]) Chloroform 3.32 1 2142.4242 (4.8) 8.18 / 8.28 1 2142.4412 ( 3.1) Toluene 3.31 1 2142.4504 ( 7.4) 6.23 1 2142.4390 ( 2.1)

PAGE 176

176 Table 6 16. Comparison of ratios of intensities of C H stretches of target shown in Figure 6 37 with the control spectra. Sample Ratio of aromatic C H stretch (~3030 cm1) to asymmetric C H alkyl stretch (~2925 cm1) Ratio of symmetric C H alkyl stretch (~2865 cm1) to asymmetric C H alkyl stretch (~2925 cm1) Control 0.27 0.81 Film 0.77 0.82 Table 6 17. Phosphorescent dendrimer thin-film PLQYs for depositions at low fluence (1 J/cm2). Solvent Mode ( m) PLQY ( PL, %) S td. Dev. (%) Chloroform 3.32 8.9 2.3 8.18 / 8.28 5.8 3.1 Table 6 18. PL quantum yields of phosphorescent dendrimer films deposited at increasing fluence. Solvent Mode ( m) Fluence (J/cm2) PLQY ( PL, %) Std. Dev. (%) Chloroform 3.32 10 8 7 1.8 20 9.1 2.1 30 7.9 2.5 Table 6 19. Measured RMS roughness of AFM images shown in Figure 640. Dendrimer film RMS roughness (nm) Pre annealed 471 Post annealed 54

PAGE 177

177 CHAPTER 7 TH ERMALLY INDUCED PHASE TRANSI TIONS AND ABLATION MECHANI SMS FROM FEL IRRADI ATION OF FROZEN SOLVENT MA TRICES 7.1 Introduction The discussion of this chapter deals with the use of the measured optical constants (Chapter 4 ) and other thermophysical properties of the solvents to explain the phase transitions and mechanisms leading to thin film dendrimer deposition. Thermal rise calcu lations are used to explain the target and film characteristics seen in C hapters 5 and 6, respectively. Note that the following discussion continues to assume that the micropulse (ps) structure of the FE L does not affect the ablation dynamics (Section 3.4.1.2), i.e. any reference to the laser pulse refers to the 4 7.2 Mechanisms for High Powe r Laser Absorption in Materials A ll optical responses of solids upon high power laser -beam absorption can be described by one of three primary mechanisms:159 (i.) the o ptical generati on of free carriers by electronic transitions ( e.g., interband (ground state -excited state) tr ansiti ons ) or impact ionization in semiconductors and insulators; (ii) n on -linear distortion of electron orbitals or of the entire molecule by the electric field of the intense beam (e.g. self -focusing3); and (iii) heat production and the associated variations in density or in the electronic characteristics of the material ( e.g. thermal self -focusing1 in transparent media). In the absence of other photophysical or photochemical processes, e.g. fluorescence and photolysis, the vibrational excited state will decay to a ground state by coupling to the solvent lattice.143, 145 T he redistribution of the vibrational energy of excited polyatomic molecules in condensed states occurs typically through the loss of energy from the s pecific vibrational mode 3 Self focusing or defocusing occurs if the real part of the refractive index of the target varies locally as a function of irradiance. Thermal self focusing is related to the temperature dependence of the refractive index.

PAGE 178

178 to some or all the other mechanical degrees of freedom ( the phonon bath) in the molecule.143144 This process ordinarily occurs on a 1012 109 s time -scale.143, 145 In the case of infrared irradiation of chloroform and toluene matrices, the FEL energy is deposited into anharmonic, vibrational modes with lifetimes in the order of ~ 10 200 ps.144, 160162 Figure 7 1 shows different possible mechanisms for the redistribution of vibrational energy.143 For a laser macropulse of 4 s, the energy of the FEL i s thermalized in a time scale of ~100 ps.163 Therefore, applying a t emperature rise model to the frozen solven t matrix system should explain t he ablation characteristics seen for the different resonance modes. 7.3 Confinement Conditions 7.3.1 Thermal Confinement Once the laser energy has been efficiently thermalized, it is appropriate to consider if the energy is confined to the focal volume aff ected by t he FEL macropulse (Table 4 16). Using the thermal and optical properties for each solvent, a characteristic thermal confinement time T) for each resonance mode can be calculated This quantity describes the time needed for the thermalized en ergy to begin diff using out of the focal volume.42, 164 In Equation 7 1: T optical T 2) ( (7 1) T is the thermal conductivity of the frozen solvent (Table 3 3) and optical is the optical penetration depth, previously defined in Section 4.2. Thermal confinement is achieved when the ratio of the laser pulse length ( p) to the thermal diffusion time fulfills the condition ( p / T ) 1.42 Table 7 1 tabulates the constants for the selected solvent modes. It is evident that the thermal confinement condition is satisfied at all resonance modes for a laser pulse of ~4 x 106 s Therefore, the peak temperatures in the focal volume will not be

PAGE 179

179 s ignificantly reduced nor the deposited energy be redistributed over a larger volume during FEL pulse irradiation 7.3.2 Stress Confinement Rapid heating of the frozen targets leads to the generation and propagation of thermoelast ic stresses until the affected volume returns to its equilibrium state.42 The release of these thermoelastic stresses are primarily limited by : (i) the depth of the disturbance (e.g. vapor bubbles) from the target surface and (ii) the speed of sound (US) in the solvent which determines the speed of propagation of the stress wave across the heated volume A relevant character istic time constant describing the m inimum time required for the stress wave to propagate outside of the focal volume is known as the stress confinement time S).42, 164 S optical SU (7 2) The condition known as stress confinement occurs wh p) is less S. With respect to the frozen matrices, no data was available in the literature for the speed of sound in the solid solvents. However, Rao established an empirical relati onship between the tempe rature coefficient of the speed of sound and the coefficient of thermal expansion for a liquid.165 166 Equation 7 3 may be used to approximate US at 77 K. Note that Toti, et al. found little variation of the speed of sound in a solvent upon polymer loading ,167 suggesting minimal effect from the dissolved dendrimer in the solvent matrices ) 298 ( exp ) 298 ( T k K U T UR S S (7 3) For the US values listed in Table 3 3 and the optical penetration de pths for the frozen (and S) vary between 1 10 ns. Therefore, the stress confinement condition is not satisfied for the FEL laser pulse.

PAGE 180

180 7.4 Assumptions 7.4.1 Negligible Plume Shielding The d iscussion in this chapter assumes that a plume does not develop during the a blation process of any of the targets; or if one forms its absorption coefficient is negligible. The average macropulse irradiance for a fluence of 1 J/cm2 is 105 W/cm2. At this rela tively moderate irradiance (i.e. below ~ 106 W/cm2) the vapor is expected to be tenuous and essentially transparent;159 therefore ionization of the desorbed molecules (formation of a plasma) should be negl igible. In fact, the good structural fidelity shown by the films and the targets deposited from chloroform, even at highe r fluences, suggest that minimal ionization of the solvent occurs. By contrast a previous report on UV (193 nm) laser irradiation of frozen chloroform matrices showed a reduction in molecular weight and structural alteration of deposited films of polyethylene glycol.168 The authors attributed the results to the detected presence of highly reactive species, Cland CHCl2 by mass spectrometry. T he highest FEL photon energy used in our experiments was 0.38 eV, while that of the UV laser is 6.4 eV. Since the typical covalent bond energies of organic compounds are ~3 4 eV, FEL irradiation does not provide enough energy for photolysis of the solvent. 7.4.2 Steady State Ablation Model Different models have been discussed for the representation of pulsedlaser ablation.42 For laser pulses in the time scale of s, material removal typically occurs simultaneously with the irradiation of the target (for the condition of no stress confinement) In such cases, a steady -state model has been used most often to describe the continuous material ejection proportional to the deposited energy in the target (fluence, irradiation). This model implies a fixed enthalpy of ablation (ha) and that material removal starts right after the pulse starts, and proceeds throughout

PAGE 181

181 the entire laser pulse. The deposited energy must overcome a threshold energy ( th), after which a linear dependence follows between the etch -depth and incident fl uence (Equation 7 4 ): a th etchh 0 (7 4 ) 0 is the experimental ablation fluence. Since the targets in this work re -melt ed following laser processing, surface etchdepth data was not available. However, from consideration of the film -yield (profilometry) data for each resonance mode and the incremental fluence results (Chapter 6 ), a linear dependence of average yields with deposited energy can be observed Note that even when ha is assumed constant for this model, the quantity may vary from one resonance mode to the next in our solvent systems, as it contains both sensible ( cp T ) and latent heat (e.g. enthalpy of melting, vaporization) components. The latter deals only with the energy required to raise the temperatur e of the matrix, but this temperature rise is dependent on the absorption coefficient of the mode (Section 7.5 ). The latent heat deals with phase transitions. Th us dependin g on the thermal properties of the solvent the absorption coefficient of the mode and the FEL pulse irradiance the target solvent may: (i) increase in temperature and remain a solid, (ii) mel t, and (iii) melt and vaporize. However, if the rate of rise in temperature is fast enough, the solvent may melt and subsequently superheat, a s will be discussed in further sections. The steady -state model has inherent limitations.42 It fails to describe systems where the ablation process is characterized by a succession of stages with different ablation enthalpies, e.g. a process that begins as a surface vapori zation but later takes on characteristics of bulk material ejection. How ever, for simplification, th e rate of rise of temperature is assumed to be such that there is a predominant ablation mechanism with a fixed ablation enthalpy (instantaneous thermal ri se) determining the film yields for each mode.

PAGE 182

182 7.5 Temperature Rise in Solvent The time -dependent heat diffusion equation (Equation 7 5 ) may be used to solve for the thermal rise in the frozen targets upon irradiation: 2 2 2 2 2 2z T y T x T c Q t TT p (7 5 ) where Q is the heat source and T is the thermal diffusivity ( T = kT / cp, ratio of thermal conductivity to the volumetric heat capacity) .13 14, 159, 169 Considering the laser beam spot size is much larger than the optical penetration depth ( optical), the heat effectively pro pagates in one dimension ( z -direction). Moreover, due to the low thermal conductivity of the solvents (~ 103 W cm1 K1 both in frozen and liquid states ) and the satisfaction of the thermal confinement condition ( Sectio n 7.3.2), the entire thermal diffusion term in Equation 7 5 can be eliminated. Equation 7 5 can then be further reduced to: pc Q t T (7 6 ) where the heating source Q is the thermalized energy from the target irradiation That is, I Q (7 7 ) where I is the irradiance (power per unit area) and is the modes absorption coefficient. Combining Equations 7 6 and 7 7 with further integration of the heat transfer equation leads to an approximate expression for the temperature rise at the irradiated solvent matrices: p pc c t I T 0 (7 8 ) wher 0 is the experimental ablation fluence. Equation 7 8 assumes that the thermal confinement condition is satisfied at all times, which is not the case once ablation has

PAGE 183

183 commenced and the deposited energy is dissipated. Additionally, it assumes that the frozen solvent matrix is isotropic and the thermophysical properties (absorption coefficient, heat capacity and density) are uniform and constant with time throughout the entire specimen vol ume The latter assumption is not valid. H owever, assuming no phase transitions occur during FEL irradiation, maximum temperature rises can be calculated for both starting frozen and melted matrices using the measured absorption coefficients (Chapter 4 ) and temperaturedependent thermophysical properties (Table 3 3 ) The re sults are tabulated in Table 7 -2 assuming an initial temperature of 77K Th e temperature rises in Table 7 2 correspond to heating rates on the order of 108 K /s. Normal phase transition s at such high heating rates are kinetically -limited, therefore superheating in t he frozen or the melted target is likely to occur .170172 T he volumetric energy density required to melt the focal volumes (Table 4 16) for solvent at a fluence of 1 J/cm2 were also calculated. Ta ble 7 3 lists the values obtained from dividing the enthalpy of melting ( Hsl) by the molar volume of the solvent in the solid state (Table 3 3). Considering that the average volumetric energy density deposited during FEL irradiation is ~ 105 J/cm3, only ~1 % or less of the deposited energy will be consumed in the melting transformation. Therefore, the maximum temperature rises in Table 7 2 should be valid; for all m odes except the C H stretch mode in frozen toluene, they result in values very close or excee ding the critical temperature of the solvent. These temperatures are thermodynamically unstable and experimentally impossible to attain, as will be discussed in later sections Therefore, a superheated melted target must undergo a phase transition at som e point prior to reaching the critical temperature corresponding to the ambient pressure. Section 7.6 discusses

