USF Libraries
USF Digital Collections

Thermal analyses of hydrophilic polymers used in nanocomposites and biocompatible coatings

MISSING IMAGE

Material Information

Title:
Thermal analyses of hydrophilic polymers used in nanocomposites and biocompatible coatings
Physical Description:
Book
Language:
English
Creator:
Mohomed, Kadine
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Poly(2-hydroxyethyl methacrylate)
Poly(2 -- 3-dihydroxypropyl methacrylate)
Dielectric analysis
Hydrogel
Dissertations, Academic -- Chemistry -- Doctoral -- USF
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSRACT: This research focuses on two hydrophilic polymers that form hydrogels when they sorb water: Poly(2-hydroxyethyl methacrylate) (PHEMA) and Poly(2,3-dihydroxypropyl methacrylate) (PDHPMA). Present work in the field obviated the need to properly characterize the thermal and dielectric properties of these materials.The dielectric permittivity, e', and the loss factor, e", of dry poly(2-hydroxyethyl methacrylate) and poly(2,3-dihydroxypropyl methacrylate) were measured using a dielectric analyzer in the frequency range of 0.1Hz to 100 kHz and between the temperature range of -150 °C to 275°C. The dielectric response of the sub-Tg gamma transition of PHEMA has been widely studied before but little to no DEA data above 50°C is present in the literature. This study is the first to present the full range dielectric spectrum of PHEMA, PDHPMA and their random copolymers up to and above the glass transition region. The electric modulus formalism and several mathematical proofs were us ed to reveal the gamma, beta, alpha and conductivity relaxations. Dielectric analysis gives insight into the network structure of the polymer; it has been shown through thermal analyses that as the DHPMA content increased in HEMA-DHPMA copolymers the polymer matrix increased in available free volume and facilitated the movement of ions in its matrix. This is of significance as we then investigated the feasibility of using PHEMA, PDHPMA and their random copolymers as materials for a biocompatible coating for an implantable glucose sensor. The biocompatibility of hydrogels can be attributed to the low interfacial tension with biological fluids, high gas permeability, high diffusion of low molecular weight compounds, and reduced mechanical and frictional irritation to surrounding tissue. Once the biocompatibility of the hydrogels was established, the task to coat the polyurethane (PU)/epoxy coated metal glucose sensor was addressed. Plasma polymerization was found to be the most feasible^ technique for the application of the biocompatible hydrogel as a coating on the implantable glucose sensor. It has also been shown that thermal analysis techniques provide a mode of investigation that can be used to investigate the interfacial interactions of a novel hydroxylated, self-assembled nanoparticle with two functionally different polymers, poly(2-dihydroxyethyl methacrylate) and poly(methyl methacrylate).
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2006.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Kadine Mohomed.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 242 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001788184
oclc - 131042952
usfldc doi - E14-SFE0001442
usfldc handle - e14.1442
System ID:
SFS0025761:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Thermal Analyses of Hydrophilic Poly mers Used in Nanocomposites and Biocompatible Coatings by Kadine Mohomed A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Chemistry College of Arts and Sciences University of South Florida Major Professor: Ju lie P. Harmon, Ph.D. Mohamed Eddaoudi, Ph.D. Jerome E. Haky, Ph.D. Abdul Malik, Ph.D. Michael J. Zaworotko, Ph.D. Date of Approval: April 3, 2006 Keywords: poly(2-hydroxyethyl methacrylat e), poly(2,3-dihydroxyprop yl methacrylate), dielectric analysis, hydrogel Copyright 2006 Kadine Mohomed

PAGE 2

Dedication This work is dedi cated to my mom and dad. Thank you

PAGE 3

Acknowledgments I would like to extend my appreciation and thanks to my research advisor, Dr. Julie P. Harmon. Her support and guidance has helped me realize my potential as a research scientist. I would also like to thank my committee members: Dr. Abdul Malik (USF), Dr. Mohamed Eddaoudi (USF), Dr. Je rome Haky (FAU), and Dr. Michael J. Zaworotko (USF). Your input and guidance in the progress of my research has been invaluable. Special thanks to Dr. Haky w ho took time off his schedule to drive from Florida Atlantic University, Boca Raton, Fl to attend my annual progress meetings. I acknowledge Dr. Heba Abourahma, Dr. Michael J. Zaworotko, and John Perry for contributions to the nanoball project, Jessica Wilson, Dr. Poddar Pankaj and Dr. Hariharan Srikanth for their collaborations on the nanoIron project, and Dr. Francis Moussy, Dr Yvonne Moussy, Dr. Bazhang Y u, Dr. Nathan Long and Young Min Ju for their outstanding support on the biocompatible hydrogel coating for an implantable glucose sensor project. I am indebted to Dr. Patrick Benz and Dr. Jose Ors at Benz Research & Development (Sarasota, FL). I thank you for the generous supply of ultra high purity monomers and for the precious time spent discussing the possibilities of my work. I thank Lou Fierro at March Plasma (St. Petersburg, FL) for his help and contributions to the synthesis of the hydrogel plasma films. I am grateful to the Department of Ch emistry at USF for providing me with teaching assistantships, and to Dr. Julie Ha rmon and Dr. Francis Moussy for providing me with an NIH research assistantship (Grant # 5R01EB001640-02). I would like to acknowledge the great people at the chemistr y department: Mike Arias, Sam Valenti, Nina Goode, Roberto Avergonzado, Mike Piazza Cheryl Graham, John Seals, Sarala Rao and the many more I cannot name. I would like to thank my co-workers and friends from the Polymer Materials lab: Dr. Shelli S. Tatro, Dr. Lanetra M. Clayton, Dr. Timofey Gerasimov, Dr. Patricia Muisener, Yang Liu, Bernard “Butch” Knudsen, Krystal McCann, Chunyan Wang and Shisi Liu. Thank you for your support and encouragement. Finally, I would like to thank my friends and family, especially Fazir and Hanna, who have seen me through this chapter in my life.

PAGE 4

Note to Reader The original of this document contains color that is necessa ry for understanding the data. The original disse rtation is on file with the USF library in Tampa, Florida.

PAGE 5

i TABLE OF CONTENTS LIST OF TABLES v LIST OF FIGURES vi LIST OF ABBREVIATIONS xix ABSTRACT xx CHAPTER 1: INTRODUCTION 1 Biomaterials: Biocompatible Hydrogels 1 Polymer Nanocomposites 3 Thermal Analysis of Polymers and Polymer Composites 5 CHAPTER 2: POLYMER CHEMISTR Y AND INSTRUMENT THEORY 7 Polymer Synthesis 7 Free Radical Initiation 8 Gamma Radiation Initiation 10 Plasma Initiation 12 Instrumentation Theory 13 Thermal Analysis 14 Differential Scanning Calorimetry (DSC) 14 Thermogravimetric Analysis (TGA) 18 Dielectric Analysis (DEA) 21 Dynamic Mechanical Analysis (DMA) 30 Microhardness 34 Spectroscopy and Microscopy 35 Ultra Violet-Visible Spectroscopy (UV-Vis) 35 Scanning Electron Microscopy (SEM) 36 Transmission Electron Microscopy (TEM) 38 CHAPTER 3: A BROAD SPECTRUM ANALYSIS OF THE DIELECTRIC PROPERTIES OF POLY (2-HYDROXYETHYL METHACRYLATE) 39 Introduction 39 Poly (2-hydroxyethyl methacrylate) (PHEMA) 39 Dielectric Theory and Analysis 41 Experimental 43 Materials 43 Synthesis of PHEMA 44

PAGE 6

ii Differential Scanning Calorimetry 44 Thermogravimetric Analysis 45 Dynamic Mechanical Analysis 45 Dielectric Analysis 45 Results and Discussion 46 Polymerization Scheme for PHEMA 46 Differential Scanning Calorimetry and Thermogravimetric Analysis 47 Dynamic Mechanical Analysis 50 Relaxation 50 Relaxation 50 Dielectric Analysis 53 Relaxation 53 and Relaxation 59 Conductivity Relaxation 62 Conclusion 74 CHAPTER 4: NANOSTRUCTURE MATRIX INTERACTIONS IN METHACRYLATE COMPOSITES 75 Introduction 75 Polymer Nanocomposites 75 The Hydroxylated Nanoball 77 Experimental 80 Poly (2-hydroxyethyl methacrylate)-Nanoball Nanocomposites 80 Poly (methyl methacrylate)-Nanoball Nanocomposites 80 Differential Scanning Calorimetry 80 Dynamic Mechanical Analysis 82 Dielectric Analysis 82 Sample Molding 82 Microhardness 82 Soxhlet Extraction 83 UV-Vis Spectroscopy 83 Transmission Electron Microscopy 83 Results and Discussion 84 UV-Vis and TEM 84 Relationship between Glass Transi tion Temperature, Gel Fraction and Microhardness. 87 Dynamic Mechanical Anal ysis (DMA) of PMMA-Nanoball Nanocomposites. 98 Dielectric Analysis 103 Schematics of Proposed Polymer-Nanoball Interactions. 129 Conclusion 130

PAGE 7

iii CHAPTER 5: DIELECTRIC ANALYSES OF A SERIES OF POLY(2HYDROXYETHYL MET HACRYLATE-co-2,3DIHYDROXYPROPYL METHACRYLATE) 132 Introduction 132 Experimental 135 Materials 135 Synthesis of Poly(HEMA-co-DHPMA) Copolymer Series 136 Differential Scanning Calorimetry 136 Dielectric Analysis 136 Results and Discussion 137 Differential Scanning Calorimetry 137 Dielectric Analysis 144 Relaxation 144 and Relaxation 157 Conductivity Relaxation 165 Conclusion 183 CHAPTER 6: BIOCOMPATIBLE HYDROGEL COATING FOR AN IMPLANTABLE GLUCOSE SENSOR 184 Foreword 184 Introduction 184 Implantable Sensors 184 Microdialysis versus Electrochemical Sensors 186 Tissue Interactions with Implantable Sensors 189 BioMaterials 190 Experimental 191 Materials 191 Synthesis of UV-Polymerized Copolymers 192 Water Equilibrium Content, Gel Fraction and Biocompatibility Studies 192 Coating Trials: Dip coating, UV-Polymerization In Situ 193 Gamma Irradiation Grafting 194 Plasma Polymerization 194 Results and Discussion 196 Equilibrium and Biocompatibility Studies 196 Drawbacks of Dip Coating 201 Drawbacks of Gamma Irradiation Grafting 205 Plasma Polymerization of Ultra-Thin PHEMA Films 209 Conclusion 220 CHAPTER 7: SUMMARY A ND FUTURE WORK 222 REFERENCES 228

PAGE 8

iv APPENDICES 239 Appendix A: Chapter 3 240 Appendix B: Chapter 4 241 Appendix C: Chapter 5 242 ABOUT THE AUTHOR End Page

PAGE 9

v LIST OF TABLES Table 3.1 DEA vs. DMA for the transition. 54 Table 4.1 Glass transition temperature, gel fraction and Vickers hardness number of the polymer nanocomposites. 89 Table 4.2 Comparison of ac tivation energies of the transition for the PMMA nanocomposites as determined from DEA and DMA. 98 Table 4.3 DEA data: Activation energies for the transition for the PHEMA and PMMA nanocomposites. 106 Table 4.4 DEA data: Comparison of the dielectric constant, ', measured at 10Hz for the polymer-nanoball nanocomposites at 25, 100 and 125C. 119 Table 4.5 DEA data: Ionic conductivity and ionic conductivity activation energies for the polymer nanocomposites. 120 Table 5.1 DSC data: Glass transition temperature, Tg, of the HEMA-DHPMA copolymer series. 143 Table 5.2 DEA data: Activation energy and movement of the relaxation. 146 Table 5.3 DEA data: Activation energy of the relaxation. 159 Table 5.4 DEA data: Ionic conductiv ity activation energy. 177 Table 6.1 % Water equilibrium conten t of the HEMA-DHPMA copolymer series. 197 Table 6.2 Gel fraction of c opolymer series. 198 Table 6.3 Contact angle measurements of HEMA on glass and PU/epoxy coated glass surface. 203 Table 6.4 FTIR spectral band assignments for PHEMA. 207 Table 6.5 Contact angle measurements of water on glass, PU/epoxy coated glass surface and HEMA graft on PU/epoxy coated glass surface. 207

PAGE 10

vi Table 6.6 Plasma processes overview. 212 Table 6.7 Various conditions and prot ocols, successful and unsuccessful, used to plasma polymerize a thin film of PHEMA. 214 Table 6.8 Results of plasma polymeriza tion of PHEMA thin films using the various protocols and experimental conditions. 215

PAGE 11

vii LIST OF FIGURES Figure 2.1 Thermal decomposition of benzoyl peroxide (BPO) to the benzoyloxy free radical. 8 Figure 2.2 Thermal decomposition of ,-azobis(isobutyronitrile) (AIBN) to the dimethylcyano free radical. 8 Figure 2.3 UV decomposition of 2-hydr oxy-2-methyl-1-phenyl-1-propanone (Benacure 1173 , Mayzo). 8 Figure 2.4 Free radical initia tion of styrene monomer. 9 Figure 2.5 Propagation of a polystyrene chain. 9 Figure 2.6 Termination of a polymer chain via a) coupling and b) disproportionation. 10 Figure 2.7 A schematic of a JL Shepard Mark cesium-137 irradiator. 12 Figure 2.8 A schematic of a RF glow discharge plasma system. 13 Figure 2.9 Cross-section of a heat flux DSC cell. 15 Figure 2.10 Hermetically sealed DSC sample pan. 17 Figure 2.11 Polymer transitions as characterized by DSC. 18 Figure 2.12 A schematic of a TGA balance assembly. 20 Figure 2.13 Electrical phase sh ift signal response of a dielectric material. 22 Figure 2.14 Capacitance and conductive com ponents of the measured current. 23 Figure 2.15 Dipole and ion alignmen t in an electric field. 25 Figure 2.16 A schematic of parallel plat e sensor, ram, and furnace assembly. 27 Figure 2.17 A schematic of ceramic singl e surface sensor, ram, and furnace assembly. 28

PAGE 12

viii Figure 2.18 A plot of permittivity and loss factor versus temperature for Poly(methyl methacrylate). Data po ints were collected for various frequencies ranging from 1Hz-100kHz. 29 Figure 2.19 Mechanical response of ma terials. (a) Sinusoidal stress ( ) = sinusoidal strain ( ), (b) Perfectly elastic in-phase response, (c) Perfectly viscous out of pha se response, (d) Combinatorial visco-elastic response of polymeric materials. 30 Figure 2.20 Mechanical phase a ngle shifts for a polymer. 31 Figure 2.21 A conceptual diagram of st ored energy, E', vs. loss energy, E". 32 Figure 2.22 Dynamic mechanical tension film clamp. 33 Figure 2.23 A plot of the storage modul us and loss modulus of Poly(ethylene terephthalate) 33 Figure 2.24 Optical system of UV-V IS diode array spectrometer. 36 Figure 2.25 A schematic of a scanning electron microscope 37 Figure 3.1 Structure and relaxations in poly (2-hydroxyethyl methacrylate). 40 Figure 3.2 DSC data: Glass transition temperature, Tg, of neat PHEMA. 48 Figure 3.3 TGA data: Decomposition temperature of neat PHEMA 49 Figure 3.4 DMA data: Mechanical loss peaks at 1Hz for PHEMA. 51 Figure 3.5 DMA data: Mechanical loss p eaks at 1Hz for PHEMA.[Kolarik] 52 Figure 3.6 DMA data: Mechanical loss peaks at 1Hz for PHEMA. [Gates] 52 Figure 3.7 DEA data: Plot of permittivity ( ') versus temperature for PHEMA at various frequencies. 55 Figure 3.8 DEA data: Plot of loss factor ( ') versus temperature for PHEMA at various frequencies. 56 Figure 3.9 DEA data: Plot of electric loss modulus ( M ) versus temperature for PHEMA at various frequencies. 57 Figure 3.10 DEA data: Arrhenius plot of relaxation in PHEMA. 58

PAGE 13

ix Figure 3.11 DMA data: Arrhenius plot of relaxation in PHEMA. 58 Figure 3.12 DEA data: FrequencyTemperature dependence of the and relaxations in PHEMA. 60 Figure 3.13 DEA data: Dielectric loss functions of PHEMA at 6 kHz. 61 Figure 3.14 DEA data: Argand plot derived from the relaxation region (-110 C). 64 Figure 3.15 DEA data: Argand plot deri ved from the conductivity relaxation region (200 C). 65 Figure 3.16 DEA data: Dependence of M on frequency in the conductivity relaxation region (165 C). 67 Figure 3.17 DEA data: Dependence of M on frequency in the conductivity relaxation region (165 C). 68 Figure 3.18 DEA data: Dependence of M on frequency at a temperature below Tg (60 C). 69 Figure 3.19 DEA data: Dependence of M on frequency at a temperature below Tg (60 C). 69 Figure 3.20 Frequency dependence of ac conductivity for PHEMA at temperatures above Tg. 71 Figure 3.21 DEA data: Arrhenius plot of log dc conductivity vs the inverse of temperature. 72 Figure 3.22 DEA data: Arrhenius plot of frequency-temperature dependence of the conductivity M peak. 73 Figure 4.1 Structure of [(DMSO)(MeOH)Cu2(benzene-1,3-dicarboxylate-5-OH)2]12, a.k.a. the hydroxylated nanoball. 77 Figure 4.2 Square seconda ry building unit of [(DMSO)(MeOH)Cu2(benzene-1,3-dicarboxylate-5-OH)2]12, a.k.a. the hydroxylated nanoball. 78

PAGE 14

x Figure 4.3 Triangular sec ondary building unit of [(DMSO)(MeOH)Cu2(benzene-1,3-dicarboxylate-5-OH)2]12, a.k.a. the hydroxylated nanoball. 79 Figure 4.4 Calculated width a nd length of HEMA monomer. 79 Figure 4.5 In situ ultrasonic polymerization t echnique developed for the synthesis of the Poly(met hyl methacrylate)nanoball nanocomposites. 81 Figure 4.6 TEM images of a) HEMA-N anoball and b) Meth anol-Nanoball. 85 Figure 4.7 Optically transparent discs (1mm) of the PMMA-nanoball nanocomposites produced via in situ ultrasonic polymerization (1st three discs) and a sample of a 0.05% nanoball-PMMA composite produced by another method (4th disc). 85 Figure 4.8 PMMA-nanoball nanocomposite produced via in situ ultrasonic polymerization. 85 Figure 4.9 UV-VIS comparison of P MMA-Nanoball nanocomposite and PHEMA-Nanoball nanocomposite 86 Figure 4.10 DSC data: Glass transition temperature, Tg, of neat PMMA. 90 Figure 4.11 DSC data: Glass transition temperature, Tg, of 0.05% NanoballPMMA composite. 91 Figure 4.12 DSC data: Glass transition temperature, Tg, of 0.1 % NanoballPMMA composite. 92 Figure 4.13 DSC data: Glass transition temperature, Tg, of neat PHEMA. 93 Figure 4.14 DSC data: Glass transition temperature, Tg, of 0.1 % NanoballPHEMA composite. 94 Figure 4.15 DSC data: Glass transition temperature, Tg, of 0.5 % NanoballPHEMA composite. 95 Figure 4.16 DSC data: Glass transition temperature, Tg, of 0.9 % NanoballPHEMA composite. 96 Figure 4.17 DSC data: Glass transition temperature, Tg, of 1.5 % NanoballPHEMA composite. 97

PAGE 15

xi Figure 4.18 DMA data: Loss Modulus, E" vs. temperature for the PMMANanoball composites at 10Hz. 99 Figure 4.19 DMA data: Loss Modulus, E", vs. temperature for neat PMMA. 100 Figure 4.20 DMA data: Arrhenius plot of transition for neat PMMA. 100 Figure 4.21 DMA data: Loss Modulus, E" vs. temperature for 0.05% NanoballPMMA composite. 101 Figure 4.22 DMA data: Arrhenius plot of transition for 0.05% NanoballPMMA composite. 101 Figure 4.23 DMA data: Loss Modulus, E", vs. temperature for 0.1% NanoballPMMA composite. 102 Figure 4.24 DMA data: Arrhenius plot of transition for 0.1% NanoballPMMA composite. 102 Figure 4.25 DEA permittivity, ', and loss factor, ", of A) neat PHEMA and B) neat PMMA. 104 Figure 4.26 DEA data: Loss Factor, vs. temperature for neat PHEMA. 106 Figure 4.27 DEA data: Arrhenius plot of transition for neat PHEMA. 107 Figure 4.28 DEA data: Arrhenius plot of transition for neat PHEMA. 107 Figure 4.29 DEA data: Loss Factor, vs. temperature for 0.1% NanoballPHEMA composite. 108 Figure 4.30 DEA data: Arrhenius plot of transition for 0.1% NanoballPHEMA composite. 109 Figure 4.31 DEA data: Arrhenius plot of transition for 0.1% NanoballPHEMA composite. 109 Figure 4.32 DEA data: Loss Factor, vs. temperature for 0.5% NanoballPHEMA composite. 110 Figure 4.33 DEA data: Arrhenius plot of transition for 0.5% NanoballPHEMA composite. 111 Figure 4.34 DEA data: Arrhenius plot of transition for 0.5% NanoballPHEMA composite. 111

PAGE 16

xii Figure 4.35 DEA data: Loss Factor, vs. temperature for 0.9% NanoballPHEMA composite. 112 Figure 4.36 DEA data: Arrhenius plot of transition for 0.9% NanoballPHEMA composite. 113 Figure 4.37 DEA data: Arrhenius plot of transition for 0.9% NanoballPHEMA composite. 113 Figure 4.38 DEA data: Loss Factor, vs. temperature for 1.5% NanoballPHEMA composite. 114 Figure 4.39 DEA data: Arrhenius plot of transition for 1.5% NanoballPHEMA composite. 115 Figure 4.40 DEA data: Arrhenius plot of transition for 1.5% NanoballPHEMA composite. 115 Figure 4.41 DEA data: Loss Factor, vs. temperature for neat PMMA. 116 Figure 4.42 DEA data: Arrhenius plot of transition for neat PMMA. 116 Figure 4.43 DEA data: Loss Factor, vs. temperature for 0.05% NanoballPMMA composite. 117 Figure 4.44 DEA data: Arrhenius plot of transition for 0.05% NanoballPMMA composite. 117 Figure 4.45 DEA data: Loss Factor, vs. temperature for 0.1% NanoballPMMA composite. 118 Figure 4.46 DEA data: Arrhenius plot of transition for 0.1% NanoballPMMA composite. 118 Figure 4.47 DEA data:Frequency dependence of ac conductivity for neat PHEMA. 118 Figure 4.48 DEA data: Arrhenius plot of ionic conductivity activation energy for neat PHEMA. 121 Figure 4.49 DEA data: Frequency depe ndence of ac conductivity for 0.1% Nanoball-PHEMA composite. 122 Figure 4.50 DEA data: Arrhenius plot of ionic conductivity activation energy for 0.1% Nanoball-PHEMA composite. 122

PAGE 17

xiii Figure 4.51 DEA data: Frequency depe ndence of ac conductivity for 0.5% Nanoball-PHEMA composite. 123 Figure 4.52 DEA data: Arrhenius plot of ionic conductivity activation energy for 0.5% Nanoball-PHEMA composite. 123 Figure 4.53 DEA data: Frequency depe ndence of ac conductivity for 0.9% Nanoball-PHEMA composite. 124 Figure 4.54 DEA data: Arrhenius plot of ionic conductivity activation energy for 0.9% Nanoball-PHEMA composite. 124 Figure 4.55 DEA data: Frequency depe ndence of ac conductivity for 1.5% Nanoball-PHEMA composite. 125 Figure 4.56 DEA data: Arrhenius plot of ionic conductivity activation energy for 1.5% Nanoball-PHEMA composite. 125 Figure 4.57 DEA data: Frequency dependence of ac conductivity for neat PMMA. 126 Figure 4.58 DEA data: Arrhenius plot of ionic conductivity activation energy for neat PMMA. 126 Figure 4.59 DEA data: Frequency depe ndence of ac conductivity for 0.05% Nanoball-PMMA composite. 127 Figure 4.60 DEA data: Arrhenius plot of ionic conductivity activation energy for 0.05% Nanoball-PMMA composite. 127 Figure 4.61 DEA data: Frequency depe ndence of ac conductivity for 0.1% Nanoball-PMMA composite. 128 Figure 4.62 DEA data: Arrhenius plot of ionic conductivity activation energy for 0.1% Nanoball-PMMA composite. 128 Figure 4.63 A schematic of the plasticiz ation effect of nanoballs in PMMA. 129 Figure 4.64 A schematic of the crossli nking effect of nanoballs in PHEMA. 129 Figure 5.1 Chemical structure of a) 2-hydroxyethyl methacrylate (HEMA) and b) 2,3-dihydroxypropyl methacrylate (DHPMA). 133

PAGE 18

xiv Figure 5.2 A histology image of a HEMA -DHPMA copolymer subcutaneously implanted in an animal specimen. 134 Figure 5.3 DSC data: Glass tr ansition temperature, Tg, of neat PHEMA. 138 Figure 5.4 DSC data: Glass tr ansition temperature, Tg, of 75% HEMA: 25% DHPMA copolymer. 139 Figure 5.5 DSC data: Glass tr ansition temperature, Tg, of 50% HEMA: 50% DHPMA copolymer. 140 Figure 5.6 DSC data: Glass tr ansition temperature, Tg, of 25% HEMA: 75% DHPMA copolymer. 141 Figure 5.7 DSC data: Glass tr ansition temperature, Tg, of neat PDHPMA. 142 Figure 5.8 DSC data: Glass transition temperature dependency on HEMA content. 143 Figure 5.9 DEA data: Loss Modulus, E plot for neat PHEMA. 147 Figure 5.10 DEA data: Electric Loss Modulus, M plot for PHEMA. 148 Figure 5.11 DEA data: Arrhenius plot of transition for neat PHEMA. 148 Figure 5.12 DEA data: Loss Modulus, E plot for 75% HEMA: 25% DHPMA copolymer. 149 Figure 5.13 DEA data: Electric Loss Modulus, M plot for 75% HEMA: 25% DHPMA copolymer. 150 Figure 5.14 DEA data: Arrhenius plot of transition for 75% HEMA: 25% DHPMA copolymer. 150 Figure 5.15 DEA data: Loss Modulus, E plot for 50% HEMA: 50% DHPMA copolymer. 151 Figure 5.16 DEA data: Electric Loss Modulus, M plot for 50% HEMA: 50% DHPMA copolymer. 152 Figure 5.17 DEA data: Arrhenius plot of transition for 50% HEMA: 50% DHPMA copolymer. 152 Figure 5.18 DEA data: Loss Modulus, E plot for 25% HEMA: 75% DHPMA copolymer. 153

PAGE 19

xv Figure 5.19 DEA data: Electric Loss Modulus, M plot for 25% HEMA: 75% DHPMA copolymer. 154 Figure 5.20 DEA data: Arrhenius plot of transition for 25% HEMA: 75% DHPMA copolymer. 154 Figure 5.21 DEA data: Loss Modulus, E plot for neat PDHPMA. 155 Figure 5.22 DEA data: Electric Loss Modulus, M plot for neat PDHPMA. 156 Figure 5.23 DEA data: Arrhenius plot of transition for neat PDHPMA. 156 Figure 5.24 DEA data: Electric Loss Modulus, M vs. temperature for A) PHEMA homopolymer; B) 75%HEMA: 25% DHPMA copolymer; C) 25%HEMA: 75%DHPMA copolymer; and D) PDHPMA homopolymer. 158 Figure 5.25 DEA data: Comparison of M at 6000 Hz for PHEMA, PDHPMA and the copolymers. 159 Figure 5.26 DEA data: Frequencytemperature dependency of the and relaxations in neat PHEMA. 160 Figure 5.27 DEA data: Frequencytemperature dependency of the and relaxations in the 75% HEMA: 25% DHPMA copolymer. 161 Figure 5.28 DEA data: Frequencytemperature dependency of the relaxation in the 50% HEMA: 50% DHPMA copolymer. 162 Figure 5.29 DEA data: Frequencytemperature dependency of the relaxation in the 25% HEMA: 75% DHPMA copolymer. 163 Figure 5.30 DEA data: Frequencytemperature dependency of the relaxation in neat PDHPMA. 164 Figure 5.31 DEA data: Argand plot deri ved from the conductivity relaxation region (200 C) for neat PHEMA. 166 Figure 5.32 DEA data: Argand plot derive d from the glass transition region (100 C) for neat PHEMA. 166 Figure 5.33 DEA data: Argand plot deri ved from the conductivity relaxation region (200 C) for 75% HE MA: 25% DHPMA copolymer. 167

PAGE 20

xvi Figure 5.34 DEA data: Argand plot derive d from the glass transition region (100 C) for 75% HEMA: 25% DHPMA copolymer. 167 Figure 5.35 DEA data: Argand plot deri ved from the conductivity relaxation region (200 C) for 50% HE MA: 50% DHPMA copolymer. 168 Figure 5.36 DEA data: Argand plot derive d from the glass transition region (100 C) for 50% HEMA: 50% DHPMA copolymer. 168 Figure 5.37 DEA data: Argand plot deri ved from the conductivity relaxation region (200 C) for 25% HE MA: 75% DHPMA copolymer. 169 Figure 5.38 DEA data: Argand plot derive d from the glass transition region (100 C) for 25% HEMA: 75% DHPMA copolymer. 169 Figure 5.39 DEA data: Argand plot derive d from the conductivity relaxation region (200 C) for neat PDHPMA. 170 Figure 5.40 DEA data: Argand plot derive d from the glass transition region (100 C) for neat PDHPMA. 170 Figure 5.41 DEA data: Dependency of M' on frequency in the conductivity region for neat PHEMA. 172 Figure 5.42 DEA data: Dependency of M on frequency in the conductivity region for neat PHEMA. 172 Figure 5.43 DEA data: Dependency of M' on frequency in the conductivity region for 75% HEMA: 25% DHPMA copolymer. 173 Figure 5.44 DEA data: Dependency of M on frequency in the conductivity region for 75% HEMA: 25% DHPMA copolymer. 173 Figure 5.45 DEA data: Dependency of M' on frequency in the conductivity region for 50% HEMA: 50% DHPMA copolymer. 174 Figure 5.46 DEA data: Dependency of M on frequency in the conductivity region for 50% HEMA: 50% DHPMA copolymer. 174 Figure 5.47 DEA data: Dependency of M' on frequency in the conductivity region for 25% HEMA: 75% DHPMA copolymer. 175 Figure 5.48 DEA data: Dependency of M on frequency in the conductivity region for 25% HEMA: 75% DHPMA copolymer. 175

PAGE 21

xvii Figure 5.49 DEA data: Dependency of M' on frequency in the conductivity region for neat PDHPMA. 176 Figure 5.50 DEA data: Dependency of M on frequency in the conductivity region for neat PDHPMA. 176 Figure 5.51 DEA data: Frequency dependence of ac conductivity for neat PHEMA. 178 Figure 5.52 DEA data: Ionic conductivit y activation energy for PHEMA. 178 Figure 5.53 DEA data: Frequency dependence of ac conductivity for the 75% HEMA: 25% DHPMA copolymer. 179 Figure 5.54 DEA data: Ionic conductivit y activation energy for the 75% HEMA: 25% DHPMA copolymer. 179 Figure 5.55 DEA data: Frequency dependence of ac conductivity for the 50% HEMA: 50% DHPMA copolymer. 180 Figure 5.56 DEA data: Ionic conductivit y activation energy for the 50% HEMA: 50% DHPMA copolymer. 180 Figure 5.57 DEA data: Frequency dependence of ac conductivity for the 25% HEMA: 75% DHPMA copolymer. 181 Figure 5.58 DEA data: Ionic conductivit y activation energy for the 25% HEMA: 75% DHPMA copolymer. 181 Figure 5.59 DEA data: Frequency dependence of ac conductivity for neat PDHPMA. 182 Figure 5.60 DEA data: Ionic conductivity activation energy for neat PDHPMA. 182 Figure 6.1 Medtronic MiniMe d Guardian RT system. 186 Figure 6.2 A schematic of the Roche Micr odialysis System, a) Microdialysis probe implanted in subcutaneous adipose tissue, and b) Fluid being pumped to a glucose sensor outside the body. 187 Figure 6.3 A schematic diagram of the co il-type implantable electrochemical glucose sensor. 188 Figure 6.4 UV polymerized polymer rod with rounded, smooth edge. 197

PAGE 22

xviii Figure 6.5 Explantation of the subcut aneously implanted hydrogel rods. Forceps point to the area wh ere the hydrogels are located. 199 Figure 6.6 Histology image of PHEM A rod, explanted after 28 days. Dark purple outline indicates scar tissue formation (fibrosis). 199 Figure 6.7 Histology image of 80%HEMA: 20%DHPMA rod, explanted after 28 days. Minimal to no fibrosis. 200 Figure 6.8 Histology image of 100% PDHP MA rod, explanted after 28 days. Highly fragile sample fibrosis induced. 201 Figure 6.9 Contact angle measurement of HEMA on the PU/epoxy coated glass slide. 203 Figure 6.10 Dip coating, followed by UV polymerization. 204 Figure 6.11 Pipetting the HEMA mono mer unto the sensor, followed by UV polymerization. 205 Figure 6.12 FTIR spectrum of th ermally prepared PHEMA. 208 Figure 6.13 FTIR spectrum of PHEMA prepared from the irradiated HEMA/water solution. 209 Figure 6.14 A schematic of a RF electric discharge plasma system. 211 Figure 6.15 FTIR spectra of 1) conven tional PHEMA (red) and 2) plasma PHEMA, sample 6 (blue). 216 Figure 6.16 FTIR spectra of 1) conven tional PHEMA (red) and 2) plasma PHEMA, sample 7 (blue). 217 Figure 6.17 AFM image of plasma polymer film (sample 7). 217 Figure 6.18 AFM image and film thickness section analysis of plasma PHEMA film (sample 7). 218

PAGE 23

xix LIST OF ABBREVIATIONS Absolute permittivity of free space AC conductivity ac Activation energy Ea Angular frequency Avogadro’s number NA Benzene-1,3-dicarboxylate bdc Cohesive energy density CED Complex permittivity DC conductivity dc Dielectric analysis DEA Dielectric constant (permittivity) Dielectric Loss Factor Differential Scanning Calorimetry DSC Dynamic Mechanical Analysis DMA Dynamic Loss Modulus E Dynamic Storage Modulus E' Electric Loss Modulus M Gas constant R Gel Permeation Chromatography GPC Glass Transition Temperature Tg Interpenetrating Network IPN Maxwell-Wagner-Sillars MWS Methyl methacrylate MMA Monomethyl ether hydroquinone MEHQ Number Average Molecular Weight Mn Poly (2-hydroxyethyl methacrylate) PHEMA Poly (2,3-dihydroxypropyl methacrylate) PDHPMA Polydispersity Mw/Mn, PD Poly (methyl methacrylate) PMMA Scanning Electron Microscopy SEM Secondary building units SBU Thermogravimetric Analysis TGA Ultraviolet UV Ultraviolet Visible Spectroscopy UV-Vis Vickers Hardness Number HV Weight Average Molecular Weight Mw Williams-Landel-Ferry WLF

PAGE 24

xx Thermal Analyses of Hydrophilic Po lymers Used in Nanocomposites and Biocompatible Coatings Kadine Mohomed ABSTRACT This research focuses on two hydrophilic polymers that form hydrogels when they sorb water: Poly(2-hydroxyethyl methacryl ate) (PHEMA) and Poly(2,3-dihydroxypropyl methacrylate) (PDHPMA). Present work in the field obviated the need to properly characterize the thermal and dielectri c properties of these materials. The dielectric permittivity, ', and the loss factor, ", of dry poly(2-hydroxyethyl methacrylate) and poly(2,3-dihydroxypropyl me thacrylate) were measured using a dielectric analyzer in the frequency range of 0.1Hz to 100 kHz and between the temperature range of -150 C to 275 C The dielectric response of the sub-Tg transition of PHEMA has been widely studied befo re but little to no DEA data above 50 C is present in the literature This study is the first to present the full range dielectric spectrum of PHEMA, PDHPMA and thei r random copolymers up to an d above the glass transition region. The electric modulus formalism and se veral mathematical proofs were used to reveal the , and conductivity relaxations. Dielectr ic analysis gives insight into the network structure of the polym er; it has been shown through thermal analyses that as the DHPMA content increased in HEMA-DHPMA copolymers the polymer matrix increased in available free volume and facilitated the movement of ions in its matrix. This is of significance as we then inves tigated the feasibility of using PHEMA, PDHPMA and their random copolymers as mate rials for a biocompatible coating for an implantable glucose sensor. The biocompatibil ity of hydrogels can be attributed to the low interfacial tension with biological flui ds, high gas permeability, high diffusion of low molecular weight compounds, and reduced m echanical and fricti onal irritation to surrounding tissue. Once the biocompatibility of the hydrogels was established, the task

PAGE 25

xxi to coat the polyurethane (PU)/epoxy coated metal glucose sensor was addressed. Plasma polymerization was found to be the most feas ible technique for the application of the biocompatible hydrogel as a coating on the implantable glucose sensor. It has also been shown that thermal analysis techniques provide a mode of investigation that can be used to investig ate the interfacial in teractions of a novel hydroxylated, self-assembled nanoparticle with two functionally different polymers, poly(2-dihydroxyethyl methacrylate) and poly(methyl methacrylate).

