|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 UNDERSTANDING PRECISION POLYOLEFINS BY SOLID STATE NU C LEAR MAGNETIC RESONANCE SPECTROSCOPY AND X RAY SCATTERING By YUYING WEI A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010
2 2010 Yuying Wei
3 To my family
4 ACKNOWLEDGMENTS During my four year study in the Graduate School at the University of Florida, I have received wonderful help, support, encouragement, advice, guidance, and friendship from a great number of people. I would not have been able to finish my PhD study smoothly or not at all, without help from these people. First and foremost, I would like to acknowledge my advisor, Professor Ken Wagener, for his guidance, advice, encouragement, patience, and understanding during the course of my graduate life I thank him for opening the door to let me join the research group for always being around whenever he is needed for his advice and suggestions extracted from his wisdom and experience, for his unde rstanding of culture differences, for his constant support and positive influence on not only my research but also my view of the world, fo r all the time he spent with me talking about both science and life, and for his confidence in my abilities as a chemist Just as he said, he is not my boss, he is my advisor. I received advice from Professor Wagener regarding all aspects of my life in Gainesville. I will be thankful forever for his presence in my life. I also thank Prof essor Russ Bowers It has been a great privilege for me to work with him i n solid state NMR. Prof essor Bowers not only introduc e d me solid state NMR techniques but more importantly all the fun of being a physical chemist. His encouragement, kindness, friendship, and guidance have been one of the sources direct ing me to my goals. I want to thank the members of my committee (Dr. David Powell, Professors Charles Martin, and Elliot Douglas) for their helpful discussions and help during my time at Florida. In addition, I want to thank Professor Valentin Cracium for the help on X ray scattering technique. Furthermore, I want to thank Dr Kathryn Williams for her helpful
5 discussion s of the DSC results and for edit ing my manuscripts. Special thanks go to Dr. Ben Smith and Mrs. Lori Clark in the graduate office for all their help and advice. I thank Mrs. Sara Klossner, and Mrs. Gena Borre ro on the Butler floor for their help and friendship. It is Sara who treats trivial and ordinary daily business with beautiful smiles. I appreciated all the fun that I had at Genas house. Further, I would like to thank Dr. Tammy Davidson and Dr. James Horvath for their kindness and help when I taught labs under their supervision. I thank all of the excellent employees in the IT shop, stock room, business office, and main office for all their help and support. During my four year stay in Gainesville, I par ticularly acknowledge all the previous and present members in Wagener group. When I just joined the group, I received both a warm welcome and detailed help from Dr. Giovanni Rojas, Dr. Erik Berda, Dr. James Leonard, and Dr. Emine Boz. Each of them shared t heir synthetic knowledge as well as friendship with me. I want to thank Professor Fabio Zuluaga, Bora Inci, and Brian Aitken for their great help on synthesis. Being a member in the Wagener research group has been my pleasure. I thank Dr. Kate Opper, Paula Delgado, Sam Popwell, Pascale Atallah, Diane Turek, Chet Simocko, Ashlyn Dennis, Luke Fisher, and Chip Few for their company and friendship. I also want to thank all the members on the Butler floor, especially Dr. Jianguo Mei and people in the lab of 309 Sisler Hall: Frank Arroyave, Erik Shen, David Liu, and Dr. Chad Amb. I would like to express my appreciation to members in Bowers research group. Special acknowledgement goes to Dr. Chi Yuan Cheng for his help on solid state
6 NMR operations and friendship. I thank Dr. Caroline Pointer Keenan, Ryan Wood, Chris Reeg, Chris Akel, and Dr. Amrish Menjoge for their friendship and help. Since I spent six months in total at Mainz, Germany, I want to offer my special acknowledgement to the people at Max Planck Institute for Polymer Research. First of all, I want to thank Prof. Dr. Hans Spiess and Prof. Dr. Klaus Mllen for offering me the chance to do research at Mainz. Also, I want to thank Professor Spiess for his great guidance, advice, and vision in solid st ate NMR field. Moreover, I thank Dr. Robert Graf for his detailed education and help on the operations of NMR spectrometers at AKSpiess group, for his solid knowledge on NMR for his strict attitude towards science, and for his warm friendship. I also want to thank all the members of AK Spiess group for their help and friendship, special acknowledgement goes to Dr. Matthias Junk, Dr. Shu Jie, Dr. Michael Hansen, Dr. Anne Bohle, and Dr. Mujeeb Khan. I also want to thank Dr. Werner Steffen, Dr. Guenter Auern hammer, Miao Wang, and Dr Ingo Lieberwirth for their friendship and collaborations. I want to thank I would like to thank my parents for their endless love, belief, devotion, support and sacrifices they have made that allowed me to become a person that I am today. I want to thank my husband Kevin and my daughter Abbie They have been everything for me. W ithout their love, care, support and belief in me, I would be incomplete. I want to thank my two brothers for their love and care of the families I have in China. I thank all my friends in and out of Gainesville for their friendship and all the help! my friends outside of AK Spiess group: Katja Nillis, Robert Haschick, Jiawei Shen, and Bin Miao, for their friendship.
7 TAB LE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES .......................................................................................................... 10 LIST OF FIGURES ........................................................................................................ 11 LIST OF ABBREVIATIONS ........................................................................................... 14 ABSTRACT ................................................................................................................... 16 CHAPTER 1 INTRODUCTION .................................................................................................... 18 2 COMMERCI AL POLYETHYLENE AND PRECISION POLYOLEFINS ................... 21 Commercial Polyethylenes ..................................................................................... 21 Classifications of Commercial Polyethylenes ................................................... 21 Industrial Manufacturing/Processing of Polyethylene ....................................... 25 Relationship of Microstructure to Physical Properties ...................................... 29 Density ....................................................................................................... 29 Crystallization and m orphology .................................................................. 31 Th ermal p roperties ..................................................................................... 35 Other p hysical p roperties ........................................................................... 37 Precision Polyolefin ................................................................................................ 39 Structure of Precision Polyolefin ....................................................................... 39 Synthesis of Acyclic Diene Metathesis ( ADMET ) Precision Polymers .............. 40 Monomer synthesis .................................................................................... 40 Polymer synthesis ...................................................................................... 42 Synthesis of A DMET Random Polymers .......................................................... 44 Comparison of Commercial Polyethylene ( PE ) to ADMET PE ................................ 45 Structure Modeling ........................................................................................... 46 Comparison of Physical Properties .................................................................. 49 3 SOLID STATE NMR SPECTROSCOPY AND X RAY SCATTERING TECHNIQUES ........................................................................................................ 54 Polymer Characterization Methods ......................................................................... 54 Solid State NMR Spectroscopy ............................................................................... 56 Introduction to NMR ......................................................................................... 56 Resonance and chemical shift ................................................................... 57 Chemical s hift a nisotropy ........................................................................... 60 NMR i n teractions ....................................................................................... 63 Cross p olarization and m agic a ngle s pinning ............................................. 64
8 NMR Spectrometer ........................................................................................... 66 Introduction to Deterium Solid State NMR ........................................................ 68 Pulse Sequences ............................................................................................. 70 Quadrupole e cho ....................................................................................... 70 SUPER ...................................................................................................... 71 REDOR ...................................................................................................... 7 2 REREDOR ................................................................................................. 73 X ray Scattering Techniques ................................................................................... 74 Intr oduction ....................................................................................................... 74 WAXS and SAXS ............................................................................................. 75 Braggs Law ...................................................................................................... 76 4 INFLUENCE OF BRANCH FREQUENCY ON DYNAMICS AND STRUCTURE OF PRECISION POLYMERS ................................................................................. 78 Introduction ............................................................................................................. 78 Polymer Synthesis .................................................................................................. 79 Monomer Synthesis .......................................................................................... 79 Polymer Synthesis ............................................................................................ 80 Characterization of Deuterated Polymers ............................................................... 80 Basic Characterization ..................................................................................... 81 Morphology Measurement ................................................................................ 81 Deuterium Solid State NMR Measurement ....................................................... 84 13C Solid State NMR Measurement .................................................................. 88 Motional Model ................................................................................................. 92 Analysis of NMR Powder Line shapes ............................................................. 94 Conclusion of Branch Frequency Effect .................................................................. 97 5 M ORPHOLOGY AND DYNAMICS BY DEUTERIUM NMR LINESHAPE ANALYSIS .............................................................................................................. 98 Introduction ............................................................................................................. 98 Theories of Li neshape Analysis .............................................................................. 99 Experimental ......................................................................................................... 100 Results & Discussion ............................................................................................ 101 Deuterium T1 Measurement ........................................................................... 101 Fully Relaxed and Amorphous 2H NMR Measurements ................................. 106 Theoretical Lineshape Fitting ......................................................................... 108 Conclusion ............................................................................................................ 115 6 EFFECTS OF BRANCH IDENTITY ON THE MORPHOLOGY OF PRECISION POLYOLEFINS WITH ALKYL BRANCHES .......................................................... 116 Introduction ........................................................................................................... 116 Thermal Properties ............................................................................................... 117 Branch Displ acement Determination by WAXS .................................................... 118 Experimental .................................................................................................. 119
9 Results and Discussion .................................................................................. 119 Investigation of Dynamic by SS NMR ................................................................... 124 Experimental .................................................................................................. 124 Results and Discussion .................................................................................. 125 Morphological Models ........................................................................................... 128 7 ADMET PRECISION POLYETHYLENE WITH LOW BRANCH FREQUENCY .... 131 Introduction ........................................................................................................... 131 Experimental ......................................................................................................... 131 Results and Discussion ......................................................................................... 133 Thermal Behavior ........................................................................................... 133 Solid State NM R Investigation ........................................................................ 137 Conclusion ............................................................................................................ 144 8 SUMMARY AND OUTLOOK ................................................................................ 146 Summary .............................................................................................................. 146 Structural Modeling ........................................................................................ 146 Polymer Synthesis .......................................................................................... 147 Branch S pacing E ffect .................................................................................... 147 Branch I dentity E ffect ..................................................................................... 148 Outlook ................................................................................................................. 149 APPENDIX A LAMELLAR THICKNESS DETERMINATION ....................................................... 151 B D E UTERIUM 2D EXCHANGE NMR SPECTRA ................................................... 152 LIST OF REFERENCES ............................................................................................. 153 BIOGRAPHICAL SKETCH .......................................................................................... 161
10 LIST OF TABLES Table page 2 1 General properties for various classes of commercial polyethylenes ................. 24 2 2 Unit cell dimensions of HDPE and ADMET PE .................................................. 52 3 1 Polymer characterization methods used in this dissertation ............................... 55 3 2 List of three nuclei investigated in this dissertation. ............................................ 58 4 1 Basic characterization data of deuterated and protonated precision polymers ... 81 6 1 Basic characterization data of PE21R ............................................................. 116 6 2 Recoupling constants derived from REREDOR experiments. .......................... 127 7 1 Molecular weight and thermal data for ADMET precision polyethylene with branch spacer of 39, referenced to linear ADMET PE. ..................................... 133 7 2 Quantitative data of chemical shifts and peak information for PE39CH3hmw. 142
11 LIST OF FIGURES Figure page 2 1 Schematic representations of commercial polyethylenes. .................................. 22 2 2 Plot of density and degree of crystallinity as functions of branch content for compressionmolded HDPE and LLDPE resins .................................................. 30 2 3 Typical unit cell structures and their characteristic dimensional values .............. 32 2 4 Three unit cell structures in commercial PE ........................................................ 34 2 5 Plot of heat flow of three types of PE samples as function of temperature/heating rates .................................................................................. 35 2 6 Relationship of structureproperty performance of polymeric materials .............. 38 2 7 General chemical structure of precision polyolefins. ........................................... 39 2 8 Precision monomer synthesis methodology starting with malonate .................... 40 2 9 Precision monomer synthesis methodology via cyanide chemistry .................... 41 2 10 Acyclic diene metathesis polymerization scheme in general. ............................. 42 2 11 Metathesis and hydrogenation Catalysts ............................................................ 42 2 12 Hydrogenation of unsaturated ADMET precision polymers ................................ 43 2 13 Laboratory synthesis of linear ADMET PE ......................................................... 44 2 14 Synthesis of ADMET random polyolefins with irregular branch spacing ............. 45 2 15 Synthesis of ADMET random polyolefins with unequal branch identities ........... 45 2 16 Correlation of ADMET PEs to commercial PE .................................................... 48 2 17 Peak melting temperatures of commer cial PE, ADMET PEs. ............................. 50 2 18 Orthorhombic crystal structure of linear ADMET PE. .......................................... 52 3 1 Schematic illustration of chemical shift mechanism. ........................................... 59 3 2 The definition of tensor orientation in diff erent coordinate systems ................... 61 3 3 Special CSA lineshapes ..................................................................................... 63 3 4 Th e crosspolarization pulse sequence. ............................................................. 65
12 3 5 The magic angle spinning experiment setup. ..................................................... 66 3 6 Schemetic overview of a NMR spectrometer ...................................................... 68 3 7 Pake pattern of 2H NMR .................................................................................... 69 3 8 The Quandrupole echo pulse sequence ............................................................. 71 3 9 The SUPER pulse sequence under CP condition ............................................... 72 3 10 The rotational echo double resonance (REDOR) pulse sequence. .................... 73 3 11 The pulse sequence of CP based Rotor encored REDOR (REREDOR) ............ 74 3 12 Schematic illustration of X ray scattering for solid samples ................................ 76 3 13 Scheme of Braggs Law ...................................................................................... 77 4 1 Structure of precisely branched polymers .......................................................... 78 4 2 Synthetic methodology used to produce perdeuterated monomers .................... 79 4 3 Synthetic methodology used to produce precision polyolefins via ADMET chemistry. ........................................................................................................... 80 4 4 TEM images for ADMET precision polymer PE21CD3 ...................................... 84 4 5 Temperature dependence of 2H spectra ............................................................. 86 4 6 13C isotropic and anisotropic chemical shifts for precise polymers. .................... 88 4 7 Schematic model for rotational dynamics ........................................................... 92 4 8 Computed 2H NMR and 13C NMR powder line shapes. ...................................... 96 5 1 Theoretical 2H lineshapes in static and fast motional limits .............................. 100 5 2 T1 determination from inversion recovery using various fit functions. ............... 102 5 3 Spin lattice rel axation data for ADMET precision polymer PE21CD3 .............. 103 5 4 Stretching parameter n at various temperatures ............................................... 104 5 5 Non crystalline fraction acquired from bimodal function EK at various temperatures for polymer PE21CD3. ............................................................... 105 5 6 Degree of crystallinity acquired from WAXS ..................................................... 105
13 5 7 Plot of temperature dependent fully relaxed (experimental), amorphous (experimental), as well as crystalline spectra (calculated) ................................ 107 5 8 Fitting of experimental amorphous spectra ....................................................... 110 5 9 Fitting of calculated crystalline spectra ............................................................. 112 5 10 Fitting of fully relaxed spectra ........................................................................... 114 6 1 Plot of melting temperature as a function of branch identity ............................. 118 6 2 WAXS results of PE21 R .................................................................................. 120 6 3 Plot of scattering angle of two strong reflections as a function of branch identity .............................................................................................................. 121 6 4 Peak assignment of slices from REREDOR spectra ........................................ 126 6 5 Flow chart of Branch displacement ................................................................... 128 6 6 Schematic morphology of precision polymers with large defect branches ........ 130 7 1 Synthetic methodology used to produce precision methyl branched (18,18) monomers. ........................................................................................................ 132 7 2 Plot of melting point as a function of branch content ........................................ 134 7 3 Correlation of melting temperature and lamellar thickness ............................... 136 7 4 13C SS NMR spectrum of PE39CH3 at melted state. ....................................... 138 7 5 Temperature dependent 13C spectra of PE39 CH3 under SPE ........................ 139 7 6 Temperature dependent 13C spectra of PE39 CH3hmw using SPE with dipolar decoupling. ........................................................................................... 140 7 7 Plot of chemical shift difference between crystalline and noncrystalline CH2 as a function of temperature. ............................................................................ 142 7 8 13C spectra of PE39 CH3hmw from various crystallization conditions ............. 143
14 LIST OF ABBREVIATION S PE Polyethylene ADMET Acyclic diene metathesis NMR Nuclear magnetic resonance SSNMR Solid state nuclear magnetic resonance WAXS Wide angle X ray scattering HDPE High density polyethylene LDPE Low density polyethylene UHMWPE Ultra high molecular Weight LLDPE Linear low density polyethylene VLDPE Very low density polyethylene XRD X ray diffraction DSC Differential scanning calorimeter TGA T hermal gravimetric analysi s MS Mass spectrometry MALDI Matrix assisted laser desorption/ionization TOF Time of fly IR Infrared spectroscopy GPC Gel permeation chromatography TEM Transmission electron microscopy CP Cross polarization MAS Magic angle spinning SPE Single pulse excitation AFM Atomic force microscopy Tm Melting temperature
15 TgX Glass transition temperature cT Degree of crystallinity 1T Spin lattice relaxation time 2SUPER S eparation of undistorted powder patterns by effortless recoupling Spin spin relaxation time REDOR Rotational echo doubleresonance REREDOR Rotor encoding rotational echo doubleresonance CSA Chemical shift anisotropy 1D One dimensional 2D Two dimensional 3D Three dimensional FID Free induced decay CR Crystalline region NCR Non crystalline region AR Amorphous region
16 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy UNDERSTANDING PRECISION POLYOLEFINS BY SOLID STATE NU C LEAR MAGNETIC RESONANCE SPECTROSCOPY AND X RAY SCATTERING By Yuying Wei August, 2010 Chair: Kenneth B. Wagener Major: Chemistry In commercial polyethylene (PE), the microstructure, i.e. the branching determines the macroscopic properties of th e polymers as well as its wide applications. To better understand the structureproperty performance relationship, precision polyolefins prepared via acyclic diene metathesis (ADMET) polycondensation chemistry are used as model structures Due to the intri nsic nature of ADMET polymerization, precision polyolefins can be prepared having, without equivocation, known branch identity and exact location of the branches along the polyolefin backbone. Precision polyethylenes with alkyl branches are selected as str uctural models to investigate the influence of branch identity and frequency on polymer morphology and chain dynamics. Several analytical methods are employed, mainly solid state nuclear magnetic resonance (SS NMR) spectroscopy and wide angle X ray scatter ing (WAXS) techniques. Deuterium SS NMR is most ly appropriate for investigat ing chain dynamics because of the quandrupolar interaction. Deuterated precision polymer samples synthesized by selectively labeling the site of interest are studied using advanced 2H and 13C SS NMR. A unique spectrum appears for ADMET precision polymer with a CD3
17 branch on each and every 15th carbon along the main chain: a regular Pake pattern exists with a half static line width, indicating the presence of axial rotation sufficiently fast due to the motions of methyl groups embedded in the crystalline regions. The axial rotation is also observed by 13Besides SS NMR, powder X ray scattering has also been applied to precision pol ymers. W ide angle X ray scattering provides important information about crystallinity and the unit cell, as well as the lamellar structure. Combining the results from solid state NMR and X ray scattering, a clear understanding of morphology has been achiev ed. When the side chain switche s from the methyl group to a bigger substituent, such as the butyl group, the thermal behavior of polymer changes significantly. The influences of the branches on crystallinity and chain packing are investigated by SSNMR and XRD. It is found that the scattering patterns of methyl branched and butyl branched polymers are different from that of the linear ADMET PE. Small branches such as the methyl group, are incorporated in the unit cell, while the branches as large as or larger than the propyl group are excluded from the unit cell. C NMR. For lower branch content the twist motions are decoupled (pinned defects), while for higher branch contents collective motion (rotator phase) is found. For the first time, the presence of the rotator phase in polyethylene crystalline region has been evidenced by solid state NMR.