PAGE 184

184 the relevant phase transitions involved in the heating of the solvent system in order to understand the implication of the calculated heating r ates to the ablation characteristics. 7.6 Thermally -Induced Phase Transitions in Solvent Matrices The following section applies thermodynamic and kinetic relations i n order to discuss the phase transitions involved in the solvent system (target). These re lations will be further used to construct the solvent phase diagrams and understand the corresponding energy dissipation mechanisms leading to dendrimer ablation. For simplification, the dissolved dendrimer (0.5 wt. %) is assumed to have a negligible effe ct on the p hase transitions of the solvent; therefore a pure substance is assumed. Figure 7 2 represents a schematic phase di agram showing equilibrium ranges between the solid, liquid and vapor phases of a pure substance The solid liquid and the liquid -v apor transition at ordinary temperatures are first -order transitions described by the latent hea ts of melting ( Hsl) and vaporization ( Hlv), respectively. The liquid vapor curve in the phase diagram ends at the critical point beyond which a supercritical fluid exists, where there is no distinction between the liquid and vapor. At the critical point the latent h eat and the volume change approach zero as ( Tc T )1/2, therefore there is no distinction between the vapor and liquid densities.159 Beyond Tc, evaporation proceeds as a continuous decrease in density, without a phase transi tion and without thermodynamic stability. 7.6. 1 Melting Melting transitions can be described by looking at the phase equilibrium between the solid and its melt and equating their free energies (chemical potentials) .159, 173 The change in free energy with respect to temperature can be written as: dT dP V S dT dP P G T G dT dG (7 9 )

PAGE 185

185 following from the thermodynamic relation G = U + PV TS The previous equation must hold for every phase individually. At equilibrium, the temperature and the pressure i nside the solid and the liquid must be the same. Therefore, sl sl sl s l s lV T H V V S S T d P d (7 10) where Vsl = ( Vl Vs) is the volume change upon evaporation and the entropy change Ssl = ( Sl Ss) has been replaced by the ratio of latent heat of melting to the solid liquid equilibrium temperature ( Hsl / Tsl). If Vsl in the previous equation is assumed to be constant, after integrating the previous equation it follows : s l sl sl slT T V H P ln (7 11) Equation 7 11 is valid in the range Tsl < T < Tlv. 7.6. 2 Evapor ation Similar to the case of melting, since evaporation typically occurs from the liquid state, the phase equilibrium between a melt and its vapor may be described by equating their free energies. However, since vapors are compressible, the equilibrium conditions depend on pressure. In this case, two additional approximations can be applied to Equation 7 10: (i) the volume of the vapor is much larger than the volume of the liquid ( Vlv v) and (ii ) the vapor may be approximated as an ideal gas. This leads to the Clausius -Clapeyron equation which describes the dependence of the vapor pressure on temperature, or, if written in the following form: lv lv lv lvH V T P d dT (7 12) also describes the dependence of the equilibrium temperature on pressure.

PAGE 186

186 An expression for the equilibrium (saturated) vapor pressure as a function of temper ature is obtained by integrating Equation 7 12: RT G P T T R H P T Plv lv lv lv lvexp 1 1 exp (7 13) where Plv is the correspo nding vapor pressure at Tlv. The previous equation is a valid approximation for Tlv < T < Tcr and is based on the empirical fact that the ratio of the latent heat to the difference in compressibility between the vapor and the liquid is approximately const ant in this temperature range (Equation 7 10). 7.7 Phase Diagrams of Solvents Used a s Matrices The phase diagrams of the solvents were constructed by applying the thermodynamic relations described in Section 7.6.2 in combination with an equation of state f or a gas. Figure s 7 3 and 7 4 depict the liquid -gas phase diagram s of chloroform and toluene respectively The curve s labeled as the binodal correspond to equilibrium vaporization These were calculated using the Clausius -Clapeyron equation (Section 7. 6 .2 ) and show the liquid at equilibrium with its saturated vapor at pressure i.e. P = P (T) The representative curve s labeled as vaporization with Knudsen layer acc ount for vaporization with a so -called Knudsen layer ( KL ) above the m elted target surface .174 176 This condition implies that solvent molecules are no longer desorbing from the target surface via an effusion -like process, i.e. the molecules mean free path is no longer much larger th an inter molecular radius. A Maxwell Boltzmann distribution for the velocity profile is no longer appropriate If th e quantity of molecules desorbing per unit time is large (on the order of 0.5 monolayer per nanosecond), they collide sufficiently and com e to equilibrium in a region known as the KL A very small number of collisions (~ 3) is sufficient to establish the KL T he vaporized particles, initially having only positive velocities normal to the surface, develop

PAGE 187

187 negative velocities upon KL formati on. However, in order that momentum be conserved, the particles also develop a positive center -of -mass (or flow) velocity.176 If the number of collisions is more than that needed for KL formation, then an Unsteady Adiabatic Expansion occurs so that, with increasing distance from the target, the gas kinetic velocities decrease while the flow (center of mass) velocity increases beyond that at the KL A passage to free -flight finally occurs and the velocities then persist unchanged. The condition of KL formation is relevant in the case that a plume was to develop above the ablated solvent target. However, as stated in Section 7.4.1, for simplicity this condition has been assumed to not be satisfied Normal heating refers to heating the solution target at the chamber pressure P0 until P equals P0 and the system. Assuming a liquid (melted target) and that the heating is slow enough the solvent undergoes normal boiling at Tlv. The curved labeled superheating refers to heating which is carried out sufficiently rapidly that the system passes beyond Tlv while the liquid remains in a metastable state at a constant pressure. An upper limit superheating of a metastable liquid exists the spinodal which is the boundary of thermodynamic phase stability. The estimation of the sp inodal locus for a pure solvent requires a ( P, V, T ) equation of stat e177 178 or equivalent e xperimental information. Since literature data for the latter is not a vailable for either solvent the spinodal was here estimated using the Redlich -Kwong equation of state, a modified van der Waals equation Th is equation may be used to approximately express the deviation of an actual gas from ideal gas behavior and has been shown to predict superheat limits very well .179 The superheat limit represents the p ractical, experimentally attainable stability limit, i.e. the deepest possible penetration of a liquid in the domain of metastable states.180 It is kinetically rather than thermodynamically controlled

PAGE 188

188 (by contrast to the spinodal) and therefore it is not an infinitely sharp sta bility boundary like the spinodal.179 The superheat limit at a constant composition is s pecified by the criteria:180 0 TV P and 0 PT S (7 14, 7 15) The previous equations are a consequence of the basic extreme principle of thermodynamics, which states that the entropy of a pure substance is at a maximum in a stable equilibrium state. T he spinodal curve should always lie beyond the observable, kinetically controlled superheat limit.179 In its general form, the Redlich -Kwong equation reads: 2 / 1T b V V a b V RT Pm m m (7 16) where P is p ressure, T is temperature, R is the gas constant, Vm is the molar volume, and the numeri cal constants are determined from the following criticality conditions: c cP T R a2 / 5 242748 0 and c cP RT b 0862 0 (7 17, 7 18) The previous equation of state ap proaches the ideal gas law as P decreases or T increases. As the liquid approaches the spinodal, fluctuations in local density increase rapidl y. This leads to fulfillment of the conditions 7.14 and 7.15, resulting in a loss of thermodynamic stability. A pplying such conditions, the equation of state becomes: 3 / 2 22599 0 1 2599 0 1 2599 0 2 1804 1 r r r r rT (7 19) and r r r r r r r r rP 5198 0 0676 0 1 5198 0 0676 0 1 2599 0 1 2599 0 1 2599 0 2 5412 33 / 2 3 / 1 5 (7 20)

PAGE 189

189 where Pr, Tr, and r are the reduced quantities of pressure, temperature and density, respectively .179 T he Redlich -Kwong spinodal is substance independent as it follows the reduced density and not the absolute solvent prope rties A dditionally, a pure substance has been assumed when calculating the spinodal i.e. ignoring the 0.5 wt. % solute concentration of dendrimer Finally, the curves labeled phase explosion onset refer to the limit at which rapid homogeneo us nucleation of vapor bubbles leads to a high pressure buildup with in the metastable (superheated) liquid. Upon crossing this limit, a violent equilibration of the system foll ows through a gas explosion that ejects liquid droplets towards the target The thermodynam ics and kinetics of this process will b e further discussed in Section 7.8.2.2.1. 7.8 Ablation Mechanisms from FEL Irradiation of Solvent Matrices T he following discussion examine s the three primary types of energy dissipation mechanisms leading to solvent/ dendrimer removal from the frozen matrices as a consequence of FEL -induced (thermal) phase transitions. These are: (i) surface evaporation (normal vaporization), (ii) volume evaporation (normal boiling), and (iii) phase explosion (e xplosive boiling ).181 The relevance of the three processes depends on the FEL pulse duration as well as the temperature distribution of the matri x following laser irradiation. The following treatment considers only thermal processes, disregarding other possibilities such as optical breakdown182 and plasma form ation.183 Other possible photomechanical effects due to stress confinement are ignored, since the condition is not satisfied in our experiments (Section 7.3.2) 7.8.1 Surface Evaporation (Normal Vaporization) Surface evaporation (also referred to as normal vaporization) refers to the transition from a condensed phase (solid or liquid) to vapor via emission of solvent molecules from the extreme outer surf ace of the target. In this sense evaporation also includes the process of sublimation. Under laser irradiation it can divided into internal and external processes.184 External processes

PAGE 190

190 deal with the motion of the vapor, plasma formation and the absorption of energy by the evaporated molecules. Internal pro cesses encompass what occurs within the condensed phase: the heating and recession of the target surface. The further discussion is limited to the latter, in particular the case of equilibrium without superheating of the solvent. U nder equilibrium conditi ons, the rate of particles evaporating from an open solid or liquid surface is equal to the rate of particles condensing on it from their saturated vapor. The rate of condensation, and hence also e vaporation, is obtained by inte gration over the Maxwellian velocity distribu tion of the vapor molecules According to the kinetic theory of gases, when the solvent evaporates into a gas, the resultant molar evaporation flux (moles / m2s) near the liquidvapor interface is defined by the Hertz Langmuir -Knudsen formula :174, 184186 T T R H T MR SP T R G T MR SP T MR T P S Jlv lv lv lv lv v1 1 exp 2 exp 2 2 (7 2 1) where M is t he molar mass of the solvent, R is the universal gas constant, S is an accommodation or evaporation c oefficient T is the temper ature of the evaporating surface P (T ) is its equilibrium vapor pressure, Plv is the corresponding vapor pressure at Tlv, the boiling temperature under normal conditions (Table 3 3) Generally, the term S varies from 0.7 0.8 to 1,185 the first range corresponding to a regime of developed surface evaporation, while the latter is commonly used for no surface re adsorption during evaporation. If S = 1 is assumed t he previous equation assumes that there is no vapor present in the ambient and no recondensation. Note that the previous equation assumes the Clausius -Clapeyron equation is applicable, i.e. it is a reasonable approximation for the temperature range Tlv < T < Tcr as long as no superheating occurs. If the actual pressur e in the liquid melt at temperature T is less than P (T ) the n the liquid melt is

PAGE 191

191 superheated. Conversely, t he same metastable condition will occur if the temperature of the surface is higher than the temperature corresponding to th e equilibrium vapor pressure P From Equation 7 2 1, the velocity of t he liquid vapor interface may be defined during surface vaporization by: l v l v zJ M V J t z u 0 (7 2 2) where Vl and l are the molar volume and the density of the li quid melt, respectively. If a substantial vapor pressure or plasma develops, the interface velocity is reduced due to backflow and condensation of solvent molecules. The representative curve in the solvents phase diagrams (Figure s 7 3 and 74) labeled v aporization with Knudsen layer shows how the saturated vapor pressure drops as a result of this Knudsen layer. Using Equation 7 2 2, the surface recession (solvent removed) by normal vaporization was calculated after 1 ns at different temperatures from Tlv to 0.9 Tc for the solvents studied in this work (Table 7 4 ). From the previous calculations, it follows that surface evaporation will be a significant mechanism for heat dissipation and material removal provided the temperature and time are sufficient. W hen the FEL irra diates a target and the resulting surface temperature exceeds Tlv, a few nanoseconds at such temperature will be sufficie nt to cause significant solvent removal from the target surface However, t he previous analysis assumes an equilibrium saturated vapor pressure at the surface. If the target material is superheated, the actual pressure at the surface will be considerably less than P (T ) thus decreasing the velocity of t he liquid -vapor interface. In fact, Song and Xu have shown t hat the surface pressure was far less than the press ure predicted by the Clausius Clapeyron equation.187 The surface temperature-pressure relation

PAGE 192

192 during laser ablation deviates significantly fro m the equilibrium relation during high-power laser irradiation.188 189 Note that for FEL induced surface evaporation, dendrimer removal from the target relies on the large macromolecules attaining sufficient kinetic energy from collective collisions with evaporating solvent molecules. Consequently, although the surface recession rates in Table 74 might be high, the dendrimer deposition rates will be considerably lower. N ormal vaporization is a very ineffici ent abl ation method for polymers via RIM -PLA; suc h efficiency can increase considerably upon dendrimer degradation (pyrolysis) into lower molecular weight species.4 7.8.2 Spontaneous Nucleation If a phase transition at equilibrium is kinetically limited, superheating may occur. A competing process that prevents superheating is spontaneous nucleation of the stable vapor phase within the metastable liquid. Spontaneous nucleation can occur (i) at preferential nucleation sites facilitating the cr eation of the bubb le (heterogeneous nucleation) or (ii) within the liquid melt if the free -energy necessary for the formation of critical -size nuclei is provided (homogeneous nucleation). T he eff ects of both mechanisms on the ablation of the solvent targe ts through heating past the binodal are now considered 7.8.2.1 Volume e vaporation (normal b oiling) Volume evaporation (also referred to as normal boiling) involves heterogeneous nucleation i.e. vapor bubble s which, in the case of liquids initiate heterog eneously from a variety of disturbances such as solvent impurities or an underlying solid surface. Once formed, the bubbles tend to diffuse and may, given enough time at a temperature above Tlv and escape out of the surface of liquid. Heterogeneous bubbl es may form (i) at the outer surface of the liquid, (ii) in the bulk of the liquid (aided by another surface) and (iii) at the under lying frozen solvent surface.