PAGE 26

1 CHAPTER 1 Introduction Biomaterials: Biocompatible Hydrogels Hydrogels are materials that can sorb a nd retain a considerable amount of water within its structure without dissolving in water; it is a continuous three-dimensional network that is held together by chemical (covalent) or physical (non-covalent) bonds. [Gates 2003, Ratner and Hoffman 1976, LaPort e 1997]. Chemical gels are formed by the introduction of covalent crosslinks and they do not dissolve in organic solvents even upon the addition of heat; whereas, physical gels are held together by secondary molecular forces and they will eventually dissolve in solvents or melt upon the addition of heat [LaPorte 1997]. Natural hydrogel materi als include crosslinked gelatin and starch agar gel, but hydrogels can also be synthe thic. Synthetic hydrogels are crosslinked hydrophilic polymers that are characterized by solubilizing pendant groups (e.g., -OH, COOH, -CONH2) incorporated into the hydrogel st ructure. The high percentage of oxygen (O), either in the main chain of the pol ymer or in the pendent groups attached to the main chain, contributes to the hydrophili c nature of the polym er. Oxygen is strongly electronegative and, even after forming two covalent bonds, will stil l consists of two pairs of free electrons. These two pairs of electrons contribute to hydrogen bonding with neighboring molecules. When exposed to water the number of polymer-solvent interactions will be high, resu lting in solubility and coil ex pansion of the polymer chains. Nitrogen (N) also contributes to the hydrophi licity on a polymer via the same reasoning [LaPorte 1997]. Hydrogels have been found to be biocompa tible; they are soft, moist and flexible, and resemble in their physical properties livi ng tissue more so than any other biomaterial [Ratner and Hoffman 1976]. The biocompatibility of hydrogels can be attributed to the low interfacial tension with biological flui ds, high gas permeability, high diffusion of low molecular weight compounds, and reduced m echanical and fricti onal irritation to surrounding tissue [Gomez Ribelles et. al 1999, LaPorte 1997, Ratner and Hoffman 1976, Hench and Ethridge 1982, Shtilman 2003].

PAGE 27

2 Some common synthetic hydrogels being used in bioappli cations include: poly (2hydroxyethyl methacrylate (PHEMA), pol y(glyceryl methacrylate) (PGMA), poly(acrylamide) (PAAm), poly(N-vinyl-2-pyr rolidone) (PNVP), and poly(vinyl alcohol) (PVA). and have been used as materials in contact lenses and drug delivery capsules; other medical applications include dermal wound healing, and impl antation in the body of a human or animal patient to improve th e interfacial tissue interaction of medical implants [LaPorte 1997, Ratner and Hoffm an 1976, Hench and Ethridge 1982, Shtilman 2003]. A range of preparation techniques are us ed depending on the application for the hydrogel. This project investigates the use of a hydrogel material as a biocompatible coating for an implantable glucose sensor de vice. This coating should be permeable to allow glucose, oxygen and hydrogen peroxide to diffuse freely, reduce adsorption of protein from surrounding cell and plasma, result in minimal fibrosis by having an interface that is compatible with the tissue. In additi on, it should be non-toxic and physically stable in vivo The monomers investigated in this res earch include 2-hydroxyethyl methacrylate (HEMA) and 2,3-dihydroxypropyl methacrylate ( DHPMA) which were crosslinked with ethylene glycol dimethacrylate. DHPMA is also commonly known as glyceryl methacrylate (GMA). These monomers can be polymerized via free radical polymerization; their proper ties were investigated as homopolymers and as random copolymers of HEMA and DHPMA. Hydr oxyl containing hydr ogels (HEMA and DHPMA) were chosen over amide containi ng hydrogels (like NVP) since the hydroxyl group binds water stronger that the amide group. The water equilibrium content will be higher and its resistance to dehydration will be better. Because their role is solely as a surface coating material, the deposition technique chosen will be of major importance. Co mmon coating techniques for hydrogel coatings include: 1) dip coat in prepolymer and so lvent, 2) dip coat in monomer and then polymerize with catalyst and heat, 3) preactiv ate surface and then add monomer and heat to polymerize, and 4) irradiate substrate wh ile in contact with m onomer vapor or liquid solution of the monomer [Rattner and Hoffm an 1976]. For best results in terms of

PAGE 28

3 stability, the hydrogel film, or coating, should be covalently bonded to the substrate. This new composite which will be formed will possess the mechanical strength of the base substrate, but the composite material will ha ve the biocompatibility of the hydrogel. This is important since hydrogels tend to loose mech anical strength as water content increases and issues of delamination of the hydrogel co ating from the substrate can be avoided [Rattner and Hoffman 1976]. Various depositi on techniques that were investigated have been presented in chapter 6. Polymer Nanocomposites Polymer matrix composites have been studi ed and used commercially as early as the 1950’s [Kusy 1986]. Much effort has been placed on improving the mechanical, optical, electronic and magnetic properties of polymers by making polymer blends, and by adding fillers to the polymeric matrix [Varga et. al. 2003, Clay ton et. al. 2005, Wilson et. al. 2004]. In recent years, great stride s have been made to better understand the polymer-filler interface, to develop methods for enhancing interfacial adhesion and to characterize filler dispersion. Polymer nanocompo sites are of particular interest; due to the large interfacial area inherent of nanos cale fillers, polymer nanocomposites access new properties and exploit the unique synergis m between the matrix and filler [Chabert et. al. 2004]. Many techniques have been developed to disperse nanoparticles in polymeric matrices. Some techniques involve in situ and intercalation polymerization and in situ sol-gel, and other techniques involve disp ersion after polymerization, such as melt blending [O’Rourke Muisener et. al. 2002, Tatro et. al 2004, Xiong et. al. 2002, Park and Jana 2003, Chen et. al. 2001, Rong et. al. 2001, Park et.al. 2002] Each technique has its advantages and disadvantages. For instance, in situ ultrasonic polymerization developed in our laboratory, which involves sonication to break up and disperse the nanoparticles during polymerization is a tech nique that is difficult to sc ale-up for industrial production even though it produces good di spersion [Mohomed et. al. 20 05]. On the other hand, melt blending is a technique that has been su ccessfully used in large scale composite

PAGE 29

4 production but it has limitations in terms of its ability to separate the agglomeration clusters efficiently. Nanosized metal particles have properties that are different from those of macrosized bulk metals. Their size influences chemical, magnetic, optical and electronic properties [Carotenuto and Nicolais 2003, Kulkarni, John Thomas and Rao 2002]. Nanosizing also induces changes in the fundame ntal properties, such as the melting point and boiling point, as well as in the material’s shape and crystalline structure. For instance, bulk silicon does not emit light; however, nanosilicon emits light as a result of the quantum confinement effect which cause s a change in the materials optical gap [Luterov et. al. 2005]. Similarly, ferroma gnetic materials on the nanoscale show remarkably different properties especially wh en their particle size is less than a single domain size. Within this size range, the nanomagnetic particles show interesting dynamics and coercivity behavior. The incr eased surface to volume ratio influences changes in their high frequency properties, magnetic anisotro py etc. [Poddar et. al. 2005, Cattaruzza et. al. 1998]. The nanoparticle being investig ated in this study is of pa rticular interest. Due to its unique molecular structure it is the first-known reported nanoscale Kagom lattice to be synthesized by the pioneering research of Zaworotko and co-workers. The structure is made up of both square secondary buildi ng units (SBU) and tr iangular secondary building units. The open nanoporous ne twork is constructed using Cu( II ) dimers positioned at the lattice points which are bri dged using organic ligands. In the square SBUs, the moments of the individual dimers (a .k.a. the spin) cancel each other leading to antiferromagnetic coupling. The unique magnetic re sponse of this nanoparticle is directly related to the presen ce of the triangular SBU. The triangular SBU introduces spin frustration in the structure; whereby, a fe rromagnetic-like respons e leading to magnetic hysteresis is observed [Srikanth et al. 2003, Moulton et.al. 2002]. This nanoparticle and its counterparts have th e potential to be used in a variety of electromagnetic and drug delivery applicati ons. Its influence in a polymer matrix is important to study as the nanoparticle may be us eful as part of a coating or capsule. In this study (chapter 4), we examined the eff ects of the interactions taking place between a

PAGE 30

5 self-assembled nanostructure with two functionally different polymers: poly(2hydroxyethyl methacrylate) (PHEMA) and poly(methyl methacrylate) (PMMA). Thermal Analysis of Polymers and Polymer Composites Apart from knowing the chemical structure of a polymer, it is of great importance to know and understand the physical properties of polymers over a range of temperature and induced stress. The propertie s of a polymer material are determined by the structure, the additives and the processing conditions [Gedde 1995]. By understanding the behavior of polymers under various testing conditions, th e end-use application will be determined. A number of thermal analysis techniques, being able to measure and record structural changes unique to substances composed of large extended chain molecules, are particularly suited to the study of polymeric materials. For example, polymeric materials exhibit broad molecular weight distributions and viscoelast ic behavior and may contain both amorphous and crystalline regions within the same matrix. Thermal analysis is inclusive of several methods which have unique capabilities but which also overlap in their ability to provide a complete pictur e of a material's properties [Sepe 1995]. Polymers are first classified either as a thermoplastic or a thermoset. Thermoplastics are composed of linear or branched chains and can be molded; whereas, thermosets are crosslinked polymers that do not melt. They can then be divided into another subcategory of being either amorphous or semi-crystalline. Atactic and highly branched polymers are amorphous polymers; th e polymer chains are highly disordered. Amorphous polymers exhibit a gl ass transition temperature (Tg) which is the temperature at which the polymer loses its glasslike properties and assumes those more commonly identified with a rubber [Malcolm 1999]. Se mi-crystalline polymers show crystalline Bragg reflections, and consist of both cr ystalline and amorphous domains. Semicrystalline polymers exhibit a first-order thermodynamic melt and a very weak glass transition that depends on th e degree of crystallinity. The molecular relaxations of polymers are not limited to the glass transition and the melt transition; sub-Tg transitions exist in polymers th at have pendent groups attached to the main chains. The movement and rotati on of pendent groups off the main chain are

PAGE 31

6 termed secondary relaxations and they can be observed using thermal analysis techniques. By studying the behavior of the secondar y and primary relaxations in polymers, copolymers, polymer blends and composites, one can gain an understanding of the interfacial interactions, network structure and overall end use for the material. This study attempts to understand the thermal propert ies of the hydrophilic polymers, PHEMA and PDHPMA, in their use as biocompatible co atings and in nanocomposite materials.

PAGE 32

7 CHAPTER 2 Polymer Chemistry and Instrumentation Theory A brief introduction into vari ous aspects of polymer chemistry, in particular the synthesis of polymers via free radical polymeriz ation, will be presen ted in this section. This will then be followed by a description of the theory and operation of several techniques employed in this research in the characterization of polymer and polymer composites. This section has been deemed n ecessary to facilitate better understanding of the data presented in future chapters of this thesis. In the world of materials, civilization has progressed from utilizing simple wood and stone to the developmen t of metallurgy. Beginning in the early 1900’s scientists began synthesizing plastics which lead to th e birth of a new age. Since the 1950s, plastics have grown into a major industry that affect s all of our lives -from providing improved packaging and new textiles, to permitting th e production of wondrous new products and cutting edge technologies. Plastics even allow doctors to replace worn-out body parts, enabling people to live more productive and long er lives. In fact, since 1976, plastic has been the most used material in the world [Stevens 1990]. Plastics, elastomers, coatings and adhesives are some of the few classe s belonging to the gr oup of materials known polymers. Polymer Synthesis A synthetic polymer by definition is a larg e molecule made up of repeating units with a molecular weight of at least 100 tim es greater than that of the repeating unit [Seymour and Carraher, Jr. 1987]. A homopolym er is made up of one repeating unit; whereas, a copolymer is made of two or more repeating units. Polymers may be synthesized either by an addition polymer ization or a condensation polymerization reaction. In this section chai n-reaction addition polymerization will be considered using free radical initiation, gamma irradiation in itiation and plasma initiation of vinyl monomers.

PAGE 33

8 Free Radical Initiation A few monomers can polymerize on heati ng without the aide of an initiator; however, most monomers requir e an initiator to jump start the polymerization process. Free radical initiators can be, but are not limited to, peroxides, hydroperoxides, azo compounds such as azobis(isobutyronitrile) (A IBN), and benzoins such as 2-hydroxy-2methyl-1-phenyl-1-propanone (Benacure 1173 , Mayzo). Initiators can decompose to produce free radicals either thermally or photolytically. Figures 2.1, 2.2 and 2.3 show the decomposition of benzoyl peroxide (BPO), ,-azobis(isobutyronitr ile) (AIBN), and 2hydroxy-2-methyl-1-phenyl-1-propanone (B enacure 1173 , Mayzo), respectfully. O O O O O O 2 Figure 2.1. Thermal decomposition of benz oyl peroxide (BPO) to the benzoyloxy free radical [Bradley 1998]. N N N N N2N C + Figure 2.2. Thermal decomposition of ,-azobis(isobutyronitrile) (AIBN) to the dimethylcyano free radical [Bradley 1998]. O OH O OH + Figure 2.3. UV decomposition of 2-hydr oxy-2-methyl-1-phenyl-1-propanone (Benacure 1173 , Mayzo) [Bradley 1998].

PAGE 34

9 Once the initiator has decomposed to produce the free radical, the free radical reacts with a vinyl monomer or a strained-ri ng cyclic monomer to begin the initiation step of polymerization. Figure 2.4 shows the in itiation of styrene monomer using the benzoyloxy free radical. O O CH2 CH O O C + Figure 2.4. Free radical initia tion of styrene monomer. This step is then followed by a propagation step where the radical activated monomer reacts with a monomer unit to begin building the polymer chain. This step is shown in figure 2.5. O O C CH2 CH O O C + Figure 2.5. Propagation of a polystyrene chain.

PAGE 35

10CH2 C CH2 C CH CH CH2 C CH2 C CHCH CH2CH2 ^^^^ Y + ^^^^ Y ^^^^ ^^^^ Y Y ^^^^ Y + ^^^^ Y ^^^^ ^^^^ ^^^^ Y + Y a) Coupling b) Disproportionation Figure 2.6. Termination of a polymer chain vi a a) coupling and b) disproportionation. The final termination step occurs as two active chains react and this can occur by either coupling or disproportionation depe nding on the monomer(s) involved [Stevens 1999, Seymour and Carraher Jr. 1987]. For instan ce, the termination route for polystyrene occurs mostly through coupling where the molecular weight effectively doubles; whereas, the termination route for methacryl ates follows disproportionation where the molecular weight is unaffected [Stevens 1990]. Polymers produced via free radical polymerization can be made using different techniques; the most common techniques include bulk, suspension, so lution and emulsion. In this research, bulk free radical polymerization was utilized for its simplicity and the lack of contaminants usually added in the other techniques. The following initia tion procedures, gamma radiation and plasma polymerization, can be considered sub-categor ies of free radical polymerization as the processes involve the generation of free radicals in one way or the other. Gamma ( ) Irradiation Initiation High energy radiation, such as and particles, and x-rays, induces free radical polymerization [Stevens 1999]. In this research, irradiation was used for two specific purposes: sterilization and graft polymerizati on. The theory of radiation and its role in free radical polymerization will be discussed. A JL Shepherd Mark cesium-137 irradiator was used for this research (fi g. 2.7); the University of South Florida (USF)

PAGE 36

11 owns two JL Shepherd Mark cesium-137 irradiators that are maintained by the University of South Floridas Radiation and Safety Office. Cesium-137 is an unstable atom and decays to Barium-137. Its half life, T or T1/2, which is the time it takes to decay to ha lf the amount present, is 30.17 years. As it decays, a neutron in the nucleus changes in to a proton. To maintain the charge an electron is emitted as a beta ( ) particle. This particle is very small and is only able to penetrate only small thicknesses of tissue. The major issue with particles is that it causes secondary emissions known as Bremsstrahlung radiation. Bremsstrahlung radiation can be shielded using low atomic number materials. The Barium-137 produced is metastable and has a half life of 2.6 minut es; this entity becomes stable by emitting a gamma ( ) ray. The ray is very penetrating; however, if an absorber such as lead is used the total fraction of rays passing through an absorber decreases exponentially as the thickness of the absorber is increased. In order to monitor radi ation exposure, (CaF2/Mn) thermoluminescent ribbon dosimeters provided by the USF Radiation and Safety office were used. Background radiation, measured using a Ludham survey meter, is approximately 0.01mRad/hr and the radiation in front the shielded irradiator measures approximately 0.05mRad/hr. 55Cs137 56Ba137 (metastable) 56Ba (stable) Eq. 2.1 In order to calculate the dosage requ ired the following equation was used Tt o t oteNeNN6930 Eq. 2.2 where Nt is the number of nuclei remaining after a time interval, t, No is the number of nuclei at some original time and is the decay constant = 0.693/T When a monomer or polymer is bombarded by radiation the collision results in the ejection of an electron from the molecules where the reactive radical intermediate is formed [Jansen and Ellinghorst 1979, Park and Nho 2003, de Lange et al. 1994]. The radicals in the system can be made up of primary and secondary alkyl radicals (H2, CH2H-), and peroxy radicals (CH-O-O) if oxygen is present. These radicals will then initiate the polymerization process. Polymers exposed to irradiation result in scission

PAGE 37

12 and crosslinking of the polymer [Tatro 2002, Janik et. al. 2002, Skaja and Croll 2003, Kim and Urban 2000. Chain scission is the br eaking of a molecular bond causing the loss of a side group or shortening of the overall chain, and crosslinking is when individual polymer chains are linked together by cova lent bonds to form one insoluble network. However, usually one process dominates the ot her, and this is dependent on the polymer structure, atmosphere, temperature et c. [Tatro 2002, Clough and Shalaby 1996]. Figure 2.7. A schematic of a JL Shepard Mark cesium-137 irradiator. Plasma Polymerization Polymers can also be synthesized us ing plasma, which is an ionized gas containing ions, excited molecules and ener getic photons. Plasma can be generated by combustion, nuclear reaction, shocks and elec tric glow discharge. For the purpose of experiments conducted in this research, a RF electric glow discharge was utilized to produce the plasma at March Plasma Systems (St Petersburg, FL). Non-plasma forming gases such as Ar, Ne, N2 and O2 can be combined with vinyl monomers to produce polymer films which can potentially resemble polymers formed via

PAGE 38

13 conventional polymerization methods with th e added benefits of better adhesion and coating unto metal and glass s ubstrates. Figure 2.8 show a schematic of a plasma system. Such variables as reactivity of the monomer, monomer flow rate, frequency of excitation signal, discharge power and system pressure are some of the few factors affecting the deposition of a plasma film. Plasma polymerizat ion will be discussed in greater detail in chapter 6. Figure 2.8. A schematic of a RF gl ow discharge plasma system. [ www.astp.com AST Products, Inc.] Instrumentation Theory This section will briefly describe the theory and operation behind the major techniques employed throughout this researc h. These techniques include differential scanning calorimetry (DSC), thermogravim etric analysis (TGA), dynamic mechanical analysis (DMA), dielectric analysis (D EA), microhardness testing, UV-VIS spectroscopy, scanning electron microscopy (SEM) and tr ansmission electron microscopy (TEM). Techniques such as the DSC, DEA and DMA will be discussed in greater detail as these

PAGE 39

14 techniques are not as common as the othe r techniques. In add ition, the techniques’ application towards polymer charact erization will be discussed. Thermal Analysis Differential Scanning Calorimetry Differential scanning calorimetry (DSC) is a powerful thermal analysis technique used to measure the heat fluxes emitted or absorbed by a sample as a function of temperature and time. When a thermal transiti on occurs the enthalpy change is recorded. In addition to measuring the basic phase ch anges like the glass transition temperature (Tg) and melt temperature (Tm) other valuable quantitative properties can be determined. These include, but are not limited to, percent crystallinity, heats of crystallization and fusion in semi-amorphous polymers and or ganic-inorganic compounds, degree of cure and reaction kinetics in thermosets, oxi dative stability, thermal conductivity, decomposition and crosslinking [Stevens 1990, Gedde 1995, TA Instruments DSC 2910 2000]. DSC is a versatile technique that can be used for polymer, organic and inorganic analysis, of which the sample can be in the form of a solid, liquid or gel. Throughout this study a TA Instruments DS C 2910 with a standard cell was used. There are two types of DSC systems: the heat-flux DSC and the power-compensation DSC [Bershtein and Egorov 1994]. In the power-compensation DSC two individual heaters and temperature sensors are used; howev er, in the heat flux DSC one heat source is employed. The TA Instruments DSC follows the heat-flux design. Reference and sample pans are placed on raised platforms as shown in figure 2.9. The DSC cell is enclosed in a heating block which transfers he at to the reference and sample pans via a constantan disc. Two area thermocouples ma de at the junction of the chromel waferconstantan disc sit underneath the two plat forms. These thermocouples measure the differential heat flow between the reference and sample pans. Two other thermocouples formed at the junction of the chromel wa fer-alumel wire independently measure the sample and reference temperature. When a transition occurs the sample temperature will either lag behind the reference temperature for endothermic processe s, or surge for exothermic processes. The

PAGE 40

15 electrical power needed to k eep the temperature constant is recorded [Stevens 1990]. The TA Instruments DSC 2910 takes into consid eration only the differential heat flow measurement as shown in equation 2.3. DR T dt dQ Eq. 2.3 Where dQ/dt is the heat flow (W/g), T is the difference in temperature between the sample and reference and RD is the thermal resistance of the constantan disk [TA Instruments DSC 2910 2000]. Figure 2.9. Cross-section of a heat fl ux DSC cell. [TA Instruments DSC 2910 2000, Reprinted with prior permission from TA Instruments, Delaware] The basis of their measurement assu mes that the thermal resistance and capacitance of the sample and reference calorimeters are identical. These assumptions are exactly what they are: assumptions. The TA instruments DSC 2910 model does not take into consideration the heat capacity effects of the pan a nd calorimeter, nor the thermal resistance imbalance between the sample sens or and furnace and the reference sensor and furnace. A more accurate equation to determin e heat flow is shown in equation 2.4. This

PAGE 41

16 equation is used in the TA Instruments TzeroTM Q-series DSC models [TA Instruments TA-273]. d T d C d dT C C R R T R T qr s s r r s r 1 10 Eq. 2.4 Prior to data collection the DSC has to be calibrated to ensure accurate experimental results. Baseline slope, ce ll constant and temperature calibrations are performed. The baseline slope calibration involv es heating the empty DSC cell within the temperature range and heating rate needed fo r the experiment. The heat flow signal is measured. This heat flow signal should be zer o, since there is no sample in the cell, and it should have a slope of zero. The calibration program calculates the slope and offset values needed to flatten the baseline and ze ro the heat flow signa l [TA Instruments DSC 2910 2000]. The cell constant and temperature ca librations can be performed in one run. A pure metal, such as indium, tin, lead or zinc is sealed in a sample pan and heated to a temperature above its melting temperature, Tm. The experimental Tm is compared to the actual literature value Tm and the difference is calculated for the temperature calibration. A one point calibration is minimal; how ever, more standards of various Tm values can be used to perform a multi-point calibration. Th e last calibration involves determining the cell constant and onset slope. The cell constant is a ratio of th e calculated heat of fusion for the standard metal over the theoretical va lue. The thermal resistance is a measure of the temperature drop that occurs in a melting sample in rela tion to the thermocouple. The thermal resistance between these two points is calculated as the onset slope and is used for kinetic and purity calculations [TA Instruments DSC 2910 2000]. In DSC the sample (2-10mg) is placed in a sample pan. The empty sample pan should have a mass identical to that of the reference pan. These pans can be made of aluminum, gold, platinum, copper or graphi te; the material used will depend on the experimental conditions necessary. The sample and reference pans are hermitically sealed as shown in figure 2.10. and placed on the raised platforms inside the cell. The experiment is run under an inert atmosphere, using dry argon, helium or nitrogen gas with a flow rate of ca. 50ml/min. The DSC can be configured to run from sub-ambient

PAGE 42

17 temperatures with the aide of a liquid nitr ogen cooling accessory (a.k.a. LNCA) or the DSC cooling can. Experimental data collecti on may encompass a temperature range of 150 to 725C. Figure 2.10. Hermetically sealed DSC sa mple pan. [TA Instruments DSC 2910 2000, Reprinted with prior permission from TA Instruments, Delaware] The thermal properties of polymers are highly dependent on processing history. For polymers, the glass transition temperature, Tg, is usually taken from the second heat cycle of a heat-cool-heat regi men in which the sample is heated to a temperature beyond its Tg or Tm using a specific heating rate, cooled to at least 20 degrees below the Tg (quench-cooled or at a controlled ra te), and then reheated to above Tg using a known heating rate. This regimen is followed so that the Tg for each sample will have identical thermal histories. The value of the glass transition temperature is dependent on the heating rate, the manner in which the sample underwent annealing prio r to data collection and experimental conditions. It is important to always state the heating rate and cooling conditions used when reporting th e glass transition temperatur e, as well as whether the Tg was taken at the onset point or inflection poi nt of the transition. Figure 2.11 shows a DSC scan indicating the various thermal transitions recorded in polymers.

PAGE 43

18 Figure 2.11. Polymer transitions as charac terized by DSC. [TA Instruments DSC Brochure 2004] Thermogravimetric Analysis Thermogravimetric analysis (TGA) measur es the change in mass of a material either with respect to time, temperature or both. It is a useful te chnique which can be employed to determine the chemical and physical changes that induce a weight change in a material. Weight changes may occur as a result of such processes as decomposition, oxidation and dehydration. In a controlled atmos phere a sample can be heated at a known rate, or may be kept isothermally as a func tion of time. Information obtained from TGA data can be used to determine the percent weig ht change in a material as a description for thermal stability, the evolved gases can be used to correlate chemical structure if coupled with a mass spectrometer and composition de termination of metal inorganic-organic composites can be made. Like DSC, this t echnique is useful in providing intrinsic property information which can dictate enduse performance. Throughout this study a TA Instruments Hi-ResTM TGA 2950 was used to assess thermal stability, water and inorganic metal composition. Very often, in the dynamic TGA mode decomposition transitions overlap due to the time-dependent nature of the reaction. This can be partially resolved by using a very

PAGE 44

19 slow heating rate or by increasing the temp erature of the furnace to the decomposition temperature and then holding the temperat ure isothermally until the transition is complete, followed by raising the temperatur e again until the next transition. The high resolution (Hi-ResTM) option for the TGA provides enha nced transition resolution and faster scans. It is a comb ination of both dynamic and isothermal thermogravimetry. The Hi-ResTM option automatically makes the adjust ments in the procedure to optimize weight change resolution. The TA Instruments Hi-ResTM TGA 2950 operates on a ze ro (null) balance principle. The sample pan made of platinum aluminum or ceramic, is hung in place by a hang-down wire which is attached to the balance arm as shown in figure 2.12. The balance arm is maintained in a reference pos ition by an optically actuated servo loop. A balance meter movement is used to physic ally keep the balan ce in a null position. The null position is dictated by a constant curre nt infrared light emitting diode (LED) and two photosensitive diodes. When the balance is in a null position, a flag located on top of the balance arm blocks an equal amount of light to each of the photodiodes. Mass changes in the sample due to such processes as deco mposition, oxidation or dehydration cause an unbalanced amount of light to hit the photodiode s. The instrument compensates for this by supplying current to the meter movement so that it can move back into its original reference position. The change in current necessary to accomp lish this task is directly proportional to the change in mass of the sample This current is conv erted to the weight signal [TA Instruments TGA 2950 2000].

PAGE 45

20 Figure 2.12. A schematic of a TGA balance assembly. [TA Instruments TGA 2950 2000, Reprinted with prior permission from TA Instruments, Delaware] The TGA covers a temperature range of 25 to 1000 C, and has two possible weight ranges from 1 g to 1000mg and 0.1 g to 100mg. An inert purge gas is used to remove the evolved gases to prevent diffusi on and contamination of these evolved gases in the balance chamber. Prior to collecting da ta in the TGA certain calibration steps must be performed to ensure accurate results. Th ese include temperature, weight and sample platform calibrations. To perform a temperature cal ibration of the TGA, the curie temperature of a high purity magnetic standard is determined, and then compared to the correct value. A temperature calibration table is construc ted in which the observed and correct temperatures entered correspond to the expe rimental and theoretical transition curie temperatures of the calibration standard. From one to five temperature calibration points can be entered in the calibration table. As in DSC, a multiple-point calibration is more accurate than a one-point calibration. The weight calibration calibrates both the 100 mg and 1 gram weight ranges and stores the calibration parameters internally in the instrument. The last calibration, the sample platform adjustment, is performed if the

PAGE 46

21 sample hang-down wire fails to pick up th e sample pan during an automatic loading procedure [TA Instruments TGA 2950 2000]. TGA is not only used to determine polymer stability but may be used to characterize percent composition of copolymers and fillers, the effect of additives, as well as the volatility of plastic izers and diluents present in the polymer. The TGA can be coupled with DSC to better characterize the processes taking place in the polymer, such as water loss and high temperature melt transitions [TA Inst ruments TGA Brochure 2004]. Dielectric Analysis Dielectric analysis (DEA) is a technique used to determine the molecular motions and structural relaxations present in materials possessing permanent dipole moments [McCrum, Read and Williams 1967, Avakian, Starkweather, Jr. and Kampert 2002]. It measures the electrical response of a materi al with respect to time, temperature and frequency. Unlike DSC, DEA can be used to id entify the secondary relaxations present in a polymer as long as the secondary group has a net dipole moment. DE A can also be used to monitor cure kinetics, resi n flow and ionic conductivity. In dielectric measurements the material is exposed to an alternating electric field which is generated by applying a sinusoidal voltage; this process causes alignment of dipoles in the material which results in polarization. The polarization will cause the output current to lag behind the applied electric field by a phase shift angle, as shown in figure 2.13. The magnitude of the phase sh ift angle is determined by measuring the resulting current. The capacitance and c onductance are then calculated from the relationship between the app lied voltage, measured curr ent and phase shift angle [McCrum, Read and Williams 1967, Avakian, Starkweather, Jr. and Kampert 2002, TA Instruments DEA 2970 Diel ectric Analyzer ].

PAGE 47

22 Figure 2.13. Electrical phase shif t signal response of a dielectric material. [TA Instruments DEA 2970 1997, Reprinted with pr ior permission from TA Instruments, Delaware] The electrical properties of the material s response is measured over a range of temperature and frequency. These properties in clude the capacitance which is the ability of the material to store elect rical charge and conductance which is the ability of the material to transfer electrical charge. The relationship between the conductive and capacitive components of the measured curr ent is shown in figure 2.14. The capacitance (C) and the conductance (1/R) can be calcula ted from the voltage (V), current (I), frequency (ƒ) and phase shift angle () as shown in equations 2.5 and 2.6. f V I farads Capplied measured 2 sin ) ( Eq. 2.5 cos ) ( / 1 applied measuredV I ohms R Eq. 2.6

PAGE 48

23 Figure 2.14. Capacitance and conductive com ponents of the measured current. [TA Instruments DEA 2970 1997, Reprinted with pr ior permission from TA Instruments, Delaware] The capacitance and conductance are re lated to the dielectric permittivity, ', and the dielectric loss factor, ", respectively. The dielectric permittivity, ', represent the amount of dipole alignment (both induced and permanent) and the loss factor, ", measures the energy required to align dipoles or move ions. Equations 2.7 2.8 show the relationship between capacitance, conductance, dielectric permittivity and dielectric loss factor, where d is the plate spaci ng, A is the electrode plate area and o is the absolute permittivity of free space (8.85 10-12 F/m): A Cdo Eq. 2.7 of RA d 2 Eq. 2.8 The dielectric permittivity and the loss factor are the real and imaginary components of the complex permittivity, *, given by loss angle Iconductive Ica p acitive Imeasure d phase shift angle

PAGE 49

24 i Eq. 2.9 The dielectric permittivity, also referred to as the dielectric constant, is a combination of the permittivity that is due to induced dipoles and the permittivity due to the alignment of permanent dipoles, as represented by the classic Debye equation 2.10. 22 1 fu r u Eq. 2.10 = permittivityinduced dipoles + permittivitydipole alignment The term u is the unrelaxed permittivity at hi gh frequency due to induced dipoles and the term 22 1 fu r represents the permittivity due to dipole alignment, where r is the relaxed permittivity occurring at low frequency, 2 ƒ is the angular frequency and is the molecular relaxation time. The permittivitie s for polymers is low at low temperatures and below transitions as the dipoles are “locked” in place and does not have enough energy to align in the electri cal field; however, as the temp erature increases to and above the secondary and primary relaxations the permittivity increases. The dielectric loss factor, ", represents the energy requi red to align the dipoles or move ions through the polymer matrix and therefore the Debye expression for the dielectric loss factor consists of two terms: the dipole loss factor and ionic conduction as shown in equation 2.11 where is the ionic ac c onductivity (S/m). o u rf f f 2 2 1 22 Eq. 2.11 = dipole loss factor + ionic conduction Ionic conduction becomes predominant when the polymer undergoes the glass transition. It is related to viscosity where the ionic impurities can more easily move through the semi-fluid sample. The bulk i onic conductivity can be calculated using equation 2.12. of 2 Eq. 2.12

PAGE 50

25 Figure 2.15. Dipole and ion alignment in an electric field. [TA Instruments DEA 2970 1997, Reprinted with prior permission from TA Instruments, Delaware]

PAGE 51

26 The DEA can accommodate testing of various forms of samples, i.e. gels, liquids, solids, powder, and thin films. Four types of sensors exist for the TA Instruments DEA: parallel plate, single surface, sputter coated and remote single surface. In this study, the parallel plate and single surface sensor s were used. Figures 2.16 and 2.17 show schematics of these two sensors. The DEA covers a frequency range of 0.1Hz to100kHz and a temperature range of -150 C to 500 C. Various calibrations must be performed to ensure accurate experimental results; these include temperature, electronic and sensor calibrations.

