Introduction to Polymer Analysis

Introduction To Polymer Analysis T.R. Crompton Introduction to Polymer Analysis T.R. Crompton iSmithers – A Smither...

Author: T. R. Crompton

546 downloads 2472 Views 4MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD

Introduction To Polymer Analysis

T.R. Crompton

Introduction to Polymer Analysis

T.R. Crompton

iSmithers – A Smithers Group Company Shawbury, Shrewsbury, Shropshire, SY4 4NR, United Kingdom Telephone: +44 (0)1939 250383 Fax: +44 (0)1939 251118 http://www.rapra.net

First Published in 2009 by

iSmithers Shawbury, Shrewsbury, Shropshire, SY4 4NR, UK

©2009, Smithers Rapra

All rights reserved. Except as permitted under current legislation no part of this publication may be photocopied, reproduced or distributed in any form or by any means or stored in a database or retrieval system, without the prior permission from the copyright holder. A catalogue record for this book is available from the British Library.

Every effort has been made to contact copyright holders of any material reproduced within the text and the authors and publishers apologise if any have been overlooked. Every effort has been made to contact copyright holders of any material reproduced within the text and the authors and publishers apologise if any have been overlooked.

ISBN: 978-1-84735-384-9 (hardback) 978-1-84735-385-6 (softback)

Typeset by Argil Services Printed and bound by Lightning Source Inc.

P

reface

The aim of this book is to familiarise the reader with all aspects of plastic analysis. It covers the analysis of the main types of commercial plastics currently in use. Practically all of the major newer analytical techniques (and many of the older classical techniques) have been used to examine plastics and their additive systems. Because so many polymers are now used commercially, it is also advisable when attempting to identify a polymer to initially classify it by carrying out at least a qualitative elemental analysis and possibly a quantitative analysis (Chapter 2) and, in some cases, depending on the elements found, to carry out functional group analysis (Chapter 3). Copolymers contain two or more different monomer units built up into a high molecular weight material. It is often important to determine the weight ratios of the monomer units in copolymers (Chapter 4). Identification of a polymer (particularly copolymers or terpolymers) is often not as simple as this, and obtaining a detailed picture of the microstructure of the polymer is necessary. Techniques that may be used, in addition to elemental and functional group analysis, include spectroscopic techniques such as infrared, nuclear magnetic resonance, proton magnetic resonance, and systematic investigations by pyrolysis–gas chromatography. The sequential order of different monomer units in a polymer or copolymer can be controlled during manufacture. This can have a very important bearing on the mechanical, electrical and other properties of the polymer (Chapter 5). Another property of polymers that can have an important bearing on the structure and properties of a polymer is stereoisomerism, geometrical or regioisometric configurations in the polymer structure (Chapters 6 and 7). Other important features of polymer microstructure are end-groups and different forms of unsaturation, determination of which are discussed, respectively, in Chapters 8 and 9. A very important aspect of polymer microstructure that is currently being studied is branching of side chain groups attached to the polymer backbone. Important

i

Introduction to Polymer Analysis conclusions reached in this work are discussed in Chapter 10. In many cases, considerable experience and innovative skills are required by the analyst to successfully identify polymers by these techniques, and it is hoped that this book will assist the analyst in developing such skills. The book gives a thorough exposition of the current state-of-the-art of polymer analysis and, as such, should be of great interest to those engaged in this subject in industry, university research, and general education. It is also intended for undergraduate and graduate chemistry students, and those taking courses in plastics technology, engineering chemistry, materials science and industrial chemistry. It will be a useful reference work for manufacturers and users of plastics, the food and beverage packing industry, engineering plastics industry, plastic components manufacturers, pharmaceutical industry, and the cosmetics industry. Before proceeding to the first two chapters which deal, with the determination of elements and functional groups, respectively, Chapter 1 discusses briefly the various types of polymers used commercially, and their properties and applications. Roy Crompton June 2009

ii

C

ontents

1 Types and Properties of Polymers

1.1 1.2 1.3

Production of Synthetic Resins................................................ 2 Polycondensation Reactions.................................................... 2 Polymerisation Reactions ........................................................ 4

2 Determination of Elements

2.1

1

7

Non-metallic Elements ............................................................ 7 2.1.1

Halogens ............................................................................ 8 2.1.1.1 Combustion Methods ........................................ 8 2.1.1.2 Oxygen Flask Combustion ................................ 8 2.1.1.3 Alkali Fusion Methods ...................................... 9 2.1.1.4 Physical Methods for Determining Halogen .... 10 2.1.2 Sulfur ............................................................................... 10 2.1.2.1 Combustion Methods ...................................... 10 2.1.2.2 Sodium Peroxide Fusion .................................. 10 2.1.2.3 Oxygen Flask Combustion .............................. 11 2.1.3 Phosphorus....................................................................... 11 2.1.3.1 Acid Digestion ................................................. 11 2.1.4 Nitrogen ........................................................................... 11 2.1.4.1 Combustion Methods ...................................... 11 2.1.4.2 Physical Method for Determination of Total Nitrogen ................................................. 13 2.1.5 Silicon .............................................................................. 13 2.1.6 Boron ............................................................................... 13 2.1.7 Total Organic Carbon ...................................................... 13 2.1.8 Total Sulfur/Total Halogen ............................................... 14 2.1.9 Nitrogen, Carbon, and Sulfur ........................................... 14 2.1.10 Carbon, Hydrogen, and Nitrogen..................................... 15

iii


2.1.11 Oxygen Flask Combustion: Ion Chromatography ............ 15 2.1.12 XRFS................................................................................ 15 2.1.13 Thermogravimetric Analysis ............................................. 18

2.2

Metals .................................................................................. 18 2.2.1

2.2.2

Destructive Techniques ..................................................... 18 2.2.1.1 Atomic Absorption Spectrometry (AAS) .......... 18 2.2.1.2 GFAAS ............................................................ 19 2.2.1.3 Atom Trapping Technique ............................... 21 2.2.1.4 Vapour Generation Atomic Absorption Spectrometry (VGAAS) ................................... 21 2.2.1.5 Zeeman AAS ................................................... 22 2.2.1.6 ICP-AES .......................................................... 24 2.2.1.6.1 Hybrid Inductively Coupled Plasma Systems ............................ 28 2.2.1.6.2 Chromatography–ICP .................. 28 2.2.1.6.3 Flow Injection with ICP ............... 28 2.2.1.6.4 Inductively Coupled Plasma with Atomic Fluorescence Spectrometry (ICP-AFS) ............... 28 2.2.1.7 Inductively Coupled Plasma Optical Emission Spectrometry–Mass Spectrometry (ICP-MS) .... 29 2.2.1.8 Pre-concentration AAS Techniques .................. 30 2.2.1.9 Applications of Techniques Detailed So Far ..... 30 2.2.1.9.1 Elemental Analysis of Polymers .... 31 2.2.1.9.2 Trace Metals in Polymers ............. 31 2.2.1.10 Pressure Dissolution Technique ....................... 34 2.2.1.11 Visible and UV Spectroscopy ........................... 36 2.2.1.12 Polarography and Voltammetry ....................... 36 2.2.1.13 Ion Chromatography ....................................... 37 Non-destructive Methods ................................................. 40 2.2.2.1 XRFS............................................................... 40 2.2.2.2 NAA................................................................ 43

3 Functional Groups

3.1

Hydroxy Groups .................................................................. 51 3.1.1

iv

51

Chemical Methods ........................................................... 51

Contents

3.1.2 3.1.3 3.1.4

3.1.5

3.2

Carboxyl Groups .................................................................. 64 3.2.1 3.2.2 3.2.3

3.3

3.3.4 3.3.5

IR Spectroscopy................................................................ 74 Derivatisation Methods .................................................... 75 3.4.2.1 Spectrophotometric methods ........................... 76

Ether Groups ........................................................................ 76 3.5.1

3.6

Saponification................................................................... 69 Hydriodic Acid Reduction–Gas Chromatography ............ 70 IR Spectroscopy................................................................ 71 3.3.3.1 Determination of Free and Combined Vinyl Acetate Groups in Vinyl Chloride-Vinyl Acetate Copolymers ..................................................... 71 3.3.3.2 Determination of Bound Vinyl Acetate in Ethylene-Vinyl Acetate Copolymers................. 71 NMR................................................................................ 72 Pyrolysis–Gas Chromatography ....................................... 73

Carbonyl Groups .................................................................. 74 3.4.1 3.4.2

3.5

NMR................................................................................ 64 Titration Procedures ......................................................... 64 IR Spectroscopy................................................................ 67

Ester Groups ......................................................................... 68 3.3.1 3.3.2 3.3.3

3.4

3.1.1.1 Acetylation and Phthalation Procedures .......... 52 Spectrophotometric methods ............................................ 59 Direct Injection Enthalpimetry ......................................... 60 IR Spectroscopy................................................................ 61 3.1.4.1 Determination of Hydroxy Groups in Dinitropropyl Acrylate Prepolymer ................. 61 NMR................................................................................ 62

Cleavage Gas Chromatography ........................................ 76

Alkoxy Groups ..................................................................... 77 3.6.1 3.6.2 3.6.3

IR Spectroscopy................................................................ 77 NMR Spectroscopy .......................................................... 79 Pyrolysis-Based Method ................................................... 80 3.6.3.1 Alkoxy Groups in Ethylene Oxide-Propylene ..... Oxide Condensates.......................................... 80 3.6.3.2 Miscellaneous Methods ................................... 81

v


3.7

Oxyalkylene Groups ............................................................. 81 3.7.1 3.7.2 3.7.3 3.7.4

3.8 3.9

Cleavage–Gas Chromatography ....................................... 81 Pyrolysis–Gas Chromatography ....................................... 83 IR Spectroscopy................................................................ 84 NMR Spectroscopy .......................................................... 84

Anhydride Groups ................................................................ 84 Total Unsaturation ................................................................ 85 3.9.1 3.9.2 3.9.3 3.9.4 3.9.5 3.9.6 3.9.7 3.9.8

Hydrogenation Methods .................................................. 85 Halogenation Methods ..................................................... 85 Iodine Monochloride Procedure ....................................... 88 IR Spectroscopy................................................................ 92 NMR Spectroscopy .......................................................... 93 Pyrolysis–Gas Chromatography ....................................... 97 Derivitivisation–Gas Chromatography ............................. 97 Radiochemical Methods ................................................. 100 3.9.8.1 Determination of unsaturation in butyl rubber .................................................. 100

3.10 Alkyl and Aryl Groups........................................................ 100 3.10.1 Alkali Fusion Reaction–Gas Chromatography................ 100

3.11 Oxirane Rings .................................................................... 101 3.12 Amino Groups .................................................................... 101 3.13 Amino and Imido Groups ................................................... 102 3.13.1 Alkali Fusion–Gas Chromatography .............................. 102

3.14 Nitrile Groups .................................................................... 105 3.15 Silicon Functions ................................................................ 106 4 Determination of Monomer Ratios in Copolymers

4.1

IR Spectroscopy .................................................................. 117 4.1.1 4.1.2 4.1.3

4.1.4 vi

117

Ethylene Propylene Copolymers ..................................... 117 Ethylene–Vinyl Acetate ................................................... 122 Styrene-based Copolymers.............................................. 123 4.1.3.1 Styrene Acrylic Acid ...................................... 123 4.1.3.2 Styrene–acrylate and styrene methylacrylate copolymers .................................................... 123 Vinyl Chloride–Vinyl Acetate Copolymers...................... 123

Contents

4.2

NMR Spectroscopy............................................................. 124 4.2.1 4.2.2 4.2.3

4.2.4 4.2.5 4.2.6

4.2.7

4.3

Pyrolysis–Gas Chromatography .......................................... 135 4.3.1

4.3.3 4.3.4

4.4

Ethylene Propylene Copolymers ..................................... 124 Ethylene–hexane-1 ......................................................... 129 Styrene-Based Copolymers ............................................. 129 4.2.3.1 Styrene Methacrylate ..................................... 129 4.2.3.2 Styrene methyl acrylate.................................. 130 4.2.3.3 Styrene–acrylic acid ....................................... 131 Benenyl Acrylate–Vinyl Acetate [27]............................... 132 Vinyl Acetate–Methylacrylate ......................................... 132 Hexafluoropropylene–Vinylidene Fluoride Copolymer ... 132 19 F-NMR ....................................................... 132 4.2.6.1 4.2.6.2 Pyrolysis–Gas Chromatography .................... 133 Acrylamide–Methacryloyl Oxy-ammonium Chloride ..... 133 Ethylene-Based Copolymers ........................................... 135 4.3.1.1 Ethylene–butene-1 ......................................... 135 4.3.1.2 Ethylene - butadiene ...................................... 138 Vinylidene Chloride – Vinyl Chloride Copolymers ......... 138 Acrylonitrile-cis (or trans) Penta 1,3 diene ...................... 141

Pyrolysis IR Spectroscopy ................................................... 141 4.4.1

Olefin Copolymers ......................................................... 141

5 Sequencing of Monomer Unit in Polymers

5.1

Sequencing in Homopolymers............................................. 147 5.1.1 5.1.2

5.1.3 5.1.4

5.2

147

NMR Spectroscopy ........................................................ 147 Pyrolysis Gas Chromatography (Py-GC) ........................ 151 5.1.2.1 Polyolefins ..................................................... 152 5.1.2.2 Polyisoprene .................................................. 153 5.1.2.3 Polyvinyl Chloride (PVC) .............................. 154 SIMS .............................................................................. 159 5.1.3.1 Polystyrene .................................................... 159 Ozonisation Technique ................................................... 161 5.1.4.1 Polybutadiene ................................................ 162 5.1.4.2 Polyisoprene .................................................. 168

Sequencing in Copolymers .................................................. 171 vii


5.2.1 5.2.2

5.2.3

5.2.4 5.2.5

IR Spectroscopy.............................................................. 171 5.2.2.1 Styrene–methacrylonitrile .............................. 171 NMR Spectroscopy ........................................................ 174 5.2.2.1 Styrene acrylate and styrene acrylic acid ........ 174 5.2.2.2 Propylene-1-butene........................................ 175 5.2.2.3 Vinylidene chloride–methacrylonitrile and vinylidene–cyanovinyl acetate copolymers ..... 177 5.2.2.4 Acrylonitrile–butyl acrylate copolymer .......... 177 Py-GC ............................................................................ 180 5.2.3.1 Ethylene–propylene diene .............................. 181 5.2.3.2 Hydrogenated acrylonitrile–butadiene copolymers (NBR) ......................................... 185 5.2.3.3 Butadiene–acrylonitrile–methacrylic acid–terpolymer............................................. 187 5.2.3.4 Styrene-n-butyl acrylate ................................. 190 5.2.3.5 Ethylene oxide condensates ........................... 196 SIMS .............................................................................. 198 5.2.4.1 Polydimethyl siloxane–urethane .................... 198 Ozonolysis Techniques ................................................... 203 5.2.5.1 Butadiene–propylene ..................................... 203 5.2.5.2 Styrene butadiene copolymers ....................... 203

6 Stereoisomerism and Tacticity

6.1 6.2 6.3 6.4 6.5 6.6

Tacticity of Polypropylene .................................................. 212 Tacticity of Syndiotactic Polystyrene (sPS) .......................... 226 Tacticity of Polyvinyl Chloride (PVC) ................................. 230 Tacticity of Poly(n-butyl methacrylate) ............................... 233 Identification of Diastereoisomeric Tetramers in the Pyrograms of polymethyl methacrylate................................................. 237 Tacticity of Poly(1-chloro-fluoroethylene) ........................... 242

7 Regioisomerism

7.1 7.2 7.3

viii

211

253

Polypropylene ..................................................................... 253 Propylene-1-Ethylene Copolymer........................................ 257 Polybutadiene-1-ethylene .................................................... 259

Contents

7.4 7.5 7.6 7.7 7.7 7.8

Poly-2,3-dimethyl Butadiene ............................................... 260 Polybutadiene ..................................................................... 260 Polyisoprene ....................................................................... 262 Polypropylene Glycol .......................................................... 266 Polyepichlorohydrin ........................................................... 269 Other Polymers ................................................................... 272

8 Determination of End Groups

8.1 8.2 8.3

Polypropylene Oxide .......................................................... 276 Polyvinyl chloride (PVC)..................................................... 277 Polystyrene (PS) .................................................................. 278 8.3.1 8.3.2 8.3.3

8.4 8.5 8.6

8.6.3

Py–GC ............................................................................ 287 MALDI-ToF-MS............................................................. 299 Dye Partition Methods ................................................... 299

Terminal Epoxides .............................................................. 300 8.8.1

8.9

Tert-chlorine Terminated PIB .......................................... 285 Olefin-terminated PIB ..................................................... 286 8.6.2.1 Anisotropic Effect.......................................... 286 Hydroxy-terminated PIB ................................................ 286

Polymethylmethacrylate ...................................................... 287 8.7.1 8.7.2 8.7.3

8.8

NMR Spectroscopy ........................................................ 278 Pyrolysis – Gas Chromatogarphy (Py-GC)...................... 280 Dye Partition Methods ................................................... 281

Polyethylene (PE) ................................................................ 282 Polyethylene Terephthalate ................................................. 283 Polyisobutylene (PIB) .......................................................... 284 8.6.1 8.6.2

8.7

275

IR spectroscopy .............................................................. 300

Poly(2,6-dimethyl 1,4, phenylene oxide) ............................. 301 8.9.1

NMR spectroscopy......................................................... 301

8.10 Miscellaneous End Groups ................................................. 304 9 Types of Unsaturation

9.1

313

Unsaturation in Homopolymers.......................................... 313 9.1.1

Polybutadiene Unsaturation ........................................... 313

ix


9.1.1.1 9.1.1.2

9.1.2

9.1.3 9.1.4

9.2

Unsaturation in Copolymers ............................................... 332 9.2.1 9.2.2 9.2.3

9.3

Infrared spectrometry .................................... 313 Nuclear Magnetic Resonance (NMR) Spectroscopy ................................................. 317 Polyisoprene Unsaturation.............................................. 321 9.1.2.1 IR Spectroscopy............................................. 321 9.1.2.2 Nuclear Magnetic Resonance (NMR) spectroscopy .................................................. 325 Polyethylene Unsaturation .............................................. 325 Polypropylene Unsaturation ........................................... 329 Styrene–divinyl benzene IR Spectroscopy ....................... 332 Poly(trimethylolpropane trimethacrylate) (TRIM) .......... 335 Miscellaneous Copolymers ............................................. 336

Ozonolysis Techniques........................................................ 336

10 Polymer Branching

353

10.1 IR Spectroscopy .................................................................. 354 10.1.1 Methyl Branching in Polyethylene .................................. 354

10.2 NMR Spectroscopy............................................................. 355 10.2.1 Ethyl and Higher Alkyl Groups Branching in Polyethylene ................................................................... 357 10.2.2 Branching in Ethylene–propylene Copolymers ................ 358 10.2.3 Branching Ethylene–Higher Olefin Copolymers .............. 361 10.2.4 Polystyrene ..................................................................... 369 10.2.5 Polyvinyl Chloride .......................................................... 369 10.2.6 Polyvinyl Fluoride .......................................................... 369

10.3 Vacuum Radiolysis ............................................................. 371 10.3.1 Ethylene Copolymer ....................................................... 371

10.4 Pyrolysis-based Techniques ................................................. 375 10.4.1 Elucidation of Short Chain Branching in Polyethylene.... 375 10.4.2 Short Chain Branching in Ethylene–Higher Olefin Copolymers .................................................................... 383 10.4.2 Branching in Ethylene–propylene Copolymers ................ 386 10.4.2.1 Microstructure of Ethylene-Propylene Copolymers ................................................... 386

x

Contents

10.4.3 Branching in PVC ........................................................... 386

10.5 Size-Exclusion Chromatography (SEC) ............................... 387 10.5.1 Polyethylene ................................................................... 387 10.5.2 Other Polymers .............................................................. 388 11 Block Copolymers

395

Abbreviations

401

Index

407

xi


xii

1

Types and Properties of Polymers

Synthetic resins, in which plastics are also included, vary widely in their chemical composition and physical properties. The number of synthetic resins that can be made is vast, but relatively few have commercial importance. Well over 90% of all synthetic resins made today comprise no more than 20 types, although there are certain variations to be found within each type. Synthetic resins are familiar to most people as plastics, but they have other uses, such as in the manufacture of surface coatings, glues, synthetic fibres and textile fibres. The rapid growth of the synthetic resin industry is because ample supplies of the necessary raw materials are available from petroleum. Synthetic resins may be divided into two classes: ‘thermosetting’ and ‘thermoplastic’. Each class differs in its behaviour on being heated. The former do not soften; the latter soften, but regain rigidity on cooling. Both types are composed of large molecules (‘macromolecules’), but the difference in thermal behaviour is due to differences in internal structure. The large molecules of thermoplastics have a long-chain structure, with little branching. They do not link with each other chemically, although they may intertwine and form a cohesive mass with properties ranging from those of hard solids to those of soft pliable materials, in certain cases resembling rubber. On heating, the chain molecules can move more or less freely relative to each other, so that, without melting, the material softens and can flow under pressure and be moulded to any shape. On cooling, the moulded articles regain rigidity. Some resins require the addition of liquid plasticisers to improve the flow of the plastic material in the mould. In such cases, moulded articles are usually softer and more flexible than the products made from the unplasticised resins. Macromolecules of thermosetting resins are often strongly branched chains and are chemically joined by crosslinks, thus forming a complex network. On heating, there is less possibility of free movement, so the material remains rigid.

1


1.1 Production of Synthetic Resins Production of these resins also falls into two groups because there are, in general, two main types of chemical reaction by which they are made: ‘polycondensation’ and ‘polymerisation’.

1.2 Polycondensation Reactions In a polycondensation reaction, two or more chemicals are brought together and a reaction between them is initiated by using heat or a catalyst or both. The reaction proceeds with the elimination of water and the molecules joined by chemical bonds to form macromolecules, long-chain or crosslinked structures of the thermoplastic or thermosetting types, respectively. Many resins obtained by polycondensation are the thermosetting type. In the manufacture of these resins, chemical reactions are arrested at an intermediate stage in which the resins are temporarily thermoplastic; they are set in their final shape by the application of heat and pressure. At this stage the interlinking of the molecules occurs. Important thermosetting synthetic resins made by polycondensation, using petroleum chemicals as raw materials, include the phenol-formaldehyde (‘Bakelite’), ureaformaldehyde, alkyd- and epoxy- types. Resins with long-chain macromolecules obtained by polycondensation have thermoplastic properties. Polyesters (‘Terylene’) and polyamides (Nylon) are examples of polycondensations. The synthetic fibre Terylene (known as ‘Dacron’ in the USA) is a polyester formed by the reaction of ethylene glycol with terephthalic acid; the latter is obtained from p-xylene by oxidation:

2


Nylon-type fibres (polyamides) are manufactured from adipic acid, which can be made from cyclohexane or phenol. Adipic acid is condensed with hexa-methylene diamine, which is a derivative of adipic acid:

3


1.3 Polymerisation Reactions Resins produced by polymerisation reactions, known technically as ‘high polymers’, are rapidly increasing in number and importance as compared with polycondensation resins. High polymers are usually made by joining together into long chains several molecules with the same type of reactive points or groupings in their structure. These individual molecules are usually olefins or other compounds with double bonds, and are called ‘monomers’. The polymer molecule often contains hundreds of monomer units. The manufacture of high polymers therefore takes place in two stages: (i) production of the monomer, or repeating chemical unit; and (ii) polymerisation to a resin. If we take preparation of polyvinyl chloride as an example, we have: First stage

4


Second stage

It is possible to form polymers from two or even three monomers that may differ from one another in chemical form and yet be capable of linking end-to-end to form mixed monomer chains. These are known as ‘copolymers’, and they form the basis of the most important types of synthetic rubber. Further examples of polymerisations are: Styrene butadiene copolymer:

Polymethylmethacrylate:

Buna N rubber:

5

2

Determination of Elements

2.1 Non-metallic Elements Non-metallic elements such as boron, halogens, nitrogen, oxygen, phosphorus and sulfur can occur in polymers as major constituents present as impurities, or as components of low-percentage additions of additives containing the element. For example, addition of 0.5% dilauryl thiodipropionate antioxidant to a polymer during processing introduces parts per million (ppm) concentrations of sulfur into the final polymer. Other sources of non-metallic elements in polymers are catalyst residues and processing chemicals. It is advisable when starting polymer analysis to determine the content of various non-metallic and metallic elements first. Initially, these tests could be qualitative, simply to indicate the presence or absence of the element. All that is required is that the test is of sufficient sensitivity such that elements of importance are not missed. If an element is found in these tests, it may be necessary to determine it quantitatively. The analytical methods used to determine elements should be sufficiently sensitive to determine about 10 ppm of an element in the polymer. That is, in a polymer, the method must detect a substance present at 0.01% and containing down to 10% of the element in question. This requirement is met for almost all the important elements by using optical emission spectroscopy and X-ray fluorescence spectrometry (XRFS). The latter is applicable to all elements with an atomic number >12. Using these two techniques, all metals and non-metals down to an atomic number of 15 (phosphorus) can be determined in polymers at the required concentrations [1–4]. Nitrogen can be determined by micro Kjeldahl digestion techniques. In addition to the polymer, the polymer additive system may contain elements other than carbon, hydrogen, and oxygen. Detection of an element such as boron, halogens, nitrogen, phosphorus, silicon, or sulfur in a polymer indicates that the element originates in the polymer and not the additive system if the element is present at relatively high concentrations (e.g., several percent). For example, a high-density polyethylene may contain 0.2–1% chlorine originating from

7

Introduction to Polymer Analysis polymerisation residues and a polyvinyl chloride (PVC) homopolymer which contains >50% chlorine. Instrumentation available for the determination of the following total elements is the subject of this chapter: halogens; sulfur; halogens and sulfur; nitrogen; nitrogen, carbon, and sulfur; carbon, hydrogen, and nitrogen; and total organic carbon (TOC).

2.1.1 Halogens 2.1.1.1 Combustion Methods The Dohrmann DX 20B furnace combustion system is based on combustion of a sample to produce an hydrogen halide, which is then swept into a microcoulometric cell and estimated. It is applicable at total halide concentrations up to 100 μg/l, with a precision of ±2% at the level of 10 μg/l. The detection limit is about 0.5 μg/l. Analysis can be done in five minutes. A sample boat is available for carrying out analysis of solid samples. Mitsubishi also supplies a microprocessor-controlled automatic total halogen analyser (model TOX-10) which is very similar in operating principles to the Dohrmann system discussed previously, i.e., combustion at 800–900 ºC followed by coulometric estimation of the hydrogen halide produced.

2.1.1.2 Oxygen Flask Combustion Oxygen flask combustion methods have been used to determine chlorine in PVC [5] and traces of chlorine in polyolefins and chlorobutyl rubber [6]. Traces of chlorine have been determined in polyolefins [5] at 0–500 ppm. The Schöniger oxygen flask combustion technique requires a 0.1 g sample and a one litre conical flask. Chlorine-free polyethylene (PE) foil is employed to wrap the sample, which is then supported in a platinum wire attached to the flask stopper; water is the absorbent. Combustion takes place at atmospheric pressure in oxygen. The chloride formed is potentiometrically titrated in nitric acid/acetone medium with 0.01 M mercuric nitrate solution. In the method for determining chlorine in chlorobutyl and other chlorine-containing polymers, [6] the sample is combusted in a 1–2 litre oxygen-filled combustion flask containing 0.01 M nitric acid. After combustion, the flask is allowed to cool and 0.01 M silver nitrate added. The combustion solution containing silver chloride is evaluated

8

Determination of Elements turbidometrically at 420 mm using a grating spectrophotometer. Alternatively, to determine bromine, chlorine, iodine, or mixtures thereof, the combustion solution can be titrated with dilute standard silver nitrate solution, or can be evaluated by ion chromatography. Determination of fluorine in fluorinated polymers such as polytetrafluoroethylene (PTFE) is based on decomposition of the sample by oxygen flask combustion followed by spectrophotometric determination of the fluoride produced by a procedure involving the reaction of the cerium(III) complex of alizarin complexan (1,2-dihydroxy-anthraquinone-3-ylmethylamine N,N-diacetic acid). The blue colour of the fluoride-containing complex (maximum absorption, 565 nm) is distinguishable from the yellow of the free dye (maximum absorption, 423 nm) or the red of its cerium(III) chelate (maximum absorption, 495 nm). A method has been described [7] for the determination of chlorine in polymers containing chlorine, fluorine, phosphorus and sulfur that involves oxygen flask combustion over water, ethanol addition, and titration to the diphenylcarbazide indicator end point with 0.005 M mercuric nitrate:

2HCl + Hg(NO3)2 = HgCl2 + 2HNO3 Using this method, Johnson and Leonard [7] obtained from PTFE, 75.8% of fluorine using a silica or boron-free glass combustion flask against a theoretical value of 76%. They obtained a low fluorine recovery of 72.1% using a borosilicate glass combustion flask.

2.1.1.3 Alkali Fusion Methods Sodium peroxide is a useful reagent for the fusion of polymer samples preparatory to analysis for chlorine [8, 9] and bromine. The polymer is intimately mixed with sodium peroxide in an open crucible or with a mixture of sodium peroxide and sucrose in a micro-Parr bomb. Chlorine can be determined after acidification with nitric acid [9]. In a method for the determination of traces of bromine in polystyrene in amounts down to 100 ppm bromine, a known weight of polymer is mixed with pure sodium peroxide and sucrose in a micro-Parr bomb, which is then ignited. The sodium bromate produced is converted to sodium bromide by the addition of hydrazine as the sulfate:

2NaBrO3 + 3NH2NH2 = 2NaBr + 6H2O + 3N 2 The combustion mixture is dissolved in water and acidified with nitric acid. Bromine

9

Introduction to Polymer Analysis content of this solution is determined by potentiometric titration with standard silver nitrate solution. Fusion with sodium carbonate is a very useful method for the fusion of polymers that, upon ignition, release acidic vapours, e.g., PE containing traces of chlorine or PVC, both of which, upon ignition, release anhydrous hydrogen chloride. To determine chlorine accurately in the polymer in amounts down to 5 ppm, hydrogen chloride must be trapped in a solid alkaline reagent (e.g., sodium carbonate). In this method, PE is mixed with pure sodium carbonate and ashed in a muffle furnace at 500 ºC. The residual ash is dissolved in aqueous nitric acid, and then diluted with acetone. This solution is titrated potentiometrically with standard silver nitrate.

2.1.1.4 Physical Methods for Determining Halogen Manatt and co-workers [10] used carbon-13 nuclear magnetic resonance (13C-NMR) to determine chlorine in polystyrene. Williams and co-workers [11] determined bromine in brominated polystyrene and poly(2,6 dimethyl 1,4 phenylene oxide) using 13 C-NMR. X-ray emission analysis has been used to determine the ratio of chlorine sulfur in copolymers based on poly(3-methyl thiophen) [12].

2.1.2 Sulfur 2.1.2.1 Combustion Methods The Mitsubishi trace sulfur analyser models TS-02 and TN-02(S) involve a microcombustion procedure in which sulfur is oxidised to sulfur dioxide, which is then titrated coulometrically with tri-iodide ions generated from iodide ions:

SO2 + I3- + H2O 3I - I3- +3e-

SO 3 + 3I- + 2H +

2.1.2.2 Sodium Peroxide Fusion Colson described an alkali fusion for the determination of down to 500 ppm of sulfur in polymers [13, 14], in which the sulfate in the digest is determined by titration with N/100 sodium hydroxide or by photometric titration with N/100 barium perchlorate.

10


2.1.2.3 Oxygen Flask Combustion To determine sulfur in amounts down to 500 ppm in polyolefins, the sample is wrapped in filter paper and burnt in a closed conical flask filled with oxygen at atmospheric pressure. The sulfur dioxide produced in the reaction reacts with dilute hydrogen peroxide solution contained in the reaction flask to produce an equivalent amount of sulfuric acid [14]:

H2SO 3 + H2O2 = H2SO 4 + H2O The sulfuric acid is estimated by visual titration with M/500 or M/50 barium perchlorate using Thorin indicator. The repeatability of this method is ±40% of the sulfur content determined at the 500 ppm sulfur level, improving to ±2% at the 1% level. Concentrations of chlorine and nitrogen in the sample may exceed the sulfur concentration several times over without causing interference. Fluorine does not interfere unless present in concentrations exceeding 30% of the sulfur content. Phosphorus and metallic constituents interfere when present in moderate amounts. A photoelectric method of end-point detection overcomes the difficulties associated with visual end-point detection because assessment of end-point is independent of the observation of colours from individual operators.

2.1.3 Phosphorus 2.1.3.1 Acid Digestion Phosphorus has been determined [15, 16] in thermally stable polymers by mineralisation with a nitric acid – perchloric acid mixture and subsequent titration with lanthanum nitrate or by photometric determination of the phosphomolybdic blue complex [17].

2.1.4 Nitrogen 2.1.4.1 Combustion Methods Mitsubishi supply two total nitrogen analysers: the model TN-10 and the model TN-05 microprocessor-controlled chemiluminescence total nitrogen analysers. These analysers measure down to micrograms per litre amounts of nitrogen in solid and liquid samples.

11

Introduction to Polymer Analysis The sample is introduced into the combustion tube packing containing oxidative catalyst under oxygen carrier gas. High-temperature oxidation (800–900 °C) occurs and all chemically bound nitrogen is converted to nitric oxide (NO): R–N→CO2 + NO. Nitric oxide then passes through a drier to remove water formed during combustion and moves to the chemiluminescence detector, where it is mixed with ozone to form excited nitrogen dioxide (NO2*):

Rapid decay of the NO2* produces radiation in the range 590–2900 nm. It is detected and amplified by a photomultiplier tube. The result is calculated from the signal produced and is given in milligrams per litre or as a percentage. Dohrmann also supplies an automated nitrogen analyser with video display and data processing (model DN-1000) based on similar principles that is applicable to the determination of nitrogen in solid and liquid samples down to 0.1 mg/l. The Dohrmann DN-1000 model can be converted to the determination of sulfur and chlorine by adding the MCTS 130/120 microcoulometer detector modules. The control module, furnace module, and all the automated sample inlet modules are common to both detectors. The system automatically recognises which detector and sample inlet is present and sets the correct operating parameters for fast, simple conversion between detection of nitrogen, sulfur, and chlorine. Equipment for automated Kjeldahl determinations of organic nitrogen in water and solid samples is supplied by Tecator Limited. Its Kjeltec system 1 streamlines the Kjeldahl procedure, resulting in higher speed and accuracy compared with classic Kjeldahl measurements. Apart from the chemical Kjeldahl digestion procedure for the determination of organic nitrogen, acid digestion of polymers has found little application. One of the problems is connected with the form in which the polymer sample occurs. If it is in the form of a fine powder, or a very thin film, digestion with acid may be adequate to enable the relevant substance to be quantitatively extracted from the polymer. Low nitrogen results would be expected for polymers in a larger granular form, and classic microcombustion techniques are recommended for the analysis of such samples. Hernandez [17] described an alternative procedure based on pyro-chemiluminescence which he applied to the determination of 250–1500 ppm nitrogen to PE. Nitrogen in the sample is subject to oxidative pyrolysis to produce nitric oxide. This, when contacted with ozone, produces a metastable nitrogen dioxide molecule which, as it relaxes to a stable state, emits a photon of light. This emission is measured quantitatively at 700–900 nm.

12


2.1.4.2 Physical Method for Determination of Total Nitrogen Wirsen [18] used size exclusion chromatography, and infrared spectroscopy and size exclusion chromatography low-angle light scattering to determine the nitrogen content of cellulose nitrate.

2.1.5 Silicon Silica has been determined in PE films by a method based on near-infrared spectroscopy. A single baseline point at the minimum near 525 cm–1 was found to be best for measurements of peak height. An additional baseline point below 430 cm–1 gave poorer results because of the increased noise at longer wavelengths due to atmospheric absorption. Peak area measurements were confined to the range 525–469 cm–1 for the same reason [19]. Measurements of height and area gave an error index close to 1%, but derivative methods were considerably poorer. In general, derivative spectra show increased noise levels so they are unlikely to be useful except when they are overlapping bands. Results obtained with the ratio program also showed a higher error index. The band index ratio method avoids uncertainty associated with measuring film thickness, but in this case the error resulting from using a rather weak reference band appears greater. Combustion in a Parr bomb with sodium peroxide, sucrose and benzoic acid in a gelatin capsule is the basis for determining silicon in polymers. Sulfur, halogens, phosphorus, nitrogen and boron do not interfere.

2.1.6 Boron Yoshizaki [20] demonstrated a method for determining boron in which 0.1 g of polymer is digested with concentrated nitric acid in a sealed ampoule to convert organoboron compounds to boric acid. The digest is dissolved in methyl alcohol and boron is estimated flame-photometrically at 595 nm. Chlorine and nitrogen do not interfere.

2.1.7 Total Organic Carbon Dohrmann supplies a TOC analyser. Persulfate reagent is continuously pumped at a low flow rate through the injection port (and the valve of the autosampler) and then into the ultraviolet reactor. A sample is acidified, sparged, and injected directly into the reagent stream. The mixture flows through the reactor where organics are oxidised by the photon-activated reagent. The light-source envelope is in direct contact with the

13

Introduction to Polymer Analysis flowing liquid. Oxidation proceeds rapidly, the resultant carbon dioxide is stripped from the reactor liquid and carried to the carbon dioxide-specific non-dispersive infrared detector. The Shimadzu TOC-500 total organic carbon analyser is a fully automated system capable of determining between 1 μg/l and 3000 μg/l TOC. OIC Analytical Instruments produce the fully computerised model 700 TOC analyser. This is applicable to solids. Persulfate oxidation at 90–100 °C followed by nondispersive infrared spectroscopy is the principle of this instrument.

2.1.8 Total Sulfur/Total Halogen The Mitsubishi TSC-10 halogen–sulfur analyser expands the technology of the TOX-10 to include measurement of total chlorine and total sulfur. Model TSX-10, which consists of the TOX-10 analyser module and a sulfur detection cell, measures total sulfur and total chlorine in liquid and solid samples over a sensitivity range of milligrams per litre to a percentage. Dohrmann also produces an automated sulfur and chlorine analyser (models TCTS 130/120). This instrument is based on combustion microcoulometric technology.

2.1.9 Nitrogen, Carbon, and Sulfur The NA 1500 analyser supplied by Carlo Erba can determine these elements in 3–9 minutes down to 10 mg/l with a reproducibility of ±0.1%. A 196-position autosampler is available. ‘Flash combustion’ of the sample in the reactor is a key feature of the NA 1500. This results when the sample is dropped into the combustion reactor which has been enriched with pure oxygen. The normal temperature in the combustion tube is 1020 ºC and reaches 1700–1800 °C during flash combustion. In the chromatographic column, combustion gases are separated so that they can be detected in sequence by the thermal conductivity detector. The output signal is proportional to the concentration of the elements. A data processor plots the chromatogram, automatically integrates the peak areas, and gives retention times, percentage areas, baseline drift, and attenuation for each run. It also computes blank values, constant factors, and relative average elemental contents.

14


2.1.10 Carbon, Hydrogen, and Nitrogen Perkin Elmer supplies an analyser (model 2400 CHN or PE 2400 series II CHNS/O analysers) suitable for determining these elements in polymers [21–26]. The sample is first oxidised in a pure oxygen environment. The resulting combustion gases are then controlled to exact conditions of pressure, temperature, and volume. Product gases are separated under steady-state conditions and swept by helium or argon into a gas chromatograph for analysis of the components.

2.1.11 Oxygen Flask Combustion: Ion Chromatography Combustion of polymers in an oxygen-filled flask over aqueous solutions of appropriate reagents converts elements such as halogens, phosphorus and sulfur into inorganic ions. For example: chlorine, bromine, iodine → chloride, bromide, iodide sulfur → sulfate phosphorus → phosphate Subsequent analysis of these solutions by ion chromatography [27] enables the concentrations of mixtures of these anions (i.e., original elements) to be determined rapidly, accurately, and with great sensitivity. Figure 2.1(a) shows separation of halides, nitrate, phosphate and sulfate obtained in six minutes by ion chromatography using a Dionex A54A anion exchange separator. A further development is the Dionex HPLC AS5A-SU analytical anion exchange column. Quantification of all the anions in Figure 2.1(b) would require at least three sample injections under different eluent conditions.

2.1.12 XRFS The X-ray fluorescence (XRF) technique has been applied extensively to the determination of macro- and micro-amounts of non-metallic elements in polymers.

15


(b) Gradient: 0-50 mM NaOH over 18 minutes

I−

HPO42−

0

2

4 6 Minutes

8

10

0

2

4

Phosphate

Fluorine

3 μS

Nitrate Malate

F

HPIC/AS4A 2 ml/min 1 mM tyrosine, 3 mM NaOH 10 μS background AMMS

Bromide

SO42−

−

Sulfate

(a)

Chlorine Nitrite

NO2− BR− NO3− Cl−

6 Minutes

8

10

12

Figure 2.1 Ion chromatograms obtained with Dionex instrument using (anodic) AMMS and (cathodic) CMMS micromembrane suppression. (a) anions with micromembrane suppressor (Dionex A54A column); (b) multi-component analysis by ion chromatography (Dionex A55A, 5 μm column). Source: Author’s own files.

An interesting phenomenon has been observed in applying the XRF method to the determination of ppm of chlorine in hot-pressed discs of low-pressure polyolefins. In these polymers, the chlorine is present in two forms, organically bound and inorganic, with titanium chloride compounds resulting as residues from the polymerisation catalyst. The organic part of the chlorine is determined by XRF without complications. During hot processing of the discs, there is a danger that some inorganic chlorine will be lost. This can be completely avoided by intimately mixing the powder with alcoholic potassium hydroxide, then drying at 105 °C before hot pressing discs. The results (Tables 2.1 and 2.2) illustrate this effect. Considerably higher total chlorine contents are obtained for alkali-treated polymers. Another example of the application of XRFS is determination of tris(2,3dibromopropyl) phosphate on the surface of flame-retardant polyester fabrics [28]. The technique involves fabric extraction with an organic solvent followed by solvent analysis by XRF for surface bromine and by high-pressure liquid chromatography (HPLC) for molecular tris(2,3-dibromopropyl) phosphate. The technique has been applied to the determination of hydroxy groups in polyesters [29, 30].

16


Table 2.1 Comparison of chlorine contents by X-ray method and chemical method (averages in parenthesis) X-ray on discs

Chemical methods on same discs as used for X-ray analysis

Chemical method on powder*

865, 841 (840)

700

786, 761 (773)

535, 570 (522)

606

636, 651

785, 675 (730)

598

650, 654 (652)

625, 675 (650)

600

637, 684 (660)

895, 870 (882)

733

828, 816 (822)

*Analysis carried out on samples which have been treated with alcoholic potash to avoid losses of chlorine when preparing discs. Source: Author’s own files.

Table 2.2 Determination of chlorine by X-ray procedure A: Polymer not treated with alcoholic potassium hydroxide before analysis, (ppm chlorine), X-ray fluorescence in polymer discs. Average of 2 discs (A)

B: Polymer treated with alcoholic potassium hydroxide before analysis, X-ray fluorescence in polymer discs. Average of 2 discs (B) (ppm chlorine)

Difference between average chlorine contents obtained on potassium hydroxidetreated and untreated samples (B) - (A) (ppm chlorine)

510

840

330

422

552

130

440

730

290

497

650

153

460

882

422

Source: Author’s own files

With n = 1 – 100 and x = 2 (polyethylene terephthalate) or 4 (polybutylene terephthalate) and ester-interchange elastomers of 4-polybutylene terephthalate and

17

Introduction to Polymer Analysis polypropylene glycol. The hydroxyl groups in these products are determined by acetylation with an excess of dichloroacetic anhydride in dichloroacetic acid, and measurement of the amount of acetylation by a chloride determination carried out on the derivative. The XRF method of Wolska [31] has been applied to the determination of bromine and phosphorus in polymers. Various other workers have applied this technique to the determination of chlorine and sulfur [32] and various other elements [33, 34]. Niino and Yabe [35] used XRF to determine the chlorine content of products obtained in the photo-irradiation of polyvinylidene chloride film.

2.1.13 Thermogravimetric Analysis This technique was used by Coulson and co-workers [36] to determine chlorine in chlorinated rubbers.

2.2 Metals Different techniques have evolved for trace-metal analysis of polymers. In general, techniques come under two broad headings: destructive techniques and non-destructive techniques. Destructive techniques are techniques in which the sample is decomposed by a reagent and then the concentration of the element in the aqueous extract determined by a physical technique. These include atomic absorption spectrometry, graphite furnace atomic absorption spectrometry (GFAAS); atom-trapping atomic absorption spectrometry, cold-vapour atomic absorption spectrometry (CVAAS), Zeeman atomic absorption spectrometry (ZAAS); inductively coupled plasma atomic emission spectrometry (ICP-AES), visible spectrometry, or polarographic or anodic scanning voltammetric techniques.