18 CHAPTER 1 INTRODUCTION Since it was first produced in large scale in 1936, industrial polyethylene (PE) has been the most manufactured polymer worldwide. Depending on the synthetic methodology and polymer processing, PE can be classified into various types, such as high density polyethylene (HDPE), low density polyethylene (LDPE) linear low density polyethylene (LLDPE), ultra high molecular polyethylene (UHMPE), etc. E ach type exhibits distinct macroscopic properties, and correspondingly various performances in broad applications. To understand the structureproperty performance relation ships of these materials, precision polyolefins are used as model structures for commercial polyethylenes. In precision polyethylene, the branch identity and branch frequency can be precisely controlled, while in commercial PE, there are many variables, including the branch size, branch spacer, and branch distribution. By limiting the freedom of the system, a better understanding of branching in PE can be achieved. It is believed that the macroscopic properties are inherently dependent on the micro structural features. For instance, the thermal behavior of PE is determined by the chainpacking in the crystalline region; likewise the mechanical propert ies of PE are correlated to the dynamics of both the main chain and the side branches. Various analy tical methods have been employed to investigate the structures and physical properties of precision polyolefins This research focuses on two of these methods: solid state nuclear magnetic resonance spectroscopy (SS NMR) and wide angle X ray scattering (W AXS). SSNMR is recognized for its capability of understanding structure and dynamics of condensed matter. Coupled with a high magnetic field, the modern pulse sequences have been used to study many nuclei,
19 such as 1H, 2H, and 13This dissertation is organized as follows: i n Chapter 2, commercial polyethylene (different structures and their synthesis), and precision polyolefins (synthesis and uniqueness) are briefly reviewed. The comparison of commercial PE to precision PE is presented. P recision polyolefins with various branches are discussed to investigate the influence of branching on the thermal, mechanical, dynamic, and morphological properties of these unique structures. C by offering NMR signals with good resolution and high sensitivity. WAXS is wellknown for investigat ing ordered structures. Information related to the degree of crystallinity, crystal structure, and unit cell dimension can be obtained directly Other analytical techniques, such as differential scanning calorimet ry (DSC), transmission electron microscopy (TEM), are also used to understand the thermal behaviors and morpholog ies of the materials respectively. Chapter 3 describes the two main analytical techniques that are employed in the dissertation: SS NM R and WAXS including the fundamental principles, the instrumentation, operation, and data interpretation. Chapter 3 elucidates the analytical tools used to understand the effects of branch identity and frequency on polymer macroscopic properties. In Chapter 4, the branch f requency effect is discussed using three precision polyolefins as examples: PE21 CD3, PE15 CD3, and PE9CD3. The 2H NMR spectra under static conditions and 13C NMR spectra with magic angle spinning are explored. By combining the results from these two measurements, a motional model is proposed, and the concept of rotator phase presenting in the crystalline region of polyethylene is
20 evidenced. The results indicate that the branch frequency indeed has significant influence on the structur es. Lineshape analysis of 2H SS NMR is specifically described in Chapter 5. By applying a deuterium T1In Chapter 6, the influence of the branch identity on morpholog y is investigat ed by sampling the precision polyolefins with alkyl branches on each and every 21 filter prior to the solid echo pulse sequence, experimental amorphous spectra can be acquired. The crystalline spectra are calculated by subtracting the experimental amorphous signals from the fully relaxed total spectra. With software simulation, the experimental spectra can be fit to various combinations of motional models. stIn Chapter 7 the ADMET precision polyolefin s with reduced branch frequency PE39 R with the alkyl group R on each and every 39 carbon along the ethylene backbone. WAXS and SS NMR data imply that the smaller branches, such as the methyl and the ethyl groups are incorporated in the crystalline unit cell, while the bulkier groups, propyl and larger are excluded from the crystall ine unit cell. th backbone carbon, are explored. Th is series of polymers exhibit more complicated crystalline structures. The SS NMR shows a broad peak representing the methyl branch, and a splitting of chemical shift belonging to carbon on the backbone. The chain folding and packing in PE39 CH3Chapter 8 presents a summary of the dissertation and suggestions for future research. implies the presence of more than one crystalline structure
21 CHAPTER 2 COMMERCIAL POLYETHYL ENE AND PRECISION PO LYOLEFINS 2 1 Commercial Polyethylenes 2 1 .1 Classifications of Commercial Polyethylenes P olyethylene (PE) has the simplest molecular structure in commercial polymers, characterized by its long backbone of covalently linked CH2 units with methyl groups as terminated chain ends. In spite of its deceptively simple chemical composition, PE has been widely used in many applications since it was first synthesized accidently by the German chemist Hans von Pechmann in 1898.1 It exhibits diverse applications as a result of low cost, flexibility, ductlility clarity, lightness, nontoxicity, good mechanical and thermal stability, and excellent electric resistance, etc. The versatility of this polymeric material arises chiefly from branching, which act s as defect s in the crystalline regions and modif ies the nature of the structure. For instance, the introduction of alkyl branches in PE lower s its crystallinity level. In principle, the degree of crystallinity is in versely proportional to the degree of defects in the structure, i.e., the more the defects, the lower the crystallinity level will be Since the chain packing in crystalline regions is denser than that in the noncrystalline regions, the overall density of polymers is determined mostly by the portion of crystalline domains. Less branching (f ewer defects ) results in denser materials. In general, the macroscopic properties, such as density, are strongly related to branching and crystallinity of the material. For a pure polyethylene, specifically, PE with no branching at all its density is 1 g/cm3; for a totally amorphous PE, which possesses high level of branching, its density is 0.85 g/cm3B ased on its density and branching, PE is roughly classified into four different categories, high density polyethylene (HDPE), low density polyethylene (LDPE), linear
22 low density polyethylene (LLDPE), and very low density polyethylene (VLDPE), with their schematic structures shown in Figure 21. Figure 2 1 Schematic representations of commercial polyethylenes: (a) high density polyethylene; (b) low density polyethylene; (c) linear low density polyethylene; (d) very low density polyethylene. High density polyethylene (HDPE) or low pressure polyethylene, as shown in of Figure 2 1 (a), has a typical density falling in the range of 0.94 0.97 g/cm3. C onsi sting of negligible branching to perturb the crystalline organization, HDPE has the chemical structure which is the closest to pure polyethylene. It is also termed as linear polyethylene (LPE) due to its low level of branching and highly linear structure. HDPE has rather high crystallinity compared to any other types of polyethylene. Thus it appears opaque, behaves tough and stiff, and withstands rather high temperature2Low density polyethylene (LDPE), or high pressure polyethylene, is so named because of its relatively low density falling in the range of 0.90 0.94 g/cm ( see mechanical and thermal data shown in T able 21 ) HDPE is utilized widely as carrier bags, pipes, bottles, toys, etc. 3 (a) (b) (c) (d) which
23 result s from the low crystallization level due to the substantial concentrations of branching. The branches are primarily ethyl and butyl groups together with some rather long chain branches. Determined by the nature of polymerization process at the high pressure at which LD PE is manufactured, the short chain bra n ches are clustered randomly and separated by the long runs of nonbranched ethylene sequence. Long chain branches, which are displaced randomly along the backbone, can themselves be branched as well, illustrated in Figure 21 (b). Compared to HDPE, LDPE contains more and larger defects, which undoubtedly inhibit the chains packing into a crystalline structure. Consequently LDPE has weaker intermolecular (instantaneous dipoledipole) interactions, as well as less temperature and chemical resistance. It appears to be softer and more flexible than HDPE Major applications for LDPE consist of heavy duty sacks, refuse sacks, films and for general packaging.3Linear low density polyethylene (LLDPE) is a linear version of polyethylene, with a significant concentration of short chain branches, seen in Figure 21 (c). The incorporation of a high number of branches in LLDPE discourages the formation of crystalline structures in terms of crystal size and perfection. The big difference between LLDPE and LDPE is the absence of long chain branching, arising from different polymerization processes of these two types of materials. LLDPE has relatively narrower distribution of molecular weight, more linear structure, and apparently different mechanical, thermal, and rheological responses ( data shown in Table 21 ) LLDPE is used for almost all of the traditional applications for polyethylene, including plastic bags, sheets, food wrap, toys, pipes, buckets, and containers. However, the predominant
24 usage of LLDPE is in the film industry due to its good mechanical property and relative transparency.3Table 21 General properties for various classes of commercial polyet hylenes Property HDPE LDPE LLDPE VLDPE Branch Lengths and Content Few or no branches SCB & LCB Numerous SCB no LCB Large numbers of SCB Density, g/cm 3 0.94 0.97 0.91 0.94 0.90 0.94 0.86 0.90 X c (% from density) 62 82 42 62 34 62 4 34 X c ( % from calorimetry) 55 77 30 54 22 55 0 22 T m C 125 132 98 115 100 125 60 100 Heat of Fusion, cal/g 38 53 21 37 15 43 0 15 Thermal Expansivity, 106 in/in/ C 60 100 100 200 70 150 150 270 Heat Distortion Temperature, C at 66 psi 80 90 40 44 55 80 -Flexural Modulus, k psi at 73 F 145 225 35 48 40 160 <40 Tensile Modulus, k psi 155 200 25 50 38 130 <38 Tensile Yield Stress, k psi 2. 6 4 .5 1 3 2 8 1 1 2 8 <1 1 Tensile Strength at Break, k psi 3.2 4 5 1 2 4 5 1 9 4 5 2 5 4 5 Tensile Elongation at Break, % 10 1,500 100 650 100 950 100 600 Industrial Production Low pressure Ziegler process High pressure, radical reaction Metallocene process Metallocene process Number of Branches per 1000 Carbon Atoms 4 <4 ( Phillips ), 5 7 (Ziegler) 20 30 (methyl), 3 5 ( n butyl) -Numerous D ata adapted from Handbook of polyethylene: structures, properties, and applications .2 Xc denotes the degree of crystallinity. TmSCB: short chain branches; LCB: longchain branches. denotes the melting temperature. Very low density polyethylene (VLDPE), also known as ultra low density polyethylene (ULDPE), is a special class of linear low density polyethylene, defined by
2 5 its density falling in the range of 0.860.90 g/cm3. VLDPE has much higher level of short chain branches, illustrated in Figure 21 (d). The extremely high concentration of branches effectively inhibits crystal organization o f the polymer chains, resulting in quite low crystallinity level or even no crystals at all, i.e., completely amorphous solids, as shwon in Table 21. A typical separation of branches would fall in the range of 725 backbone carbon atoms.2There are some other classification criteria. For instance, based on molecular weight, PE can be divided into high molecular weight PE (HMWPE), ultra high molecular weight PE (UHMWPE); based on the primary structure, PE can be classified to be linear PE, branched PE, grafted PE, crosslinked PE Herein, we focus on the branching in polyethylene and choose the classification by macromolecular density and nature of branches. Like LLDPE, VLDPE is more flexible and less brittle in comparison with HDPE. VLDPE is used mainly for hose s and tubing, food containers and packaging 2 1 .2 Industrial Manufacturing/Processing of Polyethylene T he first laboratory polyethylene was created accidently in 1898 by Hans von Pechmann when heating diazomethane .1 After 35 years, on March 29th 1933, Eric Fawcett and Reginald Gibbon at the British company Imperial Chemical Industries (ICI) made the first industrial PE by accident (again) when reacting ethylene with benzaldehyde at high pressure.2 However, it was not until 1935 when Michael Perrrin determined the experimental conditions for consistent produc ting high pressure polyethylene, known today as low density polyethylene.5 Before World War II PE was almost exclusively utilized as an electrical insulator due to its high dielectric strength, low loss factor, and easy processing. In the years following, Union Carbide, du Pont,
26 and ICI made significant improvements and discoveries that led to revolutions in polyethyl ene production. One of the most important insights was the determination of high content branching in high pressure PE revealed by infrared spectroscopy.6 The branches were identified to be ethyl and butyl groups with content of around twenty branches for every 1000 backbone carbons.2With the development of industrial polyethylene, the processes of making this material can be divided into five relatively distinct routes: (a) highpressure process; (b) Ziegler process; (c) Phil l ips Petroleum (Indiana) process; (d) Standard Oil process; and (e) metallocene process. This insight was greatly appreciated i n the PE industry because subsequent investigation s indicated considerable effects of branching on the physical, rheological, and mechanical behaviors. Such an important correlation allows the tailoring of commercial polymers to meet particular application requirements by controlling the polymerization process, thus the content of branches. (a) High Pressure Process Low density polyethylene resins are made exclusively via free radical polymerization by a high pressure process The conditions required to make the high level branched polyethylene is quite simple : a suitable amount of appropriate free radical, controlled polymerization conditions of high pressure at 100 to 300 MPa and temperature at 80 to 300 C.2, 3 Commonly used initiators include benzoyl peroxide, azodi isobutyronitrile or oxygen.3 This polymerization process involves initiation, chain propagation, chain branching, chain transfer, and termination steps. Chain branching occurs when radical transfer occurs intramolecularly (resulting in short chain branching) or intermolecularly (giving rise to long chain branching).
27 (b) Ziegler Process Ziegler Natta catalysis and metal oxide catalysis are the two industrial processes used to produce high density polyethylene. The Ziegler process occurs at relatively low pressure and temperature. The typical operati ng temperature is around 70 C, which varies with the mode of the starting material of a gaseous reaction mixture or a slurry, the production facilit y is controlled at below the melting point around 30 100 C; for solution reactors, higher operation temperatures can be reached, typically 100200 C.2 Typical operat ing pressure ranges from atmospher ic to as high as 2.1 MPa varying with the modes of the reaction mixture as well F or the gas/s lurry mode, the operati ng pressure is higher than that for the solution mode. One advantage of the Ziegl er process lies in the flexibility and variety of Ziegler catalysts, which consist chiefly of a transition metal halide and alkyl aluminium compounds. The one commonly used in production of HDPE is the complex of triethyl aluminium with titanium tetrachloride.2(c) Phil l ips Petroleum Process The Phillips process involves the use of a metal oxide catalyst at medium pressure ( 1.4 3.5 MPa ) and temperature (130160 C).3 The typical industrial Phi llips catalyst is chromium oxide on silica gel. Commercial polymers produced by this process have a rather low melt flow index of 0.25 and exhibits the highest material density among any commercial PE, as high as 0.96 g/cm3. Due to the nature of polymeriz ation mechanism involving metal oxide catalysts, the resulting polymers are highly ster e ospecific. Temperature plays a very important role in the conversion rate and molecular weight of the final product. For instance, when the operating temperature in a t ypical Phillips process increases from 140 C to just over 170 C, the melt flow index increases 40
28 times and correspondingly the molecular weight decreases. Therefore, in industry, polyethylene with high molecular mass can be realized by the low temperature Phillips process. (d) Standard Oil (Indiana) Process Besides the Ziegler and Phillips process es the Indiana process has been demonstrated to be another applicable industrial process in the production of high density polyethylene. This process operates at a pressure of 410 MPa and temperature of 200 300 C, with supported molybdenum alumina as catalysts and sodium or calcium as promoters.3, 7 The I ndiana process shares many similarities to the Phillips process : They are capabl e of producing HDPE with densities as high as 0.96 g/cm3(e) Metallocene Process both are sensitive to operating temperatures. Characterized by the usage of metallocene compounds as catalysts, this relatively new process brings unique structural and physical properties to the polyethylene industry. In contrast to the Ziegler Natta catalysts which are heterogeneous and contain multiple catalytic sites, the metallocene catalyst s are homogeneous and have only a single catalytic site This is the reason why metallocene catalysts are also referred to as single site catalysts in the industrial environment A typical metallocene molecule consists of an organometallic coordination com pound in which one or two cyclic ligands are bonded to a central transition metal, usually zirconium or titanium. The uniqueness of metallocene process lies in the catalysts which have sound productivity and activity, giving rise to narrow molecular weight distributions and simple tailoring of the final products by controlled modification of the catalyst. Interestingly, stereospecific
29 polymers, such as isotactic and syndiotactic polymers, can be manufactured by using chiral metalloce ne catalysts. LLDPE from metallocene process is denoted as mLLDPE, produced by copolymerization of ethylene with a small amount of higher alpha olefins like 1 butane, 1hexene, and 1 octene. Compared to the traditional LLDPE s, mLLDPE materials show superio r mechanical response and narrower molecular weight distributions Metallocene grades of VLDPE, denoted as mVLDPE, first introduced by Exxon Mobile in 2010,4 exhibit more desir able mechanical and molecular properties in contrast to the conventional VLDPE. However, the metallocene process has not demonstrated much impact on mHDPE, which is the metallocene version of HDPE.22 1 3 Relation ship of Microstructure to Physical Properties P olyethylene is the simplest member in the family of thermoplastic materials, composed of a long chain of aliphatic hydrocarbons. For all grades of commercial PE, the physical properties differ in one way or another, largely related to the microstructure. Structural variables that significantly influence the macroscopic properties include : (a) degree of short chain branching, (b) degree of long chain branching, (c) the distributions of chain lengths and branch concentration, (d) molecular weight and its dis tribution, (e) residue of comonomer, and (f) impurities including residue of catalysts.32 1 3.1 Density As mentioned previously, t he density is the chi ef physical property used in industry to classify the polyethylene grades. Governed by the polymerization process, density is strongly related to the type and concentration of branches. HDPE produced via low pressure process has negligible numbers of branc hes. Hence, the alltrans ethylene sequence has no difficulty pack ing into large crystalline structures. Moreover the crystalline region is denser than the noncrystalline region: the former has a density of 1
30 g/cm3, while the latter of 0.85 g/cm3.4 D ensity can then be correlated to branc h content and degree of crystallinity, as seen in Figure 22. Two types of PEs are sampled: LLDPE s with weight averaged molecular weight s in the range of 65,000130,000 g/mol, and HDPE resin with weight averaged molecular weight of 61,000 g/mol. From Figur e 2 2, two trends can be observed: Density and crystallinity decrease as the branch concentration increases; the slow ly cooled polymers obtain higher density compared to quenched materials indicating an influence from the thermal process. The molecular we ight and its distribution also play an important role in the determination of the density of PE. In general, density drops in some extent with the large increase of molecular weight. Figure 22. Plot of density and degree of crystallinity as functions of branch content for compressionmolded HDPE and LLDPE resins ( F igure adopted from ref. 2)
31 2 1 3.2 Crystallization and m orphology T he physical properties of PE, such as density, are determined by the semi crystalline nature. The presence of three phases crystalline, interphase, and amorphous phase offers unique physical properties to PE and to other semi crystalline polyolefins. Such simple evidence lies in the fact that PE is both tough and resilient: the former property origin ate s from the ordered packing in the crystalline region, while the latter comes from the highly disordered amorphous region. Crystallinity is, in turn, determined by the polymerization process, catalyst, temperature, pressure, cooling rate, etc. The crystallinity level, usually termed as degree of crystallinity ( ) is used to characterize the percentage of crystal structure in the whole material. In fact, the degree of crystallinity is intrinsically controlled by the microstructure, i.e., t he branching in PE. Figure 22 clearly shows the relationship between crystallinity level and branch content In general, the degree of crystallinity drops as branch content arises, particularly at a low level of branching. This observation is not difficul t to understand. The introduction of branches in PE hinders the main chain from pack ing into an ordered crystalline array in terms of both size and perfection. It is commonly believed that the branches are preferentially excluded from the crystalline phase, and the ability of commercial PE to crystallize largely depends on the distribution of alltrans ethylene segments .2, 810 Th erefore, the size of the crystal structure of PE is constrained by the branch spacer, or in other words, the defect free ethylene sequence between the adjacent branches. The higher the branch content, the shorter the available main chain that can crystallize, consequently, the smaller the crystal structure, the lower the crystallinity level.
32 The relative arrangement of the crystalline phase and noncrystalline phase, their portions, sizes, shapes spatial orientations, and connectivity modes, determine the physical properties of PE as a universal material. I n crystallography, any crystal structure is built up from many periodic repeating unit cell s, which are the smallest structural motif s, much like monomers are the construction unit s for polymers. A specific unit cell can be determined by six parameters: lengths of three axes and angles between phases seen in Figure 23 (a). (a) (b) (c) (d) (e) Figure 23. Typical unit cell structures and their characteristic dimensional values: (a) General unit cell; (b) Orthorhombic unit cell; (c) Monoclinic unit cell; (d) Triclinic unit cell; (e) Hexagonal unit cell There are three type s of unit cells present in commercial polyethylene: orthorhombic, monoclinic, and hexagonal crystals as shown in Figure 23 (b), (c) and (e). The polymer chain arrangement and the quantitative dimensional data of the three PE crystals are illustrated in Figure 24 (a), (b), and (c), respectively. The o rthorhombic unit cell is a cuboid with three unequal axes and three plane angles being right angle, seen in Figure 23 (b). With the pioneered contributions from C. W. Bunn11 and V. Vand12 c a b the crystal structure of commercial PE has been accepted = = = 90 = = 90 120 90 = = = 90 = 120
33 to be orthorhombic. Figure 24 (a) gives the chain orientation and crystal dimensions of this structure measured for commercial high density polyethylene at r oom temperature.2 It is worth noticing that the structure shown herein is an idealistic case without consideration of illdefined branching. When structural defects are introduced in to the polymers, low density polyethyl ene and linear low density polyethylene display increased and axis values. Interestingly, the length with covalent main chain bonds, ( i.e., axis dimension) shows less dependence on branching.13, 14The monoclinic and hexagonal unit cells are metastable crystal forms for PE. Upon stretching, 15 pressing,16 heating, or inject molding processing,17 the metastable monoclinic crystal can be obtained by a transition from the more stable orthorhombic crystal. And when the defor mation conditions are removed, the equilibrium orthorhombic crystal form will return. Originated from high pressure and high temperature crystallization, the hexagonal crystal form of PE is sometimes referred to as the rotator phase, in that some individual chain segments are capable of rotational motion at random phase angles relative to their adjacent stems.18, 19 The configuration of polyethylene chains and the sizes of metastable crystal forms are displayed in Figure 24 (b) and (c).
34 Figure 24 Three unit cell structures in commercial PE and their characteristic dimension al values: (a) Orthorhombic unit cell; ( b Monoclinic unit cell; and ( c) Hexagonal unit cell ( figures adapted from reference2 ) (a) (b) (c)
35 2 1 3.3 Thermal p roperties As a semi crystalline polymer, PE displays characteristic thermal properties with melting occurring within a range of temperatures associated with the crystalline structure and glass transition behaviour associated with the amorphous phase. Figure 25 Plot of heat flow of three types of PE samples as function of tem perature/heating rates. (a) Slow cooled high density polyethylene of moderate molecular weight; (b) quenchcooled high density polyethylene of moderate molecular weight; (c) quenchcooled linear low density polyethylene of moderate molecular weight; (d) quenchcooled low density polyethylene ( F igure adopted from ref. 2) Polymeric material s consist of numerous unequal polymer chains, which have different chain lengths and branching from each other, causing a distribution of
36 crystalline structure in terms of size and orientation. Melting is a thermal transition in which atoms on segmental chains in the crystalline region vibrate so violently that they leave the equilibrium position within the crystal lattice resulting in collapse of the crystal structure. Thi s melting process is accomplished by absorbing energy, normally by thermal heating. The m elting temperature is related to the thickness of lamella e formed in the crystalline phase of PE: the thicker the lamella e the higher the melting temperature. The dis persed lamellar thickness due to the random displacement of branches along the backbone causes the dispersion in melting temperature, and consequently the breadth of melting range. This melting range for commercial PE can span from less than 10 C up to 70 C when undergoing the transition from semi crystalline solid state to molten state. Since crystallization is thermal process dependent t he melting range of the crystalline structure is also related to the thermal history. Figure 25 illustrates the melting behaviours of two type PE samples undergoing various thermal treatments. It is clear that the slowly cooled and quenched HDPE s show significantly different melting ranges. The slowly cool ed HDPE exhibits a sharper melting peak and narrower melting range, as a consequence of thicker lamella e and narrower distribution due to more perfect organization of chain packing in crystalline regions. Also, when other param eters, such as molecular weight remain constant, the melting is a function of branching. From (b), (c), and (d) in Figure 25, all three PE samples (HDPE, LLDPE, LDPE) possess various branch content: HDPE < LLDPE < LDPE. Their melting peaks when all quench ed from the molten state show the reverse order : HDPE > LLDPE > LDPE.
37 In addition, semi crystalline polymers including PE undergo several secondary thermal transitions arising from the localized order of structure. The glass tra nsition is such a second ord er process. Thermoplastic samples are extremely rigid and brittle below a critical temperature, termed as glass transition temperature Tg. It is commonly accepted that below Tg, the polymer chains in the highly disordered regions hav e little or no freedom to move and the polymer show s glasslike behavior Above the Tg, by absorbing more energy, those chains in the amorphous phase have greater freedom to move to certain extent and polymer behaviour is rubber like. The transition from t he glass like state to the rubber like state is referred to as glass transition or transition, which is associated with the noncrystalline components in semi crystalline polymers. Commercial PE samples have Tg2 1 3.4 Other physical p roperties ranging from 130 C to 100 C, a rang e that is well below room temperature and thus facilitates the universal applications of PE. Similar to the melting process, the glass transition is also dependent on processing history and measuring techniques/routes. Other physical properties, especially mechanical, rheological, and optical, are essentially correlated to the branching in PE.2, 9, 11, 2024 For instance, regarding the mechanical behaviour, the crystal line regions in PE show modul i two orders of magnitude greater than those of the noncrystalline regions,2The rheological properties of molten PE are crucial in processing the final merchandise In the molten state, PE is rather viscous and considered as a viscoelastic liquid. The viscosity and elasticity of polyethylene materials are strongly correlated to indicating a strong relationship between the mechanical strength and the degree of crystallinity, which, as demonstrated previously, is determined by the type and number of branches.
38 the chain entanglement in both molten and solid states i n turn a function of molecular weight, its distribution, and the nature of branching of PE resins, with detailed mechanical data shown in Table 21. So far, combining the essence of polyethylene and its manufactur e a logic al relationship of structureproperty performance can be established, illustrated in Figure 2 6 The microscopic properties, such as branching, molecular weight and i ts distribution, chain packing, and chain dynamics originat s from the industrial production or lab synthesis. Moreover, the macroscopic behaviors, like morphological, thermal, mechanical, morphological, and rheological properties of polymeric materials ar e dependent on the intrinsic microstructure, i.e., the nature of branching. Guided by the macroscopic properties, it is possible to select polymers for various applications. Figure 26 Relationship of structureproperty performance of polymeric materi als Studying carefully the relationship chain in Figure 26 one has no difficult y to find that microscopic properties are key to linking synthesis/man ufacturing with polymer behaviors. Any fundamental understanding of branching effects will be of great importance to guide manufactur ing and to predict the material s applications. A great deal of research has been perform ed on the subject. However, due to the nature of randomness for commercial polyethylene, specifically, the random sizes of branches Synthesis Microscopic Properties : Branching Chain Packing Chain Dynamics Macroscopic Properties : Thermal Mechanical Morphological Rheological Applications
39 the random distribution of branches along the backbone, and the random spac ing between two adjacent branches, the fundamental relationship between branching and the macroproperties is not yet clear. Therefore, there is an obvious need for model polyethylenes. The basic requirement of t hese model polyethylenes is to limit the randomness, i.e., to study the influence of one parameter by fixing some other varia bles in commercial PE. This can be realized via a family of precision polyolefins to be discussed in detail in the following section. 2 2 Precision Polyolefin s 2 2 .1 Structure of Precision Polyolefins T he nature of branching in commercial PE samples is random, i.e., the type, number, and frequency of the branches are not defined. To better understand the influence of microstructure on macroscopic properties precision PEs are introduced as structural models which reduce the number of variables in the PE system P recision polyolefins are defined as polyolefins in which certain featural structures are defined. T he primary structure s of these polymers can be controlled by one type of metathesis chemistry, specifically, acyclic diene metathesis (ADMET) polycondensat ion chemistry The basic structure of these model polymers is illustrated in Figure 27 Compared to commercial structures, the pendant side groups R are displaced regularly on the polyethylene backbone. The branch identity R and the branch spac er x are de fined precisely Figure 27 General c hemical structure of precision polyolefins. Branch identity R is displaced regularly along the pol yethylene backbone. Branch spacing the distance between the adjacent branches is defined.