PAGE 52

27 Figure 2.16. Schematic of parallel plate sensor, ram, and furnace assembly. [TA Instruments DEA 2970 1997, Reprinted with pr ior permission from TA Instruments, Delaware]

PAGE 53

28 Figure 2.17. Schematic of ceramic single su rface sensor, ram, and furnace assembly. [TA Instruments DEA 2970 1997, Reprinted with pr ior permission from TA Instruments, Delaware]

PAGE 54

29 0.0 0.5 1.0 1.5 2.0[ ] Loss Factor 2 4 6 8 10 12 14 16Permittivity -150-100-50050100150200Temperature (C) Figure 2.18. A plot of permittivity and loss f actor versus temperature for Poly(methyl methacrylate). Data points were collected for various frequencies ranging from 1Hz100kHz.

PAGE 55

30Dynamic Mechanical Analysis Dynamic mechanical analys is (DMA) is used to measure the viscoelastic properties of materials. When an oscillating st ress, or load, is applied to a material it will respond by deforming sinusoidally. This de formation, or strain, will depend on how much viscous and elastic beha vior the material possesses. When a 100% elastic material at its Hookean limit is subjected to a stress it will respond by deforming in an in-phase sine wave strain (no time lag, = 0). When the stress is removed it will return to its original shape. When a 100% viscous material is subjected to a stress it will respond by deforming in an out of phase sine wave ( = 90). When the stress is removed it will not return to its original shape (fig. 2.19) Polymers, on the other hand, exhibit a combinatorial time dependent response that is visco-elastic. The strain that is recovered in the polymer is a result of the elastic propert ies and the strain that is not recovered is a result of the viscous properties of the material; the phase shift angle will be between 0 and 90. Figure 2.19. Mechanical response of materials. (a) Sinusoidal stress ( ) = sinusoidal strain ( ), (b) Perfectly elastic in -phase response, (c) Perfec tly viscous out of phase response, (d) Combinatorial visc o-elastic response of polymer ic materials. [Perkin Elmer Instruments PETech-90]

PAGE 56

31 Figure 2.20. Mechanical phase angle shifts for a polymer. [TA Instruments DMA 2980 2002, Reprinted with prior permission from TA Instruments, Delaware] The ratio of the stress to strain is defined as the complex modulus, E*, as shown in figure 2.20. E* defines a materials resistance to de formation and can be separated into two components: the real storage modulus E', and the imaginary loss modulus, E". E i E E Eq. 2.13 The storage modulus, E', is related to th at portion of the polymer structure that fully recovers when an applied stress is removed; in polymers the storage modulus decreases as the temperature increases to a nd above the glass transition region. The loss modulus, E", is a measure of the ability of a material to dissipate mechanical energy by converting it into heat. The absorption of mechanical energy is often related to the movements of molecular segments within the ma terial and is often seen as a mechanical loss peak. Figure 2.21. shows a conceptual diagram of E' and E".

PAGE 57

32 Figure 2.21. A conceptual diagram of stored energy, E', vs. loss energy, E". [Perkin Elmer Instruments PETech-90] The DMA can be run under three different modes: dynamic multi-frequency oscillatory mode, stress rela xation mode and creep mode. In dynamic mode an oscillating stress is applied to the material and the resu lting strain is measured; from this mode one can obtain data such as storage and loss modul us with respect to time, temperature and frequency. In stress relaxation mode, a strain is instantaneously applied to the sample, and the stress required to maintain that strain is measured as a function of time; the stress relaxation modulus can be determined and the sample recovery can be monitored with time upon release of the strain to obtain % recovery. In a creep test, a constant stress is applied to the sample and the resulting strain is measured as a function of time; the creep compliance and % recovery can be obtained. Using time-temperature superposition one can use these various modes to do short term measurements and generate master curves from which long term behavior can be predicted. In this study, the tension film clamp was used to obtain the viscoelastic properties of the studied materials (fig. 2.22) Figure 2.23 shows representative E" and E' data obtained from performing a multi-frequency sweep test using the tension film clamp.

PAGE 58

33 Figure 2.22. Dynamic mechanical tension film clamp. [TA Instruments DMA 2980 2002, Reprinted with prior permission from TA Instruments, Delaware] 0 100 200 300 400Loss Modulus (MPa) 0 1000 2000 3000 4000 5000Storage Modulus (MPa) 20406080100120140160Temperature (C) Universal V3.4C TA Instruments Figure 2.23. A plot of the storage modul us and loss modulus of Poly(ethylene terephthalate).

PAGE 59

34 Viscoelastic measurements are therefore us ed to determine the internal structure of the material and can be used to inve stigate structure-prope rty relationships [TA Instruments DMA 2980 2002]. In DMA, the modulus is measured with respect to time, temperature and frequency. In this study, a TA Instruments DMA 2980 was used and it can record measurements within the temperatur e range of -150 C to 500 C and within a frequency range of 0.1Hz to 100Hz. Various ca librations must be performed to ensure accurate experimental results; these include temperature, instrument, clamp and position calibrations. Microhardness The hardness (H) of a material is a measur e of its resistance to surface deformation [Stevens 1990, Chandler 1999, Balt Calleja and Fakirov 2000]. Hardness can be determined in several ways; however, for the purpose of the experiments in this study a static indentation test was employed. Static indentation test s involve indentation of a steel ball (Brinell test), diamond c one (Grodzinski test) or diamond pyramid (Berkovich, Knoop and Vickers tests) into the surface of the materia l; the relationship of the area (A) of the imprint with respect to the applied load (F) gives the hardness number of the material as represented by H = F/A [Leica 1999]. Microhardness testing involves measurements with force loads that are less than 1N (Balt Calleja and Fakirov 2000). In this study, a Leica Vickers Microhardness Tester (VMHT) MOT equipped with a square Vick ers indenter was employed. The Vickers indenter is a four sided pyramid which has an angle, between non-adjacent faces of the pyramid of 136. The Vickers hardness number (HV, kgf/mm2 or MPa) for each sample is determined via equation 2.14, where d is the diagonal leng th of the imprint. Eq. 2.14 2 24 1854 2 sin 2d F d F A F HV

PAGE 60

35 Microhardness testing of polymers is depe ndent on the viscoela stic behavior. It has been documented that the glass transiti on is linearly related to cohesive energy density (CED) by the following equation 1 22C mR Tg Eq. 2.15 where 2 is the CED, m is a parameter that describes the internal mobility of the groups in a single chain, R is the gas constant and C1 is a constant [Balt Calleja and Fakirov 2000]. CED is also the main f actor in determining hardne ss which results in a linear relationship between the gla ss transition and hardness. Spectroscopy and Microscopy UV-VIS Molecular Absorption Spectroscopy Various molecules can absorb ultraviolet or visible light via the presence of chromophores in the chemical structures of those molecules. A chromophore is generally a group of atoms having delocali zed electrons of low excitati on energy such as seen in C=C and C=O bonds. Upon excitation of th ese electrons to high energy non-bonding orbitals, several electronic tran sitions can occur; these include *, n *, n or transitions. These transitions each require different amounts of energy and absorb in different regions of the electromagnetic spectrum. A UV-VIS spectrometer measures the absorbance, or transmittance, of a material. It can be used to determine the concentration of an analyte in a so lution using Beer’s law which states that the concentration of the anal yte is linearly related to the absorbance. In particular for polymers, it can be used to determine in situ cure kinetics of two reactive species, the presence of unreacted monomer, inhibitors and antioxidants, as well as compositional variations in copolymers [Stevens 1999]. An Agilent Technologies 8453 UV-VIS diode array spectrometer with Agilent ChemStation software was used to determin e optical transparency of various polymer composites in this study. Figure 2.24 show s a schematic of the optical system.

PAGE 61

36 Figure. 2.24 Optical system of UV-VIS diode array spectrometer. [Agilent Technologies 2000] Deuterium and tungsten lamps encompa sses the entire UV-VIS spectrum where the deuterium lamp covers the ultra-violet wavelengths and the tungsten lamp extends into the visible region. The scan range fo r this instrument is 190nm to 1100nm. As a single beam of light passes through the source lens and then through the sample absorption by various molecular species may occur. The light is then separated by a grating unto a diode array wher e the absorbance of light will then be quantified with respect to wavelength. The multichannel di ode-array technology allows for much more precision, sensitivity, and reproducibili ty [Agilent Technologies 2000]. In this study UV-VIS sp ectroscopy was employed to look at the optical transparency to investigate the interfaci al interactions ta king place between twocomponent polymer systems. Scanning Electron Microscopy Microscopy is the use of radi ation, whether it be optical or electronic, to study the structure and morphology of materials [S awyer and Grubb 1996]. The image may be obtained all at once as in optical lens t echniques or point by point as in scanning techniques. Scanning electron microscopy (SEM) uses a beam of elect rons to scan the topography of a surface [Sawyer and Grubb 1996, Bieber 2004].

PAGE 62

37 Figure 2.25. Schematic of a scanning electron microscope. A suitable electron source, such as a tungsten field emitter or lanthanum hexaboride (LaBr6), is used to produce a beam of el ectrons which is accelerated to the sample by an electrostatic potential. These primary electrons bombard the sample causing emission of secondary electrons, backscattered electrons and x-rays from the sample. As the intensity of the primary electron beam increases the further it will penetrate the surface of the material; however, secondary electrons are emitted at very low energy (< 50eV) and can only escape from the first 10-20 atomic layers of the surface therefore one can only examine the near-surface region of th e material. This beam continuously scans the sample surface. As shown in figure 2. 25 a secondary electron detector, or SED, placed in the specimen chamber collects thes e secondary electrons and measures the intensity of the electrons. The measured signa l is then converted into an image using a cathode ray tube (CRT). The syst em is kept in vacuum as air tends to scatter electrons [Sawyer and Grubb 1996, Bieber 2004]. SEM applications include looking at th e surface structur e and morphology of biological samples, metals, thin films and polymers. Samples which are not electrically conductive such as polymers and biological samp les need to undergo a pretreatment. This pretreatment involves coating th e sample with a thin conductive film. This must be done to prevent the build-up of electrons on the surface of the materials; this event is

PAGE 63

38 commonly called charging. Charging causes scattering of the electron beam which will hinder imaging and analysis. A Hitachi S-800 scanning electron micr oscope was employed to obtain SEM images in this study. It has a guaranteed reso lution of 20 , but has been found to detect nanoparticles as small as 5 nm in diamet er, and can go up to 300,000 X magnification. Images were taken of the polymer’s fractured cross section and coat ed with 10-15 nm of gold/palladium alloy using a Hummer X spu tter coater. The Hitachi S-800 scanning electron microscope is located at the Na nomaterials and Nanoma nufacturing Research Center in the Department of Engineering (U niversity of South Flor ida). We gratefully acknowledge Jay Bieber for his help a nd expertise with obtaining SEM images throughout this study. Transmission Electron Microscopy Transmission electron microscopy (TEM) is a technique where the sample is illuminated by an electron beam; it is unlike SEM where the image is obtained by scanning the sample. TEM is used for anal yzing the surface structure, or morphology, whether it is amorphous or crystalline, as we ll as the composition of the material [Sawyer and Grubb 1996. TEM gives better resolution th an SEM and can be used to detect particles as small as 0.5 nm in size. The el ectron beam is produced by an electron gun as in SEM. However, unlike SEM where the image is formed by the reflected secondary electrons, in TEM the beam strikes the sample and a portion of the beam is transmitted through the sample. As the electrons pass thr ough the sample the image is formed by an objective lens which can be magnified by proj ector lenses. The final image is projected onto a fluorescent screen [Bozzola and Russell 1992]. In this study a Philips CM10 TEM was used to obtain images of the dispersion of nanoparticles in polymer composites. This TEM has a resolution of 0.5nm and a magnification range of 20X to 510,000X. It can be used in both imaging and electrondiffraction mode. The Philips CM10 TEM is located in the Electron Micoroscope Facility in the Department of Pathology at th e University of South Florida.

PAGE 64

39CHAPTER 3 A Broad Spectrum Analysis of th e Dielectric Properties of Poly(2-hydroxyethyl methacrylate) Introduction Poly(2-hydroxyethyl methacrylate) Poly(2-hydroxyethyl methacrylate) (PHEMA ) belongs to the class of polymers known as hydrogels. When such polymers are cro sslinked they swell in water and retain a significant fraction of water without dissolving [Ratner and Hoffman 1976, Meakin, Hukins, Imrie and Aspden 2003]. PHEMA is a widely studied polymer which has found its niche in the bioapplications field; it is included as materials for contact lenses, bioadhesive gels for drug deliv ery applications, and thromboand fibro-resistant coatings [Gates et. al. 2003, Craig and Tamburic 1997, LaPorte 1997, Shtilman 2003]. PHEMA also has great potential as a protective/intera ctive coating on the surface of implantable sensors; this will be discussed in greater detail in chapter 6. The aspect of biocompatibility together with new applica tions in nanocomposite host-guest systems (chapter 4) obviated the need to further characterize the dielectric behavior of neat PHEMA. In this study, the dielectric re sponse of dry PHEMA from -150 C to 275 C is presented. The dielectric response of dry a nd hydrated PHEMA have been studied before but data obtained above 50 C have not been previously reported [Gates et. al. 2003, Diaz Calleja 1979, Gomez Ribelles and Diaz Calleja 1985, Russell et. al. 1980, Pathmanathan and Johari 1990, Johari 1991, Janacek 1973]. In mechanical studies dry PHEMA exhibits two sub-Tg secondary relaxations and a primary glass transition (Tg). The transitions are termed , and proceeding from the high temp erature transition to the low temperature transition, as show n in figure 3.1. The primary glass transition marks the onset of large scale segmental motion of the ma in chain, or polymer backbone, and in the case of hydrogels it is affected by factors such as degree of crosslinking and water content. The relaxation corresponds to the rotati on of the ester side group and the

PAGE 65

40 relaxation is associated with the rotation of the hydroxyeth yl group [Gates, Harmon, Ors and Benz 2003, Janacek 1973, Kolari k 1982]. An additional relaxation, sw, is observed in hydrated PHEMA at a temperat ure slightly greater than the transition; sw corresponds to the motion associated with the interaction of the wate r molecules with the side groups in the polymer [Gates et. al. 2003, Janacek 1973, Kolarik 1982, Kyritsis et. al. 1994, Pathmanathan and Johari 1994]. Mechanical studies have shown that the relaxation is very pronounced whereas the relaxation is relatively weak. The relaxation often appears as a shoulder to the peak and may even be unresolvable [Gates et. al. 2003, Russell et. al. 1980, Janacek 1973, Kolarik 1982]. CH2 C H3C C CH2 C CH2 C O O CH2 CH2 OH C O O CH2 CH2 OH C O O H2C H2C OH H3C CH3 n Figure 3.1. Structure and relaxations in poly (2-hydroxyethyl methacrylate). In 1979, Diaz Calleja extensively studie d the lower region of the dielectric spectrum of PHEMA in which the relaxation was character ized [Diaz Calleja 1979]. Due to instrument constraints, high temper ature data points were unattainable but the presence of a second loss peak was detect ed and it was suggested that the higher

PAGE 66

41 temperature peak observed may be attributed to the relaxation. Then in 1984, Gomez Ribelles and Diaz Calleja became the firs t to present dielectric data on the relaxation of PHEMA in which they observed a dielectric lo ss peak at ca.50 C (0.02 Hz) with an activation energy of 29 kcal/mol [Gomez Ribelles and Diaz Calleja 1985]. The intramolecular hydrogen bonding between the polar –OH groups attached to the polymer chains hinders motion of the ester moiety a nd requires a higher energy input to onset the relaxation which is evidenced by a high temper ature loss peak and high activation energy [Gomez Ribelles and Diaz Calleja 1985, Russell et. al. 1980]. Three different processes were observed in this study taking place at ca. 50 C and above, and due to the paucity of dielectric da ta in literature covering this temperature range an attempt was made to decipher th e meaning of the diel ectric spectrum of dry PHEMA. This study is important because diel ectric behavior gives insight into the structural property and relaxati ons present in the polymer, as well as it can be used to investigate the conductivity and in teraction of the polymer with nanofillers. This aspect is examined in chapter 4. Dielectric Theory and Analysis DEA is an informative technique used to determine the molecular motions and structural relaxations present in poly meric materials posse ssing permanent dipole moments [McCrum, Read and Williams 1967, Avakian, Starkweather, Jr. and Kampert 2002]. The technical aspect of its operation has been discussed in chapter 2. In dielectric measurements, the material is exposed to an alternating electric field which is generated by applying a sinusoidal voltage; th is process causes alignment of dipoles in the material which results in polarization. The polarizati on will cause the output current to lag behind the applied electric field by a phase shift angle, The magnitude of the phase shift angle is determined via measuring the resulti ng current. The capacitance and conductance are then calculated from the relationship between the applied voltage, measured current and phase shift angle [McCrum, Read and Willia ms 1967, Avakian, Starkweather, Jr. and Kampert 2002, TA Instruments DEA 2970 1997] The capacitance and conductance of the material is measured over a range of temp erature and frequency, and are related to the

PAGE 67

42 dielectric permittivity, ', and the dielectric loss factor, ", respectively. The dielectric permittivity, ', represent the amount of dipole alignment (both induced and permanent) and the loss factor, ", measures the energy required to align dipoles or move ions. The dielectric permittivity and the loss factor ar e the real and imaginary components of the complex permittivity, *, given by i Eq.3.1 In polymeric materials it has been observed that the loss factor term is a combination of two processes which are depend ent on temperature, pressure and density: 1. The rotational reorientation of the permanent dipoles present on the side chains off the polymer backbone, known as a dipolar relaxation. This process is viscoelastic and usually exhi bits a loss peak that is close to symmetric in shape and obeys Arrhenius behavior for seconda ry relaxations [Ambrus, Moynihan and Macedo 1972, Johari and Pathmanathan 1988, Bergman et. al. 1998]. The glass transition also contributes to the loss f unction as a result of the induced dipoles created by the redistribution of electrons shared between the bonded atoms on the main chain. 2. The translational diffusion of ions which causes conduction is seen as a conductivity relaxation. In glass forming polymers this process takes place with increasing viscous flow and usua lly overpowers the viscoelastic process in the dielectric loss factor spectrum [Johari and Pathmanathan 1988, Bergman et. al. 1998, Macedo, Moynihan and Bose 1972, Starkweather, Jr. and Avakian 1992]. As temperature increases it has been shown that the loss factor becomes inversely proportional to frequency. The ac conductivity, ac, is given by equation 3.2, where is the angular frequency and is the absolute permittivity of free space (8.854 x 10-12 F/m). o ac Eq.3.2 McCrum et al. have formulated a ma thematical treatment of the complex permittivity, *, which is used to resolve the viscoelastic process from the conductivity

PAGE 68

43 effects [McCrum, Read and Williams 1967]. By taking the inverse of the complex permittivity, *, one can obtain the electric modulus given by equation 3.3. 2 2 2 2" " ' 1 i iM M M Eq.3.3 Plots of the electric loss modulus, M ", versus temperature show a significant difference from those of versus temperature with respect to the separation of the viscoelastic and conductivity relaxations, but technically cont ain the same information. Due to the placement of the dielectric constant, ', in the denominator of the equation, its effects in dominating M and M are reduced. This allows a more comprehensive analysis of the dielectric data. The conductivity relaxation possesses pr operties very different from the viscoelastic relaxations present in polymers The conductivity relaxation corresponds to the model of a Debye process having a si ngle relaxation time whereas viscoelastic relaxations are known to exhibit a distributi on of relaxation times [McCrum, Read and Williams 1967, Avakian, Starkweather Jr. and Kampert 2002]. Various mathematical treatments will be applied to reveal both the viscoelastic and conductivity relaxations present in the dielectri c spectrum of PHEMA. Experimental Materials 2-hydroxyethyl methacrylate monomer wa s generously donated by Benz R&D (Sarasota, FL). It was used as received w ithout further purification. The free radical initiator employed for the polymerization was Vazo 52 [2,2,’-azobis(2,4dimethylpentane nitrile)]. Vazo 52, obtained from Dupont (Wilmington, DE), is a low temperature polymerization initiator that d ecomposes to form a cyanoalkyl radical.

PAGE 69

44Synthesis of PHEMA 0.2 wt% of the [2,2,’-azobis(2,4dimethylpentane nitrile)] Vazo 52 initiator was added to the monomer which was then dega ssed with dry nitrogen gas. The monomer was polymerized for 8 hours at 60 C and then post cured at 110 C for 4 hours. Before thermal, mechanical and dielectric analysis the PHEMA sample was oven dried at 110 C to constant weight under vacuum and stored under vacuum in the presence of phosphorous pentoxide. It should be noted that the monomer c ontained a small amount of dimethacrylate impurity which resulted in the crosslinking of the polymer. As a result of crosslinking the polymer had the ability to be molded but not dissolved. Sample Molding Samples were compression molded using a Carver Press equipped with a heating element at a temperature of 135 C for 5 minutes; it was then ai r cooled under pressure to room temperature. DEA samples were molded into rectangular disks with dimensions of 25mm x 20mm x 1mm. The DMA samples were molded into rectangular pieces of 30mm x 6mm x 1mm. Molded samples were then vacuum oven dried at 60 C until constant mass and then stored under vacuum in th e presence of phosphorous pentoxide until ready to use. Differential scanning calorimetry (DSC) Experiments were performed on a TA Instruments DSC 2920 to determine the glass transition temperature, Tg, of PHEMA. The previously dried sample (4-10mg) was hermetically sealed in an aluminum pan a nd a heat-cool-heat cycl e was performed. The DSC cell, which was calibrated with indium a nd kept under an inert nitrogen atmosphere, was heated using a ramp rate of 5deg/min to 140 C, quench cooled with liquid nitrogen and then reheated at the same rate. The Tg was taken from the second heating cycle.

PAGE 70

45Thermogravimetric analysis (TGA) A TA Instruments HiRes TGA 2950 was used to determine both the decomposition temperature of PHEMA as well as to determine if the drying technique used resulted in complete removal of absorb ed water from the polymer. The data was obtained under a dry nitroge n purge at a ramp rate of 20 C/min from 30 C to 400 C. Dynamic mechanical analysis (DMA) Dynamic mechanical analysis was c onducted on a TA Instruments DMA 2980. The instrument and clamps were calibrated and the experiments we re run under tension mode. Measurements with an oscillating amplitude of 5m were taken from -150 C to 200 C in 5 degree increments through a freque ncy range of 1-100 Hz. A preload force of 0.010N was used to maintain sample tensi on and the force tracking option of 125% was used to automatically adjust the force as the sample changed modulus in order to minimize sample deformation. The storage modulus (E'), loss modulus (E") and mechanical loss tangent (tan ) were obtained. Dielectric analysis (DEA) Single surface dielectric analysis was performed using a TA Instruments DEA 2970. The sample was heated to 135 C to embed the sample into the channels of the single surface sensor and then taken down to cryogenic temper atures with liquid nitrogen. A maximum force of 250N was applied to th e sample to achieve a minimum spacing of 0.25mm. Measurements were taken in 5 degree increments from -150 C to 275 C through a frequency range of 0.1 Hz to 100 kHz under a dry helium at mospheric purge of 500ml/min. Capacitance and conductance we re measured as a function of time, temperature and frequency to obtain the dielectric constant, or permittivity ( '), the dielectric loss ( ") and the loss tangent (tan delta = "/ ').

PAGE 71

46Results and Discussion Polymerization Scheme for PHEMA Synthesis Bulk free radical polymerization was used to synthesize poly(2-hydroxyethyl methacrylate). This process invol ves four major steps: 1) form ation of the initiator radical which is the rate determining step, 2) addition of the initiator radical to the monomer, 3) propagation of the polymer chain and 4) term ination of the polymer chain. The above steps are schematically illustrated below. Step 1. Thermal Initiation of [2,2,’-azobis(2,4-dimethylpentane nitrile)] Vazo 52 N CN N CN CN C N2 2 + Step 2. Initiation of HEMA Monomer O OH O CN C CN C O OH O +

PAGE 72

47 Step 3. Propagation of Polymer Chain O OH O CN C O OH O CH2 C O OH O CN O OH O + Step 4. Termination of Polymer Chain CH2 C O OH O O OH O CH2 C O O H O O O H O CH2 C O OH O ^^^^ ^^^^ + n Differential Scanning Calorimetry and Thermogravimetric Analysis DSC was used to monitor the drying pro cess since the presence of water in the hydrophilic polymer is known to act as a plasticizer whic h will decrease the glass transition temperature. The drying process was complete when the Tg remained constant even after additional heating under vacuum. DSC was used to determine the glass transition temperature of PH EMA, it was found to have a Tg of 99.2 C (Fig.3.2). A decomposition temperature of 319 C was determined by thermogravimetric analysis (Fig.3.3). Minimal water content was observe d as there was only a 0.5% weight loss up to120 C. The dielectric analysis was taken up to 275 C, a temperature at which there was a 6% weight loss.

PAGE 73

48 99.15C(I) 94.76C 102.65C Neat PHEMA -0.22 -0.12 -0.02Heat Flow (W/g) 20406080100120140160Temperature (C) Exo UpUniversal V3.4C TA Instruments Figure 3.2. DSC data: Glass transition temperature, Tg, of neat PHEMA.

PAGE 74

49 43.08% (3.406mg) 318.98C 359.01C Neat PHEMA 40 60 80 100Weight (%) 3080130180230280330380Temperature (C) Universal V3.4C TA Instruments Figure 3.3. TGA data: Decompositi on temperature of neat PHEMA.

PAGE 75

50Dynamic Mechanical Analysis The mechanical viscoelastic relaxati ons in PHEMA have been previously reported [Gates et. al. 2003, Pathmanathan and Johari 1990, Johari 1991, Janacek 1973, Kolarik 1982, Nicolais et. al. 1974]. Dry PHEMA e xhibits two subTg relaxations, the relaxation which is associated with the rotation of the hydroxyethyl group and relaxation corresponding to the rota tion of the ester side group. relaxation : Our DMA experiment confirms a transition occurring between a temperature range of -135 C to -116 C for the frequency range of 1-100 Hz. It follows Arrhenius behavior and has an activation energy of 10.6 kcal/mol (44.4 kJ/mol). This is compared to previously reported values of a transition occurring at -133 C (1Hz) with an activation energy of 10.7 kcal/mole ( 44.8 kJ/mol) and -132 C (1Hz) with an activation energy of 7.5 kcal/mole (31.4 kJ/mol) (Gates et. al. 2003, Kolarik 1982). relaxation : The relaxation is only obs erved at 1Hz as it is overlapped by the relaxation as shown in figur e 3.4. Kolarik observed the transition in dry PHEMA at 26.9 C (1Hz) (fig. 3.5) and Gates observed the transition at 28 C (1Hz) (fig. 3.6) (Gates et. al. 2003, Kolarik 1982). DMA in correlation with DEA have been us ed to best describe the relaxations exhibited in PHEMA. The mechanical and di electric relaxations in PHEMA are not as closely related as one would think. The relaxation has been observed to be more pronounced in DEA than in DMA; this point is discussed in greater detail in a later section.

PAGE 76

51 1Hz10Hz 100Hz 0 200 400 600 800 1000 1200 1400Loss Modulus (MPa) -150-100-50050100150200Temperature (C) Universal V3.4C TA Instruments Figure 3.4. DMA data: Mechanical loss peaks at 1Hz for PHEMA.

PAGE 77

52 -150-100-50050100 0 20 40 60 80 100 120 140 Shear Loss Modulus, G" (MPa)Temperature (oC) Figure 3.5. DMA data: Mechanical loss p eaks at 1Hz for PHEMA. [Kolarik 1982] -150-100-50050100150 100 150 200 250 300 350 Loss Modulus (MPa)Temperature (oC) Figure 3.6. DMA data: Mechanical loss peaks at 1Hz for PHEMA. [Gates 2003]

PAGE 78

53Dielectric Analysis DEA analysis of PHEMA revealed anom alous behavior which has not been reported by researchers who studi ed the dielectric properties of this polymer. Most of the work published present data up to 50 C in which detailed analyses of the transition are presented. The , possible (or merge) and the conductivity relaxations present in PHEMA have been identified with DEA. Figure 3.7 shows the dielectric permittivity plot, figure 3.7 shows the loss factor plot and figure 3.8 shows the electric loss modulus plot of PHEMA over a wide range on temperature and frequency. The transition is clearly observed; however, the occurrence of ionic conduction in the sample has hidden the and transitions in the plot. By applying the electric modulus formalism the and relaxation are revealed. relaxation : The relaxation appears as a strong peak in both the loss factor and electric loss modulus plots. It obeys Arrhenius beha vior where the peak temperature maxima increased linearly with frequency as shown in the Arrhenius plot of ln frequency vs. the reciprocal of temperature (f ig. 3.10); the slope of which was used to determine the activation energy from the relationship [McCrum, Read and Williams 1967, Gomez Ribelles and Diaz Calleja 1984, Gedde 1995] of RT E f fa o ln ln eq.3.4 The relaxation occurs within a temperature range of -147 C to -60 C (0.1Hz-100 kHz) and has an activation energy of 6.9 kcal/mol (28.9 kJ/mol) as determined from the electric loss modulus temperature maxima Arrhenius dependence. Both the activation energy, as well as the temperature, of the dielectric relaxation is lower than the measured mechanical relaxation as shown in table 1. This occurrence has been reported previously by Gates et al. and Janacek. It can be explained by the concept of mechanical activation versus dielectric activation. Rotation of the –OH side group in PHEMA is observed as a result of 1) slow viscoela stic deformation on the application of a mechanical load and 2) slow orientation pol arization on the applica tion of an electric field. The viscoelastic deformation is weak ly dependent on the dipole moment of the –

PAGE 79

54 OH side group whereas the orientation polarizat ion is strongly dependent on the dipole moment [Hartwig 1994, Mohsen, Craig and Filisko 2000]. The dipole moment of the – OH group is large and appears to be more easil y aligned in the electr ic field, whereas in DMA the energy needed to overcome the disp ersive Van der Waals forces to allow rotation of the –OH group is greater. Previously reported activati on energy values for the relaxation range from 6.9 kcal/mol to 16 kcal/mol. As mentioned by Pathmanathan and Johari this may be caused by the different crosslinking density of the polymer; the highe r the crosslinking density the higher the activation energy needed to overcome hindered rotation of the –OH side group [Pathmanathan and Johari 1990, Johari 1991]. Table 3.1. DEA vs. DMA for the transition Properties DEA DMA peak at 1Hz (obtained from tan delta plot) -130.14oC -124.56oC 2) Ea (obtained from ”, E’ plots) 6.9kcal/mol 10.6kcal/mol

PAGE 80

55 Figure 3.7. DEA data: Plot of permittivity ( ') versus temperature for PHEMA at various frequencies.

PAGE 81

56 Figure 3.8. DEA data: Pl ot of loss factor ( ') versus temperature for PHEMA at various frequencies.

PAGE 82

57 Figure 3.9. DEA data: Plot of electric loss modulus ( M") versus temperature for PHEMA at various frequencies.

PAGE 83

58 y = -3486.2x + 27.925 R2 = 0.99 Ea = 6.9 kcal/mol0 2 4 6 8 10 12 14 00040.00450.0050.00550.0060.00650.0070.00750.0081/T (K)ln frequency Figure 3.10. DEA data: Arrhenius plot of relaxation in PHEMA. y = -5332x + 38.329 R2 = 0.9985 Ea = 10.6 kcal/mol 0 1 2 3 4 5 0.0060.00620.00640.00660.00680.0070.0072 1/T (K)ln frequency Figure 3.11. DMA data: Arrhenius plot of relaxation in PHEMA.

PAGE 84

59 and relaxations : Until now the dielectric relaxation in PHEMA has only been reported by Gomez Ribelles and Di az Calleja in which they re ported a loss peak at 50 C (0.02Hz) with an activation energy of 29 kcal/mol (121 kJ/mol) [Gomez Ribelles and Diaz Calleja 1985]. Further data at higher temperatures and frequencies were not presented. As observed in the lo ss factor plot (fig. 3.8) the and relaxations were obscured by conductivity effects so the electric modulus fo rmalism was used; in the plot the peak was only observed at high frequencies (6 kHz to 100 kHz) between ca. 145 C to 160 oC. It was interesting to observe the anomalous behavior exhibited in the electric loss modulus vs. temperature plot as shown in fi gure 3.9. Frequency scans from 0.1 Hz to 10 Hz show a symmetric, single electric modulus loss peak between the temperature range of 66 C-113 C. This peak follows Arrheniu s behavior in which the peak temperature maxima increased linearly with frequency to give an activation energy of 20.7 kcal/mol (86.7 kJ/mol). One may argue that this is the peak corresponding to the glass transition temperature but experimental data prove otherwise. The symmetry and Arrhenius relationship are characterist ic of secondary relaxations [McCrum et. al. 1967]. The frequency-temperature dependence of the and peaks is shown in figure 3.12. As the frequency is increased two M peaks are apparent. The first peak appears first as a shoulder to the second peak for frequencies 300 Hz to 1 kHz and then as a separate peak from 3 kHz to 100 kHz. The first M peak occurs at a peak height significantly lower than the one M peak observed in the lower frequencies and is attributed to the or possible merge. It is not symme tric and does not follow Arrhenius behavior. One can reason that the relaxation requires a higher temperature to initiate the rotation of the lateral side group due to th e presence of intramolecular bonding. In poly(methylmethacrylate), the relaxation is faster moving than the relaxation and tends to merge with the relaxation at a temperature above Tg [McCrum, Read and Williams 1967, Bergma n et. al. 1998]. In PHEMA, the relaxation may have overlapped with the relaxation to form the merge which is seen as the first M peak

PAGE 85

60 in the higher frequency scans. Figure 3. 13 shows the electrical loss functions for comparison of dry PHEMA at 6 kHz. Figure 3.12. DEA data: FrequencyTemperature dependence of the and relaxations in PHEMA.

PAGE 86

61 Figure 3.13. DEA data: Dielectric lo ss functions of PHEMA at 6 kHz.

PAGE 87

62 Conductivity relaxation : Upon mathematically treating the data to obtain the electric loss modulus ( M ") several changes occur. The dielectric permitivitty, ', increases dramatically with increasi ng temperature and frequency; in electric modulus the placement of the dielectric constant, ', in the denominator of the equation prevented it from dominating M and M ". It is also observed that the M peaks, especially for the transition, occurred at temperatures lower than the peaks. By taking the electric modulus the space charge effects are suppr essed and an ionic conductivity peak is revealed [Starkweather Jr. and Avakian 1992, Tsangaris, Psarras and Kouloumbi 1998, Pissis and Kyritsis 1997]. This is seen as the second M peak in the spectra for the higher frequency scans. The fact that this is a c onductivity relaxation and not a viscoelastic relaxation can be proven in several ways. Proof 1 : The dielectric permittivity and loss factor for a relaxation with a single relaxation time can be descri bed by equations 3.5 and 3.6, 2 21 'E U R U eq. 3.5. 2 21 "E E U R eq. 3.6. where E is the dielectric relaxation time, is the angular frequency, and U and R represents the high frequency, unrelaxed state and the low frequency, relaxed state, respectively. By manipulating equations 3.5 and 3.6 equation 3.7 is derived. 2 2 22 2 U R U R eq. 3.7. Cole and Cole proposed that by plotting against at a particular temperature a semicircle of radius ( R – U)/2 should be obtained [McCrum, Read and Williams 1967]. In this case, analogous Argand plots of M vs. M were made according to equation 3.8. 2 2 22 2 R U R UM M M M M M eq. 3.8.