PAGE 193

193 Bubble diffusion may occur at the surface or inside the melted solvent. In general, a diffus ion coefficient may be defined for motion in all three directions as: 2 26 1R D (7 2 3 ) where is the jump frequency (s1), is the mean molecular spacing (may be approximated by the size of the molecule), and R is the individual RMS jump dis tance (in units of ). From this equation, surface and volume diffusion coefficients may be derived by making some reasonable assumptions .181, 190 The following analysis presupposes that the concentration of the dendrimer in the solvent (0.5 wt. %) is minimal; therefore a p ure solvent was assumed This assumption has been shown to be very reasonable .191 193 Surface d iffusion : In the surface diffusion mechanism the center of gravity of the bubble is randomly displaced due to the motion of solvent molecules at the target surface. Thus, the bubble diffusion coefficient ( Db s) depends on the solvents surface diffusion coefficient ( Ds) through the following relationship: s s b s b s bD r R D4 2 22 3 6 1 (7 2 4) where r is the bubble radius. Since there is no available data on the surfa ce self -diffusion coefficient for any of the two solvents used in this research, the bu bble diffusion coefficients could not be evaluated. However, upon consider ing heterogeneous nucleation, most bubbles can be assumed to be form ed within the volume as the normal vaporization rates in Table 7 4 are high. Possible nucleation sites within th e focal volume are (i) at the underlying frozen solid matrix (ii) at the interface with suspende d impurities, or (iii) at a solvent -dendrimer interface Since a ll solvents used were highly pure ACS grade significant nucleation at suspended impurities is not expected Additionally, due to the low concentration (0.5 wt. % ~2.4 x 104

PAGE 194

194 dendrimer molecules per solvent molecule ) of dendrimer, heterogeneous nucleation at the solvent -dendrimer interface is assumed negligible for the homogeneous solution The case of bubble diffusion inside the solvent volume is now considered. Volume d iffusion : The basic transport process in the volume diffusion mechanism of bubble motion might be described as molecular motion from one position of the bubble surface to another by diffusion in the surrounding bulk liquid.181, 190 F rom Equation 7 2 3, an analytical expression may be obtained for the bubble diffusion coef ficient in a volume diffusion mechanism ( Db vol): vol vol bD r D32 3 (7 2 5) In Equation 7 25, Dvol is the solvents vo lume self -diffusion coefficient and r is in units of rather than being a true length. This equation was used to evaluate the contri bution of volume evaporation (normal boiling) in FEL ablation by using Dvol values from the literature.193195 Table 7 5 includes the relevant parameters for the calculation of Db vol for both liquids as well as the values of t he mean diffusion distance for the FEL macropulse duration The calculations a ssume a value of r of 20, i.e. a bubble radius of 20 times the size of the solvent molecule. Table 7 5 co nsider s temperatures minimally higher than Tlv, a requirement for heterogeneous bubble formation in volume evaporation The calculation assumes a step -profile for the temper ature rise at the focal volume during the FEL pulse duration, i.e. immediate thermalization of the irradiance (as discussed in Section 7. 2). From these results, it follows that for an average optical penetration depth of 1020 m for chloroform and toluen e only the bubbles in the very top surface will diffuse to the surface Since most heterogeneous nucleat ion in the melted target occurs at the underlying solid matrix (Section 7.8.2.1.1) the diffusion of

PAGE 195

195 those deeply entrenched bubbles will be kinetical ly limited. Consequently volume evaporation can not be considered a primary mechanism leading to material removal from the melted target 7.8.2.2 Macroscopic sputtering via phase explosion (explosive b oiling) The following discussion concentrates on the c ase proposed initially by Martynyuk in 1974 when he laid down the principles of phase explosion through his studies from discharging a condenser into a metal wire.170172 The underlying premise of phase explosion or explosive boiling is that for heating rates from 108 to 1010 K/s, superheating of a solid or a liquid to a metastable state occurs.170172 As de termined in Section 7.5 t hese are relevant heating rates for the FE L macropulse when tuned to the selected resonant modes of chloroform and toluene At such high heating rates, m elting of the target may follow depending on the thermophysical properties of the solvent followed by norma l vaporization However, superheati ng of the solid or the melted liquid is more likely to occur until another, more intense transition into the vapor phase begins. This transition is known as phase explosion (explosive boiling) and occurs i f the resulting target temperature u pon FEL irradi ation is sufficiently close to the solvents critical temperature where the solvent is thermodynamically unstable (see Section 7.7 ). The vapor explosion within the metastable target liquid leads to the ejection of different -sized solid particles or liqui d droplets carrying dendrimer towards the target. In this ablation me chanism, rapid homogeneous nucleation and growth of vapor bubbles follows from density stratification (fluctuations ) in the superheated liquid. As a result the hot region near the surf ace breaks down in a very short time (~0.1 2 .0 s) into vapor plus equilibrium liquid droplets. Thermodynamics and kinetics of phase e xplosion : V apor bubbles via homogeneous nucleation can arise only in a superheated liquid.172, 196 Superheating is necessary for reducing the work required for critical bubble formation. In the usual process of heterogeneo us boiling, bubbles form and grow at preferential nucleation sites (previously dissolved bubbles, suspended

PAGE 196

196 impur ities, underlying solid surface ). As the work of formation of such embryos is small, heterogeneous boiling takes place under small superheating. The concentration of preferential nucleation sites in a pure liquid is comparatively small, so the intensity of such boiling is small. U nder a sufficiently high r ate of heating, the liquid evaporating through its free surface and through bubbles which formed at preferential nucleation sites as explained previously, is kinetically limited. Consequently, for an irradiated target that reaches temperatures close to the spinodal, homogeneous nucleation of stable vapor bubbles becomes an efficient energy dissipation process Under high superheating, homogeneous nucleation of vapor bubbles is possible at any point ins ide the liquid volume, primarily induced by density fluct uations (Equation 729) The foundations of homogeneous nucleation theory are found in the classic works of Volmer and Weber, Becker and D ring, Volmer, Zeldovich and Frenkel ; other s ignificant cont ributions to the theory were made by Kagan and Deryagin .197 198 The following discussion applies this theory to the solvents studied in this work The free energy associated with bubble formation is given by:199 L V B v L V fT k P r P P r r G 3 3 23 4 3 4 4 (7 2 6) where the first term describes the energy required for the formation of a liquid -vapor interface, the second shows the work necessary to overcome the pressure forces, and the third gives the chemical potential driving force for bubble for mation. In the previous equation, r is the bubble radius, is a surface tension (energy) term, T the saturation temperature of the liquid, and PV, PL the pressure inside the bubble (i.e. the pressure on the binodal at temperature T) and the ambient pressure of the liquid, respectively, and V and L are the chem ical po tentials of the different phases. Based on this equation, subcritical embryos dissolve spontaneously while supercritical

PAGE 197

197 embryos grow spontaneously. For small r the surface term dominates leading to a positive Gf, prohibiting bubble formation. Only for sufficiently large r G is negative The critical radius rcr for spontaneous growth is specified by satisfaction of mechanical equilibrium: cr L Vr P P2 (7 2 7) and thermodynamic equilibrium: sat V L LP P (7 2 8) where Psat is the saturation pressure of the liquid phase. It follows that the critical free energy for bubble formation is given by: 2 3 ,1 ) ( 3 16 L V L V cr fP P G (7 2 9) if the formation is induced by density fluctuations .172, 196 In Equation 7 29, L and V are the densities of the liquid and the vapor in the binodal. Assuming V << L, the last term inside the bracket may be approximated as 1. According to nucleation theory, for a stationary process, the frequency of homogeneous formation of vapor bubbles of critical size at temperature T is then equal to: 2 3 0 0) ( 3 16 exp expL V B B cr crP P T k J T k G J J (7 3 0) In Equation 7 30, 2 / 1 03 m SN JL (7 3 1) wh ere S is a coefficient that accounts for the possibility that nuclei larger than the critical size may decay and NL is the numbe r of liquid molecules per unit volume .172, 179180, 199200

PAGE 198

198 Assuming S = 1 (i.e. a small supercritical nuclei decay) Equation 7 30 can be used to calculate the rates of formation of critical size vap or bubbles within a superheated volume of the melted target Figure s 7 5 and 7 6 show the temperature dependence of Jcr for chloroform and toluene respectively, at two differ ent liquid pressures: (i) the average chamber pressure during deposition (103 t orr) and (ii) a chamber pressure with no vacuum (1 atm). The res ults of these calculations are listed in Table 7.6 The calculations show that the region of intensive homogeneous nucleation of vapor in the superheated liquid solvent is located in a narrow temperature interval, with thresholds4 (Tpe) between 0.81 0.85 Tr, close to previously reported values for the explosive boiling in metals.170172 Note that the threshold for chloroform is considerably lower (~ 70 degrees) than that of toluene. The maximum nucleation rates for both solvents do not vary and are close to ~1020 nuclei/cm3. These results are in the vicinity of the estimated maximum universal nucleation rate s of 1028 cm3 s1 and 1030 cm3 s1 calcula ted by Lienhard and Pavlov in separate works .180 Martynyuk also observed major fluctuations in volume and enthalpy at this temperature r ange, implying dramatic changes in heat capacity cP for metals.172 In the case of FEL irra diat ion, with increasing fluenc es (temperatures), due to the sharp decrease of and the increase of the ( PV PL)2 factor, Jcr increases sharply. At a sufficient degree of superheating, the number of interconnected bubbles and high pressure exerted by them result in a violent (supersonic beam like) material ejection from the target.180 In fact, molecular dynamics simulations164, 201 of thermally -confined laser irradiation (33 7 nm) of targets suggest that as the metastable liquid reaches 0.85 Tc the target spontaneously decomposes into a mixture of stable liquid and vapor. Consequently, depending 4 The threshold is defined as the temperature at which the nucleation rate becomes an integer.