2.2.1 Destructive Techniques 2.2.1.1 Atomic Absorption Spectrometry (AAS) AAS has been the standard tool employed by analysts for the determination of trace levels of metals since shortly after its inception in 1955. A fine spray of the analyte is passed into a suitable flame, usually oxygen–acetylene or nitrous oxide–acetylene, which converts the elements to an atomic vapour. Through this vapour is passed

18

Determination of Elements radiation at the appropriate wavelength to excite ground-state atoms to the first excited electronic level. The amount of radiation absorbed can be measured and directly related to the atomic concentration: a hollow cathode lamp is used to emit light with the characteristic narrow-line spectrum of the analyte element. The detection system consists of a monochromator (to reject other lines produced by the lamp and background flame radiation) and a photomultiplier. Another key feature of this technique involves modulation of the source radiation so that it can be detected against the strong flame and sample emission radiation. This technique can determine a particular element with little interference from other elements, but has two major limitations: (i) it does not have the highest sensitivity; and (ii) only one element at a time can be determined. This has reduced the extent to which it can be used. Increasingly, due to their superior sensitivity, AAS instruments can implement graphite furnace techniques. Figure 2.2(a) and (b) show the optics of one particular single-beam flame spectrometer (Perkin Elmer 2280) and a double-beam instrument (Perkin Elmer 2380).

2.2.1.2 GFAAS The GFAAS technique was first developed in 1961 by L’vov. It was an attempt to improve detection limits. Instead of being sprayed as a fine mist into the flame, a measured portion of the sample is injected into an electrically heated graphite boat or tube, allowing a larger volume of sample to be handled. By placing the sample on a small platform inside the furnace tube, atomisation is delayed until the surrounding gas within the tube has heated sufficiently to minimise vapour phase interferences, which would otherwise occur in a cooler gas atmosphere. The sample is heated to a temperature slightly above 100 °C to remove free water, then to a temperature of several hundred degrees centigrade to remove water of fusion and other volatiles. The sample is heated to a temperature near 1000 °C to atomise it and the signals produced are measured by the instrument. The problem of background absorption in this technique is solved by using a broadband source, usually a deuterium arc or a hollow cathode lamp, to measure the background independently and subsequently to subtract it from the combined atomic and background signal produced by the analyte hollow cathode lamp. By interspersing the modulation of the hollow cathode lamp and ‘background corrector’ sources, measurements are done apparently simultaneously.

19


Photomultiplier

Monochromator D2ARC

Beam splitter

Primary source

(a)

Photomultiplier

Monochromator D2ARC

Chopper (b)

Primary source

Figure 2.2 Optics Perkin Elmer Model 2280 single-beam atomic absorption spectrometer; (b) Optics Perkin Elmer 2380 double-beam atomic absorption spectrometer. Source: Author’s own files.

Graphite furnace techniques are about one order of magnitude more sensitive than direct injection techniques. Lead can therefore be determined down to 50 μg/l using the graphite furnace modification of the technique.

20


2.2.1.3 Atom Trapping Technique The sensitivity difference between direct flame atomisation and furnace atomisation has been bridged via the general method of atom trapping as proposed by Watling [37]. A silica tube is suspended in the air–acetylene flame. This increases the residence time of the atoms within the tube and therefore within the measurement system. Further devices such as water-cooled systems that trap the atom population on cool surfaces and then subsequently release them by temporarily halting the coolant flow are sometimes employed. The application of atom-trapping AAS for the determination of lead and cadmium has been discussed by Hallam and Thompson [38].

2.2.1.4 Vapour Generation Atomic Absorption Spectrometry (VGAAS) In the past, certain elements, e.g., antimony, arsenic, bismuth, germanium, lead, mercury, selenium, tellurium, and tin, were difficult to measure by direct AAS [39–45]. A novel technique of atomisation, known as ‘vapour generation via generation of the metal hydride’, has evolved. This technique has increased enormously the sensitivity and specificity for these elements [41–43, 45]. In these methods, the hydride generator is linked to an atomic absorption spectrometer (flame graphite furnace) or inductively coupled plasma optical emission spectrometer (ICP-OES) or an inductively coupled plasma mass spectrometer (ICP-MS). Typical detection limits achievable by these techniques range from 3 μg/l (arsenic) to 0.09 μg/l (selenium). This technique makes use of the property that these elements exhibit: formation of covalent, gaseous hydrides that are unstable at high temperatures. Antimony, arsenic, bismuth, selenium, tellurium, and tin (and to a lesser degree germanium and lead) are volatilised by the addition of a reducing agent such as sodium tetrahydroborate(III) to an acidified solution. Mercury is reduced by stannous chloride to the atomic form in a similar manner. Automating the sodium tetrahydroborate system based on continuous flow principles represents the most reliable approach in the design of commercial instrumentation. Pahlavanpour and co-workers [46] described a simple system for multi-element analysis using an ICP spectrometer based on the sodium tetrahydroborate approach. PS Analytical Limited developed a reliable and robust commercial analytical hydride generator system along similar lines using different pumping principles from those discussed by Pahlavanpour and co-workers [46]. A further major advantage of this range of instruments is that different chemical procedures can be operated in the instrument with little (if any) modification. In

21

Introduction to Polymer Analysis addition to using sodium tetrahydroborate as a reductant, stannous chloride can be used for the determination of mercury at very low levels. The main advantage of hydride generation AAS for the determination of antimony, arsenic, and selenium is its superior sensitivity. Low concentrations of mercury, arsenic and selenium in solution, down to 10–20 ppm of these elements, can be determined in polymer digests.

2.2.1.5 Zeeman AAS The Zeeman technique, though difficult to establish, has an intrinsic sensitivity perhaps five-times greater than that of the graphite furnace technique (e.g., 1 μg/l detection limit for lead). The Zeeman effect is exhibited when the intensity of an atomic spectral line, emission or absorption, is reduced when the atoms responsible are subjected to a magnetic field, with nearby lines arising instead (Figure 2.3). This makes the Zeeman effect a powerful tool for the correction of background attenuation caused by molecules or particles that do not normally show such an effect. The technique is to subtract from a ‘field-off’ measurement the average of ‘field-on’ measurements made just beforehand and just afterwards. The simultaneous, highly resolved graphic display of the analyte and the background signals on a video screen provides a means of reliable monitoring of the determination and simplifies methods development. The stabilised temperature platform furnace eliminates chemical interferences to such an extent that in most cases personnel- and cost-intensive sample preparation steps, such as solvent extractions, as well as the time-consuming method of additions, are no longer required. The advantages of Zeeman background correction are: •

Correction over the complete wavelength range.

•

Correction for structural background.

•

Correction for spectral interferences.

•

Correction for high background absorptions.

•

Single-element light source with no possibility of misalignment.

22

Determination of Elements Magnet ‘off’

(a)

Background Absorption profile

Emission line ( – oriented component)

(b)

Magnet ‘on’

Absorption on the line component ( polarised) component ( polarised)

component ( polarised)

Emission line ( – oriented component)

Figure 2.3 Zeeman patterns. (a) analyte signal plus background. (b) background only. Source: Author’s own files The analytical range must also be considered when assessing overall performance with a Zeeman-effect instrument. For most normal class transitions, S components will be completely separated at sufficiently high magnetic fields. Consequently, the analytical curves will, in general, be similar to those obtained by standard AAS, but some overlap may occur for certain anomalous transitions. In these cases, curvature will be greater and may be so severe as to produce double-valued analytical curves. Figure 2.4, which shows calibration curves for copper, illustrates the reason for this behaviour. The Zeeman pattern for copper (324.8 nm) is particularly complex due to hyperfine structure. The dashed lines represent the separate field-off and field-on absorbance measurements. As sample concentration increases, field-off absorbance begins to saturate as in standard AAS. The S absorbance measured with the field-on saturates at higher concentrations because of the greater separation from the emission line. When the increase in S absorbance exceeds the incremental change in the field-off absorbance the analytical curve (shown as the solid line) rolls over back towards the concentration axis. This behaviour can be observed with all Zeeman designs regardless of how the magnet is positioned or operated. Roll-over introduces the possibility of ambiguous results, particularly if peak area is being measured. 23

Introduction to Polymer Analysis 2 1

ZAA signal (field off minus field on)

Magnetic field off

Magnetic field on

0.1

0.01 0.1

1

10

100

Cu concentration (ng)

Figure 2.4 Copper calibration curves (24.8 nm) measured with a Zeeman spectrometer. Source: Author’s own files.

2.2.1.6 ICP-AES Inductively coupled plasma is formed by coupling the energy from a radio frequency (1–3 kW or 27–50 MHz) magnetic field to free electrons in a suitable gas. The magnetic field is produced by a two- or three-turn water-cooled coil, and the electrons are accelerated in circular paths around the magnetic field lines that run axially through the coil. The initial electron ‘seeding’ is produced by a spark discharge but, once the electrons reach the ionisation potential of the support gas, further ionisation occurs and stable plasma is formed. Neutral particles are heated indirectly by collisions with the charged particles upon which the field acts. Macroscopically, the process is equivalent to heating a conductor by a radiofrequency field, the resistance to eddy current flow producing joule heating. The field does not penetrate the conductor uniformly and therefore the largest current flow is at the periphery of the plasma. This is the so-called ‘skin’ effect and, coupled with suitable gas-flow geometry, it produces an annular or doughnut-shaped plasma. Electrically, the coil and plasma form a transformer, with the plasma acting as a oneturn coil of finite resistance. If mass spectrometric determination of the analyte is to be incorporated, then the source must also be an efficient producer of ions. Greenfield and co-workers [47] were the first to recognise the analytical potential of annular ICP. 24

Determination of Elements Wendt and Fassel [48], reported early experiments with a ‘teardrop’-shaped inductively coupled plasma, but later described the medium-power, 1–3 kW, 18 mm annular plasma favoured in modern analytical instruments [49]. The present generation of ICP emission spectrometers provides limits of detection in the range 0.1–500 μg/l of metal in solution; a substantial degree of freedom from interferences; and a capability for simultaneous multi-element determination facilitated by a directly proportional response between the signal and the concentration of the analyte over a range of about five orders of magnitude. The commonest method of introducing liquid samples into the ICP is by using pneumatic nebulisation [50], in which the liquid is dispensed into a fine aerosol by a high-velocity gas stream. The fine gas jets and liquid capillaries used in ICP nebulisers may cause inconsistent operation and even blockage when solutions containing high levels of dissolved solids, or particular matter, are used. Such problems have led to the development of new types of nebuliser, the most successful being based on a principle originally described by Babington. In these, the liquid is pumped from a wide-bore tube and then to the nebulising orifice by a V-shaped groove [51] or by the divergent wall of an over-expanded nozzle [52]. Such devices handle most liquids and even slurries without difficulty. Two basic approaches are used for introducing samples into the plasma: (i) indirect vaporisation of the sample in an electrothermal vaporiser, e.g., a carbon rod or tube furnace or heated metal filament as commonly used in AAS [53–55]; and (ii) inserting the sample into the base of the ICP on a carbon rod or metal filament support [56, 57]. There are two main types of ICP spectrometer systems. The first is the monochromator system for sequential scanning. This consists of a high-speed, high-resolution scanning monochromator viewing one element wavelength at a time. Figure 2.5(a) shows a onechannel air path double monochromator design with a pre-monochromator for order sorting and stray light rejection, and a main monochromator to provide resolution of up to 0.02 nm. The air path design can measure wavelengths in the range 190–900 nm. The wide wavelength range enables measurements to be done in the ultraviolet (UV), visible, and near-infrared regions of the spectrum (allowing determinations of elements from arsenic at 173.70 nm to caesium at 852.1 nm). A second design (Figure 2.5(b)) is a vacuum monochromator design allowing measurements in the wavelength range 160–500 nm. The exceptionally low wavelength range enables determination of trace levels of non-metals such as bromine at 163.34 nm as well as metals at low UV wavelengths, such as the extremely sensitive aluminium emission line at 167.08 nm. Boron, phosphorus or sulfur can be routinely determined using interference-free emission lines.

25


M2 (a)

PMT L1

Hg Lamp

G1 S1 S2

G2

S3 RP

M1

M3

Ar (b)

Hg lamp

G R

F L1

PMT M

Vacuum

Ar

Figure 2.5 (a) A double monochromator consisting of an air-path monochromator with a pre-monochromator for order sorting and stray light rejection to determine elements in the range 190–900 mm; (b) Vacuum UV monochromator: an evacuated and argon purged monochromator to determine elements. Source: Author’s own files.

26


Interfaces - Labnet network - RS 232 C - HEEE 488

Printer

Central processor Mass storage

Data acquisition system model 1000 PX

Stepper motor

keyboard

Concave grating

ICP source 160 mm

Vacuum polychromator

Arc/spark source Up to 64 channels

Stepper motor

820 mm

Vacuum scanning monochromator

Stop-flow GMK nebuliser

Figure 2.6 Polychromator system for inductively coupled plasma atomic emission spectrometer. Source: Author’s own file.

The sequential instrument, equipped with either or both monochromators facilitates the sequential determination of up to 63 elements in turn, at a speed as fast as 18 elements per minute in a single sample. Having completed the analysis of the first sample, usually in less than one minute, it proceeds to the second sample, and so on. The second main type of system is the polychromator system for simultaneous scanning. The polychromator systems scan many wavelengths simultaneously, i.e., several elements are determined simultaneously at higher speeds than are possible with monochromator systems. It then moves on to the next sample. A typical system is shown in Figure 2.6.

27

Introduction to Polymer Analysis Briseno and co-workers [58] quantified inorganic dopants in polypyrrole films by a combination of electrochemistry and ICP-AES.

2.2.1.6.1.Hybrid Inductively Coupled Plasma Systems 2.2.1.6.2 Chromatography–ICP Direct introduction of a sample into ICP produces information on only total element content. It is now recognised that information on the form of the element present, or trace element speciation, is important in various applications. One way of obtaining quantitative measurement of trace element speciation is to couple the separation power of chromatography to the ICP as a detector. Because most interesting trace metal speciation problems concern non-volatile or thermally unstable species, HPLC becomes the separation method of choice. HPLC as the separation technique requires introduction of a liquid sample into the ICP with the attendant problem of sample introduction.

2.2.1.6.3 Flow Injection with ICP A steady-state signal is obtained when a solution of an element is nebulised into the plasma in conventional ICP-OES. In flow injection [59] a carrier stream of solvent is fed continuously through a 1 mm id tube to the nebuliser using a peristaltic pump, and into this stream is injected, via a sampling valve, a discrete volume of a solution of the element of interest. When the sample volume injected is suitably small, a transient signal is obtained (as opposed to a steady-state signal which is obtained with larger sample volumes) and it is this transient signal that is measured. Very little sample dispersion occurs under these conditions, the procedure is very reproducible, and sample rates of 180 samples per hour are feasible.

2.2.1.6.4 Inductively Coupled Plasma with Atomic Fluorescence Spectrometry (ICP-AFS) Atomic fluorescence is the process of radiation activation followed by radiation deactivation, unlike atomic emission which depends on the collisional excitation of the spectral transition. For this, ICP is used to produce a population of atoms in the ground state and a light source is required to provide excitation of the spectral transitions. Whereas a multitude of spectral lines from all the accompanying elements are emitted by the atomic emission process, the fluorescence spectrum is relatively simple, being confined principally to the resonance lines of the element used in the excitation source.

28

Determination of Elements ICP is a highly effective line source with a low background continuum. It is optically thin – it obeys Beer’s law – and therefore exhibits little self-absorption. It is also a very good atomiser and the long tail flame issuing from the plasma has such a range of temperatures that conditions favourable to the production of atoms in the ground state for most elements are attainable. It is therefore possible to use two plasmas in one system: (i) source plasma to supply the radiation to activate the ground state atoms; and (ii) another to activate the atomiser. This atomic fluorescence (AFS) mode of detection is relatively free from spectral interference, the main drawback of ICP-OES. Good results have been obtained using a high-power (6 kW) ICP as a source and a low-power (3 mg should be avoided. Small sample sizes necessitate more sensitive types of gas chromatograph detectors such as flame ionisation. In particular, circumstances in which the occurrence of a microstructural features is being studied, failure to use a sufficiently sensitive detector could result in the pyrolysis product being missed. A limited increase in sample weight, say from 1–2 mg to 4 mg, may be permitted to improve sensitivity. In-line hydrogenation is a useful innovation for simplifying the pyrogram obtained for polymers that produce complicated mixtures, but should be used with caution in fundamental studies. More information may be obtained by carrying out studies with and without in-line hydrogenation. The type of gas chromatography separation column used should be the subject of close scrutiny. The information gained in pyrolysis studies is only as good as the degree and type of separation achieved on the column and, certainly in the early stages of investigation work, various columns should be studied. Quantitative measurements of the amounts of various pyrolysis products can, in many instances, be correlated with the percentage composition of a copolymer, or with the concentration or a particular microconstituent in the polymer.

4.3.1 Ethylene-Based Copolymers van Schooten and Mostert [35] and van Schooten and co-workers [36] applied their pyrolysis–hydrogenation–gas chromatography technique to copolymers, an analysis which presents difficulties in solvent solution–IR methods, especially with samples that are only partly soluble in suitable solvents (e.g., CCl4). Because the hydrogenation pyrogram of polyethylene consists almost exclusively of normal alkanes and that of polypropylene isoalkanes, the ratio of the peak heights of a n-alkane to an iso-alkane is a good measure of copolymer composition. The ratio n-C7 (2-methyl C7:4-methyl C7) was found to be good measure of ethylene–propylene ratio in copolymers. 4.3.1.1 Ethylene–butene-1 A Py-GC method has been described [37–39] for the determination of the composition or an ethylene–butene-1 copolymer containing up to about 10% butane.

135

Introduction to Polymer Analysis This technique has been applied to the gas chromatography of ethylene–butene copolymers [37]. Pyrolysis was carried out at 410 °C in an evacuated gas vial and the products swept into the gas chromatograph. Under these pyrolysis conditions, it is possible to analyse the pyrolysis gas components and obtain data within a range of about 10% relative. The peaks observed on the chromatogram were methane, ethylene, ethane, combined propylene and propane, isobutene, 1-butene, trans-2-butene, cis2-butene, 2-methyl-butene and n-pentane. A typical pyrolysis chromatogram for polyethylene is shown in Figure 4.7.

6

5

C 2H 6

C3H6 + C3H8

Determination of functional groups

Response

4

0

CH4 × 1600 C2H4

n-C4H10 C4H6-1 i-C4H10

1

C4H8-2 (trans)

2 C4H8-2 (cis)

2-Methylbutane

n-C5H12

3

Start

Retention time

Figure 4.7 Typical ethylene-butene-1 pyrolysis chromatogram. Reprinted with permission from F.W. Neumann and H.G. Nadeau, Analytical Chemistry, 1963, 10, 1454. © 1963, ACS

The relationship between the amount of ethylene produced on pyrolysis and the amount of butane in the ethylene–butene copolymer was determined by an IR analysis for ethyl branches. The y intercept of 16.3% ethylene should represent that amount of ethane which would result from a purely linear polyethylene. An essentially unbranched Phillips-type polyethylene polymer yielded 14.5% ethylene, which is fairly close to the predicted 16.3%.

136

Determination of Monomer Ratios in Copolymers Wang and Smith [40] also studied the composition and microstructure of styrene/ methyl methacrylate and styrene–n-butyl acrylate [41] copolymers. The composition was quantified by Py-GC using monomer peak intensity. Because of the poor stability of methyl methacrylate oligomers, neither methyl methacrylate dimer nor methyl methacrylate trimers were detected under normal pyrolysis conditions. The numberaverage sequence length for styrene was determined by pure and hybrid trimer peak intensities. The number-average sequence length of methyl methacrylate was determined using formulae that incorporate composition and the number-average sequence length of styrene. This method is a new approach for the investigation of the microstructure of copolymers that do not produce dimer and trimer peaks upon pyrolysis. Figure 4.8 shows a typical pyrogram of a 39:60 wt% styrene–methyl methacrylate copolymer. The identification of all dimer and trimer peaks was accomplished by comparing chromatogram retention times with a literature chromatogram [41], as well as by comparing mass spectra obtained from Py-GC–mass spectrometry in the electron ionisation (EI) mode and chemical ionisation (CI) mode. The distinction of hybrid trimer peaks of styrene–styrene–methyl methacrylate and styrene–methyl methacrylate–styrene was accomplished by comparing chromatograms of styrene– methyl methacrylate homogenous copolymer and styrene–methyl methacrylate alternating copolymer.

s M

SS

FID signal intensity (arbitary unit)

SM

SSM SMS

SSM SSS

20

0

10

25

30

20 30 Retention time (min)

35

40

40

45

50

Figure 4.8 Pyrogram of 39.61% styrene (S)–methyl methacrylate (M) copolymer. Hybrid trimer peaks SSM, SMS and other dimer and trimer peaks. Reproduced with permission from F.Y. Wang and P.B. Smith, Analytical Chemistry, 1996, 68, 17, 3033. © 1996, ACS

137


4.3.1.2 Ethylene - butadiene Krishen [42] obtained the products listed in Table 4.10 by pyrolysis of ethylene– butadiene rubber and ethylene–propylene–diene terpolymer. He showed that the 2-methyl–2-butene peak was linear with the natural rubber content of the sample. Styrene-butadiene rubber was determined from the peak area of the 1,3-butadiene peak. The ethylene–propylene–terpolymer content was deducted from the 1-pentane peak area of the pyrolysis products.

Table 4.10 Peak identification and relative retention data (tricresylphosphate column at 35 oC Peak number in figure

Compound

Relative retention (nonane = 1.0000)

1

Methane

0.0009

2

Ethane + Ethane

0.0082

3

Propane

0.0268

4

Propane

0.0.347

5

2-Methylpropane

0.0849

6

Propadine + Butane

0.0948

7

1-Butane + 2-Methylpropane

0.1048

8

Trans-2-Butene

0.1409

9

Cis-2-Butene

0.1649

10

1,3-Butane

0.1769

11

2-Methyl-1-Butene

0.2074

12

1-Pentene

0.3038

13

2-Methyl-1-Butene

0.2074

14

Trans-2-Pentene

0.3848

15

Cis-2-Pentene

0.3993

16

2-Methyl-2-Butene

0.4476

17

Isoprene

0.5397

Reproduced with permission from A. Krishen, Analytical Chemistry, 1972, 44, 3, 494. © 1972, ACS

4.3.3 Vinylidene Chloride – Vinyl Chloride Copolymers Wang and co-workers [43] used Py-GC to elucidate the composition of and carry out structural studies on vinyl chloride–vinylidene chloride copolymers. The number

138

Determination of Monomer Ratios in Copolymers average sequence length, which reflects monomer arrangement in the copolymer, was calculated using formulae that incorporate pure trimer and hybrid trimer peak intensities. Due to the difference in reactivity between vinyl chloride and vinylidene chloride monomers, the structure of the polymer was further investigated on the basis on the percentage of grouped monomers (i.e., number average sequence length for vinyl chloride and vinylidene chloride repeat units). The results obtained for compositional analysis achieved by this method and by 1H-NMR were in excellent agreement. In the method, 2.5 mg of sample was pyrolysed in a quartz tube, equilibrated for 5 minutes at 180 ºC, then pyrolysed at 700 °C for 20 seconds using a pyroprobe CD5190 with platinum coil. Gas chromatography was carried out using a flame ionisation or mass spectrometric detector. In further work Wang and Smith [40] and Wang and co-workers [43] and Wang and Smith [44] used Py-GC to study the composition and structure of vinylidene chloride/vinyl chloride copolymers. The composition and number average sequence length, which reflects the monomer arrangement in the polymer, were calculated using formulae that incorporate the pure trimer peak intensities and hybrid trimer peak intensities. The structure of the polymer was further investigated on the basis of the percentage of grouped monomers, i.e., the number average sequence length for vinyl chloride and vinylidene chloride repeat units. The composition and number average sequence length elucidated from the Py-GC study were compared with the product composition specification and/or the composition measured by 1H-NMR. Figure 4.9 shows the typical pyrogram of a vinylidene chloride/vinyl chloride copolymer. Identification of all four trimers was accomplished by comparing retention times with those of standard compounds, as well as identification by Py–GC/MS in EI mode. Benzene, chlorobenzene, dichlorobenzene and trichlorobenzene are four major products formed in the pyrolysis of vinylidene chloride/vinyl chloride copolymer. To make the composition calculation, the first assumption is that all trimer peak intensities generated from the Py-GC after correction for pyrolysis efficiency and detection efficiency accurately represent the triad distribution of the vinylidene chloride/vinyl chloride copolymer. If a close relationship exists between the triad distribution in the polymer chain and the production of trimers in pyrolysis, the composition and number average sequence length can be calculated on the basis of the trimer production in the pyrolysis. Results were in good agreement with those obtained by 1H-NMR (Table 4.11). Copolymers containing 11 wt% and 5 wt% vinyl chloride and 5% and 89% vinylidene chloride were successfully analysed.

139


Dichlorobenzene (DDC, CDD, DCD) 16 14 12 Trichlorobenzene (DDD)

Chlorobenzene (CCD, DCC, CDC)

10 8 6

Benzene (CCC)

4 2 0 5

10

15

35

25 30 20 Retention time (min)

40

45

Figure 4.9 Pyrogram of vinylidene chloride–vinyl chloride copolymer showing four trimer peaks of pyrolysis products. Reprinted with permission from F.C-Y. Wang and P.B. Smith, Analytical Chemistry, 1996, 68, 425. © 1996, ACS

Table 4.11 Composition results calculated from pyrolysis peak intensities compared with 1H-NMR results of five compositions of vinylidene chloride (VDC)/vinyl chloride (VC) copolymer Sample A

B

C

D

E

F

Pyrolysis wt%

VC(C) VDC(D)

11 89

12 88

14 86

17 83

50 50

95 5

1

VC(C) VDC(D)

11 89

12 88

14 86

17 83

48 52

95a 5a

H-NMR wt%

a

Weight percentage data from commercial product specification.

Reproduced with permission from F.C-Y. Wang and P.B. Smith, Analytical Chemistry, 1996, 68, 3, 425. © 1996, ACS

140

Determination of Monomer Ratios in Copolymers

4.3.4 Acrylonitrile-cis (or trans) Penta 1,3 diene Petit and Neel [45] investigated the possibility of using the method of flash Py-GC for quantitative determination of the composition of cis- or trans-1,3-pentadiene– acrylonitrile copolymers prepared by free radicals, and for the evaluation of their comonomer sequence distributions in terms of run numbers. The experiments (sample weight: 50 μg, pyrolysis time: 4 seconds) were carried out under a flow of helium at a thermolysis temperature ranging from 450 °C to 900 °C with a Curie-point pyrolyser. The Py-GC characterisation of the primary structure of copolymers was studied, between 500 °C and 800 °C through the quantitative treatment of the corresponding liberated monomers which appeared on the pyrograms. By applying the both-side boundary effect theory on the molar amounts of these degradation products, which depend upon copolymer composition and triad sequence distributions in the chain, the relative values of the monomer formation probability constants were calculated. The composition and the run number of each pyrolysed sample were determined using these parameters. The analytical data obtained by means of the procedure suggested are in very good agreement with those predicted, from reactivity ratios, by the usual theory of copolymerisation (terminal-unit model) and with the evaluations provided by 13C-NMR spectroscopy.

4.4 Pyrolysis IR Spectroscopy 4.4.1 Olefin Copolymers In addition to polyethylene and polypropylene, a wide range of olefin comonomers are produced which consist of copolymers of C2 to C8 olefins. A case in point is a copolymer of ethylene and butane-1 containing up to 10% butane-1. The preparation of calibration standards presents a difficulty in IR methods for analysing such copolymers. Physical blends of the two homopolymers, PE and polybutene-1 will not suffice because these have a different spectrum to a true copolymer with the same ethylene– butane ratio. An excellent method for preparing such standards is to copolymerise blends of ethylene and 14C-labelled butane-1 of known activity. From the activity of the copolymer determined by scintillation counting, its butane-1 content can be calculated. Standards prepared by this method are suitable for the calibration of the more rapid IR method, which involves measurements of the characteristic absorption of the ethylbranches at 769 cm–1 (13 μm). Absorbance at 769 cm-1 (13 μm) is directly proportional to the concentration of ethyl branches up to 10 per 1000 °C.

141

Introduction to Polymer Analysis Brown and co-workers [46] showed that pyrolysis of ethylene–propylene copolymers at 450 °C produces derivatives rich in unsaturated vinyl and vinylidene groups, similar to the pyrolysis of the natural rubber and styrene–butadiene rubber mixture [47], which produces vinyl groups derived from the butadiene part of the molecule and the vinylidene groups from the methyl branches of the isoprene units. This unsaturation exhibits strong absorption in the IR region. The ratio of the absorption of the vinyl groups to that of vinylidene groups varies with the mole fraction of propylene in saturated ethylene–propylene copolymers [48]. Making use of this ratio, they developed an analytical method for determining propylene in raw and vulcanised ethylene–propylene copolymers. The vinyl group absorbs at about 909 cm–1 (11.00 μm) and the vinylidene at about 889 cm–1 (12.25 μm) [48, 49]. The values of the ratio, R(×100) range from 9.977 to 0.0290, respectively, for 0 mole% to 100 mole% propylene for raw samples, and from 5.440 to 0.0431, respectively, for 10 mole% to 100 mole% propylene for vulcanised samples. The common logarithm of the ratio, R, can be represented by a linear function of the mole% of propylene in the copolymer. Table 4.12 lists the results, expressed as common logarithms of 100R, for unvulcanised samples.

Table 4.12 Log 10 (100R) for polymer pyrolysates (raw samples) Sample

Log10 (100R) at various propylene concentrations (mole%) 0

10

20

31

40

50

100

1

2.827

2.584

2.309

2.041

1.931

1.620

0.695

2

2.840

2.525

2.309

2.048

1.096

1.614

0.743

3

2.946

2.604

2.318

2.060

1.940

1.592

0.596

4

2.999

2.587

2.376

2.047

1.908

1.596

0.580

5

2.996

2.552

2.238

2.097

1.886

1.589

0.542

6

2.954

2.568

2.327

2.055

1.896

1.588

0.432

7

2.989

2.562

2.315

2.063

1.902

1.589

0.542

8

2.951

2.578

2.301

2.048

1.916

1.582

0.461

9

2.897

2.567

2.340

2.053

1.933

1.620

0.591

10

2.964

2.561

2.362

2.068

1.904

1.588

0.658

Average

2.9365

2.5687

2.3193

2.0579

1.9114

1.5979

0.5840

Reproduced with permission from J.E. Brown, M. Tryon and J. Mandel, Analytical Chemistry, 1963, 35, 13, 2172. © 1963, ACS

142

Determination of Monomer Ratios in Copolymers

References 1. G. Tosi and T. Simonazzi, Die Angewandte Makromolekulare Chemie, 1973, 32, 1, 153. 2. P.J. Cornish and M.E. Tunnicliffe, Journal of Polymer Science, 1964, C7, 187. 3. S. Davison and G.L. Taylor, British Polymer Journal, 1972, 4, 65. 4. H.V. Drushel and F.A. Iddings, Analytical Chemistry, 1963, 35, 28. 5. W. Kimmer and R. Schmolke, Plaste und Kautschuk, 1968, 15, 807. 6. G. Natta, G. Mazzanti, A. Valvassori and A. Pajaro, Chimica e l’Industria (Milan), 1957, 29, 773. 7. H.V. Drushel and F.A. Iddings, Analytical Chemistry, 1963, 35, 1, 28. 8. H.V. Drushel and F.A. Iddings in Proceedings of the 142nd ACS Meeting, Atlantic City, NJ, USA, 1962, Paper No.20. 9. P.E. Wei, Analytical Chemistry, 1961, 33, 2, 215. 10. T. Gössl, Die Makromolekulare Chemie, 1961, 42, 1. 11. P.J. Corish, R.M.B. Small and P.E. Wei, Analytical Chemistry, 1961, 33, 12, 1798. 12. J.N. Lomonte and G.A. Tirpak, Journal of Polymer Science Part A: General Papers, 1964, 2, 2, 705. 13. J.R. Paxson and J.C. Randall, Analytical Chemistry, 1978, 50, 13, 1777. 14. F. Ciampelli, G. Bucci, A. Simonazzi and A. Santambreglio, Chimica e l’Industria (Milan), 1962, 44, 489. 15. M.E. Johnson-Plaumann, H.P. Plaumann and S. Keeler, Rubber Chemistry and Technology, 1986, 59, 4, 580. 16. N.N. Partnov, S.F. Salova, M.G. Matveer and V.S. Shein, Vysokomolekulyarnye Soedineniya Seriya B, 1987, 29, 243. 17. S. Pallacini, T. Porro and J. Pavlek, American Laboratory, 1991, 23, 38. 18. R.W. Jones and J.F. McClelland, Analytical Chemistry, 1990, 62, 19, 2074.

143

Introduction to Polymer Analysis 19. S.H. Kandil and M.A. El-Gamal, Journal of Polymer Science, Polymer Chemistry Edition, 1986, 24, 11, 2765. 20. M.W. Urban, J.L. Koenig, L.B. Shih and J.R. Allaway, Applied Spectroscopy, 1987, 41, 4, 590. 21. D.G. Anderson, K.E. Isakson, D.L. Snow, D.J. Tessari and J.T. Vandebery, Analytical Chemistry, 1971, 43, 7, 894. 22. G.H.J. van Doremaele, A.L. German, N.K. de Vries and G.P.M. van der Velden, Macromolecules, 1990, 23, 19, 4206. 23. J. Helmroth, Polyvehromarium Plast, 1973, 73. 24. H.N. Cheng and M. Kakugo, Macromolecules, 1991, 24, 8, 1724. 25. X. Zhang, H. Chen, Z. Zhou, B. Huang, Z. Wang, M. Jiang and Y. Yang, Macromolecular Chemistry and Physics, 1994, 195, 3, 1063. 26. Y.V. Kissin and A.J. Brandolini, Macromolecules, 1991, 24, 9, 2632. 27. H.N. Cheng, Polymer Bulletin, 1991, 26, 3, 325. 28. D.L. Evans, J.L. Weaver, A.K. Nukherji and C.L. Beatty, Analytical Chemistry, 1978, 50, 7, 857. 29. W. Shouting and G.W. Poehlein, Journal of Applied Polymer Science, 1993, 49, 6, 991. 30. B. Subrahmanyam, S.D. Baruah, H. Rahman, J.N. Baruah and N.N. Dass, Journal of Polymer Science, Polymer Chemistry Edition, 1992, 30, 10, 2273. 31. A.S. Brar and S. Charan, Journal of Applied Polymer Science, 1994, 53, 13, 1813. 32. J.T. Blackwell, Analytical Chemistry, 1976, 48, 13, 1883. 33. E.G. Brame, Jr., and F.W. Yeager, Analytical Chemistry, 1976, 48, 4, 709. 34. C. Vu and J. Cabestany, Journal of Applied Polymer Science, 1991, 42, 11, 2857. 35. J. Van Schooten and S. Mostert, Polymer, 1963, 4, 135. 36. J. van Schooten, E.W. Duck and R. Berkenbosch, Polymer, 1961, 2, 357.

144

Determination of Monomer Ratios in Copolymers 37. E.W. Neumann and H.G. Nadeau, Analytical Chemistry, 1963, 35, 10, 1454. 38. J.C. Verdier and A. Guyot, Macromolekulare Chemie, 1974, 175, 5, 1543. 39. E.M. Barrall, R.S. Porter and J.F. Johnson, Journal of Applied Polymer Science, 1965, 9, 9, 3061. 40. F.C-Y. Wang and P.B. Smith, Analytical Chemistry, 1996, 68, 17, 3033. 41. S. Tsuge and H. Ohtani, Pyrolysis Gas Chromatography of High Polymers, Fundamentals and Data Compilation, Techno-Systems, Tokyo, Japan, 1989, p.104. 42. A. Krishen, Analytical Chemistry, 1972, 44, 3, 494. 43. F.C-Y. Wang, B.B. Gerhart and P.B. Smith, Analytical Chemistry, 1995, 67, 19, 3536. 44. F.C-Y. Wang and P.B. Smith, Analytical Chemistry, 1996, 68, 3, 425. 45. A. Petit and J. Neel, Journal of Applied Polymer Science, 1990, 41, 1-2, 267. 46. J.E. Brown, M. Tryon and J. Mandel, Analytical Chemistry, 1963, 35, 13, 2172. 47. N. Tyron, E. Horowicz and J. Mandel, Journal of Research of the National Bureau Standards, 1955, 55, 219. 48. L.H. Cross, R.B. Richards and H.A. Willis, Discussions of the Faraday Society, 1950, 9, 235. 49. D.C. Smith, Industrial Engineering & Chemistry, Analytical Edition, 1956, 48, 7, 1161.

145


146

5

Sequencing of Monomer Unit in Polymers

A very important aspect of microstructure is the sequence of monomer units in a polymer. This applies whether the polymer is based on a single monomer which is capable of polymerising in different ways, e.g., head-to-head or head-to-tail polymerisation, or whether it is based on two or more different monomers when many variants of monomer sequence are possible. Sequence distribution has an important bearing on the tacticity and other properties of polymers, as will be discussed later. Three major techniques have been used to study sequence problems in polymers, they are pyrolysis–gas chromatography, nuclear magnetic resonance (NMR) and, more recently, secondary ion mass spectrometry (SIMS). As might be expected, groups at either end of a polymer chain differ from those in the main polymer chain. Various methods are available for determining low concentrations of such end-groups in polymers.

5.1 Sequencing in Homopolymers 5.1.1 NMR Spectroscopy Inoue and co-workers [1], Zambelli and co-workers [2] and Randall [3] have shown that 13C-NMR is an informative technique for measuring stereochemical sequence distributions in polypropylene. These workers reported chemical shift sensitivities to configurational tetrad, pentad and hexad placements for this polymer. The sequence lengths of sterochemical additions in amorphous and semi-crystalline polypropylene were accurately measured using 13C-NMR [4]. This method has some limitations for addition polymers having predominantly isotactic sequences. Randall [4] described work on the application of quantitative 13C-NMR to measurements of average sequence length of like stereochemical additions in polypropylene. He describes sequence lengths of stereochemical addition in vinyl polymers in terms of the number-average lengths of like configurational placements. Under these circumstances, a pure syndiotactic polymer has a number-average sequence length of 1.0; a polymer with 50:50 meso-racemic (m-r) additions has a

147

Introduction to Polymer Analysis number-average sequence length of 2.0, and polymers with more meso than racemic additions have number-average sequence lengths >2. Amorphous and crystalline polypropylenes were examined using 13C-NMR as examples of the applicability of the average sequence length method. The results appear to be accurate for amorphous and semi-crystalline polymers, but limitations are present when this method is applied to highly stereoregular vinyl polymers containing predominantly isotactic sequences [5]. Randall measured the 13C-NMR spin lattice relaxation times of isotactic and syndiotactic sequences in amorphous polypropylene. Spin-lattice relaxation times for methyl, methylene, and methane carbons in an amorphous polypropylene were measured as a function of temperature from 46 °C to 138 °C. Carbons from isotactic sequences characteristically exhibited the longest spin relaxation times of those observed. The spin relaxation time differences increased with temperature, with the largest differences occurring for methane carbons, where a 32% difference was observed. Randall determined activation energies for the motional processes affecting spin relaxation times for isotactic and syndiotactic sequences. Essentially no dependence upon configuration was noted. High-resolution NMR spectra of isotactic and syndiotactic polypropylene have been used by Cavalli and co-workers [6] to provide conformational information. It is known that, on the surface of the heterogeneous Ziegler–Natta catalysts promoting the isotactic polymerisation of propene, active centres are also present which may give rise to the formation of significant amounts of ‘r-rich’ sequences (‘syndiotactic’ or ‘syndiotactoid’) [7]. The terms ‘isotactoid’ and syndiotactoid are used to indicate macromolecules or sequences of monomeric units with statistical distributions of configurations such that the content of meso (m) diads or of racemic (r) diads, respectively, is significantly higher than 50% (typically, in the interval 70–90%), but not equal or nearly equal to 100% (in which cases the terms ‘isotactic’ and syndiotactic can be applied). Syndiotactic polypropene was isolated for the first time as an ‘impurity’ from samples of isotactic polypropene prepared in the presence of catalyst systems such as, e.g., A- or G-TiCl3 in combination with Al(C2H5)2F or LiAlCH4 [8]. As representative examples, Busico and co-workers [9] selected two fractions (fraction A, diethyl ether-soluble; fraction B, hexane-soluble/pentane-insoluble) of a polypropene sample prepared in the presence of the catalyst system MgCl2/TiCl4TMP/Al(C2H5)3 (TMP = 2,2,6,6-tetramethylpiperidine). The methyl region of the 150 MHz Figure 5.1.

13

C-NMR spectrum of fraction B is shown in

When not assigned in the literature [10, 11], resonances were attributed on the basis of chemical shift calculations according to the G-gauche effect [12] and by comparative analysis with the 150 MHz 13C-NMR spectra of samples of isotactic

148

Sequencing of Monomer Unit in Polymers and syndiotactic polypropylene prepared with homogeneous Group IV metallocenebased catalysts.

22.0

21.5

21.0

20.5

rm rrm m

mm rr mm

rm rrm r

rrr r mr + m r r r m m

r rrr mm

m r r rm r

rm rm

m rrr r m

r r rrr m

m mr m + r m r r

mmrr rm m r

rm m mmr

mm m m m r

mmmr

rrrrrr

mmmmmm

Complete resolution was achieved for the resonances arising from the mmmm-centred and rr-centred heptads, the latter showing a fine structure reaching the nonad or even the undecad level.

20.0

PPM

Figure 5.1 150 MHz 13C-NMR spectrum of hexane soluble pentane insoluble fractions of polypropylene. Reproduced with permission from V. Busico, P. Corradini, R. De Biazo, L. Landriani and A.L. Segre, Macromolecules, 1994, 27, 16, 4521. © 1994, ACS 149

Introduction to Polymer Analysis In the first column of Table 5.1 and Table 5.2, Busico and co-workers [9] report the experimental stereosequence distributions of the two fractions A and B as evaluated from the spectral integration.

Table 5.1 Experimental stereosequence distribution for fraction A and bestfitting distributions calculated according to the two statistical models described in the text Stereosequence mmmmmm

% (exptl)

% (calc, ‘twosite’)

% (calc, ‘Coleman-Fox two-site’)

8.8

10.2

8.4

mmmmmr + rmmmmr

6.6

7.5

7.4

mmmr

11.0

11.7

12.1

rmmr

3.6

3.0

3.0

mmrr

15.1

12.8

14.1

mmrm + rmrr + rmrm

19.5

21.5

19.4

rrrrrr

8.6

8.9

8.5

rrrrrm + mrrrrm

7.4

6.3

7.8

mrrrmr

1.5

1.8

1.8

rrrrmr + mrrrmm

5.6

5.9

5.7

rrrrmm

5.1

3.0

5.1

rmrrmr

1.2

1.8

1.6

rmrrmm

3.2

2.3

2.4

mmrrmm

2.8

3.3

2.7

S = 0.76

S = 0.80

Pr = 0.81

Pr = 0.86

w = 0.69

w = 0.65 pi/s = 0.088 (ps/i = 0.17)

3

a

2 b

a

10 3 = 2.1

103 3a = 0.48

104 7b = 1.9

104 7b = 0.48

2

3 = 3(yr–yi) . 7 = 3(yr–yi) /(n – m) where n (=14) is the number of independent experimental data and m is the number of adjustable parameters. Reproduced with permission from V. Busico, P. Corradini, R. De Biazo, L. Landriani and A.L. Segre, Macromolecules, 1994, 27, 16, 4521. © 1994, ACS

150


Table 5.2 Experimental stereosequence distribution for fraction A and bestfitting distributions calculated according to the two statistical models described in the text Stereosequence mmmmmm

% (exptl)

% (calc, ‘twosite’)

% (calc, ‘Coleman-Fox two-site’)

17.9

18.1

17.6

mmmmmr + rmmmmr

8.3

8.6

8.5

mmmr

10.0

11.5

11.5

rmmr

2.4

1.8

1.7

mmrr

13.6

12.0

12.4

mmrm + rmrr + rmrm

13.6

14.3

13.1

rrrrrr

14.1

14.1

14.0

rrrrrm + mrrrrm

5.3

5.3

6.3

mrrrmr

0.7

0.9

0.9

rrrrmr + mrrrmm

4.1

4.7

4.6

rrrrmm

4.3

2.2

3.5

rmrrmr

1.0

0.9

0.8

rmrrmm

1.5

1.7

1.7

mmrrmm

3.2

3.9

3.4

S = 0.82

S = 0.84

Pr = 0.88

Pr = 0.90

w = 0.70

w = 0.67 pi/s = 0.028 (ps/i = 0.057)

3

a

a

10 3 = 1.1

103 3a = 0.67

104 7b = 1.0

104 7b = 0.67

2 b

3 = 3(yi – yi) . 7 = 3(yi – yi)/(n – m) where n (=14) is the number of independent experimental data and m is the number of adjustable parameters. Reproduced with permission from V. Busico, P. Corradini, R. De Biazo, L. Landriani and A.L. Segre, Macromolecules, 1994, 27, 16, 4521. © 1994, ACS

5.1.2 Pyrolysis Gas Chromatography (Py-GC) Before discussing the applications of this technique to microstructural studies of polymers, one must understand the principles of the technique, and the factors which

151

Introduction to Polymer Analysis affect the results obtained. These aspects are discussed first, followed by some examples of the application of Py-GC which show its usefulness in microstructural studies. A small quantity of the polymer is mounted on an inert metal support and an electrical current is passed through the support (filament method) or external heat is supplied to the support (furnace method) so as to rapidly heat up and break down (i.e., pyrolyse) the polymer into a mixture of smaller molecules which, under standard pyrolysis conditions, are characteristic of the polymer being examined. Products are swept from the pyrolysis chamber by a stream of carrier gas onto a gas chromatographic column and separated into their individual components before passing through the detector, which records their retention time (time taken, under standard conditions, to travel from pyrolysis chamber to detector) and quantity (peak height under standard conditions). This is the essence of the Py-GC technique. It is possible to then pass the separated pyrolysis products one at a time into a mass spectrometer to obtain definitive information regarding their precise identity, i.e., pyrolysis–mass spectrometry.