PAGE 88

63 In M vs. M plots the values proceed from lowe r to higher frequencies whereas in vs. plots the values proceed from higher fr equency to lower frequency. The Argand plots are shown in figures 3.14 and 3.15. Semici rcular behavior is characteristic of the Debye model, in particular molecular liquids and small rigid molecules [McCrum, Read and Williams 1967, Emran et. al. 1999]. Polymers on the other hand deviate from semicircular behavior in wh ich they exhibit a distributi on of relaxation times and are often characterized by modi fied Cole-Cole expressions [McCrum, Read and Williams 1967]. Figure 3.14 shows the Argand plot in wh ich data points were taken in the relaxation region. The plot doe s not follow semicircular beha vior; this was expected as this is a viscoelastic relaxation where entang lements due to chain interactions result in a distribution of relaxation times Figure 3.15 shows the Argand plot constructed with data taken at a temperature above Tg where the 2nd M peak is observed. This plot reveals a true semicircular arc which can be interpreted to mean that it is indeed not a viscoelastic relaxation. Johari and Pathmanathan, together with others, have stated that conductivity relaxations in ionic conductors exhibit si ngle relaxation times [Ambrus, Moynihan and Macedo 1972, Johari and Pathmanathan 1988, Macedo, Moynihan and Bose 1972].

PAGE 89

64 Figure 3.14. DEA data: Arga nd plot derived from the relaxation region (-110 C).

PAGE 90

65 Figure 3.15. DEA data: Ar gand plot derived from the conductivity relaxation region (200 C).

PAGE 91

66 Proof 2 : Ambrus et al. presente d the electric modulus in terms of time, frequency and modulus [Ambrus, Moynihan and Macedo 1972]. De rivations have been shown in detail in various papers in which an e xpression for the electric modulus ( M ), eq. 3.9, was determined under the assumption of conditions where ionic conducti on is purely due to the diffusion of ions and independent of viscoelastic, dipolar relaxation [Ambrus, Moynihan and Macedo 1972, Johari and Pathmanathan 1988, Macedo and Moynihan 1972, Starkweather, Jr. and Avakian 1992, Tsa ngaris, Psarras and Kouloumbi 1998]. This assumption implies that under the stat ed conditions the electric modulus ( M) will have a relaxation with a single relaxation time, 2 2 21 1 ) 1 ( s sMi M i i MsM eq. 3.9. In equation 3.9, Ms = 1/ s where s occurs at a value of that is independent of temperature. Starkweather Jr. et al. showed that plots of log M" and log M' vs. log frequency will reveal slopes of 1 and 2, respectively [Starkweather, Jr. and Avakian 1992]. In this study the dependence of M', M" on frequency in the conductivity relaxation region is shown in figures 3.16 and 3.17. As exp ected the plots reveal slopes of 1 and 2 at temperatures in the region of the conductivity relaxation. Similar plots were not obtained for temperatures in the glass transition region and below (figs. 3.18, 3.19).

PAGE 92

67 Figure 3.16. DEA data: Dependence of M on frequency in the conductivity relaxation region (165 C).

PAGE 93

68 Figure 3.17. DEA data: Dependence of M" on frequency in the conductivity relaxation region (165 C).

PAGE 94

69 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 -10123456 log frequencylog M' Figure 3.18. DEA data: Dependence of M on frequency at a temperature below Tg (60 C). -3 -2.5 -2 -1.5 -1 -0.5 0 -10123456 log frequencylog M" Figure 3.19. DEA data: Dependence of M" on frequency at a temperature below Tg (60 C).

PAGE 95

70 Proof 3 : As mentioned earlier two processes contribute to the loss factor. When viscoelastic effects are negligible, the loss f actor is described by e quation 3.2. [Pissis and Kyritsis 1997, Pissis et. al. 2000, Henn et. al. 2000]. Figure 3.20 shows a plot of the frequency dependence of ac conductivity ( ac) for temperatures above Tg where conductivity is predominan t. Dc conductivity ( dc) was obtained by extrapolation to zero frequency. At low frequencies ac is independent of fre quency from 110-200 C. As temperature is increased, the frequency de pendence of ac conductiv ity plateaus and is independent of all frequencies measured. dc increased with increasing temperature and its Arrhenius relationship is expressed by equation 3.10, where E is the apparent activation energy, k is Boltzmann’s constant and o is the pre-exponentia l factor [Polizos et. al. 2000]. ) exp( log log kT Eo dc eq. 3.10 Pissis et al. reported that the ionic conductivity peak s hows the same temperature dependence as dc conductiv ity; figures 3.21 and 3.22 are used to compare the temperature dependence of the M peak and dc conductivity [Pissis and Kyritsis 1997, Pissis et. al. 2000]. The apparent activation energies de termined from both plots are very close in value where the activ ation energy from the second M peak observed at high frequencies is 13.7 kcal/mol (57.4 kJ/mol) as compared to 11.2 kcal/mol (46.9 kJ/mol) obtained from the dc conductivity plot. On ly three frequencies (3000, 6000 and 10000 Hz) were used to construct the Arrhenius plot for figure 13 since these are the only frequencies in which the two M peaks were clearly separated. Similar results have been reported in other systems [Pissis a nd Kyritsis 1997, Pissis et. al. 2000].

PAGE 96

71 Figure 3.20. Frequency dependence of ac conductivity for PHEMA at temperatures above Tg.

PAGE 97

72 Figure 3.21. DEA data: Arrhenius plot of log dc conductivity vs. the inverse of temperature.

PAGE 98

73 Figure 3.22. DEA data: Arrhenius plot of frequency-temperature dependence of the conductivity M" peak.

PAGE 99

74 Conclusion The dielectric spectrum of PHEMA has b een examined in which the electric modulus formalism has been applie d to the analysis of data. The relaxation region has been previously reported on by various author s. This study has presented analysis of the dielectric spectra in a temper ature region up to and above th e glass transition temperature to reveal the secondary relaxation, the primary relaxation and the conductivity relaxation. Several approaches were successfu lly applied to verify the presence of the conductivity relaxation. Further development and understanding of ionic conductivity in polymer composites will be discussed in chapte r 4. This analysis will also be used to characterize the dielectric spectra of 2-hydroxyethyl methacrylate (HEMA) and 2,3dihydroxypropyl methacrylate (DHP MA) copolymers used as biocompatible coatings for an implantable glucose sens or in chapter 5 and 6.

PAGE 100

75CHAPTER 4 Nanostructure Matrix Interactio ns in Methacrylate Composites Introduction Polymer Nanocomposites Polymer matrix composites have been studi ed and used commercially as early as the 1950’s [Kusy 1986]. Much effort has b een placed on improving the mechanical, optical, electronic and magnetic properties of polymers by making polymer blends, and by adding fillers to the polymeric matrix [V arga et. al. 2003, Clay ton et. al. 2005, Wilson et. al. 2004]. In recent years, great stride s have been made to better understand the polymer-filler interface, to develop methods for enhancing interfacial adhesion and to characterize filler dispersion. Polymer nanocompo sites are of particular interest; due to the large interfacial area inherent of nanos cale fillers, polymer nanocomposites access new properties and exploit the unique synergis m between the matrix and filler [Chabert et. al. 2004]. Many techniques have been developed to disperse nanoparticles in polymeric matrices. Some techniques involve in situ and intercalation polymerization and in situ sol-gel, and other technique s involve dispersion after polymerization, such as melt blending [O’Rourke Muisener et. al. 2002, Tatro et. al 2004, Xiong et. al. 2002, Park and Jana 2003, Chen et. al. 2001, Rong et. al. 2001, Park et.al. 2002]. Each technique has its advantages and disadvantages. For instance, in situ ultrasonic polymerization developed in our laboratory, which invol ves sonication to break up a nd disperse the nanoparticles during polymerization is a tech nique that is difficult to sc ale-up for industrial production even though it produces good disp ersion [Mohomed et. al. 2005] On the other hand, melt blending is a technique that has been su ccessfully used in large scale composite production but it has limitations in terms of its ability to separate the nanoagglomeration clusters efficiently. Nanosized metal particles have properties that are different from those of macrosized bulk metals. Their size influences chemical, magnetic, optical and electronic

PAGE 101

76 properties [Carotenuto and Nicolais 2003, Kulkarni, John Thomas and Rao 2002]. Nanosizing also induces changes in the fundame ntal properties, such as the melting point and boiling point, as well as in the material’s shape and crystalline structure. For instance, bulk silicon does not emit light; however, nanosilicon emits light as a result of the quantum confinement effect which cause s a change in the materials optical gap [Luterov et. al. 2005]. Similarly, ferrom agnetic materials on the nanoscale show remarkably different properties especially wh en their particle size is less than a single domain size. Within this size range, the nanomagnetic particles show interesting dynamics and coercivity behavior. The incr eased surface to volume ratio influences changes in their high frequency properties, magnetic anisotro py etc. [Poddar et. al. 2005, Cattaruzza et. al. 1998]. The nanoparticle being investig ated in this study is of pa rticular inte rest. Due to its unique molecular structure it is the first-known reported nanoscale Kagom lattice to be synthesized by the pioneering research of Zaworotko and co-workers. The structure is made up of both square secondary buildi ng units (SBU) and tr iangular secondary building units. The open nanoporous ne twork is constructed using Cu( II ) dimers positioned at the lattice points which are bri dged using organic ligands. In the square SBUs, the moments of the indivi dual dimers (a.k.a. the spin) ca ncel each other leading to antiferromagnetic coupling. The unique magnetic response of this nanoparticle is directly related to the presen ce of the triangular SBU. The triangular SBU introduces spin frustration in the structure; whereby, a fe rromagnetic-like respons e leading to magnetic hysteresis is observed [S rikanth et. al. 2003, Moulton et.al. 2002]. This nanoparticle and its counterparts have th e potential to be used in a variety of electromagnetic and drug delivery applicati ons. Its influence in a polymer matrix is important to study as the nanoparticle may be us eful as part of a coating or capsule. In this study, we examined the effects of th e interactions taking place between a selfassembled nanostructure with two functiona lly different polymer s: poly(2-hydroxyethyl methacrylate) (PHEMA) and poly( methyl methacrylate) (PMMA).

PAGE 102

77The Hydroxylated Nanoball Various noncovalent interactions exist in polymeric systems such as those that arise from hydrogen bonding, el ectrostatic attractions and interactions [Kilbinger and Grubbs 2002, Porter 2005, Jiang et. al 1999]. Synthesis of novel hydroxylated nanoparticles has been described earlier (Abourahma et. al. 2001]. The prototypal nanoballs have formula [L2Cu2(bdc)2]12 (L = solvent, or substituted pyridine, bdc = benzene-1,3-dicarboxylate). They can be functionalized in multiple ways at their surface; for example, groups that can engage in strong hydrogen bonding, e.g. sulfonate, methoxy and hydroxyl, can be positioned on each of th e twenty four bdc ligands. The axial ligands L can also be substituted. These supramolecular nanostructures are ideal for probing polymeric interactions because they offer the potential for functionalization at multiple sites. The polyhedral structures arise via self assembly and, when crystallized from DMSO, form discrete single crystals. The specific crystal structure of interest is [(DMSO)(MeOH)Cu2(benzene-1,3-dicarboxylate-5-OH)2]12. This nanoparticle is rhombihexahedral in shape, with both squa re and triangular secondary building units (SBU) and possesses 24 hydroxy groups on the surface as shown in figure 4.1. Figure 4.1. Structur e of [(DMSO)(MeOH)Cu2(benzene-1,3-dicarboxylate-5-OH)2]12, a.k.a. the hydroxylated nanoball. [Abourahma et. al. 2001 Reproduced by permission of The Royal Society of Chemistry]

PAGE 103

78 The nanoparticle, commonly referred to as the nanoball, has an internal volume of 1nm3. The square SBU windows have sides of 12.749 and 5.888 in length and the triangular SBU windows have si des that are 5.861, 9.303 and 12.716 in length, as shown in figures 4.2 and 4.3. It has been shown that MeOH ligands actively bond to the metal ions in the interior surface of thes e structures [Abourahma et. al. 2001]. It is important to note that the HEMA monomer can likewise act as a ligand due to the presence of the pendant –OH group. More over, the PHEMA chains may intertwine amongst the nanoballs and act as “poly-ligands” resulting in s upramolecular structures. In addition, the HEMA monomer which is approx imately 5 in width (figure 4.4), can find its way into the interior of the nanoball through the porous st ructure/windows. In this case the PHEMA-nanoball nanocomposites may fo rm structures similar in concept to pseudo-rotaxanes. Figure 4.2. Square secondary building unit of [(DMSO)(MeOH)Cu2(benzene-1,3dicarboxylate-5-OH)2]12, a.k.a. the hydroxylated nanoball. [Abourahma et. al. 2001, Reprinted with prior permission from Dr. H. Abourahma]

PAGE 104

79 Figure 4.3. Triangular secondary building unit of [(DMSO)(MeOH)Cu2(benzene-1,3dicarboxylate-5-OH)2]12, a.k.a. the hydroxylated nanoball. [Abourahma et. al. 2001, Reprinted with prior permission from Dr. H. Abourahma] Figure 4.4. Calculat ed width and length of HEMA monomer. A series of PHEMA-nanoball and P MMA-nanoball nanocomposites were synthesized in situ A comparison study was made between the PHEMA-nanoball nanocomposites and PMMA-nanoball nanocompo sites. It was anticipated that the nanoballs would have minimal interaction with the methyl methacrylate and the composites would exhibit pr operties different from those of the PHEMA-nanoball nanocomposites. It was presumed that the fa vorable polar-polar in teraction between the HEMA and nanoball would resu lt in a network structure containing possible physical crosslinks. This was confirmed by the thermal, mechanical and dielectric data collected.

PAGE 105

80Experimental Poly(2-hydroxyethyl methacrylate)-Nanoball Nanocomposites HEMA monomer was obtained from Benz R&D (Sarasota, FL). 0.2wt% of the free radical initiator 2,2'azobis(2,4-dimethylvaleronitrile) (Vazo52, DuPont) was added to the monomer, degassed with dry N2 gas and polymerized at 60C for 6 hours, followed by a post cure session at 110C for 4 hours. Various concentrations by wt% of the nanocomposite were made by dissolving the nanoballs in the HEMA monomer prior to polymerization. It should be noted th at the monomer contained a small amount of ethylene glycol dimethacrylate impurity whic h resulted in crosslinking of the polymer. Poly(methyl methacrylate)Nanoball Nanocomposites The nanoballs have minimal affinity for methyl methacrylate and were dispersed throughout the matrix via in situ ultrasonic polymerization (fig. 4.5). The in situ ultrasonic polymerization techni que, developed in our laborat ories, did not require any solvents. Using a Branson Sonifier 450, the monomer and nanoballs were sonicated in an ice bath under a nitrogen atmosphere for 1hour. 0.2wt% of Vazo52 was added to the mixture and sonicated under a nitrogen atmosphere and in an oil bath at 80C until the mixture became viscous. The sonicator pr obe was removed and polymerization was allowed to continue in the h eated oil bath for 24 hours. Th e samples were post-cured at 120C for 4 hours. Differential Scanning Calorimetry (DSC) Experiments were performed on a TA Instruments DSC 2920 to determine the glass transition temperature, Tg. Samples (4-10mg) were hermetically sealed in aluminium pans and a heat-c ool-heat cycle was performe d. The DSC cell, which was calibrated with indium and kept under an in ert nitrogen atmosphere, was heated using a ramp rate of 10deg/min to 140C, quench cooled with liquid nitrogen and then reheated at the same rate. The Tg was taken at the inflection point and from the second heating cycle.

PAGE 106

81 Figure 4.5. In situ ultrasonic polymerization technique developed for the synthesis of the Poly(methyl methacrylate)nanoball nanocomposites. ADDITION OF PARTICLES TO MONOMER SONICATION Step 1 Sonicate in ice bath THERMOPLASTIC: 1) POLYMERIZE 2) CRUSH AND HOTPRESS INTO MOLD 3) TESTING: DSC, DEA, DMA, PPMS, UV-VIS SELECTION OF MONOMER AND NANOPARTICLES Step 2 Sonicate in heated bath

PAGE 107

82Dielectric Analysis (DEA) Dielectric analysis was performed usi ng a TA Instruments DEA 2970. The sample was heated to 140C and then taken down to cr yogenic temperatures with liquid nitrogen. A maximum force of 250N was applied to th e sample to achieve a minimum spacing of 0.25mm. Measurements were taken in 5 de gree increments from -150 C to 200C through a frequency range of 0.1Hz to 100 kH z under a dry nitrogen atmospheric purge of 500ml/min. Capacitance and conductance were measured as a function of time, temperature and frequency to obtain the dielectric constant, or permittivity (') and the dielectric loss ("). Parallel plate sensors were used. Dynamic Mechanical Analysis Dynamic mechanical anal ysis (DMA) was conducted on a TA Instruments DMA 2980. The instrument and clamps were calib rated and the experiments were run under tension mode. Measurements with an oscillat ing amplitude of 5m were taken from -150 C to 200 C in 5 degree increments thr ough a frequency range of 1-100 Hz. The storage modulus (E'), loss modulus (E") and tan delta were obtained. Sample Molding Samples were compression molded using a Carver Press equipped with a heating element at a temperature of 135C for 5 minutes; it was then air cooled under pressure to room temperature. DEA samples were molded into 2.5cm diameter circular disks with a thickness of 1mm. Molded PHEMA sample s were then vacuum oven dried at 60 C until constant mass and then stored under vacuum in the presence of phosphorous pentoxide until ready to use. Microhardness A Leica Vickers Microhardness Tester (VMHT) MOT equipped with a square Vickers indenter, which has an angle between non-adjacent faces of the pyramid of 136 was used to perform microindentation. The Vickers hardness number (HV) for each

PAGE 108

83 sample was determined. The values were taken from the average of five indents. A load of 500g and a dwell time of 10s were used. Each sample was approximately 1mm thick and measurements were made at room temperature. Soxhlet Extraction In order to study the degree of cross-li nking (sol-gel ratio) and to identify the extent of polymerization of the monomer, the standard extraction technique has been applied. Gel fraction (fgel) was obtained via Soxhlet extraction using methanol as the extracting solvent. A set of three samp les (~0.3g each) was prepared; they were encapsulated in Whatmann 2 filter paper e nvelopes and the dry weight was obtained before and after extraction. The extracti on was performed for 7 days. Samples were vacuum oven dried before and after extracti on at 60C for 8 hours. The gel fractions (fgel) were calculated from the following equation: 0 gel gelw w f Eq. 4.1 ,where w0 and wgel are dry weights of the samples before and after extraction, respectively [Gerasimov 2002]. UV-VIS Spectroscopy An Agilent Technologies 8453 UV-VIS diode array spectrometer was used to determine optical transparency of 1mm th ick samples. The scan range was 190nm to 820nm and air was used as the background. Transmission Electron Microscopy A 0.5 wt% solution of nanoball-HEMA monomer and 0.5 wt% solution of nanoball-methanol were prepared. Droplets of each solution were placed on a grid and a Philips CM10 TEM was used to obtain micrographs.

PAGE 109

84Results and Discussion UV-VIS and TEM The hydroxylated nanoballs dissolved read ily in the HEMA monomer and were polymerized in situ TEM images (fig. 4.6) of th e nanoballs dispersed in HEMA monomer show that the presence of nanoa gglomerates was minimal as each particle measured approximately 4nm in diameter. Pr ior calculations estimate a diameter of 3.1nm. A TEM image of the nanoball in meth anol, a reported solv ent system for the nanoball, revealed the presence of nanoclu sters. The PHEMA nanocomposites exhibited high optical transparency in the blue region of visible light which resulted from the excellent dispersion and interf acial interaction of the nanoba lls in the polymeric matrix. Light scattering due to agglomerations was not observed. Since the nanoballs did not dissolve in th e methyl methacrylate monomer, an in situ ultrasonic polymerizati on technique was developed to disperse the nanoparticles in situ This technique produced samples that were optically transparent but still contained agglomerates. Figure 4.7 and 4.8 illustrates the optically transpar ent discs produced via the in situ ultrasonic polymerizati on technique, as well as a sample of non-uniform dispersion. This non-uniform sample was pr oduced by sonicating the nanoballs in methyl methacrylate, followed by polymerization wi thout sonication; it is apparent that sonication during polymerization is important to the fabricat ion process. UV-VIS spectra of both polymer systems are shown in figure 4.9.

PAGE 110

85 Figure 4.6. TEM images of a) HEMA -Nanoball and b) Me thanol-Nanoball. Figure 4.7. Optically transparent discs (1mm) of the PMMA-nanoball nanocomposites produced via in situ ul trasonic polymerization (1st three discs) and a sample of a 0.05% nanoball-PMMA composite produced by another method (4th disc). Figure 4.8. PMMA-nanoball nanocomposite produced via in situ ultrasonic polymerization.

PAGE 111

86 Figure 4.9. UV-VIS comparison of PM MA-Nanoball nanocomposite and PHEMANanoball nanocomposite.

PAGE 112

87Relationship between Glass Transiti on Temperature, Gel Fraction and Microhardness The trend observed in the change of the glass transition temperature, Tg, is the key to understanding the primary interaction taking place between th e nanoballs and the polymer matrix. It was shown that the gl ass transition temperature increased with nanoball concentration in th e PHEMA nanocomposites; wher eas, it decreased in the PMMA nanocomposites (figs. 4.10 4.17). This change in Tg suggests that changes in the available free volume of the polymer matrix are taking place. An attempt was made to remove the nanoballs from the PHEMA matrix via Soxhlet extraction in methanol. After one week in the extraction a pparatus, no nanoballs were detected in the methanol since the PHEMA samples were crosslinked; by cont rast, all the nanoballs were extracted from the PMMA samples. Data obtained from the Soxhlet extraction of the PHEMA nanocomposites was used to calculate the gel fr action (Table 4.1). Th e gel fraction is the ratio of the dry weight of the sample afte r extraction and before extraction. Because nanoballs were not detected in the extracti ng solvent the cal culated gel fraction values were normalized for the nanocomposites. The increase in the gel fr action of the PHEMA nanocomposites is characteristic of an increase in the crosslinking density of the polymer network; this is directly related to the re duction of available free volume resulting in an increase in the Tg [Molyneux 1991]. It is well known th at most physical crosslinks in polymers are labile to dissolution in the pr oper solvent environment [Nam et. al. 2004, Ilmain et. al. 1991, Gedde 1995], so it is significant that this self-assembled suprastructure persists. At this point it is evid ent that the nanoballs were playing a role in increasing the crosslinking density either by hydrogen bonding or by entanglements. This interaction is absent in the PMMA nanocomposites. The hardness ( H ) of a material is a measur e of its resistance to surface deformation [Balt Calleja et. al. 2000, Stevens 1990, Chandler 1999]. Microhardness data also confirm the existing trend in which the hardness number increased with nanoball concentration in the PHEMA nanocomposites (Table 4. 1). This was expected as it has been documented by Balt Calleja and Fakirov that the Tg is linearly related to cohesive energy density (CED ) by the following equation

PAGE 113

88 1 22 C mR Tg Eq. 4.2 where 2 is the CED, m is a parameter that describes the internal mobility of the groups in a single chain, R is the gas constant and C1 is a constant. CED is also the main factor in determining hardness which results in an almost linear relationship between Tg and H. The increased resistance to surface deform ation of the PHEMA nanocomposites may be due to the decreasing free volume content of the matrix associated with the apparent physical crosslinking and/or entanglements ta king place. Researchers have previously presented data in which surface modification of their material induced crosslinking which increased the surface hardness [Said-Galiyev et. al. 1993, Tretinnikov et. al. 1999]. The opposite effect is observed in the PMMA nanocomposites in which the nanoballs appear to act as pl asticizers. The decrease in Tg is indicative of an increase in the free volume available in the matrix. The nanoballs were removed from the PMMA matrix as a result of minimal interaction be tween the two components. When a load is applied to the surface as in micro-indentation experiments the polymer chains are able to slide past each other more easily resulting in a decrease in the surface hardness number [Lorenzo et. al. 1993]. This action was obser ved in the PMMA-nanoball nanocomposites.

PAGE 114

89 Sample Tg ( C) Gel Fraction Hardness Number, HV (MPa) Neat PHEMA 99 0.82 + 0.02 236.5 + 3.4 0.1% NB-PHEMA 100 0.84 + 0.01 276.9 + 2.1 0.5% NB-PHEMA 101 0.91 + 0.01 294.5 + 3.1 0.9% NB-PHEMA 104 0.94 + 0.02 325.6 + 1.2 1.5% NB-PHEMA 105 0.96 + 0.01 406.4 + 11.0 Neat PMMA 113 NA* 305.3 + 8.2 0.05% NB-PMMA 109 NA 242.6 + 5.5 0.1% NB-PMMA 107 NA 232.2 + 3.4 Gel fraction was not calculated fo r the PMMA-nanoball nanocomposites. Table 4.1. Glass transition temperature, ge l fraction and Vickers ha rdness number of the polymer nanocomposites.

PAGE 115

90 112.98C(I) 107.76C 116.14C Neat PMMA -0.4 -0.3 -0.2 -0.1 0.0 0.1Heat Flow (W/g) 20406080100120140160Temperature (C) Exo UpUniversal V3.4C TA Instruments Figure 4.10. DSC data: Glass transition temperature, Tg, of neat PMMA.

PAGE 116

91 109.38C(I) 103.80C 112.42C 0.05% Nanoball-P MMA composite -0.4 -0.3 -0.2 -0.1 0.0 0.1Heat Flow (W/g) 20406080100120140160Temperature (C) Exo UpUniversal V3.4C TA Instruments Figure 4.11. DSC data: Glass transition temperature, Tg, of 0.05% NanoballPMMA composite.

PAGE 117

92 107.20C(I) 102.47C 110.54C 0.1 % Nanoball-PMMA composite -0.4 -0.3 -0.2 -0.1 0.0 0.1Heat Flow (W/g) 20406080100120140Temperature (C) Exo UpUniversal V3.4C TA Instruments Figure 4.12. DSC data: Glass transition temperature, Tg, of 0.1 % NanoballPMMA composite.

PAGE 118

93 98.77C(I) 96.31C 103.66C Neat PHEMA -0.5 -0.3 -0.1 0.1Heat Flow (W/g) 20406080100120140Temperature (C) Exo UpUniversal V3.4C TA Instruments Figure 4.13. DSC data: Glass transition temperature, Tg, of neat PHEMA.

PAGE 119

94 100.06C(I) 96.37C 104.98C 0.1% Nanoball-PH EMA composite -0.4 -0.3 -0.2 -0.1 0.0 0.1Heat Flow (W/g) 20406080100120140Temperature (C) Exo UpUniversal V3.4C TA Instruments Figure 4.14. DSC data: Glass transition temperature, Tg, of 0.1 % NanoballPHEMA composite.

PAGE 120

95 101.12C(I) 99.04C 106.72C 0.5% Nanoball-PHEMA composite -0.4 -0.3 -0.2 -0.1 0.0 0.1Heat Flow (W/g) 20406080100120140Temperature (C) Exo UpUniversal V3.4C TA Instruments Figure 4.15. DSC data: Glass transition temperature, Tg, of 0.5 % NanoballPHEMA composite.

PAGE 121

96 104.32C(I) 102.87C 108.62C 0.9% Nanoball-PHEMA composite -0.8 -0.6 -0.4 -0.2 0.0Heat Flow (W/g) 20406080100120140Temperature (C) Exo UpUniversal V3.4C TA Instruments Figure 4.16. DSC data: Glass transition temperature, Tg, of 0.9 % NanoballPHEMA composite.

PAGE 122

97 104.62C(I) 100.84C 106.17C 1.5% Nanoball-PHEMA composite -0.5 -0.3 -0.1 0.1Heat Flow (W/g) 20406080100120140Temperature (C) Exo UpUniversal V3.4C TA Instruments Figure 4.17. DSC data: Glass transition temperature, Tg, of 1.5 % NanoballPHEMA composite.

PAGE 123

98Dynamic Mechanical Analysis (DMA) of PMMA-Nanoball Nanocomposites DMA is used to measure the viscoelas tic properties of polymers. The loss modulus, E", is a measure of the ability of a material to dissipate mechanical energy by converting it into heat. The absorption of mechanical energy is often related to the movements of molecular segments within the material [TA Instruments DMA 2002]. The following E" plot (fig.4.18) represents a comparison between neat PMMA and the two nanoball-PMMA nanocomposites at 10Hz. Neat PMMA exhibits three mechanical relaxations within the temperature range measured. The first, primary relaxation is referred to as the transition and it corresponds to th e glass transition. The secondary relaxation corresponds to the rotation of the ester side group and the relaxation results from the rotation of the methyl group attached to the main chain. The activation energies for the transition were obtained from Arrhenius plots of ln frequency versus 1/Temperature (figs. 4. 19-4.20) and are listed in table 4.2. In both DEA and DMA, the activation energies for the transition decreased with increasing nanoball concentration. This is common in plasticized ma terials and is a result of increased free volume in the matrix; the side ester moiety is sterically less hindered and requires less energy to rotate more freely. Th e plasticization effect is also easily observed by lowering of the glass transition temp erature and suppression of the secondary relaxation peaks. DMA was not performed on the nanoball-PHEMA composites as these samples broke easily in the instrument clamps. Sample DEA (kJ/mol), (kcal/mol) DMA (kJ/mol), (kcal/mol) Neat PMMA 78.3, 18.7 69.9, 16.7 0.05% Nanoball-PMMA 72.0, 17.2 66.2, 15.8 0.1% Nanoball-PMMA 67.4, 16.1 66.6, 15.9 Table 4.2. Comparison of activation energies of the transition for the PMMA nanocomposites as determined from DEA and DMA.

PAGE 124

99 Figure 4.18. DMA data: Loss Modulus, E", vs. temperature for the PMMANanoball composites at 10Hz.

PAGE 125

100 Neat PMMA 0 100 200 300 400 500Loss Modulus (MPa) -150-100-50050100150Temperature (C) Universal V3.4C TA Instruments Figure 4.19. DMA data: Loss Modulus, E" vs. temperature for neat PMMA. Arrhenius Plot of Transition for Neat PMMA (1-100 Hz) y = -8423.5x + 29.786 R2 = 0.9795 -1 0 1 2 3 4 5 0.0030.00310.00320.00330.00340.00350.0036 1/Tln Freq Ea = 69.9kJ/mol Figure 4.20. DMA data: Arrhenius plot of transition for neat PMMA.

PAGE 126

101 0.05% Nanoball-PMMA composite 0 100 200 300 400 500Loss Modulus (MPa) -150-100-50050100150200Temperature (C) Universal V3.4C TA Instruments Figure 4.21. DMA data: Loss Modulus, E" vs. temperature for 0.05% NanoballPMMA composite. Arrhenius plot of Transition for 0.05% Nanoball-PMMA Composite (1-100 Hz)y = -7995.5x + 29.096 R2 = 0.9982 -1 0 1 2 3 4 5 0.0030.00310.00320.00330.00340.00350.00360.0037 1/Tln Freq Ea = 66.2kJ/mol Figure 4.20. DMA data: Arrhenius plot of transition for 0.05% NanoballPMMA composite.

PAGE 127

102 0.1% Nanoball-PMMA composite 0 100 200 300 400Loss Modulus (MPa) -150-100-50050100150200Temperature (C) Universal V3.4C TA Instruments Figure 4.23. DMA data: Loss Modulus, E", vs. temperature for 0.1% NanoballPMMA composite. Arrhenius Plot of Transition for 0.1% Nanoball-PMMA compositey = -8076.5x + 29.127 R2 = 0.9962 0 1 2 3 4 5 0.00290.0030.00310.00320.00330.00340.00350.00360.0037 1/Tln Freq Ea = 66.6kJ/mol Figure 4.24. DMA data: Arrhenius plot of transition for 0.1% Nanoball-PMMA composite.

PAGE 128

103Dielectric Analysis (DEA) Plots of and for both neat PHEMA and n eat PMMA are shown in figure 4.25. In mechanical studies, PHEMA and PMMA exhibit two subTg secondary relaxations and a primary glass tran sition. The transitions are termed , and proceeding from the high temperature transition to the low temperature transition. The primary transition marks the onset of large scale segmental motion of the main chain, or polymer backbone, and in the case of hydrogels it is affected by f actors such as degree of crosslinking and water content. The relaxation corresponds to the rotation of the ester side group and the relaxation is associated with the rotation of the hydroxyethyl group in PHEMA and with the methyl group rotation in PMMA [McCrum et. al. 1967, Gates et. al. 2003, Janacek 1973, Kolarik 1982]. The relaxation for PMMA does not exhibit any net dipole change and as a resu lt is dielectrically inactive; whereas it is clearly observed in PHEMA.

PAGE 129

104 Figure 4.25. DEA permittivity, ', and loss factor, ", of A) neat PHEMA and B) neat PMMA.

PAGE 130

105 In the dielectric spectra of loss factor versus temperature for PHEMA the relaxation was observed in the temperatur e range of -125 to 0 C; whereas, the and relaxations appeared to merge and are obscured by conductivity effects. When conductivity effects were subtracted out via the electric modulus formalism the relaxations were resolved [McCrum et. al 1967, Macedo et. al. 1972]. Activation energy for the relaxation for the PHEMA and PMMA nanocomposites were determined via Arrhenius plots of ln frequency vs. the recipr ocal of temperature (figs. 4.26-4.46); the slope of which was used to determine the activation energy via the following equation: RT E f fa o ln ln Eq. 4.3 The data obeyed Arrhenius behavior where the peak temperature maxima increased linearly with frequency. As shown in table 4.3, the activ ation energy required for the alignment of the ester side chain moiety increased with nanoball concentration for the PHEMA nanocomposites. This suggests that there is hindered mobility of the side group; this is possibly due to either th e persistent hydrogen interactions and/or entanglements we believe is taking place. Whereas, the activation energy required for the alignment of the ester side chain moiety d ecreased with nanoball concentration for the PMMA nanocomposites. This is common in pl asticized materials and is a result of increased free volume in the matrix; the side ester moiety is sterically less hindered and requires less energy to rotate more freely.

PAGE 131

106 Sample Activation Energy for Transition (kJ/mol), (kcal/mol) Activation Energy for Transition (kJ/mol), (kcal/mol) Neat PHEMA 39.8, 9.5 86.2, 20.6 0.1% NB-PHEMA 34.3, 8.2 88.8, 21.2 0.5% NB-PHEMA 29.7, 7.1 95.9, 22.9 0.9% NB-PHEMA 28.9, 6.9 103.4, 24.7 1.5% NB-PHEMA 27.6, 6.6 109.3, 26.1 Neat PMMA NA 78.3, 18.7 0.05% NB-PMMA NA 72.0, 17.2 0.1% NB-PMMA NA 67.4, 16.1 Table 4.3. DEA data: Activation energies for the transition for the PHEMA and PMMA nanocomposites. Figure 4.26. DEA data: Loss Factor, ", vs. temperature for neat PHEMA.