PAGE 199

199 on the extent of penetration into the metastable region, superheated liquids ev olve very suddenly from a condition of apparent stability to one in which the new phase is formed very rapidly. This condition has been labeled in the previously discussed solvent phase diagrams as the phase explosion onset (Section 7.7). Process l eading to phase e xplosion s pinodal decay : The following discussion examine s the temperature T vs. entropy S phase diagram for a single component substance (Figure 7 5 ),172, 202 in order to understand the driving force for the high rate of homogeneous nucleation leading to phase explosion. In Figure 7 7 K corresponds to the solvents critical point, while curves kKl and mKn represent the binodal and the spinodal, respectively. The critical isobar Pc passes through the critical point. For subcritica l isobars, P < Pc the loci abcde and ace show different paths for transitions from liquid to vap or. For the latter locus the corresponding isotherm is Tlv, the normal (binodal) temperature for boiling. Ts is the temperature on the spinodal, while Sl a nd Sv are the entropies of the liquid and the vapor, respectively, at the boiling temperature Tlv. The region of superheated liquid is located in between the branch kK of the binodal and the branch mK of the spinodal. Conversely, the region of supersatur ated vapor is located in between the branches lK and nK The binodal kKl bounds the region of two -phase states. The isobar P > Pc shows a liquid vapor transition that occurs continuously without decay into two phases. In Figure 7 7 the spinodal bounds the existence region of unstable states. Thermodynamic instability of a phase is characterized by negative values of the following derivatives (stability coefficients): 0 TV P and 0 P Pc T S T (7 32, 7 33)

PAGE 200

200 In this metastable regio n, under a negative heat capacity cp, the thermal diffusivity T is also negative ( T = kT / cp). Therefore, the heat transfer equation also has a negative sign: 2 2 2 2 2 2z T y T x T t TT (7 34) With T < 0, instead of ordinary decreasing solutions descri bing the equalization of temperature, Equation 7 34 leads to exponentially increasing solutions. This means that, while in the metastable region, a local increase in the temperature in the unstable phase does not dissipate, but spontaneously increases. T his leads to the stratification of the metastable liquid into separate phases (spinodal decay ) with subsequent equalization of temperature after the limiting Kn is reached. The process of stratification of a system into two phases inside the spinodal is re lated to the motion of matter and with heat transfer. For a single -component system, the first process does not require molecular transport, and is realized by heat transfer through the densification and rarefication of the metastable liquid, therefore it is more rapid than mass transfer. In the case that the critical point refers to a binary mixture, the stratification is related to the separation of a homogeneous solution into two phases of different composition, thus the limiting process will not be he at transfer, but mass transfer (v ia molecular diffusion). In this work, at such a low concen tration (0.5 wt. %), a one component system is assumed and spinodal decay is expected to be limited by heat transfer. Heat transfer limited spinodal decay is a spo ntaneous (non activated) process that gives rise to a continuous non uniform system consisting of domains of dense and raref ied solvent (stratification), between which a large temperature difference T originates. As the surface tension is zero for the metastable liquid, the growth of these domains is not limited by surface

PAGE 201

201 partition. Under further development of the spinodal decay, the system gets out of the region of instability (reaches the super saturated vapor region line Kn in Figure 7 7 ). Spinodal decay is then replaced by the nucleation and growth of a new phase, vapor bubbl es, limited by surface partition For the system to reach equilibrium, homogeneous nucleation proceeds rapidly in ord er to decrease T within the domains of the liquid. In the case of the FEL melted targets, t he large number of interconnected bubbles builds sufficient pressure within the liquid, resulti ng in violent explosive boiling leading to dendrimer film deposition. Zeldovich and Todes202 estimated the time of stratification (spinodal decay) after consideration of the heat conduction equat ion and classical theory of capillarity near the critical point. The stratification is the fastest, the farthest from the critical point K, where | cp| ,203 204 and the smaller the critical dimension of the domai ns of different density Their analysis yielded an approximation for the order of magnitude of the time constant for stratification, based on the substances critical constants, as: ) ( 12T T P k Tc c T c sd (7 35) where is the liquids surface tension, kT is its thermal conductivity, while Tc and Pc are the critical parameters of the solvent. Conversely, Martynyuk suggests another expression for the time constant:172 T sdI 2 (7 36) where I is the characteristic dimension of the unstable phase domain. Spinodal d ecay and phase explosion from FEL pulsed heating : To understand the materi al removal (ablation) induced by pulsed laser heating, it i s necessary to consider the time scales relevant to spinodal decay and phase explosion: (i) the time of stratification (sd) and (ii)

PAGE 202

202 the time required for vapor embryo to grow to a critical nucleu s (n) which is referred to as the time lag for homogeneous nucleation. From Equation 7 35, the time constant for stratification at different temp eratures for the matrix solvents can be calculated (Table 7 7 ). The results in Table 7 7 show that the time c onstants for stratification are shorter than the FEL pulse duration ( 4 s). T herefore, spinodal decay is not kinetically l imited in FEL irradiation for either solvent. Spinodal decay in chloroform takes place one order of magnitude faster than for toluene i.e. for similar time lags for homogeneous nucleation ( n), explosive boiling in chloroform will occurs 10 times faster than in toluene. Equation 7 3 0 can be modified to account f or the time lag for homogeneous nucleation (phase explosion), n: t T k G J Jn B cr crexp exp0 (7 37) where t is the time duration at which the liquid remains superheated. The time lag has been estimated to be:199 2 2 / 14 2L V V nP P P RT M (7 38) where M is the molar mass, is the liquids surface tension, R is the universal gas constant, T is the temperature, and PV, PL the pressure inside the bubble (i.e. the pressure on the binodal at temperature T) and the ambient press ure of the liquid, respectively. Skripov performed cal culations based on Equation 7 3 8 for liquid metals and estimated the time lag to be approximately 1 10 ns.199 Applying Equation 7 3 8 to estimate the time lags for chloroform and toluene at different temperatures (around the th reshold (Tpe) described in Sectio n 7.8.2.2.1 ) yields Table 7 8 The calculations assume PL = 1 atm.

PAGE 203

203 The results from Table 7 8 show that th ere is no significant difference between solvents in the time lags before homogeneous nucleation proceeds following spinodal decay. Assuming the tar get stays at the high temperature for 100 ns, the time term in Equation 7 29 leads to 90% of the maximum nucleation rate (Jcr *) at that temperature T he time lag s for homogeneous nucleation are much less than the FEL macropulse length, therefore phase exp losion by FEL irradiation is not kinetically -limited. Moreover, phase explos io n will proceed three orders of magni tude faster than spinodal decay. 7.9 Correlation to Ablation Results The discussions in the previous sections have established the ground w or k for explaining the target and film characterization results described in Chapter 5 and 6 respectively. The calculations from Section 7.8.2.2.3 showed that phase explosion above threshold (Tpe) is limited by the spinodal decay, which occurs very rapidly (0.1 1 s). Therefore, target degradation (pyrolysis) is less likely to be obse rved for temperatures inside a focal volume exceeding Tpe as the thermalized energy is diss ipated almost instantaneously. In a fully or partially melted focal volume at t he temperature range Tlv T Tpe, dendrimer deposition will be limited to those macro molecules desorbing from the surface that have gain ed sufficient kinetic e nergy from collective collisions with the evaporating solvent molecules (Figure 7 8) However, as noted in Section 7.8. 1., normal vaporization is a ve ry inefficient ablation method; such effici ency can increase upon dendrimer degradation (pyrolysis) into lower molecular weight species.4 Any vapor nucleation in the melt at this temperature range will be through normal (heterogeneous) boiling; diffusion of such nuclei is kinetically limited during the laser pulse (Section 7.8.2.1.2) For temperatures T Tpe any surface evaporation will be accompanied by t he more violent phase explosion (Figure 7 8) as the maximum t hermal rise occurs at the surface of the focal volume (recall the FEL intensity distribution is Gaussian).

PAGE 204

204 In the case of no melting inside the focal volume at superheated temperatures T Tpe, homogeneous nucle ation of vapor (sublimation) can also occur in the frozen solid (Figure 7 8) The volumetric energy density required to ach ieve sublimation transitions for both solvents is o n the orde r of 102 J /cm3, which represents ~1% of the total deposited volumetric energy density. Therefore, sublimation inside the solid via homogeneous nucleation could lead to dendrimer spallation i.e. ejection of solid pieces of matrix The escape of the vap or bubbles from within the volume w ill be dependent on the mechanical p roperties of the solid particularly the fracture toughness ( KIC) of the frozen solvent. The fracture toughness describes the ability of a material containing a crack to resist fracture.205 206 A growing vapor bubble exerting pressure within the solid may be assumed to be a crack initiation point. No data is currently available in the l iterature for the fracture toughness of frozen chloroform or toluene. However, the mechanical properties of ice have been well documented and those values can be assumed to be in the range of those of other frozen solvents.207 Generally, the fracture toughness of ice is in the range of 50 150 kPa m1/2. By way of comparison, the fracture toughness of glass is typically 700 1000 kPa m1/2. T hus, ice has roughly one tenth the fracture toughness of g lass. A vapor bubble will form inside the frozen solvent if upon superheating the solid reaches Tpe. T he pressure exerted on the surrounding solid by the bubble will be in the range of ~ 8 M Pa at threshold, as calculated from the binodal (Clasius -Clapeyron equation).5 The critical crack length (lc) i.e. critical vapor nuclei size, t hat will cause brittle failure in the frozen solvent can be calculated:205 2 IC cK l (7 38) 5 This calculation assumes that the saturated vapor pressure is the same for both condensed phases: solid and liquid.

PAGE 205

205 where is a constant related to the crack geometry and is the applied stress, which is assumed to be the va por pressure inside the bubble. The values discussed above yield a critical crack length of 5 x 1015 m ; therefore, catastrophic failure will occur instantaneously from the initial stages of nucleation Consequently, the high pressure inside the homogeneously nucleated vapor bubbles will lead to solid spallation and large amounts of dendrimer being ejected towards the target (depending on the depth of the disturbance) It is pos sible that some frozen solvent agglomerates st rike the substrate; leaving small circular featur es as they re -melt on the previously deposited dendrimer. 7.9.1 Chloroform Results Recall from the con clusions of Sections 5.63 and 5.73 that all the chloroform targets showed the leas t deg radation for the two solvents as determined by NMR spectroscopy. Consequently, the majority of the FEL irradiated dendrimer must not have been exposed to extreme t em peratures long enough to observe molecular pyrolysis. The conclusions from Section 5 .63 and 5.73 also detail how all the chloroform films showed very dense dendrimer beads and strings for both resonance modes resulting in high roughness and thickness variations F rom Table 7 2 the peak temperature rises in the frozen state for the C H al kyl stretch and the C H bending mode s were 569 and 504 K, respectively. These temperatures exceed by far the critical threshold for phase explosion ( Tpe = 434 K) Therefore, explosive boiling should be the predominant mechanism for both modes. The previ ous section ha s also demonstrated that a vapor bubble in the frozen solvent may propagate and crack the solid solvent In fact, observation of the chloroform targets upon laser irradiation showed small pieces of frozen matrix being ejected for all chlorof orm depositions Although the focal spot size of the target is small

PAGE 206

206 (~100200 m) the fast rastering rate of the laser (Section 3.4.2.1) and the large scanned cross section of the target (Section 3.4.2.3) explain how this behavior could be observ ed The spinodal decay constants (times of stratification) (sd) and time lags of nucleation (n) calculated in Sectio n 7.8.2.2.3 showed than only ~100 ns are needed for homogeneous vapor nucleation to occur in chloroform. Consider ing this very short time scale, melting is likely to be kinetically limited, therefo re phase explosion should occur directly from the solid state. Moreover, the low density of circular patterns in the films support the conclusion that liq uiddroplet ejection is minimal; the small size of the circular features (5 15 m) occur from the re melting of solid matrix particles after they have adhered to the target substrate. This ablation mechanism explains the high ablation yields (fast and continuous material removal) and high roughness (thickness variations) for all the chloroform deposition s. It also explains the good luminescence of the chloroform ablated films, as the deposition is rapid enough such that pyrolysis of the dendrimer molecule is kinetically limited Since there is a temperature rise distribution in the focal volume, not all the irradiated target solvent will reach the threshold for phase explosion ( Tpe). Therefore, the deeper section of the focal volume will be exposed to temperatures that might lead to some molecular degradation. Consequently the NMR spectra of the entire target solution show a small degree of molecular alteration relative to the control (Section 5.6.3). However, most of the irradiated volume is successfully ablated, as supported by the almost constant proton counts as a function of irradiation time (Sect ion 5.6.3). The temperature rise calculations from Table 7 2 assume that all the irradiated en ergy (f luence) efficiently couples to the vibrational mode. However, the laser -mode couplin g efficiencies (Section 4.4 ) and the affected focal volumes (Se ction 4 .5 ) were different for all studied modes. Consequently, the effective irradiance, or more appropriately the effective

PAGE 207

207 volumetric energy density was different for each mode The effective volumetric energy densities for both modes of chloroform were calculated by normalizing the modes volumetric energy density with its laser -mode coupling efficiency (Table 7 9) The values in Table 7 9 show that the effective volumetric energy density of the C H alkyl stretch mode is ~3.9 times that of the C -H bending mod e, although both modes have similar absorption coefficients (peak temperature rises) in the solid state. C onsidering the variations in l aser -mode coupling efficiencies and affected focal volumes t he higher volumetric energy density c orrelates well with t he higher deposition rates (average film thickness ) of the C H stretch mode (Table 6.11). This correlation is more appropriate than simply comparing the lifetimes of the resonance modes (FWHM) as suggested by B ubb, et al. (Section 4.3.5);51 such analysis would lead to expect ~11 times higher deposition rates for the C -H alkyl stretch ablation. The increasing fluence (10, 20 and 30 J/cm2) results for the chloroform fil ms (Se ction 5.7.1 ) do not indicate a change in ablation mechanism from phase explosion. Considering Equation 7 8, an increase in fluence should lead to a proportional increase in the peak and subsurface temperatures. However, any calculated rise above the cri tical temperature of the solvent is meaningless as it is experimentally unattainable More importantly, at higher fluences much larger portion s of the focal volume s will attain temperatures above Tpe, resulting in higher ablation yields. The deeper sect ion s of the focal volume s above Tpe lead to more violent explosions ejecting larger amounts of material some of it melted Consequently, a higher density of circular (drying) patterns of larger sizes (Figure 6 31) is seen as the fluence is increased Th e similar PLQYs indicate little effect of increasing fluence on dendrimer degradation (Section 6.6)