5.1.2.1 Polyolefins An example of the results obtainable by Py-GC is shown in Figure 5.2, which compares the pyrograms of polyethylene, polypropylene and an ethylene–propylene copolymer. To obtain these results, the sample (20 mg), in a platinum dish, was submitted to controlled pyrolysis in a stream of hydrogen as carrier gas. Pyrolysis products were then hydrogenated at 200 °C by passing through a small hydrogenation section containing 0.75% platinum on 30/50 mesh aluminium oxide. The hydrogenated pyrolysis products were then separated on squalane on a fireback column, and the separated compounds detected by a katharometer. Under the experimental conditions used in this work, only alkanes up to C9 could be detected. Major differences occur between the pyrograms of these three similar polymers. Polyethylene produces major amounts of normal C2 to C8 alkanes, and minor amounts of 2-methyl and 3-methyl compounds such as isopentane and 3-methylpentane, indicative of short chain branching on the polymer backbone. For polypropylene, branched alkanes predominate, these peaks occurring in regular patterns, e.g., 2-methyl, 3-ethyl and 2,4-dimethyl configurations. Particularly noticeable are the large peaks due to 2,4-dimethylpentane and 2,4-dimethylheptane, which are almost absent in the polyethylene pyrolysate. Minor components obtained from polypropylene are normal paraffins present in decreasing amounts up to normal hexane. This is to be contrasted with the pyrogram of polyethylene, where n-alkanes predominate. The ethylene– propylene copolymer, as might be expected, produces normal and branched alkanes;

152

Sequencing of Monomer Unit in Polymers 2,4-dimethylpentane and 2,4-dimethylheptane concentrations are lower than those seen in polypropylene.

1. 2. 3. 4. 5. 6. 7. 8.

Time

22 1x

22 1x 21 21

20

1x

9. 10. 11.

1x 1x

12. 13. 14.

18

19 1x 18

1x

1x

15. 16. 17. 18. 19. 20.

17 17 16

17 1x

1x

1x 1x 15

15

14

1x 13

13

1x

11

11

2x

2x

2x

7 6 5 4 31 2

5x 100x

1x

1x

11

1x

10 9 8

1x

13 12

1x

10 9 8

1x 1x 1x 5x

5x

20x 100x 100x

Recorder deflection Chromatogram of pyrolyzate of polyethylene

7 5 4 31 2

20x 100x

100x

20x

Recorder deflection Chromatogram of pyrolyzate of polyethylene

10 9 8 7 6 5 4 31 2

21. 22.

E Pre Isobulane n-bulane Isopetane n-pentane 2-methylpentane and/or cyclopentane 3-methylpentane n-hexane 2.4-dimethylpentane and/or methylcyclopentane 2-methythexane 3-methythexane and/or cyclohexane 1.3-dimethylcyclopentane-C or - trans n-heptane 2.5 -dimethythexane 2.4-dimethythexane and/or toluene 2-methylheptane, 4-methylheptane 3-methylheptane 1.3-dimethylcyclohexane1.4-dimethylcyclohexane-trans n-octane 2.4-dimethytheptane

1x 1x

100x

20x 5x 20x 100x 100x

1x 2x 5x

Recorder deflection Chromatogram of pyrolyzate of ethylene propylene copolymer

Figure 5.2 Gas chromatograms of pyrolysates of polyethylene, polypropylene and ethylene–propylene copolymer. Source: Author’s own files

5.1.2.2 Polyisoprene Polyisoprene (hydrogenated natural rubber) is a completely alternating ethylene propylene copolymer (i.e., does not have ethylene or propylene blocking) and is therefore an interesting substance for Py-GC studies. The surface area of the main peaks up to C13 obtained by van Schooten and Evenhuis [13, 14] indicate that the unzipping reaction which would yield equal amounts of ethylene and propylene in the hydrogenated pyrolysate takes place to some extent, but is less important than the hydrogen transfer reactions:

The large numbered peaks produced upon Py-GC reflect the many possible transfer reactions for this polymer, some of which are illustrated below. Hydrogen transfer

153

Introduction to Polymer Analysis reactions which occur from the fifth carbon atom predominate as indicated by the large butane, 3-methyl hexane and 2-methyl heptanes peaks: Hydrogen transfer also occurs from the ninth carbon atom, which is shown by the size of the 3-methyl nonane, 3,7-dimethyldecane and the 2,6-dimethylydecane peaks:

5.1.2.3 Polyvinyl Chloride (PVC) Wang and Smith [15] developed pyrolysis followed by the GC technique for determining up to four monomer units in vinylidene chloride–vinyl chloride–(C/D) copolymers. The major mechanism of producing oligomers with pyrolysis can be attributed to thermal degradation. The intensity of the various oligomer peaks in a pyrolysis gas chromatogram will reflect the monomeric sequence and polymer structure when the formation of pyrolysis products is proportional to their existence in the copolymer. Determining composition by pyrolysis usually depends on the monomer production after polymer chain scission. Vinyl chloride and vinylidene chloride are gaseous monomers and not well retained on a capillary gas chromatography column under normal conditions. Other gases that result from pyrolysis of this system, such as hydrogen chloride and butadiene, interfere with monomer detection. Composition analysis utilising the monomers of vinyl chloride/vinylidene chloride copolymer through pyrolysis is therefore a poor approach. Wang and Smith [15] therefore used trimer peak intensities to achieve the composition quantitative analysis as well as number average sequence determination for the vinylidene chloride/vinyl chloride copolymer system. The unique phenomenon in the pyrolysis of vinylidene chloride/vinyl chloride copolymer is trimer formation. Under pyrolysis conditions, the polymer will directly undergo

154

Sequencing of Monomer Unit in Polymers the thermal dehydrochlorination to form a conjugated polyene [16]. The polymer will then unzip, followed by a radical cyclisation to form benzene, chlorobenzene, dichlorobenzene and trichlorobenzene. The mechanism can be expressed as: 1. dehydrochlorination

- (CH2-CHCl) -

-(CH = CH)-

- (CH2-CCl2)-

-(CH = CCl)-

2. unzipping

Polymer chain

CCC CCD,DC, CDC

3. cyclisation CCC

benzene

CCD, DCC, CDC

chlorobenzene

DDC, CDD, DCD

dichlorobenzene

DDD

trichlorobenzene

Where C = vinyl chloride monomer D = vinylidene chloride monomer Because these chlorinated aromatics are so stable, the trimer formation pathway is the major pyrolysis pathway for the vinyl chloride/vinylidene chloride copolymer. Two major factors dominate the relationship between triad distribution and trimer production. The first is pyrolysis efficiency, which represents the probability/efficiency of breakdown of a specific triad configuration to produce the corresponding trimer. The second is detection efficiency, which results in variable flame ionisation detection (FID) responses for the trimers. These two factors cannot be separated in the vinylidene chloride/vinyl chloride copolymer composition and structure determination case. The relationship between trimer production and triad distribution can be expressed as: Experimental trimer peak intensity x Kn m triad distribution in the polymer

155

Introduction to Polymer Analysis Where Kn is the combination of pyrolysis efficiency and detection efficiency. Because the trimers CCD, DCC, and CDC all form chlorobenzene, the second assumption must be made that Kn is the same for CCD, DCC, and CDC. The same assumption must be made for dichlorobenzene with the trimers of DDC, CDD, and DCD. The triad distribution in the polymer and the trimer peak intensities from pyrolysis can be written as: Benzene peak intensity × K1 = CCC distribution in the polymer Chlorobenzene peak intensity × K2 = CCD, DCC, CDC distribution in the polymer Dichlorobenzene peak intensity × K3 = CDD, DDC, DCD distribution in the polymer Trichlorobenzene peak intensity × K4 = DDD distribution in the polymer K1, K2, K3 and K4 values can be calculated by pyrolysing four different known compositions of vinylidene chloride/vinyl chloride copolymer standards. The composition calculation from the trimers will be as follows: mol% of C = normalised/corrected benzene peak intensity + 2/3 normalised/corrected chlorobenzene peak intensity + 1/3 normalised/corrected dichlorobenzene peak intensity. mol% of D = normalised corrected benzene peak intensity + 2/3 normalised/corrected dichlorobenzene peak intensity + 1/3 normalised/corrected chlorobenzene peak intensity. The determination of number average sequence lengths for the vinyl chloride and vinylidene chloride is challenging because six of eight trimers are not resolved by PyGC. As mentioned previously, there is no way to know how much chlorobenzene peak intensity is contributed from triad CCD and DCC or from CDC. The same situation exists for the dichlorobenzene peak from the triad of DDC, CDD and DCD. They utilise the equations listed below but it is necessary to make a third assumption to obtain all six terms of triad intensities. Number average sequence length of C and D is as follows:

nC NCCC NCCD DCC NDCD

156


¥1´ ¦ µ N CCD DCC N DCD §2¶

nD

NDDD NCDD DDC NCDC ¥1´ ¦ µ NCDD NC DD NCDC §2¶

Where nC and nD are the number average sequence lengths of monomers C and D. NCCC, NCCD + OCC, NDCD, NCDC, DDC and NDDD are the experimentally derived triad molar fractions or numbers of molecules. From the formulae above, if all six triad molar fractions or numbers of molecules can be generated, then the number average sequence lengths of monomers C and D can be calculated. The third assumption results from the known reactivity difference between vinyl chloride and vinylidene chloride monomers in the copolymerisation process. In the polymer molecules formed at the beginning of polymerisation, vinyl chloride exists in the polymer chain as a single unit among many vinylidene chloride units. In the polymer molecules formed in the latter part of polymerisation, vinylidene chloride exists in the polymerisation chain as a single unit among many vinyl chloride units. Because of the relative reactivity ratio (r1, r2) of vinylidene chloride and vinyl chloride monomers, the probability of forming an alternating monomeric unit in the copolymer molecules is minimal. With this assumption, the polymer molecules have a vinylidene chloride/vinyl chloride distribution as follows: Beginning part of polymerisation: --- DDDDDDDCDDDDDD ----- DDDDCDDDDCDDDD ----- DDDCDDDCDDDCDD ----- DDCDDCDDCDDCDD ----- CCDCCDCCDCCDCC ----- CCCDCCCDCCCDCC ----- CCCCDCCCCDCCCC --Latter part of the polymerisation: --- CCCCCCCDCCCCCC --157

Introduction to Polymer Analysis The triad distribution can be expressed as follows: NCCC = normalised/corrected benzene peak intensity NCCD + NDCC = 2/3 normalised/corrected chlorobenzene peak intensity NCDC = 1/3 normalised/corrected chlorobenzene peak intensity NCDD + NDDC = 2/3 normalised/corrected dichlorobenzene peak intensity NDCD 1/3 normalised/corrected dichlorobenzene peak intensity These terms are subsequently used for the determination of number average sequence length. Table 5.3 shows a comparison of Py–GC and 1H-NMR values for the number average sequence length calculation of vinylidene chloride/vinyl chloride copolymer samples A–F. The number average sequence lengths for vinyl chloride, N(C), and vinylidene chloride, N(D), of samples C, D, and E are all within one unit difference (1.53 versus 1.40, 1.78 versus 1.50, and 3.69 versus 3.00 for vinyl chloride, 6.17 versus 5.70, 5.51 versus 5.10, and 2.40 versus 2.20 for vinylidene chloride). These values indicate that the results from Py–GC have a very good agreement with 1H-NMR results.

Table 5.3 Composition results calculated from pyrolysis peak intensities compared with the 1H-NMR results of five compositions of vinylidene chloride/ vinyl chloride copolymera Sample A

D

E

Corrected normalised peak intensity

CCC CCD DCD DDD DDC CDC

0.115 0.317 0.568 0.748 0.222 0.031

0.138 0.330 0.532 0.737 0.228 0.035

0.172 0.349 0.479 0.719 0.238 0.043

0.253 0.371 0.375 0.697 0.243 0.060

0.604 0.250 0.146 0.360 0.448 0.192

0.997 0.002 0.002 0.876 0.100 0.024 392.9

Py-GC NASL

N(C) N(D) wt% C wt% D

1.38 7.06 11.16 88.84

1.43 6.70 12.12 87.88

1.53 6.17 13.79 86.21

1.78 5.51 17.25 82.75

3.69 02.40 49.73 50.27

13.53 94.93 5.07

1

N(C) N(D) wt% C wt% D

1.40 5.70 13.66 86.34

1.50 5.50 14.95 85.05

03.00 02.10 46.77 53.23

H-NMR NASL

a

B

C

F

Where NASL represents number average sequence length

C = Vinyl chloride monomer unit D = Vinylidene chloride monomer unit Reproduced with permission from F.C.Y. Wang and P.B. Smith, Analytical Chemisty, 1996, 68, 3, 425. © 1996, ACS

158


5.1.3 SIMS SIMS spectra of polymers have been confined to the low-mass range (m/z b500) mainly due to the mass analysers used (e.g., quadropole). With the advent of hightransmission time-of-flight mass analysers coupled to sensitive detection systems (e.g., post-acceleration and single-ion counting), primary ion dosages have been reduced considerably, minimising fragmentation; allowing detection of high-mass ions (up to m/z ~ 5000). A series of aliphatic polyamides (Nylons) was studied by time-of-flight secondary ion mass spectrometry (ToF-SIMS) [17]. Cationisation of the repeat unit with Ag+ and Na+ produced high-mass ions characteristic of the type of Nylon and the repeat unit sequence in the polymer chain. Polymer fragments cationised with Ag+ and Na+, containing as many as 24 repeat units (Nylon 6) and as high as m/z ~ 3500 (Nylon 66 (A6) were detected.

5.1.3 Polystyrene Bletsos and co-workers [18] present secondary ion mass spectra of diverse polymers: including polystyrene. Spectra were obtained by a time-of-flight secondary ion mass spectrometer, equipped with a mass-selected pulsed primary ion source, and angle- and time-focusing time-of-flight analyser, and a single-ion-counting detector. Fragmentation in the low-mass range provided some structural information about the repeat unit. Fragmentation patterns were unique for polymers having different repeat units but of equal mass; distinguishing between such polymers was possible. Oligomer distributions obtained from mass spectra compared well with distributions determined by other techniques (e.g., gel permeation chromatography (GPC)) for the same polymers. Several polystyrenes (PS) having various substituent groups at different positions on the hydrocarbon backbone or the benzene ring were studied. Part of the ToF-SIMS spectrum of PS is shown in Figure 5.3 as a typical example. PS fragments cationised with Ag+, (nR + Ag)+ (R = amu), produce the most intense peaks in the spectrum above m/z = 500; the spacing between them corresponds to the repeat unit of the polymer. The pattern of fragment ion peaks within the spacing of one repeat unit is consistent throughout the spectrum and is characteristic of the repeat unit and its various substituent groups. For example, the peaks of the (NR + Ag)+ series due to fragments containing 10 repeat units (n = 10) for poly(p-tert-butylstyrene) (R = 160 amu) and poly(4-methoxystyrene) appear at m/z of 1707 and 1447. The spacing between the (RR + Ag)+ for polyCP-tert-butyl styrene is 160 amu, and between the poly(4 methoxy styrene) is 134 amu. Therefore, from the peak position and the spacing between peaks, the repeat units of polymers can be determined.

159


counts/chammel

Poly (A-methylstyrene) (P(A-MS) and poly(4-methyl-styrene) (P(4-MS)) were studied to establish the effect of location of a substituent group on the ToF-SIMS spectrum. Specifically, the effect of substituting a methyl group on a phenyl group and on the chain backbone was determined. The repeat units of (P(A-MS) and (P(4-MS)) have equal masses, R = 118 amu; the most prominent peaks, due to Ag+ cationised fragments (nR + Ag)+, for (P(A-MS)) and (P(4-MS) appear to have exactly the same m/z values. Therefore, it is not possible to distinguish between (P(A-MS)) and (P(4-MS)) by the positions and the spacings of these peaks. The (nR + Ag)+ are surrounded by a series of peaks of varying intensity space at ±n'$m, where n' = 1, 2, 3, and 4 and $m = 14–16 mass units. Data are shown for the two polymers in Table 5.4. The positions of the smaller peaks and their relative intensities are different in the two spectra, permitting one to distinguish readily between (P(4-MS)) and (P(A-MS)).

×103 POLYSTYRENE

1.5

(−CH2−CH)n

(10R + Ag)+

(11R + Ag)+ (12R + Ag)+ (13R + Ag)+

1.0

+

(14R + Ag)+ (15R + Ag) (16R + Ag)+ (17R + Ag)+ (18R + Ag)+

1.5

0.0 1200

1300

1400

1500

1600

1700

1800

1900 2000 mass [am ]

Figure 5.3 Part of ToF-SIMS spectrum of polystyrene cationised with Ag+ (nR + Ag+) showing the most intense peaks in the spectrum where m/z = 500. Reproduced with permission from I.V. Bletsos, D.M. Hercules, D. Van Leyen and A. Benninghoven, Macromolecules, 1987, 20, 2, 407. © 1987, ACS

The differences in fragmentation are due to the different positions of the –CH3 substituent group in the two polystyrenes. It was generally observed that substituent groups at different locations in the backbone or the benzene ring produce different fragmentation patterns for the polystyrenes. Bond cleavage seems to occur statistically, and the chemical stability of the fragments produced is reflected by the peak intensities.

160

Sequencing of Monomer Unit in Polymers Mass spectrometry is also a useful technique for the determination of molecular weight distributions of low molecular weight polymers, and can provide information about the structure of the repeat unit and the number of repeat units for individual oligomers. Classical techniques used for determining molecular weights (e.g., GPC, vapour pressure osmometry, light scattering, NMR) measure average properties of an oligomer mixture and do not yield information on different types of oligomers present.

Table 5.4 Relative Intensities of Cluster Peaks for P(A-MS) and P(4-MS) Appearing at ± n'$m of the (nR + Ag)+ Series -4$m

-3$m

-2$m

-$m

(nR + Ag)+

+$m

+2$m

+3$m

+4$m

P(A-MS)

34

91

87

80

100

19

16

13

13

P(4-MS)

50

100

91

67

51

48

43

Polymer

Reprinted by permission from I.V. Bletsos, D.M. Hercules, D. Van Leyen and A. Benninghoven, Macromolecules, 1987, 20, 2, 407. © 1987, ACS

Molecular weight distributions obtained from ToF-SIMS spectra were evaluated by Bletsos and co-workers [18] using polymer standards having known molecular weight distributions. Below m/z 3500, the most intense peaks are due to Ag+ cationised polymer fragments corresponding to the series (nR + Ag)+. In the range m/z 3500–7000, intact oligomers are detected, giving rise to the (nR + C4H9 + H + Ag)+ series. These peaks are the most intense in this range, and their intensity distribution reflects the number average molecular weight distribution of the PS standard. Loss of a terminal – C4H9 group results in the (nR + Ag)+ series, having lower intensity than (nR + C4H9 + H + Ag)+. The higher intensity of the (nR + C4H9 + H + Ag)+ series indicates that even if some fragmentation occurs, the dominant processes in the m/z 3500–7000 range is desorption of intact polymer molecules cationised with Ag+. The value Mn = 4550 was calculated from peak intensities without isotopic abundance corrections. The calculated Mn value = 4550 is within 8% of Mn = 4964 determined by GPC.

5.1.4 Ozonisation Technique Ozonisation is an extremely useful technique for the elucidation of sequencing in unsaturated homopolymers. Use of the technique is illustrated below by a discussion of results that have been obtained by applying the technique to polybutadiene and polyisoprene. 161


5.1.4.1 Polybutadiene Polybutadiene contains three isomeric units: cis 1,4, trans 1,4 and trans 1,2. cis 1,4:

trans 1,4:

trans 1,2:

The variety of sequence distributions of these isomeric units and the tacticities of 1,2 units are governed by the initiator and polymerisation solvent. The control of these structural factors is significantly important for getting a good performance of rubber properties. Several approaches have been used for the determination of the average sequence length and tacticity of 1,2 units by spectroscopic or chromatographic methods. 13C-NMR studies on hydrogenated polybutadiene showed that syndiotactic 1,2 sequences were allowed to crystallise with longer than 3,7 racemic succession of 1,2 units [19]. GC measurements of ozonolysis products or metathesis products from polybutadiene were applied to the analysis of some diad and triad sequences of 1,2 units [20, 21]. The structural information thus obtained was restricted in principle only to short sequences, i.e., diad, triad, and so on. If the alignment of 1,2 units is characterised as a distribution from short to long sequences including the tacticity, it will provide definitive evidence on the relationship between the microstructure and physical properties, as well as that of the polymerisation conditions and microstructure. With this in mind, Tanaka and co-workers [22] proposed a new method for the characterisation of the sequence distribution of styrene (STY) units in styrene– butadiene copolymers by a combination of selective ozonolysis of the double bonds

162

Sequencing of Monomer Unit in Polymers in butadiene units and GPC measurements of the resulting products. His method is based upon high-resolution GPC analysis of the alcohols corresponding to styrene sequences obtained by scission of all the carbon–carbon double bonds of butadiene units. The ozonolysis–GPC method has been proven to be a very powerful tool to characterise the sequence distribution of styrene units and the tacticity in random, partially blocked, and triblock styrene–butadiene copolymers [23–27]. In this study, a new analytical method of the sequence distribution of 1,2 units in polybutadiene was investigated on the basis of the ozonolysis–GPC method. In this method, ozonisation was carried out by blowing an equimolar amount of ozonated oxygen (1.3%) to carbon–carbon double bonds into a 0.4% w/v chloroform solution of the polybutadiene at 30 °C. Reductive degradation of the resulting ozonide was done by addition of a small amount of water, after the ozonide had reacted with 4 mol of lithium aluminium hydride in ethyl ether. After reductive degradation neutralisation was carried out by adding 1.5 mol of trifluoroacetic acid (TFA) into the resulting LiOH and Al(OH)3. The resulting product was distilled off at atmospheric pressure followed by distillation at reduced pressure. Trifluoroacetates for GPC were prepared by allowing the reductive degraded products suspended in dry chloroform to react with 5 mol of trifluoroacetic anhydride (TFAA) in the presence of a catalyst mixture consisting of 1 mol% 4-(dimethylamino)pyridine and an equimolar amount of triethylamine based on hydroxyl groups at 38 °C for 14 hours. Upon the completion of the reaction, the precipitate was filtered. Trifluoroacetates were obtained after chloroform had been removed at atmospheric pressure.

163

Introduction to Polymer Analysis The following ozonisation reactions occur at 1,2- and 1,4-butadiene units:

Polyols corresponding to 1,4–(1,2)n = 1,4, n = 0–3 and so on, sequences are the products obtained by reductive degradation of polybutadiene ozonide with lithium aluminium hydride (LiAlH4). Polyols were converted into chloroform-soluble trifluoroacetates via esterification with TFAA. Figure 5.4 shows a high-resolution GPC curve of the trifluoroacetates obtained from the ozonolysis products of polybutadiene. Accordingly, a model compound corresponding to the 1,4-1,2-1,4 sequence was prepared by ozonolysis of 4-vinyl-1-cyclohexene, followed by esterification with TFAA. The GPC elution volume of this model compound was found to be 171 ml. Therefore, the corresponding peak observed in Figure 5.5 can be assigned to the 1,41,2-1,4 sequence. The peak appearing at the elution volume 187 ml is assigned to the 1,4-1,4 sequence because it has the same elution volume as trifluoroacetate derived from 1,4-butanediol as a model compound corresponding to the 1,4-1,4 sequence.

164

Sequencing of Monomer Unit in Polymers The other peaks that appeared in order of decreasing elution volume in Figure 5.4 are presumed to be n = 2–9. A plot of log Mw of n = 0-9 versus elution volume gave a straight line, showing that the structural assignment is valid.

n=0 n=5

(× 16) n=1

n=6 n=7 n=2

n=8 n=9

n=3 n=4 n=5

120

130

140

150 160 170 180 Elution volume (ml)

190

200

Figure 5.4 High-resolution GPC curve of the trifluoroacetates formed by reaction of trifluoroacetates of products of fractions n = 1 to n = 9. Reproduced with permission from Y. Tanaka, S. Kawahara, T. Ikeda and H. Tamai, Macromolecules, 1993, 26, 19, 5253. © 1993, ACS Parts a and b of Figure 5.5 show the 13C-NMR spectra of the compounds corresponding to n = 1 and 2, respectively. The former was prepared by ozonolysis of 4-vinyl-1cyclohexene, and the latter obtained by GPC fractionation of the ozonolysis product of polybutadiene. It is clear that all the hydroxyl groups have been esterfied with TFAA. The trifluoroacetates in a chloroform solution were found to have a maximum absorption wavelength of 270 nm as measured by a multichannel UV-vis detector. Therefore, the above GPC measurements can also be carried out with a UV detector. The peak of the 1,4-1,4 sequence was not detected because it overlapped with an intense impurity peak. A small shoulder observed at the peak of 164 ml is presumed to arise from unreacted hydroxyl groups that partially remained in the acetate derivatives. Therefore, the reactivity of acetic anhydride toward polyols is lower than that of TFAA. The GPC chromatogram of acetate derivatives of the ozonolysis products from polybutadiene showed a poorer separation than that of TFAA derivatives due to the

165

Introduction to Polymer Analysis smaller molecular weight of acetate groups. In contrast with trifluoroacetates, acetate derivatives in a chloroform solution were found to have no maximum absorption in the UV region as shown by a multichannel detector. The findings clearly indicate that the oxonolysis-GPC method can also be applied to the analysis of the sequence distribution of 1,2 units in polybutadiene by successive derivation of polyols to trifluoroacetates. It is remarkable that the peak corresponding to the 1,4-1,4 sequence is clearly observed in this measurement in addition to 1,4(1,2)n-1,4 sequences. On the basis of the relative intensity of each peak, quantitative measurement of the sequence distribution can be made by compensation with an appropriate correction factor for the refractive index or UV absorptivity of each fraction. −CH2O−TFA (a)

150

100

50

0

Chemical shift ( ppm from Me4Si ) −CH2O−TFA (b)

150

100

50

0

Chemical shift ( ppm from Me4Si )

Figure 5.5 13C-NMR spectrum of compounds corresponding to n = 1 (curve a) and n = 2 (curve b), eluted GPC fractions as illustrated in Figure 5.4. Reproduced with permission from Y. Tanaka, S. Kawahara, T. Ikeda and H. Tamai, Macromolecules, 1993, 26, 19, 5253. © 1993, ACS 166

Sequencing of Monomer Unit in Polymers The oxidation of double bonds in polymers in a non-aqueous solvent leads to the formation of ozonides which, when acted upon by water, are hydrolysed to carbonyl compounds: O

1

1

CH CH R + O3 = R- CH

R

O R

O CH R O

O 1

1

CHR +H2O = RCHO + R CHO + H2O2

CH O

Triphenyl phosphine is frequently used to assist this reaction. When applied to complex unsaturated polymers, this reaction has great potential for the elucidation of the microstructure of the unsaturation. Examination of the reaction products, for example by conversion of the carbonyl compounds to carboxylates then esters followed by gas chromatography, enables identifications of these products to be made. An example of the value of the application of this technique to a polymer structural problem is the distinction between polybutadiene made up of consecutive 1,4-1,4 butadiene sequences I, and polybutadiene made up of alternating 1,4 and 1,2 butadiene sequences II, i.e., 1,4-1,2-1,4: 1,4

I ~ CH2

CH2

II ~ CH2

CH2

1,4

CH

CH

CH2

CH

CH

CH2

1,2

CH2 CH

CH CH2

CH 1,4

CH2

CH

CH

CH CH2-

Upon ozonolysis, followed by hydrolysis, these produce succinaldehyde (CHO–CH2– CH2CHO) in the case of 1,4-1,4 sequences, and produce formyl 1,6-hexane-dial and formaldehyde in the case of 1,4-1,2-1,4 sequences:

Analysis of the reaction product for concentrations of succinaldehyde and 3-formyl 1,6-hexane dial shows whether the polymer is 1,4-1,4 or 1,4-1,2-1,4, or whether it contains both types of sequence. The 3-formyl-1,6-hexane dial content is directly proportional to the 1,2 (vinyl) content of polymers containing 1,4-1,2-1,4 butadiene

167

Introduction to Polymer Analysis sequences. Table 5.5 shows the results obtained in the ozonisation of polymers having 98% cis-1,4 structure, 98% trans-1,4 structure, and a series of polymers containing, from 11% to 75%, the 1,2 structure. The final products obtained from these polymers were succinaldehyde, 3-formyl-1,6 hexane-dial, and 4-octene-1,8dial. Model compounds were ozonised and the products compared with those from the polymers. A smooth relationship was obtained between the 1,2 content of polybutadiene as measured by infrared (IR) and NMR spectroscopy, and the amount of 3-formyl-1,3 hexane dial obtained on ozonisation (Figure 5.6).

Fraction 3-formyl-1.6-hexanedial

The amount of 1,4-1,2-1,4 sequences in polybutadienes can be estimated from the amounts of the different ozonolysis products (Table 5.5) if one considers the amount of 1,4 structure not detected. Because the ozonolysis technique cleaves the centre of a butadiene monomer unit, one-half of a 1,4 unit remains attached to each end of a block of 1,2 units after ozonolysis; these structures do not elute from the gas chromatographic column. Using random copolymer theory, the maximum amounts of these undetected 1,4 structures can then be calculated. Tanaka and coworkers [28] also discussed the determination by ozonolysis of 1,2 butadiene units in polybutadiene.

0.70 0.60 0.50 0.40 0.30 0.20 0.10 0

10

20

30

40 50 Vinyl (%)

60

70

80

Figure 5.6 Relationship between 1,2 content of butadiene and its amount of 1,6 hexane diol in the ozonisation products. Source: Author’s own files

5.1.4.2 Polyisoprene Ozonolysis has been applied to sequencing studies on polyisoprene. The particular sample used in this study had nearly equal amounts of 1,4 and 3,4 structures which, upon ozonolysis and hydrolysis, yield large amounts of laevulinaldehyde (CHO-

168

Sequencing of Monomer Unit in Polymers CH2CH2COOH3), succinaldehyde (CHOCH2CH2CHO) and 2,5 hexane dione (CH3CO-CH2CH2COCH3).

Table 5.5 Microozonolysis of polybutadiene Sample

1,4 Vinyl (cis + (1,2), trans) (%)

Area from GC (%) Succinaldehyde

3-formyl 1,6hexanedial

4-octene 1,8-dial

1, 2 units occurring in 1,4-1, 2-1,4 sequences

1

98.0

2

50

1

49

0.5

2

89.1

10.9

30

10

60

5

3

89.0

11.0

43

7

50

3

4

81.0

19.0

34

14

52

6

5

76.2

23.8

36

25

39

11

6

71.8

28.2

33

27

40

11

7

69.7

30.3

48

26

26

10

8

67.7

32.3

36

26

38

9

9

64.2

35.8

38

31

31

10

10

62.8

37.2

45

27

28

8

11

50.5

49.5

26

41

33

12

12

56.0

44.0

30

39

31

11

13

26.0

74.0

33

64

3

5


Isoprene monomer:

The 1,4 structures of polyisoprene can exist in three structural forms: Head-to-tail 1,4:

169

Introduction to Polymer Analysis Tail-to-tail 1,4:

Head-to-head 1,4:

Each of these configurations upon ozonolysis/hydrolysis produces a different oxygenated product: Head-to-tail: 1,4 laevulinaldehyde CHO-CH2OH2COCH3 Head-to-tail: 1,4 succinaldehyde CHOCH2CH2CHO Head-to-head: 1,4,2,5 hexane dione CH3COCH2CH2COCH3 These ozonolysis products indicate the presence in polyisoprene of large amounts of 1,4 and 3,4 structures. Boochathum and co-workers [29] applied ozonisation-GPC to a study of the structure of solution-grown trans 1,4-polyisoprene (TPI) crystals. Crystallisation of synthetic trans-1,4-polyisoprene at –20 °C in hexane and amylacetate solutions gave chain-folded A-type crystals with 53% and 57% crystallinity, respectively. Selective ozonolysis degradation of the isoprene units in the surface folds associated with high-resolution GPC measurement was used to determine the crystalline stem length, stem length distribution and fold surface structure of TPI crystals. The oligomer fraction, obtained by ozonolysis of TPI crystals grown in hexane solution followed by reduction with lithium aluminium hydride was found by 1H- and 13CNMR measurements to comprise a series of homologues of the trans 1,4-isoprenoid compounds of the following structure:

170

Sequencing of Monomer Unit in Polymers The whole products obtained were composed of high molecular weight fractions (assumed to come from the piled-up lamellae) and also oligomer parts of single traverses and double traverses (assumed to come from monolayers and bilayers, respectively). The purified oligomer fractions of single traverses were subjected to high-resolution GPC to determine the molecular weight of each fraction via the standard calibration curve. As ozone uptake increased, the isoprene units of the main fraction of oligomer products were found to decrease from 13, reaching a constant value at 11 for hexane and 12 for amylacetate. The polydipersity of the oligomer fractions was 1.01–1.02. These findings suggested that the chain fold structure may be tight folding with slightly irregular folds, with a stem length of 11 monomers and a dispersion of 2.3 for lamellae grown from hexane, and 12 monomers with dispersion of 2.3 for those grown from amylacetate.

5.2 Sequencing in Copolymers 5.2.1 IR Spectroscopy Earlier work on the application of this technique to sequencing studies on copolymers was fairly inconclusive [30–48]. These measurements were difficult and complex.

5.2.2.1 Styrene–methacrylonitrile More recently, Dong and Hill [49] used FT infrared (FT-IR) spectroscopy to study copolymer composition and monomer sequence distribution in styrenemethacrylonitrile copolymers. They determined the dependence of the frequencies of the individual spectral peaks on the copolymer composition, in particular, the vibration frequencies for the nitrile group is discussed. Correlations were established to relate changes in the peak positions to changes in the copolymer composition and monomer sequence distribution. Vibration band frequencies for blends of poly(methacrylonitrile) and polystyrene were examined to compare the effects of inter- and intra-chain interactions in these bands. An important characteristic of polymethacrylonitrile and its copolymers is strongly polar nitrile groups. Nitrile groups can interact with their surroundings in various ways. These different interactions between the nitrile groups and their surroundings may lead to a change in the stretching frequency of the CN bond. Therefore, the stretching frequency of the CN bond may provide information about the microstructure of styrene–methacrylonitrile copolymers, providing next neighbour effects are dominant. 171


Absorbance

The peak located at –4.48 μm, which is the CN bond stretching band for a styrene– methacrylonitrile copolymer with YM = 0.189, shifts to higher frequency with increasing methacrylonitrile content in the styrene–methacrylonitrile copolymers (Figure 5.7).

2270

2260

2250

2240

2230

2220

2210

2200

Wavenumber (cm−1)

Figure 5.7 Dependence of CN stretching bond at 2229.7 cm–1 on methacrylonitrile content of styrene–methacrylonitrile copolymers. Reproduced with permission from L. Dong and D.J.T. Hill, Polymer Bulletin, 1995, 34, 323. © 1995, Springer

The high dipole moment (3.9 Debye) for the nitrile group can give rise to a strong attraction or a strong repulsion (according to orientation) with similar groups or other substituents in a copolymer that has a high dipole moment [50]. The intra- and inter-molecular forces in a polymethacrylonitrile polymer chain result predominantly from these types of dipolar interactions [51, 52]. The adjacent nitrile groups in polymethacrylonitrile repel one another and force the polymer chain to adopt a conformation in which these repulsive forces are minimised. Nitrile groups on adjacent chains may be involved in attractive interactions. The incorporation of styrene units into a polymethacrylonitrile chain reduces the extent of the repulsive interactions of neighbouring nitrile groups, and enhances the mobility of the polymer chain segments. This allows more attractive interactions to occur between the polar nitrile groups, as well as other groups in the copolymers. At low concentrations of nitrile groups in a styrene–methacrylonitrile copolymer, the frequency of the CN resonance lies in the range 4.484–4.482 μm, depending on

172

Sequencing of Monomer Unit in Polymers the nature of the matrix of the polymer (e.g., solution or solid states). Thus, as the CN content of the polymer increases, the probability for the occurrence of adjacent methacrylonitrile–methacrylonitrite diad sequences rises, and hence the extent of repulsion between these neighbouring groups also rises. As the concentration of methacrylonitrile units in the copolymer increases, so does the vibrational frequency of the CN bond. The increasing vibrational frequency with increasing methacrylonitrile content in the copolymers is consistent with an apparently higher force constant for the CN bond. This can be rationalised in terms of the repulsive forces which exist between the carbon and nitrogen atoms of neighbouring nitrile residues along the polymer chain, and which restrict the vibration of the two atoms in each of the nitrile groups. Figure 5.8 shows the relationship between the CN bond stretching frequency and methacrylonitrile content in the copolymers for the solution samples and solid state samples in KBr discs.

CN peak frequency

2237

2235

2233

2231

2229 0.0

0.2

0.4

0.6

0.8

1.0

MAN content in copolymers (mol fraction)

Figure 5.8 Relationship between CN bond stretching frequences and methacrylonitrile content of styrene–methacrylonitrile copolymers. Reproduced with permission from L. Dong and D.J.T. Hill, Polymer Bulletin, 1995, 34, 323. © 1995, Springer

From Figure 5.8, the samples determined in dichloromethane solution have a higher stretching frequency for the CN bond than the corresponding solid state samples

173

Introduction to Polymer Analysis determined in the KBr disc form. This may be attributed to the nature of the interaction between the copolymer chains and the dichloromethane solvent. The polar dichloromethane, which has a dielectric constant of 9.7, will interact strongly with the polar CN bonds in the copolymers, leading to an increase in the polarisation of these bonds, and thus to a shift on their stretching frequency towards higher values. This study does not provide a method of determining the styrene and methacrylonitrile contents of these copolymers, but it does provide important structural information regarding sequence distributions.

5.2.2 NMR Spectroscopy 5.2.2.1 Styrene acrylate and styrene acrylic acid NMR techniques have been widely used to study copolymer composition and monomer sequence distribution in styrene–acrylonitrile and styrene acrylic acid. A question posed by Wong and Poehlein [53] regarding the monomer system of styrene–acrylic acid is whether sequence distribution information can be determined in NMR spectra. In this work, they examined 1H- and 13C-NMR spectra of S-AA copolymers to determine the resonances sensitive to the copolymer microstructure. The compositions of the copolymers were measured by NMR and the reactivity ratios calculated. The triad sequences were assigned by experiments and the Alfrey–Mayo (AM) statistics–kinetics model. In a typical 400 MHz 1H-NMR spectra of low-conversion S-AA bulk copolymer dissolved in deuterated dimethyl sulfoxide (DMSO-D6) and in DMSO-D6-CDCl3 at 50 °C, the chemical shift of different protons, respectively, are: 1–2 ppm for all methane protons and the methylene proton of the styrene units in the copolymer chain; 2.3 ppm for the methylene proton of the acrylic acid unit in the copolymer chain; 3.3 ppm for water in DMSO-D6, 5.2–6.4 ppm multi-peaks for the end groups of the copolymer chains that may be double bonds and isobutylnitrile; 6.4–7.4 ppm for aromatic protons; and 12.1 for the proton of the carboxyl group on the copolymer chain. Resonance behaviour of the carboxyl proton is heavily affected by the solvent used. A clear single resonance peak appears in the 1H-NMR spectra if DMSO-D6 is used and the baseline is straight. This sharp peak disappears if a DMSO-D6-CDCl3 solvent mixture is used. Such behaviour may result from dissociation of the carboxyl proton ion (or carboxylate ion pair) or from the formation of hydrogen bonds with the chlorides in chloroform. The 100 MHz 13C-NMR spectrum of S-AA bulk copolymer in DMSO-D6 at 90 °C

174

Sequencing of Monomer Unit in Polymers shows a resonance peak at 44 ppm due to the methylene carbon of the acrylic acid unit in the copolymer chain. All other carbons in the copolymer chain in the 36–42 ppm range were overlapped by DMSO. The C2-C5 carbons of phenyl have peaks at 125–130 ppm, the C1 carbon of phenyl at 143–144.5 ppm (tri-peaks), and the carboxyl carbon at 175–177.5 ppm (tri-peaks). Two tri-peaks caused by carboxyl carbon and C1 in phenyl may contain the information on the monomer sequence distribution. This will be discussed later. In addition to determining monomer composition in the range 30–85% styrene in copolymer, this method has been used in sequence distribution studies. With emulsion copolymers, the mole fractions of styrene triplets (FSSS) is much smaller and the mole fraction of acrylate–styrene–acrylate units (FASA) is much higher than for the corresponding FSSS and FASA values for bulk copolymers. Wong and Poehlein [53] concluded that the sequence distribution of styrene–acrylic copolymers can be measured by 13C-NMR of the carboxyl carbon and the C1 carbon of phenyl in S-AA copolymers. Low-conversion copolymer composition data obtained by NMR at different initial monomer ratios were used with the Kelen–Todos plot method to determine the reactivity ratios of ra = 0.13 and rs = 0.38. The resonance peaks split by triads were assigned and confirmed by comparing experimental triad values with those calculated from the Alfrey–Mayo statistics-kinetics model.

5.2.2.2 Propylene-1-butene Several peaks arising from different pentad and hexad comonomer sequences have been observed n the 13C-NMR spectrum of stereoregular 1-butene-propylene copolymers. The paper by Aoki and co-workers [54] demonstrated that the analytical method based on the two-dimensional (2D)-INADEQUATE spectrum and the chemical shift calculation via the G-effect is very powerful for the assignment of 13C-NMR spectra of higher A-olefin copolymers. A stereoregular 1-butene-propylene copolymer is a suitable example because reliable assignments have been proposed by a reaction probability model [55]. Aoki and co-workers [54] confirmed previous assignments of triad and tetrad sequences in this copolymer. Referring to confirmed assignments, chemical shift differences among comonomers sequences longer than pentad were predicted by the chemical shift calculation (G-effect method) based on the G-effect of the 13C chemical shift and Mark’s rotational isomeric state model modified by considering the sidechain conformation in a 1-butene unit. Assignments provided in this study agree well with Cheng’s assignments [56] by a reaction probability model. The conformational probability of the side chain in a 1-butene unit was evaluated through the chemical

175

Introduction to Polymer Analysis shift calculation. Figure 5.9 shows the 13C-NMR spectrum of a stereoregular 1-butene-propylene copolymer. On the basis of previous assignments [55], complicated peaks arising from different comonomers sequences longer than pentad are observed in the resonance regions of the methyl carbon of propylene (A), the central methylene carbon of a PP diad (B), and the side-chain methylene carbons of 1-butene (C) among propylene units. In the region of the methyl carbon in the propylene unit (21.4 ppm to 22.0 ppm), the side-chain methylene carbon in the 1-butene unit (27.5 ppm to 28.5 ppm), and the central methylene carbon of the PP diad (46.5 ppm to 47.5 ppm), the peaks arising from different comonomers sequences longer than pentad are observed. To provide assignments of these peaks, chemical shift differences among pentad and hexad comonomers sequences were calculated by the G-effect method. Table 5.6 shows the calculated chemical shift differences in the resonance regions of methyl and methylene carbons in 1-butene-propylene copolymer: CH2

C

i-I C

i

CH2

CH2 C

C

CH3 H

H j

C CH3

CH2

Planar zig-zag conformation of 1-butene-propylene copolymer methane resonance regions were excluded because of their low spectral resolution. Aoki and co-workers [54] demonstrated that spectral analysis based on the 2DINADEQUATE spectrum and the 13C chemical shift calculation via the G-effect is very useful for 13C-NMR chemical shift assignments of higher A-olefin copolymers. The successful result of this spectral analysis for a stereoregular 1-butene-propylene copolymer confirms the reliability of this method. Conformational states of the side chain in the 1-butene unit are evaluated through chemical shift calculation by considering the side-chain conformation. Therefore, this method is applicable to the analysis of the 13C-NMR spectrum of the side-chain conformation in various olefin homo- and copolymers.