PAGE 132

107 Arrhenius plot of Transition for neat PHEMAEa = 86.2kJ/mol 0 2 4 6 8 10 0.0020.00220.00240.00260.00280.0030.0032 1/Tln freq Figure 4.27. DEA data: Arrhenius plot of transition for neat PHEMA. Arrhenius plot of Transition for neat PHEMAy = -4810.3x + 32.45 R2 = 0.9971 0 2 4 6 8 10 12 14 0.0040.00450.0050.00550.0060.00650.007 1/Tln freq Ea = 39.8kJ/mol Figure 4.28. DEA data: Arrhenius plot of transition for neat PHEMA.

PAGE 133

108 Figure 4.29. DEA data: Loss Factor, ", vs. temperature for 0.1% NanoballPHEMA composite.

PAGE 134

109 Arrhenius plot of Transition for 0.1% Nanoball-PHEMA compositey = -10723x + 31.431 R2 = 0.987 0 2 4 6 8 10 0.0020.00220.00240.00260.00280.003 Temperature (1/K)Ln Freq Ea = 88.8kJ/mol Figure 4.30. DEA data: Arrhenius plot of transition for 0.1% Nanoball-PHEMA composite. Arrhenius plot for Transition for 0.1% Nanoball-PHEMA composite y = -4157x + 31.262 R2 = 0.9644 0 2 4 6 8 10 12 14 0.00450.0050.00550.0060.00650.007 1/TLnFreq Ea = 34.3kJ/mol Figure 4.31. DEA data: Arrhenius plot of transition for 0.1% Nanoball-PHEMA composite.

PAGE 135

110 Figure 4.32. DEA data: Loss Factor, ", vs. temperature for 0.5% NanoballPHEMA composite.

PAGE 136

111 Arrhenius plot of Transition for 0.5% Nanoball-PHEMA compositey = -11578x + 32.667 R2 = 0.992 0 2 4 6 8 10 0.0020.00220.00240.00260.00280.003 1/T (1/K)Ln Freq Ea = 95.9kJ/mol Figure 4.33. DEA data: Arrhenius plot of transition for 0.5% Nanoball-PHEMA composite. Arrhenius plot of Transition for 0.5% Nanoball-PHEMA composite y = -3582.1x + 28.678 R2 = 0.9851 0 2 4 6 8 10 12 14 0.00450.0050.00550.0060.00650.007 1/T (1/K)Ln Freq Ea = 29.7kJ/mol Figure 4.34. DEA data: Arrhenius plot of transition for 0.5% Nanoball-PHEMA composite.

PAGE 137

112 Figure 4.35. DEA data: Loss Factor, ", vs. temperature for 0.9% NanoballPHEMA composite.

PAGE 138

113 Arrhenius plot of Transition for 0.9% NanoballPHEMA compositey = -12485x + 34.191 R2 = 0.9943 0 2 4 6 8 0.00210.00220.00230.00240.00250.00260.00270.0028 1/T (1/K)Ln Freq Ea = 103.4kJ/mol Figure 4.36. DEA data: Arrhenius plot of transition for 0.9% Nanoball-PHEMA composite. Arrhenius plot of Transition for 0.9% Nanoball-PHEMA composite y = -3460x + 28.694 R2 = 0.9882 0 2 4 6 8 10 12 14 0.00450.0050.00550.0060.00650.007 1/T (1/K)Ln Freq Ea = 28.9kJ/mol Figure 4.37. DEA data: Arrhenius plot of transition for 0.9% Nanoball-PHEMA composite.

PAGE 139

114 Figure 4.38. DEA data: Loss Factor, ", vs. temperature for 1.5% NanoballPHEMA composite.

PAGE 140

115 Arrhenius plot of Transition for 1.5% NanoballPHEMA compositey = -13191x + 35.826 R2 = 0.9976 0 2 4 6 8 0.00210.00220.00230.00240.00250.00260.00270.0028 1/T (1/K)Ln Freq Ea = 109.3kJ/mol Figure 4.39. DEA data: Arrhenius plot of transition for 1.5% Nanoball-PHEMA composite. Arrhenius plot of Transition for 1.5% Nanoball-PHEMA compositey = -3339.7x + 27.949 R2 = 0.9969 6 7 8 9 10 11 12 0.00450.00470.00490.00510.00530.00550.00570.00590.0061 1/T (1/K)Ln Freq Ea = 27.6kJ/mol Figure 4.40. DEA data: Arrhenius plot of transition for 1.5% Nanoball-PHEMA composite.

PAGE 141

116 Figure 4.41. DEA data: Loss Factor, ", vs. temperature for neat PMMA. Arrhenius Plot of Transition for neat PMMA y = -9443.1x + 35.039 R2 = 0.9909 -1 0 1 2 3 4 5 6 7 0.0030.00310.00320.00330.00340.00350.00360.00370.0038 1/T (1/K)Ln Freq Ea = 78.3kJ/mol Figure 4.42. DEA data: Arrhenius plot of transition for neat PMMA.

PAGE 142

117 Figure 4.43. DEA data: Loss Factor, ", vs. temperature for 0.05% NanoballPMMA composite. Arrhenius plot of Transition for 0.05% Nanoball-PMMA composite y = -8688.8x + 31.887 R2 = 0.9943 0 1 2 3 4 5 6 7 0.00290.0030.00310.00320.00330.00340.00350.00360.0037 1/T (1/K)Ln Freq Ea = 72kJ/mol Figure 4.44. DEA data: Arrhenius plot of transition for 0.05% Nanoball-PMMA composite.

PAGE 143

118 Figure 4.45. DEA data: Loss Factor, ", vs. temperature for 0.1% NanoballPMMA composite. Arrhenius plot of Transition for 0.1% Nanoball-PMMA composite y = -8156.1x + 30.077 R2 = 0.9931 0 1 2 3 4 5 6 7 0.00280.0030.00320.00340.00360.0038 1/T (1/K)Ln Freq Ea = 67.4kJ/mol Figure 4.46. DEA data: Arrhenius plot of transition for 0.1% Nanoball-PMMA composite.

PAGE 144

119 The dielectric permittivity, ', represents the amount of dipole alignment, and as shown in table 4.4, the permittivit y data follow the general trend exhibited by the nanocomposites. For the PHEMA nanocomposites decreased with nanoball concentration; whereas, it increas ed in the PMMA nanocomposites. This supports the idea that there is hindered mobility of the side group in the PHEMA nanocomposites and the opposite effect in the PMMA nanocomposites. Sample @ 25C/10Hz @ 100C/10Hz @ 125C/10Hz Neat PHEMA 7.87 11.43 13.44 0.1% NB-PHEMA 7.42 14.66 15.05 0.5% NB-PHEMA 6.43 8.92 11.95 0.9% NB-PHEMA 5.25 6.73 11.49 1.5% NB-PHEMA 5.15 6.71 11.82 Neat PMMA 3.63 4.65 5.60 0.05% NB-PMMA 4.09 5.25 6.16 0.1% NB-PMMA 4.12 5.22 6.04 Table 4.4. DEA data: Comparison of the dielectric constant, ', measured at 10Hz for the polymer-nanoball nanocomposites at 25, 100 and 125C. To further substantiate the above data the ionic conductivity related to the movement of ions through the matr ix was examined. The ac conductivity, ac, is given by the equation o ac Eq.4.4 where is the angular frequency and is the absolute permittivity of free space (8.854 x 10-12 F/m) [Macedo et. al. 1972, St arkweather, Jr. et. al. 1992] Plots of the frequency dependence of ac conductivity ( ac) for temperatures above Tg where conductivity is predominant were made and the dc conductivity ( dc) was obtained by extrapolation to zero frequency. As temperature is increased the frequency dependence of ac conductivity

PAGE 145

120 plateaus and is independent of all frequencies measured. dc follows an Arrhenius relationship expressed by the equation ) exp( log log kT Eo dc Eq. 4.5 where E is the apparent activation energy, k is Boltzmann’s constant and o is the preexponential factor. The PHEMA nanocomposites exhibited a decr ease in the ionic conductivity and an increase in ionic conductivity activation energy for samples with the higher concentration of nanoballs. This is due to the immobilization of the matrix by the nanoball interaction [Dahmouche et. al. 1999]. PMMA nanocomposites consistently show the opposite effect in which there is an increase in the ionic conductivity and a decrease in the ionic con ductivity activation energy as the nanoball concentration is increased. Sample Ionic Conductivity (S/m) Activation Energy (kJ/mol), (kcal/mol) Neat PHEMA 1.95 10-5 35.2, 8.4 0.1% NB-PHEMA 3.65 10-5 31.9, 7.6 0.5% NB-PHEMA 5.02 10-5 44.1, 10.5 0.9% NB-PHEMA 1.5% NB-PHEMA 3.38 10-6 2.63 10-6 45.2, 10.8 47.0, 11.2 Neat PMMA 1.09 10-12 151.7, 36.2 0.05% NB-PMMA 1.72 10-9 95.8, 22.9 0.1% NB-PMMA 9.91 10-9 79.2, 18.9 Table 4.5. DEA data: Ionic conductivity and i onic conductivity activati on energies for the polymer nanocomposites.

PAGE 146

1211E-41E-30.010.1110100100010000100000 1E-9 1E-8 1E-7 1E-6 1E-5 AC(S/m)f(Hz) 90degC 111degC 135degC 156degC 177degC 195degC 216degC 237degC 255degC 270degC Figure 4.47. DEA data:Frequency depe ndence of ac conductivity for neat PHEMA. 2.22.32.4252.62.72.829 -95 -90 -85 -80 -75 -70 -65 -60 y=Ax + B ==> log dc = (-E/K)[1/T] + log oA= -4.23252 B= 4.70918 Ea = 35.2kJ/molLog dc (S/m)1000/T (K-1) Figure 4.48. DEA data: Arrhenius plot of ionic conductivity ac tivation energy for neat PHEMA.

PAGE 147

1220.1110100100010000100000 1E-9 1E-8 1E-7 1E-6 1E-5 dc (S/m)f (Hz) 80 100 120 140 160 180 195 Figure 4.49. DEA data: Frequency depe ndence of ac conductivity for 0.1% Nanoball-PHEMA composite. 2.12.22.32.42.52.62.7 -85 -80 -75 -70 -65 -60 -55 y=Ax + B ==> log dc = (-E/K)[1/T] + log oA= -3.8458 B= 4.43743 Ea = 31.9kJ/molLog dc (S/m)1000/T (K-1) Figure 4.50. DEA data: Arrhenius plot of ionic conductivity ac tivation energy for 0.1% Nanoball-PHEMA composite.

PAGE 148

1230.1110100100010000100000 1E-9 1E-8 1E-7 1E-6 1E-5 dc (S/m)f (Hz) 80 100 120 140 160 180 195 Figure 4.51. DEA data: Frequency de pendence of ac conductivity for 0.5% Nanoball-PHEMA composite. 2.12.22.32.42.52.62.72.82.9 -100 -95 -90 -85 -80 -75 -70 -65 -60 -55 y=Ax + B ==> log dc = (-E/K)[1/T] + log oA= -5.44695 B= 5.29971 Ea = 44.1kJ/mollog dc (S/m)1000/T (K-1) Figure 4.52. DEA data: Arrhenius plot of ionic conductivity ac tivation energy for 0.5% Nanoball-PHEMA composite.

PAGE 149

1240.1110100100010000100000 1E-10 1E-9 1E-8 1E-7 1E-6 1E-5 ac (S/m)f (Hz) 80 100 120 140 160 180 195 Figure 4.53. DEA data: Frequency de pendence of ac conductivity for 0.9% Nanoball-PHEMA composite. 2.12.22.32.42.52.62.72829 -10.5 -10.0 -9.5 -9.0 -8.5 -8.0 -7.5 -7.0 -6.5 -6.0 y=Ax + B ==> log dc = (-E/K)[1/T] + log oA= -5.42942 B= 5.47088 Ea = 45.2kJ/molLog dc (S/m)1000/T (K-1) Figure 4.54. DEA data: Arrhenius plot of ionic conductivity ac tivation energy for 0.9% Nanoball-PHEMA composite.

PAGE 150

1250.1110100100010000100000 1E-11 1E-10 1E-9 1E-8 1E-7 1E-6 ac (S/m)f (Hz) 140 150 160 170 180 190 195 Figure 4.55. DEA data: Frequency de pendence of ac conductivity for 1.5% Nanoball-PHEMA composite. 2.12.22.32.42.52.62.72.829 -10 -9 -8 -7 -6 y=Ax + B ==> log dc = (-E/K)[1/T] + log oA= -5.68081 B= 5.57953 Ea = 47.0kJ/mollog dc (S/m)1000/T (K-1) Figure 4.56. DEA data: Arrhenius plot of ionic conductivity ac tivation energy for 1.5% Nanoball-PHEMA composite.

PAGE 151

1260.1110100100010000100000 1E-11 1E-10 1E-9 1E-8 1E-7 1E-6 ac (S/m)f (Hz) 140 150 160 170 180 190 195 Figure 4.57. DEA data: Frequency de pendence of ac conductivity for neat PMMA. 2.152202252.302.352.402.45 -110 -105 -100 -95 -90 -85 -80 -75 y=Ax + B ==> log dc = (-E/K)[1/T] + log oA= -18.24882 B= 11.96409 Ea = 151.7kJ/molLog dc (S/m)1000/T (K-1) Figure 4.58. DEA data: Arrhenius plot of ionic conductivity ac tivation energy for neat PMMA.

PAGE 152

1270.1110100100010000100000 1E-10 1E-9 1E-8 1E-7 1E-6 ac (S/m)f (Hz) 140 150 160 170 180 190 Figure 4.59. DEA data: Frequency de pendence of ac conductivity for 0.05% Nanoball-PMMA composite. 2.152.202252302.352.402.45 -10.0 -9.5 -9.0 -8.5 -8.0 -7.5 -7.0 y=Ax + B ==> log dc = (-E/K)[1/T] + log oA= -11.51932 B= 8.76423 Ea = 95.8kJ/molLog dc (S/m)1000/T (K-1) Figure 4.60. DEA data: Arrhenius plot of ionic conductivity ac tivation energy for 0.05% Nanoball-PMMA composite.

PAGE 153

1280.1110100100010000100000 1E-10 1E-9 1E-8 1E-7 1E-6 ac (S/m)f (Hz) 140 150 160 170 180 190 Figure 4.61. DEA data: Frequency de pendence of ac conductivity for 0.1% Nanoball-PMMA composite. 2.152.202.252302352.402.45 -10.0 -9.5 -9.0 -8.5 -8.0 -7.5 y=Ax + B ==> log dc = (-E/K)[1/T] + log oA= -9.52603 B= 8.00396 Ea = 79.2kJ/molLog dc (S/m)1000/T (K-1) Figure 4.62. DEA data: Arrhenius plot of ionic conductivity ac tivation energy for 0.1% Nanoball-PMMA composite.

PAGE 154

129 Schematics of Proposed Nanoball-Polymer Interactions Figure 4.61. A schematic of the plastic ization effect of nanoballs in PMMA. Figure 4.62. A schematic of the cross linking effect of nanoballs in PHEMA.

PAGE 155

130Conclusion The hydroxylated nanoparticle and its counterparts have the potential to be used in a variety of electromagnetic and drug delivery applications and therefore its interaction with a polymer matrix is important to study as the nanoparticle may be useful as part of a coating or capsule. In this study, the effects of the intera ctions taking place between a self-assembled nanostructure with two functionally different polymers: poly(2hydroxyethyl methacrylate) (PHEMA) and poly(methyl methacrylate) (PMMA) was examined. The PHEMA-nanoball nanocomposites endur ed in a hostile swelling and extraction environment. It is well known that most physical crosslinks in polymers are labile to dissolution in the pr oper solvent environment, so it is significant that these selfassembled suprastructures persisted. The da ta showed that the crosslinking density increased in the PHEMA nanocomposites. This observation suggests that there is an interaction taking place betw een the nanoball and HEMA. Fu rther evidence gained by DSC and DEA data support this phenomenon as the glass transition temperature and the ionic conductivity ac tivation energy increased with nanoba ll concentration. It is believed that this interaction may be the result of phys ical threading of PHEM A chains through the nanoball windows, in which the HEMA mono mer may be drawn by H bonding to the internal ligands in the nanoball. The possibility of a number of different schemes exists but in order to be more conclusive inve stigations should be carried out by further characterizing the interaction using linear PHEMA and othe r polymer systems with the nanoball. By contrast, data derived for the PMMA nanocomposites indicate that there is minimal interaction between the nano ball and the matrix where the PMMA nanocomposites consistently show the opposite effect. There is an increase in the ionic conductivity and a decrease in the ionic conductiv ity activation energy as the nanoball concentration is increased. This phenomenon is due to the lack of immobilization of the polymer matrix which consequently enhances the rotational movement of the side chain

PAGE 156

131 moiety and the translational diffusion of ions in the matrix. Further DSC and microhardness data verify the plastici zation effect of the PMMA matrix.

PAGE 157

132CHAPTER 5 Dielectric Analyses of a Series of Poly(2-hydroxyethyl methacrylate-co-2,3dihydroxypropyl methacylate) Copolymers Introduction The full range dielectric response of neat poly(2-hydroxyethyl methacrylate) (PHEMA) from -150 C to 275 C was presented in chapter 3. Previously, the dielectric response of dry and hydrated PHEMA had been studied before but data obtained above 50 C had not been reported [Gates et. al 2003, Diaz Calleja 1979, Ribelles and Diaz Calleja 1985, Russell et. al. 1980, Pathmantha n and Johari 1990, Janacek 1973]. It was important to decipher the dielectric spectrum of PHEMA to further i nvestigate the effects of the novel hydroxylated nanoparticle on the poly mer matrix as presented in chapter 4. The electric modulus formalism was employe d to reveal the various structural and conductivity relaxations present in the polymer composites. The effects of crosslinking and plasticization in the polymer matrices we re monitored by the characterization of the molecular relaxations present in the polymer and by the ionic diffusion in the polymer matrix. Using dielectric spectroscopy, it wa s determined that the activation energy needed to bring about the molecular relaxa tion of the pendant groups in composites was highly dependent on the available free volume and that the ionic conductivity activation energy generally increased as the degree of crosslinking increased and it decreased as plasticization effect increased [Damouche et. al. 1999]. This phenomenon is due to the immobilization (or lack thereof) of the matrix which consequently hinders (or enhances) the rotational movement of the side chain moiety and the translational diffusion of ions in the matrix [Eloundou et. al. 2002]. Dielectric anal ysis proved to be a us eful tool to better understand the polymer-filler interface. In this study, the dielec tric spectra of several random copolymers of 2hydroxyethyl methacrylate (HEMA) and 2, 3-dihydroxypropyl methacrylate (DHPMA) will be analyzed. The structures of these monomers are shown in figure 5.1. Both of these materials belong to the class of polymers known as hydrogels, and have found a role in biomedical applications for such materials as contact lenses, bioadhe sive gels for drug

PAGE 158

133 delivery and as a thromboand fibroresistan t coating for implantable sensors [Gates et. al. 2003, Craig and Tamburic 1997, LaPorte 1997, Shtilman 2003]. Gates et. al. was the first to report the dielectric response of poly(HEMA-DHPMA) copolymers in 2003 [Gates et. al. 2003]; the hydrogel samples were prepared as powder sandwiched between polyethylene wafers. As a result, the transition was not resolved since the glass transition of PHEMA and PDHPMA occurred at a temperature close to the melt temperature of the polyethyl ene (Marlex 6000) matrix (Tm = 120 C). Figure 5.1. Chemical structure of a) 2-hydroxyethyl methacrylate (HEMA) and b) 2,3-dihydroxypropyl methacrylate (DHPMA). Poly (2,3-dihydroxypropyl methacrylate) (PDHPMA) is also known as glyceryl methacrylate (GMA) and is th e major component of Benz-G materials; the advantage of these materials is that it remains 100-percen t saturated when in contact with the eye [Benz and Ors 2000, 1999]. The increased wate r equilibrium content of PDHPMA and its biocompatibility properties have impacted its use as a biomaterial. The recent

PAGE 159

134 development of poly(HEMA-co-DHPMA) copolym ers for use as biocompatible coatings for implantable sensor devices in our laborator y has also prompted this study; research concerning the application of these hydrogels as sensor coatings will be presented in chapter 6. Figure 5.2 shows a histolo gy image of a pre-hydrated HEMA-DHPMA copolymer subcutaneously implanted in an anim al specimen where it is observed that the copolymer induced minimal to no fibrosis. Th is present dielectric study attempts to fortify previous work to better understand th e thermal and dielectric response of these materials up to and above th e glass transition region. Figure 5.2. A histology im age of a HEMA-DHPMA c opolymer subcutaneously implanted in an animal specimen. Dielectric analysis is an informative t echnique used to determine the molecular motions and structural relaxations present in polymeric materials possessing permanent dipole moments [McCrum 1967]. In dielectric m easurements, the material is exposed to an alternating electric fiel d which is generated by appl ying a sinusoidal voltage; this process causes alignment of dipoles in the material which results in polarization. The capacitance and conductance of th e material is measured over a range of temperature and frequency, and are related to the dielectric permittivity, ', and the dielectric loss factor, ", respectively. The dielectric permittivity, ', represent the amount of dipole alignment

PAGE 160

135 (both induced and permanen t) and the loss factor, ", measures the energy required to align dipoles or move ions. In polymeric materials it has been observed that the loss factor term is a combination of two processes: the rotationa l reorientation of th e permanent dipoles present on the side chains off the polymer backbone, known as a di polar relaxation and the translational diffusion of ions which cau ses conduction and is seen as the conductivity relaxation (see eq.5.1, 5.2 and 5.3). ion dipole Eq.5.1 2 21 "E E U R dipole Eq.5.2 o ac ion Eq.5.3 Various mathematical treatments will be applied to reveal both the viscoelastic and conductivity relaxations present in th e dielectric spectra of the poly(HEMA-coDHPMA) copolymers. The reader is referred to chapters 2 and 3 to obtain an in-depth explanation of dielectric th eory and its application in characterizing polymers. Experimental Materials 2-hydroxyethyl methacrylate and 2,3-di hydroxypropyl methacrylate monomers were generously donated by Benz R&D (Saras ota, FL). The monomers were used as received without further pur ification. The free radical initiator employed for the polymerization was Vazo 52 [2,2,’-azobis(2,4-dimethyl pentane nitrile)]. Vazo 52, obtained from Dupont (Wilmington, DE), is a low temperature polym erization initiator that decomposes to form a cyanoalkyl radical.

PAGE 161

136Synthesis of Poly(HEMA-co-DHPMA) Copolymer Series A series of HEMA-DHPMA random copol ymers were synthesized using free radical polymerization. 0.2 wt % of the [2,2,’-azobis(2,4-di methylpentane nitrile)] Vazo 52 initiator was added to the monomer which was then degassed with dry nitrogen. The monomers were polymerized for 8 hours at 60 C and then post cured at 115 C for 4 hours. Before thermal and dielectric analys is, the polymer sample s were oven dried at 110 C to constant weight under vacuum and stored under vacuum in the presence of phosphorous pentoxide. The properties of the two homopolymers: PHEMA and PDHPMA, together with three random copolymers of HEMA and DHPMA were investigated. Differential scanning calorimetry (DSC) Experiments were performed on a TA Instruments DSC 2920 to determine the glass transition temperature, Tg, of the polymers. The previously dried sample (4-10mg) was hermetically sealed in an aluminium pa n and a heat-cool-heat cycle was performed. The DSC cell, which was calibrated with i ndium and kept under an inert nitrogen atmosphere, was heated using a ramp rate of 5deg/min to 140 C, quench cooled with liquid nitrogen and then reheat ed at the same rate. The Tg was taken from the second heating cycle. Dielectric analysis (DEA) Single surface dielectric analysis was performed using a TA Instruments DEA 2970. The sample was first chilled with li quid nitrogen and then ground into a fine powder using a Bel Art micromill. The powde r was placed on the sensor, heated to 135 C to embed the sample into the channels of the single surface sensor and then taken down to cryogenic temperatures with liqui d nitrogen. A maximum force of 250N was applied to the sample to achieve a mini mum spacing of 0.25mm. Measurements were taken in 5 degree increments from -150 C to 275 C through a frequency range of 0.6 Hz to 100 kHz under a dry helium atmospheric purge of 500 ml/min Capacitance and conductance were measured as a function of temperature and frequency to obtain the

PAGE 162

137 dielectric constant, or permittivity ('), the dielectric loss (") and the loss tangent (tan delta = "/'). Results and Discussion Differential scanning calorimetry (DSC) The glass transition temperatures for the HEMA and DHPMA homopolymers, as well as the random copolymers were determin ed using differential scanning calorimetry. Differential scanning calorimetry was also us ed to monitor the dr ying process since the presence of water in hydrophilic polymers is known to act as a plasticizer which will decrease the glass transition temperature, Tg. The drying process was complete when the Tg remained constant even after additional h eating under vacuum. The results are listed in table 5.1 and figures 5.3-5.7 show the DSC scans for the samples. The presence of one glass transition in the copolymer is indicativ e of the miscibility of the two monomers. Unlike previous data reported by Gates et al., the glass transition temperature for this set of copolymers decreased linearly as the DHP MA content increased (with a R-squared value of 0.9741) (fig. 5.8). Gates et. al. repo rted a glass transition temperature of 105 C for both the HEMA and DHPMA homopolymer a nd the copolymers as well [Gates et. al. 2003]. This difference in reported glass transi tion temperature may be a result of varying crosslinker content between the samples. The syntheses of the HEMA and DHPMA monomers often result in th e production of ethylene glycol dimethacrylate (EGDMA) as an impurity which acts as a crosslinking agen t. The glass transiti on of the hydrogel will be dependent on the polymerization process, EGDMA concentration and water content present in the polymer. EGDMA is often adde d to the hydrogel for certain applications where dissolution of the hydroge ls needs to be avoided, as in contact lens. Equation 5.4 was used to calculate the theo retical glass transition temperatures of the copolymers based on the experimental Tg’s of the homopolymers, where w is the mole fraction of the indivi dual polymer present in the copolymer [Gedde 1995]. Table

PAGE 163

138 5.1 shows a close semblance between the calcul ated glass transition temperatures for the copolymers to the actual values: 2 2 1 11 Tg w Tg w TgCopolymer Eq. 5.4 101.41C(I) 96.37C 106.04C -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2Heat Flow (W/g) 20406080100120140160Temperature (C) Exo UpUniversal V3.4C TA Instruments Figure 5.3. DSC data: Glass transition temperature, Tg, of neat PHEMA.

PAGE 164

139 95.60C(I) 91.37C 103.37C -0.4 -0.3 -0.2 -0.1 0.0 0.1Heat Flow (W/g) 20406080100120140160Temperature (C) Exo UpUniversal V3.4C TA Instruments Figure 5.4. DSC data: Glass transition temperature, Tg, of 75% HEMA: 25% DHPMA copolymer.

PAGE 165

140 89.10C(I) 86.90C 97.31C -0.4 -0.3 -0.2 -0.1 0.0 0.1Heat Flow (W/g) 20406080100120140160Temperature (C) Exo UpUniversal V3.4C TA Instruments Figure 5.5. DSC data: Glass transition temperature, Tg, of 50% HEMA: 50% DHPMA copolymer.

PAGE 166

141 87.17C(I) 85.45C 95.54C -0.4 -0.3 -0.2 -0.1 0.0 0.1Heat Flow (W/g) 20406080100120140160Temperature (C) Exo UpUniversal V3.4C TA Instruments Figure 5.6. DSC data: Glass transition temperature, Tg, of 25% HEMA: 75% DHPMA copolymer.

PAGE 167

142 84.41C(I) 82.69C 94.77C -0.4 -0.3 -0.2 -0.1 0.0 0.1Heat Flow (W/g) 20406080100120140160Temperature (C) Exo UpUniversal V3.4C TA Instruments Figure 5.7. DSC data: Glass transition temperature, Tg, of neat PDHPMA.

PAGE 168

143 Table 5.1. DSC data: Glass transition temperature, Tg, of the HEMA-D HPMA copolymer series. Polymer Molar Ratio HEMA:DHPMA Actual Tg (C) Calculated Tg (C) 100% HEMA 1:0 101.4 101.4 (act.) 75% HEMA: 25% DHPMA 3:1 95.6 96.5 50% HEMA: 50% DHPMA 1:1 89.1 92.1 25% HEMA: 75% DHPMA 1:3 87.2 88.1 100% DHPMA 0:1 84.4 84.4 (act.) Figure 5.8. DSC data: Gl ass transition temperature dependency on HEMA content.

PAGE 169

144 Dielectric Analysis (DEA) Mechanical studies show that PH EMA and PDHPMA exhibit two sub-Tg secondary relaxations and a primary glass transition [Gates et. al. 2003, Janacek 1973, Kolarik 1982]. The transitions are termed , and proceeding from the high temperature transition to the low te mperature transition. The primary transition marks the onset of large scale segmental motion of the main chain, or polymer backbone, and in the case of hydrogels it is affect ed by factors such as degree of crosslinking and water content. The relaxation corresponds to the rotati on of the ester side group and the relaxation is associated with th e rotation of th e hydroxyl group. Mechanical studies have also shown that the relaxation is very pronounced whereas the relaxation is relatively weak. The relaxation often appears as a shoulder to the peak and may even be unresolvable [Gates et. al. 2003, Russell et. al. 1980, Janacek 1973, Kolarik 1982]. Dielectric spectroscopy also id entifies all th ree relaxations as the structural groups involved possess dipole mome nts that interact with the electrical field. An interpretation of the dielectric spect rum of neat PHEMA in which the electric modulus formalism was employed to reveal as pects of the spectrum that is ordinarily hidden as a result of conductiv ity effects caused by ionic impurities was presented in chapter 3. In this section, a similar approach will be used to ch aracterize the dielectric spectra of PDHPMA and the random copolymers of HEMA and DHPMA. Relaxation It was found that peak was pronounced for the PHEMA, PDHPMA and copolymer samples in both the loss factor and electric loss modulus plots (fig. 5.9-5.23). McCrum et al. formulated a mathematical treatment of the complex permittivity, *, which is used to resolve the viscoelastic process from the conductivity effects [McCrum 1967]. By taking the inverse of the complex permittivity, *, one can obtain the electric modulus, M given by equation 5.5. 2 2 2 2" " ' 1 i iM M M Eq.5.5

PAGE 170

145 Plots of the electric loss modulus, M ", versus temperature show a significant difference from those of versus temperature with respect to the separation of the viscoelastic and conductivity relaxations, but technically contai n the same information [Starkweather and Avakian 1992]. Due to the placement of the dielectric constant, ', in the denominator of the equation, its effects in dominating M and M are reduced [Ambrus et. al. 1972, Starkweather and Avakian 1992]. This allows a more comprehensiv e analysis of the dielectric data. The relaxation obeyed Arrhenius behavior wh ich is characteristic of secondary relaxations in polymers. The Arrhenius plot of ln frequency vs. the reciprocal of temperature showed that the peak temperatur e maxima increased linearly with frequency (figs. 5.11, 5.14, 5.17, 5.20, 5.23); the slope of which was used to determine the activation energy from: RT E f fa o ln ln Eq.5.6 The orientation polarization of the – OH side group in PHEMA and PDHPMA is strongly dependent on the dipole moment; the dipole moment of the –OH group is large and is easily aligned in the electric field. The general trend observed was an increase in the activation energy of the transition from 8.9 to 15 kcal/mol as the molar concentration of DHPMA increased. It was also observed that the temperature of the peak max increased with frequency as well as with DHPMA concentration from -122.3 C to -79.8 C at 10 Hz,, as shown in table 5. 2. As the DHPMA content increased, the region also broadened. This data is in agreement w ith Gates et. al. 2003, and is explained by the greater energy needed to overcome the inte rmolecular interactions brought about by the hydroxyl groups in DHPMA to allo w rotation of these groups.

PAGE 171

146 Table 5.2. DEA data: Activation energy and movement of the relaxation. Polymer Activation Energy, EA (kcal/mol) Tmax(C) at 10 Hz Tmax (C) at 100 Hz Tmax (C) at 1000 Hz PHEMA 8.9 -122.3 -110.0 -94.9 3 HEMA: 1 DHPMA 10.3 -109.9 -95.0 -80.7 1 HEMA: 1 DHPMA 12.4 -94.9 -79.8 -64.9 1 HEMA: 3 DHPMA 13.2 -87.4 -70.0 -55.5 PDHPMA 15.0 -79.8 -64.5 -52.4

PAGE 172

147 0 2Loss Factor -150-5050150250Temperature (C) Universal V3.4C TA Instruments Figure 5.9. DEA data: Loss Modulus, E", plot for neat PHEMA.

PAGE 173

148 Figure 5.10. DEA data: El ectric Loss Modulus, M", plot for PHEMA. 0.00450.00500.0055000600.00650.0070 2 4 6 8 10 12 y = -4515x + 32.348 R2 = 0.9985 Ea = 8.9 kcal/mol ln Frequency1/Temperature (K-1) Figure 5.11. DEA data: Arrhenius plot of transition for neat PHEMA.

PAGE 174

149 0 2Loss Factor -150-5050150250Temperature (C) Universal V3.4C TA Instruments Figure 5.12. DEA data: Loss Modulus, E", plot for 75% HEMA: 25% DHPMA copolymer.

PAGE 175

150 Figure 5.13. DEA data: El ectric Loss Modulus, M", plot for 75% HEMA: 25% DHPMA copolymer. 0.00450.0050000550.0060 2 4 6 8 10 12 y = -5222.4x + 34.262 R2 = 0.99859 Ea = 10.3 kcal/mol ln Frequency1/Temperature (K-1) Figure 5.14. DEA data: Arrhenius plot of transition for 75% HEMA: 25% DHPMA copolymer.

PAGE 176

151 0 2Loss Factor -150-5050150250Temperature (C) Universal V3.4C TA Instruments Figure 5.15. DEA data: Loss Modulus, E", plot for 50% HEMA: 50% DHPMA copolymer.

PAGE 177

152 Figure 5.16. DEA data: El ectric Loss Modulus, M", plot for 50% HEMA: 50% DHPMA copolymer. 0.0040 0.0045 0.0050 00055 2 4 6 8 10 12 y = -6266.4x + 36.990 R2 = 0.99585 Ea = 12.4 kcal/mol ln Frequency1/Temperature (K-1) Figure 5.17. DEA data: Arrhenius plot of transition for 50% HEMA: 50% DHPMA copolymer.

PAGE 178

153 0 2Loss Factor -150-5050150250Temperature (C) Universal V3.4C TA Instruments Figure 5.18. DEA data: Loss Modulus, E", plot for 25% HEMA: 75% DHPMA copolymer.

PAGE 179

154 Figure 5.19. DEA data: El ectric Loss Modulus, M", plot for 25% HEMA: 75% DHPMA copolymer. 0.0040 0.0045 00050 0.0055 2 4 6 8 10 12 y = -6929.9x + 39.1223 R2 = 0.99498 Ea = 13.2 kcal/mol ln Frequency1/Temperature (K-1) Figure 5.20. DEA data: Arrhenius plot of transition for 25% HEMA: 75% DHPMA copolymer.