PAGE 208

208 7.9.2 Toluene Results The discussion in Section 5.6.3 concluded that ablation of toluene targets resulted in considerable degradation of the dendrimer mole cule, as determined by NMR spectroscopy. Consequently a significant amount of the target must have been exposed to high enough temperatures to induce pyrolysis. The discussion in Section 6.3.2.3 also concluded that the films deposited from tolu ene were s ubstantially smoother and the deposition rates were on average ~10% of those of chloroform. These films also showed a high density of circular features from liquid droplets striking the substrate. From Table 7 2 the peak temperature rises in the frozen st ate for the C H stretch and the C=C aromatic stretch mode s were 283 and 592 K respectively. Only one of these temp eratures exceed the critical threshold for phase explosion in toluene ( Tpe = 503 K). Therefore, the C H stretch ablated targets may melt ( Tsl = 178 K) but not vaporize ( Tl v = 384 K) However upon melting, the absorption coefficient of this mode increases drastically resulting in a peak thermal rise of 2655 K, well above Tpe. Consequently, ablation for this mode occurs from the liquid state. T he average deposition rate for these films w as very low (~ 0.150.2 nm/s) indicating that normal vaporization is the likely mechanism for this mode. The high peak surface temperatures of the melted target allow for explosive boiling, which explains the high density of circular features resulting from droplets striking the substrate. In fact, observation of the chloroform targets upon laser irradiation showed melting at the FEL focal spot, which would refreeze as the FEL rastered away. Relative to chlor oform, the melting temperature of toluene is ~30 degrees lower ( Tsl = 178 K) and its solid-state specific heat capacity is 30% lower. Accordingly, it is not surprisin g that toluene melts rapidly The peak temperature rise for the C=C aromatic stretch mode of chloroform (592 K) initially suggest s that a blation occur s from the solid state which would indicate film properties

PAGE 209

209 similar to that of chloroform However, the average deposition rate at this mode was ~0.2 nm/s 0 = 1 J/cm2) and the films did not s how any significant difference from those ablated at the C H stretch mode. This supports the conclusion that ablation at this mode occurs from the liquid state after the target melts. T he absorption coefficient at this mode decreases considerably upon me lting, leading to lower peak temperatures (518 K) in the liquid. This peak surface temperature is only slightly above the phase explosion threshold ( Tpe = 503 K). Consequently, the ablat ion mechanism at the C=C aromatic stretch mode continued to be norma l vaporization with a small rate of phase explosion, evidenced by the circular features from liquid droplet strikes. As discussed in Section 6.3.2.2, the density of these features is much less than that of the C -H stretch ablated films. However, the feat ures are considerably larger (> 66%) indicating that the homogeneous nucleation rate is lower, but when formed, the vapor nuclei grow to much larger diameters than those formed at the C H stretch mode. The lower rate of liquid explosions explains the slig htly lower deposition rate fo r the C=C aromatic stretch mode, as phas e explosion contributes more to material deposition than normal vaporization. Relative to chloroform, the much lower deposition rates for both toluene modes can be explained by the charac teristic times for phase explosion, particularly the rate limiting time of stratification (spinodal decay) ( sd). The calculated times in Table 7 7 show that ~1 2 s are needed for spinodal decay to occur the precursor to vapor nucleation. Those values are one order of magnitude higher than those of chloroform and indicate that phase explosion will not occur until after about half of the macropulse (~4 s) has elapsed. Consequentl y, the efficiency for phase explosion in toluene is much lower than that of chloroform. Therefore, normal vaporization is the predominant mechanism for both toluene modes, which is a very inefficient method for dendrimer deposition and continuously subjects the material to extreme temperatures

PAGE 210

210 As discussed in Section 7.8.1, this efficiency can increase considerably upon dendrimer degradation (pyrolysis) into lower molecular weight species, wh ich explains the high degree of degradation seen for both toluene modes by NMR spectroscopy (Sections 5.6.3 and 5.7.3). This degradation coupled with the very low deposition rates also explains the minimal photoluminescence quantum yields s een for these films (Section 6.6). Similar to the analysis performed with chloroform, the effective volumetric energy dens ities for both toluene modes can be compared (Table 7 10). The values in Table 7 10 show that there is minimal change in the effective vo lumetric energy density, considering the variations in laser -mode coupling e fficiencies and affected focal volumes. These results correlate well with the minimal changes in deposition rates (average film thickn ess ) for both resonance modes (Table 6.11). The average deposition rates in toluene also correlate very well to the lifetimes of the resonance modes (FWHM) as suggested by Bubb, et al. (Section 4.3.5);51 such analysi s would lead to expect almost the sam e depositi on rates as the FWHM of the modes only vary by ~4% The increasing fluence (10, 20 and 30 J/cm2) results fo r the toluene films (Section 5.7.2) do not indicate a change in ablation mechanism from normal vaporiz ation, although phase explosion becomes more predominant as the fluence is increased. The corresponding increase in target sub -surface temperatures (above Tpe) leads to a higher rate of phase explosion, evidenced by the higher deposition rates and film ro ughness T he increase d rate of phase explosion is also confirmed by the higher density of circular (drying) features in the films (Figure 6 33) The size of these features increased proportionally to fluence a result of vapor nucleation and pressure bui ldup within deeper sections of the focal volume. A lthough phase explosion becomes more predominant as the fluence was increased, the targets did not show a decrease in mo lecular

PAGE 211

211 degradation (Section 5.6.3 ). Th is confirms that the time required for spinodal decay is long enough, which submits the dendrimer molecules to extreme temperatures lon g enough to induce pyrolysis. A ll quantum yields for these films were extremely low and assumed unreliable (Table 6.6 ). 7.10 Conclusion The model discuss ed in this c hapter explains the experimental results from RIM -PLA of electroluminescent dendrimers through a complete thermodynamic and kinetic analysis. Three principal mechanisms of thermally induced laser ablation are presented namely normal vaporization at the s urface, normal boiling, and phase explosion. During high power pulsedlaser irradiation, the latter mechanism based on homogeneous nucleation of vapor bubbles will be the most significant contribution to high deposition rates of intact luminescent dendrim er Phase explosion occurs when the solvent system is rapidly heated and approache s its critical temperature. Spontaneous density stratification within the condensed metastable phase leads to rapid homogeneous nucleation of vapor bubbles As the se vapor bubbles interconnect, large pressures buildup within the condensed phase leading to target explosions and dendrimer deposition at a near substrate Phase explosion is a temperature (fluence) threshold limited process, while su rface evaporation can occur from very low fluence doses.

PAGE 212

212 Figure 7 1. Possible vibrational energy relaxation mechanisms in solvent molecules (from Deak, et al):143 (a) the e nergy of the vibrational mode is lost to the crystal phonon bath characterized by a maximum fundamental frequency D; (b) the vibration (with energy ) decays by exciting a phonon of frequency of equal energy ph = ; (c) the vibration decays via simultaneous emission of several phonons (multiphonon emission); (d) the excited vibration decays via a ladder process, ex citing a lower energy vibration 1 and a small number of phonons; (e) intramolecular vibrational relaxation where many lower energy vibrations are simultaneously excited; (f) a vibrational cascade consisting of many steps down the vibrational ladder, each step involving the emission of a small number of phonons; the lowest energy vibration decays directly by exciting phonons. Figure 7 2 Typical phase diagram for a pure substance

PAGE 213

213 Figure 7 3. Liquid gas p hase diagram of pure chloroform. Figure 7 4. Liquid gas p hase diagram of pure toluene.

PAGE 214

214 Figure 7 5. Temperature dependence of the rate of homogeneous nucleation for metastable liquid chloroform at two different pressures. Figure 7 6. Temperature dependence of the rate of homogeneous nuclea tion for metastable liquid toluene at two different pressures.

PAGE 215

215 Figure 7 7. Temperature-entropy phase diagram of a single component material with three different isochores.172 Figure 7 8. Ablation scenarios at different temperature ranges for RIM -PLA targets. Table 7 T) for selected solvent resonance modes. Solvent Mode Mode peak ( m) T (s) Chloroform CH alkyl stretch 3.32 3.2 x 10 4 CH bending 8.28 2.1 x 102 Toluene CH stretch 3.31 3.3 x 103 C=C aromatic stretch 6.23 1.0 x 103

PAGE 216

216 Table 7 2. Peak temperatures in the frozen and melted solvent matrices for selected resonance modes in chloroform and toluene (1 J/cm2). Solvent Mode FEL peak (m ) Tmax (K) (solid) Tmax (K) (liquid) Chloroform CH alkyl stretch 3.31 569 412 CH bending 8.22 504 2655 Toluene CH stretch 3.31 283 551 C=C aromatic stretch 6.23 592 518 Table 7 3. Required volumetric energy density for a melting transition in the frozen targets. Solvent Volumetric energy density (J cm3) Chloroform 1.57 x 102 Toluene 2.13 x 102 Table 7 4. Calculated recession at the target surface from FEL -induced surface evaporation. Solvent Temperature (K) Surface evaporation: Recession after 1 ns (nm) Chloroform 334 ( T lv ) 0.22 375 (0.7 Tc) 0.68 429 (0.8 Tc) 2 .10 483 (0.9 Tc) 5.04 Toluene 384 ( Tlv) 0.19 414 (0.7 Tc) 0.39 473 (0.8 Tc) 1.21 533 (0.9 Tc) 2.90 Table 7 5. Volume diffusion coefficients and diffusion lengths for FEL pulse duration. Solvent Tlv (K) T for c alculation (K) Db vol x 10 5 (cm2 / s ) (Db vol t ) 1/2 (nm) for 4 s Chloroform 334 339 23.9 309 Toluene 384 388 43.6 417 *At saturation vapor pressure.193 194

PAGE 217

217 Table 7 6. Critical temperature thresholds ( Tpe) and maximum rates ( Jcr max) fo r homogeneous nucleation in the superheated solvents. Solvent Threshold ( Tpe) Jcr max (nuclei cm3) Chloroform 0.81 Tc (434 K) 4.99 x 10 20 Toluene 0.85 Tc (503 K)* 1.65 x 1020 These temperatures correspond to the intersection of the curve P/P0 with th at labeled phase explosion onset in the solvent phase diagrams (Figures 7 3 and 74) Table 7 7. Time of stratification ( sd) of superheated liquid solvents. Temperature Chloroform ( s) Toluene ( s) 0.8 Tc 0.10 0.75 0.85 Tc 0.13 1.00 0.9 Tc 0.20 1.50 Table 7 8. Time lag (n) for homogeneous nucleation of superheated matrix solvents. Temperature Chloroform (ns) Toluene (ns) 0.8 Tc 4.9 6.9 0.85 Tc 2.9 4.0 0.9 Tc 1.8 2.5 Table 7 9. FEL effective volumetric energy densities for solid -state chloroform (1 J cm2). Mode FEL peak ( m ) Effective volumetric energy density (J cm3) CH alkyl stretch 3.31 3.28 x 105 CH bending 8.22 8.45 x 104 Table 7 10. FEL effective volumetric energy densities for solid -state toluene (1 J cm2). Mode FEL peak ( m ) Eff ective volumetric energy density (J cm3) CH stretch 3.31 2.65 x 105 C=C aromatic stretch 6.23 2.14 x 105

PAGE 218

218 CHAPTER 8 CONCLUSIONS Research on pulsed laser abl ation for polymer deposition has considerably progressed over the last 25 years ; the relevant ablation (material removal) mechanisms for different laser sources and target compositions have been examined.12, 38 However, no full study had yet correlated the deposited fil m and target characteristics from Mat rix -Assisted Resonant Infrared Pulsed Laser Ablation (RIM -PLA) to all possible mechanisms for the dissipation of the deposited free electron laser (FEL) energy. The primary goal of this dissertation has been to shed more light into these relevant ablation mechanisms by studying different model solvent matrices ( resonance modes ) and polymer solutes. The use of therm odynamics and kinetic equations and a temperature rise model matched to the experimental results led to a consistent unders tanding of these mec hanisms. The semi -empirical model can be tran sposed to new frozen matrix -polymer systems. Selected vibrational -mode optical constants (absorption coefficients, optical penetration depths, mode widths) of two model solvent RIM -PLA matrices (chloroform and toluene) were measured in the liquid and solid state. T he results indicated that the optical constants ar e highly temperature-dependent, but are not altered ( 10 %) by dendrimer loading (0.5 weight %) Crystallization of chloroform also induces large ch anges in the mode spectra. The large FEL spectral bandwidths relative to the absorption modes result ed in high (~40 90%) offresonance irradiation, which lead to lower effective volumetric energy densities. A temperature rise model based on the time -depen de nt heat diffusion equation a nd the measured optical constants was used to calculate peak temperatures in the frozen and melted targets for the selected resonance modes These surface temperature rises correspond to heating rates on the order of 108109 K/s. Phase diagrams for the model solvents were constructed to