176


(A)

(B)

(C)

45

40

35 a

30 25 20 from Tetramethylsilane

15

10 ppma

Figure 5.9 13C-NMR spectrum of stereoregular propylene-1-butene copolymer, A: Methyl carbon of propylene B: Central methylene carbon of a polypropylene diad C: Side chain methylene carbon of 1-butene among propylene units. Reproduced with permission from A. Aoki, T. Hayashi and T. Asakura, Macromolecules, 1992, 25, 1, 155. © 1992, ACS

5.2.2.3 Vinylidene chloride–methacrylonitrile and vinylidene–cyanovinyl acetate copolymers Montheard and co-workers [57] studied the structure of these copolymers by 13CNMR spectroscopy. They showed that macromolecules prepared from equimolar amounts of the monomers are mostly alternating, with only a small proportion of homo sequences of methacrylonitrile or cyanovinyl acetate. The nitrile groups of p-vinyl chloride and of p-vinyl acetate give three peaks which can be referred to the three configuration of triads denoted in m, mr and rr (m = meso, r = racemic).

5.2.2.4 Acrylonitrile–butyl acrylate copolymer Brar and Sunita [58] described a method for the analysis of acrylonitrile–butyl acrylate (A/B) copolymers of different monomer compositions. Copolymer compositions were determined by elemental analyses and comonomers reactivity ratios were determined using a non-linear least squares errors-in-variables model. Terminal and penultimate reactivity ratios were calculated using the observed distribution determined from 13C(1H)NMR spectra. The triad sequence distribution was used to calculate diad concentrations, conditional probability parameters, number-average sequence lengths and block character of the copolymers. The observed triad sequence concentrations determined from 13C(1H)-NMR spectra agreed well with those calculated from reactivity ratios.

177


Table 5.6 Calculated 13C NMR Chemical Shift Differences in the Resonance Regions of Methyl and Methylene Carbons of a 1-Butene-Propylene Copolymer Carbona

a

Comonomer sequencesa

Chemical shift differences, b ppm

BBPPBB

0.000

BBPPBP PBPPBP BBPPPB BBPPPP PBPPPB PBPPPP BPPPPB BPPPPP PPPPPP BBPB PBPB BBPP PBPP BBBB BBBP PBBP PPBPP BPBPP BPBPB PBBPP BBBPP PBBPB BBBPB PBBBP BBBBP BBBBB PPPPP BPPPP BPPPB PBPPP PBPPB BBPPP BBPPB PBPBP BBPBP BBPBB PBP BBP BBB

–0.048 –0.096 –0.185 –0.222 –0.233 –0.273 –0.370 –0.410 –0.451 0.000 –0.087 –0.183 –0.270 0.000 –0.096 –0.203 0.000 –0.032 –0.064 –0.133 –0.170 –0.170 –0.201 –0.265 –0.313 –0.350 0.000 –0.032 –0.074 –0.154 –0.191 –0.196 –0.239 –0.318 –0.360 –0.403 0.000 –0.026 –0.052

C, P, and B denote the carbon atom, the propylene unit, and the 1-butene unit, respectively. Chemical shift differences are expressed by ppm relative to those of the peaks appearing at the lowest field, set to be 0.000 ppm Reprinted by permission of A. Aoki, T. Hayashi and T. Asakura, Macromolecules, 1992, 25, 1, 155. © 1992, ACS

b

178

Sequencing of Monomer Unit in Polymers The 13C(1H)-NMR spectrum of the A/B copolymer (A = 55.0 mol%) recorded in a mixture of CDCl3 and DMSO-D6 at room temperature is shown in Figure 5.10. The various resonance signals have been assigned by comparing the copolymer spectrum with NMR spectra of homopolymers. For polyacrylonitrile, –CH2 and –CH carbons appeared at around D = 33.4 ppm and 27.8–28.1 ppm, respectively. The nitrile (–CN) carbon in polyacrylonitrile appeared as a multiplet in the region D = 118.2–120.8 ppm using high-magnetic-field NMR. For polybutylacrylate (PBA), resonance signals around D = 64.2 and 174.2 ppm can be assigned to –OCH2 and >C = O carbons, respectively.

CN

CN

CN

CN

CN

OAc

CN

OAc

CN

OAc

OAc mm

OAc

OAc

OAc mr

CN

OAc

CN rr

OAc

Figure 5.10 13C(1H)-NMR spectrum of acrylonitrile-n-butyl acrylate copolymer in CDCl3-DMSO-D6 mixture. Reproduced with permission from A.S. Brar and A. Sunita, Polymer, 1993, 34, 3391. © 1993, Elsevier

The –ACH and –BCH2 in polybutyl acrylate appeared at around D = 41.3 and 34.4–36.3 ppm. The sharp resonance singlets around D = 30.5, 18.9 and 13.5 ppm can be attributed to C2, C3 and (–CH3) carbons, respectively. In the A/B copolymer, resonance signals at around D = 13.4, 18.7 and 30.1 and 64.4 ppm can be assigned to (–CH3)B (C3)B , (C2)B and (O–CH2)B carbons of the butyl acrylate monomer. Resonance signals around D = 26.8–28.2 and 33.4–35.1 ppm can be attributed to (–ACH)A and (–CN2)A carbons, respectively, but could not be used for the sequence analysis because of poor resolution. The (–CH2)B and (–CH)B carbons overlapped with the solvent DMSO-D6 signals (D = 38.8–42.0 ppm) and, therefore, could not be used for the analysis of B-centred sequences. The carbonyl carbon in the A/B copolymer appeared as a multiplet around D = 172.6–174.2 ppm, indicating that the splitting of the >C = O signals is due to its sensitivity towards the compositional sequences. The nitrile carbon of the A unit appeared as a well-resolved multiplet

179

Introduction to Polymer Analysis around D = 119.2–121.2 ppm, showing its sensitivity towards different monomer placements. For A/B copolymer, a shift occurs in the position of various functional groups of A and B units compared with that in homopolymers; this is due to the change in the nature of adjacent monomeric units in the copolymer, which changes the chemical shifts of A- and B-centred triads. The carbonyl carbon (>C = O) and nitrile carbon (–CN ) expansion of the A/B copolymer (A = 55.0 mol%) are shown in Figure 5.10. PBA shows a singlet centred around D = 174.2 ppm. As the concentration of acrylonitrile in the copolymer increases, signals characteristic of polybutylacrylate decrease, whereas a set of signals centred at around D = 173.5 ppm start appearing. These signals, with a further increase in the acrylonitrile content, increase to a maximum and then decrease, whereas a third new set of resonance signals appears at around D = 173.0 ppm. The three sets of signals whose intensities change with copolymer composition can be assigned to the carbonyl carbon of a central B unit in BBB, ABB (BBA) and ABA triad sequences form low to high field. Figure 5.11 shows the plots of normalised acrylonitrile (A) and butyl centred triad concentrations versus the mole fractions of acrylonitrile (fA) and butyl acrylate (fB) in the copolymers, respectively. The increase in the concentration of acrylonitrile in the copolymers increases the fraction of the AAA triad, whereas it decreases the fraction of the BAB triad. The fraction of the AAB triad initially increases with the increase in concentration of acrylonitrile, passes through a maximum value, and then starts decreasing. The maximum fractions of AAB and BBA triads are obtained at 0.60 and 0.55 mole fractions of the respective monomers.

5.2.3 Py-GC Yamada and co-workers [59] pointed out that spectroscopic methods such as IR spectroscopy and NMR spectroscopy previously used in sequencing studies on ethylene–propylene–diene and hydrogenated acrylonitrile butadiene rubbers often encountered the same difficulties experienced with the analysis of vulcanised rubbers, i.e., their insolubility. Due to the recent developments of excellent pyrolysers and highly efficient fused silica capillary columns, Py-GC has become an efficient tool to give unique information about sequencing and polymer structure. The technique has the practical advantage of being applicable to insoluble vulcanised rubbers.

180


1.0

(a)

0.9

(b)

0.8

Triad Fractions

0.7

BAB

ABA

0.6 0.5 0.4

AAB

ABB

0.3 0.2 AAA

0.1 0

BBB

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Figure 5.11 Normalised acrylonitrile (A) and butyl acrylate (B) centred triad concentration plotted against mole fraction of acrylonitrile (fA) and butyl acrylate (fB). Reproduced with permission from A.S. Brar and A. Sunita, Polymer, 1993, 34, 3391. © 1993, Elsevier

5.2.3.1 Ethylene–propylene diene The high-resolution Py-GC pyrograms obtained by Yamada and co-workers [59] were interpreted with regard to ethylidene norbornene (ENB) content. Several characteristic peaks of the degradation product were interpreted in terms of sequence distribution and ethylene–propylene composition. The composition of the polymers examined in this study is listed in Table 5.7. Sample A is a random copolymer whereas the remainder are random terpolymers. Propylene contents were determined by the IR method using the absorbance ratio of A at 1150 cm–1 (8.69 μm) to A at 720 cm–1 (13.89 μm) [60].

181


Table 5.7 EPM and EPDM samples

a

Samplea

Propylene (P), wt%b

ENB (D) wt%c

Composition, monomer ratio

A

49

-

E/P = 2/1

B-1

43

2.8

E/P/D = 2/1/0.02

B-2

43

4.5

E/P/D = 2/1/0.04

B-3

43

7.1

E/P/D = 2/1/0.06

B-4

43

12.3

E/P/D = 2/1/0.1

C-1

37

7.1

E/P/D = 2.5/1/0.04

C-2

28

7.1

E/P/D = 4/1/0.04

A: EPM, B-1, C-2: EPDM. Sample A is EPM, the others are EPDM.

b

The value of propylene content is obtained by regarding the total of the propylene and ethylene content as 100%, and is determined by IR (within 5% of coefficient of variance).

c

ENB content in the sample is determined by the iodine value method (within 5% of coefficient of variance). Reproduced with permission from Reproduced with permission from T. Yamada, T. Okumoto, H. Ohtani and S. Tsuge, Rubber Chemistry and Technology, 1990, 63, 2, 191. © 1990, ACS Rubber Division

Typical polymer structures are shown in Figure 5.12.

−E−

−P−

−D−

− ( CH2 − CH2 ) i − ( CH2 − CH ) m − ( CH − CH ) n − CH3 CH − CH3 mol ratio E/P/D = 2/1/0.1

PEEPEEEPEPEEPE D E P E E

Figure 5.12 Fundamental structure and sequence distribution of a typical ethylene-propylene-ethlydene-norbornene rubber. Reproduced with permission from T. Yamada, T. Okumoto, H. Ohtani and S. Tsuge, Rubber Chemistry and Technology, 1990, 63, 2, 191. ©1990, ACS Rubber Division

182

Sequencing of Monomer Unit in Polymers Figure 5.13 shows the pyrograms of EPDM with various ethylene/propylene (E/P) compositions and a constant ENB content (7.1 wt%). The fact that linear olefin peaks (LE) are observed up to C20 suggests that more than the heptad of the ethylene sequence (–CH2–CH)7–) exists in the polymer chain, even though both sides of the terminal groups in linear olefins incorporate propylene units (see Scheme 1)

Scheme 1 – Linear olefins

Relative peak intensities of LE above C10 increase with the rise of ethylene content in the EPDM. The fact that the peak of the propylene trimer (P3) is observed while the tetramer peak (P4) is not, suggests that at least the triad propylene sequence (-CH2-CH(CH3)-)3 exists in the polymer chain. Furthermore, intensity of the P3 peak decreases as the ethylene content rises in the EPDM. The peaks in the C6, C7, C8 and C10 regions were examined in detail because these peaks are expected to contain the information with respect to the E–P sequence distributions. Among these, it was proved that 1-O (C6), 1-O (C7) and 2-MO (C7) and 2-MO (C8) reflect the triad sequences such as –EEE–, –PEE–, and –PEP–, respectively. In a detailed pyrogram of an ethylene-propylene diene terpolymer (EPDM) (Sample B-3) in the C7 region the characteristic peaks identified are summarised in Table 5.8. Assuming that formation of the degradation products is primarily formed through random bond cleavages, several peaks are assigned to the triad sequence (Table 5.8). Among these, 1-O (C7) is the strongest peak, and 2-MO (C7) is the second strongest for all EPDM examined. Both of these products reflect the propylene-ethylene-ethylene (PEE) sequence in EPDM. Considering the most provable tetrad, the propylene-ethyleneethylene-propylene (PEEP) sequence in EPDM with the E/P composition raging from 2/1 to 3/1, these two products may be mainly formed through the following backbiting mechanisms (Schemes 2 and 3):

183

A

175.8103 B

175.407

176.0905

175.2785


D

C

SA - 80

SA - 100

SA - 80

SA - 50

SA - 40 (A)

(B)

SA - 20

SA - 10

Figure 5.13 Pyrograms of EPDM with various E/P compositions. Source: Author’s own files.

Scheme 2 – Mechanism for the formation of 1-O (CT)

184


Scheme 3 – Mechanism for the formation of 2-MO (CT) A relationship exists between the peak intensities of the characteristic products such as 1-O (C6), 2-MO (C8), and 1-O (C10) and the E/P content in EPDM. In general, as the ethylene content increases in the EPDM, the ethylene sequence length becomes longer, and the amount of the isolated ethylene units decreases. As described previously, 1-O (C6) and 1-O (C10) are formed from the longer ethylene sequences than the triad. Intensities of 1-O (C6) and 1-O (C10) monotonously increase with the rise of the ethylene content in the EPDM. Conversely, 2-MO (C8), which is correlated to the isolated ethylene unit, monotonously decreases as the ethylene content increases. Consequently, these relations can be used to estimate the E/P composition in EPDM.

5.2.3.2 Hydrogenated acrylonitrile–butadiene copolymers (NBR) In work by Kondo and co-workers [61], the microstructures of hydrogenated NBR were investigated by spectroscopic methods such as IR and NMR, and by high-resolution Py–GC. Degrees of hydrogenation were calculated from the intensities of characteristic peaks in IR and 1H-NMR spectra of the samples, and the results compared with those determined by an iodine value method. Pyrograms of hydrogenated NBR were interpreted with regard to the degree of hydrogenation. Several peaks of larger degradation products were correlated to long sequences in the polymer chain. Newly assigned characteristic peaks in a high-resolution 13C-NMR spectrum were interpreted in terms of the sequence distribution and hydrogenation mechanisms. Pyrograms of NBR samples at 550 °C before and after hydrogenation (a) N-37(0), (b) N-37(440), and (c) N-37(98). Table 5.9 summarises the characteristic thermal degradation products observed on the pyrograms of NBR and hydrogenated NBR. Mass spectral data of the characteristic degradation products enabled many of them to be identified.

185


Table 5.8 Characteristic degradation products for the C7 region Peak

Compound

Structure

bp, oC

Sequence

3-MO

3-methyl-1-hexene

C=C-C-C-C-C

84

EPE

85

PEE

86

EPE

92

PEE

C 5-MO

5-methyl-1-hexene

C=C-C-C-C-C C

4-MO

4-methyl-1-hexene

C=C-C-C-C-C C

2-MO

2-methyl-1-hexene

C=C-C-C-C-C C

1-0

1-heptene

C=C-C-C-C-C-C

94

PEE

P

n-heptene

C-C-C-C-C-C-C

98

PEE

2-0

2-heptene

C-C=C-C-C-C-C

98–98.5

PEE

Reproduced with permission from T. Yamada, T. Okumoto, H. Ohtani and S. Tsuge, Rubber Chemistry and Technology, 1990, 63, 2, 191. © 1990, ACS Rubber Division

Characteristic peaks in the pyrogram of N-37(0) are butadiene (BD) monomer, BD dimer, and acrylonitrile (ACN) monomer. Those of hydrogenated NBR consists of a series of linear mononitriles (MN(A)) up to C12, each of which consists of a doublet corresponding to an A-olefinic MN(A) (the former) and a saturated MN(A) (the latter). Another series of mononitrile positional isomers (MN(B)) are also observed. C11 mononitrile (MN(C)) and C9 and C10 dinitriles (DN), which reflect the alternate arrangement of ACN and hydrogenated BD units (BD–ACN–BD and ACN–BD–ACN, respectively), are also observed. A series of peaks of hydrocarbons (HC) are observed, which reflect the methylene chains produced by hydrogenation. HC peaks of each carbon number consist of a triplet corresponding to an A, W-di-olefin, an A-olefin, and a n-alkane. The fact that up to C12-HC peaks are observed suggests that at least a hydrogenated BD–BD–BD sequence exists in the polymer chain. The peak intensity of the unsaturated hydrocarbon C7-MN(A) i.e., CH2 = CH(CH2)4 – C y N obtained by Py-GC provided a practical calibration curve applicable to ever highly hydrogenated acrylonitrile–butadiene copolymers.

186


Table 5.9 Characteristic degradation products from hydrogenated NBR Compound Butadiene

Abbreviation

Structure

Sequence

BD

CH2═CH-CH=CH2

B

Butadiene dimer (4-vinyl cyclohexene)

VCH

Acrylonitrile

ACN

Hydrocarbons

Mononitriles

BB

CH2═CHCN

A

HC

MN(A) MN(B)

Dinitriles

MN(C)

EAE

MN(D)

BA

DN

NyC(CH2)7CyN

AEA AEA

B = 1,4-butadiene unit; A = acrylonitrile; E = hydrogenated 1,4-butadiene unit. Reproduced with permission from A. Kondo, H. Ohtani, Y. Kosugi, S. Tsuge, Y. Kubo, N. Asada, H. Inaki and A. Yoshioka, Macromolecules, 1988, 21, 10, 2918. © 1988, ACS

5.2.3.3 Butadiene–acrylonitrile–methacrylic acid–terpolymer Rao and co-workers [62] applied Py-GC and 13C-NMR to the determination of sequence distribution of butadiene (B) – acrylonitrile (A) – methacrylic acid (M) terpolymers. Sequence distribution was described in terms of six triads (BBB, ABA, ABB, BBA, MBR and AMB) and found to vary with the mode of addition of methacrylic acid monomer. NMR data were found to be in good agreement with a mechanism of polymerisation in which methacrylic acid is preferentially involved in initiation reactions by a

187

Introduction to Polymer Analysis thiyl radical arising from the reaction of the chain modifier, 1-dodecanethiol, and cyanoisopropyl radical generated from azo-bis-isobutyro-nitrile initiator. Binders obtained by curing the liquid terpolymers with an epoxy resin showed widely varying mechanical properties, with tensile strength varying from 4.3 kg/cm2 to 0.6 kg/ cm2, and elongation at break from 130% to 450%. Tensile strength increased and elongation decreased with the number of acrylonitrile units between the methacrylic acid crosslinks. Good correlation was obtained between triad population ratio [ABB + BBA] [BBvD]/[ABA][MBB] and the mechanical properties of the binders. Pyrolysis–GC data at 550 °C and 600 °C confirmed the results obtained from 13C-NMR, and the mole ratio of butadiene to arylonitrile in the pyrolysates showed correlation with the properties of binders. Rao and co-workers [62] classified into (I) triads containing butadiene (b), (II) triads containing butadiene (B) and acrylonitrile (A) and (III) triads containing B, A and M units. These workers discuss, under separate headings, the following types of triads: BB, ABA, ABB, BBA, MBM, MBB and BBM under olefinic carbon resonances; and BBB, ABA, ABB triads under nitrile carbon resonances. The main conclusions that can be drawn from the 13C-NMR studies of the terpolymers can be summarised as follows: (i) The average configuration of the butadiene units is essentially identical in all the polymers. (ii) The incorporation of 1-dodecyl thiyl moiety into the polymer chain is quite considerable for all the polymers. (iii) The number of butadiene units connected to methacrylic acid unit is not commensurate with the concentration of the acid when compared with the number of butadiene units connected to acrylonitrile units. The former increases from polymer I (Mw = 4978, acrylonitrile 11.8%) to polymer V (Mw = 2781, 11.8% acrylonitrile) (Table 5.10). (iv) Polymer I (Mw = 4978, 11.8% acrylonitrile) contains triads of the type ABM or BMA. (v) The fractional population of the triads ABA increases from polymer I (Mw = 4978, 11.8% acrylonitrile) to V (Mw = 2781, 11.8% acrylonitrile). (vi) The concentration of the terminal methyl groups varies from polymer to polymer.

188

Sequencing of Monomer Unit in Polymers All these observations can be rationalised by assuming the mechanism of polymerisation to be similar to free-radical addition of thiols to olefins as outlined next. In summary, the average microstructure of the terpolymers as established by Rao and co-workers [62] can be represented as: C12H25SM (BBB)a (BBA)b (BBM)c (BMA)d (ABA)e CH2 – CH = CH – CH3 The essential differences between the structures of polymers I and V are the: (i) number of BBA units between two methacrylic acid units decreases from polymer I to V, as indicated by the decrease in the population ratio of (ABB + BBA) and ABA triads (Table 5.10); and (ii) fraction of MBB triads increases and consequently the fraction of AMB or BMA triads decreases from polymer I to V, as reflected by the decrease in the ratio of BBvB and MBB sequences. Terpolymer samples (0.6–0.7 mg) were pyrolysed at 550 °C and 600 °C. The major products of pyrolysis were found to be ethylene, propylene, butadiene, benzene, toluene, and vinyl cyclohexene from the butadiene part, and acrylonitrile and acetonitrile from the acrylonitrile part of the polymers.

Table 5.10 Triad intensity ratios for terpolymers Sample Polymer [ABB] + [BBA] number [ABA]

a

[BBB]a [BBvB]b [BBvB]b [ABB] + [BBA] [MBB] [MBB] [t-CH3]c [ABA]

1

I

6.2

8.2

5.0

5.1

31.0

2

II

5.7

5.5

2.7

2.2

15.4

3

III

4.7

4.8

2.2

3.1

10.3

4

IV

4.4

4.7

2.3

2.0

10.1

5

V

3.7

4.2

2.1

1.8

7.8

× [BBvB]

[MBB]

Calculated from Sp2 carbon resonances, [BBB] = ttc/t

b

Calculated from Sp3 carbon resonances

c

t-CH3 = terminal CH3

A = acrylonitrile B = butadiene M = methacrylic acid units Reproduced with permission from M.R. Rao, T.V. Sebastian, T.S. Radhakrishnan and P.V. Ravindran, Journal of Applied Polymer Science, 1991, 42, 3, 753. © 1991, Wiley

189

Introduction to Polymer Analysis The composition of the products of pyrolysis at 550 °C (Table 5.11) shows that mole ratios of butadiene/vinylcyclohexane and butadiene/acrylonitrile increases from polymer I to V, although the change in butadiene/vinylcyclohexane ratio is only marginal. Similar trends are also observed at the pyrolysis temperature of 600 °C (Table 5.11), although the differences narrow down considerably. These ratios are sensitive to the sequence distribution of butadiene and acrylonitrile units in the polymer chain. Formation of vinyl cyclohexane, a dimer of butadiene, requires two butadiene moieties adjacent to each other and, consequently, in copolymers of butadiene. The vinylcyclohexane/butadiene ratio is dependent on the distribution of comonomers and becomes almost zero for an alternating copolymer. Hence, the marginal increase in vinylcyclohexane/butadiene ratio from polymer I to V reflects the increasing alternating nature of the placement of acrylonitrile units in the chain. This is consistent with the observation from 13C-NMR studies that the population of ABA triads increase from polymer I to V. The mechanism of degradation of polybutadiene, as suggested by Golub and Gargiulo [63], involves main-chain scission to give radical ends, which can undergo depolymerisation (to yield mainly vinyl cyclohexane and butadiene) [64] or cyclisation to give cyclised polybutadiene. For copolymers of butadiene and acrylonitrile, cyclisation will be more facile due to the pendant nitrile group as in the case with polyacrylonitrile polymers [65, 55]. This mechanism of decomposition suggests that the greater the alternating nature of the polymers, the greater will be the extent of cyclisation reactions. The cyclised polymers would undergo further thermal degradation to produce various products.

5.2.3.4 Styrene-n-butyl acrylate Wang and Smith [66] described a Py-GC method to investigate the microstructure of emulsion polymers. The number-average sequence length, which reflects the monomer arrangement in the polymer, was calculated using the formulae that incorporate the pure trimer peak intensities and hybrid trimer peak intensities. In this study, styrene and n-butyl acrylate copolymer systems were used to measure the ‘degree of structure’ (i.e., number-average sequence length for styrene and n-butyl acrylate repeat units) and compared with a homogeneous non-structured (or random) copolymer. Numberaverage sequence length information was further extended to calculate composition. For the emulsion polymers examined in this study, the composition elucidated from the number-average sequence length matched the preparation recipe and/or what was measured by 13C-NMR.

190

1.4

1.5

1.5

1.4

2.4

2.4

2.3

2.2

2.4

2.2

III

IV

V

VII

I

II

III

IV

V

VII

7.8

8.6

8.2

8.2

8.6

8.8

7.4

7.9

7.4

7.7

7.3

4.8

5.5

5.2

5.1

5.6

5.4

4.5

5.0

4.4

4.5

4.5

4.7

C3 (2.2)

21.4

23.1

22.0

20.8

20.9

20.9

25.7

26.5

25.9

26.2

23.2

24.3

BD (3.5)

ACN (4.7)

C5 (5.6)

BZ (9.1)

2.0

1.8

2.1

2.1

2.2

3.6

7.7

7.1

7.1

7.4

7.4

7.3

10.3

9.0

8.6

9.7

8.5

9.3

1.5

1.0

1.2

1.1

1.4

1.3

2.5

2.3

2.5

2.3

2.6

3.6

8.0

7.4

7.5

7.9

7.9

8.1

8.1

7.3

7.6

7.3

7.5

7.6

Pyrolysis Temperature: 600 oC

1.1

0.8

1.2

1.2

1.1

0.4

Pyrolysis Temperature: 550 oC

AcN (4.0)

Values in brackets are the retention times in minutes

1.4

II

7.5

C2 (1.3)

Wt%a

7.2

7.0

7.0

6.8

6.9

6.8

6.3

6.2

5.2

5.3

5.4

5.4

T (15.2)

4.1

3.5

3.5

3.7

3.4

3.5

6.7

5.8

6.4

6.2

5.7

6.7

VCH (24.1)

11.2 + 0.3 12.6 + 0.4 13.2 + 0.4 10.4 + 0.3

8.6 + 0.3 9.9 + 0.3 8.4 + 0.2

7.7 + 0.2

12.6 + 0.4

8.9 + 0.3

9.1 + 0.3

14.4 + 0.4

12.3 + 0.3

8.1 + 0.2

12.1 + 0.3

7.9 + 0.2

8.5 + 0.3

12.3 + 0.4

11.9 + 0.3

8.1 + 0.2

10.3 + 0.3

5.7 + 0.2

7.2 + 0.2

BD/VCH

6.6 + 0.2

BD/ACN

Mol%

Reproduced with permission from M.R. Rao, T.V. Sebastian, T.S. Radhakrishnan and P.V. Ravindran, Journal of Applied Polymer Science, 1991, 42, 3, 753. © 1991, Wiley

a

1.3

C1 (1.0)

I

Polymer

Table 5.11 Composition of major pyrolysates


191

Introduction to Polymer Analysis Pyrolysis followed by gas chromatographic separation uses thermal energy to break down a polymeric structure to monomers and oligomers, and separation of those units for quantification. Because of the temperature limitations of the common silicone capillary column, only the dimer and trimers of the system studied can be reliably separated and detected. The major mechanism of producing dimers and trimers with pyrolysis can be attributed to thermal degradation. A relatively small amount of dimers and trimers is formed as a result of a recombination of monomers. This mechanism is demonstrated as follows. The intensity of the various dimer and trimer peaks in a pyrolysis gas chromatogram reflect the monomer sequence. Pyrolysis of an emulsion polymer is done on the dried film. The liquid emulsion is heated in the pyrolysis chamber at 250 °C for 10 minutes and allowed to coalesce to a solid. A volatility experiment showed there were no detectable materials released during this period. The response of the FID detector is assumed equal for all three styrene-centred trimers and for all three n-butyl acrylate-centred trimers in this study. Essentially, FID is a carbon atom counter; any components having the same number of carbon atoms should have the same response. The styrene-centred trimers have 22–24 carbon atoms; the difference in carbon atoms is less than ±5% around those trimers. This fact makes the equal response assumption valid. The same argument also applied to n-butyl acrylate-centred trimers. A pyrolysis temperature of 500 °C was chosen to obtain a higher yield of trimer for styrene and n-butyl acrylate (BA). Figure 5.14 shows the typical pyrogram of a 50%/50% by weight STY/BA homogeneous emulsion polymer. Figure 5.15 is an expansion of the trimer area in Figure 5.14. Figure 5.16 shows the pyrogram of the trimer area for five compositions of STY/nBA. The number-average sequence lengths were calculated for all five polymers. Peak areas were normalised on the basis of the summation of NSSS, NSSBA+BASS, and NBASBA equalling 1, and the summation of NBABABA, NSBABA+BASS, and NSBAS equalling 1. All normalised peak areas were then used to calculate the number-average sequence length for styrene and n-butyl acrylate (Table 5.12).

192

Sequencing of Monomer Unit in Polymers 22

Styrene

20 18 16 14 12 10 n-Ba

8

Trimers

6 4

dimers

2 -0 5

10

15 20 25 30 35 Rotantion Time (minutes)

40

45

Figure 5.14 GC pyrogram at 500 oC of a 50:50% styrene-n-butyl acrylate homogenous emulsion polymer. Source: Author’s own files. SBaS

7.0

6.0

5.0

4.0 BaBaS.SBaBa 3.0

BaSBa SSBa + BaSS

2.0 BaBaBa

SSS

1.0

0.0 38.0

38.5

38.5

39.0

39.5

40.0

40.5

41.0

41.5

42.0

42.5

43.0

43.5

44.0

Rotantion Time (minutes)

Figure 5.15 Expansion of trimer area at retention time 40–44 min in Figure 5.14 (S = styrene, BA = butyl acrylate).

193

Introduction to Polymer Analysis Wang and Smith [66] conclude that by applying the appropriate statistical formula and the data obtainable from Py-GC, the number-average sequence length, as well as the monomer composition of an emulsion copolymer, can be explored. The structure of a copolymer of two monomeric types can be quantified by deriving the percent of grouped monomers and the number-average length of grouped monomers. This method could be extended to any copolymer system as long as all six trimer peaks can be identified and the peak intensities obtained by assuming that these intensities represent the polymer compositions. This method extends the capabilities of pyrolysis not only in the quantitative study of monomer composition, but also in the realm of polymer structure investigation.

Latex E

Latex D SSS SSBa + BaSS BaSBa SBaS

Latex C

BaBaS + SBaBa BaBaBa Latex B

Latex A

35

36

37

38

39

40

41

42

43

44

45

46

Retention Time (minutes)

Figure 5.16 GC pyrogram of the trimer area for five compositions of styrene-butyl acrylate copolymer. Source: Author’s own files.

194

Sequencing of Monomer Unit in Polymers In further work, Wang and Smith [15] applied Py-GC to the determination of the number-average trimer sequence lengths of grouped monomers in STY/BA copolymers. The method can be applied to any copolymer system as long as all six trimer peaks are identified and peak intensities obtained.

Table 5.12 Number-average sequence length for different compositions of homogeneous emulsion polymers from the pyrolysis gas chromatography method A

B

C

D

E

SSS SSBA+BASS BASBA

0.069 0.058 0.872

0.116 0.221 0.663

0.168 0.322 0.511

0.610 0.305 0.085

0.733 0.210 0.057

N(S)

1.11

1.29

1.49

4.21

6.17

BABABA BABAS+SBABA SBAS

0.374 0.401 0.224

0.076 0.379 0.545

0.058 0.323 0.620

0.005 0.123 0.872

0.031 0.106 0.863

N(BA)

2.35

1.36

1.28

1.07

1.09

mol% S BA

32 68

49 51

54 46

80 20

85 15

Experimental wt S BA

28 72

44 56

49 51

76 24

82 18

Standard wt% S BA

25 75

43 57

50 50

74a 26a

82 18

Grouped % S BA

13 78

34 46

49 38

91 13

94 14

Grouped N(S) N(BA)

4.39 3.86

3.05 2.40

3.04 2.36

6.00 2.08

8.98 2.59

Normalised peak intensity

a

Weight percentage determined by 13C-NMR analysis

Source: Author’s own files.

195


5.2.3.5 Ethylene oxide condensates Ishida and co-workers [67] applied reactive pyrolysis in the presence of cobalt sulfate as a catalyst to the evaluation of ethylene oxide sequence up to E7 in copolymers on the basis of peak intensities of the cyclic ethers containing ethylene oxide unit found on the pyrogram. Pyrolysis in the presence of cobalt sulfate gives a pyrogram in which cyclic ethers containing ethylene oxide (E) and oxymethylene units (F) are predominant. In this case, much larger cyclic ethers are observed up to the E7F and (E2F)3 using a capillary column. The peaks of E2F and F4 overlapped completely when using a non-polar column such as poly(dimethylsiloxane). By using mildly polar poly(methylphenylsiloxane), they were sufficiently resolved. The distributions of E units, as well as the E contents of the polymers, can be determined from the intensities of peaks due to these cyclic ethers obtained if the peaks on the pyrogram reflect the chemical structures in the original polymer chain. Three model polymers were subjected to Py–GC analysis in the presence of cobalt sulfate to confirm that the sequence distributions estimated by the cyclic ethers on the pyrograms reflect those in the original polymer chain. Figure 5.17 shows the pyrograms of (a) polyoxymethylene homopolymer, (b) polyoxymethylene–1,3-dioxolane (DO) copolymer, and (c) polyoxymethylene–1,3,6-trioxocane copolymer at 400 °C in the presence of 5 wt% cobalt sulfate. As would be expected from the polymer structure, in the pyrogram of the polyoxymethylene homopolymer (a), formaldehyde (F) and cyclic compounds comprising only F units are observed, whereas no cyclic ether containing E unit(s) is observed. This result suggests that the formation of E unit(s) from oxymethylene sequences does not occur during the reactive pyrolysis of the polyoxymethylene homopolymer sample. Similarly, in the pyrogram of polyoxymethylene–1,3dioxolane copolymer (b) (CH2O) n–(CH2CHO)–(CH2CH2O)m (5.38% ethylene oxide units sequence length = 1) only cyclic ethers reflected isolated E units, such as EF, EF2 and FE3, are observed in addition to the cyclic ethers formed from the F sequences. In the pyrogram of the polyoxymethylene trioxocane copolymer (c) (CH2O)n– CH2CH2O– CH2– CH2) (CH2)m (7.4% ethylene oxide units, sequence length of ethylene oxide units = 2), cyclic ethers reflecting EE diads such as E2F, (E2F)2, and (E2F)3 are characteristically observed. Here, the fact that a small peak of EF is observed suggests that the degradation of the EE diad in the polymer chain to form EF occurs to some extent. Because the peak intensity of EF is much smaller than those of cyclic ethers reflecting EE diads such as E2F, (E2F)2, and (E2F)2, the contribution

196

Sequencing of Monomer Unit in Polymers of the degraded EE diad could be negligible. These results observed for the model polymers indicate that the sequence distributions of E units in the original polymer chain are almost quantitatively reflected in the cyclic ethers formed through the reactive pyrolysis in the presence of cobalt sulfate.

1 F

(a)

F1

6

F3

3 F5

8

(a) 2

EF2

EF

(b)

4

F

1 6 7 EF3

F3

F1

3

(c)

5 (E2F)2

E2F

11

0

15 (E2F)3

EF F3

2 3

6 F1

F

1

20

40

60

Retention time (min)

Figure 5.17 Pyrogram of polyoxymethylene (a) homopolymer, (b) polyoxymethylene-1,3 dioxolane copolymer (c) polyoxymethylene 1,3,6 trioxocane copolymer at 400 oC in presence of 5 wt% cobalt sulfate. Reproduced with permission from Y. Ishida, H. Ohtani, K. Abe, S. Tsuge, K. Yamamoto and K. Katoh, Macromolecules, 1995, 28, 19, 6528. © 1995, ACS

197

Introduction to Polymer Analysis The method was applied to the determination of ethylene oxide sequences up to 7 units long in copolymers. On the basis of peak intensities of cyclic ethers containing ethylene oxide units in the programme, pyrolysis results were in reasonably good agreement with those obtained by hydrolysis–gas chromatography (Table 5.13).

Table 5.13 Sequence distributions of polyacetal copolymers estimated by pyrolysis–gas chromatography POM-EO-1 [98.6/1.4]

POM-EO-2 [95.3/4.7]

POM-EO-E [91.1/8.9]

Py-GCa

Hydrolysisb

Py-GCa

Hydrolysisb

Py-GCa

Hydrolysisb

-FEF-

70.1 (74.5)

79.3

60.0

59.0

44.3

47.1

-FE2F-

25.5 (22.3)

17.2

33.0

31.6

37.1

32.2

-FE3F-

4.4 (3.2)

3.5

5.6

8.7

9.5

14.9

1.4

0.7

5.2

4.9

-FE5E-

2.7

0.9

-FE6E-

0.9

-FE7F-

0.3

-FE4F-

Total

100.0

100.0

100.0

a

100.0

100.0

100.0

o

Sequence distribution obtained by Py-GC through the reactive pyrolysis at 400 C in the presence of 5 wt% cobalt sulfate. Values in parentheses are obtained in the presence of 1 wt% cobalt sulfate. b Sequence distribution obtained from hydrolysis followed by gas chromatography. POM-EO = Polyoxymethylene - ethylene oxide E = Ethylene oxide units F = Oxymethylene units Reproduced with permission from Y. Ishida, H. Ohtani, K. Abe, S. Tsuge, K. Yamamoto and K. Katoh, Macromolecules, 1995, 28, 19, 6528. © 1995, ACS

5.2.4 SIMS 5.2.4.1 Polydimethyl siloxane–urethane

198

Sequencing of Monomer Unit in Polymers So far in this chapter on sequencing, two techniques have predominated: NMR spectroscopy and Py–GC. An exception is the work of Zhuang and co-workers [67] who used ToF-SIMS in their study of the distribution of polydimethylsiloxane (PDMS) segment lengths at the surface of PDMS–urethane-segmented copolymers. Their aim was to establish whether, at the copolymer surface, the distribution of segment or chain lengths is different from that in the bulk of the polymer. SIMS has been emerging as a potential tool, particularly with ToF detection technologies. ToF technologies can yield increased resolution, mass range, and transmission. Before the work of Zhuang and co-workers [60], ToF-SIMS had not provided information on segment length distribution at the surface of a multicomponent polymer, particularly in the form of a thick film. The difficulties in accomplishing this are primarily due to: (a) lack of structurally well-defined copolymers, (b) charging effects at the surfaces upon ion beam perturbation [68–71], (c) lack of cationising ions, and (d) polymer chain entanglement and interchain and/or intrachain interactions in a polymer. In many cases, complete charge compensation for the analysis of an insulating sample can be attained in ToF-SIMS by flooding the surface with low-energy (10 eV), electrons pulsed between ion pulses [72–76]. Figure 5.18 (a) and (b) show the positive ToF-SIMS spectra from the thick PU-PDMS film in the range of m/z 0–300 and 1000–2500. Comparing Figure 5.18 (b) and 5.18 (c), a remarkable resemblance is seen between the spectrum from the copolymer and that originating from pure PDMS, except that the peak at m/z = 116 from the PDMS end cap is noticeably lower in the PU-PDMS spectrum (which may arise from the fact that, after incorporation in the PU-PDMS spectrum, it requires two bonds to be broken to form the 116 fragment). This is a strong indication that the PDMS segment compositionally dominates the surface of the thick PU-PDMS film. The peaks in Figure 5.18(b) are separated from each other by 74 Da. According to a ‘simple statistical model’ for chain scission [73] stating that only main-chain scission occurs, the masses of all possible fragments formed by any two chain cleavages along the PU-PDMS copolymer backbone were calculated and compared with the mass values of the peaks in Figure 5.18(b). Only fragments having the structure represented in Scheme 1 (see page 200) agree with the series observed in Figure 5.18(b). At this point, it is not known what caused the stability of this particular fragment ion structure, but it is certain that the spectrum observed is not consistent with any other type of fragmentation considered from bond breaking. All other ion structures from bond breaking near or within the urea linkage were rejected because they were not consistent with the repeating mass pattern or the isotopic pattern of the repeating cluster.

199


CH3 HN

CH2

3

Si CH3

O

CH3 O

Si

CH2

3

NH C

HN

CH3

CH2 NH

ion beam

NH

O CH2 NH C

CH3 NH CH2

3

Si CH3

CH3 O

Si

CH2

3

HN

CH3 CH2 NH

NH

Scheme 1 Fragmentation mechanism of the PU-PDMS copolymer in the form of a thick film Based on the assigned fragment structure, the length distribution (relative intensities versus m values) was constructed. It was calculated that Mn = 1131.1, Mw = 1171.4, and Mw /Mn = 1.1, very close to those values of the pure PDMS prepolymers. This result suggests that the PDMS segment segregated at the surface of the PU-PDMS copolymer with a PDMS nominal Mw of approximately 1000 Da is essentially identical with that in the bulk in terms of PDMS segment length distribution. The distribution of PDMS segment lengths segregated at the surface was nearly identical with that in the bulk of the PU-PDMS copolymer with PDMS of a nominal Mw of approximately 1000 Da. This accomplishment enables study of segment length distributions at the surface of other siloxane copolymers as a function of bulk segment length distribution and polymer processing. By comparing the low-mass (m/z = 0–300) spectra from the submonolaye m/z PDMS prepolymer film on Ag and the thick PU-PDMS copolymer film on A1, it was noted that the PDMS segment compositionally dominated the surface of the thick PU-PDMS film. This observation agrees well with earlier electron spectroscopy for chemical analysis (ESCA) results. Ions detected and assigned to fragments in the low-mass range (m/z b300) provided structural information about the repeat units and the end groups. The high mass spectrum of the PDMS homopolymer yielded a series of ions assigned to Ag+ cationised oligomers; this enabled determination of the molecular weight distribution. In the highmass (m/z = 800–2500) spectra of thick PU-PDMS films, the peak series was assigned to a simple fragmentation process. That process would yield ions where the intact PDMS segment is present; it therefore can be used to evaluate the PDMS segment length 200

Sequencing of Monomer Unit in Polymers distribution at the surface of the copolymer. Distribution of PDMS segment lengths segregated at the surface of the thick film was almost identical with that in the bulk of PU-PDMS with PDMS nominal Mw of approximately 1000 Da. These results allow the development of an analysis of ion structure and a stepwise procedure for evaluating the segment length distributions in the near-surface region of siloxanes. (a) 28

20000

73

counts/channel

15000 43 10000 1

147

5000 15

59

133

207

281

0 0

50

100

150 m/z

200

250

300

(b) 0.3

counts/channel

0.2

0.1

0.0 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 m/z

(c) 50

counts/channel

40 30 850

20

900

950 1000

* [PDMS oligomer + Ag]+

10 0 600

800

1000 1200 1400 1600 1800 2000 m/z

Figure 5.18 ToF-SIMS spectra from polyurethane (PU)-PDMS film in range m/z (a) of 0–300 amu, (b) of 1000–2500 amu (c) of 600–2000 amu (PDMS oligomer + Ag+). Reproduced with permission from H. Zhuang, J.A. Gardella, Jr., and D.M. Hercules, Macromolecules, 1997, 30, 4, 1153. © 1997, ACS

201

CO CH2CH2CH2CH3

CN BBA

(B)

O

DMSO - d6 (CH2)B (C2)B

(A)

B

AA

AA

A

BA

B

ABA BBB

(OCH2)B

(CN)A

(x = 0)B

160

140

120

(CH2)A (CH)B

176 175 174 173 172 171 123 122 121 120 119 118 117 (ppm)

(CH)B

(CN)A

(x = 0)B

180

(C3)B

(CH2)CH)(CH2CH)

(CH3)B


(CDCl)3

100

80

60

40

20

0

Figure 5.19 ToF-SIMS spectra from PU-PDMS film in the range m/z (a) of 0–300 amu, (b) of 1000–2500 amu (c) of 600–2000 amu (PDMS oligomer + Ag+). Reproduced with permission from H. Zhuang, J.A. Gardella, Jr., and D.M. Hercules, Macromolecules, 1997, 30, 4, 1153. © 1997, ACS

Wide-angle ESCA has been used to determine the length of PDMS segments in PDMS. These results were evaluated using a concentration–depth profile deconvolution programme, and continuous concentration–depth profiles of the hard segment were reported in the near-surface region [77–79]. This yields a more intuitive understanding of the compositional features of PU-PDMS copolymers. Each of the aforementioned

202

Sequencing of Monomer Unit in Polymers cases showed a PDMS-rich surface due to the incompatibility of the soft segment and hard segment, and the lower surface energy of the soft segment in the segmented copolymers.