PAGE 180

155 0 2Loss Factor -150-5050150250Temperature (C) Universal V3.4C TA Instruments Figure 5.21. DEA data: Loss Modulus, E", plot for neat PDHPMA.

PAGE 181

156 Figure 5.22. DEA data: El ectric Loss Modulus, M", plot for neat PDHPMA. 0.00400.00420.004400046000480.00500.0052 2 4 6 8 10 12 y = -7561.7x + 41.2389 R2 = 0.99756 Ea = 15.0 kcal/mol ln Frequency1/Temperature (K-1) Figure 5.23. DEA data: Arrhenius plot of transition for neat PDHPMA.

PAGE 182

157 and Relaxations The dielectric spectrum of PHEM A showing the occurrence of the and merge was covered in detail in chapter 3. For PHEMA, at low frequencies 2 M" peaks were seen, of which one corresponded to the relaxation and the other was the relaxation. The peak was symmetrical in shape and followed Arrhenius dependency having an activation energy of 24.8 kcal/m ol. At frequencies above 6 kHz, a 3rd M" peak was observed; going from low temper ature to high temperature the 1st M" peak corresponded to the relaxation, the 2nd M" peak represented the merge and the 3rd M" peak was proven to be the conductivity relaxation. The merge occurred at higher temperatures and frequencie s and exhibited non-linear depe ndency between frequency and temperature. The relaxation was not completely resolved and in agreement with McCrum et. al. and Bergman et. al., the relaxation in methacrylate polymers was faster moving than the relaxation and tended to merge with the relaxation [McCrum et.al. 1967, Bergman et. al. 1998]. The fact that the 3rd M" peak was a conductivity relaxation based on ionic conduction and not related to a ny molecular relaxation in the polymer is proven in three ways. The following section shows these proofs but the reader is once again referred to chapter 3 for a complete explanation. Figure 5.24 show the full spectr a of electric loss modulus, M", for PHEMA, PDHPMA, and two copolymers; obvious differen ces can be seen. In neat PHEMA, three M" peaks were seen, as the DHPMA content in creased to 25% (molar), three peaks can still be seen; however, the merge is less resolved at hi gh frequencies. As the content increased to 50 and 75 % DHPMA one can notice that the 2nd low frequency M" peak is no longer symmetrical as it was in PHEMA, it has broadened and has a right shoulder. Conductivity tests prove that this peak is due to viscoelastic relaxa tion as it does not fit the conditions for a conductivity peak (figs. 5.32, 5.34, 5.36, 5.38, 5.40). Temperaturefrequency plots show that the low frequenc ies (from 0.6 Hz to 10 Hz) followed a linear Arrhenius relationship which may be indicative of the region. However, as frequency increased the relationship deviated from linear ity (figs. 5.26-5.30). This non-linear region is most likely the merge. Activation energies calculated for the relaxation using the

PAGE 183

158 low frequencies follow a trend that the act ivation energy decreased as DHPMA content increased (Table 5.3). If the assumption is made that this peak is made up of a cooperative motion between the and relaxations drawing from the observation that the peak is not entirely symmet rical as secondary peaks usually are, then this data would support the fact that th e glass transition temperature also decreased with DHPMA content as seen in DSC; therefore, less energy woul d be needed to bring about the transition. Figure 5.25 shows the trend observed as DHPMA content increased in the copolymer at 6 kHz. As DHPMA content increased conductivity effects became more pronounced as it became difficult to resolve the and relaxations. Figure 5.24. DEA data: El ectric Loss Modulus, M", vs. temperature for A) PHEMA homopolymer; B) 75%HEMA: 25% DHPMA copolymer; C) 25%HEMA: 75%DHPMA copolymer; and D) PDHPMA homopolymer.

PAGE 184

159 Figure 5.25. DEA data: Comparison of M" at 6000 Hz for PHEMA, PDHPMA and the copolymers. Table 5.3. DEA data: Activation energy and movement of the relaxation. Polymer Activation Energy (kcal/mol) (0.6Hz to 10Hz) PHEMA 24.8 3 HEMA: 1 DHPMA 24.2 1 HEMA: 1 DHPMA 21.4 1 HEMA: 3 DHPMA 20.0 PDHPMA 19.1

PAGE 185

160 Figure 5.26. DEA data: Frequencytemperature dependency of the and relaxations in neat PHEMA.

PAGE 186

161 Figure 5.27. DEA data: Frequencytemperature dependency of the and relaxations in the 75% HEMA: 25% DHPMA copolymer. Looking at figure 5.26, one will notice that the merge in the 75% HEMA copolymer occur at lower frequencies and in a shorter range of frequencies as compared to neat PHEMA.

PAGE 187

162 Figure 5.28. DEA data: Frequencytemperature dependency of the relaxation in the 50% HEMA: 50% DHPMA copolymer. The merge in the 50% HEMA: 50% DHPMA copolymer became irresolvable as frequency increased; the relaxation temperature – fre quency dependency could only be obtained from low frequencies (0.6 Hz to 10 Hz). The same was observed as DHPMA content increased. Therefore, only the relaxation, not the or merge, will be depicted for the 50% DHPMA, 75% DMPMA and 100% DHPMA polymers. It is known that the M" peak at these frequencies are due to viscoelastic relaxation; whereas as the frequency increased the M" peak exhibited conductivity relaxation characteristics.

PAGE 188

163 Figure 5.29. DEA data: Frequencytemperature dependency of the relaxation in the 25% HEMA: 75% DHPMA copolymer.

PAGE 189

164 Figure 5.30. DEA data: Frequencytemperature dependency of the relaxation in neat PDHPMA.

PAGE 190

165Conductivity Relaxation Three different proofs were shown in ch apter 3 verifying that the anomalous 2nd M" high temperature peak observed in the loss modulus plot of PHEM A was in fact not a contribution of viscoelastic re laxation but a result of ioni c conduction. The translational diffusion of ions which causes conduction is seen as a conductivity relaxation and in glass forming polymers this process takes pla ce with increasing viscous flow and usually overpowers the viscoelastic process in the dielectric loss factor spectra. Proof 1 Proof 1 explains that if the Argand pl ot, obtained in the region where the 2nd high temperature M peak is observed, reveals a true semi circular arc it can be interpreted to mean that it is indeed not a viscoelastic relaxation. Equation 5.7 below describes the behavior of a molecule, or ri gid polar liquid, having a sing le relaxation time. The semicircular arc is characteristic of the Debye model. Both th e homopolymers and the series of copolymers exhibited semi-circular Debye plots at temperatures above the glass transition region. Viscoelastic relaxations in polymers, on the othe r hand, deviate from semicircular behavior in wh ich they exhibit a distributi on of relaxation times and are often characterized by modified Cole-Col e expressions [McCrum 1967]. Figures 5.315.40 show the Argand plot for the polymer seri es where the values proceed from lower to higher frequencies. The plots show data de rived from the conductivity relaxation region and the glass transition region. 2 2 22 2 R U R UM M M M M M Eq.5.7 Comparing the Argand plots of the copolymer series one will observe two things: 1) the Argand plot generated from the conduc tivity relaxation region (200 C) is semicircular following the Debye model; whereas the plot in the glass tr ansition region (100 C) deviates from Debye beha vior and 2) as DHPMA conten t increases the Argand plot in the glass transition region appears to look more like a se mi-circle. This is another indication that the region in high DHPMA content copolymers is affected by conductivity more than in hi gh HEMA content copolymers.

PAGE 191

166 Figure 5.31. DEA data: Ar gand plot derived from the conductivity relaxation region (200 C) for neat PHEMA. Figure 5.32. DEA data: Argand plot derived from the gl ass transition region (100 C) for neat PHEMA.

PAGE 192

167 Figure 5.33. DEA data: Ar gand plot derived from the conductivity relaxation region (200 C) for 75% HE MA: 25% DHPMA copolymer. Figure 5.34. DEA data: Arga nd plot derived from the glass transition region (100 C) for 75% HEMA: 25% DHPMA copolymer.

PAGE 193

168 Figure 5.35. DEA data: Ar gand plot derived from the conductivity relaxation region (200 C) for 50% HE MA: 50% DHPMA copolymer. Figure 5.36. DEA data: Arga nd plot derived from the glass transition region (100 C) for 50% HEMA: 50% DHPMA copolymer.

PAGE 194

169 Figure 5.37. DEA data: Ar gand plot derived from the conductivity relaxation region (200 C) for 25% HE MA: 75% DHPMA copolymer. Figure 5.38. DEA data: Arga nd plot derived from the glass transition region (100 C) for 25% HEMA: 75% DHPMA copolymer.

PAGE 195

170 Figure 5.39. DEA data: Ar gand plot derived from the conductivity relaxation region (200 C) for neat PDHPMA. Figure 5.40. DEA data: Arga nd plot derived from the glass transition region (100 C) for neat PDHPMA.

PAGE 196

171Proof 2 The second proof involved fitting the data to equation 5.8, an equation derived by Ambrus et al. in which the electric modulus is presented in terms of time, frequency and modulus [Ambrus et. al. 1972]. Starkweather Jr et. al. also employe d this equation to show that plots of log M and log M' vs. log frequency will reveal slopes of 1 and 2, respectively, if the electric modulus ( M) is due purely to ioni c conduction as a result of ionic diffusion and independent of viscoelastic, dipolar rela xation [Avakian et. al. 2002, Starkweather Jr. et. al. 1992]. Please refer to ch apter 3 for a detail explanation. Both the homopolymers and the series of copoly mers revealed slopes of 1 and 2 for M", M dependence on frequency at temperatures above the glass tr ansition region. 2 2 21 1 ) 1 ( s sMi M i i MsM Eq.5.8 Figure 5.41 to 5.50 show plots of M', M" dependency for neat PHEMA, neat PDHPMA and the HEMA:DHPMA copolymers. It is interesting to note that as the DHPMA content increased the slope value appr oached the ideal value. For example, the actual slope for the M' plot and the M" plot for neat PHEMA is a 1.69 (ideal = 2) and 0.96 (ideal = 1); whereas the actual slope for the M' plot and the M" plot for neat PDHPMA is a 1.77 (ideal = 2) and 0.99, respectively. This fact establishes the interpretation that conductiv ity effects are more dominant in DHPMA than HEMA.

PAGE 197

172 Fig. 5.41: DEA data: Dependence of M' on frequency in the conductivity relaxation region (165 C) for neat PHEMA. Fig. 5.41: DEA data: Dependence of M" on frequency in the conductivity relaxation region (165 C) for neat PHEMA.

PAGE 198

173 Fig. 5.43: DEA data: Dependence of M' on frequency in the conductivity relaxation region (165 C) for 75% HEMA: 25% DHPMA copolymer. Fig. 5.44: DEA data: Dependence of M" on frequency in the conductivity relaxation region (165 C) for 75% HEMA: 25% DHPMA copolymer.

PAGE 199

174 Fig. 5.45: DEA data: Dependence of M' on frequency in the conductivity relaxation region (165 C) for 50% HEMA: 50% DHPMA copolymer. Fig. 5.46: DEA data: Dependence of M" on frequency in the conductivity relaxation region (165 C) for 50% HEMA: 50% DHPMA copolymer.

PAGE 200

175 Fig. 5.47: DEA data: Dependence of M' on frequency in the conductivity relaxation region (165 C) for 25% HEMA: 75% DHPMA copolymer. Fig. 5.48: DEA data: Dependence of M" on frequency in the conductivity relaxation region (165 C) for 25% HEMA: 75% DHPMA copolymer.

PAGE 201

176 Fig. 5.49: DEA data: Dependence of M' on frequency in the conductivity relaxation region (165 C) for neat PDHPMA. Fig. 5.50: DEA data: Dependence of M" on frequency in the conductivity relaxation region (165 C) for neat PDHPMA.

PAGE 202

177Proof 3 When viscoelastic effects are negligible the loss factor is described by equation 5.3. Figures 5.51 to 5.60 show plots of th e frequency dependence of ac conductivity ( ac) for temperatures above Tg where conductivity is predomin ant for both the homopolymers. Dc conductivity ( dc) was obtained by extrapolation to zer o frequency. As temperature is increased, the frequency dependence of ac conduc tivity plateaus and is independent of all frequencies measured. Dc conductivity ( dc) follows an Arrhenius relationship expressed by the equation 5.9, where E is the apparent activation energy, k is Boltzmann’s constant and o is the pre-exponential factor [Polizos et. al. 2000]. ) exp( log log kT Eo dc Eq.5.9 Table 5.4 shows the ionic conductivity act ivation energy for the copolymers. The ionic conductivity ac tivation energy is the energy required to bri ng about the translation diffusion of ions in the polymer matrix. As shown in table 5.3, the activation energy decreased, from 10.1 to 5.6 kcal/mol, as DHP MA content increased. Therefore it can be concluded that DHPMA facilitates ionic m ovement through the polymer matrix better than HEMA; a conclusion also determined by Gates et. al. whose ion transport studies showed higher ion diffusion (of both Na+ and K+) in PDHPMA than PHEMA [Gates et. al. 2003]. Table 5.3. DEA data: Ionic conductivity activ ation energy. Polymer Ionic Conductivity Activation Energy (kcal/mol) PHEMA 10.5 3 HEMA: 1 DHPMA 9.9 1 HEMA: 1 DHPMA 7.1 1 HEMA: 3 DHPMA 6.3 PDHPMA 5.6

PAGE 203

178 Figure 5.51. DEA data: Frequency depe ndence of ac conductivity for neat PHEMA. Figure 5.52. DEA data: Ionic conductiv ity activation energy for PHEMA.

PAGE 204

179 Figure 5.53. DEA data: Frequency depe ndence of ac conductivity for the 75% HEMA: 25% DHPMA copolymer. Figure 5.54. DEA data: Ionic conductivit y activation energy for the 75% HEMA: 25% DHPMA copolymer.

PAGE 205

180 Figure 5.55. DEA data: Frequency depe ndence of ac conductivity for the 50% HEMA: 50% DHPMA copolymer. Figure 5.56. DEA data: Ionic conductivit y activation energy for the 50% HEMA: 50% DHPMA copolymer.

PAGE 206

181 Figure 5.57. DEA data: Frequency depe ndence of ac conductivity for the 25% HEMA: 75% DHPMA copolymer. Figure 5.58. DEA data: Ionic conductivit y activation energy for the 25% HEMA: 75% DHPMA copolymer.

PAGE 207

182 Figure 5.59. DEA data: Frequency depe ndence of ac conductivity for neat PDHPMA. Figure 5.60. DEA data: Ionic conductivit y activation energy for neat PDHPMA.

PAGE 208

183Conclusion The dielectric spectra of a series of copolymers of 2-hydroxyethyl methacrylate (HEMA) and 2,3-dihydroxypropyl methacrylate (DHPMA) have been investigated. Chapter 3 presented an interpretation of the dielectric spectrum of PHEMA where the electric modulus formalism was used to reveal the viscoelast ic and conductivity relaxations present in the polymer. This study looked at the effects on the dielectric behavior as a result of 2, 3-dihydroxypropyl methacrylate add ition. To the best of the authors’ knowledge, this is the first study presenting the dielectric response of these materials up to and above the glass transiti on region. It was important to study this as DHPMA has been proven to be an excellent ma terial for bio-applications, and is often used as a co-monomer unit with HEMA. Several notable changes were observe d as 2,3-dihydroxypropyl methacrylate concentration increased. The glass transition temperature decreased, the activation energy increased, the activation energy decreased and ionic conductivity increased with DHPMA content. Overall, it was noted as DHPMA content increase d conductivity effects became more pronounced as it became difficult to resolve the and relaxations. Also it was recorded that DHPMA facilitates the m ovement of ions through its matrix more efficiently than in HEMA.

PAGE 209

184CHAPTER 6 Biocompatible Hydrogel Coating fo r an Implantable Glucose Sensor Introduction Foreword The objective of this research project is to formulate, modify and characterize a biocompatible coating for an implanted gluc ose sensor device. The project has been funded by the National Institute of Health (Grant # 5R01EB001640-02) and the research has been conducted under the supervision of Principal Investigator (PI), Dr. Francis Moussy of the Department of Chemical En gineering (USF), and co-PI, Dr. Julie Harmon of the Department of Chemistry (USF). Some of the work presented in this dissertation chapter has been conducted by Dr. Moussy a nd his research group, and are included in this section to lend an un derstanding for the overall obj ective of the project. Implantable Sensors Implantable medical devices have been around for many years; for example, the first implantation of a heart pacemaker in a human occurred in 1960 [Jeffrey 2001]. In recent years, the market for medical electroni cs has grown rapidly as the medical sector has turned to more sophisticated solutions fo r the identification and treatment of illnesses, and the improvement of patient care. An emerging trend is the move toward miniaturization of equipment and implan table sensor devices [Ake Oeburg 2004]. Implantable sensor devices include blood gl ucose monitoring systems, insulin pumps, and body temperature sensors; other implants ra nge from defibrillators to neurological stimulators, pacemakers and cochlear hearing aids. These products not only simplify the testing, monitoring, and treatment processes, but also help to improve the quality of life for the patient. Implantable devices help by minimizing the time patients spend in hospitals and often provide automatic, c ontinuous treatment of chronic conditions. Dr. F. Moussy and his research group in the Department of Chemical Engineering, USF, have developed an im plantable biosensor for the monitoring of glucose. Glucose monitoring is an important step towards controlling the metabolic

PAGE 210

185 disease known as diabetes. There are two type s of diabetes: type I and type II. Type I diabetes is the result of the immune syst em destroying the body’s insulin-producing cells of the pancreas; glucose control can only be achieved by insulin injections. Type II diabetes is the result of the body not produci ng sufficient insulin which results in reduced uptake of glucose by the cells in the body; as a result, the sugar leve l in the blood remains elevated and this can lead to complications wi th the eyes, kidney, nerves and heart. It is important for diabetic patients to maintain their glucose concentration to near-normal levels to reduce the occurrence of diabetes complications [Heller 1999]. Conventional glucose testing involves pricking the patien ts’ finger with a lancet (a small, sharp needle), putting a drop of blood on a test strip and then placing the strip into a meter that displays the blood sugar (g lucose) level. Meters vary in features, readability (with larger displays or spoken instructions for the visually impaired), portability, speed, size and cost. Current devi ces provide results in less than 15 seconds and can store this information for future use. These meters can also calculate an average blood glucose level over a period of time [Haines 2005]. Unfort unately, the pain associated with finger-stick assays deter many patients from frequent monitoring which usually should be measured several times a day. A recent article in www.DiabetesSelfManagement.com covered the topic of children who manage their diabetes often falsify their gl ucose record in their log books for many reasons; the first reason being the pain and inconvenience of the finger-stick testing [Roemer 2004]. Alternative glucose sensors are currently being investig ated, the first continuous FDA approved glucose sensor/i nsulin pump combo has been introduced by Medtronic Diabetes, a Minimed monitor is shown in fi gure 6.1. In this system the sensor is implanted subcutaneously, a lead is connect ed through the skin to a radio-frequency transmitter that is taped onto the skin and, then this transmitter sends a signal to the monitor. This system allows for continuous glucose measurement; however, the sensor can only be worn for up to 72 hours due to loss of sensitivity to glucose in vivo [Kerner 2001]. This type of device is for initial assessm ent of the patients’ glucose profile and is not for long term use. Problems such as bleedi ng, swelling, irritation and infection at the

PAGE 211

186 insertion site are possible risks associated wi th inserting the sensor and may result from improper insertion and maintenance of the insertion site. Figure 6.1. Medtronic MiniMed Guardian RT system [www.minimed.com, Medtronic MiniMed 2005]. Another current product on the market, al though not an implanted sensor, is the GlucoWatch Automatic Glucose Biographer by Cygnus Inc. It works by applying an electrical potential to the skin which causes gluc ose to travel to the surface of the skin via an electro-osmotic flow, the glucose is then measured by an enzyme electrode [Kerner 2001]. GlucoWatch is not a replacement for fi nger-stick arrays; in fact, the makers of GlucoWatch insist on concomitant us e with finger-stick glucose sensors [www.glucowatch.com]. The GlucoWatch has a 15 minute lag time and often results in irritation to the skin. Microdialysis Versus Electrochemical Sensors The aim of current research endeavors is to produce a sensor that implants subcutaneously, defends against the body’s na tural fouling attempts and resists loss of sensitivity to glucose over time. At the moment, two systems are used for glucose sensors; sensors are based upon either an el ectrochemical system or a microdialysis system. Microdialysis technology aims to simula te the action of capillaries. In the Roche Microdialysis System ( www.roche.com ) a catheter which contains a thin dialysis fiber is implanted into the patient’s subcutaneous fa tty tissue. The subcutaneous fatty tissue has

PAGE 212

187 been shown to have a glucose concentrati on that very closely resembles the glucose concentration in venous plasma [Thomas et. al 1998]. The fiber is irrigated with isotonic glucose-free Ringer fluid. This irrigating fluid is in a state of constant interchange with the interstitial fluid surrounding the catheter. As a result of the prevailing concentration gradient, glucose migrates from the intersti tial fluid into the glucose-free Ringer fluid. Fig. 6.2. A Schematic of the Roche Micr odialysis System, a) Microdialysis probe implanted in subcutaneous ad ipose tissue, and b) Fluid being pumped to a glucose sensor outside the body [Roche 2002].

PAGE 213

188 The glucose-enriched Ringer solution is pumped to a glucose sensor which is connected outside the body, where the glucos e concentration is measured continuously. Current research shows that as long as th e blood glucose concentr ation stays constant then the glucose supplied to tissues via th e capillaries and theref ore the microdialysis probe will be equal to blood glucose [Wientje s et. al. 1998]; however, abrupt changes in blood glucose levels can cause the glucose levels in the capillaries and interstitial fluid to differ. In addition to a physio logical time lag, a physical time lag of 30 minutes must also be taken into account in microdi alysis measurements. The adva ntage to this is that the probe is unaffected by the body’s fouling attemp ts which means that the system is not subjected to loss of sensitivity. The disadvant age is that the implan ted probe leads to a sensor which is situated outsi de the body and, as with th e Medtronic Minimed monitor, there is a risk for infect ion and complications. In electrochemical systems, an amperometric measurement of hydrogen peroxide, generated by enzymatic oxidation of glucose by glucose oxidase, is used to calculate the glucose concentration in vivo [Linke et al. 1999, Heller 1999, Yu et al. 2005, Pickup et. al. 1988]. The sensor designed by Moussy and his research group is an electrochemical amperometric sensor; a schematic diagram of the sensor is shown in figure 6.3. Fig. 6.3. A Schematic diagram of the coil -type implantable elec trochemical glucose sensor based on a coiled Pt-Ir wire. 1elec trically-insulating sealant; 2-Teflon-covered platinum wire; 3outer membrane; 4cotton fiber with enzyme gel; 5stripped platinum wire; 6enzyme layer. [Yu et. al 2005 Reproduced by permission of Frontiers in Bioscience ]

PAGE 214

189 This sensor design utilizes a novel exce ss-enzyme loading technique that has shown promising results in extending the lifetime of the sensor; the lifetime of the sensor using this technique increased to 60, up to 120, days in vitro depending on the composition of the outer membrane. Loss of sensor functionality occurs when the outer membrane deteriorates. Membrane defects such as micocracks and pinholes, which can be caused by non-uniform coating application, result in erroneous glucose readings [Yu et. al. 2005]. Improving the stability and bioc ompatibility of the outer membrane on the sensor should help to reduce adverse tissue re actions and potentially extend the life of the sensor in vivo Tissue Interactions with Implantable Sensors When an implant device is placed insi de the body it is don e so via invasive surgical procedures. These procedures cause cell, tissue and, possible, organ injury depending on the implantation site. The injury triggers the body’s natural response to repair the damaged area. This remarkable complex response involves a sequence of interdependent processes that overlap in time; however, simplistically it can be viewed as a two step process [Dee et. al. 2003, Hickey et al. 2002]. This two-st ep process involves 1) inflammation and 2) wound healing. When damage to blood vessels in vascularized tissue occur a fibrin mesh, commonly known as a blood clot, plugs the injury. The blood clot provides a temporary protection for the w ound and also acts as a matrix for cells to attach and migrate into during the healing pr ocess. The process of blood coagulation and activation of various chemical reactions init iate the inflammation stage. Macrophages and phagocytes clean up the damaged area of any dead cells, extracellular debris and bacteria by engulfing and ingesting the unwanted mate rial. Fibroblast, plat elet and vascular endothelial growth factors are released from the macrophages to begin the second step of wound healing [Dee et. at. 2003]. Fibroblasts begin to synthesize an extr acellular matrix made up primarily of collagen; vascularization of this newly fo rmed tissue follows. Inhibitors of protein synthesis provide the controlli ng balance in this process. In the end the new tissue will have blood vessels and cells necessary for the specific function of that tissue in the body.

PAGE 215

190 In the case where that process becomes unc ontrolled, as is the case in chronic tissue interaction with the implant, extensive tissu e fibrosis occurs and the implanted sensor becomes encapsulated in scar tissue. The resu lting scar tissue has less vascularization than normal tissue and as a result the concentration of glucose and oxygen in the surrounding scar tissue is lower [Linke et. al. 1999, Hickey et. al. 2002]. To maintain the sensitivity of implanted glucose sensors it is imperative to reduce the tissue interaction with the sensor. Differe nt approaches are be ing investigated to control the inflammation process and the encap sulation of the sensor. Some researchers are looking into mediated anti-inflammatory drug release [Hickey et. al. 2002, Patel et. al. 2006, Zhong et. al. 2005, Hahn et. al. 2004], wh ile others are focusing on coating the implant with a biocompatible coating [Karpm an et. al. 2001, Bottcher 2000, Lugscheider et. al 1991]. Biomaterials Biomaterials are used in numerous medical applications; for the most part it is a material that will replace a part, or functi on, of the body [Hench and Ethridge 1982]. As a result, it will have direct contact form ing an interface between non-living and living substances. Its interaction with the body will de termine its long term stability and its final end use as a product. The type of material th at can be used as a biomaterial ranges from metal to ceramic to polymeric. These biomat erials are used primarily for orthopedic implants, but new and innovative materials are being used to build artificial organs, and promote bone regeneration. Achieving a high degree of biocompatibility and unique surface properties will lead to a new generation of materials for applications in both short and long term implantable devices. These new materials wi ll provide satisfactory performance for specific applications in contact with cel ls, tissue, or blood [Tavakoli 2005]. For the purpose of this research, we will look at biocompatible thin films as coatings for implantable devices. Biocompatible thin films, used to date include films made from polyurethane, polyvinyl alcohol, polyethylene glycol and other hydrogel form ing polymers [Santerre et.

PAGE 216

191 al. 2005, Ali et. al. 1993, Lai and Baccei 1991, Mansur et. al. 2004]. Hydrogel materials show promise as bio-coatings because of thei r capacity to act as humectants in the wound area; its high water equilibrium content make s it soft and flexible, plus its high porosity allows the diffusion of analytes through its matrix to the sensor [Kejlova et. al. 2005]. The objective in this research is to formulate, modify and characterize a biocompatible coating for an implanted gl ucose sensor device. This coating should 1) be permeable to allow glucose, oxygen and hydrogen peroxide to diffuse freely, 2) reduce adsorption of protein from surrounding cell and plasma, 3) result in minimal fibrosis by having an interface that is compatible with the tissue. For the purpose of this research random copolymers of 2-hydroxyethyl methacrylate (HEMA) and 2,3-dihydroxy propyl methacrylate (DHPMA) will be used to develop a thin, biocompatible coating for the implantable glucose sensor that was designed by Dr. F. Moussy and his research group. Experimental Materials 2-hydroxyethyl methacrylate and 2,3-di hydroxypropyl methacrylate monomers were generously donated by Benz R&D (Sarasot a, FL). They were used as received without further purification. Ethylene glycol dimethacrylate a crosslinking agent, was obtained from Aldrich (St. Louis, MO). Th e free radical initiator employed for the UVinitiated polymerization was 2-hydroxy-2methyl-1-phenyl-1-propanone (Benacure 1173 ) by Mayzo (Norcross, GA). Benacure 1173 is a highly efficient, non-yellowing liquid photoinitiator that is recommended for UV inks and coatings. Please refer to figure 2.3. for the decomposition scheme of Ben acure 1173. Phosphate buffered saline solution (PBS, pH 7.4) was obtained from Fisher Scie ntific (New Jersey) and used for the water equilibrium content study.

PAGE 217

192Synthesis of UV-Polymerized Copolymer Rods Random copolymers of 2-hydroxyethyl methacrylate (HEMA) and 2,3dihydroxypropyl methacrylate (DHPMA) have been synthesized in a series of various molar ratios. The molar concentration of the crosslinking agent, ethylene glycol dimethacrylate (EGDMA), was kept constant at 2%, and the molar ratio between the two monomers was varied. The polymerization was carried out in an inert argon atmosphere in an in-lab built UV reactor using a wave length of 254nm for 24 hours. These hydrogels were polymerized via UV initiation in Tefl on tubing plugged with wax at one end. The wax was melted and then one end of the stra ightened Teflon tubing (Voltrex tubing, SPC Technology) was dipped into the molten wax. By capillary action, the wax was drawn up the tube to produce an upper surface with a concave meniscus. The monomer mixture was injected into the tubing using a 22 gauge needle to avoid any bubble formation. The rounded (smooth) edge of the resulting copol ymer rod reduces inte rfacial interaction; thereby, minimizing tissue reaction. Samples were post-cured in a vacuum oven at 110 C for two hours. Water Equilibrium Content, Gel Fraction and Biocompatibility Studies Three samples, weighing approximately 0.5g each, of each homopolymer and copolymer were prepared for equilibrium cont ent studies. They were dried to constant weight in a vacuum oven at 110 C. The initia l dry weight was recorded and then each sample was placed in a capped 50ml glass jar co ntaining PBS; the jars were stored in an oven at internal body temperature: 36.9 + 0.5 C. The hydrated samples were weighed every 7 days until constant mass. The final water equilibrium content (% change) was then calculated. In order to study the degree of cross-li nking (sol-gel ratio) and to identify the extent of polymerization of the monomer, the standard extraction technique has been applied. Gel fraction (fgel) was obtained via Soxhlet extractio n using distilled water as the extracting solvent. A set of three samp les (~0.7g each) was prepared; they were encapsulated in Whatmann 2 filter paper e nvelopes and the dry weight was obtained

PAGE 218

193 before and after extraction. The extraction was performed for 7 days. Samples were vacuum oven dried before and after extrac tion at 100C for 8 hours. The gel fractions (fgel) were calculated from the following equation: 0 gel gelw w f Eq. 6.1 ,where w0 and wgel are dry weights of the samples before and after extraction, respectively [Gerasimov 2002]. Samples were prepared for implantation in rat specimens. The polymer rods were first washed continuously in the Soxhlet ex traction apparatus usi ng distilled water to remove any unreacted monomer, and were th en placed in capped vials containing PBS. The samples were sterilized using a Tuttn auer-Brinkmann 2340E steam autoclave for one hour at 122 C and under 16psi.The biocompa tibility studies were carried out by a certified veterinarian who is a member of Dr. F. Moussy’s research group. The samples were subcutaneously implante d in the rat. Explantation was performed 3 and 28 days after implantation; which was then followe d by histopathology slid e preparation where the tissue/hydrogel interface were sliced us ing a microtome, set on glass slides and stained with Hematoxilin & Eosin. Coating Trials: Dip Coating via UV-Polymerization In Situ Trials to coat the sensors involved di p-coating the polyurethane coated metal sensor in the monomer-initiator mixture, followed by UV initiated polymerization. The drawbacks will be discussed. Static contact angle measurements were made using a VCA Optima, AST Products, Inc. (courtesy of Tran sitions Optical Inc., St. Petersburg FL). 0.75 l of distilled water, or HEMA monomer, was used to study the wetting behavior and surface treatments of the polyurethane/epoxy coating.

PAGE 219

194Gamma Irradiation Grafting A JL Shepherd Mark cesium-137 irradiator was employed to initiate polymerization of HEMA and grafting of PHEMA unto a polyurethane thin film. The polyurethane (PU) (Selectophor e) was obtained from Sigma Aldrich/Fluka, the 5 minute general epoxy, Perma Oxy, was obtained fr om Permatex, a non-ionic surfactant Brij 30 (polyethylene glycol dodecyl ether) and st abilized Tetrahydrofuran (THF) were both obtained from Sigma-Aldrich. The polyur ethane-epoxy/THF solution was prepared by dissolving 60% PU: 40% epoxy to make a 2.5% (by wt.) THF solution. 5% Brij 30 (of total PU-epoxy mass) was then added to the solution. The solution was then pipetted unto a clean glass slide. After 30 minutes of allowi ng the THF to evaporate, the glass slide was placed in an oven at 120 C for one hour to cure. The resulting coating was opaque in appearance, hard and adhered quite well to the glass slide. These PU-epoxy coated glass slides were placed in 20ml glass vials containing 1) 100% hydroxyethyl methacrylate (HEMA), 2) 50% HEMA: 50% DI water, 3) 50% HEMA: 50% methanol, 4) 10% HEMA: 90% DI water, and 10% HEMA: 90% methanol. The radiation dosage was optimized to result in PHEMA grafting w ithout gelation of the polymer. Only HEMA monomer was used to determine experiment al conditions, as DHPMA is costly and would only be utilized when experimental c onditions and procedures are optimized. After irradiation grafting the slides were rem oved from the reaction vials and washed voluminously with distilled water and metha nol, followed by drying in a vacuum oven at 110 C for one hour. FTIR was then taken after the irradiation grafting to determine if the surface functionality of the PU -epoxy coating changed. Plasma Polymerization The plasma polymerization system was bui lt and manufactured by March Plasma Systems ( www.marchplasma.com ). Use of the instrument was provided to Dr. Harmon’s research laboratory complements of March Pl asma Systems; we are indebted to March Plasma employees, especially the applica tions manager Mr. Lou Fierro, for their unending support and kindness.

PAGE 220

195 The plasma experimental set-up consists of a reactor system containing an upper and lower electrode, a gas flow system allo wing the introduction of multiple gases, an inlet for vaporized monomer, and a 40 kHz RF power supply. The set-up also included a shielding stage and a vacuum system. Th e software developed by March Plasma, P2CIM 2000, was used to control system operations. Before any experiments, the system was purged with an oxygen burn to remove any cont amination from within the sample stage area. An oxygen purge was carried out at 0.150 L/min under an output power of 650Watts for 20 minutes. Oxygen is a non-plasma forming gas, and produces a characteristic white/violet glow. Attempts were made to deposit a thin plasma film of poly(2-hydroxyethyl methacrylate) (PHEMA). Deposition conditi ons were varied, the variables included vaporization temperature of the monomer, argon gas flow rate, output power and deposition time. The plasma films were tested using a Ni colet Avatar 320 FTIR equipped with the Smart Miracle ATR accessory for attenuated total reflectance scanning capabilities. Attenuated total reflectance infrared (ATR /IR) spectrometry can provide valuable information related to the chemical struct ure of polymer films and membranes. Midinfrared spectra are obtained by pressing the polymer film against an internal reflection element, for this particular set-up zinc seleni de (ZnSe) was used. IR radiation is focused onto the end of the element wher e light enters the element a nd reflects down the length of the crystal. At each internal reflection, the IR radiation penetrates a short distance (~1 m) from the surface of the element into the polymer membrane. It is this unique physical phenomenon that enables one to obtai n infrared spectra of samples without performing much sample preparation, such as needed with KBr pellets [Skoog et. al. 1992, MicroMem Analytical 2006]. The functionality of the pl asma films were compared to that of conventional PHEMA. A Digital Instruments Atomic force microscope (AFM) Nanoscope with Nanoscope Control 5.12rs (AFM) was used to determine the plasma film thickness. We thank Dr. Emirov from the Nanomaterials a nd Nanomanufacturing Research Center at USF for his expertise in obtaining the AFM images.