PAGE 219

219 understand the phase transitions occurring in the matrices as a result of these rises Calculations of required volumetric energy densities for melting and sublimation transitions indicated t hat FEL irradiation could result in equilibrium phase changes; however the high heating rates would lead to superheating in the matrix. Consequently, three possible mechanisms were proposed for thermally induced FEL ablation namely normal vaporization at the surface, normal boiling, and phase explosion. During high power pulsed FEL irradiation, the latter mechanism based on homogeneous nucleation of vapor bubbles will be the most significant contribution to ablation. Phase explosion occurs when the cond ensed target is rapidly heated (at the aforementioned rate s) and approaches the solvents critical temperature. A spontaneous d ensity stratification (spinodal decay) process will then occur within the condensed metastable phase leading to rapid homogeneo us nucleation and growth of vapor bubbles (system equilibration) As the se vapor bubbles coalesce large pressures build within the condensed phase resulting in solvent explosions and polymer deposition at a near substrate. Phase explosion thresholds and time lags for spinodal decay and homogeneous nucleation were estimated for the model ma trices, chloroform and toluene With these estimates, the contribution of this mechanism can be predicted. Phase explosion directly from the solid target (no melting) was shown to be the primary ablation mechanism for chloroform. The estimated peak temperatures by the model shows that the temperatures exceed the solvents phase explosion threshold ( Tpe = 434 K) by 70 and 135 degrees for the C H stretch and the C H bend ing modes, res pectively. The very high average film deposition rates for these modes (2 nm/s for C H stretch, 1 nm/s for C -H bending) are consistent with an efficient ablation mechanism. The film characteristics, deposition rates and roughness for each resonance mode were almost identical for the different polymers loaded in

PAGE 220

220 the solvent matrix. The high roughnesses, w hich increased with time, showed very large dense particles and beads of agglomerated dendrimer (as confirmed by high luminescence under U V microscopy). These high roughnesses were indicative of spallation of large quantities of polym er, which agglomerated on its way to the target. The low density of circular features and their small sizes (5 15 m) indicate that few liqui d droplets reach the target; those formed circular features were frozen particles that re-dissolve d some polymer upon melting at the substrate The deposition rates for both chloroform modes correlate well with the different effective FEL volumetric energy densities (i.e. normalized to the modes laser -mode coupling efficiency). The post ablated targets with a chloroform matrix showed slight degradation of the phosphorescent dendrimer upon comparison of the 1H NMR proton counts and peak shape s to a control. However, the relative degradation did not increase as a function of irradiation time or fluence. The almost constant amount of degraded dendrimer (i.e. that exposed to high temperatures below Tpe) retained in the target is consistent with the high ablation efficiency of phase explosion. The time lags for stratification ( spinodal decay) and homogeneous nucleation for chloroform sum to 100150 ns, a short time for pyrolysis of the dendrimer to be expected. Thicker films were deposited for longer irradiation times but s ince the temperature distribution in the focal volume wa s constant with time, the amount of degraded polymer in the target volume also remained constant. At higher fluences, the temperature distribution increase d, leading to deeper sections in the focal v olume reaching higher temperatures, some undergoing melting (as evidenced by the increase in circular features with fluence Figure 6 31). However, whether phase explosion occurs from the solid or the melt, the short time lags for chloroform continued to result in efficient ablation and almost constant relative amounts of degraded material in the

PAGE 221

221 target. Photoluminescence quantum yield measurements also showed that all chloroform film s retained significant luminescence (relative to the reported value) ,89 being consistent with the high ablation efficiency of phase explosion. At low temperatures (T > Tlv) when p hase explosion does not occur, the primary material removal mechanism will be normal vaporization at the target surface. This mechanism was predominant for FEL irradiation of toluene frozen matrices. 1H NMR c omparison of these f ilms with a control showed large changes in proton counts and peak shapes, indicative of high degradation of the phosphorescent dendrimer. The relative degradation increased with irradiation time and fluence, which is consistent with the inherent ineffici ency of this ablation mechanism That is, dendrimer deposition via this mechanism will be limited to th ose macromolecules desorbing from the surface, which have gained sufficient kinetic energy from collective collisions with evaporating solvent molecules. Consequently, the focal volume reaches high temperatures at which a high number of solvent molecules can escape the target, but few dendrimer molecules can. Although part of the focal volume may reach the phase explosion threshold ( Tpe), the long times lags for stratification (spinodal decay) and homogeneous nucleation (sum of ~1 2 s) are expected to on ly lead to phase explosion after melting in the target This is consistent with the high density of large circular features (50 -250 m diameter) indicating large droplet ejections from the target, which upon striking the substrate lead to previous film re -dissolution. Additi onally, the low deposition rates from the toluene matrices (0.150.2 nm/s about 1/10 of those of chloroform) are consistent with the ablation inefficiency of normal vaporization. Photoluminescence quantum yield measurements also sho wed that all toluene films had little luminescence (relative to the reported value),89 being consistent with the high dendrimer degradation of normal vaporization. The film characteristi cs, deposition rates

PAGE 222

222 and roughness for each resonance mode were almost identical for the different polymers loaded in the toluene matrix. This confirms that the RIM -PLA ablation characteristics at this polymer concentration were solely dictated by the sol vent and its optical constants. The same deposition rates for both toluene modes correlate well with similar effective FEL volumetric energy densities (i.e. normalized to the modes laser -mode coupling efficiency). The results in this study have show ed th at normal vaporization results in smoother film characteristics than those of phase explosion, from either a frozen or melted target. Although phase explosion resulted in the best molecular structural fide lity, the large dendrimer agglomerates deposited a nd the circular (drying) patterns in these films resulted in rough film topographies. From the standpoint of deposition of electroluminescent polymer layers in PLEDs, the lack of uniformity and excellent interfaces does not show good promise for RIM -PLA. However, the deposition technique may find promise for electrochromic coating technologies wh ere rough transparent films are used. Although it was possible to ameliorate the roughness issue with a post ablation anneal, future studies should examine the co nsequences of heating the substrate (above Tg) and altering its location relative to the target for improving film smoothness Additionally, the m aterial ejection dynamics can be further understood through plume shadowgraphy which would corroborate the v alidity of the estimated time lags for spinodal de cay and homogeneous nucleation. The temperature rise model can also be adapted to understand the distribution within the focal volume, i.e. perform calculations taking into account the Gaussian distribution of the FEL beam and Beer -Lamberts law The results from the model can also be c orroborate d by studying solvents of similar thermophysical properties in each phase i.e. decouple the effects from the matrix properties from its optical constants (ab soprt ion coefficients and lifetimes).

PAGE 223

223 Additionally, t he effects of the polymers molecular weight on the ablation characteristics can be further understood by expanding the studied range. Finally, the e ffects of low l aser -mode coupling efficiencies should be examined.

PAGE 224

224 REFERENCE LIST 1. C. W. Tang and S. A. Vanslyke, Applied Physics Letters 51 (12), 913915 (1987). 2. J. H. Burroughes, D. D. C. Bradley, A. R. Brown, R. N. Marks, K. Mackay, R. H. Friend, P. L. Burns and A. B. Holmes, Natur e 347 (6293), 539541 (1990). 3. L. S. Hung and C. H. Chen, Materials Science and Engineering R Reports 39 (5 6), 143222 (2002). 4. K. P. Gritsenko and A. M. Krasovsky, Chemical Reviews 103 (9), 36073649 (2003). 5. N. Matsumoto, H. Shima, T. Fujii and F. Kannari, Applied Physics Letters 71 (17), 24692471 (1997). 6. E. Ina, N. Matsumoto, E. Shikada and F. Kannari, Applied Surface Science 127, 574578 (1998). 7. M. A. Stevens, B. A. Weir, G. J. Denton and R. H. Friend, Synthetic Metals 101 (1 3), 234235 (1999). 8. J. M. Fitz Gerald, G. Jennings, R. Johnson and C. L. Fraser, Applied Physics A: Materials Science and Processing 80 (5), 11091112 (2005). 9. B. Toftmann, M. R. Papantonakis, R. C. Y. Auyeung, W. Kim, S. M. O'Malley, D. M. Bubb, J. S. Horwitz, J. Schou, P. M. Johansen and R. F. Haglund, Thin Solid Films 453454, 177181 (2004). 10. S. L. Johnson, C. T. Bowie, B. Ivanov, H. K. Park and R. F. Haglund, Proceedings of SPIE 6486, 64860G (2007). 11. R. Pate, K. R. Lantz and A. D. Stiff Roberts, IEEE J ournal of Selected Topics in Quantum Electronics 14 (4), 10221030 (2008). 12. D. B. Chrisey, A. Pique, R. A. McGill, J. S. Horwitz, B. R. Ringeisen, D. M. Bubb and P. K. Wu, Chemical Reviews 103 (2), 553576 (2003). 13. S. L. Johnson, K. E. Schriver, R. F Haglund and D. M. Bubb, Journal of Applied Physics 105 (2), 02490 (2009). 14. S. L. Johnson, Dissertation (Ph. D.), Vanderbilt University, 2008. 15. H. Letheby, Journal of the Chemical Society 15, 161163 (1862). 16. H. Shirakawa, E. J. Louis, A. G. MacD iarmid, C. K. Chiang and A. J. Heeger, Journal odf the Chemical Society Chemical Communications (16), 578580 (1977). 17. C. K. Chiang, C. R. Fincher, Y. W. Park, A. J. Heeger, H. Shirakawa, E. J. Louis, S. C. Gau and A. G. MacDiarmid, Physical Review Letters 39 (17), 10981101 (1977).

PAGE 225

225 18. A. J. Heeger, Reviews of Modern Physics 73 (3), 681700 (2001). 19. H. Shirakawa, Reviews of Modern Physics 73 (3), 713 718 (2001). 20. P. W. Atkins and J. De Paula, Physical chemistry 8th ed. (W.H. Freeman, New York, 2006). 21. M. Pope and C. E. Swenberg, Electronic processes in organic crystals and polymers 2nd ed. (Oxford University Press, New York, 1999). 22. A. J. Heeger, S. Kivelson, J. R. Schrieffer and W. P. Su, Reviews of Modern Physics 60 (3), 781850 (1988). 23. T. A. Skotheim, R. L. Elsenbaumer and J. R. Reynolds, Handbook of conducting polymers, 2nd ed. (M. Dekker, New York, 1998). 24. A. Sakamoto, Y. Furukawa and M. Tasumi, Journal of Physical Chemistry B 101 (10), 17261732 (1997). 25. S. V. Rakhmanova an d E. M. Conwell, Applied Physics Letters 75 (11), 15181520 (1999). 26. R. Kelly, Physica Status Solidi 21 (2), 451459 (1967). 27. J. L. Bredas and G. B. Street, Accounts of Chemical Research 18 (10), 309-315 (1985). 28. W. P. Su, J. R. Schrieffer and A. J. Heeger, Physical Review B 22 (4), 2099-2111 (1980). 29. J. Shinar, Organic light -emitting devices: a survey (AIP Press/Springer, New York, 2004). 30. F. C. Chen, G. F. He and Y. Yang, Applied Physics Letters 82 (7), 10061008 (2003). 31. M. A. Baldo, D F. O'Brien, Y. You, A. Shoustikov, S. Sibley, M. E. Thompson and S. R. Forrest, Nature 395 (6698), 151154 (1998). 32. M. A. Baldo, S. Lamansky, P. E. Burrows, M. E. Thompson and S. R. Forrest, Applied Physics Letters 75 (1), 4 6 (1999). 33. H. Yersin, H ighly efficient OLEDs with phosphorescent materials (Wiley -VCH, Weinheim, 2008). 34. W. Brutting, S. Berleb and A. G. Muckl, Organic Electronics 2 (1), 1 36 (2001). 35. B. Geffroy, P. Le Roy and C. Prat, Polymer International 55 (6), 572582 (2006). 36. Y Ohmori, A. Fujii, M. Uchida, C. Morishima and K. Yoshino, Applied Physics Letters 63 (14), 18711873 (1993).