40 The branch identity R can be any functional group that is compatible with ADMET chemistry. So far, the advances in this area have shown that branches R can be alkyl groups2537, siloxane units3842, halide groups4348, biological active groups4954, acid groups5560, conjugated functional groups, and even some ionic groups61. The branch spac ing denoting the number of carbon atoms on the main chain between two adjacent branches, can be an integer tuned from 2 to 3 8 according to the synthetic efforts to date.362 2 .2 Synthesis of Acyclic Diene Metathesis (ADMET) Precision Polyme rs The synthesis of precision polyolefins can be realized via acyclic diene metathesis (ADMET), which is a type of step growth olefin metathesis capable of preventing short chain branching distribution from chain transfer/walking which occurs in chain propagation chemistry. The soul of making regularly branched precision polyolefins lies in the synthesis of diene monomers. It is the nature of the monomers that brings the precision to the polymer structures. 2 2 .2 .1 Monomer s ynthesis There are two methodologies present i n the synthesi s of the symmetric diene monomers, shown in F igure 28 and F igure 2 9 respective ly. Figure 28. Precision monomer synthesis methodology starting with malonate
41 The monomer synthetic route #1, exampled by a methyl branched monomer, is illustrated in F igure 28 Briefly, the starting diethylmalonate 1 can be easily disubstitut ed with an alkenyl halide possessing the appropriate methylene spacing. The resulting diester 2 is saponified followed by decarboxylation to produce the pure mon oacid 4 which is then selectively reduced with lithium aluminum hydride (LAH) to yield the primary alcohol 5 The symmetric diene monomer of interest 7 can be obtained with mesylation of the primary alcohol and the subsequent reduction via hydride displacement. The same methodology has been successfully applied to synthesize ethyl branched31 and hexyl branched34 ADMET precisio n polyolefins. Figure 29 Precision monomer synthesis methodology via cyanide chemistry An optional monomer synthesis route (Figure 2 9) dialkylation of primary nitrile 8 in the presence of base such as lithium diisopropylamide, followed by cyanodiolefin 9 via radical reaction diene monomer 10. This newly designed cyanide methodology offers quantitative yield with short synthetic steps,27 and has been proved to be well suited fo r making alkyl branched symmetric diene monomer s. T o date, the alkyl branches vary from small groups, like the methyl group, to larger ones like the hexyl group, and even to the adamantyl group.29, 37 8 9 10
42 2 2 .2 .2 Polymer s ynthesis Once the symmetric diene monomer of interest is achieved, the polymerization can be readily realized via acyclic diene metathesis polycondensation chemistry, shown in Figure 2 10. Figure 2 10. Acyclic diene metathesis polymerization scheme in general. Figure 2 11. Meta thesis and hydrogenation Catalysts: (a) Shrocks molybdenum catalyst; (b) 1st generation Grubbs; (c) 2nd generation Grubbs catalyst; (d) Wilkinson catalyst (a) ( b ) ( c ) ( d )
43 Pure, moisturefree monomers are generally polymerized in bulk under mild step polycondensation conditions with the help of organometallic catalyst s, such as Shrocks [ Mo ] catalyst or 1st and 2nd generation Grubbs [ Ru ] catalysts. Considering that the 2nd generation Grubbs catalyst can cause isomerization and correspondingly disrupt s the precision o f ADMET polymers,62, 63 Shrocks and 1stThe polymerization is conduct ed under appropriate vacuum to protect the moisture and air s ensitive catalysts. Another prupose of vacuum is to remove gaseous ethylene from the reaction system and pushing the equilibrium to the right side, i.e., the polymer side, in order to accomplish the polymerization. generation Grubbs catalyst are much more widely used in ADMET chemistry. The structures of the catalysts are shown in Figure 211. The precise alkyl branched ADMET polymers can be readily produced by hydrogenation of the unsaturated products in Figure 210 with exhaustive gaseous hydrogen at relatively high pressure and temperature, as illustrated in Figure 21 2 Figure 2 12. Hydrogenation of unsaturated ADMET precision polymer s ADMET is such a useful but simple chemistry. It provides a unique way to transfer the symmetry in the diene monomer s to the unsaturated and eventually the saturated ADMET precision polymers. However, due to the nature of ADMET poly merization, the precisely branched ADMET polymers possess controlled primary structure, but undetermined stereochemistry ; the orientation of branches in principle can
44 be at any direction with respect to the ethylene backbone. Moreover, constrained by the e ssence of step condensation, the molecular weight s ( ) of these polymers is normally less than 100,000 g/mol, which are not comparable with that of commercial polyethylenes produced via chain propagation chemistry. Purely linear polyethylene can be treated as a perfect model for commercial polyethylene, as well as for ADMET precision polyolefins because of its non branched structure. The synthesis of linear ADMET PE is quite simple: ADMET polymerization of 1,9 decadiene followed by hydrogenation, s hown in Figure 2 13. Figure 2 1 3 Laboratory synthesis of linear ADMET PE. Step 1: ADMET polymerization of 1,9 decadiene; step 2: hydrogenation of unsaturated linear PE 2 2 3 Synthesis of ADMET Random Polymers To model the mechanized random polyethylene structures, the need to study the effects of randomness of branch frequency is fulfilled by the socalled ADMET random polymers, which have controlled branch identity but random distribution of branches along the main chain. These ADMET random polymers with irregular branch spacing can be easily yielded by copolymerization of two symmetric diene monomer s possessing various branch spac ing,31, 32, 37 se en in Figure 21 4 By carefully controlling the ratio of the two monomers, one can calculate the average branch concentration. Or with the required branch concentration in mind, one can easily deduce the mixing ratio of the two starting monomers. Based on this, pairs of precision and
45 random ADMET polyolefins have been successfully synthesized in order to compare their macroscopic behaviors.32, 37 Figure 2 1 4 Synthesis of ADMET random polyolefins with irregular branch spac ing Of course, to model the effect of randomness of branch identity, another type of ADMET random polyolefin can be similarly achieved by copolymerizing two diene monomers with various branch identities followed by hydrogenation. The scheme of synthesis is i llustrated in Figure 21 5 So far, this type of ADMET random polymer has generated little attention and is included for completion of randomness influence. Recalling that low density polyethylene contain s both short chain and long chain branches t he 2nd type of ADMET random polymer s will be suitable structural model s for understanding the effects of unequal branch identity in LDPE. Figure 21 5 Synthesis of ADMET random polyolefins with unequal branch identities 2 3 Co mparison of Commercial Polyethylene ( PE) to ADMET PE As mentioned, different types of commercial PE, whether Ziegler Natta catalyzed or metallocene catalyzed, are produced via conventional chain propagation chemistry. In contrast, ADMET polymers are based on step polycondensatio n chemistry. The ir
46 t otally different polymerization mechanisms result in microstructures with both similarities and variations: the same poly ethylene backbone but different branching. 2 3 1 Structure Modeling It is well established that the type, size, con tent, and distribution of branches significantly affect the physical properties of PE based materials. Although numerous researches have been undertaken on the subject, the influence of branching is still unclear due to the complexity of macromolecules. This in turn con strains the manipulation of production process es and understanding of the end products. Among the unclear fields in PE chemistry, some questions generate particular interest in this dissertation work One is the location of branches. It is commonly accepted that the sections of polyethylene main chains without defects maintain higher priority to crystallize than the sections with defects.8, 9, 22, 48, 6466 Thermodynamically, small branches like the methyl groups are capable of being incorporated into the crystalline regions.2, 8, 21The influence of branch content forms another interesting subject As illustrated in Figure 22, branch content influence s the density and degree of crystalline for PE resins. Also, the mechanical properties, such as ductility are dependent directly on the A s for larger branches, polymer scientists tend to believe that larger groups are present only in the noncrystalline regions. The question then arises. How large can the branches be and still be incorporated into the unit cells? LLDPE differs from LDPE in that the former has only short chain branches while the latter has both short chain and long chain branches. These materials exhibit essentially distinguished physical properties. There fore, unveiling the location of branches in various phases could be very helpful for understanding their effects on the macroscopic properties.
47 branch concentration. P olyethylene sequences with no or small defects/branches tend to pack into ordered struc tures Since the polymer chains are rigid in terms of folding, what is the shortest branch spac ing needed to maintain the alltrans polyethylene structure in crystalline regions? To answer the two questions mentioned above, a typical experimental design will be addressed to vary the parameter of interest while fixing the rest of the variables. For instance, to determine how large the branches can be when incorporated in to the crystalline domains, one could f ix the branch content and branch distribution and vary only the size of the branches. However, this is not feasible for merchandized polymeric material s. The advent of ADMET chemistry offer s such required structure models through determined primary structures: the branch identity and branch spac ing can be controlled, according to the synthetic scheme of ADMET polymers in Figure 27 Therefore ADMET polyolefins can serve as excellent models for commercial PE based structures. If the variables in PE based str uctures can be treated as freedoms of a spot in threedimensional (3D) space, the relationship of ADMET precision polymers, ADMET random polymers, and commercial PE can be interpreted as cartoons shown in Figure 2 1 6 Three variables in terms of branching are present in commercial polyethylene structures: the branch identity, branch spac ing, and the heterogeneity of branch distribution. The three axes + + + indicate the three variables respectively Since none of the branching variables is determined for any PEbased commercial material, the merchandized structure therefore has three types of freedom. Such a structure can
48 be cartooned as mixture of spot s with various coordinate ( ) representing (branch distribution branch spacer branch iden tity ) illustrated as the red cuboid in Figure 216 (c). Figure 21 6 : Correlation of ADMET PEs to commercial PE: (a) ADMET precision polyolefin, (b) ADMET random polyolefin, and (c) commercial PE As far as ADMET precision polyolefins are concerned, the branching freedom is actually zerofor a specific ADMET precision polyethylene structure, the branch spacer ( zaxis), branch distribution ( x axis), and branch identity ( yaxis) are fixed once the pol ymer is prepared. Therefore, ADMET precision polyethylene can be represented by one defined spot in the coordinate system, illustrated as a red spot in Figure 216 (a). The case for ADMET random polyethylene falls in middle of above two extreme conditions: it encompasses two branching variables (branch spacer and branch regularity) and one fixed branching parameter (branch identity), if the ADMET random polyolefins in Figure 21 5 are sampled. Therefore, the structure of a specific ADMET random polyethylene can be cartooned as mixture of spots on a horizontal surface ( ) (a) (b) (c) Branch identity Branch spacer Branch identity Branch spacer Branch identity Branch spacer
49 with fixed axis (branch identity is fixed), illustrated as a red rectangular surface in Figure 216 (b). The above structure modelling cartoons allow PE structures to be pictured as fo llows : from point (ADMET precision PE) to surface (ADMET random PE) to 3D structure (commercial PE). Therefore, one can declare confidently that ADMET polymers provide beautiful structur al models for the commercial PE based materials by allowing investigat ion of the branching variables separately. This understanding of ADMET polymers, both precision and random, will in no doubt decipher the complex structure property performance for the PE based commercial materials. 2 3 2 Comparison of Physical Properties ADMET polymers exhibit very different physical behaviours from the commercial PEs due to the introduction of precision in to the structure. The thermal behaviours of these two families of PE are compared in Figure 217. The peak melting temperatures of comm ercial PEs (HDPE, LDPE, LLDPE, and VLEPE) are shown on the top part. The peak s observed for several series of ADMET precision and random polyolefins sampled with methyl branched and butyl branched polymers are shown below the commercial ones. It is cl ear that by varying branch identity or spac er the s for A DMET polymers can span around 200 C Such a wide range not only cover s the melting range for commercial PEs but also extending the to even lower temperatures. The PEX Methyl a group of ADMET precision polyethylenes with methyl branches regularly displaced on every Xth carbon along the backbone, with branch spacer X ranging from 7 to 39 (corresponding to a branch content of 143 to 25 methyl
50 groups / 1000 backbone carbons), ex hibits peak melting points ranging from 60 C to 92 C, spanning around 1 5 0 C.26, 28 In contrast, the group of random versions RPE Methyl denotes a series of ADMET random polyethylenes with methyl branches irregul arly displaced on the backbone. Calculated from the ratio of the starting diene monomers (refer to Figure 214), the average branch spacer which defines the average position of the branch on each and every backbone carbon, can range from 17 to 666 (corresponding to an average branch content of 55.6 to 1.5 per 1000 bac kbone carbons ) .32 -70 -60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 R PE-X-Butyl PE-X-Butyl R PE-X-Methyl PE-X-MethylVLDPE LLDPE LDPE Melting temperature, oCHDPE The RPE Methyl polymers display peak melting points range from 50 C to 130 C. Figure 21 7 Peak melting temperatures of commercial PE, ADMET precision and random PE s. PE X Methyl : a group of ADMET precision PEs with methyl branches precisely displaced on every Xth carbon along the backbone, branch spacer X ranging from 7 to 39. RPE Methyl : a series of ADMET random PEs with methyl branches irregularly displaced on the backbone, average branch spacer ranging from 17 to 666 PEX Butyl : a group of ADMET precision PE with butyl groups, branch spacer X in the range of 15 to 39. RPE Butyl : a series of ADMET random PEs with butyl groups irregular ly pendant on the backbone, average branch spacer ranging from 23 to 400
51 As for the butyl branched ADMET polymers, PEX Butyl and RPE X Butyl represent the precision and random versions, separately. The peak melting temperatures for PEX Butyl span from 33 C to 75 C, when branch spacer X rang es from to 15 to 39 ( corresponding to branch concentrations ranging from 67 to 25 per 1000 backbone carbons ) The random polymers RPE Butyl obtain peak melting temperatures span from 10 C to 126 C, when average branch spacer varies from 23 to 400 (corresponding to bra nch content s ranges from 43.5 to 2.5 per 1000 backbone carbons ) Noticeably, the melting points of RPE Methyl when = 666 ( = 129 ) and RPE Butyl when = 400 ( = 126 ) are rather close to the melting temperature of HDPE ( = 125 ~ 132 ) indicating the extreme level of perturbation caused by the branching Moreover, unbranched ADMET PE has a peak melting temperature of 133 C, identical to that of HDPE. However, the melting curve s of ADMET polymers observed via differential scanning calori metry (DSC) are relatively sharper than those of the commercial PEs, implying a much narrower distribution of lamellar structures.28, 29, 48Wide angle X ray scattering measurement s of unbranched ADMET PE has identified the crystal unit cell to be orthorhombic and provided the quantitative unit cell data 29, 48 as illustrated in Figure 218. It is fo und that the dimensions are rather close to those of the orthorhombic HDPE, seen in Table 2 2. Recalling that the melting point of unbranched ADMET PE is equal to that of HDPE o ne can readily conclude that the chain packing behavior and morphological property of ADMET PE are similar to those of commercial HDPE, due to the complete lack of branching in both types of structures.
52 Figure 21 8 Orthorhombic crystal structure of linear ADMET PE: a = 7.48 b = 4.98 c = 2.55 Top is the orthogonal view, bottom is the view along the axis. Table 2 2. Unit cell dimensions of HDPE and ADMET PE Polymer Unit Cell a, b, c, HDPE Orthorhombic 7.42 4.95 2.55 HDPE Monoclinic 8.09 4.79 2.55 HDPE Hexagonal 8.42 4.56 <2.55 ADMET PE Orthorhombic 7.48 4.98 2.55 a= 7 48 c= 2 55 a= 7.48 b= 4 98
53 In short, ADMET polymer s can function as i deal structural model s for commercial polyethylenebased materials by avoiding the complications arising from the collection of organizational structure s at both the microand macromolecular scales. Upon introducing heterogeneity into the simpl e ethylene sequence, the carbon backbone in 3D structure is able to adopt various orientations and motions. It can fold, vibrate, twist, rotate, or entangle. All these morphological related behaviors can to be deciphered using ADMET chemistry and suitable physical measurements.
54 CHAPTER 3 SOLID STATE NMR SPEC TROSCOPY AND X RAY SCATTERING TECHNIQUES 3 1 Polymer Characterization Methods Polymer characterization techniques have been developed for decades and are quite established.2, 21, 6772 With the development of modern analytical instrumentations, materials structural, dynamic, morphological, thermal, rheological, optical, electric al mechanical information, etc. are able to be identified and acquired.4, 68, 7379Among them, solution nuclear magnetic resonance (NMR) spectroscopy, infrared (IR) spectroscopy, and mass spectrometry (MS) are widely used to determine the compositional structure of polymeric samples. Gel permeation chromatography (GPC), also termed as size exclusion chromatography (SEC) is used to achieve molecular weight and polydispersity (PDI). Transmission electron microscopy (TEM) and atomic force microscopy (AFM) are mainly used to obtain lamellar thickness of the semi crystalline samples. As for the thermal analysis, differential scanning calorimetry (DSC) is the major tool. Frequently used characterization methods used in this dissertation are listed in Table 31. Solid state NMR and X ray scattering techniques are famous for their powerful characterization abilities and diverse applications for condensed matter.76, 77 Since we are interested in modeling the commercial polyethylenebased materials, the physical propert ies of precision polyolefins at solid state are of particular interested. Th erefore, SSNMR and XRD have been selected specifically in this dissertation as two major analytical tools in order to investigate structural, dynamical, and morphological properties of ADMET polymers at condensed state. In the following sections, they will be reviewed in detail
55 Table 31. Polymer characterization methods used in this dissertation Method Principle Sample Information Limitation GPC Random coil dimensions of polymer chains are in correlation with their molecular weight solution state molecular weight and it s distribution sample in solution; quantitative information referenced to standard with known M w MS Samples are ionized in the gas phase and the ion abundance is measured as a function of mass to charge ratio of the ions. solution or solid state composition; primary structure; endgroup; molar mass complex spectra; sample to be relatively stable under laser irradiation NMR Nuclear spins interact with the external magnetic field. solution molten or solid sample structure; end group; molecular weight; stereoregularity ; dynamics; chain alignment Only non zero nuclear spins are active. IR The frequency of the absorbed radiation matches the vibration frequency of the bond or group. thin film or solid sample s tructure ; functional groups ; additives sample to be infraredactive TEM Electrons are transmitted through ultra thin sample and form images by interacting with the sample atoms. ultra thin film morphology sample preparation; data acquired by averaging several characteristic areas in the image AFM A cantilever probe interacts with the sample surface and the deviation of the cantilever is monitored versus the distance between the probe and the surface. solid thin film; in solution topographic images; force data acquired by averaging several characteristic areas in the image; slow scanning speed DSC Heat flux to the sample is monitored against time or temperature. solid sample thermal behaviors: Tm, H, T c T g C p low sensitivity compared to DMTA XRD X radiation interacts with sample atoms and scattered signals are monitored. solid sample crystalline structure not sensitive to noncrystalline structure
56 3 2 Solid State NMR Spectroscopy 3 2 1 Introduction to NMR Just as Nobel Prize winner Dr. Ernst commented on the Nobel Lecture80, Nuclear spin systems possess unique properties that predestine them for studies of molecules, it has demonstrated that modern NMR spectroscopy is well suited in studying polymer structures in solution, molten, and solid state. The principle of structur e determination by NMR lies in the fact that a nuclear spins response in the magnetic field is due to the intrinsic properties of nuclear spins and electron spins. The nuclei with nonzero spin not only interact with the applied magnetic field, but also w ith the chemical environments they encounter. Since the density of the surrounding electron cloud varies with the chemical environment of the material, the degree of shielding of the nucleus of interest is thus dependent on its chemical structure. This off ers the fundamentals of NMR spectroscopy as an excellent tool in understanding structures of polymeric materials. Some parameters of an NMR spectrum are used widely, such as chemical shift and chemical shift anisotropy, peak area, coupling constants, and r elaxation times.20, 57, 81 Unitizing NMR techniques, one can not only confirm the molecular structure of interest, but also exact quite a lot useful infor mation including molecular weight, endgroup structure, morphology, bond distance, exchanging rate, relative bond alignment, chain packing patterns, interphase, branching, crosslink, isomerization, and segmental dynamics.67, 77, 79, 80, 82Benefited from the development of quantum mechanics, n uclear magnetic resonance (NMR) was first predicted and measured by I sidor Isaac Rabi in 193 7 82 and in 1944 Rabi won the Noble Prize i n Physics for his discovery of NMR. The phenomenon of nuclear magnetic resonance in bulk material w ere first reported by
57 Edward Purcell83 and Felix Block84 independently who shared the Noble Prize in Physics in 1952 for their development of new ways and methods for nuclear magnetic precision measurements." On the way to modern NMR spectroscopy there are many major breakthroughs contributed, including t he introduction of pulsed Fourier transform (FT) techniques ,85, 86 the discovery of magic angle spinning (MAS) and cross polarization, the prediction87The basic concepts and applications of solid state NMR techniques on macromolecules have been studied and covered in several classic NMR book s and realization of Overhauser Effect the emergence of high magne tic field, and the development of modern multi pulse sequences. 77, 79, 82, 8890 and review papers91, 923 2 1. 1 Resonance and c hemical s hift Afterwards basic concepts, theories, and pulse sequences which are closely related or have been employed in the dissertation herein will be discussed. In quantum mechanics, each nucleus spins, like the electrons do. Unpaired particles, such as protons, neutrons, and electrons possess a spin of The total nuclear spin, denoted as I is the sum of proton spins and neutron spins. For instance, with one un paired proton and one unpaired electron, 1H has the total nuclear spin number of and the total electron spin number However, 2H has a total nuclear spin number I of 1 arising from one unpaired proton and one unpaired neutron. In NMR, it is the unpair ed spins matter the atoms with unpaired spins have signals in the magnetic field. For instance, 13C is NMRactive while 12C is not. Besides 1H 2H and 13C, 15N, 29Si, 31P, 19F, 35Cl, 37Cl are NMRactive nuclei and are widely used for structural and dynamic investigations.