PAGE 221

196Results and Discussion Equilibrium and Biocompatibility Studies Random copolymers of 2-hydroxyethyl methacrylate (HEMA) and 2,3dihydroxypropyl methacrylate (DHPMA) cr oss-linked with ethylene glycol dimethacrylate were synthesized in a series of various molar ratios. Figure 6.4. shows a sample of the UV polymerized polymer rod; the rounded, smooth edge was designed to reduce interfacial interaction; thereby, minimizing tissue reaction. The crosslinker content was kept constant at 2% (molar), and the HEMA:DHPMA content was varied. The equilibrium study showed that the hydrogels took approximately 7 weeks to equilibrate in the PBS solution at 37 C. Ta ble 6.1. shows the final water equilibrium content of the hydrogels at the end of the 7 weeks. An increase of DHPMA in the copolymers leads to changes in the swelling behavior, network structure, mechanical strength and polymer-water interaction in the hydrogels, viz. an increase of equilibrium water content. As observed the PDHPMA absorbed more than 3 times the amount of water than PHEMA; this is due to the ex tra hydroxyl group presen t in DHPMA. Gates et. al. reported a % water equilibrium content of 38% for neat PHEMA and 75% for neat PDHPMA at 23 C [Gate 2001]. As presented in chapter 5, the dielect ric interpretation of neat PDHPMA indicated that it possesses a more open network than PHEMA; diffusion experiments carried out by Benz R&D confirm this observation. It was noted that the high concentration DHPMA copolymers and th e 100% neat PDHPMA polymer exhibited a loss in mechanical strength integrity. Simila r observations have been reported in other studies using different mono mers where the addition of highly hydrophilic monomers often lead to the fabrication of highly frag ile materials [Jarvie et al. 1998, Seo et. al. 2004].

PAGE 222

197 Table 6.1. % Water equilibrium content of the HEMA-DHPMA copolymer series. Hydrogel Series % Water Equilibrium Content 100 % HEMA 52.88 + 0.83 80% HEMA: 20% DHPMA 77.14 + 1.09 60% HEMA: 40% DHPMA 97.06 + 0.94 40% HEMA: 60% DHPMA 118.13 + 1.31 20% HEMA: 80% DHPMA 141.27 + 1.95 100% DHPMA 166.47 + 2.21 The molar ratio of EGDMA was kept constant at 2%, the remaining 98% of the hydrogel composition was devided between HEMA and DHPMA molar % concentration. Soxhlet extraction was performed on the hydrogels for one week using methanol as the extracting solvent. This enabled us to determine the amount of unreacted monomer and linear chain polymer present in the hydrogel. It is important to rid the sample of these substances before implantation as unreacted monomer can cause adverse tissue reaction. From table 6.2, it can be seen that the 100% polyHEMA hydrogel contained ~21% of linear polymer chain and unreacted monomer that can be washed out of the crosslinked hydrogel; as the DHPMA conten t increased the sol-gel fraction in the hydrogels increased. Fig. 6.4. UV polymerized polymer rod with rounded, smooth edge. 0.8mm

PAGE 223

198 Table 6.2. Gel fraction of copolymer series. Hydrogel Series Gel Fraction (Amount of crosslinked polymer) 100% HEMA 0.79 + 0.04 80% HEMA:20% DHPMA 0.83 + 0.03 60% HEMA:40% DHPMA 0.96 + 0.03 40% HEMA:60% DHPMA 0.95 + 0.01 20% HEMA:80% DHPMA 0.96 + 0.06 100% DHPMA 0.96 + 0.06 These samples performed best in terms of biocompatibility and mech anical properties. After Soxhlet extraction, the equilibrate d hydrogels were sterilized using a steam autoclave and were then implanted into the subcutaneous layer of the rat. Explantation was performed 3 and 28 days after implantati on (figure 6.5), which was then followed by histology slide preparation. The hist ograms of the hydrogels show varying biocompatibility depending on the copolym er formulation. The 80% HEMA:20% DHPMA and 60% HEMA:40% DHPMA hydrogels gave the best results in terms of biocompatibility and mechanical properti es (figs. 6.6-6.8). High DHPMA content copolymers broke easily producing sharp edges and fragments; thereby, inducing fibrosis.

PAGE 224

199 Figure 6.5. Explantation of the subcutaneously implanted hydrogel rods. Forceps point to the area where the hydrogels are located. [Courte sy of Dr. Moussy’s research laboratory] Figure 6.6. Histology image of PHEMA rod, explanted afte r 28 days. Dark purple outline indicates scar tissue formation (fibrosis). [Courtesy of Dr. Moussy’s research laboratory]

PAGE 225

200 Figure 6.7. Histology image of 80%HEMA: 20%DHPMA rod, explanted after 28 days. Minimal to no fibrosis. [Courtesy of Dr. Moussy’s research laboratory]

PAGE 226

201 Figure 6.8. Histology image of 100% PDHPMA rod, explanted after 28 days. Highly fragile sample, fibrosis induced. [Courtes y of Dr. Moussy’s research laboratory] Drawbacks of Dip Coating The wettability and adhesion of the methacrylate monomer, HEMA, to the polyurethane-epoxy coating that is present as a coating on the metal wire sensor surface was investigated. UV, photoinitiated polymeri zation was chosen as the polymerization route for several reasons. UV radiation curing is a techniqu e that enjoys an advantage over other curing techniques used in the industr ial setting. Its use is especially noted in the coating industry where the application of thin polymer films on a variety of surfaces is used for surface protection; it is also co mmon in the dental health care industry where many of the composite fillers are cured with in seconds of high intensity UV radiation [Decker 1998, Sibold et. al. 2002]. In order for photo-polymerization to proceed the medium must absorb light to produce an initiating species, for the polymerization of methacrylate monomers a UV phot oinitiator is added. High energy radiation, whether it is UV or gamma, is known to produce ionization and excitation in polymer and mono mer molecules. The formation of ions and free radicals usually signify that the m onomer and polymer undergo dissociation,

PAGE 227

202 abstraction and addition reactions leading to ch emical stability. The stabilization process, which can occur immediately, or may take days, months or years, often result in crosslinking or chain scissi on [Tatro 2002, Janik et. al. 2002, Skaja and Croll 2003, Kim and Urban 2000. Chain scission is the breaki ng of a molecular bond causing the loss of a side group or shortening of the overall chai n, and crosslinking is when individual polymer chains are linked together by covale nt bonds to form one insoluble network. Both chain scission and crosslinking occur during the radiation of a polymer. However, usually one process dominates the other, and this is dependent on th e polymer structure, atmosphere, temperature etc. [Tat ro 2002, Clough and Shalaby 1996]. In the UV polymerization of HEMA monomer unto the surface of the polyurethane/epoxy surface, several possible reac tions can occur concurrently. Previous studies by several researcher s have shown the photodegrada tion via chain scission of polyurethane/epoxy coatings by UV radiation [Wang et. al. 2005, Kim and Urban 2000, Skaja and Croll 2003]. The commonly accepted mechanism by which chain scission occurs is through the C-N and N-H linkages pr esent in the urethane bond, which can then react with hydrogen and oxygen to promote polym er degradation. At the same time, the UV photoinitiator undergoes photolytic decompos ition; the free radical can then react with the HEMA monomer to begin the polymer ization process (see ch. 2. for free radical reaction mechanisms).The irradiation of th e PU/epoxy coating can potentially result in surface activation; followed by reaction with th e activated monomer to form a covalently bound PHEMA surface. Contact angle analysis is the m easure of the angle of contact, between a liquid and a surface. The contact angle is an inverse measure of the ability of a particular liquid to “wet” the surface. This analysis involves th e interfacial free ener gies between the three phases and is given by: lv cos = sv sl, where lv sv and sl refer to the interfacial energies of the liquid/vapor, solid/vapor and solid/liquid interfaces. Wetting occurs when SV > SL, and nonwetting occurs when SV < SL. As a result, if th e liquid droplet has a higher free energy than the surface of the s ubstrate then the liquid will bead on that surface [Gersten and Smith 2001].

PAGE 228

203 If water is used as the liquid, the sm aller the contact angl e the more hydrophilic the surface [Dee et. al. 2003]. It was found th rough contact angle measurements that the HEMA monomer produces a droplet whic h has a high contact angle, 93 + 1.5 when in contact with the polyurethane coating (tab le 6.3 and figure 6.9), this was further exemplified when attempts were made to UV polymerize HEMA monomer on the surface of the PU/epoxy coated sensor. Table 6.3. Contact angle measurements of HEM A on glass and PU/epoxy coated glass surface. Sample Surface Contact Angle HEMA Drop 1 Contact Angle HEMA Drop 2 Contact Angle HEMA Drop 3 Contact Angle HEMA Avg. + S.D. Glass 54.4 55.3 56.1 55.4 + 1.0 PU-Epoxy 91.3 94 93.8 93.0 + 1.5 PU-Epoxy 90.7 89.9 91.8 90.8 + 0.9 Figure 6.9. Contact angle measurement of HEMA on the PU/epoxy coated glass slide.

PAGE 229

204 The hydrophilicity of the PU/epoxy coati ng could be improved by increasing the concentration of Brij 30 in the formulation. Br ij 30 is a polyethylene lauryl ether which is used as a non-ionic solubiliser/dispersant. It improves the wettability of a system. By increasing the concentration of Brij 30 in the PU/Epoxy/THF formulation the resulting coating had a high affinity to HEMA; however, this result was a trade-off to the loss of adhesion of the PU/epoxy coating to the me tal sensor wire. The PU/epoxy coating is a key factor in maintaining the sensitivity a nd working order of the sensor. New coating techniques were investig ated and are presented in the upcoming sections. Figures 6.10 and 6.11 show beading of PHEMA on the sensor. This was a direct result of the low “wettabili ty” of the PU/epoxy surface by the HEMA monomer. Trials of dip-coating the PU-epoxy coated sensor in HEMA, followed by UV polymerization were unsuccessful. Simple adhesion tests show that the PHEMA beads can be easily pulled off once hydrated. Figure 6.10. Dip coating, followed by UV polymerization.

PAGE 230

205 Figure 6.11. Pipetting th e HEMA monomer unto the sensor, followed by UV polymerization.

PAGE 231

206Drawbacks of Gamma Irradiation Grafting To overcome the wetting disadvantage of the first coating technique, gamma ( ) irradiation grafting was employed. Severa l researchers have investigated the irradiation grafting of various monomers unto polymer coatings [Yuan et. al. 2004, Jansen and Ellinghorst 1984, L ee and Hwong 1997]. radiation, like UV radi ation, results in chain scission and recombination events upon radiati on of the PU surface. Pierpoint et. al. and Murphy et. al. the radiation effects of polyuretha ne and found that the polymer undergoes rapid crosslinking as a result of the carbon-centered secondary carboxyl radical that is formed when chain scis sion of the C-O, N-C and C-C bonds occur [Pierpoint et. al. 2001, Murphy and Wettela nd 2005]. The idea of irradiation grafting explained in an earlier section of this chapter is the same. For this experiment, glass slides were coated with the PU/epoxy formulation which was then cured in an oven for one hour at 120 C. Since HEMA monomer has such a high contact angle when in contact with the PU coating, the glass slides were immersed in a vial of HEMA monomer, which was then degassed with nitrogen Irradiation of the vial was then carried out. It was hoped that the preswelling of the PU coating in HEMA would result in the formation of an inte rpenetration network (IPN) and not only the surface grafting of HEMA unto the PU/epoxy surf ace. This would ultimately result in the resolution of the issue regard ing the delamination of PHEM A from the PU coating. Initially, the optimal radiation dosage n eeded to be determined. High radiation dosages, > 0.10Mrads, resulted in the gelati on and crosslinking of the HEMA monomer in vial. To optimize the experiment, the ra diation dosage needed to be high enough to initiate HEMA grafting but low enough to pr event complete gelation and polymerization of all the HEMA monomer present in the vi al. In addition to re ducing the dosage, the HEMA monomer was diluted in both wate r and methanol to reduce the total polymerization of HEMA in the vial. Two concentrations were prepared: 50% HEMA: 50% DI water, 3) 50% HEMA: 50% methanol 4) 10% HEMA: 90% DI water, and 10% HEMA: 90% methanol. These samples were irra diated at 0.10Mrad and 0.04Mrad. Both water and methanol are not free radical sc avenging solvents; therefore their presence

PAGE 232

207 would not inhibit the polymeri zation process [Okamoto et. al. 1999, Serrano Aroca et. al. 2004, Chen et. al. 2002]. It was observed that when the HEMA/wat er solution was irradiated the solution turned cloudy, white. The solution was filtered, the precipitate was washed with water and dried in a vacuum oven. The FTIR sp ectrum on the PHEMA precipitate matches the spectrum of thermally prepared PHEMA. The FTIR spectra in the wavenumber range of 450-4500 cm-1 are shown in figures 6.12 and 6.13. Table 6.4 show the spectral band assignment for PHEMA. This procedure; how ever, resulted in no observable grafting of PHEMA unto the PU/epoxy coating. The HEMA/methanol solution did not have the same result as the HEMA/water solution. A white, PHEMA precipitate did no t form when the HEMA/methanol solution was radiated in the presence of the PU/epoxy coated glass, but FT IR data and contact angle measurements (table 6.5) confir m the grafting of PHEMA unto the PU/epoxy coated glass. The contact angle for water on the surface decreased for the PHEMA-graft samples indicating an in crease in hydrophility. Table 6.4. FTIR spectra band assignme nt for PHEMA [Perova et. al. 1997]. Frequency (cm-1) Possible Assignments 3440 O-H stretching vibration 2950 CH3, CH2, CH antisymmetric and symmetric stretching vibration 1720 C=O stretching vibration 1630 H-O-H bending vibration 1480 (CH2) 1310 CH2 twist and rock 1079 C-O-C vibration stretching

PAGE 233

208 Table 6.5. Contact angle measurements of water on glass, PU/epoxy coated glass surface and HEMA graft on PU/epoxy coated glass surface. Sample Surface Contact Angle Water, Drop 1 Contact Angle Water, Drop 3 Contact Angle Water, Drop 3 Contact Angle Avg. + S.D. Glass 47.6 45.3 46.7 46.5 + 1.2 PU/Epoxy 87.6 88.0 86.9 87.5 + 0.5 PU/Epoxy 89.5 89.6 89.5 89.5 + 0.1 HEMA graft 84.3 84.4 85.8 84.8 + 0.8 HEMA graft 85.9 85.1 83.3 84.8 + 1.3 Even though grafting of PHEMA was observed through irradiation, this technique has a major drawback. When th e PU/epoxy coating is preswollen in the HEMA-solvent mixture it lose s its adhesion to the glass s lide and undergoes dissolution. If left undisturbed during the preswelling and radiation steps, followed by removal from the HEMA-solvent vial, the coating will re-h arden on the glass surface. However, if an attempt is made to wipe the pre-swollen PU/epoxy/HEMA coating off the slide before radiation the coating will come off. This obs ervation renders this te chnique unreliable in terms of reproduction. 419.26 0 466.725 516.420 747.889 848.532 897.273 943.915 1064059 1020.168 1073.567 1155.729 1273.513 1456842 1488979 1636.147 1728.664 2956.595 3438.592 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 % T r a n s m i t t a n c e 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 3200 3400 3600 3800 4000 Wavenumbers (cm-1) Figure 6.12. FTIR spectrum of thermally prepared PHEMA.

PAGE 234

209 431.57 7 593.055 651.684 818.618 900380 947885 1034979 1064.566 1079058 1310.768 194473 1434.850 1447969 1635.789 1713.007 2954933 3442.794 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 % T r a n s m i t t a n c e 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 3200 3400 3600 3800 4000 Wavenumbers (cm-1) Figure 6.13. FTIR spectrum of PHEMA prepared from the irradiated HEMA/water solution. Plasma Polymerization The search to find a suitable, reliable process to deposit a thin film of the hydrophilic polymers unto the PU/epoxy coated glucose sensor brought our attention to the process of plasma polymeri zation. Plasma polymerization is in fact a process and not a new mechanism for polymerization; it relies on the recombination of activated species on a substrate to form a continuous polyme r film [Yasuda 1985]. It has been known for many years that deposits were inherently fo rmed when an electrical discharge was operated in the presence of organic vapors or gases. The recognition that fine tuning of the deposition conditions can be applied to make hard, scra tch resistant C:H films was made in the 1950s by pioneers like Knig and Schmellenmeier [Brockers and Knig 1958, Schmellenmeier 1956]. This technique was chosen for several reasons; the reasons being 1) uniform, smooth, clear thin films can be deposited, 2) no solvents or chemical initiator additives

PAGE 235

210 are necessary, 3) the films will be crosslinke d; thereby preventing dissolution, 4) plasma polymer films have excellent adhesion to both metallic and polymer surfaces, 5) the films have good corrosion resistance, and 6) the subs trate can be three dimensional. Typically, conventional coating processes involve multiple steps of preparation, coating and curing. These steps generally include 1) synthesi s of the monomer, 2) synthesis of the prepolymer or polymer, 3) preparation of the coating solution, 4) cleaning and activation of the substrate surface, 5) a pplication of the coating, 6) dr ying of the coating, and 7) curing of the coating. The adva ntage of plasma polymerization is that a polymer film can be deposited on a substrate in one or two steps without the use of organic solvents [Yasuda 1985]. Plasma polymer films have f ound its niche in many fields as potential electronic, optical, protective and biomedi cal materials [Suwa et. al. 1996, Shi 1996, Arefl et. al. 1992]. Plasma is a state of matter that is made up of partially ionized gas. The partially ionized gas is a mixture of free radicals, pos itively and negatively charged ions, neutral species, electrons and UV photons [Yasuda and Yu 2004, Yasuda 1985, March Plasma 2006]. A plasma can be generated using an energy source such as combustion, flames, electric discharge, controlled nuclear reac tions and shocks [Yasuda 1985]. The plasma generated by electric discharge, either DC or RF, is often termed “cold”, or low temperature plasma and deposition on th e substrate typically occurs at room temperature. An electric discharge is th e most common source used to maintain a continuous plasma state over a long period of time. When gas molecules pass between two activated electrode plates in a vacuum, three events occur. These events are 1) ionization, 2) excitation, a nd 3) elastic collisions which result in no change. The excitation and fragmentation of the original molecules that is present in plasma is short lived as the activated species quickly recombine once the electric discharge is deactivated. The deposition rate of plasma films is dependent on several variables which include reactivity of starting ma terial, monomer flow rate, sy stem pressure, geometry of the plasma system, discharge power, frequenc y of the excitation si gnal, and temperature of the system [Huber and Springer 1996]. Since the plasma has highly fragmented

PAGE 236

211 species, the recombination event will produce plasma polymer films that may not contain regularly repeating units as s een in conventionally prepared polymer films; in fact, it is possible for the plasma polymer to not resemble at all the polymer formed from the same monomer under conventional means. Plasma fi lms are characteristically branched, pinhole free and randomly terminated with a high degree of crosslinking [Jeon et. al. 2004, Yasuda 1985]. The plasma system used for this expe riment was provided courtesy of March Plasma Systems, a leading company in ma nufacturing plasma system. The electric discharge plasma system used consisted of a radio frequency (RF) power supply, two parallel plate electrodes, a reaction chamber, an inlet and outlet for the gas, an inlet for the monomer, and a vacuum system. The upper electrode plate served as the excitation electrode and the lower plat e was used to ground the system. Figure 6.14 show a schematic of a RF plasma system. The gas used for the process can vary depending on the required properties of the plasma f ilm. Argon, neon, oxygen and nitrogen are nonplasma forming gases, and are ideal carrier gases to use in experiments where only the monomer will be converted into the plasma polymer film. Table 6.6 gives a plasma process overview with respect to process gas. Figure 6.14. A schematic of a RF electric discharge plasma system.

PAGE 237

212 Table 6.6. Plasma processes overview. [Courtesy of March Plasma Systems, www.marchplasma.com ] Surface Modification Process Process Gas Material Types Post Plasma Application Contamination Removal Argon (Ar) Oxygen (O2) Hydrogen (H2) Stainless Steel Aluminium Polymers 1. Ultra cleaning 2. Material removal for improved adhesion Crosslinking Argon (Ar) Polymers 1. Makes surface impermeable 2. Polymer-metal adhesion Surface Activation Nitrogen (N2) Oxygen (O2) Hydrogen (H2) Helium (He) Ammonia(NH3) Polymers Teflon Silicone 1. Bonding 2. Permeability 3. Friction 4. Wettability Etch Oxygen and Carbon Tetrafluoride (CF4) Epoxy Polyimide Silicon Silicon Dioxide 1. Wafer level applications Deposition Coating Vaporized Monomer Polymers Metal, Glass 1. Wettability 2. Biomedical The properties of the conventiona lly prepared PHEMA and PDHPMA homopolymers and copolymers have been di scussed earlier in th is chapter and in previous chapters. It has been shown that various copolymer fo rmulations exhibited excellent biocompatibility; the hydrogels also ha ve the network struct ure that facilitates the diffusion of glucose and oxygen which is necessary for the operation of the glucose sensor. Finding an appropriate coating t echnique has been a trying task, and recent developments in plasma polymerization show promise. From the above description of plasma polymerization, one may think that the technique would be inapplicable as it could produce a plasma film that may potentially bear no resemblance to the conventional PH EMA hydrogel. The plasma film would be highly crosslinked and impermeable; it would therefore lack the diffusion transport properties necessary for the operation of the sensor. All these statements are true; however, it is possible to produce a plasma film that resembles, or behaves similar to, the

PAGE 238

213 conventional polymer. The added benefit would be the ease and uniformity of the film application, the improved adhesion of the polym er to the substrate, and the avoidance of wet chemistry involving organic solvents. This task can be achieved by controlling the deposition conditions of the experiment, es pecially the energy flux (RF power). By reducing the energy flux, the monomer will undergo less frag mentation and the plasma polymer will retain the mo lecular structure of th e monomer [Yasuda 1985]. Two research groups have been able to produce PHEMA plasma films that possess physical and chemical properties si milar to those of conventional PHEMA. Tarducci and coworkers investigated varyi ng the RF flux by comparing the results from using low-power continuous wave plasma and low-power pulsed cycle plasma. Their system used a high frequency 13.56 MHz RF power supply using a power of 3W for continuous wave plasma and 40W for pulsed pl asma. The carrier gas employed was air. It was concluded that even t hough both protocols resulted in the deposition of plasma PHEMA, the pulsed RF plasma procedure re sulted in a higher degree of structural retention as measured using FTIR, XPS and NMR [Tarducci et. al. 2002]. Bodas and coworkers used a 13.56 MHz RF power supply at 75W for a deposition time of 10 and 40 minutes of continuous wave plasma. Th e carrier gas employed was argon and the monomer was vaporized over a temperature range of 50-75 C. FTIR and XPS data confirmed plasma PHEMA having an identi cal chemical composition to conventional PHEMA. The plasma film deposited for 10 minutes had an approximate thickness of 80nm and the one deposited for 40 minutes had a thickness of 200nm [Bodas et. al. 2005]. The system utilized in this project ha s a 40kHz RF power supply; therefore, a series of trial and error expe rimental conditions were tested Since the power supply is at a much lower frequency signal than the ones us ed by Tarducci et. al. and Bodas et. al. a higher power output was necessary to generate a plasma film. Table 6.7 show the various conditions and protocols, both successful and unsuccessful, used to plasma polymerize a thin film of PHEMA.

PAGE 239

214 At the beginning of each protocol the vacuum system was allowed to stabilize at the set argon flow rate (~20 minutes) without RF glow disc harge, this time is not included in the protoc ol descriptions.

PAGE 240

215 Table 6.7. Various conditions and protocols, both successful and unsuccessful, used to plasma polymerize a thin film of PHEMA. Sample Argon Flow Rate (L/min) Temperature Range (C) RF Power (W) Protocol 1 0.1 45-65 300 RF on. Monomer allowed into the system for 20 mins w/RF. RF off. 2 0.1 45-65 500 RF on. Monomer allowed into the system for 20 mins w/RF. RF off. 3 0.1 45-65 500 RF on. Monomer allowed into the system for 20 mins w/RF. RF off. Monomer flow continued for 10 mins. 4 0.15 50-75 600 Ar flow w/RF for 20 mins. Monomer allowed into the system for 40 mins w/RF. RF off. Monomer flow continued for 20 mins. 5 0.15 50-75 650 Ar flow w/RF for 20 mins. Monomer allowed into the system for 40 mins w/RF. RF off. Monomer flow continued for 20 mins. 6 0.15 50-75 800 Ar flow w/RF for 20 mins. Monomer allowed into the system for 40 mins w/RF. RF off. Monomer flow continued for 20 mins. 7 0.15 50-75 800 Ar flow w/RF for 20 mins. Monomer allowed into the system for 60 mins w/RF. RF off. Monomer flow continued for 20 mins.

PAGE 241

216 Table 6.8. Results of plasma polymerizati on of PHEMA thin films using the various protocols and experimental conditions. Sample Results 1 through 3 No visible plasma film. No detectable IR spectra. 4 through 5 Visible, yellow tinted clear film. No detectable IR spectra 6 and 7 Visible film. Positive IR spectra By varying the deposition conditions and RF power a plasma film was generated for samples 4 through 7. The FTIR spectra for samples 4 and 5 we re undetectable. The FTIR spectrum for sample 6 showed peaks a ssociated with those of PHEMA; however, the heights of the peaks rather small (fig. 6.15). By increasing the deposition time for sample 7, a positive IR spectra with better defined peaks was obtained (fig. 6.16). The plasma films from samples 6 and 7 adhered quite well to the glass slides even after continuously washing with wa ter and methanol. The FTIR spectra before and after rinsing the films were the same. AFM images obtained for sample 7 show that the plasma film is approximately 25nm thick (figs 6.17-6.18).

PAGE 242

217 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 % T r a n s m i t t a n c e 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 3200 3400 3600 3800 4000 Wavenumbers (cm-1) Figure 6.15. FTIR spectra of 1) conventi onal PHEMA (red) and 2) plasma PHEMA, sample 6 (blue).

PAGE 243

218 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 % T r a n s m i t t a n c e 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 3200 3400 3600 3800 4000 Wavenumbers (cm-1) Figure 6.16. FTIR spectra of 1) conventi onal PHEMA (red) and 2) plasma PHEMA, sample 7 (blue). Figure 6.17. AFM i mage of plasma polymer film (sample 7).

PAGE 244

219 Figure 6.18. AFM i mage and film thickness se ction analysis of plasma PHEMA film (sample 7).

PAGE 245

220Conclusion Hydrogels are materials that can sorb a considerable amount of water without dissolving. Natural hydrogel materials include cr osslinked gelatin and starch agar gel, but hydrogels can also be synthethic. Synthetic hydrogels are slightly crosslinked hydrophilic polymers that are characterized by solubi lizing pendant groups (e.g., -OH, -COOH, CONH2) incorporated into the hyd rogel structure. Some hydrogels have been found to be biocompatible. Hydrogels have been used as ma terials in contact lenses and drug delivery capsules; other medical applications include dermal wound healing, and implantation in the body of a human or animal patient to im prove the interfacial tissue interaction of medical implants. The biocompatibility of hydrogels can be attributed to the low interfacial tension with biological fluids, high gas permeability, high diffusion of low molecular weight compounds, and reduced m echanical and fricti onal irritation to surrounding tissue. This project investigated the bioc ompatibility of pol y(2-hydroxyethyl methacrylate) (PHEMA) and poly(2,3-di hydroxypropyl methacrylate) (PDHPMA) as homoand copolymers. Its application to an implantable glucose sensor is highly desirable because of its excellent biocompatib ility and diffusion transport properties. The objectives of the hydrogel coating were:1) it should be permeable to allow glucose, oxygen and hydrogen peroxide to di ffuse freely; 2) it should reduce adsorption of protein from surrounding cell and plasma; and 3) its use should result in minimal fibrosis by having an interface that is co mpatible with the tissue. It was found that PDHPMA had a water equi librium content almost triple that of PHEMA, which is attributed to the add itional hydroxyl group on the pendant moiety. Unfortunately, as it sorbed this much water the mechanical stability of the high content DHPMA copolymers and the PDHPMA homopo lymer was lost, and the samples were easily fragmented. Improved biocompatibility an d mechanical properties were seen in the 80%HEMA:20%DHPMA, and the 60%HEMA:40%DHPMA copolymers. These copolymer hydrogels were found to induce minimal to no fibrosis when implanted subcutaneously in rats.

PAGE 246

221 Once the biocompatibility of the hydrogels wa s established, the task to coat the polyurethane (PU)/epoxy coated metal sensor ne eded to be addressed. The wettability of the HEMA monomer to the PU/epoxy coati ng was found to be minimal using contact angle measurements. As a result, technique s involving dip-coating, or in situ polymerization, were not adequate as they produced non-uniform coatings on the sensor. It was also noted that the PHEMA coating easily delaminated from the PU/epoxy coating once swollen in water. Therefore, it was n ecessary to employ a technique that would not only produce a uniform, smooth hydrogel coatin g, but one where the hydrogel coating would be bound to the PU/epoxy coating to prevent loss of adhesion. Two polymerization processes were then investigated: irradiation grafting and plasma polymerization. The irradiation grafting was ruled out as a viable technique since the monomer/solvent system resulted in dissolution of the PU/epoxy coating. Plasma polymerization is a technique that is usually used to produce highly crosslinked, barrier coatings. However, it is possible to produce a plasma film that resembles, or behaves similar to, the conven tional polymer (in this case PHEMA). The added benefit of this process was the ease and uniformity of the film application, the improved adhesion of the polymer to the substrate, and the avoidance of wet chemistry involving organic solvents. This task was achieved by cont rolling the depositi on conditions of the experiment, especially the energy flux (RF power). By reducing the energy flux, the monomer underwent less fragmentation and th e plasma polymer retained the molecular structure of the monomer. FTIR data show ed that the plasma film maintained the functionality of conventional PHEMA. Further work still needs to be carried out to determ ine the physical and thermal properties of the plasma film; this will be discussed in greater detail in chapter 7. At this point, plasma polymerization appears to be a fe asible technique for the application of the biocompatible hydrogel materials for use as a coating on the implanta ble glucose sensor.