PAGE 226

226 37. F. F. So and S. R. Forrest, Physical Review Letters 66 (20), 26492652 (1991). 38. R. Srinivasan and B. Braren, Chemical Reviews 89 (6), 13031316 (1989). 39. H. M. Smith and A. F. Turner, Applied Optics 4 (1), 147148 (1965). 40. R. Srinivasan and W. J. Leigh, Journal of the American Chemical Society 104 (24), 67846785 (1982). 41. R. Srinivasan and V. Maynebanton, Applied Physics Letters 41 (6), 576578 (1982). 42. A. Vogel and V. Venugopalan, Chemical Reviews 103 (2), 577644 (2003). 43. G. B. Blanchet, Applied Physics Letters 62 (5), 479481 (1993). 44. G. B. Blanchet, C. R. Fincher, C. L. Jackson, S. I. Shah and K. H. Gardner, Science 262 (5134), 719721 (1993). 45. G. B. Blanchet and S. I. Shah, Applied Physics Letters 62 (9), 10261028 (1993). 46. A. Pique, R. A. McGill, D. B. Chrisey, D. Leonhardt, T. E. Mslna, B. J. Spargo, J. H. Callahan, R. W. Vachet, R. Chung and M. A. Bucaro, Thin Solid Films 356, 536541 (1999). 47. R. Cristescu, D. Mihaiescu, G. Socol, I. Stamatin, I. N. Mihailescu and D. B. Chrisey, Applied Physics A: Materials Science and Processing 79 (4 6), 10231026 (2004). 48. A. Gutierrez Llorente, G. Horowitz, R. Perez -Caser o, J. Perriere, J. L. Fave, A. Yassar and C. Sant, Organic Electronics 5 (1 3), 29 34 (2004). 49. A. Pique, P. Wu, B. R. Ringeisen, D. M. Bubb, J. S. Melinger, R. A. McGill and D. B. Chrisey, Applied Surface Science 186 (1 4), 408415 (2002). 50. A. L. Mer cado, C. E. Allmond, J. G. Hoekstra and J. M. Fitz Gerald, Applied Physics A: Materials Science and Processing 81 (3), 591599 (2005). 51. D. M. Bubb, S. L. Johnson, R. Belmont, K. E. Schriver, R. F. Haglund, C. Antonacci and L. S. Yeung, Applied Physics A Materials Science and Processing 83 (1), 147151 (2006). 52. D. M. Bubb, J. S. Horwitz, R. A. McGill, D. B. Chrisey, M. R. Papantonakis, R. F. Haglund and B. Toftmann, Applied Physics Letters 79 (17), 28472849 (2001). 53. D. M. Bubb, J. S. Horwitz, J. H. Callahan, R. A. McGill, E. J. Houser, D. B. Chrisey, M. R. Papantonakis, R. F. Haglund, M. C. Galicia and A. Vertes, Journal of Vacuum Science and Technology A 19 (5), 26982702 (2001). 54. D. M. Bubb, M. R. Papantonakis, B. Toftmann, J. S. Horwitz, R. A. McGill, D. B. Chrisey and R. F. Haglund, Journal of Applied Physics 91 (12), 98099814 (2002).

PAGE 227

227 55. S. L. Johnson, K. Appavoo, H. K. Park and R. F. Haglund, Proceedings of SPIE 6879, 68790Y (2008). 56. I. Apitz and A. Vogel, Applied Physics A Materials Science and Processing 81 (2), 329338 (2005). 57. Z. D. Popovic and H. Aziz, IEEE Journal of Selected Topics in Quantum Electronics 8 (2), 362371 (2002). 58. H. Aziz and Z. D. Popovic, Chemistry of Materials 16 (23), 45224532 (2004). 59. Y. F. Liew, H. Aziz, N. X. Hu, H. S. O. Chan, G. Xu and Z. Popovic, Applied Physics Letters 77 (17), 26502652 (2000). 60. S. Y. Kim, K. Y. Kim, Y. H. Tak and J. L. Lee, Applied Physics Letters 89 (13), 132108 (2006). 61. S. F. Lim, W. Wang and S. J. Chua, Materials Sci ence and Engineering B Solid State Materials for Advanced Technology 85 (2 3), 154159 (2001). 62. H. Aziz, Z. Popovic, C. P. Tripp, N. X. Hu, A. M. Hor and G. Xu, Applied Physics Letters 72 (21), 26422644 (1998). 63. J. McElvain, H. Antoniadis, M. R. H ueschen, J. N. Miller, D. M. Roitman, J. R. Sheats and R. L. Moon, Journal of Applied Physics 80 (10), 60026007 (1996). 64. P. E. Burrows, V. Bulovic, S. R. Forrest, L. S. Sapochak, D. M. McCarty and M. E. Thompson, Applied Physics Letters 65 (23), 29222 924 (1994). 65. V. V. Jarikov and D. Y. Kondakov, Journal of Applied Physics 105 (3), 034905 (2009). 66. D. Y. Kondakov, Journal of Applied Physics 104 (8), 084520 (2008). 67. D. Y. Kondakov, W. C. Lenhart and W. F. Nichols, Journal of Applied Physics 101 (2), 024512 (2007). 68. Y. Kim, D. Choi, H. Lim and C. S. Ha, Applied Physics Letters 82 (14), 22002202 (2003). 69. S. T. Lee, Z. Q. Gao and L. S. Hung, Applied Physics Letters 75 (10), 14041406 (1999). 70. J. Shen, D. Wang, E. Langlois, W. A. Barrow, P. J. Green, C. W. Tang and J. Shi, Synthetic Metals 111, 233236 (2000). 71. D. C. Zou, M. Yahiro and T. Tsutsui, Synthetic Metals 91 (1 3), 191193 (1997). 72. M. Yahiro, D. C. Zou and T. Tsutsui, Synthetic Metals 111, 245247 (2000).

PAGE 228

228 73. T. Yamada, D. C. Zou, H. Jeong, Y. Akaki and T. Tsutsui, Synthetic Metals 111, 237240 (2000). 74. H. Aziz, Z. D. Popovic, N. X. Hu, A. M. Hor and G. Xu, Science 283 (5409), 19001902 (1999). 75. V. V. Jarikov, K. P. Klubek, L. S. Liao and C. T. Brown, Journal of Applied P hysics 104 (7), 074914 (2008). 76. S. Scholz, C. Corten, K. Walzer, D. Kuckling and K. Leo, Organic Electronics 8 (6), 709717 (2007). 77. R. Meerheim, S. Scholz, S. Olthof, G. Schwartz, S. Reineke, K. Walzer and K. Leo, Journal of Applied Physics 104 (1), 014510 (2008). 78. F. Papadimitrakopoulos, X. M. Zhang, D. L. Thomsen and K. A. Higginson, Chemistry of Materials 8 (7), 13631365 (1996). 79. J. C. Scott, J. H. Kaufman, P. J. Brock, R. DiPietro, J. Salem and J. A. Goitia, Journal of Applied Physics 79 (5), 27452751 (1996). 80. B. H. Cumpston, I. D. Parker and K. F. Jensen, Journal of Applied Physics 81 (8), 37163720 (1997). 81. M. Rajeswaran, T. N. Blanton, C. W. Tang, W. C. Lenhart, S. C. Switalski, D. J. Giesen, B. J. Antalek, T. D. Pawlik, D. Y. Kon dakov, N. Zumbulyadis and R. H. Young, Polyhedron 28 (4), 835843 (2009). 82. H. Aziz, Z. Popovic, S. Xie, A. M. Hor, N. X. Hu, C. Tripp and G. Xu, Applied Physics Letters 72 (7), 756758 (1998). 83. J. R. Gong, L. J. Wan, S. B. Lei, C. L. Bai, X. H. Zhang and S. T. Lee, Journal of Physical Chemistry B 109 (5), 16751682 (2005). 84. Y. Shirota, Journal of Materials Chemistry 15 (1), 75 93 (2005). 85. K. T. Wong, Y. L. Liao, Y. T. Lin, H. C. Su and C. C. Wu, Organic Letters 7 (23), 51315134 (2005). 86. M. H alim, J. N. G. Pillow, I. D. W. Samuel and P. L. Burn, Advanced Materials 11 (5), 371374 (1999). 87. J. M. Lupton, I. D. W. Samuel, R. Beavington, P. L. Burn and H. Bassler, Advanced Materials 13 (4), 258261 (2001). 88. O. Varnavski, I. D. W. Samuel, L. O. Palsson, R. Beavington, P. L. Burn and T. Goodson, Journal of Chemical Physics 116 (20), 88938903 (2002).

PAGE 229

229 89. S. C. Lo, E. B. Namdas, P. L. Burn and I. D. W. Samuel, Macromolecules 36 (26), 97219730 (2003). 90. L. O. Palsson, R. Beavington, M. J. Fram pton, J. M. Lupton, S. W. Magennis, J. P. J. Markham, J. N. G. Pillow, P. L. Burn and I. D. W. Samuel, Macromolecules 35 (21), 78917901 (2002). 91. S. C. Lo, N. A. H. Male, J. P. J. Markham, S. W. Magennis, P. L. Burn, O. V. Salata and I. D. W. Samuel, Ad vanced Materials 14 (13 14), 975979 (2002). 92. J. P. J. Markham, S. C. Lo, S. W. Magennis, P. L. Burn and I. D. W. Samuel, Applied Physics Letters 80 (15), 26452647 (2002). 93. D. R. Lide, CRC Handbook of Chemistry and Physics 86th ed. (CRC, Boca Raton 2005). 94. R. J. Bell, Introductory Fourier transform spectroscopy (Academic Press, New York, 1972). 95. B. C. Smith, Fundamentals of Fourier transform infrared spectroscopy (CRC Press, Boca Raton, 1996). 96. P. Jacquinot, Reports on Progress in Physic s 23, 268312 (1960). 97. C. A. Brau, Free -electron lasers (Academic Press, Boston, 1990). 98. T. C. Marshall, Free -electron lasers (Macmillan, New York, 1985). 99. P. Luchini and H. Motz, Undulators and free -electron lasers (Clarendon, Oxford, 1990). 1 00. E. L. Saldin, E. A. Schneidmiller and M. V. Yurkov, The physics of free electron lasers (Springer, New York, 2000). 101. C. A. Brau, Science 239 (4844), 11151121 (1988). 102. H. Motz, W. Thon and R. N. Whitehurst, Journal of Applied Physics 24 (7), 8 26833 (1953). 103. J. M. J. Madey, Journal of Applied Physics 42 (5), 19061913 (1971). 104. G. Dattoli, L. Giannessi, A. Torre, G. Altobelli and J. Gallardo, Journal of Applied Physics 79 (2), 570572 (1996). 105. F. Ciocci, G. Dattoli, L. Giannessi and P. L. Ottaviani, IEEE Journal of Quantum Electronics 33 (2), 147 151 (1997). 106. G. S. Edwards, D. Evertson, W. Gabella, R. Grant, T. L. King, J. Kozub, M. Mendenhall, J. Shen, R. Shores, S. Storms and R. H. Traeger, IEEE Journal of Selected Topics in Qua ntum Electronics 2 (4), 810817 (1996).

PAGE 230

230 107. M. A. Mackanos, J. A. Kozub and E. D. Jansen, Physics in Medicine and Biology 50 (8), 18711883 (2005). 108. M. A. Mackanos, J. A. Kozub, D. L. Hachey, K. M. Joos, D. L. Ellis and E. D. Jansen, Physics in Medicine and Biology 50 (8), 18851899 (2005). 109. D. Pandey, R. Reifenberger and R. Piner, Surface Science 602 (9), 16071613 (2008). 110. D. Raghavan, X. Gu, T. Nguyen, M. VanLandingham and A. Karim, Macromolecules 33 (7), 25732583 (2000). 111. P. Leclere, R Lazzaroni, J. L. Bredas, J. M. Yu, P. Dubois and R. Jerome, Langmuir 12 (18), 43174320 (1996). 112. L. O. Palsson and A. P. Monkman, Advanced Materials 14 (10), 757758 (2002). 113. D. M. Price, M. Reading, A. Hammiche and H. M. Pollock, International J ournal of Pharmaceutics 192 (1), 85 96 (1999). 114. K. F. Dziubek and A. Katrusiak, Journal of Physical Chemistry B 112 (38), 1200112009 (2008). 115. O. Yamamuro, I. Tsukushi, A. Lindqvist, S. Takahara, M. Ishikawa and T. Matsuo, Journal of Physical Chemi stry B 102 (9), 16051609 (1998). 116. V. M. Shulga, F. G. Eldarov, Y. A. Atanov and A. A. Kuyumchev, International Journal of Thermophysics 7 (6), 11471161 (1986). 117. D. Harrison and E. A. Moelwynhughes, Proceedings of the Royal Society of London Serie s A Mathematical and Physical Sciences 239 (1217), 230246 (1957). 118. O. I. Purskii, N. N. Zholonko and V. A. Konstantinov, Low Temperature Physics 29 (9 10), 771775 (2003). 119. D. W. Scott, W. T. Berg, J. F. Messerly, J. P. McCullough, S. S. Todd, I A. Hossenlopp and G. B. Guthrie, Journal of Physical Chemistry 66 (5), 911914 (1962). 120. O. I. Pursky and V. A. Konstantinov, European Physical Journal B 56 (3), 205 208 (2007). 121. A. Mariano, M. Postigo, D. Gonzalez Salgado and L. Romani, Journal o f Chemical Thermodynamics 39 (2), 218224 (2007). 122. J. George and N. V. Sastry, Journal of Chemical and Engineering Data 48 (4), 977 989 (2003). 123. A. Pal and R. K. Bhardwaj, Journal of Chemical Thermodynamics 34 (8), 11571171 (2002).