58 When an atom with non zero nuclear spins is placed in a magnetic field B00= 0 (3 1) the spin vectors of the particle align themselves to the direction of the external magnetic field, which is typically the + z direction. In modern NMR spectroscopy upon absorbing energy from a pulse with appropriate radio frequency (RF), a spin is capable of jumping to a higher energy level; when the R F signal is switched off, the spin relaxes back to its lower energy state ( a process termed as procession in NMR), and energy is emitted simultaneously. This process is called nuclear magnetic resonance. The signal s in NMR spectroscopy result from the energy difference between the absorption and the emission; this energy difference is proportional to the spin population difference between the high and low energy levels, which are normally termed as spin up and spin down for a spin of quantum number Therefore, NMR obtains its sensitivity from the resonance, or exchange of energy at a suitable frequency referred to as Lar m o r frequency 0, which is determin ed by the nature of nuclear spin and the strength of the magnetic field obeying the following equation: where is the magnetogyric ratio (often termed as gyromagnetic ratio), a constant for each nucle us; 0 is the magnetic field at the site of the particle. In this dissertation, three nuclei are focused : 1H, 2H, and 13Table 32. List of three nuclei investigated in this dissertation. C. Their spin numbers, natural abundance, and values are shown in Table 32. Nucleus Ground state s pin Energy levels Natural abundance Gyromangnetic ratio rad s 1 T 1 1 H 1/2 2 99.9885% 267.5552*10 6 2 H 1 3 0.115% 41.066*10 6 13 C 1/2 2 1.07% 67.283*10 6
59 A nucleus encounters not only the external electromagnetic field but also the local ized magnetic fields induced by the surrounding electron cloud s, illustrated in Figure 31 The change of the local magnetic field results in a change of resonance frequency termed as chemical shift The sum of applied external field and the induced field generated by the electrons is the local magnetic field that a certain nucleus feels: = 0 (3 2) Figure 31. Schematic illustration of chemical shift mechanism B0 indicates the external magnetic field; Bind i s the induced magnetic field due to the surrounding electrons in the applied field; Bloc is the local ized magnetic field of the nucleus by subtracting/addition of Bind from B0 The inner circle illustrates the nuclear spin while the outside yellow circle is the surroun ding electron cloud Although the induced field is only around 104 of the applied field 0,88 = (0 0 0 ) (3 3) it is large enough to give rise to measurable chemical shifts. Since the induced electronic field is directly proportional to the strength of the external magnetic field, the local resonance frequency is rel ated to 0 as well. However, chemical shift is independent on 0, based on its definition: where 0is the Larmor frequency of the same nucleus in a reference compound (f.g. tetramethylsilane) exposed to the same applied field. Chemical shift is dimensionless. B ind B 0 B loc
60 The induced filed, has a linear relationship with the applied filed 0: = 0 (3 4) where is a 3x3 matrix, termed as chemical shift shielding tensor of nucleus site The matrix vector form of Equation 34 can be written as : = 0 0 0 = 0 0 0 (3 5) where the applied static field 0 is assumed to be along the z axis of the laboratory frame. The components of induced filed in the x, y, and z direction are represented by 0, 0,and 0, respectively. Equation 35 explains the fact that the induced filed is usually in a different direction to the applied filed. Therefore, the chemical shift at a given nuclear site depends on the molecular orientation with respect to the external field, as well as the location of the nuclear spin within the molecule. 3 1 1.2 Chemical s hift a nisotropy Based on the mechanism of chemical shift, the full form Hamiltonian of chemical shift of spin can be represented as: = = 0 = ( ) 0 ( ) 0 ( ) 0 (3 6) where the symbol is molecular orientation used to emphasize the spatial orientationdependent of the chemical shift; and are operators for the three components of nuclear spin angular momentum at x, y, and z directions Since applied field 0 is much stronger than the chemical shift interaction, Equation 36 can be simplified by applying the secular approximation:
61 ( ) 0 (3 7) where is the experimental measurable shielding tensor when the applied filed pints to z direction. Its quantity can be obtained from the following equation79 = +1 2 ( 3 2 1 + 2 22 ) (3 8) : where is the isotropic chemical shift 1 2 ( 3 2 1 + 2 22 ) is the anisotropic part with shielding anisotropy and asymmetric parameter and are the Euler angles 93 as illustrated in Figure 32. Figure 32. The definition of tensor orientation in different coordinate systems. LAB stands for laboratory frame. PAF stands for principal axis frame. is the angle between the z axis of PAF and external field 0. is the angle between the XPAF and the projection of 0 in the xy plane of PAF. The three princip al values and are defined as follows: ZLABZPAFXLABYLABXPAFYPAFzzxxyyB0
62 =1 3 + + (3 9) = (3 10) = / (3 11) where and are tensor components in the princip al axis frame (PAF), as shown in Figure 32. Since the chemical shift interaction is related to the spatial orientation with respect to the applied magnetic field, the introduction of PAF largely simplifies the problem by offering a diagonal tensor. The spectral lineshapes of semi crystalline polymers are typically called powder pattern because of the random o rientation of chemical microstr u ctures. CSA patterns are characteristic of nuclear spins spatial orientation and its chemical environment and are sensitive to segmental motions Figure 33 shows three special CSA lineshapes: (a) The three shielding components are different from each other, However, the asymmetry parameter = 1 corresponding to static motion or hop of nuclear spin between two spatial orientations. (b) When a chemical shift tensor has axially symmetry, the CSA lineshape can be signified as Figure 33 (b). For this case, and refer to the shielding along and perpendicular to the principle z axis, where = = = and = 1 This unique CSA spectrum can be assigned to an axial motion of the nuclear spin at site. (c) When the three components of chemical shift tensor are identical, = = the anisotropy part disappears and the spectrum
63 collapses to an pure isotropic lineshape with one middle peak left at frequency. This CSA pattern represents the isotropic motion of the nuclear spin at site. (a) (b) (c) Figure 33 Special CSA lineshapes : (a) static or 180 hopping; (b) axial rotation; and (c) isotropic motion 3 1 1. 2 NMR i nteractions In NMR, the total Hamiltonian can be expressed as following: = + + + + + (3 1 2 ) where Zeeman interaction and the interaction with radio frequency (RF) pulses belong to external interactions. Indirect spinspin coupling ( Jcoupling) dipole dipole interaction chemical shift and quadrupole interaction can be classified as internal interactions. describes the Hamiltonian of an isolated spin in the static, uniform magnetic field, and is much stronger than other interactions when a high magnetic field is applied. zzxxyyxx: ~50 ppm yy zz: ~34 ppm : ~15 ppm 6 0 4 0 2 0 0 p p m 60 40 20 0 ppm 6 0 4 0 2 0 0 p p m
64 It is that causes the energy splitting into 2 + 1 energy levels. Arising from the timedependent RF electromagnetic radiation is unitized to manipulate the internal spin interactions in NMR. Based on this, NMR is capable of provide a wealth of information about the structures, dynamics, interactions, etc. Representing the indirect interaction of nuclear spins through bonding electrons, in SS NMR is quite small and obtains not many attentions herein. in contrast, is the direct interaction of nuclear spins through space without involving t he electron clouds. describes the quadrupole interaction for all nuclei with a spin number > 1 /2 such as 2 For molecules in solution, due to the rapid tumbling of the molecules the CSA and dipolar interactions are rarely to be observed and line peaks can be obtained. However, in the case of solid samples the motionally averaged spin Hamiltonian are dominantly govern ed by the much stronger short range dipolar interaction, chemical shift anisotropy, and quadrupole coupling (for > 1 /2 ) which give rise to the undesired line broadening in SSNMR spectra Therefore, numerous techniques have been explored to reduce the spectra line width in SS NMR. H. It is generated from the interaction of nuclear quadrupole moment with the gradient of its surrounding electric field at the nucleus site. 3 1 1. 3 Cross p olarization and m agic a ngle s pinning To achieve high quality spectra for solid samples the spin and space manipulations of NMR interactions are usually adopted. The most useful spin manipulation technique is the cross polarization (CP), while a popular space manipulation method is the usage of magic angle spinning (MAS). The experiment by
65 combination of these two techniques, CP/MAS is widely used in SS NMR for polymeric samples. For dilute spins, such as 13C, it is hard to obtain NMR signals without any manipulation due to the problems arising from its low natural abundance: low signal to noise ratio and very long relaxation time. These problems can be solved by CP. The rich magnetization in the abundant 1H spins can be transferred to the X spins in the sample via the dipol e dipole coupling between 1 H and X spins. Since dipolar interaction is distancedependent, CP is more efficient for spins in the crystalline regions than those in the amorphous regions. Figure 34. The cross polarization pulse sequence. Magic Angle Spinning (MAS) is a routinely used method in the vast majority of SS NMR experiments. It involves rotating the s olid sample in a cylindrical rotor at high frequency The spinning axis is ori ented at the magic angle, = 54 7 with respect to the applied magnetic field 0, as shown in Figure 35. The major tasks of MAS are to 1H XContact pulse 90xy y Contact pulse y decoupling
66 eliminate the chemical shift anisotropy and heteronuclear dipolar coupling effects, and thus to reduce a powder pattern to a single line at the isotropic chemical shift. The magic angle of 54 7 is calculated from the anisotropic Hamiltonian according to the perturbation equation.88 For a sample spinning uniform ly at an angle relative to the applied field the theoretic anisotropic interaction contains the second Legendre Polynomial factor ( ) =1 2 ( 3 2 1 ) in the constant term. 77 Such a mathematic expression provides a simple but efficient way to eliminate the anisotropy broadening by setting = 54 7 to satisfy 2 1 = 0 This angle 54. 7 is the socalled magic angle. Spinning rate should be fast on the NMR timescale, i.e., the change of molecular orientation has to be fast relative to chemical shift anisotropy, dipolar coupling, etc.77, 90 Figure 35. The magic angle spinning experiment setup. The angle between the spinning axis and applied field is set to 54 7 the magic angle. 3 1 2 NMR Spectro meter NMR spectrometer has now become a very sophisticated instrument, which has been used widely in chemistry, material science, biology, physics, etc. As Figure 36 illustrated, t he spectrometer consists of several important component s: B0 R 54.7
67 A strong, homogeneous and stable magnetic fieldt he perfectly uniform magnetic field contributes to the high quality signals by avoiding inhomogeneous broadening and allowing the small changes of Larmor frequency to be resolved. Modern NMR magnet is usually accomplished by using persistent superconducting materials, typically alloys of Nb and Sn. The superconducting coil is immersed in a bath of liquid He which is isolated by a liquid N2 A probet he probe is inserted into NMR though the centre of the magnet, which is called bore Inside of the probe, the sample is surrounded by a RF coil. RF pulses are exerted on the coil as currents to generate a transient field on the sample, and the response from the sample is then detected by the coil as well. The thermal noise from the RF coil is the major noise in a well designed NMR spectrometer. reservoir. The positional and temporal inhomogeneity can be compensated by the shim coils. A n RF transmitter it is the RF transmitter that generates controlled RF pulses, either short but highpower ed, or long but low powered. A pulse programmer it is used to produce precisely timed pulses and delays; A sensitive RF receiver since the NMR signal from the RF coil is in the order of V RF receiver is used to capture and amplify the NMR signals A digitizer it is a n analoguedigital converter, used to convert the NMR signals (oscillating electrical current) into digital form (a sequence of ones and zeros) which can be stored in computer memory A computer as the heart of NMR spectrometer it co ntrol s all the components and process es the data
68 Figure 3 6 Schemetic overview of a NMR spectrometer The magnets for solution and SSNMR are the same. It is the probe that determines the state of the samples that can be examined. T ypical probes used in SS NMR include various magic angle spinning probes, static wide line probe, etc. 3 1 3 Introduction to Deterium Solid State NMR Deuterium is the isotope of hydrogen with e xtremely low natural abundance and a spin number of one. Like other nuclei with spin number I greater than 2H NMR lineshapes are mostly dominated by the quadrupole interaction, which arises from the interaction of the nuclear electric quadrupole moment with the surrounding electric field gradient (EFG) at t he nucleus. Powder NMR spectra of 2 Magnet Magnet Shim Coils Shim Coils Probe Computer RF Source Pulse Prog RF AMP RF Detector Probe Digitizer H of static samples consist of doublet patterns which arise from the two spin transitions: + 1 0 and 0 1 These doublet are called Pake patterns with their horns split by 3 4 [ 2 / ] falling in the range of 105 165 kHz for organic samples. The Pake patterns and the doublet splitting are shown in Figure 37. Transmitter
69 The quadrupole splitting between the doublets is given by: =3 2 [ 2 / ] 1 2 ( 3 2 1 ) 1 2 ( 2 2 ) (3 1 2 ) where 2 / is the quadrupole coupling constant, in the range of 140 220 kHz for organic compounds; and are the Euler angles which define the principle axis frame of the EFG relative to the laboratory frame at the deuteron site; and is the quadrupolar asymmetry parameter of the EFG Figure 37. Pake pattern of 2 H NMR The doublet splitting is due to the two allowed spin transitions. Their horns are split by of the quadrupole coupling constant. 2H NMR is particularly well suited for investigation of molecular dynamics. In polymer chemistry, by isotope labelling of the specific site of interest, 2H NMR can be very informative in terms of molecular motions. The C D bonds give birth to the electron filed gradient, which is axially symmetric around the C D bond. Such a fact allows the mo lecular motions to be monitor ed through the orientation of an individual C D bond, largely simplifying the NMR data interpretation. Moreover, the moderate width of powder patter n s, as compared to other quadrupole nuclei, and the high dynamic range of 104 t o
70 106 Hz, make 23 1 4 Pulse Sequences H NMR extensively effective and applicable in dynamics studies, via spectra lineshape analysis, relaxation time measurements, and exchange experiments. The success of NMR as a powerful characterization tool for polymer chemistry is undoubtedly related to the highly developed pulse sequences. Some of the ones that turn out to be quite useful for polymer studies will be discussed in below. 3 1 4 .1 Quadrupol e e cho The quadrupole echo, also termed as solid echo, is the most common pulse sequence used in 2Quandrupole echo has been widely used to study slow motions of polymers at condensed state. H NMR measurements As a result of the strong quadrupole coupling, deuterium spectra acquired with one pulse excitation can have extremely broad lines which display fast decaying free induced decays (FIDs). Due to the occurrence of the receiver dead time (a delay of measurement after a pulse), the loss of signal arising from the rapidly delaying FID accounts for a significant part in the total FID, which leads to severely distorted spectral lineshapes. This phenomenon is negligible in solution NMR because the FID decays much more slowly On the contrary, in solid state NMR, this effect cannot be ignor ed. Fortunately an echo sequence can resolve this problem. Q uadrupol e echo is such a method. 79, 91, 94, 95 Its pulse sequence can be seen in Figure 38 The magnetization dephasing due to the quadrupolar interaction can be refocused by the second 90 pulse and hence form a quadrupolar echo. By taking the standard Fourier t ransform of the data starting at the echo maximum, there is no signal loss due to a dead time delay .77, 91
71 Figure 38 The Quandrupole echo pulse sequence. Two 90 pulses are applied; experimental signals are acquired after echo delay 3 1 4 .2 SUPER SUPER, which stands for S eparation of U ndistorted P owder patterns by E ffortless R ecoupling, is a newly developed recoupling technique used t o obtain undistorted, quasi static CSA powder patterns under magic angle spinning .96As mentioned previously, high resolution NMR in solids can be achieved via MAS by spatially averaging anisotropic spin interactions, which act ually contain valuable dynamical and structural information. To avoid this sacrifice, recoupling techniques which can selectively reintroduce anisotropic interactions while maintaining high resolution via MAS have been explored. Among them, SUPER attracts researchers attention because of its robustness: being insensit ive to pulselength related imperfections, being compatible to a wide range of hardware, and works effectively under standard power levels and spinning rates. The pulse sequence is shown in Figure 39 SUPER can be conducted under CP condition as Figure 39 illustrated, or single pulse excitation (SPE) (not shown here) The latter can be easily realized by removing the CP part without affecting the nature of SUPER pulse sequence. 96 90 0 90 0 90x 90y
72 SUPER pulse sequence is derived from a similar sequence reported by Tycko and co workers.97 By replacing the 180 pulses with the 360 pulses, SUPER yield s two dimensional spectra with one dimension providing recoupled chemical shift anisotropy (CSA) information an d the other dimension providing the high resolution spectrum Using SUPER, isotropic and anisotropic chemical shifts can be separated effectively. Figure 39 The SUPER pulse sequence under CP condition: DD stands for dipole decoupling; CP indicates cross polarization; x and x are 360 pulses. TOSS ( total suppression of spinning sidebands ) and intergral are used to suppress the sidebands up to the fourth order.98 3 1 4 .3 REDOR Like SUPER, REDOR (rotationalecho double resonance) and REREDOR (rotor encoded REDOR) are recoupling techniques used to selectively reintroduce anisotropic interactions including dipoledipole coupling under magic angle spinning. The difference between SUPER and REDOR lies in that the former provides CSA information while the latter studies dipoled ipole interaction. The pulse scheme of REDOR is shown in Fig ure t1 preparation t2 DD DD I StR/2 t1= n tRx x x xTOSS integral CP tatbtRtbtRta P(t) 1 0
73 3 10. The major idea of this technique is to use a series of well spaced 180 pulses to prevent the heteronuclear dipolar coupling from being averag ed to zero of by MAS. Since the dipoledipole coupling depends on the interspindistance REDOR is widely used to investigate molecular structure, especially the heter o nuclear distance and angle determination. Figure 310. The rotational echo double resonance (REDOR) pulse sequence. 3 1 4 .4 REREDOR REREDOR (rotor encoded rotational echo double resonance) originates from the REDOR sequence but is more sensitive in determination of heteronuclear 1It is a two dimensional NMR experiment with high resolution spectrum in the direct dimension and spinning sideband patterns in the indirect dimension. Usually a fast MAS with rate of 25 kHz or higher is performed to offer the desired spectral resolution as well as informative sidebands. The scheme of REREDOR pulse sequence is shown in Figure 311. Start ing with the preparation of initial S magnetization via CP transfer H X dipolar couplings. The improved sensitivity arises from the dipolar coupling. 900 y 1800 1800 I S Rotor R
74 (single pulse excitation also works), REREDOR sequence is composed of two REDOR blocks f ollowed by indirect dimension t1Under very fast MAS, REREDOR is capable of studying the dipoledipole coupling of groups in either rigid or mobile systems, where can be isotopically dilute, such as after each block. After recoupling, spin S magnetization will gain several phase factors, which are related to the recoupling constant and can be calculated from signal intensity of spin S The quantitative dipolar coupling information can be extracted from the spinning sideband fitting. 13 C. This makes REREDOR a useful technique for investigation of the molecular structure in polymeric materials. Figure 311. The pulse sequence of CP based Rotor encored REDOR (REREDOR) 3 2 X ray Scattering Techniques 3 2.1 Introduction X ray diffraction (X RD) technique can be dated back to the time of Bunn.11 preparation t2 t1 DD I S CP recoupling recoupling t1 y y y Sx+SyIz Nowadays, it has become one of the mature tools used to investigate morphological properties of polymeric materials. Like light scattering and neutron scattering, X ray scattering is a technique utilizing the X ray as the incident media to study the inner
75 structure of interphase orientation of various materials, such as metals, ceramics, alloys, and polymers. Both qualitative and quantitative information can be obtained from XRD, including dimensional parameters of unit cells, degree of crystallinity, lamel lar periodicity, and degrees of orientation. X rays, like light, are electromagnetic radiation with wavelength in the range of 0.1 to 100 When X rays pass though a substance their energy is dissipated by the ejection of orbital electrons and by scattering resulting in the loss of the X ray intensity. The interaction of incident X ray and sample atoms can be complex with several processes competing : elastic scattering, inelastic scattering, fluorescent X ray, photoelectron releasing, and heat accumulat ion. The elastic scattering (also named Thomson scattering), which has the scattered X rays with the same energy, i.e., wavelength as the incident radiation, is the main type of scattering involved in XRD and this dissertation. Since the emitted X rays con tain characteristic structural information, XRD has been widely used in polymer science and industry. 3 2.2 WAXS and SAXS Wide angle X ray scattering (WAXS), sometimes termed as wide angle X ray diffraction (WAXD), defines the scattered X rays occurring at the angle of 2 within the range of 5 ~ 120 where 2 (the socalled scattering angle) is the angle between the incident and scatter ed X rays, as shown in Figure 312. In the case of PE, the most useful information extracted from the 2 range of 5 ~ 50 corresponding to the atomic spacing ranging from 20~ 2 if Cu radiation is used. WAXS measurement can provide information of atomic spacing ( d spacing), unit cell parameters, as well as the degree of crystallinity.
76 Small angle X ray scattering (SAXS) is the twin of WAXS by sharing many similarities such as the scattering theor ies and the instrumentation schemes The basic difference of these two methods lies in the scattering angle 2 : SAXS detects the 2 value smaller than 5 Typical 2 for Cu radiation falls in the range of 0 2 ~ 2 corresponding to atomic spacing of 440 ~ 44 which is much larger than the dimension that WAXS can detect. Consequently, SAXS equipment is much more sophisticated than WAXS resulted from the small scattering angles Besides WAXS and SAXS, there are several other XRD methods existing such as medium angle X ray scattering (MAXS), ultra small angle X ray scattering (USAXS), classified by the scattering angle as well. Figure 312. Schematic illustration of X ray scattering for solid samples 3 2.3 Braggs Law I n crystallography Braggs law is widely used. Since the incident X rays are always in phase and parallel, when they interact with the sample atoms, the scattered X rays interfere constructively and the differences in the beam travel path equal s to some integer multiples of the wavelength. The occurring of such constructive interference ensures the X ray beams to be diffracted at the same angle as the incident
77 beam with respect to the atomic plane. This condition is summarized in Figure 313, named as Braggs law: = 2 sin (3 1 3 ) where is the wavelength of incident radiation, is the distance between the adjacent parallel crystallite planes, and i s the integer indicating the order of diffraction. Figure 31 3 Scheme of Braggs Law d diffraction crystallite atomic planes d d
78 CHAPTER 4 INFLUENCE OF BRANCH FREQUENCY ON DYNAMICS AND STRUCTURE OF PRECISION POLYMERS 4.1 Introduction T he precision polymers exhibit significantly different thermal behaviors compared to the commercial polyethylenes as described in Chapter 2 I t is believed that the branch identity and branch frequency influence not only inter but also intrachain packing an d the corresponding macroscopic properties. H erein, the branch identity is fixed while the branch frequency varies. T o simplify the study, a seri es of methyl branched polymers possessing various branch concentration has been chose n in order to investigate how and why the branch frequency will influence the microscopic properties of precision polyolefins, therefore to help us understand the structureproperty performance relationship in commercial PEbased materials Figure 4 1 Structure of precisely branched polymers: (a) linear ADMET PE, with no branch es ; (b c) PEn CD3, with deuterated methyl group on each and every nth carbon. Four precise polyolefins including three perdeuterated polymers were prepared to study the conforma tional defects: ADMET PE, PE21 CD3, PE15 CD3, and PE9 CD3, in which the number after PE indicates the spacing of two adjacent CD3 branches along the backbone F or instance, PE21 CD3 specifies the ADMET polyethylene with CD3
79 group regularly displaced on each and every 21st4.2 Polymer Synthesis carbon along the ethylene backbone, seen in Figure 41. 4.2.1 Monomer Synthesis Figure 42 Synthetic methodology used to produce perdeuterated monomers: when x=9, monomer 1 for PE21CD3; x=6, monomer 2 for PE15CD3; and x=3, monomer 3 for PE9CD3 The model polyolefins were synthesized via ADMET step polymerization using a modification of the procedure described before.26, 31 Synthesis of the three symmetric diene monomers containing precisely displaced deuterated methyl branches was achieved based on a modification of the synthetic methodology in literature, described in Figure 42 Basically, the synthesis started with the diethylmalonate being disubstituted with an alkenyl bromide possessing the appropriate methylene spacing. The resulting diester was then saponified and decarboxylated to yield the corresponding
80 monoacid Upon selective reduction of the acid with lithium aluminum deuteride, the acid was able to be reduced to the primary alcohol, which was mesylated and then reduced via deuteride displacement to offer the symmetrical diene of interest. The frequency of the CD3 branches can be varied by choosing starting alkenyl bromides possessing various methylene spacing, e.g. when 5bromo 1 pentene (x=3) is used, the resulting symmetric diene is the monomer of PE9CD34.2.2 Polymer Synthesis The precision polymers sampl ed for this study were synthesized by ADMET step polycondensation chemistry. In general, the diene monomer is polymerized in the presence of Schrocks or Grubbs catalyst under mild step polymerization conditions to form the unsaturated precision polymer, which is then exhaustively hydrogenated to offer the saturated precision polymer, as shown in Figure 4 1 For linear ADMET PE, 1,9 decadiene was used as the starting compound for polymerization.28 Figure 43. Synthetic methodology used to produce precision polyolefins via ADMET chemistry. 4.3 Characterization of Deuterated Polymer s The sample polymers were characterized using various analytical methods. Besides the normal characterization methods, such as thermal analysis, molecular
81 weight determination, and structure confirmation, the morphology and dynamics were investigated by solid state NMR spectroscopy and scattering techniques. 4.3.1 Basic Characterization The basic characterization data of the deterated precision polymers studied herein are presented in Table 4 1 Table 4 1 Basic characterization data of deuterated and protonated precision polymers Model Polymer Methyl Branches/ 1000 Backbone Carbons M w a (kg/mol) PDI b T m c (K) m c (J/g) T g d (K) ADMET PE 0 68 2. 7 407 205 --PE21 CD 3 48 56 1.9 335 101 --PE15 CD 3 67 53 1.9 312 79 --PE9 CD 3 111 56 1.8 264 30 --PE21 C H 3 48 34 1.7 335 103 229 PE15 C H 3 67 29 1.7 312 82 229 PE 9 C H 3 111 30 1.7 259 28 230 PE7 CH 3 143 13 e --213 19 --PE5 CH 3 200 28 e --amorphous 208 a. Gel permeation chromatography (GPC), polystyrene standards b. Polydispersity index (Mw/Mn) from GPC c. Determined by differential scanning calorimetry (DSC) d. Glass transition temperature from literature99 e. number averaged molecular weight data28 4.3.2 Morphology Measurement The morphology and packing behaviour of HP15 and HP21 (protonated f ormat s of PE15 CD3 and PE21 CD3) have been published earlier .99, 100 The results by X ray scattering and electron microscopy in previous literature99 support our rotational model extracted from solid state NMR measurements in this study. The lamellar thickness (between 10 and 20nm) and powder scattering patterns of the two precisely displaced polymers show that some of the methyl branches are incorporated into the crystalline
82 region, i.e., the methyl side group functions as conformational defect within the crystal lattic e. As expected, the shorter spaced PE15 CH3The deuterium labelled versions were also invest igated by microscopy and scattering techniques to compare with the previous results from protonated ones. Consider that the sizes of deuterium and hydrogen atoms have no significant difference within the scale involved herein; the lamellar size and the chain packing behaviour should be identical for protonated and deuterated ones. TEM measurements of CD exhibits higher degree of conformational disorder. However, the motion and torsion angle of this disorder are uncertain by scattering and microscopy measurements. 3branched precision polymer show no different morphological behaviour in contrast to CH3Modern transmission electron microscopy is famous for its capability of imaging fine structures at nano scales which is wellsuited for analyzing the lamellar structures formed in semi crystalline polymeric samples. branched one, indicating the reproducibility of ADMET synthesis and correctness of the hypothesis concerning no significant effect of deuterium on packing behaviour TEM measurements were conducted at Max Planck Institute for Polymer Research (MPIP), Mainz, Germany White powder like of ADMET precision polymer PE21 CD3 as achieved from synthesis was dissolved in O xylene with a concentration of 10mg/10ml ; the mother solution was then 10 times diluted TEM sample was prepared by putting one drop of diluted solution on carbon coated grid and stored in the refrigerator at around 7 C for single crystal s to g row The O x ylene was evaporated while the grid remained in the refrigerator.