PAGE 247

222CHAPTER 7 Conclusions and Future Work Dielectric spectroscopy is an excellent th ermal analysis technique that has found its niche in more than one field. By definition, dielectric analysis is the probing of interactions in a material using a time-depe ndent electric field. The resulting polarization in the material occurs from the reorientat ion of permanent and induced dipole moments in the material; other events include translational movement of ions and interfacial Maxwell-Sillars charge buildup in heterogeneous systems. A dielectric spectrum can be recorded over a large range of frequencies from millito teraHertz; to achieve this several instruments would be required to cover this range. Diel ectric spectroscopy has been commonly used to analyze the molecular relaxations in polymers, the cure kinetics of polyurethanes and dielectric lo ss of materials. It is impor tant to know these properties so that the developer can determine if a material is a high loss material which makes it ideal for shielding and anechoic applications, or if it is a low loss material which makes it ideal for waveguide, insulating, antenna, and device interconnect applications. Recently, scientists have been using dielectric analys is for new applications, such as label-free cellular analysis, drug adsorption and releas e in polymer matrices (transdermal and implantable applications) and for monito ring water and other analyte content in agricultural grains and soil [Ciambron et.al. 2004, Li et. al. 2004, Hgerstrm et. al. 2005, Nelson et. al. 2004]. The dielectric response of a material must be accurately measured and understood in order for the material to be skillfully util ized in a given application. The thermal properties of poly (2-hydr oxyethyl methacrylate) (PHEMA) and poly (2,3-dihydroxypropyl methacrylate) (PDHPMA) have been presented. Both of these materials sorb water to form hydrogels, and ha ve found a role in biomedical applications for such materials as contact lenses, bioadhesi ve gels for drug delivery and as thromboand fibroresistant coatings for implantable sensors. In chapters 3 and 5, interpretations of the dielectric spectra of PHEMA, PDHPMA and their random copolymers have been pres ented using the electric modulus formalism

PAGE 248

223 and various mathematical formulations to ch aracterize the viscoelastic processes and the conductivity relaxation present in the homopolymers and in the random copolymers of HEMA and DHPMA. Neat PHEMA and PDHPMA exhibit two subTg secondary relaxations and a primary glass transition ( Tg). The transitions are termed , and proceeding from the high temperature transition to the low temperature transition. The primary glass transition marks the onset of large scale segmental motion of the main chain, or polymer backbone, the relaxation corresponds to the rotation of the ester side group and the relaxation is associated with th e rotation of the hydroxyethyl group. Previous studies by various rese archers presented the dielectric spectra of these materials but did not report the dielectri c properties at and above th e glass transition region (100 C). Three different processes were observed in this dissertation st udy taking place at ca. 50 C and above, and due to the paucity of di electric data in literature covering this temperature range an attempt was made to de cipher the meaning of the dielectric spectra of neat PHEMA, PDHPMA and random copolymers of HEMA and DHPMA. It was found that the relaxation in PHEMA is fast moving and at higher temperatures and frequencies it tended to merge with the transition resulting in the merge. As the temperature and frequency in creased further, ioni c conductivity effects became predominant and a loss peak was observed. Using various mathematical proofs it was shown that this peak did not exhibit a ny visco-elastic properties, but followed the Debye model for molecules that exhibit a single relaxation time. Literature states that conductivity relaxations in ioni c conductors exhibit single relaxation times [Ambrus, Moynihan and Macedo 1972, Johari and Pathmanathan 1988, Macedo, Moynihan and Bose 1972]. Using a similar approach employed in chapter 3 the dielectric properties of DHPMA was determined. To the best of the au thors knowledge this was the first study to investigate the dielectric properties of poly(HEMA-co-DHPMA) copolymers up to and above the glass transition region. It was obs erved that the glass transition temperature decreased, the activation energy increased, the activation energy decreased and ionic conductivity increased with DHPMA content. Overall, it was noted as DHPMA content

PAGE 249

224 increased conductivity effects became more pronounced as it became difficult to resolve the and relaxations, and that DHPMA facilitat es ionic movement through its matrix more efficiently than in HEMA. This study is important because diel ectric behavior gives insight into the structural property and relaxations pr esent in the polymer, this information can then be used to determine the materials end use. Understanding the dielectric spectra of PHEMA, enabled the investigation of the interaction of the polymer with a nanofiller; this aspect was examined in chapter 4. A novel self-assembled hydroxylated nanoparticle, [(DMSO)(MeOH)Cu2(benzene-1,3dicarboxylate-5-OH)2]12, has the potential to be used in a variety of el ectromagnetic and drug delivery applications. In this study, th e effects of the inte ractions taking place between the self-assembled nanostructure with two functionally different polymers: poly(2-hydroxyethyl methacrylate) (PHEMA) and poly(methyl methacrylate) (PMMA) were examined. The PHEMA-nanoball nanocomposites endur ed in a hostile swelling and extraction environment. It is well known that most physical crosslinks in polymers are labile to dissolution in the pr oper solvent environment, so it is significant that these selfassembled suprastructures persisted. The da ta showed that the crosslinking density increased in the PHEMA nanocomposites. This observation suggests that there is an interaction taking place betw een the nanoball and HEMA. Fu rther evidence gained by DSC and DEA data support this phenomenon as the glass transition temperature and the ionic conductivity ac tivation energy increased with nanoba ll concentration. It is believed that this interaction may be the result of phys ical threading of PHEM A chains through the nanoball windows, in which the HEMA mono mer may be drawn by H bonding to the internal ligands in the nanoball. The possibility of a number of different schemes exists but in order to be more conclusive inve stigations should be carried out by further characterizing the interaction using linear PHEMA and othe r polymer systems with the nanoball. By contrast, data derived for the PMMA nanocomposites indicate that there is minimal interaction between the nanoball and the matrix where the PMMA nanocomposites consistently show the opposite effect. There is an increase in the ionic

PAGE 250

225 conductivity and a decrease in the ionic conductiv ity activation energy as the nanoball concentration is increased. This phenomenon is due to the lack of immobilization of the polymer matrix which consequently enhances the rotational movement of the side chain moiety and the translational diffusion of ions in the matrix. Further DSC and microhardness data verify the plasticization effect of the PMMA matrix. This study is useful as it gives insight into the interac tions taking place between these supramolecular nanoparticles with various polymer matrices; further understanding can be gained in the future by investigating the in teractions between functiona lly different “nanoballs” and polymer systems. PHEMA and PDHPMA hydrogels have been used to formulate a biocompatible coating for an implantable glucose sensor (C hapter 6). Hydrogels are materials that can sorb a considerable amount of water w ithout dissolving. Hydrogels are slightly crosslinked hydrophilic polymers that are ch aracterized by solubi lizing pendant groups (e.g., -OH, -COOH, -CONH2) incorporated into the hydroge l structure. Some hydrogels have been found to be biocompatible. The bioc ompatibility of hydrogels can be attributed to the low interfacial tension with biologi cal fluids, high gas permeability, high diffusion of low molecular weight compounds, and reduced mechanical and fric tional irritation to surrounding tissue. This project investigated the bioc ompatibility of pol y(2-hydroxyethyl methacrylate) (PHEMA) and poly(2,3-di hydroxypropyl methacrylate) (PDHPMA) as homoand copolymers. Its application to an implantable glucose sensor is highly desirable because of its excellent biocompatib ility and diffusion transport properties. The objectives of the hydrogel coating were:1) it should be permeable to allow glucose, oxygen and hydrogen peroxide to di ffuse freely; 2) it should reduce adsorption of protein from surrounding cell and plasma; and 3) its use should result in minimal fibrosis by having an interface that is co mpatible with the tissue. It was found that PDHPMA had a water equi librium content almost triple that of PHEMA, which is attributed to the add itional hydroxyl group on the pendant moiety. Unfortunately, as it sorbed this much water the mechanical stability of the high content DHPMA copolymers and the PDHPMA homopo lymer was lost, and the samples were

PAGE 251

226 easily fragmented. Improved biocompatibility an d mechanical properties were seen in the 80%HEMA:20%DHPMA, and the 60%HEMA:40%DHPMA copolymers. These copolymer hydrogels were found to induce minimal to no fibrosis when implanted subcutaneously in rats. Once the biocompatibility of the hydrogels wa s established, the task to coat the polyurethane (PU)/epoxy coated metal sensor ne eded to be addressed. The wettability of the HEMA monomer to the PU/epoxy coati ng was found to be minimal using contact angle measurements. As a result, technique s involving dip-coating, or in situ polymerization, were not adequate as they produced non-uniform coatings on the sensor. It was also noted that the PHEMA coating easily delaminated from the PU/epoxy coating once swollen in water. Therefore, it was n ecessary to employ a technique that would not only produce a uniform, smooth hydrogel coatin g, but one where the hydrogel coating would be bound to the PU/epoxy coating to prevent loss of adhesion. Two polymerization processes were then investigated: irradiation grafting and plasma polymerization. The irradiation grafting was ruled out as a viable technique since the monomer/solvent system resulted in dissolution of the PU/epoxy coating. Plasma polymerization is a technique that is usually used to produce highly crosslinked, barrier coatings. However, it is possible to produce a plasma film that resembles, or behaves similar to, the conven tional polymer (in this case PHEMA). The added benefit of this process was the ease and uniformity of the film application, the improved adhesion of the polymer to the substrate, and the avoidance of wet chemistry involving organic solvents. This task was achieved by cont rolling the depositi on conditions of the experiment, especially the energy flux (RF power). By reducing the energy flux, the monomer underwent less fragmentation and th e plasma polymer retained the molecular structure of the monomer. FTIR data show ed that the plasma film maintained the functionality of conventional PHEMA. Further work still needs to be carried out to determ ine the physical and thermal properties of the plasma film. The dielectric spectrum of the PHEMA plasma film can be compared to that of the c onventional polymer. The activati on energies of the secondary and primary molecular relaxations and of the i onic conductivity of the matrix can be used

PAGE 252

227 to determine if the plasma film possesses a different network stru cture the conventional polymer. At this point, plasma polymerization appears to be a feasible technique for the application of the biocompa tible hydrogel materials for use as a coating on the implantable glucose sensor.

PAGE 253

228REFERENCES Abourahma, H., Coleman, A.W., Moulton, B., Rather, B., Shahgaldian, P., and M.J. Zaworotko. 2001. Chemical Communications 22: 2380-2381. Agilent Technologies 2000, Agilent 8453 UV-Visible Spectroscopy System: Service Manual. Germany. Ake Oeberg, P. 2004. S ensor Applications, Sensors in Medicine and Health Care Vol.3, New York: Wiley, John and Sons. Ali, S.A.M., Zhong, S.P., Doherty, P.J., a nd D.F. Williams. 1993. Biomaterials 14(9): 648-656. Ambrus, J.H., Moynihan, C.T., and P.B. Macedo. 1972. Journal of Physical Chemistry 76(22): 3287-3295. Anderson, S., Aplin, R.T., Claridge, T.D.W., Goodson T., Maciel, A.C., Rumbles, G., Ryan, J.F., and H.L. Anderson. 1998. Journal of the Chemical Society, Perkin Transactions 1 : 2383-2396. Arefl, F., Andre, V., Montazer-R ahmati, P., and J. Amouroux. 1992. Pure and Applied Chemistry 64(5): 715-723. Avakian, P., Starkweather, Jr. H.W., and W.G. Kampert. 2002. Dielectric analysis of polymers In: Cheng SZD, editor. Handbook of thermal analysis and calorimetry, 3, pp. 147-164, New York: Elsevier. Balt Calleja, F.J., and S. Fakirov, 2000. Microhardness of Polymers New York: Cambridge University Press. Benz PH, Ors JA. 2000. Contact lens having impro ved dimensional stability United States Patent. US6011081. Benz PH, Ors JA. 2000. Contact lens of high water content and high water balance United States Patent. US6096799. Benz PH, Ors JA. 1999. Terpolymer for contact lens United States Patent. US5891932. Bergman, R., Alvarez, F., Alegria, A., and J. Colmenero. 1998. Journal of Chemistry and Physics 109(17): 7546-7555. Bershtein, V.A., and V.M. Egorov. 1994. Differential Scanning Calorimetry of Polymers: Physics, Chemistry, Analysis, Technology New York: Ellis Horwood.

PAGE 254

229 Bieber, J.A. 2004. Synthesis of nanostructures in si ngle crystal silicon carbide by electron beam lithography. University of South Florida: Masters Thesis. Bodas, D.S., Desai, S.M., and S.A. Gangal. 2005. Applied Surface Science 245: 186-190. Bottcher, H. 2000. Journal fuer Praktische Chemie 342(5): 427-436. Bozzola, J.J., and L.D. Russell. 1992. Electron Microscopy: Principles and Techniques for Biologists Boston: Jones and Bartlett Publishers. Bradley, G., Ed. 1998. Photoinitiators for Free Radical Cationic and Anionic Photopolymerization 2nd ed. New York: John Wiley and Sons Publishers. Brockers, A., and H. Knig. 1958. Zeitschrift fuer Physik 152: 75-86. Carotenuto, G., and L. Nicolais. 2003. Polymer News 28: 174-182. Cattaruzza, E., Gonella, F., Ma ttei, G., Mazzoldi, P., Gatteschi, D., Sangregorio, C., Falconieri, M., Salvetti, G ., and G. Battaglin. 1998. Applied Physics Letters 73(9): 1176-1178. Ciambrone, G.J., Liu, V.F., Lin, D.C., McGui nness, R.P., Leung, G.K., and S. Pitchford. 2004. Journal of Biomolecular Screening 9(6): 467-480. Chabert, E., Bornert, M., Bourgeat-Lami, E., Cavaill, J.-Y., Dendiev al, R., Gauthier, C., Putaux, J.L., and A. Zaoui. 2004. Materials Science and Engineering A-Struct. 381: 320-330. Chambron, J-C., Heitz, V., and J-P. Sauvage. 1995. Bulletin de la Societe Chimique de France 132: 340-347. Chandler, H. (Ed.). 1999. Hardness Testing 2nd, ed. Materials Park (Ohio): ASM International. Chen, G-H., Wu, D-J., Weng, W-G ., He, B., and W-L. Yan. 2001. Polymer International 50: 980-985. Chen, H.H., Zhuang, R.C., Lin, J., Ye, J.L., Dai, L.Z., and Y.S. Zou. 2002. Gaofenzi Cailiao Kexue Yu Gongcheng 18(3): 53-56. Clayton, L.M., Sikder, A.K ., Kumar, A., Cinke, M., Meyyappan, M., Gerasimov, T.G., and J.P. Harmon. 2005. Advanced Functional Materials 15: 101-106.

PAGE 255

230 Clough, R.L., and S.W. Shalaby. 1996. Irradiation of Polymers Washington, DC: American Chemical Society Craig, D.Q.M., and S. Tamburic. 1997. European Journal of Pharmaceutical Biopharmacy 44: 61-70. Dad, S., Hodge, P., and S.D. Kamau. 2003. Reactive and Functional Polymers 54: 131140. Dahmouche, K., Santilli, C.V ., and S.H. Pulcinelli. 1999. Journal of Physical Chemistry B 103: 4937-4942. Decker, C. 1998. Polymer International 45: 133-141. Dee, K.C., Puleo, D.A., and R. Bizios. 2003. An Introduction to Tissue-BioMaterial Interactions New Jersey: John Wiley & Sons, Inc. de Lange, P.J., Gebben, B., and P.F. Elfrink. 1994. Surface and Interface Analysis 22 (12): 502-506. Diaz Calleja, R. 1979. Journal of Polymer Science, Part B: Polymer Physics 17: 13951401. Emran, S.K., Newkome, G.R., Weis, C.D., and J.P. Harmon. 1999. Journal of Polymer Science, Part B: Polymer Physics 37: 2025-2038. Gates, G. 2001. Material Properties of 2,3-Dihydroxypropyl Methacrylate and 2Hydroxyethyl Methacrylate H ydrogels. University of South Florida: Doctoral Dissertation. Gates, G., Harmon, J.P., Ors, J., and P. Benz. 2003. Polymer 44(1): 207-214. Gates, G., Harmon, J.P., Ors, J., and P. Benz. 2003. Polymer 44(1): 214-222. Gedde, U.W. 1995. Polymer Physics New York: Chapman & Hall. Gerasimov, T.G. 2002. Vibrational Over tone Spectroscopy of Several Monoand Polyacrylates, Azo-initiators and Poly siloxanes, Vibrationally Initiated Polymerization of Methyl Methacrylate a nd Methyl Methacrylat e – Polysiloxane Mixtures. Bowling Green State Univ ersity: Doctoral Dissertation. Gersten, J.L. and F.W. Smith. 2001. The Phys ics and Chemistry of Materials. New York: John Wiley & Sons, Inc.

PAGE 256

231 Gibson, H.W., Bryant, W.S., and S-H. Lee. 2001. Journal of Polymer Science Part A: Polymer Chemistry 39: 1978-1993. Gomez Ribelles, J.L., and R. Diaz Calleja. 1984. Journal of Macromolecular SciencePhysics B23(2): 255-269. Gomez Ribelles, J.L., and R. Diaz Calleja. 1985 Journal of Polymer Science, Part B: Polymer Physics 23: 1297-1307. Gomez Ribelles, J., Monleon Pradas, M., Galle go Ferrer, G., Peidro Torres, N., Perez Gimenez, V., Pissis, P, and A. Kyritsis 1999. Journal of Polymer Science : Part B : Polymer Physics 37 : 1587-1599. Hgerstrm, H., Strmme, M., and K. Edsman. 2005. Journal of Pharmaceutical Sciences 94(5): 1090-1100. Hartwig, G. 1994. Polymer properties at room and cryogenic temperatures New York: Plenum Press. Havriliak Jr. S., and J.S. Havriliak. 1997. Dielectric and Mechani cal Relaxation in Materials New York: Hanser Publishers. Hanh, S.K., Jelacic, S., Maier, R.V., Stayton, P.S., and A.S. Hoffman. 2004. Journal of Biomaterials Science, Polymer Edition 15(9): 1111-1119. Haines, C. 2005. The Diabetes Toolbox http://www.webmd.com/content/article/46/1667_50934.htm Heller, A. 1999. Annual Review of Biomedical Engineering 1: 153-175. Helfer, C.A., Xu, G., Mattice, W.L., and C. Pugh. 2003. Macromolecules 36: 1007110078. Hench, L.L. and E.C. Ethridge. 1982. Biomaterials: An Interfacial Approach New York: Academic Press. Henn, F., Devautour, S., Maati, L., Giuntini, J.C., Schaefer, H., Zanchetta, J.V., and J. Vanderschueren. 2000. Solid State Ionics 136-137: 1335-1343. Hickey, T., Kreutzer, D., Burgess, D.J. and F. Moussy. 2002. Journal of Biomedical Materials Research 61(2): 180-187. Huber, F., and J. Springer. 1996. Journal of Applied Polymer Science 62: 145-151. Ilmain, F., Tanaka, T., and E. Kokufuta. 1991. Nature 349: 400 – 401.

PAGE 257

232 Janacek, J. 1973. Journal of Macromolecular Science-Review M 9(1): 1-47. Janik, I., and J.M. Rosiak. 2002. Radiation Physics and Chemistry 63(3-6): 529-532. Jansen, B., and G. Ellinghorst. 1979. Journal of Polymer Science: Polymer Symposium 66: 465-473. Jansen, B., and G. Ellinghorst.1984. Journal of Biomedical Materials Research 18: 655669. Jarvie, A.W.P., Lloyd, M.C., and C.B. St. Pourcain. 1998. Biomaterials 19: 1957-1961. Jeffery, K. 2001. Machines in Our Hearts: The Cardiac Pacemaker, the Implantable Defibrillator and American Health Care Maryland: John Hopkins University Press. Jeon, H.S., Wyatt, J., Harper-Nixon, D, and D.H. Weinkauf. 2004. Journal of Polymer Science: Part B: Polymer Physics 42: 2522-2530. Johari, G.P. 1991. Journal of Molecular Structure 250: 351-384. Johari, G.P., and K. Pathmanathan. 1988. Phys Chem Glasses 29(6): 219-224. Karpman, M. G., Kovalev, A. I., Kostina, M. V., Chervyakov, A. V., and Y.D. Yagodkin. 2001. Metal Science and Heat Treatment 43(11-12): 464-465. Kejlova, K., Labsky, J., Jirova, D., and H. Bendova. 2005. Toxicology In Vitro 19: 957962. Kerner, W. 2001. Experimental and Clinical Endocrinology & Diabetes 109(2): S341S346. Kilbinger, AFM., and R.H. Grubbs. 2002. Angewandte Chemie International Edition 41: 1563-1566. Kim, H., and M.W. Urban. 2000. Langmuir 16: 5382-5390. Kolarik, J. 1982. Secondary relaxations in glassy polymers In: Cantow, H.J. et al., editors. Advances in Polymer Science, Vol. 46 pp. 119-161. New York: SpringerVerlag. Kulkarni, G. U., John Thomas, P., and C. N. R. Rao. 2002. Pure and Applied Chemistry 74(9): 1581–1591.

PAGE 258

233 Kusy, R.P. 1986. Metal-Filled Polymers: Properties and Applications In: Bhattacharya, S. K., Ed. Plastics Engineering, Vol. 11, New York: Marcel Dekker. Kyritsis, A., Pissis, P., Gomez Ribelles, J.L., and M. Monleon Pradas. 1994. Journal of Non-Crystalline Solids 172-174: 1041-1046. Lai, Y.C., and L.J. Buccei. 1991. Journal of Applied Polymer Science 42: 3173-3179. LaPorte, R.J. 1997. Hydrophilic polymer coatings for medical devices Pennsylvania: Technomic. Lee, W.F., and G.Y. Hwong. 1997. European Polymer Journal 33(9): 1499-1504. Leica. 1999. Leica VMHT MOT Operati ng Instructions. Austria: Leica. Li, Y.H., Song, C., and K.S. Zhao. 2004. Huaxue Xuebao 62(16): 1495-1502. Linke, B., Kiwit, M., Thomas, K., Krahwinkel, M., and W. Kerner. 1999. Clinical Chemistry 45(2): 283-285. Lugscheider, E., Weber, T., Kne pper, M., and F. Vizethum. 1991. Materials Science & Engineering, A: Structural Materi als: Properties, Microstructure and Processing A 139: 45-48. Luterov, K., Cazzanelli, M., Likforman, J.-P., Navarro, D., Valenta, J., Ostatnick, T., Dohnalov, K., Cheylan, S., Gilliot, P., H nerlage, B., Pavesi, L., and I. Pelant. 2005. Optical Materials 27(5): 750-755. Macedo, P.B., Moynihan, C.T., and R. Bose. 1972. Phys Chem Glasses 13(6): 171-179. Mansur, H.S., Orefice, R.L., and A. A.P. Mansur. 2004. Polymer 45: 7193-7202. March Plasma. 2006. www.marchplasma.com McCrum, N.G., Read, B.E ., and G. Williams. 1967. Anelastic and dielectric effects in polymeric solids. New York: Dover. Meakin, J.R., Hukins, D.W.L., Imrie, C.T, and R.M. Aspden. 2003. Journal of Material Science-Materials M 14: 9-15. Michels, J.J., OConnell, M.J., Taylor, P.N., Wilson, J.S., Cacialli, F., and H.L. Anderson. 2003. ChemistryA European Journal 9: 6167-6176. Micromem Analytical. 2006. Website access: http://www.micromemanalytical.com/ATRKen/ATR.htm

PAGE 259

234 Mohomed, K., Abourahma, H., Zaworo tko, M.J., and J.P. Harmon. 2005. Chemical Communications 26: 3277-3279. Mohomed, K., Gerasimov, T.G., Abourahma, H., Zaworotko, M.J. and J.P. Harmon. 2005. Materials Science and Engineering A 409 (1-2): 228-234. Mohomed, K., Gerasimov, T.G., Moussy, F.,and J.P. Harmon. 2005. Polymer 46(11): 3847-3855. Mohsen, N.M., Craig, R.G., and F.E. Filisko. 2000. Journal of Oral Rehabilitation 27: 250-268. Molyneux, P. 1991. Role of Labile Crosslinks in the Behavior of Water-Soluble Polymers In: Shalaby, S.W., McCormick, C.L., Butler, G.B. (Eds.). WaterSoluble Polymers: Synthesis, Soluti on Properties and Applications, ACS Symposium Series 467, Washington DC: American Chemical Society. Moulton, B., Lu, J., Hajndl, R., Hari haran, S., and M.J. Zaworotko. 2002. Angewandte Chemie International Edition 41(15): 2821-2824. Nam, K., Watanabe, J., and K. Ishihara. 2004. European Journal of Pharmaceutical Sciences 23: 261-270. Nelson, S.O., and S. Trabelsi. 2004. Journa l of Microwave Power and Electromagnetic Energy 39(2): 107-117. Nicolais, L., Acierno, D., Stol, M., and J. Janacek. 1974. Journal of Polymer Science, Part B: Polymer Physics 12: 2579-2581. Okamoto, Y., Nakano, T., Yamada, K., and Y. Isobe. 1999. Polymeric Materials Science and Engineering 80-83. O’Rourke Muisener, P.A., Clayton, L., D’Angelo, J., and J.P. Harmon. 2002. Journal of Materials Research 17(10): 2507-2513. Park, C., Ounaies, Z., Watson, K.A., Crooks, R.E., Smith Jr., J., Lowther, S.E., Connell, J.W., Siochi, E.J., Harrison, J.S., and T.L. St. Clair. 2002. Chemical Physics Letters 364: 303-308. Park, J.H., and S.C. Jana. 2003. Polymer 44: 2091-2100. Park, K.R., and Y.C. Nho. 2003. Radiation Physics and Chemistry 67:361-365.

PAGE 260

235 Patel, H.J., Su, S.H., Patterson, C., and K.T. Nguyen. 2006. Biotechnology Progress 22(1): 38-44. Pathmanathan, K., and G.P. Johari. 1994. J Chem Soc Faraday Trans 90(8): 1143-1148. Pathmanathan, K., and G.P. Johari. 1990. Journal of Polymer Science, Part B: Polymer Physics 28: 675-689. Perova, T.S., Vij, J.K., and H. Xu. 1997. Colloid and Polymer Science 275: 323-332. Pickup, J.C., Shaw, G.W., and D.J. Claremont. 1988. Biosensors 3(6): 335-346. Pissis, P., and A. Kyritsis. 1997. Solid State Ionics 97: 105-113. Pissis P., Kyritsis, A., Georgoussis, G., Shilov, V.V., and V.V. Shevchenko. 2000. Solid State Ionics 136-137: 255-260. Polizos, G., Kyritsis, A., Shilov, V.V., and V.V. Shevchenko. 2000. Solid State Ionics 136-137: 1139-1146. Poddar, P., Srikanth, H., Morrison, S. A., and E. E. Carpenter. 2005. Journal of Magnetism and Magnetic Materials 288: 443-451. Porter, D. 1995. Group Interaction Modelling of Polymer Properties New York: Marcel Dekker, Inc. Ratner, B.D., and A.S. Hoffman. 1976. Synthetic Hydrogels for Bi omedical Applications. In: Andrade, J.D., editor. Hydrogels fo r Medical and Relate d Applications, ACS Symposium Series, vol.31, pp. 1-36. Washington: American Chemical Society. Roche. 2002. New Approaches in Diabetes Care http://www.roche.com/rtdcmannh02.pdf Rong, J., Jing, Z., Li, H., and M. Sheng. 2001. Macromolecular Rapid Communications 22: 329-334. Roemer, J. 2004. www.DiabetesSelfManagement.com http://www.diabetesselfmanagement.com/article.cfm?AID=1895&SID=7&SSL=n &SK=9AAG&page=1 Russell, G.A., Hiltner, P.A., Gregonis, D.E., deVisse, A.C., and J.D. Andrade. 1980. Journal of Polymer Science, Part B: Polymer Physics 18:1271-1283. Said-Galiyev, E.E, and L.N. Nikitin. 1993. Mekhanika Kompozitnkh Materialov 29: 259266.

PAGE 261

236 Santerre, J.P., Woodhouse, K., Laroche, G., and R.S. Labow. 2005. Biomaterials 26: 7457-7470. Sawyer, L.C., and D.T. Grubb. 1996. Polymer Microscopy second ed. New York: Chapman & Hall. Schmellenmeier, H. 1956. Zeitschrift fuer Physikalische Chemie 205: 349-350. Seo, K.W., Kim, D.J., and K.N. Park. 2004. Journal of Industrial and Engineering Chemistry 10(5): 794-800. Sepe, M.P. 1995. Thermal Analysis of Polymers 95(8:11). Rapra Publishing. Serrano Aroca, A., Campillo Fernandez, A. J ., Gomez Ribelles, J. L., Monleon Pradas, M., Gallego Ferrer, G., and P. Pi ssis. 2004. Polymer 45(26): 8949-8955. Seymour, R.B., and C.E. Carraher Jr. 1987. Polymer Chemistry: An Introduction New York: Marcel Dekkar, Inc. Shi, F.F. 1996. Surface and Coatings Technology 82: 1-15. Shtilman, M.I. 2003. Polymeric biomaterials: Part1-Polymer implants, New concept in polymer science series Boston: VSP. Sibold, N., Madeo, P.J., Masson, S., and T.N. Pham. 2002. Polymer 43(26): 7257-7267. Skaja, A.D., and S.G. Croll. 2003. Polymer Degradation and Stability 79: 123-131. Skoog, D.A., Holler, F.J., and T.A. Nieman. 1992. Principles of Instrumental Analysis 5th ed. Philadelphia: Sau nders College Publishing. Srikanth, H., Hajndl, R., Moulton, B., and M.J. Zaworotko. 2003. Journal of Applied Physics 93(10): 7089-7091. Starkweather Jr., H.W., and P. Avakian. 1992. Journal of Polymer Science, Part B: Polymer Physics 30: 637-641. Stevens, M.P. 1990. Polymer Chemistry: An Introduction, second ed. New York: Oxford University Press. Sun, Y., Zhang, Z., and C.P. Wong. 2005. Polymer 46: 2297-2305. Suwa, T., Jikei, M., Kakimoto, M.A., Imai, Y., Tanaka, A., and K. Yoneda. 1996. Thin Solid Films 273: 258-262.

PAGE 262

237 TA Instruments, DEA 2970 Dielectric Analyz er, TA-057, TA Instruments Applications Library: www.tainst.com TA Instruments, Dynamic Mechnical An alyzer, TA-232C. 2002. TA Instruments Applications Library : www.tainst.com TA Instruments, Design of a New DSC Cell with Tzero™ Technology, TA-273, TA Instruments Applications Library: www.tainst.com TA Instruments. 1997 DEA 2970 Dielectric Anal yzer Operator's Manual. Delaware: TA Instruments. TA Instruments. 2000 DSC 2910 Differential Sc anning Calorimetry Operator's Manual. Delaware: TA Instruments. TA Instruments. 2000 TGA 2950 Thermogravim etric Analyzer Operator's Manual. Delaware: TA Instruments. Tarducci, C., Schofield, W.C.E., Badyal, J.P.S., Brewer, S.A., and C. Willis. 2002. Chemistry of Materials 14: 2541-2545. Tatro, S.R. 2002. Ionizing Radiation Effects on Polymers and Polymer Nanocomposites. University of South Florida: Doctoral Dissertation. Tatro, S.R., Clayton, L.M., O’Rourke Muisen er, P.A., Rao, A.M., and J.P. Harmon. 2004. Polymer 45: 1971-1979. Tavakoli, M. 2005. Surfaces and Interfaces for Biomaterials Vadgama, P., editor. Cambridge: Woodhead Publishing Ltd., pp. 719-744. Thomas, K., Kiwit, M., and W. Kerner. 1998. Experimental and C linical Endocrinology and Diabetes 106: 465-469. Tsangaris, G.M., Psarras, G.C., and N. Kouloumbi. 1998. Journal of Materials Science 33: 2027-2037. Varga,Z., Fehr J., Filipcse i G., and M. Zrnyi. 2003. Macromolecular Symposium 200: 93-100 Wientjes, K.J., Vonk, P., Kleh, Y.V-V., Schoonen, A.J.M., and N.W. Kossen. 1998. Diabetes Care 21(9): 1481-1488. Wilson, J.L., Poddar, P., Frey, N.A., Srikanth, H., Mohomed, K., Harmon, J.P., Kotha, S., and J. Wachsmuth. 2004. Journal of Applied Physics 95: 1439-1443.

PAGE 263

238 Xiong, M., Wu, L., Zhou, S., and B. You. 2002. Polymer International 51: 693-698. Yasuda, H. 1985. Plasma Polymerization New York: Academic Press, Inc. Yasuda, H., and Q. Yu. 2004. Plasma Chemistry and Plasma Processing 24(2): 325-351. Yu, B., Moussy, Y., and F. Moussy. 2005. Electroanalysis 17(19): 1771-1779. Yu, B., Moussy, Y., and F. Moussy. 2005. Frontiers in Bioscience 10: 512-520. Yuan, J., Chen, L., Jiang, X., Shen, J., and S. Lin. 2004. Colloids and Surfaces B: Biointerfaces 39: 87-94. Zhong, Y., and R.V. Bellamkonda. 2005. Journal of Controlled Release 106(3): 309-318.

PAGE 264

239 APPENDICES

PAGE 265

240 APPENDIX A: Chapter 3 Thermal Methods for DSC, DEA, DMA DSC Segment Description Segment 1: Data storage: off Segment 2: Equilibrate at 25.00 C Segment 3: Isothermal for 2.00 min Segment 4: Data storage: on Segment 5: Ramp 5.00 C/min to 140.00 C TGA Segment Description Segment 1: Ramp 20.00 C/min to 400.00 C DEA Segment Description Segment 1: Data storage: off Segment 2: Equilibrate at 135.00 C Segment 3: Isothermal for 3.00 min Segment 4: Equilibrate at -150.00 C Segment 5: Isothermal for 1.00 min Segment 6: Data storage: on Segment 7: Isothermal for 2.00 min Segment 8: Frequency sweep Segment 9: Increment 5.00 C Segment 10: Repeat segment 7 until 275.00 C DMA Segment Description Segment 1: Data storage: off Segment 2: Equilibrate at -150.00 C Segment 3: Isothermal for 1.00 min Segment 4: Data storage: on Segment 5: Isothermal for 1.00 min Segment 6: Frequency sweep Segment 7: Increment 5.00 C Segment 8: Repeat segment 5 until 200.00 C

PAGE 266

241APPENDIX B: Chapter 4 Thermal Methods for DSC, DEA, DMA DSC Segment Description Segment 1: Data storage: off Segment 2: Equilibrate at 25.00 C Segment 3: Isothermal for 2.00 min Segment 4: Data storage: on Segment 5: Ramp 10.00 C/min to 140.00 C DEA Segment Description Segment 1: Data storage: off Segment 2: Equilibrate at 140.00 C Segment 3: Isothermal for 3.00 min Segment 4: Equilibrate at -150.00 C Segment 5: Isothermal for 1.00 min Segment 6: Data storage: on Segment 7: Isothermal for 2.00 min Segment 8: Frequency sweep Segment 9: Increment 5.00 C Segment 10: Repeat segment 7 until 200.00 C DMA Segment Description Segment 1: Data storage: off Segment 2: Equilibrate at -150.00 C Segment 3: Isothermal for 1.00 min Segment 4: Data storage: on Segment 5: Isothermal for 1.00 min Segment 6: Frequency sweep Segment 7: Increment 5.00 C Segment 8: Repeat segment 5 until 200.00 C

PAGE 267

242APPENDIX C: Chapter 5 Thermal Methods for DSC, DEA, DMA DSC Segment Description Segment 1: Data storage: off Segment 2: Equilibrate at 25.00 C Segment 3: Isothermal for 2.00 min Segment 4: Data storage: on Segment 5: Ramp 5.00 C/min to 140.00 C DEA Segment Description Segment 1: Data storage: off Segment 2: Equilibrate at 135.00 C Segment 3: Isothermal for 3.00 min Segment 4: Equilibrate at -150.00 C Segment 5: Isothermal for 1.00 min Segment 6: Data storage: on Segment 7: Isothermal for 2.00 min Segment 8: Frequency sweep Segment 9: Increment 5.00 C Segment 10: Repeat segment 7 until 275.00 C

PAGE 268

ABOUT THE AUTHOR Kadine Mohomed received a B.S. Degree in Chemistry with a Biology minor from Florida Atlantic University in 2000. She entered the Ph.D. Chemistry program at the University of South Florida in Fall 2000 and officially joined Dr. Julie Harmon’s polymer materials research lab in Summer 2001. While in the Ph.D. program at the Univ ersity of South Flor ida, Kadine taught general, organic and instrumental chemistry laboratory classes and was nominated for the Provost’s Outstanding Teaching award for e fforts while teaching Organic chemistry laboratory classes. During her stay at USF, she was awarded the Shembekar Scholarship (2001), Florida-Caribbean Scholarship ( 2003-2006) and Tharp Award (2004, 2005). She received a National Institute of Health (N IH) Research Assistantship from 2004 to 2006, and was nominated by the USF Chemistry department to apply for the American Chemical Society (ACS) Irving Sigal Fellowshi p (2006). Kadine has al so authored four publications and coauthored two publications in various peer-reviewed scientific journals, and has made eight conference presentations at several regional chemical meetings. Upon graduation, she will join the thermal analysis comp any, TA Instruments, as a thermal applications scientist.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001788184
003 fts
005 20070518095831.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 070518s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001442
040
FHM
c FHM
035
(OCoLC)131042952
049
FHMM
090
QD31.2 (ONLINE)
1 100
Mohomed, Kadine.
0 245
Thermal analyses of hydrophilic polymers used in nanocomposites and biocompatible coatings
h [electronic resource] /
by Kadine Mohomed.
260
[Tampa, Fla] :
b University of South Florida,
2006.
3 520
ABSRACT: This research focuses on two hydrophilic polymers that form hydrogels when they sorb water: Poly(2-hydroxyethyl methacrylate) (PHEMA) and Poly(2,3-dihydroxypropyl methacrylate) (PDHPMA). Present work in the field obviated the need to properly characterize the thermal and dielectric properties of these materials.The dielectric permittivity, e', and the loss factor, e", of dry poly(2-hydroxyethyl methacrylate) and poly(2,3-dihydroxypropyl methacrylate) were measured using a dielectric analyzer in the frequency range of 0.1Hz to 100 kHz and between the temperature range of -150 ¨C to 275¨C. The dielectric response of the sub-Tg gamma transition of PHEMA has been widely studied before but little to no DEA data above 50¨C is present in the literature. This study is the first to present the full range dielectric spectrum of PHEMA, PDHPMA and their random copolymers up to and above the glass transition region. The electric modulus formalism and several mathematical proofs were us ed to reveal the gamma, beta, alpha and conductivity relaxations. Dielectric analysis gives insight into the network structure of the polymer; it has been shown through thermal analyses that as the DHPMA content increased in HEMA-DHPMA copolymers the polymer matrix increased in available free volume and facilitated the movement of ions in its matrix. This is of significance as we then investigated the feasibility of using PHEMA, PDHPMA and their random copolymers as materials for a biocompatible coating for an implantable glucose sensor. The biocompatibility of hydrogels can be attributed to the low interfacial tension with biological fluids, high gas permeability, high diffusion of low molecular weight compounds, and reduced mechanical and frictional irritation to surrounding tissue. Once the biocompatibility of the hydrogels was established, the task to coat the polyurethane (PU)/epoxy coated metal glucose sensor was addressed. Plasma polymerization was found to be the most feasible^ technique for the application of the biocompatible hydrogel as a coating on the implantable glucose sensor. It has also been shown that thermal analysis techniques provide a mode of investigation that can be used to investigate the interfacial interactions of a novel hydroxylated, self-assembled nanoparticle with two functionally different polymers, poly(2-dihydroxyethyl methacrylate) and poly(methyl methacrylate).
502
Dissertation (Ph.D.)--University of South Florida, 2006.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 242 pages.
590
Adviser: Julie P. Harmon, Ph.D.
653
Poly(2-hydroxyethyl methacrylate).
Poly(2
3-dihydroxypropyl methacrylate).
Dielectric analysis.
Hydrogel.
690
Dissertations, Academic
z USF
x Chemistry
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1442