PAGE 231

231 124. M. J. P. M uringer, N. J. Trappeniers and S. N. Biswas, Physics and Chemistry of Liquids 14 (4), 273295 (1985). 125. H. L. Cupples, Journal of Physical Chemistry 50 (5), 412417 (1946). 126. R. Fanelli, Journal of the American Chemical Society 70 (5), 17921793 (1948). 127. M. C. Hannan and M. C. Markham, Journal of the American Chemical Society 71 (3), 11201120 (1949). 128. M. D. Lechner, Data extract from Landolt -Brnstein IV: Surface Tension of Pure Liquids and Binary Liquid Mixtures (Springer, Berlin Heidelberg 2008). 129. D. N. Sathyanarayana, Electronic Absorption Spectroscopy and Related Techniques (Orient Blackswan, 2001). 130. S. Xie, H. Li and B. Li, Chinese Optics Letters 1 (1), 44 46 (2003). 131. S. L. Jacques, Applied Optics 32 (13), 24472454 (1993). 132. C. G. Murphy and S. S. Alpert, American Journal of Physics 39 (7), 834836 (1971). 133. A. Samoc, Journal of Applied Physics 94 (9), 61676174 (2003). 134. C. D. Keefe, J. K. Pearson and A. MacDonald, Journal of Molecular Structure 655 (1), 6980 (2003). 135. H. El -Kashef, Optical Materials 20 (2), 81 86 (2002). 136. T. G. Goplen, D. G. Cameron and R. N. Jones, Applied Spectroscopy 34 (6), 657691 (1980). 137. A. Kimoto and H. Yamada, Bulletin of the Chemical Society of Japan 40 (1), 243 (1967). 138. E. A. D'Alessio, E. Dodero and C. Pomposiello, Journal of Chemical Physics 57 (10), 41364142 (1972). 139. B. Andrews, A. Anderson and B. Torrie, Chemical Physical Letters 104 (1), 65 70 (1984). 140. J. E. Bertie, R. N. Jones, Y. Apelblat and C. D. Keeke, Applied Spectroscopy 48 (1), 127143 (1994). 141. M. R. Anderson, Thesis (M. S.), University of Missouri -Rolla, 2000. 142. M. Budde, G. Lupke, C. P. Cheney, N. H. Tolk and L. C. Feldman, Physical Review Letters 85 (7), 14521455 (2000).

PAGE 232

232 143. J. C. Deak, L. K. Iwaki, S. T. Rhea and D. D. Dlott, Journal of Raman Spectroscopy 31 (4), 263274 (2000). 144. E. L. Sibert and R. Rey, Journal of Chemical Physics 116 (1), 237257 (2002). 145. V. E. Bondybey, Annual Review of Physical Chemistry 35, 591612 (1984). 146. W. M. Yen and D. D. Dlott, Laser spectroscopy of solids II (Springer -Verlag, Berlin ; New York, 1989). 147. L. F. Sutcu, H. W. White and J. L. Wragg, Surface Science 249 (1 3), L343-L346 (1991). 148. D. Stanila, W. Smith and A. Anderson, Spectroscopy Le tters 35 (5), 703 713 (2002). 149. Y. Takahashi, K. Shigematsu and T. Nishizaki, Journal of Crystal Growth 297 (2), 366371 (2006). 150. F. A. Cotton, Chemical applications of group theory 3rd ed. (Wiley, New York, 1990). 151. C. A. Angell, J. M. Sare and E. J. Sare, Journal of Physical Chemistry 82 (24), 26222629 (1978). 152. A. V. Lesikar, Journal of Physical Chemistry 80 (9), 10051011 (1976). 153. C. Alba, L. E. Busse, D. J. List and C. A. Angell, Journal of Chemical Physics 92 (1), 617624 (1990). 154. V. P. Smirnov, B. S. Razbirin, A. N. Starukhin, A. V. Chugreev, Y. S. Grushko, S. Kolesnik and B. P. Konstantinov, International Journal of Quantum Chemistry 100 (4), 448451 (2004). 155. N. E. Jacobsen, NMR spectroscopy explained: simplified theory, applications and examples for organic chemistry and structural biology (Wiley Interscience, Hoboken, N.J., 2007). 156. P. J. Hore, Nuclear magnetic resonance (Oxford UP, Oxford, 1995). 157. G. Socrates, Infrared and Raman characteristic group frequencies: tables and charts 3rd ed. (Wiley, Chichester, NY, 2001). 158. M. Snels, A. Beil, H. Hollenstein and M. Quack, Chemical Physics 225 (1 3), 107130 (1997). 159. M. Von Allmen and A. Blatter, Laser -beam interactions with materials: physical principles and applications 2nd updated ed. (Springer, Berlin, 1995). 160. G. Huf Desoyer, A. Asenbaum and M. Sedlacek, Journal of Chemical Physics 70 (11), 49244927 (1979).

PAGE 233

233 161. J. E. F. Rubio, V. G. Baonza, M. Taravillo, J. Nunez and M. Caceres, Journal of Chemical Phy sics 115 (10), 46814688 (2001). 162. H. Graener, R. Zurl and M. Hofmann, Journal of Physical Chemistry B 101 (10), 17451749 (1997). 163. H. Graener, R. Zurl, M. Bartel and G. Seifert, Journal of Molecular Liquids 84 (2), 161168 (2000). 164. L. V. Zhigil ei and B. J. Garrison, Journal of Applied Physics 88 (3), 12811298 (2000). 165. M. J. Crocker, Handbook of acoustics (J. Wiley, New York, 1998). 166. N. F. Otpushchennikov, Russian Physics Journal 8 (2), 120121 (1965). 167. U. S. Toti, M. Y. Kariduragan avar, M. I. Aralaguppi and T. M. Aminabhavi, Journal of Chemical and Engineering Data 45 (5), 920925 (2000). 168. D. M. Bubb, P. K. Wu, J. S. Horwitz, J. H. Callahan, M. Galicia, A. Vertes, R. A. McGill, E. J. Houser, B. R. Ringeisen and D. B. Chrisey, Jo urnal of Applied Physics 91 (4), 20552058 (2002). 169. P. Gordon, B. Balogh and B. Sinkovics, Microelectronics Reliability 47 (2 3), 347353 (2007). 170. M. M. Martynyuk, Combustion, Explosion, and Shock Waves 13 (2), 178191 (1977). 171. M. M. Martynyuk, Thermochimica Acta 206, 55 60 (1992). 172. M. M. Martynyuk, International Journal of Thermophysics 14 (3), 457470 (1993). 173. D. V. Ragone, Thermodynamics of materials (Wiley, New York, 1995). 174. A. Peterlongo, A. Miotello and R. Kelly, Physical Review E 50 (6), 47164727 (1994). 175. R. Kelly and A. Miotello, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 122 (3), 374400 (1997). 176. R. Kelly and R. W. Dreyfus, Surface Science 198 (1 2), 263 276 (1988). 177. J. Wisniak and K. Galon, Physics and Chemistry of Liquids 38 (6), 643661 (2000). 178. M. M. Martynyuk and P. A. Tamanga, Russian Journal of Physical Chemistry 74 (7), 10451049 (2000). 179. P. G. Debenedetti, Metastable liquids: concepts and principles (Princeton UP, Princeton, N.J., 1996).

PAGE 234

234 180. C. T. Avedisian, Journal of Physical and Chemical Reference Data 14 (3), 695 729 (1985). 181. A. Miotello and R. Kelly, Applied Physics A: Materials Science & Processing 69, S67 S73 (1999). 182. H. Fujii, K. Yoshino and Y. Inuishi, Journal of Physics D: Applied Physics 10 (14), 19751983 (1977). 183. L. J. Radziemski and D. A. Cremers, Laser -induced plasmas and applications (M. Dekker, New York, 1989). 184. E. N. Sobol, Phase transformations and ablation in laser -treated solids (Wiley, New York, 1995). 185. D. M. Zhang, D. Liu, Z. H. Li, L. Guan, X. Y. Tan, L. Li, R. R. Fang, D. Z. Hu and G. B. Liu, Applied Surface Science 253 (14), 61446148 (2007). 186. R. Kelly and A. Miotello, Applied Surface Science 96-8 205215 (1996). 187. K. H. Song and X. F. Xu, Applied Surface Science 127, 111116 (1998). 188. X. Xu and K. Song, Applied Physics A: Materials Science & Processing 69, S869S873 (1999). 189. X. F. Xu and K. H. Song, Materials Science & Engi neering. A Structural Materials: Properties, Microstructure and Processing 292 (2), 162168 (2000). 190. R. Kelly, Physica Status Solidi 21 (2), 451459 (1967). 191. E. Ilyina and H. Sillescu, Polymer 36 (1), 137141 (1995). 192. B. Nanagara, R. D. Oconn or and F. D. Blum, Journal of Physics Chemistry 96 (15), 64176423 (1992). 193. S. Pickup and F. D. Blum, Macromolecules 22 (10), 39613968 (1989). 194. F. X. Prielmeier and H. D. Ludemann, Molecular Physics 58 (3), 593604 (1986). 195. K. R. Harris, H. N. Lam, E. Raedt, A. J. Easteal, W. E. Price and L. A. Woolf, Molecular Physics 71 (6), 12051221 (1990). 196. V. P. Skripov, Journal of Non-Equilibrium Thermodynamics 17 (3), 193236 (1992). 197. V. G. Baidakov and V. P. Skripov, Experimental Thermal and Fl uid Science 5 (5), 664678 (1992). 198. M. Blander and J. L. Katz, AIChE Journal 21 (5), 833848 (1975).

PAGE 235

235 199. V. P. Skripov, Metastable liquids (J. Wiley, New York, 1973). 200. O. Kokkinaki and S. Georgiou, Digest Journal of Nanomaterials and Biostructure s 2 (2), 221241 (2007). 201. L. V. Zhigilei, E. Leveugle, B. J. Garrison, Y. G. Yingling and M. I. Zeifman, Chemical Reviews 103 (2), 321347 (2003). 202. Y. B. Zeldovich and O. M. Todes, Journal of Experimental and Theoretical Physics 10 (12), 14411445 (1940). 203. N. G. Polikhronidi, I. M. Abdulagatov, J. W. Magee and R. G. Batyrova, Journal of Chemical and Engineering Data 46 (5), 10641071 (2001). 204. N. G. Polikhronidi, I. M. Abdulagatov, J. W. Magee and G. V. Stepanov, International Journal of Ther mophysics 22 (1), 189200 (2001). 205. D. J. Green, An introduction to the mechanical properties of ceramics (Cambridge University Press, Cambridge; New York, 1998). 206. W. D. Callister, (John Wiley & Sons, New York, 2003). 207. J. J. Petrovic, Journal of Materials Science 38 (1), 1 6 (2003).

PAGE 236

236 BIOGRAPHICAL SKETCH Ricardo Daniel Torres Pag n was born in San Juan, Puerto Rico on December 26, 1980. He was raised in the municipality of Trujillo Alto and attended the Academia Nuestra Se ora de l a Providen cia, a Catholic school Until graduation from high school, Ricardo traveled multiple times to the U.S. to represent the island at different international science fairs and symposia. In 1998, he moved t o Cleveland, Ohio to attend Case Western Reserve Univ ersity after receiving a Provost Special Scholarship. As an undergraduate Ricardo was involved with several student organizations including the co-founding of the local chapter of the Society of Hispanic Professional Engineers. He also worked over a yea r fo r the General Electric Company at the Global Phosphors Division where he was part of the Edison Engineering Development Program. He graduated in May 2003 with a bachelor in Macromolecular Science and Engineering receiving th e Samuel P. Maron Outstand ing Research award at commencement. In August 2003 he joined Dr. Paul H. Holloways research group at the University of Florida. As a graduate student, Ricardo worked on several original ideas, one of them ultimately leading to his dissertation. He was also a n instructor for two semesters of introductory chemistry, which he considers one of his best experiences in graduate school. He graduate d in August 2009 with a Ph.D. in Materials Science and Engineering for his work on infrared laser ablation of lum inescent polymers.