83 The solutiongrown single crystal for PE21CD3 shows lamellar crystal structure, illustrated in Figure 44. Images in (a), (b), and (c) of Figure 44 were taken from the crystals as developed from O xylene solution. Clear lamellar type crystals can be observed. The thickness of the lamellar was measured using various techniques including Pt shadowing, Electron Energy Loss Spectroscopy (EELS) and Atomic Force Microscopy (AFM), as seen in Appendix A. The thicknesses determined by these three methods are in agreement with each other: 1013 nm from TEM (Pt shadowing), 8 nm from TEM (EELS), 9 nm from AFM. The 10 nm averag lamellar thickness confirmed the pr evious result about thickness dimension for PE21CH3Figure 44 (d) and (e) are TEM images of the socalled edge on lamellar structures, which were prepared by shearing the molten sample between two glass slides and consequently the edges of the lamell ar structures can be readily imaged. Directly measuring the dimension of the lamellar edges provides an averaged thickness of around 16 nm. It is worth noticing that the lamellar thickness from the edgeon structure is not comparable with solutiongrown single crystals because in the former case, a crystalline nucleation thread (shish) is formed and thus thicker crystals are expected. The structures shown in (c) and (d) in Figure 44 can be easily identified as the famous shish kebab morphology in c rystal lography
84 Figure 44. TEM images for ADMET precision polymer PE21CD3 : (a) and (b ) solutiongrown single crystals as prepared, no contrast enhancement ; ( c) and ( d ) edgeon lamellar structures. The red arrows in ( d ) indicate a n edge of a lamellar with size of 16.5nm. 4.3.3 Deuterium Solid State NMR Measurement Molecular motions are of utmost importance for the macroscopic properties of polymers.21, 101 In linear polymers like polyethylene (PE), branches change the mobility and have pronounced effects on the mechanical properties102, 103, processablity and drawability and has, therefore, been studied in detail.9, 104, 105 In the systems studied (a) (b) (d) ( c ) 100 nm 0.5 m 0.5 m 1 m
85 previously and in particular in merchandized PE, the side chains are irregularly distributed along the main chain, which results in complexity and uncertainty of the defect influence. Therefore, it is highly desirable to study the effect of branching on molecular motions in polyolefins with exactly defined branches both with regard to the chemical nature of the branch and the distance between the branches along the chain. Once such samples are available, both rate and amplitude of chain motions can be studied site selectively with advanced solid state NMR met hods.79 Moreover, due to the sensitivity of 13C chemical shifts on conformation and torsional angles,20, 106 NMR can also probe structure and dynamics of the chain defects imposed by the branch. The effect of branching on mobility is particularly interesting in the crystalline regions of the semi crystalline polymer, provided the branches can be incorporated. This suggests methyl groups as branches. Combining stateof the art synthetic chemistry and NMR spectroscopy we can then address the following questions: What is the nature of the motion of the branches? How do the branches alter the geometry and mobility of adjacent chain segments? Does the motion of one branch influence the motion of neighboring branches (collective motion, rotator phas e)? Can gauche conformers be incorporated in the crystalline regions? Indeed, such precisely defined polyolefins have recently been prepared by acyclic diene metathesis (ADMET) polycondensation26, 31, 46, where the primary structure, i.e. branch identity and s pacer, can be controlled. Moreover, the methyl branches can be selectively deuterated, such that the mobility of the branches themselves can conveniently be studied by 2H NMR.91, 107109
86 (a) (b) (c) Figure 4 5 Temperature dependence of 2H spectra for (a) PE9CD3; (b) PE15 CD3; and (c) PE21 CD3. Spectra were acquired on a Bruker Avance 400MHz NMR spectrometer using quadrupolar echo pulse sequence with an echo delay of 30s, a 90o pulse width of 2.5 s, and a recycle delay of 1s. Samples were heated above melting points to remove the thermal history prior to collecting data every 10K. The special line shape of P E15 CD3 at T = 303 K, plotted bold, indicates the presence of fast axial rotational dynamics. Governed by the quadrupolar interaction, 2H NMR is well suited for studies of local molecular dynamics. The 2H NMR line shapes are very sensitive to segmental mo tions .91, 107109 Figure 4 5 displays temperature dependent 2H NMR spectra for the three polymer samples with deuterated methyl branches. Its easy to see that the trend of line shape with temperature is similar for all of them and shows the gradual buildup of molecular dynamics. At low temperatures Pake patterns indicative of an axially symmetric tensor with fix ed C3 axes of the branches are observed. With increasing
87 temperature, the singularities of the Pake pattern broaden and the 2H NMR line shape changes via an almost rectangular shape to a pattern resembling an asymmetric tensor (asymmetry parameter =1). Si nce the 2H NMR spectra contain signals from chains in crystalline and noncrystalline regions of the sample, which for the CD3groups cannot be separated by the standard procedures,91, 108 we mention that such patterns can also result from branches undergoing large angle motions on intermediate timescales. Upon further increasing the temperature, a rather narrow component in the centre of the spectrum emerges. Remarkably, only in PE15 CD3 at T = 303K, does t he 2H NMR spectrum show a regular Pake pattern ( =0) with half the static line width, indicating the presence of axial rotation sufficiently fast to average the 2H line shape not observed in linear PE itself.91, 107109 Therefore, we attribute this behaviour to motions of methyl groups embedded in the crystalline regions. The highly asymmetric 2H line shape observed in all cases is usually interpreted as evidence for a kink motion,91, 108An attempt measuring 2 dimentsional but is also consistent with ill defined rotations around the local chain axis ( alltrans in the crystallites) with amplitude about 40, see n in section 4.3.6. The detailed interpret ation of the deuterium NMR lineshapes is a ided by fitting to spectra lineshape simulations, which are presented in chapter 5. 2H exchange NMR indicates that there has no slow orientation on the exchange time scale due to absence of off diagonal intensities, as shown in Appendix B.
88 4.3.4 13 C Solid State NMR Measurement Figure 4 6 13C isotropic and anisotropic chemical shifts for precise polymers: (from the top) ADMET PE at 300K, PE21CD3 at 300K, PE15CD3 at 300K, PE9CD3 at 235K and 245K. Spectra were acquired on a Bruker Avance III NMR spectrometer at 213 MHz 13C Larmor frequency (850 MHz for 1H) using the SUPER96 pulse sequence with a spinning rate 4125 Hz and an initial CP set to improve 13 C spin polarization. 13C NMR experiments, observing randomly distributed 13C sites along the chain, provide a direct picture of the chain dynamics. The isotropic chemical shift in a magic angle spinning (MAS) NMR spectrum79 indicates the local conformation20, 106 and the motionally averaged chemical shift anisotropy (CSA), that can be siteselectively recorded by twodim ensional NMR,79 reflects local reorientations of CH2 groups of the
89 polymer backbone. Figure 4 6 displays both the isotropic and anisotropic chemical shift patterns of the three samples and a linear PE sample prepared by ADMET synthesis, acquired by separation of undistorted powder patterns by effortless recoupling (SUPER).96 For linear PE without methyl branches, the 13C isotropic chemical shifts in Figure 4 6 left column, show a strong signal at around 33.5 ppm, characteristic of the alltrans stems in the orthorhombic phase and a minor monoclinic contribution at around 35 ppm, as well as a very broad amorphous signal with low intensity at 3329 ppm. In the following, only the signals originating from the stems in the crystalline regions will be discusse d. The 13C static line shape for the orthorhombic signal of ADMET PE exhibits ppm and 51 ppm perpendicular to it as previously discussed in the literature.110Introducing regularly spaced methyl side groups along the polyethylene backbone changes the morphology 99 and the molecular dynamics. In the 13C MAS NMR spectra, Figure 4 6 two new signals are observed in all cases, assigned to the methyl group(~ 19 ppm) and the methine carbon (~40 ppm). Moreover, changes are observed for the signals assigned to the regular CH2 units located in crystalline and noncrystalline regions of the samples along the PE main chain. In PE21 CD3 the NMR signal at 33.5 ppm known from linear PE and assigned to all trans conformations in the crystalline regions splits into two signals at 33.6 ppm and 33.2 ppm, which are resolved in our 850 MHz NMR spectrometer. The CSA line shapes for the MAS NMR signal at 33.6 ppm shows a CSA tensor line shape similar to that of PE. The CSA powder line shape of the MAS NMR signal at 33.2 ppm, however, displays the lineshape of an almost axially symmetric CSA tensor, where the principal value of 13 ppm along the crystalline c axis
90 persists and the two other values are largely averaged. Remarkably, the averaging observed is consistent with rotations of the trans segments around the chain axis with the same amplitude ( 40) as deduced from 2H NMR for the branch (see section 4.3.6). It should be noted that the isotropic chemi cal shift of the averaged tensor is shifted by 0.8 ppm towards the gauche containing signals in the non crystalline regions. As 13C chemical shifts of polymers are sensitive even to small changes in torsional angles,106 we assign this small shift to local deviations from the regular alltrans geometry, resulting from chain twists in the vicinity of the methyl branch. This will minimize t he spatial requirements for incorporating the methyl branch in the crystals. The isotropic shift observed actually reflects an average over twisting motions of the defected chain units close to the methyl branch. Finally, the 13C NMR signal at 30 ppm in th e spectrum of PE21 CD3 is assigned to noncrystalline PE. This line is considerably narrower and more intense compared to linear PE. These differences, however, can be attributed to the different crystallinities of the two samples (30 % in PE21 CD3 vs. 85% in ADMET PE) and the nonquantitative CP MAS method used to efficiently record the 13For PE15 CD C MAS NMR spectra. 3, the signal assigned to CH2 sites in the crystalline regions is observed at 32.9 ppm with a shoulder of about half the signal heig ht at 33.5 ppm. Remarkably, all 13C signals in the NMR spectrum of PE15 CD3 show the powder line shape of well defined axially symmetric CSA tensors, even at 33.5 ppm, the isotropic chemical shift characteristic of undistorted alltrans units. Thus, whereas in PE21 CD3 the CH2 groups in proximity to the lattice perturbing methyl branches perform local motions like in pinned defects, the rotation in PE15 CD3 at T = 303 K involves all CH2
91 groups along a given polymer chain and thus shows c ollective behavior like in a rotator phase, also deduced from X ray data.99The ADMET precision polymer PE9CD 3, with the shortest spacer of only 8 CH2 units between subsequent methyl branches, exhibits at T = 235 K (about 20 K below its melting point) a 13C MAS NMR spectrum very similar to that of PE21 CD3 and PE15CD3, see Figure 4 6 All known carbon signals are observed close to the chemical shifts from the other spectra, but the line width of the NMR signals is 24 times larger. This increased line width at low temperatures is attributed to a substantial conformational disorder in the crystalline regions of the sample, induced by the dense distribution of methyl branches along the polymer backbone. At T = 235 K, the static powder line shape of the cr ystalline component observed at 32.8 ppm, resembles the pattern of an incompletely averaged CSA tensor of the chain defects in PE21CD3. At T = 245 K, however, only 10 K higher and 10 K below the melting point, all 13C MAS NMR peaks become much sharper and all signals except the methyl signal are shifted towards higher fields. Moreover, instead of static CSA powder patterns, isotropic lines are observed under CSA recoupling, as seen in the spectrum plotted as a dotted line in Fig ure 46 The findings at T = 245 K thus indicate substantial pretransitional conformational disorder in PE9CD3, which hampers the comparison of this sample with the other two. Therefore the results obtained for PE9CD3 have not been used to develop a model for the molecular dynamic s in the regular methyl branched polyethylene samples.
92 4.3.5 Motional Model Figure 4 7 Schematic model for rotational dynamics in (a) PE21CD3, ADMET precision polymer with deuterated methyl branch on every 21st carbon along the backbone and (b) PE15CD3, ADMET precision polymer with deuterated methyl branch on every 21st carbon along the backbone Combining the results for PE15 CD3 and PE21 CD3 obtained from 2H and 13C NMR, which probe the branch and the neighbouring chain defect separately, we propose a simple model for the molecular dynamics of the regular methyl branched polyethylene in the crystalline regions as shown in Figure 4 7 : In the crystalline regions of PE15 CD3 and PE21 CD3, the lattice perturbing methyl branches undergo axial oscillations around the polymer backbone. In both samples, this motion of the methyl branches gains in amplitude and rate with increasing temperature. However, in the case of PE21 CD3, this dynamic mode is restricted to CH2 groups in close proximity to the methyl branch in a twist defect, whereas the CH2 groups remote from the methyl branches are not involved in this motion as demonstrated by the regular PE CSA pattern observed at 33.6 ppm, but may undergo 180 flips as known for crystalline PE.110 The relative intensities of th e isotropic lines corresponding to the axially
93 CH2 units involved in the dynamic twist defect is estimated as 34 CH2 units on either side of the branch. Based on this pict ure of a localized dynamic perturbation of the crystal lattice caused by the methyl branches, one would expect a 1:1 ratio between rotating and rigid units for PE15 CD3. While MAS NMR signals indicate undistorted trans conformers, see Figure 46 all sites along the chain are found to exhibit similar rotational dynamics. In other words, the localized rotational motion observed in PE21 CD3 turns into collective dynamics due to the higher density of defect sites in PE15 CD3The main results of our study can then be summarized as: Nature of the motion of the branches : The 2Influence of the defects on the adjacent chain segments in the same molecule: The methyl branches lead to a twist of the adjacent chain segment Figure 4 7 manifesting itself as a component at lower H spectra Figure 4 5 clearly show that the branches perform angular restricted rotations reaching amplitudes of about 40 around the long axis of alltrans PE chains close to the melting point. 13C chemical shif t. From the intensity of this line, compared to that of the alltrans segments, the length of the defect is readily estimated to involve about 34 CH2Local vs. cooperative axial motion: The chemical shift anisotropy patterns, Figure 4 6 right column, clearly show axial motions of the twisted segments similar to those of the branch itself, whereas the CH groups at either side of the branch. 2groups of the undistorted alltrans segments in the sample with long chains between the branches, PE21 CD3, are rigid on the timescale of these experiments, as in linear PE itself. Thus, the motion of the branches and the
94 adjacent chain defects are localized and can be considered as pinned. In PE15 CD3Incorporation of gauche defects in the crystals: Incorporation of gauche defects is only observed in PE9 CD however, the twisted parts are so close that the axial motion i mposed by the defect becomes cooperative and also rotates the undistorted trans segments in between, as in a rotator phase, Figure 4 7 (b). 3, 4.3.6 Analysis of NMR Powder Line shapes the sample with the highest branch density, at temperatures close to the melting point. In static solid state NMR spectra of quadrupolar nuclei such 2 H or nuclei with a substantial chemical shift anisotropy (CSA) the line shape is governed by these angle dependent interactions. Both, the first order quadrupole coupling as well as the chemical shift anisotropy are second rank tensorial interactions, which can be described dependence of the resonance frequencies for a set of polar coordinates between the zz component of the principal axes system of the tensor and the magnetic field is given by ) 2 cos( sin 1 cos 3 ) (2 2 2 1 79 ( 4 1) Rotational motions in the fast motion limit lead to a new tensor with averaged principal values and different principal axes. The angular dependence of the NMR frequency can then still be described by an equation similar to Equation ( 4 1), but with averaged principal values and new polar angles: ) 2 cos( sin 1 cos 3 ) (2 2 2 1 ( 4 2 )
95 Processes of different geometries can give identical principal values. Moreover, in case of 2 ambiguity . In the current case, however, we are interested in the chain motion in the crystalline regions only, where the isotropic chemical shift 13C tells us that we look at motions of extended trans units, which can only rotate around their local chain axis, the crystallographic c axis of the PE. Therefore, in order to analyse the static NMR line shapes in terms of dynamics, we assume Gaussian distributed rotational fluctuation around the crystallographic c axis of the PE samples. The unique axes of the tensorial i nteraction, i.e. qZZ for 2H and CSZZ for 13C form angles of 90 and 0 ,110 respectively, with the c axis. Assuming, that the molecular dynamics close to the melting point of the crystallites is in the fast motional limit, the NMR line shape can conveniently be computed using the NMR Weblab.109Along these lines, the quadrupole and CS A line shapes obtained for a systematic figure 4 8 In the case of the 2 H NMR, the strength of the quadrupole coupling, increases with the spread of the distribution, reaching = 1 80 an axially symmetric averaged quadrupole tensor with uniq ue axis parallel to the rotation axis results. In the case of 13C NMR CSA powder line shapes for CH2 groups of extended trans conformers these fluctuations keep CSZZ constant and reduce 40 lead to a powder line shape which is difficult to distinguish from that of an axial rotation as almost vanishes already. Moreover, it should be noted that the experimental line shapes not only reflect these geometrical
96 aspects o f the local molecular dynamics. For NMR spectra recorded in a temperature range 050 K above the glass transition temperature the time scale of the dynamic processes will be another crucial parameter and distortions of the NMR line shapes due to the interm ediate motional effects will have to be taken into account to obtain agreement between ex perimental and computed spectra. Figure 4 8 : Computed 2H NMR and 13 C NMR powder line shapes for a Gaussian the crystallographic c axis of crystalline PE according the motional model presented in Figure 4 7 From Figure 4 8 it is clear that the only motional mode that is consistent with both the observed 2H and 13 C NMR line shapes recorded at temperatures close to the melting point, where the fast motion limit is a good approximation, involves fluctuations of the order of 40 used to develop our model of twisted defects.
97 4.4 Conclusion of Branch Frequency Effect Reflecting the importance for the mechanical behaviour, the nature of chain motions of the stems in the crystals has been under debate for decades. In particular, local conformational defects21, 101 as opposed to propag ating twists111 were considered as mechanisms for chain transport. Our results clearly favour twists, which require considerably less disturbance of the crystal lattice. Remarkably, recent NMR studies of chain diffusion between the crystalline and the no n crystalline regions of linear ultrahigh molecular weight PE showed that chain transport, involving cooperative motion of the stems, has different temperature dependence than the local mobility of the chains. This indicates increased presence of defects t hat do not transport the chain at higher temperatures.112115 Rotator phases in polymers were deduced from X ray scattering thermal analysis, and solid state 29Si NMR.116, 117 However, for branched polyethylenes, this is the first time that clear evidence of a rotator phase has been found, consistent with the r esults of scattering and microscopy.99 Our proof of chain twists also helps to explain the conformational motions of amorphous polymers in the melt, where twodimensional NMR has clearly shown that conformational transitions do happen, yet they occur wit h broad distributions of rotational angles.94 This suggests twisted defects in the chain as also considered theoretically.118, 119 Clearly, the occurrence of both isolated and cooperative motions in stems of single molecules could only be detected by a combination of innovative synthetic chemistry generating well def ined model systems and stateof the art solid state NMR on different nuclei probing structure and dynamics with different interactions. The implications for polymer physics and engineering of merchandized PE are obvious as irregular branching along the chain9, 104, 105 results in the possibility of all the different motions identified above to occur in the same polymer.
98 CHAPTER 5 MORPHOLOGY AND DYANM ICS BY DEUTERIUM NMR LINESHAPE ANALYSIS 5.1 Introduction Despite its structural simplicity, polyethylene (PE) is a remarkably versatile material. Branching explains its versatility While previous studies of branching in PE have been limited to relatively illdefined primary structures, the methodology now exists to synthesize polyolefins possessing uniformly spaced branches of a specified type, yielding an unequivocal pr imary structure. Here, we present the first variable temperature deuterium NMR study of CD3 branching groups in precision PE, the intent being to understand the effect of precision branch placement on morphology and chain dynamics. Since 2H NMR spectra ar e extensively dominated by the quadrupole interaction, it is wellsuited to study chain segmental dynamics in bulk. The selectively labeling of methyl branches offers a good opportunity to utilize deuteron as an effective indicator of the polymer main chai n motion by studying the C CD3For methyl branched ADMET precision polyethylenes, it has demonstrated that the methyl branches are incorporated in the crystalline phase. Therefore, for such semi crystalline polymers, th e magnetization from solid echo experiment, in fact, contains signals from both crystalline region (CR) and noncrystalline region (NCR), which makes the lineshape analysis even more troublesome. Achieving signals from mostly pure homogeneous regions is in need. In this study, it can be realized by T bond. 1 filtration experiment The quadrupole echo spectra lineshapes are analyzed by least squares fitting to deuterium motion simulations.
99 5.2 Theories of Lineshape Analysis Deuterated methyl groups are advantageous for 2 2 2 03 c os 1 s i n c os 2 H NMR spectroscopy since even at temperatures below the glass transition, where the polymer chain motion is quenched, the fast methyl rotation averages the nuclear quadrupole interaction of the methyl deuterons In the secular approximat ion, the transition frequencies are given by (5 1) where and are the motionally averaged quadrupole coupling and asymmetry parameter, respectively, and and are the Euler angles relating the lab frame to the principal axis system of the (motionally averaged) quadrupole coupling tensor. The fast rotating CD3 / 3 mimics a C D group with a scaled quadrupole coupling constant and 0 This results in a threefold greater signal and a threefold reduction in the frequency dispersion (in polycrystalline samples), yielding a substantial NMR sensitivity enhancement compared to a single static deuteron with quadrupole coupling constant Deuterium lineshapes are very sensitive to motions, as shown in Figure 51. When at low temperatures, the segmental chain motions are frozen, result ing in the classic Pake pattern (see blue line structure) with characteristic doublet splitting. At elevated temperatures, by absorbing thermal energy, the chains lose part of the motional anisotropy and result in motionally averaged lineshapes. For instance, when a threebond kink motion is involved, the 2H NMR shows a characteristic lineshape with the appearance of a central peak (see Figure 71 (b)). When more chains get involved in the segmental motions, the lineshape varies correspondingly. With such high sensitivity,
100 deuterium NMR lineshape analysis becomes very attractive for investigation of segm ental dynamics within the NMR timescale. Figure 5 1 Theoretical 2H lineshapes in static and fast motional limits : (a) the blue spectrum corresponds to the static 2 H lineshape; the green one indicates the kink motion; the red one refers to the crankshaft motion. (b) 3 bond kink motion. (c) 5 bond crank shaft motion. 5.3 Experimental Deuterium NMR measurements were performed on a Bruker Avance NMR spectrometer operating at 61.4MHz (9.4T). Temperaturedependent solid echo experiments used the 90 90 pulse sequence with = 25 and 75 s, a 90o pulse width of 2. 2 s, a recycle delay of 1 .5 s and 512 scans Deuterium spinlattice relaxation time was determined by using inversion recovery pulse sequence with 1s recycle delay, a 90o pulse width of 2. 2 s an echo delay of 30 s and 128 scans. A T1 (a) (b) (c) filter was used to selectively acquire the spectra of the noncrystalline phase. S pectra
101 were acquired by adding a 5 ms delay following the saturation by a /2 pulse train.91 The T15. 4 Results & Discussion filtered spectra were acquired with a 11ms recycle delay, echo delays of either 25 or 75 s, and 16k scans. The temperature ranges from the melt to well below the glass transition temperature. 5 4 1 Deuterium T1Inversion recovery is the most commonly used method to extract s pin lattice relaxation time (T Measurement 1). In this study, the magnetization at the symmetric cent er of th e lineshape plotted as a function of recovery time in Figure 52 to extract T1The experimental data were fit to three different functions for best fit: values. 1. Double exponential function ( EE) 120 = 11 1 + 1 12 1 + 2 (5 2) : where 1 is the initial magnetization from noncrystalline regions (NCRs) when = 0 2 is the initial magnetization from crystalline regions (CRs) when = 0 1 is a constant, 1 is the spinlattice relaxation time of NCRs, and 1 is the 1 of CRs. This function considers the presence of two phases in semi crystalline polymers: crystalline and noncrystalline phases with distinguishable relaxation times. 2 Kohlrausch or stretched exponential function ( K ) 121125 = ( 1 )1 + (5 3 ) : where n is the stretching parameter (when n =1, the usual exponential decay is recovered), which describes the stretched exponential decay in disordered systems. 3. Exponential Kohlrausch function ( EK ) : = 11 1 + 1 12 ( 1 )1 + 2 (5 4)
102 where n b1, and b2 have the physical meanings as in EE and K functions. In fact, EK fu nction is a combination of EE and K functions by taking advantages of both. EK contains not only the normal morphological parameters (crystalline and noncrystalline phases) but also the possible distribution of relaxation times. Figure 52 T1 determinat ion from inversion recovery using various fit functions: (a) 298K, and (b) 118K. EK: Exponential Kohlrausch function. EE: Double exponential function K: Kohlrausch function. Fitting of the deuterium T1 data was performed for spectra at each te mperature using three functions separately. Figure 52 sampled the fit results at two temperatures:
103 298K (a) and 118K (b). It can be seen the bimodal fitting function EK provides the best fit. Double exponential function EE favors the high temperature whil e the Kohlrausch function K works better at low temperature. Fitting at other temperatures (not shown here) also supports this conclusion. Therefore, the bimodal fitting function EK is chosen to extract the deuterium T1 data for precision polymer PE21CD3. T he relaxation time data are shown in Figure 53. As expected, the T1 value arising from the non crystalline region is much shorter than that from the crystalline region: the former is less than 4 0 ms while the latter exceeds 100 ms at temperatures above Tg. Both T1 values undergo a minimum at around 170 K. Figure 53 Spin lattice relaxation data for ADMET precision polymer PE21CD3 at various temperatures. Data were acquired from the EK bimodal fitting function. From the EK bimodal fitting f unction, the stretching parameter n is obtained in addition to the spinlattice relaxation time. Stretching parameter n indicates the width of 50 100 150 200 250 300 50 100 150 200 250 300 350 T1,non-crystallineT1,crystalline Temperature (K)T1(ms) Tm(DSC) Tg(DSC)
104 the distribution of relaxation times : the small er the n the narrower the distribution. Figure 54 shows the Kohl rausch stretching parameter as a function temperature for fits to EK function At temperatures below the glass transition, n reaches a maximum value of about 0.2, indicating a distribution of microenvironments of the methyl bran ch. Above the glass transition, molecular exchange between these different microenvironments can occur on a timescale short compared to T1 Therefore, an average relaxation time is observed. As the temperature is increased, n decreases and approaches zero at the melting temperature. Under these conditions, each spin relaxation becomes effectively homogeneous throughout the noncrystalline phase. Figure 54 Stretching parameter n at various temperatures for PE21 CD3 achieved from the EK bimodal fitting function. Parameter s b1 and b2 have significant physical meaning: the initial magnetization from noncrystalline and crystalline regions. Therefore, the degree of crystallinity Xc 100 150 200 250 300 0.1 0.2 0.3 0.4 0.5 Kohlrausch Stretching Parameter ( n) EK Bimodal Fitting Function Temperature (K)Stretching Parameter, n can be calculated from these two parameters by using the equation:
105 =21+ 2 100% (5 5) In Figure 55, the non crystalline fraction, 1 is plotted as a funct ion of experimental temperature. These results are in agreement with the degree crystallinity acquired from wide angle X ray diffraction measurements (see Figure 56 ) Figure 55 Non crystalline fraction acquired from bimodal function EK at various temperatures for polymer PE21CD3 Figure 56 Degree of crystallinity acquired from WAXS at various temperatures for polymer PE21CD3 100 150 200 250 300 0.2 0.4 0.6 0.8 1.0 Temperature (K)Non Crystalline Fraction -40 -20 0 20 40 60 6 8 10 12 14 16 18 20 22 Degree of Crystallinity, %Temperature, 0C
106 5 4 2 Full y Relaxed and Amorphous 2Herein the segmental dynamics of ADMET poly mer was explored on the 10 H NMR Measurements 7102 s time scale using variabletemperature 2H quadrupol e echo NMR. Typically the temperature range d from the melting temperature to well below the glass transition temperature. For each temperature, spectra were acquired at different echo delay time in order to compare the evolution of quadrupolar relaxation in the spectral lineshape simulations. The fully relaxed experimental spectra cle arly demonstrate that there is a mobile and a rigid fraction in the sample, which have different contributions to the line shape. The more mobile fraction is observed as the narrow signal in the cent er of the spectrum and has a shorter T1The amorphous spectra were acquired by T relaxation time ( c.a. < 40 ms ) 1At two extreme conditions, near melt ing and well below the glass transition, the amorphous and crystalline spectra are not appreciably different At 75 C, the crystalline and amorphous spectra show nearly identical Pake patterns evidenced with horns at 20 kHz The solutionlike narrow peaks at 55 C indicated the isotropic motions are dominant in both phases. However, at any temperature in between, one can distinguish the two spectra easily. A fairly broad f eature in the amorphous spectra begins to appear at 75 C and is apparent at 15 C. Such a feature does not arise in the crystall ine spectra until 15 C when the Pake powder pattern is still dominant At 25 filtration experiments as described above. The crystalline spectra, which are more difficult to measure directly are calculated by subtrac t ing the experimental amorphous signals from the fully rela xed spectra. E xperimentally acquired fully relaxed and amorphous spectra, as well as calculated crystalline spectra at various temperatures are shown in Figure 57.