5.2.5 Ozonolysis Techniques 5.2.5.1 Butadiene–propylene This technique has been applied to a study of sequencing in butadiene–propylene copolymer [80–82]. Samples of highly alternating copolymers of butadiene and propylene yielded large amounts of 3-methyl 1,6 hexane dial when submitted to ozonolysis. The ozonolysis product from 4-methyl cyclohexane-1 was used as a model compound for this structure. Ozonolysis of these polymers occurs as shown next:

Table 5.14 shows results obtained for several butadiene–propylene copolymers having more or less alternating structure.

5.2.5.2 Styrene butadiene copolymers Ozonisation followed by GPC has been employed by Tanaka and co-workers [83] to study sequencing of vulcanised styrene–butadiene copolymers. Tanka and co-workers [83] carried out the ozonolysis in methylene dichloride and examined the fractions obtained after GPC by 1H-NMR. These workers found nonad, diad and triad styrene sequences flanked by 1,4 butadiene units and long styrene sequences:

203

204

47.8

53.1

-

B

C

D

-

3.2

2.2

5.7

1,2 (%)

30

43.7

50

49.3

Propylene (mole%)


45

A

1,4 (%)

49

25

11.5

5

Succinaldehyde

38

61

85

92

3-Methyl-1,6hexane-dial

1

6

0.5

1

3-Formyl 1,6hexane-dial

12

8

3

2

4-Octene - 1,8-dial

Area from gas chromatography (%)

Table 5.14 Microzonolysis of butadiene-propylene copolymers

33

48

71

77

Alternating BD/Pr (%)


Sequencing of Monomer Unit in Polymers In further work on the configurational sequences in styrene units and the arrangement of styrene and 1,2 butadiene units in styrene–butadiene rubber, Tanaka and coworkers [84] carried out 1H- and 13C-NMR spectroscopy on products obtained by ozonisation–gel permeation chromatography and ozonisation - high performance liquid chromatography (HPLC). Ozonides were produced from diad and triad styrene sequences and from the 1,4 butadiene sequences which flanked them. The chromatograms showed 2 or 3 peaks corresponding to styrene diad and triad. Ozonolysis products obtained from styrene and 1,2 butadiene sequences were separated in up to three fractions by HPLC. The first and second of these peaks in these fractions were assigned to 1,4 and 1,2 butadiene units, and the peaks in the third fraction tube assigned to meso and racemic forms of 1,4 styrene, 1,4 butadiene sequence structure. Tanaka and co-workers [25], again using the ozonolysis–GPC technique, showed that the ozonolysis products obtained from styrene – butadiene and styrene isoprene copolymers indicated 77% to 99% styrene block sequences in linear copolymers. A linear copolymer could be distinguished from star copolymer by comparing the Mw and chemical composition of the main peak and shoulder peaks by GPC, and also by comparing the Mw of the block styrene sequences determined by ozonolysis.

References 1. Y. Inoue, A. Nishiolka and R. Chujo, Die Makromoleculare Chemie, 1972, 152, 1, 15. 2. A. Zambelli, D.E. Dorman, A.I.R. Brewster and F.A. Bovey, Macromolecules, 1973, 6, 6, 925. 3. J.C. Randall, Journal of Polymer Science: Polymer Physics Edition, 1974, 12, 4, 703. 4. J.C. Randall, Journal of Polymer Science: Polymer Physics Edition, 1976, 14, 11, 2083. 5. J.C. Randall, Journal of Polymer Science: Polymer Physics Edition, 1976, 14, 9, 1693. 6. L. Cavalli, G.C. Borsini, G. Carraro and G. Confalonieri, Journal of Polymer Science, Polymer Chemistry Edition, 1970, 8, 4, 801. 7. P. Corradini, V. Busico and G. Guerra in Comprehensive Polymer Science, Pergamon Press, Oxford, UK, 1989, 4, 29. 8. E. Schroeder and M. Byrdy, Plaste und Kautschuk, 1977, 24, 11, 757.

205

Introduction to Polymer Analysis 9. V. Busico, P. Corradini, R. De Biazo, L. Landriani and A.L. Segre, Macromolecules, 1994, 27, 16, 4521. 10. G. Di Silvestro, P. Sozzani, B. Savare and M. Farina, Macromolecules, 1985, 18, 5, 928. 11. T. Hayashi, Y. Inoue, R. Chujo and Y. Doi, Polymer, 1989, 30, 9, 1714. 12. J.A. Ewen, M.J. Elder, R.L. Jones, L. Haspeslagh, J.L. Atwood, S.G. Bott and K. Robinson, Die Makromolekulare Chemie – Macromolecular Symposia, 1991, 48–49, 253. 13. J. van Schooten and J.K. Evenhuis, Polymer, 1965, 6, 11, 561. 14. J. van Schooten and J.K. Evenhuis, Polymer, 1965, 6, 7, 343. 15. F.C.Y. Wang and P.B. Smith, Analytical Chemisty, 1996, 68, 3, 425. 16. S. Enomoto, Journal of Polymer Science: Polymer Chemistry Edition, 1969, 7, 5, 1255. 17. Y. Tanaka, H. Sato and Y. Nakafutami, Polymer, 1981, 22, 12, 1721. 18. I.V. Bletsos, D.M. Hercules, D. Van Leyen and A. Benninghoven, Macromolecules, 1987, 20, 2, 407. 19. K. Makino, M. Ikeyama, Y. Takeuchi and Y. Tanaka, Polymer, 1982, 23, 3, 413. 20. J. Furukawa, K. Haga, E. Kobayashi, Y. Iseda, T. Yoshimoto and K. Sakamoto, Polymer Journal, 1971, 2, 371. 21. H. Abendroth and E. Canji, Makromolekulare Chemie, 1975, 176, 3, 775. 22. Y. Tanaka, S. Kawahara, T. Ikeda and H. Tamai, Macromolecules, 1993, 26, 19, 5253. 23. Y. Tanaka, H. Sato, Y. Nakafutami and Y. Kashiwazaki, Macromolecules, 1983, 16, 12, 1925. 24. Y. Tanaka, H. Sato and J. Adachi, Rubber Chemistry and Technology, 1986, 59, 1, 16. 25. Y. Tanaka, H. Sato and J. Adachi, Rubber Chemistry and Technology, 1987, 60, 1, 25.

206

Sequencing of Monomer Unit in Polymers 26. Y. Tanaka, Y. Nakafutami, Y. Kashiwazaka, J. Adachi and K. Tadokoro, Rubber Chemistry and Technology, 1987, 60, 2, 207. 27. Y. Tanaka in Handbook of Polymer Science and Technology, Synthesis and Properties, Volume 1. Ed., N.P. Cheremisinoff, Marcel Dekker, New York, NY, USA, 1989, p.677. 28. Y. Tanaka, H. Sato and Y. Nakafutami, Polymer, 1981, 22, 12, 1721. 29. P. Boochathum, M. Shimizu, K. Mita and Y. Tanaka, Polymer, 1993, 34, 12, 2564. 30. H.L. McMurray and V. Thornton, Analytical Chemistry, 1952, 24, 2, 318. 31. J. van Schooten, E.W. Duck and R. Berkenbosch, Polymer, 1961, 2, 357. 32. G. Bucci and T. Simonazzi, Journal of Polymer Science, Polymer Symposia, 1964, 7, 1, 203. 33. G. Bucci and T. Simonazzi, Chimica e l’Industria, 1962, 44, 262. 34. G. Natta, G. Mazzanti, A. Valvassori, G. Sartori and D. Morero, Chimica e l’Industria (Milan), 1960, 42, 2, 125. 35. T.A. Veerkamp and A. Veermans, Makromolekulare Chemie, 1961, 50, 147. 36. J. van Schooten and S. Mostert, Polymer, 1963, 4, 135. 37. G. Natta, G. Dall’Asta, G. Mazzanti and F. Ciampelli, Kolloid - Zeitschrift, 1962, 182, 50. 38. N. Sheppard and G.B.B.N. Sutherland, Nature, 1947, 159, 739. 39. F.M. Rugg, J.J. Smith and L.H. Wartman, Journal of Polymer Science, 1953, 11, 1, 1. 40. G.B.B.N. Sutherland, Discussions of the Faraday Society, 1950, 9, 274. 41. C.Y. Liang, M.R. Lytton and C.J. Boone, Journal of Polymer Science, 1961, 54, 160, 523. 42. C.Y. Liang and W.R. Watt, Journal of Polymer Science, 1961, 51, 156, S14. 43. R.R. Stromberg, S. Straus and B.G. Achhammer, Journal of Polymer Science, 1959, 35, 129, 355.

207

Introduction to Polymer Analysis 44. R.M. Briber and E.L. Thomas, Polymer, 1985, 26, 1, 8. 45. T. Doiuchi, H. Yamaguchi and Y. Minoura, European Polymer Journal, 1981, 17, 9, 961. 46. S. Kawaguchi, T. Kitane and T. Ite, Macromolecules, 1991, 24, 22, 6030. 47. F. Danusso, M.C. Tanzi, M. Levi and A. Martini, Polymer, 1990, 31, 8, 1577. 48. N. Oi, K-I. Miyazaki, K. Moriguchi and H. Shimada, Kobunshi Kagaku, English Edition, 1972, 1, 566. 49. L. Dong and D.J.T. Hill, Polymer Bulletin, 1995, 34, 3, 323. 50. G. Henrici-Olivé and S. Olivé, Advances in Polymer Science, 1979, 32, 123. 51. W.R. Krigbaum and N. Tokita, Journal of Polymer Science, 1960, 43, 142, 467. 52. C.R. Bohn, J.R. Schaefgen and W.O. Statton, Journal of Polymer Science, 1961, 55, 162, 531. 53. S. Wong and G.W. Poehlein, Journal of Applied Polymer Science, 1993, 49, 6, 991. 54. A. Aoki, T. Hayashi and T. Asakura, Macromolecules, 1992, 25, 1, 155. 55. B.D. Coleman and T.G. Fox, Journal of Chemical Physics, 1963, 38, 5, 1065. 56. H.N. Cheng, Journal of Polymer Science: Polymer Physics Edition, 1983, 21, 4, 573. 57. J.P. Montheard, A.Mesli, A. Belfkira, M. Raihane and Q-T. Phan, Macromolecular Reports, 1994, A31, Supplements 1 and 2, 1. 58. A.S. Brar and A. Sunita, Polymer, 1993, 34, 16, 3391. 59. T. Yamada, T. Okumoto, H. Ohtani and S. Tsuge, Rubber Chemistry and Technology, 1990, 63, 2, 191. 60. H. Zhuang, J.A. Gardella, Jr., and D.M. Hercules, Macromolecules, 1997, 30, 4, 1153. 61. A. Kondo, H. Ohtani, Y. Kosugi, S. Tsuge, Y. Kubo, N. Asada, H. Inaki and A. Yoshioka, Macromolecules, 1988, 21, 10, 2918

208

Sequencing of Monomer Unit in Polymers 62. M.R. Rao, T.V. Sebastian, T.S. Radhakrishnan and P.V. Ravindran, Journal of Applied Polymer Science, 1991, 42, 3, 753. 63. M.A. Golub and R.J. Gargulio, Journal of Polymer Science: Polymer Letters Edition, 1972, 10, 41. 64. D.W. Brazier and N.U. Schwartz, Journal of Applied Polymer Science, 1978, 22, 1, 113. 65. N. Grassie in Developments in Polymer Degradation, Ed., N. Grassie, Applied Science Publishers, London, UK, 1977, p.137. 66. F.C-Y. Wang and P.B. Smith, Analytical Chemistry, 1996, 68, 17, 3033. 67. Y. Ishida, H. Ohtani, K. Abe, S. Tsuge, K. Yamamoto and K. Katoh, Macromolecules, 1995, 28, 19, 6528. 68. H.W. Werner and A.E. Morgan, Journal of Applied Physics, 1976, 47, 4, 1232. 69. H.W. Werner and N. Warmoltz, Journal of Vacuum Science and Technology, A, 1984, 2, 2, 726. 70. W. Reuter, M.L. Yu, M.A. Frisch and M.B. Small, Journal of Applied Physics, 1980, 51, 2, 850. 71. D.W. Vance, Journal of Applied Physics, 1971, 42, 13, 5430. 72. G.J. Muller, Journal of Applied Physics, 1976, 47, 317. 73. J.A. Gardella and D.M. Hercules, Analytical Chemistry, 1981, 53, 12, 1879. 74. J.E. Campana, J.J. De Corpo and R.J. Colton, Journal of Applied Surface Science, 1981, 8, 3, 337. 75. D. Briggs, M.J. Hearn and B.D. Ratner, SIA Surface Interface Analysis, 1984, 6, 4, 184. 76. P.A. Zimmerman, D.M. Hercules and A. Benninghoven, Analytical Chemistry, 1993, 65, 8, 983. 77. X. Chen, J.A. Gardella, H.Tai and K.J. Wynne, Macromolecules, 1995, 28, 5, 1635. 78. J.A. Gardella, T. Ho, K.J. Wynne and H-Z. Zhuang, Journal of Colloid and Interface Science, 1995, 176, 1, 277.

209

Introduction to Polymer Analysis 79. H. Tai and K.J. Wynne, Macromolecules, 1996, 29, 11, 3991. 80. M.J. Hackathorn and M.J. Brock, Journal of Polymer Science, Polymer Chemistry Edition, 1975, 13, 4, 945. 81. A. Kawasaki in Proceedings of the 27th Autumn Meeting of the Japan Chemical Society, Pre-prints, 1972, Volume 2, p.20. 82. M.J. Hackathorn and M.K. Brock, Rubber Chemistry and Technolnology, 1972, 45, 5, 1295. 83. Y. Tanaka, K. Nunogaki and J. Adachi, Rubber Chemistry and Technology, 1988, 61, 1, 36. 84. Y. Tanaka, Y. Nakafutami, Y. Kashiwazaaki, J. Adachi and K. Tadokoro, Rubber Chemistry and Technology, 1987, 60, 2, 207.

210

6

Stereoisomerism and Tacticity

During polymerisation, it is possible to direct the way in which isotactic monomers join on to a growing chain. This means that side groups (X) may be placed randomly (‘atactic’) or symmetrically along one side of the chain (‘isotactic’) or in a regular alternating pattern along the chain (‘syndiotactic’) as discussed below. Along with chemical composition, molecular weight, molecular weight distribution, and type and amount of gel, branching is considered to be one of the fundamental parameters needed to characterise polymers fully. This latter property, which is a microstructural feature of the polymer, has very important effects on polymer properties. Changes in branching of a given polymer such as polypropylene (PP) lead to changes in its stereochemical configuration and this, in turn, is a fundamental polymer property for formulating polymer physical characteristics and mechanical behaviour. Technically, each methane carbon in a poly(1-olefin) is asymmetric:

R

H

H

C

C

C

H

H

H

methine carbon

methylene carbons

This symmetry cannot be observed because two of the attached groups are essentially equivalent for long chains. Thus, a specific polymer unit configuration can be converted into its opposite configuration by simple end-to-end rotation and subsequent translation. It is possible, to specify relative configurational differences, and Natta introduced the terms isotactic to describe adjacent units with the same configurations, and syndiotactic to describe adjacent units with opposite configurations [1]. Tacticity is defined as the ratio of syndiotactic to isotactic structure. Although originally used to describe diad configurations, isotactic now describes a polymer sequence of any number of like configurations, and syndiotactic describes any number of alternating configurations. Diad configurations are called ‘meso’ is they are alike, and ‘racemic’ if they are not [2]. Thus, from a configurational viewpoint, a poly(1-olefin) can be viewed as a copolymer of meso and reacemic diads. 211


6.1 Tacticity of Polypropylene An interesting aspect of PP chain structure is distinct configurational isomers resulting from a pseudo asymmetric carbon atom. The polymer stereogularity or tactility, as it is termed, is quite variable, being dependent on the nature of the catalyst, presence or absence of additives, and other parameters such as temperature or reaction medium. Because the polymer morphology and hence physical properties are crucially dependent on PP tacticity, measurement of this property is of considerable interest in commercial production and fundamental investigations. Three different types of PP structure are shown next: H

H

H

H

X

H

H

C

C

C

C

C

C

C

X

H

X

H

H

H

X

atactic H

H

X

H

H

H

X

C

C

C

C

C

C

C

X

H

H

H

X

H

H

syndiotactic H

H

H

H

H

H

H

C

C

C

C

C

C

C

X

H

X

H

X

H

X

isotactic

Three-dimensionally, atactic, and isotactic PP (iPP) may be represented as shown in Figure 6.1. Multiple sequences of syndiotactic or isotactic units can exist in PP. Thus, diads and triads of iPP would have the structures:

Diad

Traid

212

H

H

H

H

C

C

C

C

Me

H

Me

H

H

H

H

H

H

H

C

C

C

C

C

C

Me

H

Me

H

Me

H

Stereoisomerism and Tacticity CH3

H

C

CH3

H

C

C

H

H

CH3

C

C

H

CH3

H

C

C

H

H

H

H

(a) Atactic polypropylene CH3

H

C

CH3

H

C

C

H

C

C

H

H

CH3

H

H

CH3

H

C

C

H

H

(b) Isotactic polypropylene

Figure 6.1 Structure of atactic and isotactic polypropylene. Source: Author’s own files.

Similarly, a pentad and hexad of syndiotactic PP (sPP) would have the structures:

Pentad

Hexad

H

H

Me

H

H

H

Me

H

H

H

C

C

C

C

C

C

C

C

C

C

Me

H

H

H

Me

H

H

H

Me

H

H

H

Me

H

H

H

Me

H

H

H

Me

H

C

C

C

C

C

C

C

C

C

C

C

C

Me

H

H

H

Me

H

H

H

Me

H

H

H

As well as this, the monomer units in a polymer can exist in head-to-head, head-totail and tail-to-tail configurations as illustrated below in the case of iPP (discussed later under ‘regioisomerism’, Chapter 7):

213


Processes for the manufacture of ethylene–propylene copolymer can produce several distinct types of polymer which, although they may contain similar proportions of the two monomer units, differ appreciably in their physical properties. The differences in these properties lie not only in the ratio of the two monomers present but also, and very importantly, in the detailed microstructure of the two monomer units in the polymer molecule. Ethylene–propylene copolymers may consist of mixtures of the following types of polymer: (i) Physical mixture of ethylene homopolymer and propylene copolymer: E-E-E-E-E- P-P-P-P (ii) Copolymers in which the propylene is blocked, for example: E-E-P-P-P-P-P-E-E-E-E-E-E-P-P-P(iii) Copolymers in which the propylene is randomly distributed, for example: -E-P-E-P-E-P-E-P (alternating e.g., pure cis-1,4-polyisoprene) or -E-E-E-P-E-E-E-E-E-P-E-E-E-E (iv) Copolymers containing random (or alternating) segments together with blocks along the chains, i.e., mixtures of (iii) and (ii) (random and block) or (iii) and (ii) (alternating and block), for example: -E-P-E-P-E-P-

214

Stereoisomerism and Tacticity (v) Containing tail-to-tail propylene units in propylene blocks:

i.e., Head-to-head and tail-to-tail addition giving even-numbered sequences of methylene groups. (vi) Graft copolymers:

Although molecular symmetry is well understood, until the development of proton nuclear magnetic resonance (NMR) spectroscopy, and later 13C-NMR, a study of this aspect of polymer structure presented problems. The advantages of 13C-NMR in measurements of polymer stereochemical configuration arise primarily from a useful chemical shift range, which is approximately 20 times that obtained by proton NMR. Structural sensitivity is enhanced through well-separated resonances for different types of carbon atoms. Overlap is generally not a limiting problem. The low natural abundance (0.8) region. Burfield and Lo [21] consider the 10.28 μm band as the most appropriate reference and consequently data for this band are described, i.e., ratios A10.02/A10.28 (Figure 6.5) and A11.89/A10.28 (Figure 6.6). A further advantage of calibrations involving the 10.28 μm reference is the possibility of representation as the simple linear equation: Absorbance ratio = m(mn) + c 1.0

A998 / A973

0.8

0.6

0.4

0.2 0.3

0.5

0.7 mm

0.9

Figure 6.5 Calibration curve for IR absorbance ratio A998/A973 versus NMR triad isotacticity. (●) = high temperature annealed, (○) = hot pressed only, ($) = ambient temperature annealed. Reproduced with permission from D.R. Burfield and P.S.T. Loi, Journal of Applied Polymer Science, 1988, 36, 2, 279. ©1988, Wiley 224

Stereoisomerism and Tacticity In a direct comparison of results obtained by this method with those obtained with Luongo’s [19] widely used calibration, it is apparent that the agreement is rather poor if the percent atactic values are compared with NMR triad readings. If Luongo’s atactic and isotactic samples are assigned values of mm = 0.35 and mm = 1.00, respectively, then a good agreement is obtained: i.e., Absorption ratio = m x 0.35 + c (for aPP) and Absorption ratio = m + c (for iPP) Syndio and isotacticity studies have been conducted on PP manufactured by various processes. These included metallocene-catalysed syndiotactic PP (sPP), Zeigler–Nattacatalysed iPP, and metallocene-catalysed iPP [37, 38]. Structural differences between PP homopolymers and PP copolymers and their effect on polymer properties have been discussed [38, 39]. 1.0

A841 / A973

0.8

0.6

0.4

0.2 0.3

0.5

0.7 mm

0.9

Figure 6.6 Calibration curve for IR absorbance ratio A841/A973 versus NMR triad isotacticity. (●) = high temperature annealed, (○) = hot pressed only, ($) = ambient temperature annealed. Reproduced with permission from D.R. Burfield and P.S.T. Loi, Journal of Applied Polymer Science, 1988, 36, 2, 279. ©1988, Wiley 225


6.2 Tacticity of Syndiotactic Polystyrene (sPS) Fully sPS was discovered relatively recently, and polymers are now manufactured commercially. Only limited structural information is available [40–47]. It has four main crystalline polymorphic forms: A, B, G and D. The attractive and interesting physical characteristics of this polymer can be summarised as follows: (1) a melting temperature of 270 °C, (2) a fully trans planar zig-zag backbone (A and B forms) and (3) a solid phase transition [40, 42–45]. Because of its inherent backbone stiffness and strong intermolecular interactions, macroscopic properties such as modulus and strength are expected to exceed those of most polymers, even those of some liquid-crystalline polymers. Various crystal forms have been suggested, including a helical conformation upon crystallisation from dilute solution and an all-trans conformation [42] with annealing. The helical phase has been proposed to have a TTGG or T3GT3G1 conformation. Reynolds and co-workers [48] were interested in the nature of these crystalline forms, and the amorphous state and the transition between them. They used vibrational spectroscopy as the primary characterisation technique. Its sensitivity to local conformation and changes in chain packing allowed them to observe microstructural changes with annealing, orientation, or solvent treatments. They examined structural differences between samples of different tacticities observed from their IR spectra, and evidence is presented for two conformational forms and the transition between them caused by thermal treatment or orientation. Large differences in the IR and Raman data for atactic, isotactic and syndiotactic PS related to the different chain conformations of these isomers are observed, especially in the regions of 18.51, 10.33, 11.11, 9.34 and 8.33–7.14 μm. Reynolds and co-workers [48] also found that vibrational spectra can be perturbed significantly by thermal annealing. The spectra obtained for the sPS generally contain bands that are sharp (approximately 6 cm–1 in half-width) as compared with the relatively broad features observed for isotactic or atactic isomers. From the intensity decrease in the helical bands and the corresponding sharpening of the spectroscopic features observed upon annealing, these workers intended to show that annealed sPS is of high crystallinity, and has a planar zigzag backbone conformation. One of the primary objectives was to seek explicit evidence of vibrations that can be assigned to the all-trans planar zigzag backbone. In the 8.33–7.14 μm region, several conformation-sensitive skeletal vibrations are present. It is also quite likely that these bands are sensitive to chain packing. The IR spectrum of the cast film contains spectroscopic features that disappear when sample temperature is raised. Some of the weak features (9.27, 9.21 and 9.60 μm)

226

Stereoisomerism and Tacticity that are hard to observe at room temperature are seen quite clearly at liquid nitrogen temperature. The intensity and position of these weak features are especially sensitive to thermal annealing. One of the more interesting features observed for sPS is the 9.71 μm band. This band, assignable to the combination of CH in-plane bending, CC ring stretching, and CCC ring-bending vibrations [49], seems to be sensitive to chain packing and clearly splits into two components at 9.71 μm and 9.72 μm at low temperature. A 9.35 μm band is present as a broad feature in the cast film. However, after annealing, the band sharpens but remains as a singlet, even at low temperature. Two bands of medium intensity were observed in this region for iPS (9.50 and 9.23 μm). These bands were assigned previously to ring-backbone and ring CC stretching and to ring stretching and CH in-plane bending, respectively. In that case, they were thought to be associated with the sequence length of preferred conformations in the amorphous phase [49]. Reynolds and co-workers [48] conclude that both of these medium-intensity bands at 9.71 μm and 9.35 μm in sPS are crystalline. Atactic PS has been shown to possess a significant amount of syndiotactic trans isomers [50]. Therefore the spectrum of sPS is expected to be more similar than that of iPS to that of aPS, and this is generally observed. Reynolds and co-workers [48] assign the two bands at 10.60 μm and 10.7 μm to the helical conformation found for the cast sample and this is removed by annealing. In the 500 cm–1 region, bands are observed at 17.51, 18.25 and 18.69 μm. After annealing at 200 °C, only a single band at 18.55 μm remains. In iPS, a single band at 17.64 μm is observed and assigned to the N26b skeletal out-of-plane mode of the aromatic ring [50, 51]. The sPS spectra in this region are consistent with studies of PS model compounds in which the 18.52 μm band is observed when at least four backbone carbon atoms are in a trans conformation, whereas a band at 18.05 μm is assigned to a second conformation containing gauche isomers [51]. Thus, the 18.25 μm band of sPS is consistent with a syndiotactic all-trans structure, whereas the cast film exhibits the 18.25 μm band, indicating gauche conformers. aPS exhibits a broad band at 18.48 μm, suggesting a broad conformation distribution. Extrusion of a cast film at 100 °C produces spectral changes similar to those observed on annealing. The drawing process would also transform the helical form in the cast film to the more extended all-planar zigzag form, and this is observed. Bandwidths are broader for this oriented sample than for the annealed film, indicating that the thermal treatment produced greater structural regularity than extruding the sample to a draw ratio of 4. In conclusion, it is believed that IR spectra of sPS obtained under different crystallisation and thermal conditions are characteristic of the overall structural regularity and the specific chain conformations present. Spectra of samples cast from dilute solution are

227

Introduction to Polymer Analysis consistent with previous studies, suggesting a helical conformation. Heat treatment causes a transition to an all-trans phase. Long trans sequences can be obtained only by annealing or drawing. More recently, Kellar and co-workers [52] undertook a detailed analysis of the Raman spectrum of sPS in the region 16.67–11.76 μm. Because sPS exhibits considerable polymorphism, spectra of various preparations, including melt-crystallised sPS, solvent crystallised sPS and quenched glassy materials, were studied. Peaks were assigned to conformational changes and sequences. The N1 vibration of the phenyl ring (ring breathing mode) was shown to manifest itself through two peaks resulting from local conformational changes in the alkyl backbone. The peak centred at 12.94 μm is assigned to an all-trans backbone sequence, whereas the higher frequency feature at 12.53 μm is attributed to mixed trans/gauche transformations. Comparison with aPS is made. Study of the cross section of a compressed moulded plaque exhibiting a skin/core structure revealed the continuous way the structure varied with changing crystallinity. The various physical forms of sPS and the routes for interchange amongst them are shown diagrammatically. Analysis of the Raman data obtained by Kellar and co-workers [52] and the literature enabled them to almost completely assign all the spectral features in this region of the sPS spectrum. The assignments are given in Table 6.2, using Wilson and Hertzberg nomenclature, but only the Wilson format is used in this discussion. Six fundamental vibrational modes can be identified, N1, N6b, N10a, N4 and Bas (CH2) by Wilson nomenclature. The first two are derived from in-plane vibrations of the phenyl ring, and the rest from out-of-plane modes or backbone motions. The spectral profile of the two polymers is very similar, differing only with respect to the position and relative intensity of the peaks corresponding to the N1 vibration. It has been clearly demonstrated that the highly symmetric nature of the N1 vibration is sensitive to the conformation of the backbone and that of its immediate neighbours. In the case of PS, this is shown by two strong and highly polarised peaks. Previous work on model compounds has shown that the lower frequency vibration is due to all-trans sequences (A/B crystal polymorphs) whereas the higher frequency peak results from a mixture of trans and gauche states. The latter can therefore be attributed to the amorphous component of sPS provided its position is not shifted to values >800 cm–1, which is observed only when long ttg+g+ sequences are present, resulting in a crystalline helical structure (G/D crystal polymorphs). The lower frequency peak at 770–773 cm–1 is observed to grow and narrow into an intense feature upon crystallisation. Two N1 peaks for aPS reaffirms previous reports using other methods that there is a significant syndiotactic component present with the atactic polymer.

228


Table 6.2 Vibrational Assignments for Polystyrene and Model Compounds within 600-800 cm-1 Region of Raman Spectruma Herzberg

v18

Wilson

v6b

dimer (meso)ab

v4

v2

v2

v10b + v16b

v11

v1(tt…)

v1 (t/g…)

623

740

760

dimer (racemic)b

623

740

765 sh

trimer (isotactic)b

623

740

763

trimer (heterotactic)b

623

740

763

trimer (syndiotactic), liquidb

623

740

trimer (syndiatactic), crystalb

623

iPSb aPSb

a

v8 CHCl3

v4

v11a Bas(CH2)

v10a

780 (tg+)

844

791 (g-g-)

843

787 (tg+tg+)

843

750

781 (tg+(tt), g-t(tt), g-tg-g-)

843

763

763

789 ((tt+) g+g+)

843

740

763

763

623

740

768

623

740

762

763 (ttt) 797 ((tt) n)

sPS, trans (crystal)c

622

740

756

773

796

811

841

sPS (glass)c

622

741

757

770

798

811

841

sPS, helical (crystal)c

622

745

758

772

802

812

840

aPSc

622

741

756

769

796

813

841

670

751

845

789 ((tg+)n)

844 846

Note: dimer refers to 2,4-diphenylpentane and trimer refers to 2,4,6-triphenylheptane

b

Data from Jasse and co-workers [50]

c

Author’s data.

The continuity of these observed changes was demonstrated through analysis of a glassy skin/crystalline core sample. It has been shown that Raman spectroscopy can provide valuable information about the level of crystallinity together with the type of backbone conformation present within sPS samples [53, 54]. Study of a polished cross section by taking spectra at set intervals from skin to core underlines the power of this technique.

229

Introduction to Polymer Analysis Reynolds and Hsu [55] carried out a normal vibrational analysis of sPS, comparing their calculated results with IR and Raman spectra on drawn fully annealed samples (all-trans form). The calculated frequencies agree well with observed bands, and some features such as the 8.18 μm band are identified as being unique to the syndiotactic isomer. Large intensity changes are found for some features upon annealing and drawing, with peaks at 10.60 μm and 10.70 μm (IR), and 12.5 μm (Raman) disappearing totally with such treatments. These peaks are also absent from the calculated results of crystalline sPS (A/B form), thus providing good evidence that these features are not derived from all-trans conformations. A study was carried out by Nyquist and co-workers [56] comparing the vibrational spectra (Raman and IR) for sPS (all-trans and helical) with those of iPS, aPS and toluene. From these comparisons they were able to make partial assignments for both sPS polymorphs. Some of the assignments suggested that the crystal structure of the all-trans polymorph may not be isomorphous with C2N symmetry. These workers appear to make no distinction between the two known polymorphs which have the all-trans conformation. In an article by Kobayashi and co-workers [57], various techniques were used (including Raman and IR spectroscopy) to study the differences and transformation between the helical and all-trans forms of sPS. By comparing crystalline samples, they observed a major difference in the 14.28–17.85 μm region of the Raman spectrum. For the alltrans material, a strong sharp peak was present at 12.99 μm, together with a broad weaker feature at 12.58 μm but, in case of the helical crystal sample, the strong broad feature at 12.53 μm now dominated a much weaker peak at 13.00 μm. The sharp 12.99 μm band was attributed to the all-trans conformation and the 12.53 μm feature to gauche conformations. Various other workers [47, 58] have used IR spectroscopy in syndiotactic studies on PS. Isemura and co-workers [59] carried out stereoregularity studies on PS using pyrolysis– gas chromatography–mass spectrometry (Py-GC-MS). They detected tetramers and pentamers, and found that the minimum requirement for a disastereoisomer is the inclusion of more than two asymmetric carbon atoms in the molecule. Nonobe and co-workers [60] confirmed by the same technique with good reproducibility that tetramers and pentamer peaks reflected the original tacticity of the polymer. Other tacticity studies have been carried out on PS [61–66].

6.3 Tacticity of Polyvinyl Chloride (PVC) The study of stereochemical configuration by 13C-NMR has not been limited to the polyolefins. Schneider and co-workers [67] showed that the absorption around

230

Stereoisomerism and Tacticity 14.49 μm was proportional to the number of isotactic diads, and in the region of 16.66–15.62 μm to syndiotactic diads: H

H

H

H

H

Cl

H

Cl

C

C

C

C

C

C

C

C

H

Cl

H

Cl

H

H

H

H

isotactic diad

syndiotactic diad (h-t)

Based on this finding, they proposed a method for determining the tacticity of amorphous samples of PVC. Because some samples cannot be easily transformed into an amorphous state, Schneider and co-workers [68] devised an IR method of tacticity determination which is independent of sample crystallinity. From the temperature dependence of IR spectra of PVC samples prepared by different methods, the intensity of the band at 14.40 μm (proportional to the number of isotactic diads in the sample), as well as that of the tacticity-independent C-H stretching band, was found to be independent of sample crystallinity. These lines were applied to the tacticity determination in PVC, measured in potassium bromide pellets. The numerical tacticity value was obtained from the known values of absorbance coefficients of SCH and SHH type C-Cl stretching bands in solution, and from the shape of the spectrum. Abe and co-workers [69] investigated the NMR spectra of model compounds of PVC in the hope that these investigations may offer useful information for the analysis of vinyl polymer spectra. They studied the NMR spectra of three stereoisomers of 2,4,6-trichloroheptane as model compounds of PVC:

Me

Me

Me

H

H

H

H

H

C

C

C

C

C

Cl

H

Cl

H

Cl

H

H

H

H

Cl

C

C

C

C

C

Cl

H

Cl

H

H

H

H

Cl

H

H

C

C

C

C

C

Cl

H

H

H

Cl

Me

Me

Me

231

Introduction to Polymer Analysis Spectra were observed at 60 macrocycles/s and 100 macrocycles/s both at room temperature and at high temperatures, and spin-decoupling experiments were done. The difference in the chemical shifts of the two meso methylene protons at 60 macrocycles/s was found to be approximately 7 cycles/s for the isotactic three-unit model, whereas it was approximately 16 cycles/s for the isotactic two-unit model or heterotactic three-unit model. PVC spectra can be reasonably interpreted on the basis of this result. Observed values of vicinal coupling constants of model compounds were interpreted as the weighted means of those for several conformations, and the stable conformations of the models determined. Chemical shifts of PVC and model compounds such as meso- and racaemic-2,4dichloropentane have been measured from NMR spectra [70]:

Me

H

H

H

C

C

C

Cl

H

Cl

Me

Me

H

H

Cl

C

C

C

Cl

H

H

Me

Nakayama and co-workers [71] carried out a two-dimensional NMR characterisation study of PVC tacticity. They proposed tetrad assignments of stereosequences in PVC on the basis of the carbon-carbon connectivities revealed on the two-dimensional incredible natural abundance double quantum transfer experiment (2D-INADEQUATE) spectrum. The validity of the proposed assignments was investigated by comparing the relative peak areas observed (based on the assignments by the 2D-INADEQUATE method with those calculated by the Bernoullian propagation model). Pentad assignments were provided from the high-resolution doublet cross peaks in which the connectivities of centred methine carbons in pentads with centred methylene carbons in tetrads appear. Similarly, Bernoullian propagation statistics were used for the confirmation of the pentad assignments. It is well-known that the polymerisation of vinyl chloride proceeds under the control of the Bernoullian statistical model (selection between meso and racemo). Table 6.3 shows the comparison of the observed relative areas of methine pentad and methylene peaks with those calculated by Bernoullian propagation statistics. Relative areas of observed peaks are determined by the curve resolution method. The relative areas of observed pentad peaks agree well with calculated values, indicating the validity of their pentad assignments by the method of Dong and co-workers [61]. As for the methylene peaks, observed areas (except peak 6´) agree well with calculated values. Assuming

232

Stereoisomerism and Tacticity that the resonance of mmrmm overlaps peak 5´, the agreement of the observed and calculated areas of peak 6´ is improved. From the correlation between methine peak of rmmr and methylene peak (peak 5´) of rmmrx (x = m or r), this overlap is plausible. It is impossible for the resonances of mmrmr and rmrmr to overlap peak 5´ because this hexad should include the pentad structure, rmmr. Consequently, tactic sequence assignments of PVC are proposed (Table 6.4). In the two-dimensional spectrum of PVC (whole polymer), the peak with the assignment of mmr and mrm by 2D spinlock relay experiment should be mmmrx (x = m or r) and the overlap of rmmrx (x = m or r) and mrm, respectively.

Table 6.3 Comparison of Observed Realtive Areas of Methylene Tetrad Peaks with Those Calculated by Bernoullian Statistics Tetrad sequence

Observed

Calculateda

rrr

0.15

0.16

rmr

0.15

0.14

mrr

0.27

0.27

mmr

0.11

0.23

mrm

0.23

0.11

mmm

0.09

0.10

a

Calculated by Bernoullian propagation statistics. Source: Author’s own files

6.4 Tacticity of Poly(n-butyl methacrylate) Quinting and Cai [62] carried out high-resolution 13C-NMR and proton NMR measurements to determine the tacticity of poly(n-butyl methacrylate) (PBMA) with particular focus on the peak assignments for the n-butyl side chain. Free-radical and anionic PBMA were examined, with the former being predominantly syndiotactic and the latter isotactic. Proton NMR resonances for the n-alkyl chain of these polyacrylics show a combination of effects from configurational sensitivity and homonuclear scalar interactions. A combination of J-resolved proton NMR and proton-13C-heteronuclear correlated 2D-NMR spectra was used to characterise the long-range chemical shift effects due to tacticity.

233


Table 6.4 13C-NMR Chemical Shift Assignments of PVC in 1,2,4Trichlorobenzene at 90 oC Chemical shift (ppm)

Assignment

57.11

mrrm

57.05

mrrr

56.96 56.82

rrrr

56.47

mmrm

56.43

mmrm + mmrr

56.22

rmrm

56.04

rmrm + rmrr

55.37

mmmm

55.23

mmmr

55.10

rmmr

47.73

rrr

47.29

rmr

46.82

mrr

46.27

mmmrm + mmmrr

46.12

rmmrm + rmmrr + mmrmm

46.04

mmrmm + rmrmr

45.47 46.32

mmm

45.18 Source: Author’s own files

The n-butyl side chain resonances in the high-resolution 1H-NMR spectrum for PBMA at high (100 °C) temperature shows expected J-couplings from adjacent protons and the previously unreported chemical shift differences from monomer configurational effects. Free-radical and anionic poly(n-butyl methacrylate)s showed this effect, with the former being predominantly syndiotactic and the latter isotactic. These results suggest that high-temperature and high-resolution 13C and 1H-NMR spectra provide

234

Stereoisomerism and Tacticity rich monomer tacticity information, which should allow expanded use of proton NMR to study the tacticity of complex polymers. Figure 6.7 shows the 13C-NMR spectrum of syndiotactic PBMA (sample A) obtained at 100 °C in 1,2,4-trichlorobenzene. It had sufficiently high resolution to reveal unexpected splitting of the side-chain methylene C-2´ and methyl C-4´ peaks. Especially striking is that the relative areas of the three peaks assigned to C-4´ seemed to match the relative peak areas for the quaternary carbon: C-2. This observation leads to the suspicion that the splitting is perhaps due to previously unobserved long-range tacticity effects on chemical shift, seen not only in the carbon spectrum, but also in the proton spectrum (vide infra). The side-chain C-1´ and C-3´ peaks do not show analogous fine splitting. The carbonyl carbon C-1 region of the spectrum reveals detailed pentad chemical shift sensitivity, whereas the backbone quaternary carbon C-2 peaks show the relative areas characteristic for a triad distribution. The peaks between 52 ppm and 56 ppm show the diad and tetrad distributions for the B-methylene, C-3. One of the peaks for the methyl carbon (C-4) overlaps with the side-chain methylene C-3´ peak, and hence no information could be extracted.

4

rr

1

mr

CH3

3´

3 2 1=o

mm

o 179

177

175 173 1⸍

2⸍

4⸍ 3⸍

2´

4´

1´ 2 rr mr

3 70

65

60

55

4

mm

50

45

40 ppm

35

30

25

20

15

Figure 6.7 13C-NMR spectrum of syndiotactic PBMA. Reproduced with permission from G.R. Quinting and R. Cai, Macromolecules, 1994, 27, 22, 6301. ©1994, ACS

235

Introduction to Polymer Analysis Quinting and Cai [62] determined tacticity through analysis of the carbonyl (C-1), quaternary (C-2), side-chain methylene (C-2´), and methylene (C-4´) resonances. The 13C spectrum of isotactic PBMA obtained at 100 °C in 1,2,4-trichlorobenzene shows that the quaternary backbone carbon (C-2) resonances of isotactic polybutyl methacrylate were the only peaks intense enough for accurate curve fitting and tacticity determination. Analysis of the quaternary region gave a Pm value of 0.95. Though the methyl peak (C-4´) for the butyl group appeared at first to be a singlet, close examination revealed shoulders, tentatively attributed to the same long-range effects of tacticity observed for the syndiotactic polybutyl methacrylate sample. The splitting pattern is again analogous to that for the quaternary peak. The farthest downfield and largest of the three peaks corresponds to the mm triad, with the smaller two upfield peaks being the mr and rr peaks, respectively. Deconvolution of the C-4´ methyl carbon peaks and subsequent analysis gave a Pm value of 0.95.

1 3´ 4´ pmm

2´ 180

178

176

ppm

174

2

1´ 4 3

70

65

60

55

50

45

40 ppm

35

30

25

20

15

Figure 6.8 13C-NMR spectrum of isotactic poly-n-butyl methacrylate. Reproduced with permission from G.R. Quinting and R. Cai, Macromolecules, 1994, 27, 22, 6301. © 1994, ACS

236

Stereoisomerism and Tacticity The 1H spectrum of syndiotactic PBMA at 100 °C reveals the multiplicity for the butyl side chain resonances. These do not seem consistent with the rules for scalar couplings. For example, the methyl resonances are an apparent quartet instead of the triplet that one expects from scalar coupling to two adjacent and equivalent methylene protons. Further scrutiny reveals that the resonances for the adjacent methylene protons are an apparent heptet, even though one would expect a hextet. The case is analogous, though less obvious for the two other butyl methylene groups. Curve fitting the butyl methyl ‘quartet’ gave an area ratio for the two overlapping triplets which closely matches the rr:mr ratio as determined by analysis of the 13C spectrum. Similar ratios were obtained by analysis of the other butyl multiplets. The 1H-NMR spectrum of isotactic PBMA at 100 °C does not show the ‘quartet’ for the butyl methyl group (H-4´), which instead seems to be the expected triplet. Further scrutiny reveals smaller peaks between the peaks of the triplet. The same long-range tacticity effect exists for isotactic PBMA, but the larger triplet corresponds to the dominant mm triad sequence. The much smaller peaks correspond to the mr sequence. Ordinarily, it is possible to distinguish between mm/rr and mr/rm triad stereosequences using standard NMR experiments, but distinction between the resonances of mm and rr triads can be made only if a spectrum from a stereoregular polymer of known relative configuration is available. Triple resonance 3D-NMR techniques combined with isotopic labelling have provided powerful tools for biomolecular structure determination which have tremendous potential applications in polymer chemistry [63–66].