107 C, the powder pattern in the amorphous spectra collapses to a broad peak which gets narrower at elevated temperatures due to accelerated segmental motion. Figure 57 Plot of temperaturedependent fully relaxed (experimental in black ), amorphous (experimental in red ), as well as crystalline spectra (calculated, in blue) From the separated spectra shown in Figure 57 one may conclude that the rigid crystalline component having a longer T1 oC 55 35 25 15 5 5 45 15 25 35 45 55 65 75 relaxation time has a great contribution to the horns. The mobile amorphous component gives birth to the central broad peak The kHz kHz
108 next step is to use deuterium NMR line shape simulations to characterize the segment al motions in the crystalline and noncrystalline domains. 5 4 3 Theoretical Lineshape Fitting To simulate the motions in CD3 branched PE, we look to the NMR literature of PE, where 1H, 2H and 13C NMR have been employed to study chain dynamics in both the amorphous and crystalline phases.126130 V arious processes have been considered, including two site rotational jumps, a special case of which is the well known 180 flip flop motion associated with the translational diffusion along the chain axis, and rotational oscillations Although flip flops cannot be directly detected by 2H quadrupole echo NMR,131 their occurrence, together with small angle rotational oscillations of 8 were first deduced from a 2nd moment analysis of the proton lineshape in crystalline PE at 100 C.128130 Flip flops have also been confirmed by 13C NMR.131 An early 2H NMR study reported an increase in the amplitude of rotational oscillations from 5 to 12 over the 40100 C temperature range ( Tm=12 3 C ).127 VanderHart reported a narrowing of the 13C -13C dipolar satellites in crystalline PE at temperatures above 100 C consistent with much larger rotational motions .132 However, no distinction between continuous or discrete jump models c ould be made from this data. Rotational oscillations can be classified as harmonic, statistical or Gaussian, where the probability for angular displacements from the equilibrium positions in the crystal lattice is given by a normal distribution.128130 In addition, rotational diffusion about the chain axis should also become a plausible motional mode when the potential energy barrier to rotation is low, such as in a rotator phase,116 in the amorphous phase,114 or near the melting point in the crystalline phase of t he precisely branched polymers studied here.
109 Fits to the 2H SSNMR spectra of PE21 CD3 were performed using Eastmans Deuterium Fitting Program (DFP).133Since the amorphous spectra contain mostly highly mobile signals, it is reasonable to speculate the motion s to be nominally isotropic in character In the Dif C CD3 bond undergoes rotational diffusion which is modelled by discrete jumps between nearest neighbor vertices of a polyhedral approximation to a sphere. The results of fitting of the experimental spectra of the non crystalline domains using various motional models are shown in Figure 58. Single Dif versus Dif + Dif (two discrete rotational diffus i ve motions with different rates) model are compared. It is clear that the Dif + Dif model provides a better fitting, evidenced by the nearly overlapping experimental and simulated lineshapes The goodness of fit parameter, chi square as shown in Figure 5 9, offers quantitative support of this conclusion. The rather small chi square value of Dif + Dif model indicates the correctness of this double diffusive model. Recalling the rather small values of stretching parameter n related to the distribution of spin relaxation, this double diffusion model is consistent with structur al heterogeneity in the amorphous phase. Such heterogeneity is inevitably averaged at elevated temperatures. Even the fits by the single diffusion motion Dif are acceptable at high temperatures, due to the averaged dynamics in the amorphous phase. To explore the dynamic information in the complicated polymer system, s everal motional models are employed and compared: diffusion abbreviated as Dif describes the diffusion of deuterium atoms on a sphere; rotation, abbreviated as Rot describing the rotation about axis perpendicular to C CD3 bond; and kink motion, denoted as Kink describing the three bond kink motion (refer to Figure 51)
110 Figure 58 Fitting of experimental amorphous spectra using various motional models: single diffusion model (left) and two diffusion models. Echo delay is 25 s. The experimental spectra are in red and the fits are in blue. Polymer sample is PE21 CD3 oC 55 35 25 15 5 5 45 15 25 35 Dif 25 s, AR Dif + Dif 25 s, AR
111 Figure 59 Chisquare value as a function of fitting models for amorphous spectra. Model 1 denotes a single Dif is Dif + Dif is Dif + Dif + Dif is Dif + Dif + Dif + Dif 0.000E+00 5.000E+04 1.000E+05 1.500E+05 2.000E+05 2.500E+05 3.000E+05 3.500E+05 4.000E+05 1 2 3 4 Motional modelsChi square
112 Figure 510. Fitting of calculated crystalline spectra using various motional models: diffusion+rotation model (left) and two diffusion models. Echo delay is 25 s. The experimental spectra are in red and the fits are in blue. Polymer sample is PE21 CD3 Figure 510 shows the fitting to the calculated crystalline spectra. Inspired by the simulation results of amorphous spectra, fitting models involving more than one type of motional models are compar ed: Dif + Rot (diffusio nal and rotational dynamics with oC 55 35 25 15 5 5 45 15 25 35 Dif + Rot 25 s CR Dif + Dif 25 s CR
113 different rates and fractions) versus Dif + Dif (two diffusional dynamics). As expected, the former model offers slightly better fits compared to the latter one, indicated by the smaller chi square values The fitting results to the CR phase indicate that the motion in the chains can be described by rotational diffusion about the chain axis and isotropic diffusive motion of the C CD3 bond. The necessity to include a Dif component could arise from the inability to achieve a complete separation of the noncrystalline and crystalline subspectra due to the occurrence of the interphase or due to the tail in the distribution of T1Based on the linesha pe simulat ion results of the amorphous and crystalline spectra, the fully relaxed spectra can be readily fit into a threecomponent model: Dif + Dif + Rot Figure 5 1 1 shows the fitting results. relaxation times for the amorphous phase. The fits (in red) are superimposed on top of the experimental spectra (in black). As the temperature increases through the glass transition, the onset of dynamics becomes evident with the broadening of the singularities of the Pake pattern, with intensity building in the center. Changes in the 2H spectra are consistent with the thermodynamic data indicating that the CD3 branches introduce disorder. Upon further increasing the temperature in PE21CD3 a narrow liquidlike Lorentzian shaped peak appears, representing the relatively mobile regions of the amorphous phase.
114 Figure 5 1 1 Fitting of fully relaxed spectra using a combination of three motional models: diffusion+diffusion+ rotation model. Echo delay is 25 s. The experimental spectra are in red and the fits are in blue. Polymer sample is PE21 CD3 oC 55 35 25 15 5 5 45 15 25 35 Dif + Dif + Rot 25 s Full
115 5.4 Conclusion To conc lude, deuterium solid state NMR has been proven to be well suited in investigation of conformational dynamics. By applying T1 filtration, the amorphous spectra can be separated from the fully relaxed ones. The crystalline spectra can be simply calculated by subtracting the amorphous signals from the total. Lineshape simulation by using combinations of motional models provides new insights into the microscopic details of the segmental dynamics in PE. The experimental lineshapes are found to be compatible wit h a model in which rotation about the local chain axis occurs by a diffusive process. However, other modes of rotational motion i.e., two site jumps or harmonic oscillation still cannot be excluded, and such processes could involve either specific rota tion angles or distributions. Thus, further studies with selectively labeled, precisely branched polyethylene, are needed to more fully characterize the chain dynamics in these systems.
116 CHAPTER 6 EFFECTS OF BRANCH IDENTITY ON THE MORPHO LOGY OF PRECISION POLYOLEFINS WITH ALKYL BRANCHES 6.1 Introduction The densities of commercial PE based materials are determined by the size of the branches low density of LDPE is due to the presence of both short chain and long chain branches while relatively high density of LLDPE comes from the existence of only short chain branches. Meanwhile, the density itself origins from the different packing behavior which in turn is constrained by the size of the branches. For ADMET polymers, t he size of branches also counts. Herein the influence of branch identity on several properties including thermal property and morphology is discussed. ADMET precision polyethy lenes with various alkyl branches R on each and every 21stTable 61. Basic characterization data of PE21R, including the M carbon on the backbone, denoted as PE21R, are sampled. w data and the thermal analysis results29, 134 Polymer Branch identity Mw, a PDI g/mol T b m H C m X J/g c c ADMET PE % H 15 k 2.6 134 204 70 PE21 Methyl methyl 20.2 k 1.7 63 104 35 PE21 gemMethyl gem methyl 76 K d 45 62 21 PE21 Ethyl ethyl 50.2 k 1.9 24 65 22 PE21 Propyl propyl 41.4 k 1.7 12 60 20 PE21 isoPropyl iso propyl 46.0 k 1.7 11 37 13 PE21 Butyl butyl 4 0.3 k 1. 7 12 57 19 PE21 secButyl sec butyl 42.6 k 1.9 9 43 15 PE21 tertButyl tert butyl 32.1 k 1.7 13 50 17 PE21 Pentyl pentyl 45.8 k 1.8 14 58 20 PE21 Hexyl hexyl 46.1 k 1.7 12 49 17 PE21 cycloHexyl cyclo hexyl 3 3 6 k 1.6 9 37 13 PE21 Adamantyl adamantyl 64.7 k 1.7 8 & 17 2 & 8 0.1 & 2 a Mw acquired by GPC in THF (40 C) referenced to PS b P olydispersity index = Mw/Mn c D e gree of crystallinity = m / 293 J/g 135 d Number averaged data: Mn from literature134
117 The synthesis of this series of polymers is mature and has been reported in literatures.26, 28, 29, 31, 34, 37 The polymers studied in this chapter are mainly synthesized via the cyanide chemistry because of its short steps and quantitative yield.28, 37Table 61 lists the ba sic characterization data of PE21 R, including M w and thermal data. Note that the weight averaged molecular weight of each PE21R polymer is larger than 15,000 g/mol, after which the molecular weight expose negligible influence on polymers behaviours.266 2 Thermal Properties The thermal analysis of precision ADMET polymer series PE21R was accomplished via TGA and DSC.28, 29 Plot the melting temperature data listed on Table 6 1 as a function of the branch identity, seen in Figure 61 At the first glance, one can easi l y find that the Tm = 1 2 (6 1) decreases when introducing methyl and ethyl branches, surprisingly levels o ff when branch identity switches to the propyl group. This level off indicates the melting temperature is no longer affected by the branch identities from the poi nt of the propyl group. One realistic explanation will be that the bulkier groups are no long er present in the crystalline phase. In fact, this hypothesis can be reasonable, according to the famous Thompson Gibbs equation: w here Tm denotes the real melting temperature of a polymer, Tm o is the ideal equilibrium melting temperature, e is the free energy for each unit area of the fold surface of the c r ystal, l is the crystal thickness, f is the enthalpy of the phase transition from melt state to crystalline phase. In general, this rule tells us the actual melting temperature is proportional to the thickness of the crystal. The similar melting
118 points of bulkier group branched PE21R indicat es the similarity in crystal sizes of these polymers. If the bulkier branches were incorporated, the crystal size would have been changed with the size of branches, and consequently the melting temperature will change as well. Nevertheless, this is not the fact. Melting temperatures of PE21Propyl, PE21 Butyl, PE21 Hexyl, ect. Remain constant. It seems that the bulkier branches are indeed cannot exist in the crystalline phase. X ray and SS NMR in the next two sections will be employed to investigate this h ypothesis. PE Methyl Ethyl Propyl Butyl Pentyl Hexyl iPropyl sButyl tButyl cHexyl 280 300 320 340 360 380 400 420 Tm (K)Branches on Every 21st Backbon Carbon Figure 61 Plot of m elting temperature as a function of branch identity Tm was acquired by DSC at a rate of 10 C/min. 6 3 Branch Displacement Determination by WAXS Thermal analysis displays that the alkyl groups larger (including) propyl group on longer affect ADMET polymers melting temperature. Base on that observation, a hypothesis is proposed, saying that the bulkier alkyl groups are excluded from the crystalline regions. Therefore, changing the branch identit y doesnt change the chain
119 packing behavior in crystalline phase, and thus has no effect on the melting temperature. WAXS is a powerful technique in studying crystal structures and is applied herein to investigate the branch displacement in 3D structure. 6 3 1 Experimental Wide angle powder X ray diffraction (WAXD) data were obtained on a Bruker D8 diffractomer equipped with liquid N2 cooling system, using copper K radiation with The solid polymers (Tm > room temperature), such as linear ADMET PE and methyl branched PE, were filled into a copper container of 10 by 10 mm2 with a depth of 1mm. Prior to measurement, these solid samples were melted to remove any thermal history. As for liquid polyolefins (Tm6 3 2 Results and Discussion < room temperature), i.e., polymers with br anches longer than the ethyl group, samples were put on top of a flat copper substrate to form a thin layer with ~1mm thickness. During the measurements, in order to reduce scattering from air and avoid chemical degradation by oxygen at elevated temperatur es, medium vacuum was applied to the sample chamber. XRD data were collected at the to 45 with a step of 0.1 /min. Temperature dependent measurements were addressed at temperatures varying from below glass transition temperature to above melting points, with cooling or heating rate of 1 C /min. Wide angle X ray diffraction (WAXD) measurements further support the observation of a change in polymer morphology as a function of branch size. Six such WAXD diffractograms are shown in Figure 6 2 ; these patterns indicate that the introduction of branches leads to the lattice distortion and local conformational disorder,
120 where the type of crystal structure and polymer morphology is s trongly dependent on the branch identity. Figure 62 WAXS results of PE21 R, reference d to linear ADMET P E : ADMET PE at 27 oC (no branch, Tm= 134 oC), PE21 M ethyl at 27 oC (methyl branch, Tm= 63 oC), PE21 E thyl at room temperature ( ethyl branch, Tm= 24 oC), PE21 P ropyl at 0 oC (propyl branch, Tm= 12 oC), PE21 B utyl at 0oC (butyl branch, Tm= 12 oC), PE21 sec B utyl) at 0 oC (sec butyl branch, Tm= 9 oC), PE21 P entyl) at 0 oC (pentyl branch, Tm= 14 oC). Prior to measurements, all samples were heated to above melting temperature in order to remove thermal history, and then cooled to specific temperature at a rate of 1 o C /min. For the sake of comparison, ADMET PE is displayed at the bottom of Figure 6 2 exhibiting the typical orthorhombic crystal form wit h two characteristic crystalline peaks 10 20 30 402 Intensity ( a.u .) ADMET PE methyl propyl ethyl butyl sec butyl pentyl
121 superimposed with the amorphous halo, exactly the same as for highdensity polyethylene made by chain propagation chemistry. The more intense peak at scattering angle 21.5 and the less intense one at 24.0 correspond to reflection planes (110) and (200), respectively. Upon introducing precisely placed branches of known identity, the crystal structure loses its symmetry with the unit cell shifting from orthorhombic to triclinic. Moreover, in contrast to linear polyethy lene, scattering occurs at relatively lower scattering angles and with broad reflections being displayed, suggesting the decrease of crystallite size and crystallinity. Figure 63 Plot of scattering angle of two strong reflections as a function of branch identity The top line is for the reflection at higher angle while the bottom one is for the reflection at lower angle. In the case of the methyl branched polymers PE21 M ethyl two reflections representing a triclinic crystal orientation occur at scattering angle 19 with the Miller index (100) and 22 with the Miller index (010). Transmission electron microscopy (TEM) no branch methyl ethyl propyl butyl sec-butyl pentyl16 18 20 22 24 26 Scattering Angle (degree)Branch Identity Higher angle Lower angle
122 shows the lamellar thickness to be quite small, between 10 and 20 nm Since such a thickness is much larger than the length of 20 C H2Similar changes in crystalline unit cell identity were observed in the ethyl branched polymer, PE21 E thyl Two strong reflections shift to lower scattering angles (18 and 21) compared to the methyl branched polymer (19 and 22). On the basis of Braggs law, this observation allows us easily to draw the conclusion that the ethyl branch is incorporated in the crystal region. In contrast to the PE21 M ethyl the introduction of the ethyl side chain perturbs the crystal structure more and requires larger space to be incorporated. As a consequence, the reflections shift to lower scattering angles, which correspond to larger d spacing. For the polymers possessing bulkier branches (propyl or larger), the WAXD diffractograms (the top four graphs in Figure 6 2 ) show nearly identical scattering patterns, indicating that the crystal structure is independent of the branch identity. Moreover, these patterns are obviously different from those of polymers possessing smaller branches like methyl or ethyl branches, exhibiting larger scattering angles and even broader reflections. units, indicating that the methyl side chains are incorporated in a triclinic lattice. To understand these observations, the scattering angles of two strong diffraction peaks as a function of branch identity are illustrated in Figure 6 3 We see two trends: the decrease of scattering angles for smaller branches, followed by leveling off of the scattering angles for bulkier branches. The former trend is easily understood. The methyl/ethyl side c hains function as defects in the crystal lattice and result in the increase of d spacing, thus decreasing scattering, where the ethyl results in a smaller scattering angle. Distinct difference is the case for the bulkier branches, where the
123 angles for bulk ier branchedpol ymers, PE21 P ropyl PE21 B utyl ), PE21 P entyl and PE21 sec B utyl increase to higher degrees, 19.5 and 22.5. Recall the X ray scattering theory: scattering angles are determined by the type of unit cell while the peak intensities are based on the atom arrangements within the lattice. With this in mind, we conclude that the packing behaviors of precision polymers possessing sizes equal to or larger than propyl are different from those of the methyl and ethyl branched polymers. In all cases, the bulkier branch is excluded from the unit cell into the amorphous region. The morphology and packing behavior of these precision polymers with bulky branches are independent of the size of the branch. This assessment is in agreement with the thermal dat a in Table 6 1 where essentially identical melting temperatures are observed for all the branched polymers. Note that the reflection occurring at 19.5 is much broader and less symmetric than peaks for the methyl branched polymer PE21 M ethyl or ethyl bra nched polymer PE21 E thyl The asymmetry suggests the presence of more than one crystal lattice besides triclinic; broadening is due to a decrease in the degree of crystallinity. Of particular interest is the fact that the WAXD scattering patterns for these precision polyolefins are noticeably different from those for randomly branched polymers: the precision structures result in varying crystal lattice identity and relatively sharper scattering peaks.10, 32, 136138 The random ethylenepropylene (EB) copolymer with 20% comonomer content exhibits a dominant hexagonal crystal form,136 compared to the triclinic form present in our precision PE21 E thyl polymer. As for the random ethylenebutene (EB) copolymer, it shows orthorhombic crystal structure with low crystallinity.136, 137 In contrast, the precision ethyleneoctene (EO) polymer PE21 H exyl
124 exhibits an additional hexagonal mesophase besides the orthorhombic crystalline phase.1386 4 Investigation of Dynamic by SS NMR Th e X ray investigation for precisely sequenced polymers with alkyl branches has concluded that the smaller groups can be incorporated in the unit cell while the larger ones are excluded from the unit cell. The cut off of the branch identity is the propyl grou p, w hose size is too large to be integrated in the crystalline phase. 136 4 1 Expe rimental C SS NMR measurement leads to same conclusion as WAXS based on the knowledge that segments of polymer chains move slower in crystalline region than them in amorphous phase. The polymer samples are the same as used in X ray studies, i.e., ADMET precision polyethylenes with alkyl branches. The synthesis of these polymers has been described in section 6.1.1 as well as literatures.28, 29 SSNMR was carried out at Max Planck Institute for Polymer Research (MPIP), Mainz, Germany, utilizing a Bruker spectrometer equipped with a Bruker Avance II+ console working at 1The segmental dynamics were studied via REREDOR pulse sequence, a sensitive approach to det ermine H Larmor frequency of 850 MHz. Polymer samples were tightly and evenly packed in 2.5 mm rotors and measured in 850 MHz spectrometer using 2.5 mm MAS H X double resonance probe, which is suitable for NMR experiments under fast MAS conditions. 1H -13C dipolar decoupling constants by spinning sideband pattern analysis. The detailed introduction of this modern SS NMR technique can be found in section 126.96.36.199.