6.5 Identification of Diastereoisomeric Tetramers in the Pyrograms of polymethyl methacrylate Figure 6.9 shows a typical pyrogram of polymethyl methacrylate (PMMA) (S-2) at 500 °C obtained by flame ionisation detection (FID) [72]. The main peak is due to the monomer (about 96%) because PMMA is easily depolymerised at elevated temperatures [72–75]. On the pyrogram recorded with higher sensitivity, one can clearly recognise the tetramers (about 0.1%) and even the pentamers (about 0.03%) as well as the dimers and trimers. Among these fragment clusters, tetramers and the pentamers should contain at least two (m and r) and four (mm, mr, rm and rr) diastereoisomers because they have two and three asymmetric centres in the molecules, respectively. The chemical structures of the diastereoisomers in the tetramer region were estimated from electron ionisation (EI) and chemical ionisation (CI) mass

237

Introduction to Polymer Analysis spectra of the tetramers observed by Py-GC-MS in comparison with those of dimers and the trimers.

monomer

trimer dimer region

tetramer region pentamer region

0

10

20

30

40

50

60

70

retention time (min)

Figure 6.9 Pyrolysis–gas chromatography pyrogram of polymethylmethacrylate. Reproduced with permission from T.M. Wu, T.F. Yin and S.F. Hsu, Macromolecular Science B, 2004, B43, 329. © 2004, Taylor & Francis

In the corresponding CI spectra of two dimers and the trimer, we can clearly observe [M + 1]+ at m/z = 201 for both the dimers and at m/z = 301 for the trimer. The trimer structure confirmed is that shown in Figure 6.10 on the basis of the formation mechanism discussed later in Scheme 5. Also, the main dimer structures with MW = 200 should be those shown in the figure. Furthermore, according to the characteristic fragments, especially the prominent [M – OCH3]+ peak observed as the base peak in the CI mass spectra, we can estimate that the dimer (a) and the trimer should have the same terminals illustrated at the bottom of the figure.

238


Scheme 5 Formation mechanism of trimer through 1,5-radical transfer from primary macroradical

239


El

101

141

El 81

mode

mode

109

141 168

60

80 100 120 140 160 180 200 120

m/z:

60

80 100 120 140 160 180 200 120

Cl

169

Cl

115

mode

(M−OCH3)+

m/z:

mode

101 (M+1)+ 201 141

101

60

(M+1)+ 201

80 100 120 140 160 180 200 120

m/z:

60

80 100 120 140 160 180 200 120

m/z:

169 141 CH3 H3C

C C

O

CH3

CH3 C H

m/z = 200

C C

OCH3

H3C

C C

O

OCH3

C C CH2 H2 C O

OCH3

101

101

O

m/z = 200

OCH3

115 diamer (a)

diamer (b)

121 59

El

149 209

269

Cl mode

mode

(M−OCH3)+

241 (M+1)+ 301

101 285 300 50

100

150

200

250

300

50

m/z:

100

150

200

250

300

m/z:

269 CH3 H3C

C C

101

CH3

C C H2 O C

OCH3

CH3 C H O

OCH3

C C

m/z = 300 O

OCH3

trimer (b)

Figure 6.10 Mass spectra of dimers (a) and (b) and the trimer produced on pyrolysis of polymethylmethacrylate. Reproduced with permission from T.M. Wu, T.F. Yin and S.F. Hsu, Macromolecular Science B, 2004, 43, 2, 329. © 2004, Taylor & Francis 240

Stereoisomerism and Tacticity Although the expected quasi-molecular ions are not observed even in CI spectra, the common ions at m/z = 369 can be attributed to [M – OCH3]+. Thus, A and B should have the same MW (400). Furthermore, in the EI spectra, the teramer A shows a fairly strong peak at m/z = 301, whereas B exhibits a prominent peak at m/z = 315. The possible bond cleavages are shown at the bottom of the figure, together with the possible structures for the isomers. In this case, the relationship between the retention times and the position of the double bonds for the tetramers is also consistent with that for the dimers. The small satellite peaks (A´ and B´) appearing at slightly smaller retention times than those of the main tetramers (A and B) observed in the expanded pyrogram showed exactly the same mass spectra for A and A´ and for B and B´, suggesting that they are stereoisomers. The reason why the main trimer peak consists of only one component is easily explained by Scheme 2, where the trimer is exclusively formed through 1,5-radical transfer of the primary macroradical (II) at the fifth methylene carbon followed by B-scission because 1,5-radical transfer of the primary macroradical (II) at the fifth methyl carbon followed by B-scission yields only a dimer at best. When double backbiting occurs through 1,5- and then 5,9-radical transfers of the primary macroradical (II) at the fifth and the ninth methylene carbons, the pentamers consisting of only one chemical structure are formed in a similar manner as the trimer formation, although they comprise the associated diastereoisomers (Scheme 5). In the tetramers, two kinds of position isomers [A (or A´) and B (or B´)] can be formed depending on the paths of the double back-bitings following by B-scission. In path (a), the first back-biting as shown in Scheme 6 occurs at the fifth methyl carbon of the primary macroradical (II) and the second 1,5-radical transfer at the seventh methylene carbon followed by B-scission to yield tetramer A (or A´). In path (b), the first back-biting occurs at the fifth methylene carbon and the second 1,5-radical transfer at the ninth methyl carbon followed by B-scission to yield the tetramer B (or B´). In both paths, there is no chance for thermal isomerisation because the associated radical transfers occur only at methyl and methylene carbons. In the cases of the thermal degradation of PP and PS, the corresponding double back-bitings followed by B-scission to yield their tetramers occur mostly at asymmetric methine carbons, resulting in some thermal isomerisation. The diad tacticity determined from the relative peak intensities of the diastereomeric tetramers in the pyrograms was consistent with that obtained by proton NMR, suggesting that no appreciable thermal isomerisation occurred during pyrolysis. The thermal degradation mechanisms to yield the diastereomeric tetramers from PMMA without isomerisation open up the possibility of estimating the triad tacticityol PMMA from the distribution of diastereoisomeric pentamers.

241


primary macroradical

C

CH 2 C

O

O

O

C

OCH 3

CH 2 CH 2 C

CH 2 C CH 3

O

O

O

O

OCH 3

OCH 3

OCH 3

C

O

OCH 3

C

CH 2 C C

CH 3

C

OCH 3

CH 3

CH 3

CH 3

CH 2 C

CH 2 C

CH C

CH 2 C CH 3

O

O

O

O

C

OCH 3

C

C

OCH 3

OCH 3

OCH 3

CH 2

CH 3

CH 3

C

O

OCH 3

B scission CH 3

CH 3

CH 3

CH 3

CH 3

CH 2

C

CH C

CH 2 C

CH 2 C CH 3

O

C

O

O

O

OCH 3

second 1,5-backbiting radical transfer

CH 2 C

O

H

OCH 3

OCH 3

B scission CH 3

H

O

C

CH 3

CH 2 C C

C

CH 3

CH 2 C C

CH 3 CH 2 C C

b

a

CH 3

CH 3

C

OCH 3

second 1,5-backbiting radical transfer

CH 3

CH 3

CH 2 C

C

first 1,5-backbiting radical transfer

a

CH 2 C

OCH 3

CH 2 C

CH 3

CH 3

CH 3 CH 2 C

b

C

C

C

O

OCH 3

OCH 3

OCH 3

OCH 3

CH 3

CH 3

CH 3

CH 3

C

CH 2 C

CH 2 C

CH 2 C

C

O

O

O

OCH 3

OCH 3

C

OCH 3

C

CH 2 C C

CH 2

C

CH 2 C

CH 2 C

CH 2 C CH 3

O

C

O

O

O

OCH 3

OCH 3

C

OCH 3

CH 3

CH 3 O

OCH 3

CH 3

CH 2

C

CH 2 C

CH 2 C

C

O

O

OCH 3

tetramer A

OCH 3

C

O

OCH 3

CH 3

CH 3

C

C

C

OCH 3

CH 2 C O C

OCH 3

CH 3 O

OCH 3

tetramer B

Scheme 6 Formation mechanism of tetramers from primary macroradical

6.6 Tacticity of Poly(1-chloro-fluoroethylene) Li and Rinaldi [74, 75] used a 3D, 1H, 13C, 19F resonance NMR experiment to unambiguously determine the resonance assignments for mm, mr rm and rr triad stereosequences in poly(1-chloro-1-fluoro-ethylene) (PCFE) without resorting to the preparation of a stereoregular polymer with known relative configuration. In a later article, Li and Rinaldi [76] provided a complete description of the technique. The significantly better dispersion in 3D-NMR compared with 1D-NMR and 2D-NMR resolves additional signals, and makes unequivocal assignments of the 1H, 13C, and 19 F resonances from methylene groups in tetrads and fluorines in pentad sequences.

242

Stereoisomerism and Tacticity Figure 6.11 shows the 1H, 13C and 19F NMR spectra of PCFE. For a fluorinecontaining polymer with random stereochemistry, the NMR spectra have enormous complexity, arising from the various stereosequences found in the polymer as well as from 1H–1H, 19F–1H and 19F–13C couplings. In the 1H spectrum of PCFE with 19 F broad band decoupling. Even with the simplification achieved by elimination of 19 F–1H couplings, the spectrum is still too complex to interpret because of limited chemical shift dispersion. The 13C spectrum of PCFE with 1H decoupling shows two resonances which arise from the quaternary and methylene carbons (central triplet is the CDCl3 solvent peak). When 19F decoupling is applied, the doublet at 108 ppm resulting from the one-bond 13F–13C coupling collapses to a singlet; the broad peak at about 54 ppm sharpens into two groups of resonances. Tacticity has only a small influence on the appearance of the methylene resonances in the 13C spectrum, and no detectable influence on 13CF(Cl) resonance. In the 19F spectrum of PCFE, there are three groups of resonances. These resonances were originally assigned to rr, mr/rm, and mm in order of increasing field strength, but no justification for this assignment has been described. Because the 19F chemical shifts are more sensitive to structural differences than 1H or 13C chemical shifts, it is possible to obtain 1H and 13C resonance assignments using a 3D 19F/13C/19F chemical shift correlation NMR experiment which disperses signals based on the 19F chemical shifts. A 3D-NMR sequence can be adapted for this purpose. The low-resolution 1H/13C/19F 3D-NMR spectrum of PCFE is shown in Figure 6.12; f1f3 (1H-13C correlations) slices at the three different 19F chemical shifts are shown in Figure 6.13(a)–(c), and the relative positions of these slices within the 3D spectrum are schematically illustrated in Figure 6.12(d). At each 19F chemical shift, sets of crosspeaks to at least two different 13C resonances are observed, one for each germinal methylene group. Methylene carbons centred in m diads show correlations to the resonances of the two non-equivalent, directly bonded protons (e.g., A and B pairs of crosspeaks in Figure 6.12(a)). The methylene carbons centred in 4 diads are attached to 1H atoms, which are essentially chemically equivalent (although these protons are not rigorously equivalent unless the polymer is syndiotactic, remote stereochemistry has very little influence on the 1H chemical shifts, and separate resonances are not observed in these data) and therefore exhibit a correlation to a single 1H resonance (e.g., crosspeaks C and D in Figure 6.12(c)). The fact that methylene protons centred in m diads are non-equivalent was first used by Bovey and Tiers [25] to assign resonances in the 1H spectrum of polymethylmethacrylate. Later, this same characteristic was used in the interpretation of polymer 2D-NMR spectra.

243


59 59

102

EI mode

212 121

EI mode 142 249

277

301

315 341

369 341

369 50

100

150

200

250

300

350

400

101

m/z

50

100

150

200

250

300

350

400

101

CI mode

m/z

CI mode

369 369 50

100

150

200

250

300

350

400

m/z

50

100

150

200

H 3C

C C

O

OCH3 101

CH3 C C C* H2 H2 C O OCH3

300

350

400

m/z

315

301 CH3

250

369 341 CH3

OCH3 C O C* CH3

C H

C

CH3 H 3C

C

C O OCH3

tetramer A m/z = 400

C O

OCH3 101

CH3 C C C* H2 H2 C O OCH3

OCH3 C O C* CH3

C H2

369 341 CH3 C

CH2

C O OCH3

tetramer B m/z = 400

Figure 6.11 EI and CI mass spectra corresponding to the two strong peaks for tetramer A at 45 min and tetramer B at 46 min. Reproduced with permission from T.M. Yu, T.F. Yin and S.F. Hsu, Journal of Macromolecular Science B, 2004, 205, 10, 1351. Copyright symbol 2004, Taylor & Francis. Reproduced with permission from L.Li and P.L. Rinaldi, Macromolecules, 1997, 30, 520. © 1997, ACS

244


19F

d

- 96

- 97

- 98

- 99

- 100

- 102

ppm

c

13C

b

13C

110

100

90

80

70

60

ppm 1H

a

3.7

3.6

3.5

3.4

3.3

3.2

3.1

3.0

2.9

ppm

Figure 6.12 1D spectrum of poly(1-chloro-fluoro)ethylene: (a) 1H spectrum with 19 F decoupling; (b) 13C spectrum with 1H decoupling; (c) 13C spectrum with 1H and 19F decoupling; (d) 19F spectrum with 1H decoupling. Reproduced with permission from L. Li and L. Rinaldi, Macromolecules, 1997, 30, 520. © 1997, ACS

245

Introduction to Polymer Analysis In the slice at D19F = –98.2 ppm (Figure 6.12(a)) both carbon resonances from adjacent methylenes show crosspeaks to two proton resonances, therefore, this 19F must be centred in an mm triad type (type a fluorines in the structures in Figure 6.12). In the slice at D19F = –100.7 ppm (Figure 6.12(c)), both methylene carbon resonances show crosspeaks to single proton resonances; therefore, this 19F must be centred in an rr triad (type c fluorines). This the slice at D19F = –99.4 ppm (Figure 6.12(b)). One methylene carbon resonance shows a crosspeak to a single proton resonance and the second methylene carbon resonance shows a crosspeak to two proton resonaces; therefore, the fluorines having this shift must be centred in mr/rm triads. The 1H/13C/19F 3DNMR spectrum clearly shows four sets of crosspeaks from several possible tetrad structures (B, C, E and F in Figure 6.13(b)). Once the triad stereosequences are determined from examination of single slices, the relative stereochemistry of adjacent diads in the chain can be determined by looking for identical C–H crosspeaks in different 19F slices. For example, in Figure 6.12(a), the A pair of crosspeaks do not occur in the other two slices; therefore, type A methylenes show crosspeaks only to 19F atoms in mm. Although heteronuclear 3D-NMR experiments are typically carried out in conjunction with isotopic labelling, Li and Rinaldi [74] clearly demonstrate that useful data can be obtained without isotopic labelling, especially if high-abundance, NMR active isotopes such as 19F are in the molecule. By taking the advantage of the sensitivity of the 19F chemical shift to structural variations, 1H and 13C resonance assignments can be determined through a 1H-13C-19F 3D-NMR correlation experiment. This information could not be obtained from 1D- or 2D-NMR experiments. By dispersing resonance into three dimensions, it is possible to resolve numerous methylene resonances, where only a single signal is detected in the 1D-NMR spectrum. Once these resonances are resolved, the unique ability of 3D-NMR experiments to simultaneously relate the shifts of three coupled nuclei provides unequivocal assignments for the resonances of different stereosequences. While the results described here rely on the presence of 19F as the third nuclear in a fluoropolymer, similar results could be obtained from other NMR active nuclei such as 31P.

References 1. G. Natta and F. Danusso, Journal of Polymer Science, 1959, 34, 127, 3. 2. F.A. Bovey, Polymer Conformation and Configuration, Academic Press, New York, NY, USA, 1969, p.8. 3. J.C. Randall, Journal of Polymer Science: Polymer Physics Edition, 1974, 12, 4, 703.

246

Stereoisomerism and Tacticity 4. A. Zambelli, P. Locatelli, G. Bajo and F.A. Bovey, Macromolecules, 1975, 8, 5, 687. 5. F.C. Stehling and J.R. Knox, Macromolecules, 1975, 8, 5, 595. 6. J. Schaefer and D.F.S. Natusch, Macromolecules, 1972, 5, 4, 416. 7. D.E. Axelson, L. Mandelkern and G.C. Levy, Macromolecules, 1977, 10, 3, 557. 8. A. Provasoli and D.R. Ferro, Macromolecules, 1977, 10, 4, 874. 9. J.C. Randall in Carbon 13-NMR Polymer Science, Ed., W.M. Pasika, ACS Symposium Series No.103, ACS, Washington, DC, USA, 1979. 10. J.C. Randall, Journal of Polymer Science: Polymer Physics Edition, 1976, 14, 11, 2083. 11. R.S. Porter, Journal of Polymer Science: Polymer Chemistry Edition, 1966, 4, 1, 189. 12. E.M. Barrall II, R.S. Porter and J.F. Johnson in Proceedings of the ACS Division of Polymer Chemistry 148th National Meeting, Chicago, IL, USA, 1964, p.816. 13. E.M. Barrall II, R.S. Porter and J.F. Johnson, Journal of Applied Polymer Science, 1965, 9, 9, 3061. 14. S. Satoh, R. Chûjô, T. Ozeki and E. Nagai, Journal of Polymer Science, 1962, 62, 174, S101. 15. G.A. Reilly, Shell Chemical Company, Emeryville, CA, USA, private communication, 1964. 16. F.C. Stehling, Journal of Polymer Science Part A: General Papers, 1964, 2, 4, 1815. 17. M. Peraldo, Gazzetta Chemica Italiana, 1959, 89, 798. 18. M.P. McDonald and I.M. Ward, Polymer, 1961, 2, 341. 19. J.P. Luongo, Journal of Applied Polymer Science, 1960, 3, 9, 302. 20. C.Y. Liang and F.G. Pearson, Journal of Molecular Spectroscopy, 1961, 5, 1–6, 290.

247

Introduction to Polymer Analysis 21. D.R. Burfield and P.S.T. Loi, Journal of Applied Polymer Science, 1988, 36, 2, 279. 22. J. Boor, Jr., Ziegler-Natta Catalysts and Polymerisations, Academic Press, New York, NY, USA, 1979, Chapter 3. 23. D.R. Burfield and P.S.T. Lo in Catalytic Polymerization of Olefins, Eds., T. Keii and K. Soga, Kodansha, Tokyo, Japan, 1986, p.387–406. 24. Y.V. Kissin, Isospecific Polymerization of Olefins with Heterogeneous ZieglerNatta Catalysts, Springer Verlag, New York, NY, USA, 1985, p.439. 25. F.A. Bovey, High Resolution NMR of Macromolecules, Academic Press, New York, NY, USA, 1972. 26. H.J. Harwood in Preparation and Properties of Stereoregular Polymers, Eds., R.W. Lenz and F. Ciardelli, D. Riedel Publishing Company, Boston, MA, USA, 1978, Chapters 15 and 16. 27. J.L. Koenig, Chemical Microstructure of Polymer Chains, John Wiley, New York, NY, USA, 1980. 28. J.C. Randall, Polymer Sequence Determination: Carbon-13 NMR Method, Academic Press, New York, NY, USA, 1977. 29. G. Natta, P. Pino, P. Corradini, F. Danusso, E. Mantica, G. Mazzanti and G. Moraglio, Journal of the American Chemical Society, 1955, 77, 6, 1708. 30. H. Tadokoro, M. Kobayashi, M. Ukita, K. Yashufuku, S. Murahashi and T. Torii, Journal of Chemical Physics, 1965, 42, 4, 1432. 31. G. Natta, A. Valvassori, F. Ciampelli and G. Mazzanti, Journal of Polymer Science Part A: General Papers, 1965, 3, 1, 1. 32. J.J. Brader, Journal of Applied Polymer Science, 1960, 3, 9, 370. 33. R.H. Hughes, Journal of Applied Polymer Science, 1969, 13, 3, 417. 34. Y.V. Kissin, V.I. Tsvetkova and N.M. Chirkov, European Polymer Journal, 1972, 8, 4, 529. 35. Y.V. Kissin, Advances in Polymer Science, 1974, 15, 91. 36. Y.V. Kissin, V.I. Tsvetkova and N.M. Chirkov, Doklady Akademii Nauk SSSR, 1963, 152, 1162.

248

Stereoisomerism and Tacticity 37. D. Choi and J.L. White, Polymer Engineering and Science, 2004, 44, 2, 210. 38. B.P. Garcia, Revista de Plasticos Modernos, 2003, 83, 296. 39. X. Zhang, H. Chen, Z. Zhou, B. Huang, Z. Wang, M. Jiang and Y. Yang, Macromolecular Chemistry and Physics, 1994, 195, 3, 1063. 40. N. Ishihara, T. Seimiya, M. Kuramoto and M. Voi, Macromolecules, 1986, 19, 9, 2464. 41. A. Zambelli, P. Longo, C. Pellecchia and A. Grassi, Macromolecules, 1987, 20, 8, 2035. 42. M. Kobayashi, T. Nakaoki and M. Uoi, Polymer Preprints Japan, 1988, 37, 432. 43. O. Greis, T. Asano, J. Xu and J. Petermann, Zeitschrift für Kristallographie, 1988, 182, 58. 44. A. Immirzi, F. De Candia, P. Iannelli, A. Zambelli and V. Vittoria, Die Makromolekulare Chemie, Rapid Communication, 1988, 9, 11, 761. 45. V. Vittoria, F. De Candia, P. Iannelli and A. Immirzi, Die Makromolekulare Chemie, Rapid Communications, 1988, 9, 11, 765. 46. N. Ishihara, M. Kuramoto and M. Uoi, Macromolecules, 1988, 21, 12, 3356. 47. R.A. Nyquist, Applied Spectroscopy, 1989, 43, 3, 440. 48. N.M. Reynolds, J.D. Savage and S.L. Hsu, Macromolecules, 1989, 22, 6, 2867. 49. P.C. Painter and J.L. Koenig, Journal of Polymer Science: Polymer Physics Edition, 1977, 15, 11, 1885. 50. B. Jasse, R.S. Chao and J.L. Koenig, Journal of Raman Spectroscopy, 1979, 8, 5, 244. 51. B. Jasse and L. Monnerie, Journal of Molecular Structure, 1977, 39, 2, 165. 52. E.J.C. Kellar, C. Galiotis and E.H. Andrews, Macromolecules, 1996, 29, 10, 3515. 53. G. Guerra, V.M. Vitalgliano, C. De Rosa, V. Petraccone and P. Corradini, Macromolecules, 1990, 23, 5, 1539.

249

Introduction to Polymer Analysis 54. P. Corradini and G. Guerra, Advances in Polymer Science, 1992, 100, 183. 55. N.M. Reynolds and S.L. Hsu, Macromolecules, 1990, 23, 14, 3463. 56. R.A. Nyquist, C.L. Putzig, M.A. Leugers, R.D. McLachlan and B. Thill, Applied Spectroscopy, 1992, 46, 6, 981. 57. M. Kobayashi, T. Nakoaki and N. Ishihara, Macromolecules, 1989, 22, 11, 4377. 58. G. Conti, E. Santoro, L. Resconi and G. Zerbi, Mikrochimica Acta, 1988, 1, 1–6, 297. 59. T. Isemura, Y. Jitsugiri and S. Yonemori, Journal of Analytical and Applied Pyrolysis, 1995, 33, 103. 60. T. Nonobe, H. Ohtani, T. Usami, T. Mori, H. Fukumori, Y. Hirata and S. Tsuge, Journal of Analytical and Applied Pyrolysis, 1995, 33, 121. 61. L. Dong, D.J.T. Hill, J.H. O’Donnell and A.K. Whittaker, Macromolecules, 1994, 27, 7, 1830. 62. G.R. Quinting and R. Cai, Macromolecules, 1994, 27, 22, 6301. 63. C. Griesinger, O.W. Sorenson and R.R. Ernst, Journal of Magnetic Resonance, 1989, 84, 1, 14. 64. G.M. Close and A.M. Gronenborn, Progress in Nuclear Magnetic Resonance Spectroscopy, 1991, 23, 1, 43. 65. Two-Dimensional NMR Spectroscopy: Applications for Chemists and Biochemists, 2nd Edition, Eds., W.R. Crosamun and R.M.K. Carlson, VCH Publishers, New York, NY, USA, 1994. 66. J. Cavanagh, W.J. Fairbrother, A.G. Palmer, M. Rance and N. Skelton, Protein NMR Spectroscopy, Principles and Practice, Academic Press, New York, NY, USA, 1996. 67. B. Schneider, J. Storr, D. Daskoulova, M. Kolinsky, S. Sykora and D. Lim in Proceedings of the International Symposium on Macromolecular Chemistry, Prague, 1965. 68. B. Schneider, J. Štokr, M. Kolínský, M. Ryska and D. Lím, Journal of Polymer Science, Polymer Chemistry Edition, 1967, 5, 8, 2013.

250

Stereoisomerism and Tacticity 69. Y. Abe, M. Tasumi, T. Shimanouchi, S. Satoh and R. Chûjô, Journal of Polymer Science, Polymer Science Edition, 1966, 4, 6, 1413. 70. I. Ando, A. Nishioka and S. Watanabe, Polymer Journal, 1972, 3, 3, 403. 71. N. Nakayama, A. Aoki and T. Hayashi, Macromolecules, 1994, 27, 1, 63. 72. T.M. Wu, T.F. Yin and S.F. Hsu, Journal of Macromolecular Science B, 2004, B43, 2, 329. 73. O. Tarallo and V. Petraccone, Macromolecular Chemistry and Physics, 2004, 205, 10, 1351. 74. L. Li and P.L. Rinaldi, Macromolecules, 1996, 29, 13, 4808. 75. L. Li and P.L. Rinaldi, Macromolecules, 1997, 30, 3, 520.

251


252

7

Regioisomerism

As well as stereoisomerism and geometrical isomerism, polymers and copolymers can exhibit a third form of isomerism: regioisomerism. Head-to-head, head-to-tail and tail-to-head isomerism is well known for simple organic compounds. Thus, a dimer of styrene monomer can exist in the following three different regioisometric forms:

7.1 Polypropylene In the case of isotactic polypropylene, as shown next, six placements are possible when considering triads. 13C-NMR spectroscopy can be used to determine isolated head-to-head and tail-to-tail units in polypropylene. Polypropylenes produced using vanadyl catalysts possessed the normal head-to-tail structure [1]. More detailed examination shows that the amorphous fractions isolated from these polypropylenes show infrared (IR) absorption at 13.3 μm, pointing to methylene sequences of two units, which means only tail-to-tail arrangement of propylene units can occur. A

253

Introduction to Polymer Analysis very small absorption peak at 13.3 μm was also found in the spectrum of crystalline fractions. Polypropylenes prepared with catalysts based on VCl3 show only the normal head-to-tail arrangement in amorphous and crystalline fractions, as do polymers prepared from TiCl3 catalyst. The amount of propylene units coupled tail-to-tail was estimated to range from 5% to 15% for amorphous fractions, and from 1% to 5% for crystalline fractions. The amount of propylene units in tail-to-tail arrangements was calculated from spectra of thin films by comparing the ratio of the absorbances at 13.60 μm and 8.65 μm in the spectrum of hydrogenated natural rubber. This implies that the absorbance per CH2 group is the same at 13.30 μm for (CH2)2 sequences as at 13.60 μm for (CH2)3 sequences. The differences in amount of tail-to-tail coupled units between crystalline and amorphous fractions are to be expected because every head-to-head and tail-totail configuration disturbs the regularity of the isotactic chain. In polypropylenes, every tail-to-tail configuration must necessarily be accompanied by a head-to-head coupling:

254

Regioisomerism This would be expected to show up in an absorption peak at 8.8–9.0 μm, characteristic of the structure:

This is also found in hydrogenated poly-2,3-dimethylbutadiene, used as a model compound and in alternating copolymers of ethylene and butane-2 [2]. In the polypropylenes examined by van Schooten and Mostert [3] and in ethylene–propylene copolymers, they found an absorption band near 9.0 μm, although, unlike van Schooten and Mostert [3] (see above), it was much less sharp than the model compound. All spectra containing the 13.3-μm peak show a further small band at 10.9 μm, which is also found in the spectrum of poly-2,3-dimethylbutadiene. To summarise, the IR spectrum of amorphous polypropylene prepared using vanadyl catalysts in addition to normal head-to-tail structures –CHCH3- CH2-CH CH3-CH2- shows the following features:

255


*All ethylene propylene and polypropylene polymers showing an absorbance at 13.3 μm also show small absorbance at 10.9 μm, which is also found in poly 2,3dimethylbutadiene. 256

Regioisomerism

7.2 Propylene-1-Ethylene Copolymer Ethylene–propylene copolymers can contain up to four types of sequence distribution of monomeric units. These are propylene–propylene (head-to-tail and head-to-head), ethylene–propylene and ethylene to ethylene:

In addition, other stereochemical placements could occur, e.g., syndiotactic head-totail and head-to-head polypropylene units:

257


Sequences 1–4 depicted previously and their average sequence lengths of both monomer units can be measured by the Tanaka and co-workers [4] method. Measurements were made at 15.1 MHz. Assignments of signals were carried out using the method of Grant and Paul [5], and also by comparing the spectra with those of squalane, hydrogenated natural rubber, polyethylene and atactic polypropylene. The accuracy and precision of intensity measurements, i.e., deviation from the theoretical values and the scatter of the measurements, respectively, were checked for using the spectra of squalane and hydrogenated natural rubber, and were shown to be at most 12% for most of the signals. IR spectroscopy also provides information on regioisomerism in ethylene propylene copolymers. The fact that the polypropylenes prepared with VOCl3- or VO(OR)3containing catalysts show tail-to-tail arrangement means that tail-to-tail coupling of propylene units may also occur in ethylene–propylene copolymers. Because the content of (CH2)2 sequences in the copolymers is much higher than in the polypropylenes prepared with the same catalysts, a large part of these sequences probably stems from isolated ethylene units between two head-to-head oriented propylene units, their relative amount depending on the ratio of reaction rates of formation of the sequences:

258

Regioisomerism Absorption at 13.3 μm which is characteristic of methylene sequences of two units is characteristic of a tail-to-tail configuration of ethylene and propylene units. This absorption is also found in polybutene-2-ethylene copolymer:

Absorption at 9.0 μm is characteristic of a head-to-head configuration of polypropylene units, namely: H

H

H

H

C

C

C

C

H

CH3 CH3 H

This is found in amorphous ethylene–propylene alternating copolymers, polybutene2-ethylene alternating copolymer, and hydrogenated poly 2,3-dimethyl butadiene:

7.3 Polybutadiene-1-ethylene The IR spectrum of amorphous alternating polybutene-1-ethylene copolymer shows absorptions at 13.3 μm (characteristic of methylene sequences of two units) and at 9 μm (characteristic of the structure). Absorption at 10.8 μm, also found in hydrogenated poly 2,3-dimethyl-butadiene, confirms the above structure.

259


7.4 Poly-2,3-dimethyl Butadiene Hydrogenated poly-2,3-dimethyl butadiene has strong IR absorption at 9.0 μm, confirming a head-to-head configuration of two propylene units in the hydrogenated polymer:

The following structure for the unhydrogenated polymer is shown below:

7.5 Polybutadiene For dimers, many regioiosmeric configurations are possible. Polybutadiene unsaturation occurs in three forms: trans-1,4, cis-1,4 and vinyl-1,2:

260

Regioisomerism

No opportunity for regioisomerism exists in the cis and trans-1,4 configurations, but does for vinyl-1,2 unsaturation:

261


7.6 Polyisoprene Due to the methyl group, opportunities for regioisomerism exist for all four types of unsaturation present in polyisoprene:

262

Regioisomerism 1. Regioisomerism in trans-polyisoprene:

263

Introduction to Polymer Analysis 2. Regioisomerism in cis-1,4 polyisoprene:

264

Regioisomerism 3. Regioisomerism in 1,2 polyisoprene:

265

Introduction to Polymer Analysis 4. Regioisomerism in 3,4 polyisoprene:

7.7 Polypropylene Glycol Ozonisation followed by reduction with lithium aluminium hydride to oxyalkylene groups has been used to study regioisomerism in polypropylene glycols. Adopting the following nomenclature for propylene glycol:

266

Regioisomerism

then the following three sequences are possible in polypropylene glycol:

If these sequences occurred consecutively in polypropylene glycol, then it would have the structure:

267

Introduction to Polymer Analysis Upon ozonolysis, fission occurs to produce aldehydic and ketonic groups:

Upon reduction with lithium aluminium hydride, the following are produced:

From the relative amounts of these three glycols produced, determined by gas chromatography, the amounts of head-to-head, tail-to-tail and tail-to-head configurations can be deduced. The di-primary propylene glycol occurs in two optically active forms, 1(a) and (b):

268

Regioisomerism

7.7 Polyepichlorohydrin Epichlorohydrin (ECH) is a cyclic triad:

with one oxygen atom, one methane carbon atom with a chloromethylene substituent, and a methylene carbon atom. Because the methane carbon atom has four substituents, i.e., O, CH3, CH2 and CH2Cl, the monomer exists in R and S configurations, and can be resolved into pure stereoisomers [6]. The ring-opening polymerisation of ECH can produce PECH, -[-O-CH(CH2Cl)-CH2]-n, with a broad range of microstructures that depend on the mode of ring-opening promoted by the initiator and the optical purity of the monomer. In all cases, each of the backbone carbon atoms of PECH is adjacent to an oxygen atom and the chemical shifts observed in their NMR spectra are shifted accordingly. There are two carbon–oxygen (C–O) bonds in the monomer. The sequence in which the C–O bond (B or A in the diagram) above cleaves during polymerisation determines whether the resulting PECH will have a regular head-to-tail structure, a regular head-to-head, tail-to-tail structure, or a more random structure. The head is represented as the –OCH(CH2Cl)- end of the repeat unit, and the tail is represented as the –CH2-end of the repeat unit. Polymerisation of a racemic monomer in a regular head-to-tail manner potentially results in PECH with different stereochemical triads shown in Figure 7.1. In practice, only the isotactic crystalline polymer has been prepared. Lindfors and co-workers [7] studied a crystalline polyepichlorohydrin which had a regular (head-to-tail) isotactic structure with an excess, indicated by its optical activity, of –RRR– or –SSS– polymer chains. Regular RRR ... or SSS ... chains result in identical NMR spectra. This type of isomerism has been studied previously and was not studied further by Lindfors and co-workers [7]. The crystalline polymer was used

269

Introduction to Polymer Analysis as an aid for making assignments of regiosequence resonances in the NMR spectra of the amorphous cationic polymer. CH2CI H

H

H

CH2CI H

O O

O H H

H

H

CH2CI

H

Isotactic, RRR or SSS CH2CI

H

H

CH2CI H

O O

O

H CH2CI H H Syndiotactic, RSR or SRS

H

CH2CI H

H

H

H

H

CH2CI

O O

O H

H H

CH2CI

H

H

Heterotactic-1 RRS or SSR H

CH2CI

H

H

CH2CI H

O O

O H

H H

CH2CI

H

H

Heterotactic-2 SRR or RSS

Figure 7.1 Different stereochemical triads of polyepichlorohydrin (PECH). Source: Author’s own files. If during the ring-opening polymerisation both C–O bonds in ECH are subject to random cleavage, four regiosequence triads are possible for PECH (Figure 7.2). The regular H–T structural sequence results in the regiosequence triad 1. For simplicity and clarity of labelling to be used below, Lindfors and co-workers [7] focused on the central monomer unit of the triad and the two adjacent carbon atoms, and therefore called triad 1 T–H:T–H. That is, the regular H–T structural sequence leads to a T–H:T–H regiosequence triad. If during polymerisation, a single, isolated monomer reversal occurs, three additional regiosequence triads result. These are shown in Figure 7.2 as 2–4 or T–H:T–T, T–T:H–H, and H–H:T–H, respectively. The sequences were obtained

270

Regioisomerism by reversing the third (right) monomer unit for 2, the second (middle) monomer unit for 3, and the first (left) monomer unit for 4. By use of the techniques described below, NMR assignments for each of these triads were made and the concentration of each triad determined. This information permits the first calculation directly from NMR data of the percentage of reverse (H – H, T – T) units, in an irregular PECH. This two-dimensional (2D) NMR method can analyse PECH to ascertain its regiosequence distribution. It consists of taking its homonuclear spectroscopy (THCSCH) spectrum, integrating the peaks corresponding to the regiosequence triads, and calculating the dyad concentrations using first-order Markovian statistics. Assignments of proton and carbon chemical shifts for the four possible regiosequence triads were made. Proton chemical shifts have not been reported previously. From integration of the peaks in the THCSCH spectrum, the concentration of each regiosequence in a cationic PECH was calculated. When these experimental triads were tested by first-order Markovian statistics, a fit was found for a polymer with short blocks of four-to-five monomer units resulting from B-cleavage of the ECH (H-T PECH) and the rest of the monomer units from A-cleavage of the ECH (H-H, T-H, PECH). The cationic PECH was found to be ~63% H-t and ~37% H-H, T-T.

1. CH2CI

CH2CI

CH2CI

-O-CH-CH2-O-CH-CH2-O-CH-CH2T - H:T - H 2. CH2CI

CH2CI

CH2CI

-O-CH-CH2-O-CH-CH2-O-CH2-CHT - H:T - T 3. CH2CI

CH2CI CH2CI

-O-CH-CH2-O-CH2-CH - O - CH-CH2T T:H H 4. CH2CI CH2CI

CH2CI

-O-CH2-CH - O - CH-CH2-O-CH-CH2H H:T H

Figure 7.2 Four regiosequence triads for polyepichlorohydrin. Source: Author’s own files.

271

Introduction to Polymer Analysis To reach these conclusions, various 1D and 2D NMR spectra of a cationic PECH were obtained and analysed. Similar spectra of a crystalline PECH aided in the analysis and simplified assignment of some peaks. The DEPT experiment provided information for CH, CH2 and CH2Cl peak identification. 13C spectroscopy allowed basic assignment of the four regiosequence triads in cationic PECH. From 2D J-resolved experiments, the heteronuclear coupling constants (HC–H) and homonuclear coupling constants (JH–H) were determined and proton chemical shifts assigned for different regiosequences. From the THCSCH experiment, absolute proton assignments were correlated with their respective 13C assignments. Chemical shifts and coupling constants for cationic PECH derived from the proton decoupled 13C -NMR spectrum, and heteronuclear 2D J-resolved spectroscopy were in good agreement.

7.9 Other Polymers Regioisomerism is also exhibited by other polymers, including polyvinylidene fluoride, polyvinylidene chloride [8–11], polydienes and polyvinyl acetate [12, 13]. Head-to-tail and tail-to-tail sequences have been determined for polyvinylidene chloride [8–12] and polyvinyl acetate [12, 13].

Gädda and co-workers [14] used 1H-NMR, 13C-NMR, 19F-NMR and 29Si-NMR in the study of regioregularity in cyclotrisiloxane-based polymers such as phenyl(3trifluoromethylcyclotrisiloxane). They showed a high degree of regularity in which the trifluoromethyl electron-withdrawing groups enhance stereoregularity. Hugger and co-workers [15] discussed regio-random and regio-regular forms of poly(3-hexylthiophene). Wang and co-workers [16] also discussed regioregularity in regioregular poly(3-dodecylthiophen). Tonzola and co-workers [17] characterised regioregular polymers containing bis(phenylquinoline) and regioregular dialkyl bithiopene utilising 1 H-NMR Fourier transform - IR spectroscopy and thermometric techniques.

272

Regioisomerism

References 1. J. van Schooten, E.W. Duck and R. Berkenbosch, Polymer, 1961, 2, 357. 2. G. Natta, G. Dall’Asta, G. Mazzanti and F. Ciampelli, Kolloid Zeitschrift, 1962, 182, 50. 3. J. van Schooten and S. Mostert, Polymer, 1963, 4, 135. 4. Y. Tanaka and K. Hatada, Journal of Polymer Science: Polymer Chemistry Edition, 1973, 11, 8, 2057. 5. D.M. Grant and E.G. Paul, Journal of the American Chemical Society, 1964, 86, 15, 2984. 6. M.P. Dreyfuss in Proceedings of the 8th Central Regional Meeting of the American Chemical Society, Akron, OH, USA, 1976. 7. K.R. Lindfors, S. Pan and P. Dreyfuss, Macromolecules, 1993, 26, 11, 2919. 8. F.A. Bovey, F.C. Schilling, T.K. Kwei and H.L. Frisch, Macromolecules, 1977, 10, 3, 559. 9. K. Okuda, Journal of Polymer Science Part A: General Papers, 1964, 2, 4, 1749. 10. R. Chûjô, S. Satoh and E. Nagai, Journal of Polymer Science Part A: General Papers, 1964, 2, 2, 895. 11. J.L. McClanahan and S.A. Previtera, Journal of Polymer Science Part A: General Papers, 1965, 3, 11, 3919. 12. A. Abe and N. Nishioka, Kobunshi Kagaku, 1972, 29, 326, 402. 13. B. Ibrahim, A.R. Katritzky, A. Smith and D.E. Weiss, Journal of the Chemical Society, Perkin Transactions, 1974, 2, 13, 1537. 14. T.M. Gädda, A.K. Nelson and W.P. Weber, Journal of Polymer Science Part A: Polymer Chemistry, 2004, 42, 20, 5235. 15. S. Hugger, R. Thomann, T. Heinzel and T. Thurn-Albrecht, Colloid and Polymer Science, 2004, 282, 8, 932. 16. W. Wang, K.C. Toh and C.W. Tjiu, Macromolecular Chemistry and Physics, 2004, 205, 9, 1269.

273

Introduction to Polymer Analysis 17. C.J. Tonzola, M.M. Alam, B.A. Bean and S.A. Jenekhe, Macromolecules, 2004, 37, 10, 3554.

274

8

Determination of End Groups

Recent trends in multi- and/or higher functionalisation of polymeric materials require precise characterisation not only of the main structures, but also of microstructure. The chain-end characterisation of polymers is regarded as one of the most important and challenging subjects in polymer characterisation. It is known that initiators and chain transfer reagents are often incorporated into the polymer as chain ends, and these features can cause significant changes in polymer properties. This type of information often provides an extremely important clue to polymerisation mechanisms. Characterisation of the end groups in polymer samples with large molecular weight is extremely difficult because of their very low concentration compared with the main chain. Because end groups in polymers are generally attributed to an initiator and/or chain transfer and terminating agent incorporated into polymer chains, analysis of end groups is one of the most substantial approaches for assessing the mechanism of polymerisation. Presence or absence of specific end groups often causes significant changes in the polymer properties, and thus precise characterisation has been eagerly sought in recent multifunctionalisation of polymeric materials. The characterisation of end groups in a high molecular weight polymer sample is not an easy task. Also, available methods are not always adequate for quantitative analysis of end groups in high molecular weight polymers. The characterisation of polymer chain ends gives valuable clues to clarify the polymerisation mechanisms, and to design new polymers with improved properties [1–7]. The recent advent of various analytical techniques has made it possible to carry out practical studies of the end groups of polymers. The radioactive isotope labelling method has been used over a long period to determine the initiator fragment incorporated at chain ends by measuring the specific activities of radioactive samples prepared with 14C-labelled initiator [4, 8]. In recent years, high-field nuclear magnetic resonance (NMR) techniques have been successfully applied to study polymer chain ends [1–7, 9, 10]. Several NMR studies have looked at the end groups of various polymers prepared with the initiators isotopically enriched with NMR-active nuclei such as 13C, 2H, 19F, and 15N [2, 5, 6, 11, 12]. Hatada and

275

Introduction to Polymer Analysis co-workers polymerised totally deuterated methyl methacrylate (MMA) monomer with nondeuterated initiator to determine the content of the initiator fragments incorporated in the polymer chain by 1H-NMR, and discussed the mechanism of polymerisation in detail [13–16]. Owing to recent developments in highly specific pyrolysis devices, highly efficient separation columns for gas chromatography (GC), and specific identification of the peaks in the pyrograms by GC–mass spectrometry (GC–MS), and pyrolysis–GC (Py–GC), structural characterisation of polymeric materials has become a reality. Particularly, Py–GC is an extremely sensitive, simple and rapid technique.