125 6 4 2 Results and Discussion The slices from REREDOR experiments are shown in Figur e 64 The peaks are assigned and highlighted with different colors. For methyl branched polymer, the methyl branch marked in yellow shows up at 20.3 ppm; the peak at 28 ppm is contributed from 33.8 ppm represent the alltrans CH2 2 amorphous phase, respectively; the small peak appearing at 35 ppm (in green) can be assigned to the methine carbon. I nterestingly, a clear splitting of chemical shift occurs for carbon at around 40 ppm and 38 ppm; moreover, the ratio of the peak heights is c ( bon in crystalline region, 40 ppm ) aThe dipolar recoupling constant, D ( region, 38 ppm ) crystalline and amorphous phases, an observation in agreement with WAXS data. The spinning sideband analysis als o supports this conclusion, seen in Table 62 ij of c by simulation is determined to be 14 kHz, corresponding to a slow motion, while the Dij of a is simulated to be 7.5 kHz, implying a fast motion in amorphous phase due to less structural constrains from surrounding atoms. Note the Dij of c a are 14 kHz and 6.5 kHz separately, which confirm the correctness of peak assignment and reverse relati onship between Dij It is worth mentioning the difficulty in simulating the spinning sideband of methyl branch for PE21Methyl. By all means, the sideband pattern of methyl group is impossible to fit by any single dipolar coupling constant. However, when fitting it to one large and one small D and segmental dynamics. ij values, the problem is solved. Combined with the results from
126 methyl branches are present at both crystalli ne and amorphous phase. Figure 64 Peak assignment of slices from REREDOR spectra for (a) methyl branched and (b) propyl branched ADMET precision polymers As for the case of PE21Propyl, seen in Figure 64 (b), the peak assignment are as follows: methyl group in the propyl branch (carbon 1) shows up at 15 ppm (in yellow); CH2 15 20 25 30 35 40 ppm PE21 -CH3 c a c a a group in the propyl branch (carbon 2) appears at 20.5 ppm (in orange); the phas e (in blue); the two strong peaks at 30.5 ppm and 33.5 ppm correspond to the all15 20 25 30 35 40 ppm PE21 -C3H7 c a a a (a) (b)
127 trans CH2 2 amorphous phase, respectively; the trace peak shows at around 35 ppm may comes from the methine carbon (in green) or carbon 3 (CH2 of the propyl branch). Apparently both of them are overlapped under the strong peak of crystalline alltrans CH2The simulation results of the spinning sideband patterns for PE21Propyl is shown in the second half of Table 62 Backbone CH groups. 2 carbons display two dipolar decoupling constants of 14 kHz and 6.8 kHz, corresponding to the alltrans CH2 in the crystalline region and gauge conformation of CH2 in the amorphous region. The DijTable 6 2 Recoupling constants derived from REREDOR experiments for methyl branched and butyl branched ADMET precision polymers with branch spacer of 21. 38 ppm is determined to be 6.5 kHz, a typical value indicating fast motion. So are the values of carbon 1 and 2: 6.5 and 3.7, indicating the presence of the propyl groups only in the amorphous phase. polymer site D ij S kHz Conclusion ij PE21 Methyl c 14 ( ) 0.7 slow, CR a 7.5 ( ) 0.35 fast, AR c ( backbone CH 2 14 in CR ) 0.7 slow, CR a ( backbone CH 2 6.5 in AR ) 0.3 fast, AR 1 c/a ( CH 3 6.8/3.5 in CR/AR ) ~0.2 fast, AR / slow, CR PE21 Propyl a 6.5 ( ) 0.3 fast, AR c ( backbone CH 2 14 in CR ) 0.7 slow, CR a ( backbone CH 2 6.8 in AR ) 0.3 fast, AR 2 (CH 2 6.5 of propyl group) 0.3 fast, AR 1 (CH 3 3.7 of propyl group) <0.2 fast, AR CR: crystalline region AR: amorphous region The REREDOR experiments for PE21Ethyl and PE21Butyl result in the same conclusion as achieved from that for PE21Methyl and PE21Propyl smaller branches
128 including the methyl and the ethyl groups are integrated in the ordered 3D array, while the larger on es such as the propyl and the butyl groups are found in the noncrystalline phase only. 6 5 Morphological Models Combin ing the results from X ray and SS NMR, one can conclude the branch displacement for ADMET precision polyolefins exampled by PE21R series illustrated by the flow chart in Figure 65 Figure 65 Flow chart of Branch displacement in ADMET precision polyolefins PE21 R where R = alkyl branches on each and every 21st carbon along the ethylene backbone For high crystallinity linear ADMET polyethylene, i.e., when branch R is in fact the hydrogen atom, the crystal structure deciphered via WAXS is claimed to be orthorhombic, same the crystal form as for commercial PE based structures. When
129 small alkyl branc hes, including the methyl and the ethyl branches are introduced, the symmetric 3D array of crystal structure is disturbed by the defects and shifted to triclinic crystal structures due to the incorporation of small side groups in the ordered region. Situation changes when the branches get bigger. The propyl group is the turning point. The alkyl branches as big as or larger than the propyl groups cannot be present in the crystalline region due to the oversized perturbation, seen in Figure 65. Therefore, f or large alkyl branches, the polymer chains fold to form lamellae devoid of large defects : it is the alltrance polyethylene segments that pack into threedimensional ordered array, just as the scheme of morphology illustrated in Figure 66. However, it is noticeable that the bulkier branched ADMET polyolefins obtain different unit cell parameters than the unbranched ADMET PE. Based on the schematic morphology of large defect branches, the size of the lamellae should be determined by the branch spacer, if the polymers crystallize. For instance, if the branch spacer is 21, the lamellae thickness will then be limited by the length of 20 CH2 sequence in between of two large defects. In fact the reality should be a little off from the hypothesis, if the substantial size of the branches present at the interphase has to be considered. Due to the connection of polymer structures, the large alkyl branches at the interphase will inevitably disturb the nearby crystal segments orientation and packing. T hus, to maintain the ordered structure, the polymer chains will sacrifice some segments by transferring them from the crystalline phase into the interphase, which will result in a lamellar thickness a little bit smaller than the length of CH2 sequence betw een two branches. In principle, the bigger the branches are, the large this shrinkage will be.
130 Moreover, since the melting temperature is directly related to the lamellae thickness for semi crystalline polymers, one can comment that the melting point of PE21R series is determined by branch to branch distancein this case: 20 CH2 sequences. The fact that branches as big as or bigger than the propyl groups are placed in the noncrystalline regions perfectly explains the similar melting temperatures of PE21 Propyl, PE21 Butyl, PE21 Hexyl, and other PE21R polymers. Figure 66 Schematic morphology of precision polymers with large defect branches
131 CHAPTER 7 ADMET PRECISION POLY ETHYLENE WITH LOW BR ANCH FREQUENCY 7.1 Introduction As we have demonstrated, the size of the branch as well as the branch content significantly affects the microstructural, thermal, dynamical, morphological properties of the ADMET polyolefins. The manipulation of branching provides a n applicable way of modeling the realistic industrial PEbased materials. With high branch content, such as one branch on each and every 21st polyethylene carbon, ADMET polymers are capable of mimic king the structure of linear low density polyethylene. However, for an attempt of understanding metal locene PE, rather low content of branching is needed.24, 139145 Recently, a precision ADMET polyethylene possessing quite low frequent brancha butyl branch displaced on each and every 39th backbone carbonwas made successfully,367. 2 Experimental offer ing an applicable methodology to produce precision models with up to date maximum precision run lengths between adjacent branches Precision polyethylenes with branches on each and every 39th backbone carbon, denoted as PE39R, were synthesized via polycondensation reactionADMET chemistry, followed by exhausted hydrogenation. The key to the successful synthesis involves the prepar ation of the symmetric diene monomer with reduced branch frequency. Deuterated and protonated methyl branched precision monomers for polymers PE39 CD3 and PE39CH3 were made via the two methodologies described in Chapter 2. Specifically, the synthesis of these tw o precision monomer s used herein w as illustrated in Figure 71. The important limitation of this synthesis lies in the
132 production of commercial unavailable 18spacer alkenyl bromide,36 which can be accomplis hed by dehydrohalogenation of alkyl dibromide.36, 146 Figure 71. Synthetic methodology used to produce precision methyl b ranched (18,18 ) monomers: the top reaction was used to make trideuterated monomer while the bottom one was for the protonated monomer. Once the pure monomers with desired run length are produced, the precision polymers of interest PE39 CD3 and PE39CH3 can be readily afforded by acyclic diene metathesis polycondensation chemistry and the consecutive hydrogenation under Wilkinsons [ Rh ] catalyst. The butyl branched ADMET precision polymer, PE39C4H9, was prepared based on procedures in a recent literature.36 Table 71 lists the basic characterization data of PE39R, with unbranched ADMET PE as reference. PE39 CH3hmw was the high molecular weight version of PE39CH3Th ermal analysis was conducted via a TA Instruments d ifferential scanning calorimetry (DSC) Q1000 equipped with a controlled liquid N It was a chieved via sox h let extraction by removing the low molecular weight parts into toluene solution. 2 cooling setup at a heating/cooling rate of 10 C/min. Molecular weight information was acquir ed by gel permeation chromatography (GPC) in trichlorobenzene (TCB) at 135 C with reference to polystyrene standard. Solid state NMR measurement was performed utilizing the 850 and 700 MHz spectrometers at fast magic angle spinning. 2D X ray data were obtained
133 using a rotating anode (Rigaku 18 kW) X ray beam with a pinhole collimation and a Siemens 2D area detector. Cua double gr aphite monochromator. Except for the thermal analysis, GPC, SS NMR, and WAXS were all measured at Max Planck Institute for Polymer Research, Mainz, Germany. Table 71. Molecular weight and thermal data for ADMET precision polyethylene with branch spacer of 39, referenced to linear ADMET PE. Polymer Branch Identity M w x10 3 PDI g/mol T m H C m X J/g c ADMET PE % n/a 68.4 2.66 13 4 204 70 PE39 CD CD 3 39.7 3 2.16 88,91 124 43 PE39 CH CH 3 35.1 3 2.39 87, 91 175 60 PE39 CH 3 CH hmw 92.7 3 1.97 9 2.5 134 46 PE39 C 4 H C 9 4 H 67.0 9 1.44 a 75 a 66 22 a: molecular weight of unsaturated ADMET polymer36 7. 3 Results and Discussion 7 3 1 Thermal Behavior It has been evidenced that for semi crystalline precision ADMET polyolefins, the melting temperature obeys an inverse relationship with the branch frequency. The more the branches are introduced, the more defects in the polyethylene microstructures will affect the segmental chains packing into ordered arrays. When the branch frequency exceeds certain limitation, due to the lack of controlled stereochemistry, alkyl branches exhibit random orientation relative to main chain. This heterogeneity of side chain orientation consequently hinders the form of crystalline domains especially when the run lengths between adjacent defects are so small that packing into ordered structure results in unfavoured conformation with high internal energy Thus, totally amorphous polymer forms. This is t he case for PE5 CH3 and PE5C4H9, where there is no
134 crystalline regions exiting. In contrast, with some higher frequency of branches, PE7CH3 and PE15C4H9 0 20 40 60 80 100 120 140 160 -100 -50 0 50 100 150 Tm, oCBranches per 1000 backbone carbon Tm, methyl-branched Tm, butyl-branched linear fit, methyl-branched ) linear fit, butyl-branched ) appear semi crystalline behaviours. Figure 72. Plot of melting point as a function of branch content: black squares and red triangles are the experimental data points acquired from DSC at a heating/cooling rate of 10 C/min; linear fit s w ere done via Origin 8.0. Tm data of polymers with lower branch frequency are from literatures.26, 28, 36, 37 Figure 72 shows the correlation of melting temperature and branch content for methyl branched and butyl branched ADMPE precision polyolefins. The experimental data achieved from DSC can be better fit into linear function: = 132 34 1 3456 with correlation coefficient R2 of 0.99480 for PEx CH3, where x is the branch spacer; the linear fit for PEx C4H9: = 142 5012 2 51659 shows not so perfect fitting with R2 of 0.9 3253 due to less data points Apparently, there is a ADMET PE reciprocal relationship of melting behavior and branch frequency which is in agreement with the conclusion drawn from commercial PE. The fitting functions not only give quantitative relation of X=39 X=21 X=19 X=15 X=11 X=9 X=7 X=39 X=21 X= 15
135 melting points and branch frequency, but also can be utilized for prediction. For instance, from the linear line fit equation, one can deduce the melting temperature of PE75 CH3 (when methyl branch is displaced regularly on each and every 75th backbone carbon) to be 114 C, and Tm of PE75 C4H9 to be around 110 C The proximity of these two Tm values can be explained from Figure 72 as well: the two linear fits approach same upper limitationunbranched ADMET PE. The perfect linear ity of melting point as a function of branch frequency for PEx CH3 arises from the fact that the crystalline domains incorporate pendant methyl branches as defects. The atactic methyl branches enforce the ideal alltrans stem conformation to deviate into conf ormationally disordered crystal structures. Previous study on PE21CH3 and PE15CH3 document s that th is deviation in turn shortens the chain stem s and appears stronger in the lower frequently methyl branched precision polymer, PE15CH3.99 The shorter chain stems due to the introduction of more frequent branches then give rise to correspondently thinner lamellar structures, which appear in thermal behavior low er melting points. As for the precision polyolefins with bulkier branches, PEx C4H9, since it has been demonstrated that the butyl branches are present only in the noncrystalline regions, the lamellar thickness is limited by the run length of chain stem in between of two adjacent butyl branches. Therefore, Tm = 1 2 ( 7 1) of butyl branched precision polymer is more directly related to the branch spacer i.e. the chain length of alltrans conformation in crystalline phase. Based on Thompson Gibbs equation listed in chapter 6
136 for the family of PEx C4H9, one can reasonably assume that Tm o (ideal equilibrium melting temperature) e ( free energy for each unit area of the fold surface of the cr ystal ) and f (enthalpy of the phase transition from melt state to crystalline phase) remain constant regarding the difference in branch spacer. Thus, real melting point Tm can be correlated to lc = (7 2 ) (crystal thickness) in such a formula: 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -40 -20 0 20 40 60 80 100 120 140 Tm, oC1/lc, nm-1 Tm Linear fit of Tm Figure 73 Correlation of melting temperature and lamellar thickness for a series of butyl branched ADMET polymers. The lamellar thickness of unbranched ADMET PE was assumed to be 12 nm, value adopted from the ultra high molecular weight PE.23, 147 L inear fit formula is : = with R2 = 0.986. The length of alltrans chain stem in between two adjacent branches can be calculated from the C C bond length of 1.54 and the CH2CH2CH2 PE15 C 4 H 9 tetrahedral angle of 109.47 According to the morphology model proposed in Figure 6 6 for ADMET precision polyolefins with bulky alkyl branches, one can easily plot the experimental lc = 1.8 nm PE 21 C 4 H 9 lc = 2.5 nm PE 39 C 4 H 9 lc = 4.8 nm ADMET PE lc = 12 nm
137 melting point as a function of reciprocal of calculated lamellar thickness 1 11T, as shown in Figure 73. The linear correlation of melting points and the reciprocal of lamellar thickness further prove the reality of our morphology model for ADET precision polymers with bulky alkyl side chains. The bulk branch is, indeed, not present in the crystalline domain; the lamellar thickness is limited by the length of ethylene sequence between two branches. 7 3 2 Solid State NMR Investigation Precision polymers with reduced branch frequency, PE39R, behave quite different ly from those with high branch frequency, not only on the thermal behavior but also on how the polymer chains pack into ordered structure and how the branches orientate in term s of the backbone. Despite of the low natural abundance, 13C as an interesting nucleus has been proved to be very useful especially when coupled with a high magnetic field. Here in 13C SSNMR under either CP condition or single pulse excitation was applied on methyl branched polymer s PE39 CH3, PE39 CH3hmw, and PE39C4H9T he polymer sample PE39 CH under fast MAS conditions. The methyl branched ADMET polymers show some unique observations and will be discussed in details. 3 was heated to above its melting temperature to remove any thermal history present. The 13C spectrum of the melted polymer is shown in Figure 74. The narrow solutionlike peaks arise from the accelerated motional averaging at melt state when crystalline structure collapses and the resulting total amorphous polymer exhibits isotropic motions. Based on the chemical shifts and relative intensities, t he peaks are assigned to each group with no difficulty. If calculated
138 roughly from the integrations, one can conclude the degree of polymerization is less than 10, indicating there are quite a few short chains present in the polymer system. Figure 74 13C SS NMR spectrum of PE39 CH3 at melted state : single pulse excitation (SPE) with dipolar decoupling was acquired by a Bruker spectrometer at 1 H Larmor frequency of 850 MHz, using 2.5 mm MAS H X double resonance probe, under 8 kHz magic angle spinning at T= 383 K The top spectrum (in red) is obtained by 10 times magnification of the experimental one ( in black). Slowly cooling down from the melt, temperature dependent spectra were acquired, as shown in Figure 75. The effect of anisotropy interactions can be seen clearly at 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 ppm CH branch CH3all -trans CH2 -1 -2 -3
139 carbons are surprisingly broad, which is not observed for PE21CH3, the ADMET precision polymer with high branch frequency. One possible explanation may arise from the relatively low molecular weight in PE39 CH3 When there is considerably high ratio of short polymer chains present in the sample, individual polymer chain has less constrains in spatial orientation due to the decrease of chain length. As a conseq uence, the pendent branches may obtain increased possibility of steric displacements and chemical environments, displaying as wide distribution of chemical shifts in the spectra. Figure 75 Temperature dependent 13C spectra of PE39CH3 under SPE : cooling from the melt at a rate of 0.2 C/min The colorful spectra are expanded 10 times of the experimental ones (in black). 15 20 25 30 35 40 ppm T=363K T=343K T=323K
140 To discover whether the unusual line broadening is due to the low molecular weight or not, low molecular weight parts in PE39 C H3 is removed by using soxhlet extraction to yield PE39 CH3hmw, with higher molecular weight and narrower molecular weight distribution, as shown in Table 71. Similar experiments were employed on PE39CH3 hmw. Interestingly, the unusual line broadening f or the methyl (see Figure 76). Figure 76 Temperature dependent 13C spectra of PE39 CH3hmw using SPE with dipolar decoupling : cooling from the melt at a rate of 0.2 C/min. The spectra are normalized with reference to the amorphous CH2 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 ppm 333K 380K 370K 365K 360K 355K 350K 343K peak. 29 30 31 32 33 34 35 carbon branch CH 3 zoomed
141 The duplication of line broadening in PE39 CH3Figure 76 shows the hmw abolishes the previous hypothesis of low molecular weight effect. Then, what are the origins of these broadening? Since its known that the methyl branches exist in both the crystalline regions and the noncrystalline regions, the answer to such a question will reveal the chain packing behaviors and the orientations of the methyl branches relative to the backbone. 13C spectra of PE39CH3hmw at various temperatures when cooling slowly from the melt. Beside the unusual line broadening for the branch temperature dependent chemical shifts of the all trans conformation in crystalline regions and the gauche conformation in amorphous regions. It can be seen clearly from the expanded spectrum that the alltrans crystalline and the gauche amorphous peak move to each other when cooling i.e., the chemical shift of crystalline CH2 decreases at lower temperature while the trend for noncrystalline CH2 Since the chemical shift is very sensitive to the chemical environment for the nucleus at site, such deviations remind us that the complex crystall ine structures are possible to be represented in SS NMR by such a simple parameter as chemical shift. is the opposite. The quantitative data of these dev iations are shown in Table 72. The correlation of chemical shift difference between crystalline and noncrystalline regions versus the system temperature is illustrated in Figure 77. It is clear that the chemical shift difference shortens as temperature goes down and this trend of decreasing becomes more significant at lower temperatures.
142 Table 72. Qua ntitative data of chemical shifts and peak information for PE39CH3 T, K hmw in Figure 76. carbon, ppm FWHM Hz CH 2 (CR), ppm CH 2 (NCR), ppm ppm Ratio of CR/NCR (intensity) CH 3 branch, ppm FWHM CH3 Hz RT 1 39.05 382 32.73 31.01 1.72 12.82 21.37 595 380 37.96 69 -30.49 -0 --370 38.07 71 33.23 30.60 2.63 0.23 20.55 105 365 38.07 374 33.22 30.63 2.59 0.42 20.42 117 360 38.09 322 33.16 30.64 2.52 0.54 20.43 195 355 38.11 372 33.14 30.67 2.47 0.62 20.43 198 350 38.06 425 33.07 30.66 2.41 0.78 20.39 352 343 38.08 420 33.02 30.68 2.34 1.12 20.27 360 333 38.20 405 32.95 30.71 2.24 1.96 20.23 522 RT 2 39.17 458 32.71 30.89 1.06 8.85 20.10 632 FWHM: full width at half maximum CR: crystalline region NCR: non crystalline region = chemical shift of CR chemical shift of NCR Ratio of CR/NCR: the intensity ratio of CR over NCR RT1RT : room temperature (sample as received from synthesis) 2 290 300 310 320 330 340 350 360 370 380 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 chemical shift differencetemperature, K : room temperature (sample recrystallized from the melt ) Figure 77. Plot of chemical shift difference between crystalline and noncrystalline CH2 as a function of temperature. Data from Table 72
143 It is wellknown that the solutioncrystallized and melt crystallized polyethylenes behave differently.81, 113, 114 A more restricted local chain dynamics in noncrystalline regions of the solutioncrystallized UHMWPE and the largely isotropic motion of melt crystallized sample have been observed. Figure 78 shows the 13C spectra of PE39CH3hmw at various crystallization conditions. As expected, the noncrystalline CH2 recrystallized from the melt displays more isotropic motion than that from the synthesis (polymer was achieved via precipitation). Moreover, the broad peak s at around 21 ppm assigned from methyl branches are nearly identical, indicating the crystallization condition is among the para meters that play important roles in determining the orientation distribution of CH3 branches Figure 78. 13C spectra of PE39 CH3 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 ppm 18 19 20 21 22 23 24 ppm Melt crystallized As received from synthesis hmw from various crystallization conditions. Measurements were done at room temperature. The broad methyl branch peaks are amplified for clarity
144 7. 4 Conclusion The ADMET precision poly ethylenes with low branch frequency show unique thermal, morphological, and microstructural properties distinguished from those with high branch frequency. The rather low content of branching o ffers high melting temperature closer to the TmConsidering that the lamellar thickness of UHMWPE is around 12 nm, and that of PE21 CH of HDPE, implying the possibility of modeling HDPE by PE39 R polymers. The SS NMR measurements show interesting line broadening of the molecular weight and its distribution. Temperature variable experiments indicate that the broad methyl peak is born quickly when cooling from the melt and seems not affected by the crystallization conditions. 3 is around 10 nm, PE39CH3A morphological model can be proposed. Since the methyl branches in the amorphous regions undergo typically isotropic motions, the resulting chemical shift should be solution like, i.e., these branches do not contribute to the unusual line broadening for the case of PE39CH hmw should fold into lamellar structure with thickness at similar magnitude. The distance between two adjacent methyl branches for an all trans conformation is nearly 4.7 nm, which means for a specific chain stem, there are maximum two methyl branches can be incorporated in the crystalline regi ons prior to the occurrence of chain folding. 3. So there leaves us the methyl branches in the crystalline regions and interphase. The methyl branches in the crystalline regions actually may be in anisotropic local environment s. The remo te spacing between branches allows us to treat the branches more like the intrachain defects, just as the case in HDPE or UHMWPE. In contrast to PE21CH3, in which the branches
145 communicate with each other and behave more like normal parts composed of the unit cells, the branches in PE39CH3 are too far away to work together in disturbing the ordinary orthorhombic PE structures. Therefore, these intrachain defects suffer static irregularity and display a wide distribution of space orientation relative to the main chain. Complimentary X ray and TEM measurements will be very helpful in confirming the hypothesis.