8.1 Polypropylene Oxide Heatley and co-workers [17] and others [18-27] described methods for the qualitative and quantitative characterisation of saturated and unsaturated end groups in anionically polymerised polypropylene oxide (PPO) using 1H- and 13C-NMR at temperatures between 0 °C and 80 °C using different types and concentrations of initiator. The main features of the mechanism of the anionic polymerisation of PPO have been well established by 13C-NMR [28, 29]. Propagation proceeds principally by epoxide ring opening via anionic attack at the secondary ring carbon, giving regular head-totail monomer enchainment:

The stereochemistry of the polymer from optically inactive monomer is random, and the chains terminate in secondary alcohol groups. About 2.5% of the additions occur by attack at the tertiary carbon [29], giving a small proportion of head-to-head and tail-to-tail monomer placements. It has also been established [30] that the proton abstraction reaction occurs, leading to the initiation of a new chain via the allyl alcoholate, as well as continued growth of the original chain because alcohol and alkoxide groups are rapidly equilibrated:

276

Determination of End groups

Because new chains are initiated throughout the reaction, the abstraction reaction leads to a broadened molecular weight distribution and a lower average molecular weight than would be expected from the ratio of monomer to initial alkoxide concentrations. The allyl ether may undergo base-catalysed isomerisation to a propenyl ether [30]:

Below are two examples of end groups as they occur in polymers:

8.2 Polyvinyl chloride (PVC)

The end groups of an hydroxyl and a double bond, are a structural feature of the polymer and it may be important to identify and determine these groups. Various techniques have been used to determine end groups. These methods must be highly sensitive because, particularly with high molecular weight polymers, the percentage of end groups is very low. Thus, if 1 mole (62 g) of diethylene glycol is reacted with 20 moles (880 g) of ethylene oxide according to the following equation, then the molecular weight of the product is 62 + 880 = 942: HO–CH2CH2OH + 20CH2CH2O = HOCH2CH2(CH2CH2O)9O CH2CH2O(CH2CH2O)9CH2–CH2OH

Thus, two hydroxy end groups (34 g) occur per 997 g of polymer, i.e., 3.4% hydroxy group. Similarly, if 100 moles of ethylene oxide reacted with 1 mol of diethylene

277

Introduction to Polymer Analysis glycol, then the final product would have a hydroxy end-group content of 0.76%. One of the uses to which end-group analysis can be put is determination of molecular weight. Thus, if a polyethylene glycol–ethylene oxide condensate was found to contain 0.3% hydroxy end groups:

Percentage hydroxyl = 2 × 17/62 + n44 = 0.3 i.e., n = 3400 – 186/132 = 255 Molecular weight of the polymer is: HO(OCH2CH2)n/2–1OCH2CH2O(CH2CH2O)n/2–1 OH = HO(OCH2CH2)126.5OCH2CH2O(CH2CH2O)126.5OH = 11,226 Using a matrix-assisted laser desorption/ionisation (MALDI) technique with an ionsource Fourier transform (FT) mass spectrometer, van Rooji and co-workers [25] carried out high-resolution end-group analysis of polyethylene glycols (PEG). Jackson and co-workers [26] used MALDI combined with collision induced dissociation (CID) using a time-of-flight instrument to achieve a similar analysis.

8.3 Polystyrene (PS) 8.3.1 NMR Spectroscopy 1

H-NMR has been applied to the end-group analysis of PS formed by utilising the totally deuterated monomer technique [27, 31]. In addition, 13C-NMR has been used to characterise phenolic end groups in PS prepared by cationic polymerisation in the presence of alkylphenols [32]. The initiator-derived residues in PS prepared by using 13 C-labelled initiators have been identified and quantified by 13C-NMR, even in high molecular weight polymers [33–37]. This technique was also successfully used to evaluate the role of initiator-derived functionalities in PS for thermal degradation [36–38].

278

Determination of End groups Ito and co-workers [39] applied 13C-NMR to the determination of end groups in PS polymerised anionically with n-butyl lithium as the initiator. Polymers with molecular weights between 1000 and several million were included in this study. The 13C-NMR spectrum of the PS sample immediately after preparation and again after heating at 100 °C for 595 hours was compared. The intense resonances from main-chain carbon nuclei at D z 18 (CH3), 73 (CH2) and 76 (CH) were unaffected by heating the polymer, but there were significant changes among the minor peaks. In addition to OCH3 and CHOH carbon resonances from initiator and secondary alcohol end groups, respectively, the spectrum of the sample immediately after preparation shows resonances from an olefinic CH2 carbon at D z 116.4, an olefinic CH carbon at D z 134.7 and a CH2O carbon at D z 71.9, all of equal intensity. These three resonances are consistent with assignment to an allyl ether end-group [26]. On heating, these peaks disappear and are quantitatively replaced by two olefinic CH carbon peaks at D z 100.4 and 145.8, and a CH3 carbon peak at D z 9.0 attributable to a propenyl ether end-group [17]. Smaller peaks of unknown origin also appeared at D z 8, 29 and 65. 1

H-NMR spectra of the same PS samples show main-chain protons at D = 0.9 (CH3) and 3.1–3.5 (CHO + CH2O). Olefinic protons of the allyl group were evident in the spectrum of the sample immediately after isolation as multiplets in the region D = 4.3 to 5.8. Using the labelling scheme:

together with the fact that trans three-bond H–H spin-spin coupling constants are larger than cis [24], the chemical shifts and coupling constants were assigned as DA = 4.99, DM = 5.08, Dx – 5.70, JM = 1.5 Hz, JAX = 10.5 Hz and JMX = 17.5 Hz. CH2O protons resonated at D = 3.82, with a doublet splitting of 5.5 Hz from coupling to Hx. After heating, propenyl olefinic protons appeared at D = 4.21 (CH3CH=) and 5.81 (=CHO–), with a mutual coupling constant of 6.2 Hz, whereas propenyl methyl protons appeared at D = 1.42 with a coupling constant of 6.5 Hz to the adjacent olefinic proton. Although the propenyl group may exist in cis or trans forms, only one set of propenyl peaks appeared in 1H and 13C spectra, indicating that only one stereoisomer was produced. From the magnitude of the olefinic coupling constant [24], it was concluded that it was the cis isomer which was formed. The relative time scales of the polymerisation stage (2 hours) and the isomerisation stage (595 hours) indicate that the isomerisation reaction is two orders of magnitude slower than the polymerisation reaction. 279

Introduction to Polymer Analysis In CDCl3 solvent, the propenyl peaks were observed only in freshly prepared solutions. After storage at room temperature for two days, the peaks disappeared due to hydrolysis by trace acid impurities, producing propanal and a – CH2OH endgroup. The 1H-NMR spectrum of polystyrene shows two further small peaks of equal intensity: a sharp triplet at D = 3.00 and a broad multiplet at D = 3.72. On comparison of the spectra of samples of different molecular weight, the relative intensities of these peaks were found to vary systematically with molecular weight, and the peaks were therefore associated with end groups. To determine the origin of these peaks, a series of 13C spectra were recorded with low-power continuous-wave 1H decoupling at various frequencies in the region D = 3 to 4 in the proton spectrum. These spectra showed that placing the decoupler frequency on the small proton resonance at D = 3.72 gave maximum decoupling of the terminal CH(CH3)OH carbon resonances, and this peak was therefore assigned to the CHOH proton. 1H– 1H homonuclear spin-decoupling experiments then showed that the triplet at D = 3.00 arose from one of the non-equivalent protons in the adjacent CH2 group. Both peaks are sufficiently well resolved at 300 MHz to be of use in the quantitative characterisation of chain length.

8.3.2 Pyrolysis – Gas Chromatogarphy (Py-GC) This technique has been used to study PS [40]. Pyrolysis of PS derived from n-butyl and four from n-butyl groups and a polystyrene unit is shown in Table 8.1. Based on these results, Mn values could be estimated using relative molar intensities of these nine characteristic peaks (i = 1–9) against those of the major peaks (i = 1–22). In this case, minor peaks other than i = 1–22 were ignored because the total relative intensity of these peaks was less than a few percent. Because the degree of polymerisation (Dp) of the PS sample is defined as the number of styrene units per end group, Dp can be calculated by the following equation:

280

Determination of End groups where Ii is the intensity of peak, i in the pyrogram of a PS sample having one n-butyl end group, Mi is the number of styrene units in the i component, and ni is the molar sensitivity correction factor of the i component for flame ionisation detector (FID) response, i.e., the effective carbon number of the i component. Of the various pyrolysis products produced, one in particular, peak 8 (C4H9- CH2C(Ph) = CH2) (2-phenyl heptanes) was the most characteristic of the nine products mentioned previously. It gave a linear relationship between Mn and the relative intensity of this peak to the total intensity of all 22 peaks in the programme. Li and Rinaldi [41] used three-dimensional (3D) NMR to determine the chain-end structure of 13C-labelled PS.

8.2.2 Dye Partition Methods Ghosh and co-workers [42] carried out end-group analysis of persulfate-initiated PS using a dye partition and a dye interaction technique. Sulfate and hydroxyl end groups are usually found to be incorporated in the polymer to an average total of 1.5 to 2.5 end groups per polymer chain.

Table 8.1 Peak assignment in the pyrogram of polystyrene Structure of pyrolysates

Styrene unit (m)

Carbon number (C)

Aliphatic C=C bond (u)

Effective carbon number (n)

1

CH3CH==CH2

0

3

1

2.9

2

CH3CH2CH3

0

3

0

3.0

3

CH3CH2CH==CH2

0

4

1

3.9

4

CH3CH2CH2CH3

0

4

0

4.0

5

CH3CH2CH2CH2CH3

0

5

0

5.0

6

CH3CH2C(Ph)==CH2

1

10

1

9.9

7

C4H9─CH2CH2 (Ph)

1

12

0

12.0

8

C4H9─CH2C(Ph)==CH2

1

13

1

12.9

9

C4H9─CH==C(Ph)CH3

1

13

1

12.9

Reproduced with permission from H. Ohtani, S. Ueda, Y. Tsukahara, C. Watanabe and S. Tsuge, Journal of Analytical Applied Pyrolysis, 1993, 25, 1. © 1993, Elsevier

281

Introduction to Polymer Analysis Banthia and co-workers [43] determined sulfate, sulfonate and iso-thioronium salt end groups in PS by the dye-partition technique. Polymer polarity did not affect the results of end-group determination. Nitrile groups incorporated in PS by initiation or copolymerisation have been detected and estimated by dye-partition techniques after reduction to amino groups with lithium aluminium hydride in tetrahydrofuran [44].

8.4 Polyethylene (PE) High-density polyethylene (HDPE) prepared with Ziegler-based catalysts have predominantly n-alkyl or saturated end groups. Those prepared with chromium-based catalysts have a propensity towards more olefinic end groups. The ratio of olefinic to saturated end groups for PE prepared with chromium-based catalysts is approximately unity. End-group distribution is therefore another structural feature of interest in lowpressure polyolefins because it can be related to the type of catalyst used, and possibly to the extent of long-chain branching. It is possible not only to measure by 13 C-NMR concentrations of saturated end groups, but also of the olefinic end groups and subsequently an end-group distribution. Perez and van der Hart [45] described a 13C-NMR method for the determination of chain-ends and branches in crystalline and non-crystalline regions in PE. A high proportion of these units resided in the crystalline phase. Hammond and co-workers [46] conducted an IR study of bond rupture of carbonyl and vinyl groups formed during plastic deformation of HDPE. The relationship between end-group concentration and the time lapse between deformation and spectroscopic examination was investigated. There is no significant time dependence for vinyl group concentration, whereas the carbonyl group concentration shows a slight increase with time. The carbonyl peak centred at 1742 cm–1 reaches a maximum intensity after 48–72 hours. The vinyl out-of-phase deformation band at 909 cm–1 was used to measure vinyl end groups. Figure 8.1 shows the time dependence of this peak for two randomly selected draw ratios. The effect of temperature was examined over the range 7–60 °C; the lower limit is set by the increasing brittleness of the polymer as it is cooled, with fracture occurring before yield is reached. The results, presented in Figure 8.2, show that the concentration of vinyl and carbonyl groups, for a given draw ratio, decreases with increasing temperature.

282


Conc. (1018 groups/cm3

5 4

X

X

X

DR = 9.5

X

3X DR = 6

2 1 0 0

72

48

24

120

96

Time (h)

Conc. (1018 groups/cm3)

Figure 8.1 Relationship between concentration of (a) carbonyl groups, (b) vinyl end groups in polyethylene and the time lapse between deformation and spectroscopic examination by infrared spectroscopy. Reproduced with permission from C.L. Hammond, P.J. Hendra, P.G. Latore, W.F. Maddams and H.A. Willis, Polymer, 1988, 29, 49. © 1988, Elsevier

3 2 1

Carbonyl 1742 cm-1

Vinyl, 909 cm-1

0 0

10

20

30

40

50

60

Temperature (° C)

Figure 8.2 Effect of temperature during deformation showing decrease of vinyl and carbonyl concentrations with increase in temperature. Reproduced with permission from C.L. Hammond, P.J. Hendra, P.G. Latore, W.F. Maddams and H.A. Willis, Polymer, 1988, 29, 49. © 1988, Elsevier

8.5 Polyethylene Terephthalate Van Houwelingen [47] used coulometric bromination to determine vinyl ester end groups in polyethylene terephthalate (PET) formed by thermal chain scission:

283


The constant-current generation of bromine is carried out in a medium of dichloroacetic acid, water, potassium bromide and mercury(II)chloride. To this medium an amount of the polymer, previously dissolved in hexafluoroisopropanol and diluted with anhydrous dichloroacetic acid, is added and bromine generated. The end of the reaction is detected biamperometrically. The suitability of this method was tested against methyl vinyl terephthalate:

Additions of 14.2 μmol and 1.0 μmol of methyl vinyl terephthalate (corresponding to 30 mmol and 2 mmol of vinyl ester end groups per kilogram of polymer) were recovered quantitatively (recoveries of 99.8% and 98.5%, respectively). Nissen and co-workers [48] described an ultraviolet (UV) spectroscopic method for carboxyl end groups in PET. Hydrazinolysis led to formation of terephthalomonohydrazide from carboxylated terephthalyl residues to provide a selective analysis for carboxyl groups via UV absorbance at 240 nm. Other techniques that have been applied to end-group analysis of PET include 1 H-NMR, gel permeation chromatography–MALDI-time-of-flight (ToF) mass spectrometry [49] and MALDI-ToF combined with collision-induced dissociation (CID) [26].

8.6 Polyisobutylene (PIB) In their analysis of 1H-NMR spectra of various end-functionalised PIB, Jiaoshi and Kennedy [50] covered inductive effects (due to tert-chlorine-ended polyisobutylenes), magnetically anisotropic effects (due to olefin groups and phenyl rings) and allylic coupling (due to olefinic end groups):

284

Determination of End groups I

Chlorine-ended PIB

II Olefinic-ended PIB

III Hydrogenated PIB

8.6.1 Tert-chlorine Terminated PIB The – CH3 and – CH2– proton region of the 1H-NMR spectrum of chlorine-ended polyisobutylene (Structure I) show that chlorines exert a strong electron-withdrawing effect on neighbouring – CH3 and – CH2– groups, and shift the resonances of these groups downfield. The proton resonances of ‘normal’ –CH3 and –CH2– groups in PIB appear at 1.09 ppm and 1.40 ppm, respectively, whereas those of the first isobutylene unit adjacent to the tert-Cl appear at 1.67 ppm and 1.95 ppm (i.e., show a downshift of 0.57 ppm and 0.55 ppm), and those of the second isobutylene unit from the tertCl appear at 1.16 ppm and 1.46 ppm (i.e., with a downshift of 0.07 ppm and 0.06 ppm). With decreasing molecular weights (down to Mn = 480), resonances of the CH3– and – CH2– groups of the third isobutylene unit separate with downfield shifts of 0.01 ppm and 0.01 ppm, respectively, with regard to their normal position. That indicates that the inductive effect due to the chlorine end-group can be transmitted up to 6 S-bonds along the PIB chain.

285


8.6.2 Olefin-terminated PIB 8.6.2.1 Anisotropic Effect The 1H-NMR spectrum of an olefin-ended PIB (Structure II - a bond) shows that ‘normal’ – CH3 and – CH2– groups of PIB appear essentially at the same position as in Structure I previously (i.e., at 1.09 ppm and 1.40 ppm). Similarly, the first isobutylene unit adjacent to the double bond shows the – CH3 and – CH2 – proton resonances at 1.77 ppm and 1.98 ppm (i.e., with down shifts of 0.68 ppm and 0.58 ppm), respectively. In contrast to the first isobutylene unit, the – CH3 and – CH2 – groups of the second isobutylene unit away from the exo double bond appear at 1.01 ppm and 1.36 ppm, i.e., show upfield shifts of 0.08 ppm and 0.04 ppm, respectively. This type of shifting (i.e., the first isobutylene unit downfield, the second isobutylene unit upfield), is probably due to shielding anisotropy because of unsaturation [51, 52]. After hydroboration and oxidation, the unsaturation disappears and the resonances of the first and second isobutylene units shift back to their usual positions.

8.6.3 Hydroxy-terminated PIB In the 1H-NMR spectrum of hydroxyl-terminated PIB (Structure III) the eight peak pattern in the 3.22–3.48 ppm range is clearly due to the AB part of a typical ABX system. This splitting pattern is most likely due to prochirality: the two protons (g1 and g2) in the – CH2OH group are magnetically non-equivalent, as they are part of a prochiral – CH2– group:

Monnatt and co-workers [53] observed 1H-NMR signals for olefinic end groups – CH2C(CH3) = CH2, CH = C(Me)2, and – CH2C– (CH2) CH2– in high molecular weight PIB.

286


8.7 Polymethylmethacrylate 8.7.1 Py–GC Between 1989 and 1997, Ohtani and co-workers [40, 54–60] published a series of articles on the application of Py–GC to the determination of end groups in polymethylmethacrylate (PMMA). In earlier work, Ohtani and co-workers [54] identified end groups in PMMA by high-resolution Py–GC. Minor peaks in the chromatogram were associated with end groups derived from benzoyl peroxide polymerisation initiator or dodecane thiol chain transfer agent reactions. End-group data were related to molecular weight data. Ohtani and co-workers [58] used Py–GC at 700 °C to determine the end groups in PMMA macromonomers and their prepolymers which had been synthesised radically in the presence of azobis(isobutyronitrile) (AIBN) as initiator and mercaptoacetic acid (MCA) or mercaptopropionic acid (MPA) as chain transfer reagent. Because one of the end groups in most of the PMMA examined in this study should have a sulfur atom or a cyano group, the Py–GC system is equipped with a simultaneous multidetection system. A FID was always used in conjunction with a sulfur-selective flame photometric detector (FPD) or a nitrogen-phosphorus detector (NPD). In the method developed by Ohtani and co-workers [58], simultaneous-multidetection systems were quantitatively applied to the analysis of end groups. The simultaneous pyrograms of PMMA taken by FID and NPD in the presence of benzothiophene as an internal standard are used for the selective determination of the sulfur-containing chain ends. The simultaneous pyrograms taken by FID and NPD are interpreted in terms of the AIBN residues incorporated into the polymer chains. They compared the results observed by simultaneous multidetection Py–GC with those estimated by size exclusion chromatography (SEC) and kinetic data for the polymerisation. According to the mechanism of the radical polymerisation which is initiated by AIBN followed by the chain transfer reactions with MCA or MPA, most of the resulting prepolymers should be terminated by the corresponding carboxylic residues as follows:

287


Judging from the big differences in chain-transfer constants, chain transfer reactions with AIBN, monomer, and solvent (benzene) can be regarded negligible in the polymerisation in the presence of TGA or MPA. However, the following polymer having the terminal AIBN residue should also be formed depending on the relative feed of AIBN:

Recombination of disproportionation termination reactions may yield other polymers having different combinations of terminals. Purified pre-polymer-As were converted to the corresponding macromonomers by the reaction of the end carboxylic groups with glycidyl methacrylate in xylene at 140 °C for 6 hours in the presence of a small amount of hydroquinone and N,Ndimethyllaurylamine as follows:

288


A typical pyrogram of a prepolymer prepared in the presence of TGA or MPA as a chain transfer reaction is illustrated in Figure 8.3, where benzothiophene was used as a common internal standard for FPD and FID. PMMA has a tendency to depolymerise mostly into the MMA monomer at elevated temperatures. Therefore, the MMA monomer is the main pyrolysate (>70% of the total peak intensities except for the internal standard) on the pyrograms observed by FID). Sulfur-containing products characteristic of the TGA or MPA-chain-end residues are difficult to detect via FID because even the outstanding peaks observed at retention time up to 10 minutes are mostly assigned to hydrocarbons by GC–MS. Only sulfur-containing products formed from the end-group moiety of the TGA or MPA residues, along with the internal standard, were detected on pyrograms observed by FPD. The main component of the MMA monomer is not observed at all because it contains no sulfur atoms. Therefore, to quantify the yields of sulfur-containing compounds on the FPD pyrogram relative to the MMA-related products on the FID pyrogram, benzothiophene was used as the correlating internal standard because the peak due to this compound was observed by FPD and FID detectors. The pyrograms observed by FID for the prepolymers synthesised in the presence of TGA and MPA were almost identical. When comparing the corresponding pyrograms detected by FPD, a fairly strong peak due to CH3– CH2SH was characteristic of the various prepolymers examined.

289


COS CH SH 3 A HS CH3SCH3 2 CS2 Benzothiophone

(internal standard) S

B

0

10

20

30

40

50 min

Figure 8.3 GC Pyrogram at 700 oC of polymethylmethacrylate (PMMA) prepolymers synthesised in the presence of (a) TGA or (b) MPA chain transfer agents. Reproduced with permission from H. Ohtani, Y.F. Luo, Y. Nakashima, Y. Tsukahara and S. Tsuge, Analytical Chemistry, 1994, 66, 1438. © 1994, ACS

Table 8.2 summarises the characteristic pyrolysates that were observed on the pyrograms and identified by GC–MS. They are the MMA backbone-related products (peaks 1–24) and the AIBN-related ones (peaks B–K) which are also observed in the NPD pyrograms. AIBN-related peak D was used to normalise the MMA backbone related peaks because it was observed as a moderately strong, isolated peak. In summary, these results demonstrated that pyrolysis with simultaneous multidetection GC is an effective technique for chain-end analysis of PMMA macromonomers and their prepolymers synthesised via radical polymerisation. In this method, minute amounts of heteroatom-containing end groups in PMMA are determined using the ratios between heteroatom-containing fragments and backbone MMA-related products, which are simultaneously detected by the heteroatom-selective detector and by FID, respectively. An appropriate internal standard is used to correlate the simultaneously observed pyrograms.

290


Table 8.2 Assignment of peaks reflecting chain-end AIBN residue in the pyrogram of PMMA Peak code

Molecular weight

B

67

C

69

D

123

E

125

F

136

G

155

H

169

I

169

J

167

K

183

Structure

Reproduced with permission from H. Ohtani, Y.F. Luo, Y. Nakashima, Y. Tsukahara and S. Tsuge, Analytical Chemistry, 1994, 66, 9, 1438. © 1994, ACS

291


Table 8.3 Assignment of peaks concerned with the initiator observed in the pyrograms MMA units

End-group units a

ECNb

Source

A

0

0.5

4.0

End moiety

B

0

0.5

3.9

End moiety and main chain

C

0

1.0

8.0

Remaining initiator fragment

D

0

1.0

7.9

Remaining initiator fragment

Ec

0

1.0

8.0

Initiator

F

1

0.5

7.75

End moiety

G

1

0.5

7.65

End moiety

Hc

0

1.5

12.0

Initiator

I

0

2.0

16.0

Initiator

Peak

Chemical structure

Reproduced with permission from H. Ohtani, Y. Ito, S. Tsuge, S. Wakabayashi, T. Kawamura and J. Atarashi, Macromolecules, 1996, 29, 13, 4516. © 1996, ACS

292


Table 8.4 Assignment of common peaks concerned with the main chain of PMMA

a

Peak

Molecular weight

Chemical structure

MMA units

ECNa

1

88

CH3CH2COOCH3

1

2.75

2

102

CH3CH(CH3)COOCH3

1

3.75

3

100

CH2==C(CH3)COOCH3 (monomer)

1

3.65

4

116

C6H12O2

1

4.75

b

5

92

PhCH3

0

7.00

6

116

C6H12O2

1

4.75

7

114

C6H10O2

1

4.65

8

114

C6H10O2

1

4.65

9

114

C6H10O2

1

4.65

10

104

PhCH==CH2b

0

7.90

11

142

C8H14O2

2

6.65

12

140

C8H12O2

2

6.65

13

156

C8H16O2

2

7.65

14

140

C8H12O2

2

6.65

15

140

C8H12O2

2

6.65

16

158

C9H18O2

2

7.65

17

158

CH3OCOCH==CHCH2COOCH3

2

4.40

18

186

C9H14O4

2

6.40

19

200

C10H16O4

2

7.40

20

186

C9H14O6

2

6.40

21

200

C10H16O4

2

7.40

22

200

C10H16O4

2

7.40

23

200

C10H16O4

2

7.40

24

214

C11H18O4

2

8.40

25

214

C11H18O4

2

8.40

PhCH2CH==C(CH3)COOCH3

1

10.65

C15H24O6 (trimer)

3

11.25

26

190

27

300

b

b

ECN = effective carbon number. These peaks are only characteristics of PMMA initiated by benzoyl peroxide Reproduced with permission from H. Ohtani, Y. Ito, S. Tsuge, S. Wakabayashi, T. Kawamura and J. Atarashi, Macromolecules, 1996, 29, 13, 4516. © 1996, ACS

293

Introduction to Polymer Analysis In further work, Ohtani and co-workers [59] characterised branched alkyl end groups of PMMA polymerised radically with 2,2´-azobis(2,4,4-trimethylpentane)(ABTMP) or benzoyl peroxide (BPO) as an initiator by Py–GC. On the resulting pyrogram at 540°C, characteristic products formed from the end-group moiety due to the initiator, such as isobutane, isobutene, and so on, were clearly separated from those from the main chain. Then number-average molecular weight (Mn) of PMMA was determined by the ratio of the relative intensity of these peaks due to the end group and the main chain. After simple correction using a reference PMMA sample having different end groups, Mn values estimated by Py–GC agreed well with those obtained by SEC. Determination of end groups well supported the assumption that disproportionation was dominant in termination in this polymerisation system of MMA. Figure 8.4 shows pyrograms of (a) PMMA (Mn = 2.78 × 103) initiated by ABTMP and (b) the reference PMMA sample initiated by BPO. Because PMMA has a tendency to depolymerise mostly into the original monomer at elevated temperatures, the main pyrolysis product is MMA monomer (ca 95%). In addition, dimer peaks and a trimer peak are observed in the pyrograms. Many minor components are also observed as well-separated peaks. Among these, the five peaks A, C, D, F and G are observed only in pyrogram a (Figure 8.4). This result indicates that these components arise from the end moiety originating from ABTMP initiator. Additionally, peak 6 in Figure 8.4(a) should be mostly attributed to one of the initiator related products, although it is also observed in the pyrogram of the BPO initiator reference polymer shown in Figure 8.4(b) even with weaker intensity. The relative intensity of peak B on the pyrogram on a polymer sample of Mn = 2.78 × 103 proved to be much larger than that of a similarly initiated polymer of much lower molecular weight (Mn = 4 × 105). These peaks observed in Figure 8.4(a) (ABTMP initiator) identified by Py–GC–MS are summarised in Table 8.3, whereas the assignment of the other peaks observed in Figure 8.4(a) and 8.4(b) is shown in Table 8.4. The peaks listed in Table 8.3 can be mostly attributed to the initiator-related fragments formed from the end-group moiety of PMMA obtained with ABTMP. Number-average molecular weights measured by this Py–GC method and SEC are in fairly good agreement. This strongly supports the validity of the assumption that termination mechanisms occur exclusively by disproportionation. Ohtani and co-workers [60] determined by Py–GC end groups in anionically polymerised standard PMMA in the range of Mn of 20,000 to 1,300,000 with narrow molecular weight distributions. The characteristic fragments reflecting the end groups on the pyrogram of the PMMA were identified by comparison with those of a radically polymerised PMMA, together with the mass spectra of the characteristic peaks on the resulting pyrograms taken by a GC–MS spectrometer system. Concentrations of the end groups determined from their relative peak intensities were interpreted in terms of

294

Determination of End groups Mn, and then compared with their reference values from the manufacturers and those estimated by proton NMR. By this method, direct determination of the end groups was possible even for PMMA with Mn about 1,000,000 without using standard polymer samples.

(a)

MMA

B

D X

A

G

C

Dimers

X

Trimer

F

MMA 3

(b)

Dimers Trimer 21~25 19

B 1 2 4 0

10

20

18 5~8 10 15 13 16 20 11 14 12 17 9 30

40

27

26

50 60 70 Retention Time (min)

Figure 8.4 GC Pyrogram of PMMA (a) Mn = 2.78 × 103 initiated by ABTMP and (b) reference PMMA sample initiated by BPO. Reproduced with permission from H. Ohtani, Y.F. Luo, Y. Nakashima, Y. Tsukahara and S. Tsuge, Macromolecules, 1996, 29, 4516. © 1996, ACS

295

Introduction to Polymer Analysis The anionically polymerised standard PMMA and the radically polymerised PMMA with BPO as an initiator have common backbone structures but have different endgroup structures. Therefore, by comparing the pyrograms for the two types of PMMA, the characteristic peaks reflecting the end groups in the standard PMMA can be estimated. Because PMMA has a tendency to depolymerise, the main pyrolysis product on the pyrograms (about 95%) is the MMA monomer. Various minor products, including MMA dimers and trimers, are also observed. The peaks of products larger than trimers were negligibly small even under high-temperature GC conditions with the oven temperature up to about 400 °C. This fact suggests that almost all the PMMA chains are reflected in the observed pyrograms as the peaks up to trimers. From detailed comparison of these pyrograms, the characteristic peaks (A–G; Table 8.5), were observed only in the pyrogram of the polymer with Mn = 20,200. Among the pyrograms for polymers with an Mn in the range of 20,200 to 1,330,000, the relative intensities of peaks A–G monotonously decreased with the rise of Mn. These data suggest that the characteristic peaks (Table 8.5) originate from the anionic initiator residue incorporated into the polymer chain. The structures of these peaks reflecting the chain end were identified mostly by GC–MS. From the observed mass spectra, the relatively abundant peaks of A and B can be assigned as cumene and A-methylstyrene, respectively. These data strongly suggest that these standard polymers contained a cumyl end-group at one end of the polymer. Estimated Mn values by Py–GC and 1H-NMR are compared with reference values from the manufacturer in Table 8.6. Data from Py–GC and 1H-NMR are in fairly good agreement with the reference values for lower molecular weight samples (S–I to S–IV). Considering the signal-to-noise (S/N) ratio for the observed 1H-NMR spectra under the given spectral condition of 500 scans with a 2-second pulse delay, estimation of Mn was limited to 1000 g resin/g-eq epoxy at concentration levels of meq/l. The 4532 cm–1 combination tone is reasonably free of interferences, and can be employed to measure oxirane ring concentrations for epoxy-coating resin systems during synthesis and crosslinking. With the use of low S/N FTIR supported by computer data manipulation, chloroform solutions of five commercially available resins were analysed for epoxide-equivalent weight and correlated with results obtained by perchloric acid titrations. The near-IR technique displays linearity for epoxy concentrations of 3.6–20.7 meq/l. Similar results were obtained via a serial concentration study, indicating that the technique is not strongly affected by matrix effects. A comparison of terminal epoxide determinations by the near-IR method and by standard perchloric acid titration showed that the near-IR method is much less affected by interference by solvents and reagents than titration.

8.9 Poly(2,6-dimethyl 1,4, phenylene oxide) 8.9.1 NMR spectroscopy 13

C-NMR has been proven to be very useful in identifying repeat unit 1, end groups 2 and 3, and units 5–9 in this polymer (Figure 8.6). Due to the insensitivity of the 13 C nucleus for NMR studies and the low concentration of most of these units, it is still impractical to use it as a routine and reliable analytical tool to determine the concentration of these units.

301


O

O

OR n

2

1

O

O

3

OH

Bu

N 4

O

Bu

O

O Bu

OH

N

5

Bu

6

O

O

O

OH

7 8

O

O

OH

O

O

OH

OH

O

OH

O

OH

20 9

21

Figure 8.6 End unit structures of poly(2,6-dimethyl-1,4-phenylene oxide) Because most of the trace structural units in PPO resin contain labile phenolic hydroxyl groups, Chan and co-workers [74] thought that these functional groups could serve as a ‘handle’ for attaching a more sensitive NMR nucleus which would give stronger and well-separated signals for the different structural units. The particular nucleus

302

Determination of End groups that they have studied was phosphorus, because it is 377 times more sensitive than carbon [75]. Phosphorus has a large range of chemical shift, ~700 ppm, which ensures a good separation of signals of the 31P nuclei in different environments. It is well-known that derivatisation of a phenolic group with phosphorus halides is quantitative and rapid. This is very significant for ensuring reliable quantitative results. There are no interfering phosphorus atoms within the PPO resin, which simplifies the assignment of the spectra. Diphenyl chlorophosphate has been used previously to derivatise PPO hydroxyl groups, with some success in differentiating the various hydroxyl groups using 31P-NMR. Brevand and Granger [75] carried out a quantitative analysis on three samples which can be identified as follows: (A) PPO resin, (B) PPO and polystyrene alloy blend (1:1, w/w; PPO/PS), (C) PPO and high-impact polystyrene alloy blend (1:1, w/w; PPO/HIPS). Results are summarised in Table 8.7. On the basis of this technique, they obtained the hydroxyl concentration of normal phenolic ends 2, phenolic ends 4, and phenolic groups on the backbone 8 or 9, as well as the number-average molecular weight of the polymers. Hydroxy contents obtained by this method agree well with those obtained by IR spectroscopy.

Table 8.7 Results of 31P quantitative analyses obtained for three PPO resins using 1,/32 dioxaphosphatamylchloride (A) PPO resina

(B) PPO/PSa

(C) PPO/HIPSa

(1) % of OH on normal phenolic end

0.087

0.045

0.042

(2) % of OH on phenolic end 4

0.039

Not detected

Not detected

(3) % of OH on backbone phenolic group

Not detected

0.011

0.010

(4) % of total OH groups

0.126

0.056

0.052

(5) % N on phenol end

0.034

Not detected

Not detected

(6) Mnb

13 400

18 700

20 200

a

Percentage values are expressed as w/w. b Mn of the PPO in the blend is based on the %wt of PPO. Reproduced with permission from C. Brevand and P. Granger, Handbook of High Resolution Multinuclear NMR, John Wiley & Sons, New York, NY, USA, 1981, p.102. © 1981, John Wiley and Sons

303


8.10 Miscellaneous End Groups van Rooji and co-workers [25] carried out MALDI on an external ion source Fourier transform ion cyclotron resonance mass spectrometer equipped with a 7-T superconducting magnet to analyse the end groups of synthetic polymers in the mass range 500–5000 μm. Native, perdeutero methylated, propylated, and acetylated polyethylene glycol and polyvinyl pyrrolidone with unknown end-group elemental composition were investigated in the mass range up to 5000 μm using a 2,5-dihydroxybenzoic acid matrix. A small electrospray setup was used for the sample deposition. Two methods to process data were evaluated for the determination of end groups from the measured masses of the component molecules in the molecular weight ranges: a regression method and an averaging method. The latter is demonstrated to allow end-group mass determinations with an accuracy within 3 μm for the molecular weight range 500–1400 and within 20 μm for the molecular weight range from 3400 to 5000. This is sufficient to identify the elemental composition of end groups in unknown polymer samples. Mori [76] characterised the end groups of polyethylene sebacate as a function of molecular size by derivatisation of the hydroxyl and carboxyl groups with 3,5 dinitro benzoyl chloride and o-(p-nitrobenzyl)N,N´-diidopropyliso urea, respectively, and analyses of the derivatised polymers by SEC with IR detection. Methods for the determination of end groups in other polymers are reviewed in Table 8.8 [77-96].

Table 8.8 Methods for determination of end groups in polymers Polymer or copolymer

End group

Analysis method

Comment

Carboxy- and hydroxyl- terminated polbutadiene

-

Infrared spectroscopy

Carboxyl hydroxy equivalent weights

Styrene–acrylonitrile polymers

Acrylonitrile


Acrylonitrile determined

Terminal hydroxy


Hydroxy nitrile

Miscellaneous

Near-infrared

-

Polyvinylchloride

Unsaturated end groups

Spectroscopy FTNMR

End group-contained allylic groups

Polyvinylfluoride

CH2CH2F

19

F-NMR

-

Phenylglycidyl

End groups

13

C-NMR

-

Natural rubber

Vinyl

13

-

Butadiene-isoprene Miscellaneous

304

C-NMR


13

Poly 2,6-dimethyl-1,4 phenylene ether

End groups

Polyacrylamide

End groups

MALDI-ToF-MS

-

Polyacrylonitrile

End groups

Mass spectrometry

-

Poly (styrene sulfonic acid)

End groups

MALDI-ToF mass spectrometry

-

Terminal hydroxy

Spectrophotometry of ceric ammonium nitrate complex

-

Hydroxy end group

Ozonisation size-exclusion chromatography

-

End group

Size-exclusion chromatography

Effect of end group and solvent in size-exclusion elution

Carboxy

Potentiometric titration using alcoholic potassium hydroxide

-

-

-

-

End group

MALDI

End-group sequence distribution

Polycaplonates

Hydroxy-terminated polybutadiene Digoxyethylene

Poly-m-phenylene isophthalamine

Review of methods of end-group analysis Polyfluoride polyethers

C-NMR

-


References 1. J.C. Bevington, J.R. Ebdon, T.N. Huckerby and N.W.E. Hutton, Polymer, 1982, 23, 2, 163. 2. J.C. Bevington, T.N. Huckerby and N.W.E. Hutton, European Polymer Journal, 1982, 18, 11, 963. 3. B.K. Hunter, E. Redler, K.E. Russell, W.G. Schnarr and S.L. Thompson, Journal of Polymer Science: Polymer Chemistry Edition, 1983, 21, 2, 435.

305

Introduction to Polymer Analysis 4. J.C. Bevington, J.R. Ebdon and T.N. Huckerby, European Polymer Journal, 1985, 21, 8, 685. 5. P. Locatelli, M. C. Sacchi, I. Tritto and F. Forlini, Macromolecules, 1990, 23, 8, 2406. 6. K. Hatada, K. Ute and M. Kashiyama, Polymer Journal, 1990, 22, 10, 853. 7. J.C. Bevington, J.R. Ebdon and T.N. Huckerby in NMR Spectrometry of Polymers, Ed., R.N. Ibbett, Blackie Academic & Professional, Glasgow, UK, 1993, p.80. 8. C.A. Barson, J.C. Bevington and S.W. Breuer, European Polymer Journal, 1989, 25, 3, 259. 9. D.E. Axelson and K.E. Russell, Progress in Polymer Science, 1985, 11, 3, 221. 10. T. Kashiwagi, A. Inaba, J.E. Brown, K. Hatada, T. Kitayama and E. Masuda, Macromolecules, 1986, 19, 8, 2160. 11. G. Moad, E. Rizzardo, D.H. Solomon, S.R. Johns and R.I. Willing, Makromolekulare Chemie, Rapid Commununications, 1984, 5, 12, 793. 12. J. Devaux, D. Daoust, R. Legras, J.M. Dereppe and E. Nield, Polymer, 1989, 30, 1, 161. 13. K. Hatada, T. Kitayama, K. Fujikawa, K. Ohta and H. Yuki, Polymer Bulletin, 1978, 1, 2, 103. 14. K. Hatada, T. Kitayama and H. Yuki, Makromolekulare Chemie, Rapid Communications, 1980, 1, 1, 51. 15. K. Ute, T. Kitayama and K. Hatada, Polymer Journal, 1986, 18, 3, 249. 16. K. Hatada, T. Kitayama and E. Masuda, Polymer Journal, 1986, 18, 5, 395. 17. F. Heatley, J.F. Ding, G. Yu and C. Booth, Macromolekulare Chemie, Rapid Communications, 1993, 14, 12, 819. 28. F.C. Schilling and A.E. Tonelli, Macromolecules, 1986, 19, 5, 1337. 29. F. Heatley, Y. Luo, J. Ding, R.H. Mobbs and C. Booth, Macromolecules, 1988, 21, 9, 2713. 30. S.D. Gagnon in Encyclopedia of Polymer Science and Engineering, 2nd Edition, Volume 6, Eds., H.F. Mark and J.I. Kroschwitz, Wiley Interscience, 306

Determination of End groups New York, NY, USA, 1987, p.273. 18. J. Plucinski, R. Janik and H. Malyschok, Przemyśl Chemiczny, 1981, 60, 210. 19. F. Heatley, G. Yu, W. Sun, E.J. Pywell, R.H. Mobbs and C. Booth, European Polymer Journal, 1990, 26, 5, 583. 20. F. Heatley, G. Yu, M.D. Draper and C. Booth, European Polymer Journal, 1991, 27, 6, 471. 21. F. Heatley, G. Yu, C. Booth and T.G. Blease, European Polymer Journal, 1991, 27, 7, 573. 22. S. Dickson, G. Yu, F. Heatley and C. Booth, European Polymer Journal, 1993, 29, 2-3, 281. 23. E. Breitmeier and W. Voelter, Carbon 13 NMR Spectroscopy, 3rd Edition, VCH Publishers, New York, NY, USA, 1987, p.192. 24. R.K. Harris, Nuclear Magnetic Resonance Spectroscopy, Pitman, London UK, 1982, p.221. 25. Van Rooji, M.C. Duursma, R.M.A. Heeren, J.J. Boon and C.G. de Koster, Journal of the American Chemical Society for Mass Spectrometry, 1996, 7, 5, 449. 26. A.T. Jackson, H.T. Yates, J.H. Scrivens, G. Critchley, J. Brown, M.R. Green and R.H. Bateman, Rapid Communications in Mass Spectrometry, 1996, 10, 13, 1668. 27. K. Hatada, T. Kitayama and H. Yuki, Polymer Bulletin, 1980, 2, 1, 15. 31. K. Hatada, T. Kitayama and E. Masuda, Polymer Journal, 1985, 17, 8, 985. 32. B. Wesslén and K.B. Wesslén, Journal of Polymer Science, Part A: Polymer Chemistry, 1989, 27, 3915. 33. G. Moad, D.H. Solomon, S.R. Johns and R.I. Willing, Macromolecules, 1982, 15, 4, 1188. 34. G. Moad, D.H. Solomon, S.R. Johns and R.I. Willing, Macromolecules, 1984, 17, 5, 1094. 35. A. Zambelli, P. Longo, C. Pellecchia and A. Grassi, Macromolecules, 1987, 20, 8, 2035.