146 CHAPTER 8 SUMMARY AND OUTLOOK Summary This dissertation aimed to gain an understand ing of the relationship of structureproperty performance for commercial polyethylene utilizing acyclic diene metathesis polymers as structural models. It has been demonstrated that the macroscopic properties of ADMET precision and random polymers are strongly related to the branch identity, branch spac ing, and branch distribution. Solid state NMR and X ray scattering used herein are proved to be powerful and well suited as two major analytical tools in investigating this subject. Structural Modeling Commercial PE resins are classified based on their densities which are determined by branching. HDPE consists primarily of nonbranched ethylene sequences; LDPE is characterized by its incorporation of both short chain and long chain branches; LLDPE contains a substantial number of short chain branches; VLDPE is kn own for its much high level of short chain branches. ADMET precision polymer s maintain controlled primary structures, i.e., pre determined branch identity regularly spaced on the ethylene backbone. ADMET random polymer s possess more variables: either pre det ermined branch identity with irregular branch distribution, or precise branch distribution with heterogeneous branch identities. From the ADMET precision polymers to ADMET random polymer and then to the commercial PE based polymers, the progression of stru cture modeling can be envisioned from a point to a surface to a cuboid, corresponding to increasing freedom of branching fro m zero to two and then to three. Therefore, the investigation of commercial polyethylenebased structures by
147 fixing some of the bran ching variables while varying one parameter is highly feasible By varying the content or type of branches, the ADMET precision and random polyolefins can be effectively used to model commercial LLDPE, VLDPE, HDPE, or even LDPE. Polymer Synthesis The deuterium labeling was chosen as a means of investigation because the 2Synthesis of per deuterated polymer samples was accomplished via acyclic diene metathesis polycondensation. The starting symmetric diene monomers were made via either malonatebased or cyanide chemistry. Unsaturated ADMET precision polymers were then hydrog enated to provide saturated versions The ADMET random polymers were prepared by copolymerization of two diene monomers followed by exhaustive hydrogenation. One particular challenge was the residue of transition metal catalysts, which by all attempts could not be completely removed from the final products. Purification of ADMET polymers to eliminate trace catalyst residue has been a hot subject in metathesis chemistry. However, several questions are not yet clear and need further investigation: Does the presence of transition metal catalyst will affect some physical properties of ADMET polymers ? W hat is the relationship of catalyst residue concentration and the affected properties ? H SSNMR lineshape is extremely sensitive to segmental motions. This characteristic observation offers a powerful method to study polymer microstructure as well as chain dynamics. Branch S pac ing E ffect When the branch spac ing varies, morphology, crystallinity, and the thermal behavior of ADMET precision polymers change correspondingly. For polymers with identical branch identi ty, there is generally a qualitative relation between branch spac ing
148 and melting temperature. The smaller the branch spac ing th e lower the melting temperature. However, when the spac ing is too small to allow segments of chains to pack into ordered structures, totally amorphous polymer will be formed. Sampled with three methyl branched precision polymers, PE21CD3, PE15 CD3, and PE9CD3, 2H SS NMR and 13C SS NMR measurements w ere analyzed jointly By studying these regularly branched model polyolefins prepared though metathesis chemistry by advanced solid state NMR, twist defects around the branches in the crystall ine regions are observed. For PE with low branch content (in the case of PE15 CD3The longest branch spac ing to date, 39, is different from other precision poly mers. The WAXS determination indicates more than one crystal structures is present. ), the twist motions are decoupled (pinned defects), for higher branch contents collective motion (rotator phase) is observed. F or branched polyethylenes, this is the first time that clear evidence of a rotator phase has been found, consistent with the results of X ray scattering and electron microscopy. 13Branch I dentity E ffect C SS NMR displays the rather broader peak s for methyl branches, methine carbons, and carbons; these breadths arise from the dispersed heterogeneous orientation of these carbon atoms with respect to the ethylene backbone. In the case of these ADMET precision polyethylenes with the lowest branch frequency the branches can be considered more reasonably as defects in the ethylene packing array. Such a consideration leads us to conclude polymers with low branch frequency are rather similar to the polyethyl ene structures in reality, especially high density polyethylene. The type and size of branches are crucial as well. This dissertation employ ed a series of ADMET precision polymer s with various alkyl branches sp aced along the
149 polyethylene main chain. WAXD measurements indicat es that the smaller branches such as methyl and ethyl groups can be incorporated into the unit cells, while the larger br anches, propyl groups or bigger are excluded from the unit cells. This conclusion from XRD is in agreement with the evidence from thermal analysis and SS NMR. The latter method was used to study the dynamics of different parts in the precision polymers of interest and it was observ ed that the methyl branches in PE21Methyl display two dynamics slow and fast, representing the methyl branches in ordered crystalline regains and disordered noncrystalline regions, respectively. In contrast, t he CH3Indeed, the ADMET polyolefins provide perfect structural models for commercial polyethylenerelated materials by fixing one or two of the branching variables. The branch spac ing branch identity, and branch distribution are all significant parameters in determining microstructure and consequently the macromolecular properties and eventually the performance of the ADMET polyolefins. part of t he butyl branches in PE21Butyl exhibit only one fast recoupling constant, indicating that the butyl branches are present only in disordered noncrystalline reg ions, as expected based on XRD measurements and thermal analysis. Outlook The fundamental understanding of the structureproperty performance relationship for ADMET polyolefins offers not only reasonable structural models for industrial PE based materials, but also a potential application of ADMET PEs as additives during processing. Much like the use of additives in the industrial production of thermoplastics, a ddition of a trace amount of ADMET precision monomer/polymer may tune the structure and properties of the commercial material C onsidering the substantial effects
150 of precision versus randomness on microstructures and physical properties, one would expect that ADMET polyolefins may function well as industrial additives The branch frequency plays an important role in the organization of crystal structures and consequently the melting temperature. The linear fit of melting temperature as a function of the branch content allows us to predict the thermal behavior of the A DMET polyolefins which have yet been synthesized. This prediction will both guide future synthesis and reduce unnecessary attempts Since all attempts to remove the r esidue of transition metal catalysts present in the final polymer products failed it is then worth investigat ing the effect of catalyst on the physical properties including nuclear relaxation. The applicable methodology of purification, the acceptable content of the catalyst, and the influence of the catalyst residue on physical properties a re all of special interest. These issues can be more interesting in SS NMR due to the fact that paramagnetic metal nuclei can affect the relaxation time significantly. At present, only the primary structure is controlled precisely via ADMET chemistry. How to manipulate the precision in the secondary or even the t ertiary structures in polyolefins will be of particular interest Since a polymer is in fact a mixture of polymer chains with heterogeneous sizes, microstructures and orientations, a structure with higher level of regularity will minimize the complexity and will function as a perfect structural model in several disciplines, including the chemical industry, physics, and material science.
151 APPENDIX A LAMELLAR THICKNESS DETERMINATION Lamellar thickness measurements for PE21 CD3 (measured at Max Planck Institute for Polymer Research by Dr. Ingo Lieberwirth): (a) by TEM using Pt shadowing; (b) by TEM using ectron Energy Loss Spectroscopy ; (c) by TEM using stereoscopic method; (d) by AFM 0.2 m 0.2 m tan a = 0.485 28 nm tan a = 13.6 nm 28 nm tan a = 13.6 nm a b c d
152 APPENDIX B DEUTERIUM 2D EXCHANGE NMR SPECTRA 2 dimensional 2H exchange NMR spectra for PE21CD3 at 296 K. The absence of off diagonal intensity indicates no slow reorientations on the exchange time scale.
153 LIST OF REFERENCES (1) Von Pechmann, H. Ber. Dtsch. Chem. Ges. 1898, 31 (2) Peacock, A. J. Handbook of polyethylene: structures, properties, and applications ; New York Marcel Dekker, Inc., 2000. (3) Brydson, J. A. Plastics Materials ; Elsevier ButterworthHeinemann, 1999. (4) Vasile, C.; Pascu, M. Practical guide to polyethylene ; Smithers Rapra, 2005. (5) Fawcett, E. W.; Gibson, R. O.; Perrin, M. W.; Patton, J. G.; Williams, E. G.: B Patent, 1937. (6) Fox, J. J.; Martin, A. E Trans. Faraday Soc., 1940, 36 897911. (7) Field, E.; Feller, M. Industrial & Engineering Chemistry 1957, 49, 1883 1884. (8) Schultz, J. M. Polymer crystallization : the development of crystalline order in thermoplastic polymers ; Oxford: American Chemic al Society, 2001. (9) Alamo, R. G.; VanderHart, D. L.; Nyden, M. R.; Mandelkern, L. Macromolecules 2000, 33 60946105. (10) Alamo, R. G.; Jeon, K.; Smith, R. L.; Boz, E.; Wagener, K. B.; Bockstaller, M. R. Macromolecules 2008, 41, 71417151. (11) Bunn, C. W. Transactions of the Faraday Society 1939, 35 482 491. (12) Vand, V. Acta Crystallographica 1951 4 104105. (13) Swan, P. R. Journal of Polymer Science 1962 56, 403 407. (14) Swan, P. R. Journal of Polymer Science 1962 56, 409 416. (15) Zhan, C. M. Journal of Macromolecular Science, Part B 1991 30 63 73. (16) AbouKandil, A. I.; Windle, A. H. Polymer 2007, 48 5069 5079. (17) Werner, P. E. Journal of Applied Crystallography 1976 9 216 219. (18) Tashiro, K.; Sasaki, S.; Kobayashi, M. Macromolecules 1996, 29 74607469. (19) Bassett, D. C.; Block, S.; Piermari.Gj Journal of Applied Physics 1974, 45, 41464150. (20) Tonelli, A. E. NMR Spectroscopy and Polymer Microstructure; Wiley VCH: Weinheim, 1989.
154 (21) Strobl, G. The Physics of Polymers 2nd ed.; Springer Verlag: Berlin, 1997. (22) Bruce, H. P. Advances in Polymer Technology 1983, 3 271 275. (23) Gedde, U. W.; Mattozzi, A. LongTerm Properties of Polyolefins 2004 169 29 73. (24) Lohse, D. J.; Milner, S. T.; Fetters, L. J.; Xenidou, M.; Hadjichristidis, N.; Mendelson, R. A.; GarciaFranco, C. A.; Lyon, M. K. Macromolecules 2002, 35, 30663075. (25) Baughman, T. W.; Sworen, J. C.; Wagener, K. B. Macromolecules 2006, 39 50285036. (26) Smith, J. A.; Brzezinska, K. R.; Valenti, D. J.; Wagener, K. B. Macromolecules 2000, 33 37813794. (27) Rojas, G.; Baughman, T. W.; Wagener, K. B. Synthetic Communications 2007 37 39233931. (28) Rojas, G.; Berda, E. B.; Wagener, K. B. Polymer 2008, 49 29852995. (29) Rojas, G.; Inci, B.; Wei, Y.; W agener, K. B. Journal of the American Chemical Society 2009, 131 17376 17386. (30) Rojas, G.; Wagener, K. B. Journal of Organic Chemistry 2008, 73 49624970. (31) Sworen, J. C.; Smith, J. A.; Berg, J. M.; Wagener, K. B. Journal of the American Chemical Society 2004 126 11238 11246. (32) Sworen, J. C.; Smith, J. A.; Wagener, K. B.; Baugh, L. S.; Rucker, S. P. Journal of the American Chemical Society 2003, 125 2228 2240. (33) Sworen, J. C.; Pawlow, J. H.; Case, W.; Lever, J.; Wagener, K. B. Journal of M olecular Catalysis aChemical 2003, 194 69 78. (34) Sworen, J. C.; Wagener, K. B. Macromolecules 2007, 40 44144423. (35) John, E. S.; Kenneth, B. W. Macromolecular Chemistry and Physics 2005, 206 14611471. (36) Zuluaga, F.; Inci, B.; Nozue, Y.; Hosoda, S.; Wagener, K. B. Macromolecules 2009, 42 49534955. (37) Rojas, G.; Wagener, K. B. Macromolecules 2009, 42 19341947. (38) Matloka, P. P.; Sworen, J. C.; Zuluaga, F.; Wagener, K. B. Macromolecular Chemistry and Physics 2005, 206 218 226.
155 (39) Matlok a, P. P.; Wagener, K. B. Journal of Molecular Catalysis aChemical 2006 257, 89 98. (40) Matloka, P. P.; Kean, Z.; Greenfield, M.; Wagener, K. B. Journal of Polymer Science Part aPolymer Chemistry 2008, 46 39924011. (41) Matloka, P. P.; Kean, Z.; Wagen er, K. B. Journal of Polymer Science Part aPolymer Chemistry 2010, 48, 18661877. (42) Wagener, K. B.; Zuluaga, F.; Wanigatunga, S. Trends in Polymer Science 1996, 4 157163. (43) Boz, E.; Ghiviriga, I.; Nemeth, A. J.; Jeon, K.; Alamo, R. G.; Wagener, K. B. Macromolecules 2008, 41, 25 30. (44) Boz, E.; Nemeth, A. J.; Alamo, R. G.; Wagener, K. B. Advanced Synthesis & Catalysis 2007, 349 137 141. (45) Boz, E.; Nemeth, A. J.; Ghiviriga, I.; Jeon, K.; Alamo, R. G.; Wagener, K. B. Macromolecules 2007, 40, 65456551. (46) Boz, E.; Nemeth, A. J.; Wagener, K. B.; Jeon, K.; Smith, R.; Nazirov, F.; Bockstaller, M. R.; Alamo, R. G. Macromolecules 2008, 41 16471653. (47) Boz, E.; Wagener, K. B. Polymer Reviews 2007, 47 511 541. (48) Boz, E.; Wagener, K. B.; Gho sal, A.; Fu, R. Q.; Alamo, R. G. Macromolecules 2006, 39 44374447. (49) Leonard, J. K.; Hopkins, T. E.; Chaffin, K.; Wagener, K. B. Macromolecular Chemistry and Physics 2008, 209 14851494. (50) Hopkins, T. E.; Pawlow, J. H.; Koren, D. L.; Deters, K. S.; Solivan, S. M.; Davis, J. A.; Gomez, F. J.; Wagener, K. B. Macromolecules 2001, 34 79207922. (51) Hopkins, T. E.; Wagener, K. B. Macromolecules 2003, 36 22062214. (52) Hopkins, T. E.; Wagener, K. B. Journal of Molecular Catalysis aChemical 2004 213, 93 99. (53) Hopkins, T. E.; Wagener, K. B. Macromolecules 2004, 37 11801189. (54) Leonard, J. K.; Turek, D.; Sloan, K. B.; Wagener, K. B. Macromolecular Chemistry and Physics 2010 211, 154 165. (55) Zuluaga, F.; Valderruten, N. E.; Wagener, K. B. Poly mer Bulletin 1999, 42, 41 46. (56) Opper, K. L.; Fassbender, B.; Brunklaus, G.; Spiess, H. W.; Wagener, K. B. Macromolecules 2009, 42, 44074409.
156 (57) Opper, K. L.; Markova, D.; Klapper, M.; Mullen, K.; Wagener, K. B. Macromolecules 2010, 43, 36903698. (5 8) Opper, K. L.; Wagener, K. B. Macromolecular Rapid Communications 2009, 30, 915919. (59) Schwendeman, J. E.; Wagener, K. B. Macromolecules 2004, 37 40314037. (60) Schwendeman, J. E.; Wagener, K. B. Macromolecular Chemistry and Physics 2009, 210 18181833. (61) Aitken, B. S.; Lee, M.; Hunley, M. T.; Gibson, H. W.; Wagener, K. B. Macromolecules 2010, 43, 16991701. (62) Petkovska, V. I.; Hopkins, T. E.; Powell, D. H.; Wagener, K. B. Macromolecules 2005, 38 58785885. (63) Petkovska, V. I.; Hopkins T. E.; Powell, D. H.; Wagener, K. B. Analytical Chemistry 2006, 78 36243631. (64) Baker, C. H.; Mandelkern, L. Polymer 1966, 7 71 83. (65) Davis, G. T.; Eby, R. K.; Martin, G. M. Journal of Applied Physics 1968, 39, 49734981. (66) Swan, P. R. Journal of Polymer Science 1960 42, 525 534. (67) Koenig, J. L. Spectroscopy of Polymers ; Amsterdam: Elsevier, 1999. (68) Stuart, B. Polymer analysis. Analytical techniques in the sciences ; Wiley VCH, 2002. (69) Groenewoud, W. M. Characterisation of polymers by thermal analysis ; Amsterdam: Elsevier, 2001. (70) Cheremisinoff, N. P. Polymer Characterization Laboratory Techniques and Analysis ; William Andrew Publishing/Noyes, 1996. (71) Elias, H. G. An introduction to polymer science; Weinheim: VCH, 1997. (72) Pai nter, P. C.; Coleman, M. M. Fundamentals of Polymer Science: An Introductory Text 2nd. ed.; CRC Press, 2000. (73) Elias, H. G. Macromolecules: Volume 2: Industrial Polymers and Syntheses ; WILEY VCH, 2007. (74) Elias, H. G. Macromolecules: Volume 1: Chemic al Structures and Syntheses ; WILEY VCH, 2007.
157 (75) Richardson, T. L. Industrial plastics: Theory and application, 5th. ed.; Albany, N.Y.: Delmar, 1989. (76) David, W. I. F. Structure determination from powder diffraction data; Oxford: Oxford University Press, 2002. (77) Duer, M. J. Solid state NMR spectroscopy : principles and applications Malden, MA: Blackwell Science, 2002. (78) Kasai, N.; Kakudo, M. X ray diffraction by macromolecules ; Tokyo: Kodansha, 2005. (79) Schmidt Rohr, K.; Spiess, H. W. Multidime nsional SolidState NMR and Polymers ; Academic Press: London, 1994. (80) Ernst, R. R. Angewandte ChemieInternational Edition in English 1992 31 805 823. (81) Yao, Y.; Chen, Q.; Graham, A. W. In Annual Reports on NMR Spectroscopy ; Academic Press, 2010; V ol. Volume 69, pp 199224. (82) Martin, G. E.; Zekter, A. S. Two Dimensional NMR Methods for Establishing Molecular Connectivity ; VCH Publishers: New York, 1988. (83) Purcell, E. M.; Torrey, H. C.; Pound, R. V. Physical Review 1946, 69 37. (84) Bloch, F. Physical Review 1946, 70, 460. (85) Ernst, R. R. Review of Scientific Instruments 1965, 36 1689&. (86) Ernst, R. R.; Anderson, W. A. Review of Scientific Instruments 1965 36, 1696&. (87) Overhauser, A. W. Physical Review 1953, 92, 411. (88) Levitt, M. H. Spin Dynamics: Basics of Nuclear Magnetic Resonance ; John Wiley & Sons, LTD, 2001. (89) Keeler, J. Understanding of NMR Spectroscopy ; Wiley: Chichester, 2006. (90) Duer, M. J. Introduction to solidstate NMR spectroscopy Oxford, UK ; Malden, MA : Blackwell, 2004. (91) Spiess, H. W. Colloid and Polymer Science 1983, 261 193209. (92) Saalwachter, K.; Schnell, I. Solid State Nuclear Magnetic Resonance 2002, 22 154187. (93) Rose, M. E. Elementary theory of angular momentum ; Wiley: New York, 19 57.
158 (94) Spiess, H. W. Journal of Polymer Science Part aPolymer Chemistry 2004, 42, 50315044. (95) Khuong, T. A. V.; Dang, H.; Jarowski, P. D.; Maverick, E. F.; Garcia Garibay, M. A. Journal of the American Chemical Society 2007, 129 839 845. (96) Liu, S. F.; Mao, J. D.; SchmidtRohr, K. Journal of Magnetic Resonance 2002, 155, 15 28. (97) Tycko, R.; Dabbagh, G.; Mirau, P. A. Journal of Magnetic Resonance 1989, 85 265274. (98) Eden, M. Concepts in Magnetic Resonance Part A 2003, 18A 24 55. (99) Lieser G.; Wegner, G.; Smith, J. A.; Wagener, K. B. Colloid and Polymer Science 2004, 282 773 781. (100) Qiu, W. L.; Sworen, J.; Pyda, M.; Nowak Pyda, E.; Habenschuss, A.; Wagener, K. B.; Wunderlich, B. Macromolecules 2006, 39 204217. (101) Sumpter, B. G.; N oid, D. W.; Liang, G. L.; Wunderlich, B. Atomistic Modeling of Physical Properties 1994, 116 27 72. (102) Piel, C.; Starck, P.; Seppala, J. V.; Kaminsky, W. Journal of Polymer Science Part a Polymer Chemistry 2006, 44 16001612. (103) Islam, A.; Hussein, I. A. Journal of Applied Polymer Science 2006, 100 50195033. (104) Alamo, R. G.; Blanco, J. A.; Agarwal, P. K.; Randall, J. C. Macromolecules 2003, 36, 15591571. (105) Hussein, I. A.; Hameed, T.; Williams, M. C. Journal of Applied Polymer Science 2006, 102 17171728. (106) Born, R.; Spiess, H. W. Macromolecules 1995 28 7785 7795. (107) Greenfield, M. S.; Ronemus, A. D.; Vold, R. L.; Vold, R. R.; Ellis, P. D.; Raidy, T. E. Journal of Magnetic Resonance 1987, 72, 89 107. (108) Hentschel, D.; Sillescu, H.; Spiess, H. W. Macromolecules 1981, 14, 16051607. (109) Macho, V.; Brombacher, L.; Spiess, H. W. Applied Magnetic Resonance 2001 20 405432. (110) Vanderhart, D. L. Journal of Magnetic Resonance 1981, 44, 117 125. (111) Mansfield, M.; Boyd, R. H. Jo urnal of Polymer Science Part B Polymer Physics 1978, 16 12271252.
159 (112) Schmidtrohr, K.; Spiess, H. W. Macromolecules 1991, 24 52885293. (113) Yao, Y. F.; Graf, R.; Spiess, H. W.; Lippits, D. R.; Rastogi, S. Physical Review E 2007, 76 060801(R). (114) Yao, Y. F.; Graf, R.; Spiess, H. W.; Rastogi, S. Macromolecules 2008 41, 25142519. (115) Yao, Y. F.; Graf, R.; Spiess, H.; Rastogi, S. Macromolecular Rapid Communications 2009, 30, 11231127. (116) Sirota, E. B.; Herhold, A. B. Polymer 2000 41 87818 789. (117) Kaji, H.; Horii, F. Macromolecules 2007, 40, 54205423. (118) Helfand, E.; Wasserman, Z. R.; Weber, T. A.; Skolnick, J.; Runnels, J. H. Journal of Chemical Physics 1981, 75 44414445. (119) He, Y. Y.; Lutz, T. R.; Ediger, M. D.; Ayyagari, C.; Bedrov, D.; Smith, G. D. Macromolecules 2004, 37, 50325039. (120) Cory, D. G.; Ritchey, W. M. Macromolecules 1989, 22, 16111615. (121) Okuom, M. O.; Metin, B.; Blum, F. D. Langmuir 2008, 24 25392544. (122) Metin, B.; Blum, F. D. Journal of Chemical Physics 2006, 124 (123) O'Connor, R. D.; Ginsburg, E. J.; Blum, F. D. Journal of Chemical Physics 2000, 112, 72477259. (124) Blum, F. D.; Xu, G.; Liang, M. H.; Wade, C. G. Macromolecules 1996 29, 87408745. (125) Borsa, F.; Torgeson, D. R.; Martin, S. W.; Patel, H. K. Physical Review B 1992 46 795. (126) Hentschel, D.; Sillescu, H.; Spiess, H. W. Makromolekulare ChemieMacromolecular Chemistry and Physics 1979 180 241 249. (127) Hentschel, R.; Spie ss, H. W. Journal of Magnetic Resonance 1979 35, 157162. (128) Olf, H. G.; Peterlin, A. Journal of Polymer Science Part a2 Polymer Physics 1970 8 791 &. (129) Olf, H. G.; Peterlin, A. Journal of Polymer Science Part a2 Polymer Physics 1970 8 771 &. (130) Olf, H. G.; Peterlin, A. Journal of Polymer Science Part a2 Polymer Physics 1970 8 753 &.
160 (131) Hu, W. G.; Boeffel, C.; SchmidtRohr, K. Macromolecules 1999, 32, 16111619. (132) Vanderhart, D. L. Journal of Magnetic Resonance 1976, 24, 467 470. (133) Eastman, M. A.; Nanny, M. A. Journal of Magnetic Resonance 2007 184 302 314. (134) Schwendeman, J. E.; Wagener, K. B. Macromolecular Chemistry and Physics 2005, 206 14611471. (135) Wunderlich, B. In Macromolecular Physics ; Academic Press inc.: New York, 1973; Vol. 1, pp 401407. (136) Hu, W.; Srinivas, S.; Sirota, E. B. Macromolecules 2002, 35 50135024. (137) Hu, W.; Sirota, E. B. Macromolecules 2003, 36 51445149. (138) Androsch, R.; Blackwell, J.; Chvalun, S. N.; Wunderlich, B. Macromolecul es 1999, 32, 3735. (139) Suhm, J.; Schneider, M. J.; Mulhaupt, R. Journal of Polymer Science Part aPolymer Chemistry 1997, 35, 735 740. (140) Schneider, M. J.; Suhm, J.; Mulhaupt, R.; Prosenc, M. H.; Brintzinger, H. H. Macromolecules 1997, 30, 31643168. (141) Suhm, J.; Schneider, M. J.; Mulhaupt, R. Journal of Molecular Catalysis aChemical 1998, 128 215 227. (142) McKnight, A. L.; Waymouth, R. M. Chemical Reviews 1998, 98, 2587 2598. (143) Schwerdtfeger, E. D.; Irwin, L. J.; Miller, S. A. Macromolecules 2008, 41, 10801085. (144) Salata, M. R.; Marks, T. J. Macromolecules 2009, 42 19201933. (145) Hadjichristidis, N.; Xenidou, M.; Iatrou, H.; Pitsikalis, M.; Poulos, Y.; Avgeropoulos, A.; Sioula, S.; Paraskeva, S.; Velis, G.; Lohse, D. J.; Schulz, D. N.; Fetters, L. J.; Wright, P. J.; Mendelson, R. A.; Garcia Franco, C. A.; Sun, T.; Ruff, C. J. Macromolecules 2000, 33, 24242436. (146) Baughman, T. W.; Sworen, J. C.; Wagener, K. B. Tetrahedron 2004 60 10943 10948. (147) Organ, S. J.; Keller, A. Journal of Materials Science 1985, 20 15711585.
161 BIOGRAPHICAL SKETCH Yuying Wei was born in the city of Xian, Shaanxi province, China, in 1973. She is daughter of Zhanxian Wei and Huamei Li, sister of Yuning Wei, wife of Liwen Jin, and mother of Abbie Jin. Yuying received the Bachelor of Science i n Chemical Engineering from Northwest University Xian, China. After that she worked as a lec turer in Xian aerotechnique College for two years. In 1996, she joined the graduate school in University of Petroleum Beijing, China, where she received the Master of Science i n Chemical Engineering and met Liwen Jin. They got married in 1999. In 199920 01, Yuying worked as a chemical engineer at Beijing Research Institute of Chemical Industry China Petroleum & Chemical Corporation, designing and producing industrial catalyst. Later on, she and her husband moved to the United States for better understanding of the world and life. In 2002, their first child was born at Greenville Memorial Hospital, located at Greenville, South Carolina. Yuying studied polymer physics at the Chemistry Department, University of Clemson for two years. In 2006, Yuying moved to Gainesville, Florida and entered the Ph.D. program in the Chemistry Department at the University of Florida, majoring in analytical division. She joined Prof. Ken Wageners research group in the same year, working on synthesis and characterization of deut erated precision polyolefins. In 2008 & 2010, Yuying visited Max Planck Institute for Polymer Research, Mainz, Germany, and worked with Prof. Dr. Hans Spiess for 6 months in total. She learned modern solid state NMR techniques, and opened her insights of doing research. I n August 2010, Yuying will have completed her PhD study at the University of Florida.