307

Introduction to Polymer Analysis 36. G. Moad, D.H. Solmon and R.I. Willing, Macromolecules, 1988, 21, 3, 855. 37. J. Krstina, G. Moad and D.H. Solomon, European Polymer Journal, 1989, 25, 7-8, 767. 38. A.K. Banthia, B.M. Mandal and S.R. Palit, Journal of Polymer Science, Polymer Chemistry Edition, A1, 1977, 15, 4, 945. 39. Y. Ito, H.O. Ohtani, S. Ueda, Y. Nakashima and S. Tsuge, Journal of Polymer Science, Polymer Chemistry Edition, 1994, 32, 2, 383. 40. H. Ohtani, S. Ueda, Y. Tsukahara, C. Watanabe and S. Tsuge, Journal of Analytical Applied Pyrolysis, 1993, 25, 1. 41. L. Li and P.L. Rinaldi, Macromolecules, 1997, 30, 3, 520. 42. P. Ghosh, S.C. Chadha and J. Palit, Polymer Science Part A: General Papers, 1964, 2, 10, 4441. 43. A.K. Banthia, R.M. Mandell and S.R. Palit, Journal of Polymer Science, 1977, 15, 4, 945. 44. C. Kanjilal, B.C. Mitra and S.R. Palit, Makromolekular Chemie, 1977, 178, 6, 1707. 45. E. Perez and D.L. van der Hart, Journal of Polymer Science, Part B: Polymer Physics Edition, 1987, 25, 8, 1637. 46. C.L. Hammond, P.J. Hendra, P.G. Latore, W.F. Maddams and H.A. Willis, Polymer, 1988, 29, 1, 49. 47. G. D.B. van Houwelingen, Analyst, 1981, 106, 1057. 48. D. Nissen, V. Rossbach and H. Zahn, Journal of Applied Polymer Science, 1974, 18, 7, 1953. 49. D. Tillier, H. Lefebvre, M. Tessier, J.C. Blais and A. Fradet, Macromolecular Chemistry and Physics, 2004, 205, 5, 581. 50. J. Si and J.P. Kennedy, Journal of Polymer Science, Polymer Chemistry Edition, 1994, 32, 11, 2011. 51. T.I. Atta-ur-Rahman, Nuclear Magnetic Resonance: Basic Principles, SpringerVerlag, New York, NY, USA, 1986, p.18. 52. D.E. Leyden and R.H. Cox, Analytical Applications of NMR, John Wiley & 308

Determination of End groups Sons, New York, NY, USA, 1977, p.115–185. 53. S.L. Mannatt, J.D. Ingham and J.A. Miller, Organic Magnetic Resonance, 1977, 10, 198. 54. H. Ohtani, S. Ishiguro, M. Tanaka and S. Tsuge, Polymer Journal, 1989, 21, 1, 41. 55. H. Ohtani, M. Tanaka and S. Tsuge, Journal of Analytical Applied Pyrolysis, 1989, 15, 167. 56. H. Ohtani, M. Tanaka and S. Tsuge, Bulletin of the Chemical Society of Japan, 1990, 63, 4, 1196. 57. Y. Tsukahara, Y. Nakanishi, Y. Yamashita, H. Ohtani, Y. Nakashima, Y.F. Luo, T. Ando and S. Tsuge, Macromolecules, 1991, 24, 9, 2493. 58. H. Ohtani, Y.F. Luo, Y. Nakashima, Y. Tsukahara and S. Tsuge, Analytical Chemistry, 1994, 66, 9, 1438. 59. H. Ohtani, Y. Ito, S. Tsuge, S. Wakabayashi, T. Kawamura and J. Atarashi, Macromolecules, 1996, 29, 13, 4516. 60. H. Ohtani, Y. Takehana and S. Tsuge, Macromolecules, 1997, 30, 9, 2542. 61. A.T. Jackson, H.T. Yates, J.H. Scrivens, M.R. Green and R.H. Bateman, Journal of the American Society of Mass Spectrometry, 1997, 8, 12, 1206. 62. S.R. Palit and P. Ghosh, Microchemical Journal Symposium Series, 1961, 2, 663. 63. P.D. Bartlett and K. Nozaki, Journal of Polymer Science, 1948, 3, 2, 216. 64. P. Ghosh, A.R. Mukherjee and S.R. Palit, Journal of Polymer Science Part A: General Papers, 1964, 2, 6, 2807. 65. S. Maiti and M.K. Saha, Journal of Polymer Science, Polymer Chemistry Edition, 1967, 5, 1, 151. 66. S.R. Palit, Makromolekulare Chemie, 1960, 36, 1, 89. 67. S.R. Palit, Makromolekulare Chemie, 1960, 38, 1, 96. 68. S. Maiti, A. Ghosh and M.K. Saha, Nature, 1966, 210, 513. 69. P. Ghosh, P.K. Sengupta and A. Pramanik, Journal of Polymer Science Part A:

309

Introduction to Polymer Analysis General Papers, 1965, 3, 5, 1725. 70. M.K. Saha, P. Ghosh and S.R. Palit, Journal of Polymer Science Part A: General Papers, 1964, 2, 3, 1365. 71. R.F. Goddu and D.A. Delker, Analytical Chemistry, 1958, 30, 12, 2013. 72. H. Dannenburg, SPE Transactions, 1963, 3, 78. 73. M.C.P. Peck, R.O. Carter and S.B.A. Quaderi, Journal of Applied Polymer Science, 1987, 33, 1, 77. 74. K.P. Chan, D.S. Argyropoulos, D.M. White, G.W. Yeager and A.S. Hay, Macromolecules, 1994, 27, 22, 6371. 75. C. Brevand and P. Granger, Handbook of High Resolution Multinuclear NMR, John Wiley & Sons, New York, NY, USA, 1981, p.102. 76. S. Mori, Analytical Chemistry, 1986, 58, 17. 77. R.D. Law, Polymer Science, Polymer Chemistry Edition, 1971, 9, 3, 589. 78. Ch. Tsvetanov, I. Panayotov and B. Erussalimsky, European Polymer Journal, 1974, 10, 7, 557. 79. V.I. Valnev, R.A. Shiykher, T.S. Gunitrieva and I.B. Tsvetovskii, Zhurnal Analiticheskoi Khimii, 1975, 30, 1236. 80. R.B. Roy and C. Kradjel, Journal of Polymer Science: Polymer Chemistry Edition, 1988, 26, 7, 1733. 81. E.C. Bezdedea, Journal of Polymer Science: Physics Edition, 1977, 15, 4, 611. 82. D.W. Ovenall and R.E. Uschold, Macromolecules, 1991, 24, 11, 3235. 83. J.C. Ronda, A. Serra, A. Mantecon and V. Cadiz, Macromolecular Chemistry and Physics, 1994, 195, 10, 3445. 84. D.J. Patterson and J.L. Koenig, Applied Spectroscopy, 1987, 41, 3, 441. 85. T. Usami, F. Keitoku, H. Ohtani and S. Tsuge, Polymer, 1992, 33, 14, 3024. 86. A. Favier, C. Ladaviere, M.T. Charreyre and C. Pichot, Macromolecules, 2004, 37, 6, 2026. 87. A.R. Dolan and T.D. Wood, Synthetic Metals, 2004, 143, 2, 243.

310

Determination of End groups 88. P.O. Danis, D.E. Karr, F. Mayer, A. Holle and C.H. Watson, Organic Mass Spectrometry, 1992, 27, 843. 89. M.H. Motorina, L.G. Kalinina and E.I. Metalkina, Plasticheskie Massy, 1973, 6, 74. 90. Y. Ishida, S. Kawaguchi, Y. Ito, S. Tsuge and H. Ohtani, Journal of Analytical Applied Pyrolysis, 1997, 40-41, 321. 91. H. Ohtani, Y. Ito, H. Ogasawara, C. Kawaguchi and S. Sage in Proceedings of the International Symposium on Chromatography, 35th Annual Research Group, Hanei Tedetors World Scientific, Singapore, 1985, p.813–820. 92. M. Ramarao, K.J. Scariah, P.V. Ravindran, G. Chandrasekharan, S. Alwan and K. Sastri, Journal of Applied Polymer Science, 1993, 49, 3, 435. 93. J.R. Craven, H. Tyrer, S.P.L. Li, C. Booth and D. Jackson, Journal of Chromatography, 1987, 387, 233. 94. L.N. Kreshbov, L.N. Shvelsovr and E.A. Emelin, Soviet Plastics, 1968, 10, 53. [Chemical Abstracts, 1969, 70, 21345q] 95. G. Montaudo, Trends in Polymer Science, 1996, 4, 3, 81. 96. M.S. de Vries and H.E. Hunziker, Applied Surface Science 1996, 106, 466.

311


312

9

Types of Unsaturation

9.1 Unsaturation in Homopolymers 9.1.1 Polybutadiene Unsaturation 9.1.1.1 Infrared spectrometry Infrared (IR) spectroscopy is a very useful technique for the measurement of different types of unsaturation in polymers. Polybutadiene (PBD) has the following structure: Its polymers can contain the following types of unsaturation:

Fraga [1] developed an IR–near-infrared method of analysis of carbon tetrachloride solutions of polybutadienes suitable for the evaluation of cis-1,4 at 5000–714.2 cm–1 (2–14 μm), trans-1,4 at 9708 cm–1 (10.3 μm) and vinyl at 9091 cm–1 (11.0 μm). Only polybutadiene is required for calibration. The method is applicable to carbon

313

Introduction to Polymer Analysis tetrachloride-soluble polybutadienes containing 0–97% cis-1,4 structure, 0–70% trans-1,4 structure, and 0–90 % vinyl structure. Some typical spectra are shown in Figure 9.1 for a high cis-1,4 polybutadiene, and high trans-1,4 polybutadiene, and a high 1,2 (atactic) polybutadiene, and other samples of different cis and trans compositions.

1

2

778 cm-1 3 1070 cm-1

1075 cm-1

4

Figure 9.1 Infrared spectra of (1) cis-1,4 PBD, (2) trans-1,4 PBD (3) mixed PBD structure (4) atactic 1,2 PBD. All samples in CS2, 2 cm light path cell with NaCl window. Source: Author’s own files

Figure 9.2 shows IR spectra of various kinds of polybutadienes [2, 3], which illustrate the usefulness of IR spectroscopy for distinguishing between different types of unsaturation.

314


Transmittance

100

cis 1, 4 polybutadiene 100

trans 1, 4 polybutadiene

Transmittance

100

1, 2 polybutadiene 100

Emulsion polybutadiene

0

2

3

4

5

6

7

8 9 10 11 Wavelength (μ)

12

13

14

15

Figure 9.2 Infrared spectra of various kinds of PBD. Source: Author’s own files

Fraga [1] also described an IR thin-film area method for the analysis of styrene–butadiene copolymers. The integrated absorption area between 1515 cm–1 and 1389 cm–1 (6.6 μm and 7.2 μm) has been found to be essentially proportional to total bound butadiene, and is independent of the isomeric-type butadiene structure present. This method can be calibrated for bound styrene contents ranging from 25% to 100%. IR [4–8] and pyrolysis–IR [9] methods have been tried for determination of the composition of vulcanisates, but both methods have serious disadvantages.

315

Introduction to Polymer Analysis Albert [10] compared determinations of butadiene in high-impact polystyrene by an IR method and by the iodine monochloride method described by Crompton and Reid [11]. The IR method is based on a characteristic absorbance in the IR spectrum associated with the transconfiguration in polybutadiene:

Because different grades of high-impact polystyrene may contain elastomers with different trans-butadiene contents, calibration curves based on the standard rubber are not always suitable for analysing these products. The results obtained by the two methods for several high-impact grades are compared in Table 9.1. The rubber content of high-impact polystyrene sample 1 determined by titration is lower than the value obtained by the IR method. This is expected for interpolymerised polymers because of crosslinking, which reduces the unsaturation of the rubber. The other polymers (except sample 3), appear to contain diene 55 type rubber of high trans-butadiene content because reasonable agreement was obtained between the iodine monochloride and IR methods. High-impact polystyrene 3 must contain a polybutadiene of high cis content to explain the low (1.2 wt%) amount of rubber found by the IR method compared with the 9.0 wt% found by the titration method.

Table 9.1 Rubber content of high-impact polystyrenes (based on PBD) Sample

Polybutadiene by the iodine monochloride method (wt%)

Polybutadiene by an IR method (wt%)

Standard: 6.0 wt% diene 55

6.2

-


12.2

-


14.8

-

High-impact polystyrene 1

8.6

9.7


5.6

5.8


9.0

1.2


11.2

11.4


5.8

5.9


316

Types of Unsaturation More recent studies include the use of near-IR spectroscopy to determine cis 1,4, trans 1,4 and 1.2 butadiene units in polybutadiene and styrene butadiene copolymers [12] and Fourier transform Raman spectroscopy to determine cis 1,4, trans 1,4 and vinyl 1,2 contents of polybutadienes [13–15].

9.1.1.2 Nuclear Magnetic Resonance (NMR) Spectroscopy Various workers [16–18] have used the aliphatic 13C resonances [16, 17] (assuming Bernoullian statistics) to give indirect information on the relative abundance of the three base units. The first method (i.e., measurement of 1,2 vinyl/cis 1,4 and trans 1,4) necessitates two independent measurements (1H/13C-NMR); the second method for cis 1,4/trans 1,4 (13C-NMR) is severely hampered by questionable assignments [19, 20] and different compositionally induced diads for the methylene carbons and traids for the methane carbons [18]. T1 and nuclear Overhauser effects (NOE) effects are not necessarily equal for these different signals. With the above mentioned assignments, Elgert and co-workers [21] outlined a simple method for measuring the isomeric distribution along the polybutadiene chains. Sequence analysis can, therefore, be used to quantify the microstructure of polybutadiene. Assignments for the 13C-NMR signals of the olefinic main chain cis 1,4 and trans 1,4 carbons in butadiene are given in Table 9.2. IR methods rely, more or less, on the availability of isometrically pure reference polymers each containing relatively high concentrations of one of the three kinds of unsaturation units. Proton NMR spectroscopy does not require samples containing pure vinyl 1,2, cis 1,4 and trans 1,4 units. Except for polymers that contain only two base units, i.e., vinyl 1,2 units and cis 1,4 or trans 1,4, it is not possible by 1HNMR to obtain information about the amounts of the three base units present in polybutadienes containing a significant fraction of all three types of unsaturation [22–25] even if measurements are conducted at 400 MHz. 13

C-NMR spectroscopy offers more information because detailed assignments have been described for the aliphatic [16–19] and olefinic carbons [17, 21]. This lead van der Velden and co-workers [25] to attempt a quantitative analysis, excluding the necessity of using model polymers, by measuring the 1,2 vinyl/cis 1,4, and trans 1,4 ratio by 1H-NMR [19] and the ratio cis/trans 1,4 via 13C-NMR (aliphatic carbon resonances). The combination of these two techniques results in values for 1,2 vinyl, cis 1,4 and trans 1,4.

317


Table 9.2 Assignments for the 13C-NMR signals of the olefinic main-chain cis-1,4 and trans-1,4 carbons in polybutadiene (in Figure 9.3) Carbon atom ─C==C*─

─*C==C─

Resonance 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Triad assignment vtv ctv, ttv vcv vtt, vtc ccv, tvc ctc, ctt vcc, vct ttc, ttt ctv, ttv ccc, tcc cct, tct ccv, tcv vtc vtt vtv vcc vct vcv

Chemical shift (ppm) 131.79 131.35 130.69 130.56 130.24 130.08 129.88 129.67 129.48 129.34 128.51 128.37 128.23 128.08 127.90 127.75


Figure 9.3 shows the 13C-NMR spectra of the five polymers arranged in order of increasing vinyl 1,2 content. The olefinic region can be subdivided into two parts: a. Resonances at approximately 114 and 143 ppm have been assigned to the two different vinyl 1,2 carbon atoms, the methylene and methane, surrounded by neighbouring 1,2 vinyl, cis 1,4 and trans 1,4 units. Besides compositional sequence splitting, also configurational splitting (tacticity) occur. This can be seen in Figure 9.3, especially for polymer E. b. The complex resonance pattern between 127–133 ppm, depicted in Figure 9.4, is due to compositional splitting of the two olefinic carbons in central cis 1,4 (c) or trans 1,4 (t) units, present in different combinations of homotriads (ccc and ttt), heterotriads (ccv, ttv) and symmetric and non-symmetric isolated triads (tct, vcv, tcv, vct) [17, 18, 21]. When the negligible influences of tacticity effects is ignored, the theoretically expected number of resonances on a triad level is 36, i.e., 2(xcy + xty), x, y = c, t, v. The homotriads and the symmetric isolated triads contain two magnetically equivalent carbons, so the number of resonances to be

318

Types of Unsaturation observed cannot exceed 24. Experimentally, 16 to 18 different carbon resonances are reasonably well resolved [17] (Figure 9.4). Van der Velden and co-workers [25] used the assignment of Elgert and co-workers [21] because it is possible to quantify small amounts of 1,2 vinyl groups and the absence of trans 1,4 units by this technique. Other workers who have investigated the applications of NMR spectroscopy to the analysis of unsaturation in polybutadiene and styrene–butadiene copolymers include Carlson and Altenau [26], Carlson and co-workers [27], Binder [28, 29], Braun and Canji [30, 31], Hast and Deur Siftar [32], Silas and co-workers [33], Cornell and Koenig [15], Neto and Di Lauro [34], Binder [28], Clark and Chen [35] and Harwood and Ritchey [36].

A B C

D 8 (ppm)

E 140 120 100 80

60 40

20

0

Figure 9.3 13C-NMR spectra of five polybutadienes of increasing 1,2 vinyl content. Reproduced with permission from G. van de Velden, C. Didden, T. Veermans and J. Beulen, Macromolecules, 1987, 70, 6, 1252. © 1987, ACS

319


10

A × 10 A

3

1314 15 16 17 18

4

B×5 34 8 6 11 10

34

2

ccv + tcv ctc + ctt 1

D

5,6

13 14 15 17

11 9 14 10 4 15 12 3 13 161718 5,6 A4 A3 7, 8

2

ttv + ctv vcv vtc + vtt

A1 A2

vtc

7, 8

2 1

ctv + ttv ccc + tcc cct + tct ccv + tcv vtc vtt vtv vcc vct vcv

C

vcc+vct+ ttc + ttt

1 2

B

A5

A6

10 9

11

12

15 14 16 13 17 18

6 (ppm)

E 134

133

132

131

130

129

128

127

126

Figure 9.4 13C-NMR spectra of PBD15 showing complex resonance pattern between 127 ppm and 133 ppm, revealing the 16–18 different carbon resonances. Reproduced with permission from G. van de Velden, C. Didden, T. Veermans and J. Beulen, Macromolecules, 1987, 20, 6, 1252. © 1987, ACS

320


9.1.2 Polyisoprene Unsaturation 9.1.2.1 IR Spectroscopy Isoprene has the structure:

Its polymers can contain the following four types of unsaturation:

321

Introduction to Polymer Analysis Vodchnal and Kossler [37, 38] reported an IR method for analysis of polyisoprenes suitable for polymers with a high content of 3,4 addition and relatively small amounts of cis-1,4 and trans-1,4 structural units. Absorptivities of the bands commonly used for the determination of the amount of 1,4 structural units are about 50-times lower than the absorptivity of the band at 888 cm–1 (11.26 μm) which is used for the determination of the amount of 3,4-polyisoprene units. Therefore, in analyses of samples with high content of 3,4-polyisoprene units, it is necessary to use two concentrations or two cuvettes with different thicknesses. Application of the 1780 cm–1 and 3070 cm–1 band (5.62 μm and 3.26 μm) offers the possibility of using only one cuvette and one concentration. The 1780 cm–1 and 3070 cm–1 (5.62 μm and 3.26 μm) absorption bands do not overlap with absorption bands of other structural forms, the accuracy of analyses thus being increased. Besides exact determination of the amount of 3,4 structural units, it is possible to estimate an approximate amount of 1,4 addition from the 840 cm–1, 572 cm–1 and 600 cm–1 (11.90, 17.48 and 16.66 μm) absorption bands. Values of apparent molar absorptivities of 3,4-polyisoprene, Hevea and balata in carbon disulfide solutions for the 572 cm–1, 840 cm–1, 888 cm–1, 1780 cm–1, and 3070 cm–1 (17.48, 11.90, 11.26, 5.62 and 3.26 μm) absorption bands are summarised in Table 9.3. The results of measurements of the samples in carbon disulfide solutions, obtained using absorptivities from Table 9.3, are summarised in Table 9.4. From the value of absorption at 840 cm–1 (11.90 μm), the minimum amount of 1,4 structural units was estimated assuming that all 1,4 units are cis. Analysis using the 572 cm–1 and 980 cm–1 (17.48 μm and 10.20 μm) bands was inapplicable due to the cyclic structure.

Table 9.3 Apparent molar absorptivities (km) for CS2 solutions 1,4 units

3,4 units

Sample

572 cm-1 (17.48 μm)

840 cm-1 (11.90 μm)

888 cm-1 (11.26 μm)

1780 cm-1 (5.62 μm)

3070 cm-1 (3.26 μm)

Hevea

5.7

16.6

1.72

0

0

Balata

2.7

7.6

0.58

0

0

3,4-Polyisoprene

6.5

0

110

3.46

30.6

Reproduced with permission from J. Vodchnal and I. Kossler, Collection of Czechoslovak Chemical Communications, 1964, 29, 2428 [37]. © 1964, Collection of Czechoslovak Chemical Communications

322

Types of Unsaturation The results of analyses of the samples in potassium bromide pellets are presented in Table 9.5. In these analyses it was possible to utilise the 572 cm–1 and 600 cm–1 (17.48 μm and 16.66 μm) absorption bands only for an approximate estimation of the relative abundance of 1,4 structural units.

Table 9.4 Analyses using various infrared absorption bands 3,4 units (%) –1

–1

3070 cm

88 cm

1,4 units (%) –1

1780 cm

Average

840 cm–1

31.6

32.9

-

32

43

37.0

39.0

-

38

42

37.2

40.2

-

39

47

39.0

2.3

-

41

41

-

49.7

51.3

51

-

-

54.6

59.6

58

28


Table 9.5 Results of polyisoprene analyses cis-1,4 (%) -1

trans-1,4 (%)

3,4 units (%)

572 cm (17.48 μm)

-1

–1

840 cm (11.90 μm)

888 cm (11.26 μm)

3070 cm–1 (3.26 μm)

1780 cm–1 (5.62 μm)

0

10

-

57

62

60

9

15

11

-

45

17

-

20

-


Fraga and Benson [39] investigated a thin-film IR method for the analysis of polyisoprene. They emphasise that clear, smooth and uniform films are necessary, and that these can be cast from a toluene solution of the polymer. Film thickness should be maintained to provide between 0.5 and 0.7 absorbance units at the peak near 1370 cm–1 (7.3 μm). Binder [40] found a direct correlation between the intensity of the 742

323

Introduction to Polymer Analysis cm–1 (13.48 μm) and the percentage net cis-1,4 for various high cis-1,4 polyisoprenes. Synthetic cis-1,4 polyisoprenes prepared with Zeigler catalysts or lithium catalysts contain a small percentage of the 3,4 structure. Naturally occurring polyisoprenes such as natural rubber (Hevea), gutta percha, balata, and chicle consist exclusively of the 1,4 structure. The differences between the thermal and mechanical properties of the natural and synthetic polyisoprenes have been attributed to the amount of cis1,4 units. It is reasonable to expect that the physical properties of the polyisoprenes are also affected by the distribution of the isomeric structure units along the polymer chain, as well as the composition of the polymers. Various other IR methods [41–45] have been reported for the analysis of polyisoprene with predominantly cis-1,4 and trans-1,4 structural units with low amounts of 3,4 addition. Maynard and Moobel [45], and Ferguson [46] described IR methods, for determining cis-1,4, trans-1,4, 3,4 and 1,2 structures and 1,4 structures in polyisoprene. Kossler and Vodchnal [47] concluded that the IR spectra of polymers containing cis1,4, trans-1,4 and 3,4 or cyclic structural units are not additively composed of the spectra of stereoregular polymers containing only one of these structures. It is known that stereoregular cis-1,4 polyisoprene (Hevea) and stereoregular trans-1,4 polyisoprene (balata) have absorption bands at 1130 cm–1 and 1150 cm–1 (8.84 μm and 8.69 μm), respectively. These workers found that a polymer having a high content of 3,4 structural units, in addition to the 1,4 structural units, has no absorption band at 1130 cm–1 or 1150 cm–1 (8.84 μm or 8.69 μm) but does have a band at 1140 cm–1 (8.77 μm). They attribute this band to the C– CH3 vibration of the –C(CH3)=CH structural unit separated by other structural units. The appearance of the absorption band at 1140 cm–1 (8.77 μm) in some synthetic polyisoprenes has been mentioned by Binder [43, 48, 49] with the comment that the origin of this band is not known [49]. A similar phenomenon has been discovered by analysis of a polymer having approximately 20% trans-1,4 in addition to about 75% cis-1,4 structural units, as estimated by an analysis using the absorption bands at 572 cm–1 and 980 cm–1 (17.48 μm and 10.20 μm) [50]. The band at 1130 cm–1 (8.85 μm) was shifted towards higher values. In a mixture of Hevea and balata with the same content of 20% trans-1,4 structural units, the 1130 cm–1 and 1150 cm–1 (8.85 μm and 8.69 μm) bands are quite distinct. Behaviour of the 1130 cm–1 and 1150 cm–1 (8.85 and 8.69 μm) bands is in agreement with the finding of Golub [50, 51], who has shown that during the cis–trans isomerisation of polyisoprene the 1136 cm–1 (8.80 μm) absorption band appear instead of the 1126 cm–1 (8.88 μm) band in cis-1,4 isomers or the 1149 cm–1

324

Types of Unsaturation (8.70 μm) band in trans-1,4 isomers. The statement of Maynard and Moobel [45] that the small amount of trans-1,4 structural units may be better detected using the band pair near 1307 cm–1 (7.65 μm) rather than the bands at 1131 cm–1 and 1152 cm–1 (8.84 μm and 8.68 μm) is also in good agreement with the findings of Kossler and Vodchnal [47]. These results suggest that only polyisoprenes having long sequences of cis-1,4 or trans-1,4 units have absorption bands at 1130 cm–1 and 1150 cm–1 (8.85 μm and 8.69 μm), respectively. It is also evident that the analysis of synthetic polyisoprenes using these absorption bands leads to distorted results. Kossler and Vodchnal [47] obtained better results using the absorption bands at 572, 980 and 888 cm–1 (17.48, 10.20 and 11.26 μm) for cis-1,4 and trans-1,4 and 3,4 polyisoprene structural units, respectively. Various combinations of different absorption bands permits one to conclude if a polymer is more of the block copolymer type or a mixture of stereoregular polymers.

9.1.2.2 Nuclear Magnetic Resonance (NMR) spectroscopy Tanaka and co-workers [52] determined the distribution of cis-1,4 and trans-1,4 units in 1,4 polyisoprenes using 13C-NMR spectroscopy. Thay found that cis-1,4 and trans-1,4 units are distributed almost randomly along the polymer chain in cis–trans isomerised polyisoprenes, and that chicle is a mixture of cis-1,4 and trans1,4 polyisoprenes. These workers [53] also investigated the 13C-NMR spectra of hydrogenated polyisoprenes and determined the distribution of 1,4 and 3,4 units along the polymer chain for n-butyl lithium catalysed polymers. They confirmed that these units are randomly distributed along the polymer chain. Polymers did not contain appreciable amounts of head-to-head or head-to tail 1,4 linkages. The diad distribution of cis-1,4 and trans-1,4 units in low molecular weight 1,4polyisoprene has been determined from the 13C-NMR spectra at 77 oC K by MareseSeguela and co-workers [54]. Gronski and co-workers [55], Beebe [56], Dolinskaya and co-workers [57] and Duch and Grant [58], used the chemical shift correction parameters for linear alkanes in the aliphatic region of 13C-NMR spectra to determine the relative amounts of 3,4 and cis-1,4 units of polyisoprene. Microstructure studies have been carried out.

9.1.3 Polyethylene Unsaturation Much useful information regarding the types of unsaturation present in polyethylene can be gained by IR spectroscopy. Polyethylenes can contain various types of

325

Introduction to Polymer Analysis unsaturation of great importance from the microstructural point of view. These include external vinylidene (a), terminal vinyl (b) and internal cis (d) and trans (c) unsaturation:

In branched polyethylene, most of the unsaturation is of the external vinylidene type, whereas trans olefinic end groups and terminal vinyl groups are relatively low in concentration. In linear polyethylene, most of the unsaturation is terminal vinyl with relatively small amounts of external vinylidene and trans olefinic end groups. The electron irradiation of linear and branched polyethylenes causes several molecular rearrangements in the polymer structure [59]. In addition to the significant changes in the type and distribution of unsaturated groups, IR comparison of the radiationinduced chemical changes that occur in air and in a vacuum showed that oxygen has a marked influence on the structural rearrangements that occur on irradiation. Figure 9.5 shows the IR spectra of the branched polyethylene before and after irradiation in vacuum and air. Strongly absorbing trans-type unsaturation (CH=CH) bands at 964 cm–1 (10.37 μm) appear in the vacuum and air-irradiated sample spectra. Vinylidene decay on irradiation is shown by the decrease in the R1R2C-CH2 band at 888 cm–1 (11.26 μm).

326

Types of Unsaturation Irradiation in a vacuum produces a significant decrease in methyl (–CH3) content (1373 cm–1) (7.28 μm), whereas in the bombardment in air there appears to be only a negligible decrease in –CH3 (if any). Comparison of the 720–730 cm–1 (13.89–13.70 μm) doublet shows that only the 720 cm–1 (13.89 μm) component in the air-irradiated sample remains in the spectrum of the vacuum samples, whereas there is only a slight decrease of the 730 cm–1 (13.69 μm) component in the air-irradiated samples. Additional evidence of structural changes is shown in the spectra of the air-irradiated sample. Here –OH and C=O bands appear, and there is a general depression of the spectrum background from 1300 cm–1 to 900 cm–1 (7.69–11.11 μm).

R1R2C

Before irradiation

CH2

CH3 Absorption

In vacuum CH

CH

In air OH C

4000

3000

O

2000 1800 1600 1400 1200 1000 800 Frequency (cm-1)

Figure 9.5 Effect of irradiation (500 Mrad) in air and vacuum on branches in polyethylene. Reproduced with permission from J.P Luongo and R.J. Salovay, Journal of Applied Polymer Science, 1963, 7, 2307. © 1963, Wiley Figure 9.6 shows the unsaturation region of the spectra. The top two traces show this region for branched polyethylene before and after 6 Mrad irradiation. The lower traces are those of linear polyethylene before and after similar irradiation. In branched polyethylene before irradiation, most of the unsaturation is of the external vinylidene type. After a dose of 6 MR, trans-unsaturation at 964 cm–1 (10.73 μm) increases and the vinylidene (at 888 cm–1, 11.26 μm) decreases.

327


R1R2C

CH

CH2

DYNK Before irradiation

CH2 After 6 MR

H C

C

Absorption

H

CH

MARLEX Before irradiation

CH2 After 6 MR

H C

C H

POLYMETHYLENE Before irradiation After 6 MR H C

C H

1050 1000 950 900 850 Wavenumber (cm-1)

Figure 9.6 Unsaturation region of three polyethylenes before and after irradiation at 6 Mrad. Reproduced with permission from J.P. Luongo and R.J. Salovay, Journal of Applied Polymer Science, 1963, 7, 2307. © 1963, Wiley

In linear polyethylene, almost all the saturation is of the terminal vinyl type (CH= CH2) as shown by the bands at 990 cm–1 and 910 cm–1 (10.10 μm and 10.99 μm). Here again, after only a 6-MR dose, the trans groups form rapidly and the vinyl groups at 990 cm–1 and 910 cm–1 (10.10 μm and 10.99 μm) decrease. Because of the rapid increase of the trans groups during irradiation and the simultaneous decrease of the other unsaturated groups, it may appear that the trans groups are being formed from a reaction involving the sacrifice of the other unsaturated groups in the polymer. To determine the validity of this observation, Luongo and Salovay [60] exposed to similar doses of irradiation a sample of polymethylene which has no IR-detectable

328

Types of Unsaturation unsaturation or branching. In Figure 9.6 (lower curves), in polymethylene, the trans unsaturation band still forms strongly after irradiation in air or vacuum. This means that the trans groups come from a reaction that is independent of unsaturation or branching. As for the vinyl and vinylidene decay, although there is no conclusive mechanism to explain their disappearance, they probably become saturated by atomic hydrogen in the system or become crosslinking sites. Unsaturation in low-density polyethylene has been estimated to be ± 0.003 C=C 103 carbon atoms by compensating with brominated polymer of the same thickness [61]. Rueda and co-workers [62] used IR spectroscopy to measure vinyl, vinylidene and internal cis or trans olefinic end groups in polyethylene. Dankovics [63] determined the degree of unsaturation of low-density polyethylene. The total degree of unsaturation in polyethylene was determined by summing the vinyl, vinylene and vinylidene unsaturation derived from the differential infrared spectra using an unbrominated polyethylene film as the sample and a brominated film as the reference [62]. Hammond and co-workers [64] used IR spectroscopy to determine vinyl and carbonyl groups in high-density polyethylene.

9.1.4 Polypropylene Unsaturation The types of unsaturation that can occur in polypropylene are listed next:

329

Introduction to Polymer Analysis Identification of the free radicals produced when polymers are irradiated (usually by gamma-radiation from 60Co or by fast electrons from an electron accelerator tube) can sometimes give vital information regarding structural features, saturated and unsaturated, of the original polymer. Slovakhotova and co-workers [65] applied IR spectroscopy to a study of structural changes in polypropylene in vacuum exposed to fast electrons from an electron accelerator tube (200 kV accelerating field) and with gamma-radiation from 60 Co. They found that the IR spectrum of irradiated polypropylene contains absorption bands in the 6.08, 11.23 and 13.60–13.51 μm regions. The first two bands correspond to RRĆ = CH2 vinylidene groups and the band in the 13.60– 13.51 μm region to propyl branches, R-CH2CH2CH3. When polypropylene is degraded thermally, these groups are formed by disproportionation between free radicals formed by rupture of the polymer backbone. Under the action of ionising radiations the polymer backbone ruptures, with formation of two molecules, with vinylidene and propyl end groups, at a temperature as low as that of spectrum of polypropylene irradiated at –196 °C and measured at –130 °C. When polypropylene is irradiated with dosages >350 Mrad, a band appears at 10.99 μm, corresponding to vinyl groups, R – CH = CH2, i.e., degradation of polypropylene can also involve simultaneous rupture of two C–C bonds in the main and side chains. The strength of the vinyl-group band in the spectrum of irradiated polypropylene is lower than that of the vinylidene group, although the extinction coefficients of these bands are approximately the same [65]. In the spectrum of amorphous polypropylene irradiated with a dosage of 4000 Mrad at –196 °C and measured at –130 °C, in addition to the band at 6.08 μm, a weaker band appears with a maximum near 6.00 μm, possibly due to internal bonds:

When the spectrum of the specimen is recorded after it is heated to +25 °C, this maximum disappears, leaving only a shoulder on the strong band at 6.08 μm. The extinction coefficient [65] of the band at 6.00 μm is less than that of the 6.08 μm band by a factor of 6.7. Supplementary evidence of the formation of internal double bonds in irradiated polypropylene is provided by the presence in the spectrum of bands in the 12.27–11.69 region. In this region lie bands due to deformation vibration of CH at double bonds in groups, existing in various conformations [43]. The appearance of bands at 12.27 μm and 11.69 μm in the spectrum of irradiated polypropylene can be regarded as an indication of the formation of internal double bonds in the polymer:

330


A study of the electron spin resonance (ESR) spectra of irradiated polypropylene has shown that the alkyl radicals formed during irradiation at –196 °C undergo transition to alkyl radicals when the specimen is heated, i.e., on heating the radical centres migrate to internal double bonds with the formation of stable allyl radicals. Irradiation at room temperature leads immediately to the formation of allyl radicals. It is very probable that the decrease in intensity of the internal double bond valency vibration band at 6.00 μm and the broadening of its maximum after a specimen irradiated at a low temperature is heated to room temperature, is associated with the formation of allyl radicals because interaction of the P-electrons of the double bond with the unpaired electron also lowers the frequency of the double-bond vibration. Comparison of the intensities of the terminal vinyl double bonds at 11.23 μm and 10.99 μm with the band at 6.08 μm in the spectrum of irradiated isotactic polypropylene shows that the intensity of absorption in the 6.08 μm region does not correlate with the intensity of absorption in the 11.23 μm regions. Thus, according to the known extinction coefficients for these bands, the ratio of their optical densities should be: D11.23/D6.08 = 3.7, D10.99/D6.08 = 3.2 and (D11.23 + D10.99)/D6.08 = 3.5 In the latter case, D6.08 is the sum of the optical densities of the vinylidene and vinyl absorption bands in this region. An optical density ratio for these bands of approximately this value was found (1.75 to 3.3) for the products of thermal degradation of polypropylene. It is seen that only in the case of amorphous polypropylene irradiated with gamma-radiation from 60Co is the ratio (D11.23 + D10.99)/D6.08 close to the value calculated from the extinction coefficients of these bands. In the spectra of irradiated isotactic polypropylene, the intensity of the 6.08 μm band is greater than would be expected if only vibration of terminal double bonds contributes to absorption in this region. This increase in absorption in the 6.08-μm region can be related to absorption by the internal double bond in the allyl radical, the vibrational frequency of which is lowered by conjugation of the P-electrons of the double bond

331

Introduction to Polymer Analysis with the unpaired electron of the radical. In amorphous polypropylene irradiated at room temperature, the alkyl radicals can combine rapidly; therefore there is little formation of allyl radicals. This explains the fact that the ratio of the optical densities of the terminal double bond bands in the 11.10 μm and 6.08 μm regions is close to the calculated value. Conjugated double bonds in irradiated polypropylene are indicated by the following: (1) in the spectra of isotactic polypropylene irradiated with dosages of 2000–4000 Mrad at room temperature, there is a band at 6.21 μm, which is the region in which polyene bands occur, whereas this band is absent from the spectrum of isotactic polypropylene irradiated with the same dosages at –196 °C; (2) in the electronic spectra of these polypropylene specimens, the boundary of continuous absorption is shifted to a region of longer wavelength in comparison with the spectra of polypropylene irradiated with the same dosages at –196 °C. It has been shown from ESR spectra [66] that when specimens of isotactic polypropylene are heated above 80 °C they contain polyenic free radicals:

and this also indicates the possibility of migration of double bonds along the polymer chain.

9.2 Unsaturation in Copolymers 9.2.1 Styrene–divinyl benzene IR Spectroscopy During the early stages of copolymerisation of styrene and divinylbenzene, linear and branched primary macromolecules are formed at the divinylbenzene repeat units [67–72]. Intermolecular reactions of pendent vinyl groups with growing radicals form crosslinks and lead to gelation. Intramolecular cyclisations of pendent vinyl groups do not form crosslinks. The extent of reaction at gelation of styrene–divinylbenzene copolymers is greater than predicted on the basis of copolymer reactivity ratios. The delay of gelation could be due to extensive intramolecular cyclisation or to low copolymer reactivity of the second double bond of divinylbenzene. The relative contributions of these factors to the delay of gelation could be determined if the cyclic structures in the copolymers could be analysed, and if the fraction of pendent vinyl groups could be analysed accurately enough to determine the reactivity ratios involving the second double bond.

332

Types of Unsaturation Pendent vinyl groups in styrene–divinylbenzene copolymers have been analysed by IR and Raman spectroscopy [73–84] and by wet chemical methods [83, 85] after extents of reaction varying from before gelation to after nearly complete conversion of monomers. Periysamy and Ford [86] report a new analytical approach to the problem. Copolymers of styrene with methane-13C-labelled p-divinylbenzene were analysed by liquid-state and solid-state cross-polarisation magic angle spinning (CP–MAS) 13C-NMR methods. p–Divinyl benzene (DVB), 13C-labelled at the methane carbon of the vinyl group was copolymerised in suspension with styrene at 70 ºC, 70–95 ºC and 135–155 ºC using azo bis(isobutyronitrile) (AIBN) as initiator. In this method, the number of unreacted vinyl groups in each copolymer was determined by 13C-CP–MAS NMR analysis of solid samples, direct polarisation 13C-NMR analysis of deuterochloroformswollen gels and bromination. Results from all three methods were found to agree qualitatively. 13

C-NMR spectra of the labelled networks were obtained by direct polarisation, liquid-state method with fully CDCl3-swollen samples, and by the solid-state, CP–MAS with solid samples. By both methods, the 1% crosslinked samples prepared at 70 °C and at 95 °C showed a peak at 137 ppm due to the labelled carbon of unpolymerised vinyl groups. The 137 ppm signal was not seen after further polymerisation of the 1% crosslinked sample at 135–155 °C. Polystyrenes prepared with 10% and 20% of the labelled DVB by the same procedures showed residual vinyl groups before and after the 135–155 °C post-polymerisation. With the 10% and 20% crosslinked samples, only CP–MAS spectra gave peaks of labelled carbon narrow enough to analyse. The residual vinyl groups of all labelled samples were analysed quantitatively from peak areas in the NMR spectra by two methods. First, the area of the 137 ppm vinyl peak was compared with the area of all of the aromatic carbon signals in the spectrum due to styrene and divinylbenzene carbons in natural abundance. Second, the area of the 137 ppm peak was compared with the area of all of the aliphatic carbon signals in the spectrum, which includes signals from polymerised labelled carbons of the DVB and from all other aliphatic carbons at natural abundance. It was assumed that all carbon atoms in the sample are equally detectable in each NMR spectrum (Table 9.6). All of the labelled polystyrene networks were also analysed by bromination of residual vinyl groups. The 13C-NMR and bromination methods were also applied to several commercial crosslinked polystyrenes prepared with unlabelled divinylbenzene, which usually consists of a 2:1 meta/para mixture of isomers and contains also meta and para ethylvinylbenzene. NMR results for 20–80% crosslinked macroporous polymers are in Table 9.6.

333

Introduction to Polymer Analysis Periyasamy and Ford [86] concluded that all common styrene-DVB copolymers, even those containing as little as 1% DVB, contain unreacted DVB vinyl groups. No method is available yet for accurate, quantitative analysis of the residual vinyl groups because wet chemical methods do not detect vinyl groups in the most hindered parts of the highly crosslinked, heterogeneous networks. IR and 13C-NMR analyses may contain systematic errors in peak area determinations.

Table 9.6 Fraction of divinylbenzene repeat units with a pendent vinyl group DVB (wt%)

Polymerisation temperature (oC)

1

CP-MASa

Direct polarisationb

Bromine additionc

Vinyl versus aliphatic

Vinyl versus aromatic

Vinyl versus aliphatic

Vinyl versus aromatic

70

0.34

0.39

0.54

0.67

0.45 + 0.10

1

95

0.16

0.16

0.27

0.19

0.13 + 0.05

1

155

0.00

0.00

0.18

0.17

0.05 + 0.02

4

95

0.19

0.15

0.10 + 0.03

10

70

0.21

0.14

0.28 + 0.11

10

155

0.18

0.20

0.18 + 0.02

20

70

0.28

0.14

0.33 + 0.05

20

155

0.28

0.19

0.26 + 0.04

d

0.54

0.39

0.09 + 0.01

e

0.63

0.47

0.13 + 0.01

f

0.81

0.51

0.10 + 0.01

20

50

80 a

Dry solid samples.

b

CDCl3-swollen samples.

c

Average and standard deviation of the three measurements.

d

Rohm & Haas XAD-1.

e

Rohm & Haas XAD-2.

f

Rohm & Haas XAD-4.

Reproduced with permission from M. Periyasamy and W.T. Ford, Journal of Polymer Science: Polymer Chemistry Edition, 1989, 27, 7, 2357. © 1989, Wiley

1

H-NMR spectroscopy has been used to determine residual vinyl groups in polydivinylbenzenes [87].

334


9.2.2 Poly(trimethylolpropane trimethacrylate) (TRIM)

There are several methods for the detection of carbon–carbon double bonds, including IR spectroscopy [88–91]: IR [87–90], Raman spectroscopy [92], and chromatography together with the formation of Pt(II) complexes [91] and addition of bromine [93, 94]. The high crosslinking density of polyTRIM and its rich IR spectrum make these analyses quite difficult. With the advent of high-resolution solid-state NMR (CP-MAS NMR), a powerful technique to analyse insoluble polymers has become available. In an unreacted acrylate group, the carbonyl bond is conjugated with a double bond, which should shift the 13C carbonyl resonance about 10 ppm upfield compared with the reacted units. This approach has also been used to determine the amount of unreacted units in several different polymers obtained from multi-functional acrylates and methacrylates [95–97], including TRIM [98]. Hjertberg and co-workers [99] used CP–MAS 13C-NMR to determine the double bonds in TRIM and compared results with those obtained by standard bromine addition methods. They also examined the possibility of utilising different relaxation parameters obtained by NMR measurements to study the mobility of unreacted units. A detailed analysis of the cross-polarisation behaviour showed that quantitative results can be obtained. The amount of unreacted units, typically 0–15%, was found to depend on the polymerisation parameters. Conditions favouring mobility, i.e., higher temperatures or increased solvent quality, resulted in lower content of residual double bonds. Bromine addition values are 2–3% higher than NMR data. Reactivity toward bromine further indicates that the mobility is reasonably high. This has also been confirmed by measurements of the rotating-frame relaxation time constant (T1R(13C). Most likely T1R is dominated by spin–lattice processes; i.e., it can be interpreted in terms of molecular dynamics. The values obtained for C=O and >C*- CH2 in unreacted units are about twice that of C=O in reacted units, indicating increased mobility. The reactivity of the remaining double bonds in a radical polymerisation with a chiral monomer was also demonstrated.

335

Introduction to Polymer Analysis The percentage fraction of unreacted methacrylic groups in TRIM was found to depend on polymerisation temperature and range from 8.4% at 60 °C to 2.7% at 90 °C. 13

C CP–MAS NMR has also been used in studies on poly(tetraethylene glycol dimethacrylate) [96].

9.2.3 Miscellaneous Copolymers Pyrolysis gas chromatography and IR spectroscopy have been used to determine vinyl groups in ethylene–vinyl acetate copolymers [100]. ESR spectroscopy has been applied to studies of unsaturation and other structural features in a wide range of homopolymers including: polyethylene [101–110], polypropylene [111–121], polybutenes [115], polystyrene [122–124], PVC [125, 126], polyvinylidene chloride [127], polymethyl methacrylate [128–137], polyethylene glycol polycarbonates [137–140], polyacrylic acid [136–139, 141, 142], polyphenylenes [143], polyphenylene oxides [143], polybutadiene [144], conjugated dienes [145, 146], polyester resins [146], cellophane [143, 147] and also to various copolymers including styrene grafted polypropylene [148], ethylene–acroline [149], butadiene–isobutylene [150], vinyl acetate copolymers [151] and vinyl chloride–propylene.

9.3 Ozonolysis Techniques The oxidation of double bonds in organic compounds and polymers in a non-aqueous solvent leads to the formation of ozonides which, when acted upon by water, are hydrolysed to carbonyl compounds:

Triphenyl phosphine is frequently used to assist this reaction. When applied to complex unsaturated organic molecules or polymers, this reaction has great potential for the elucidation of the microstructure of the unsaturation. Examination of the reaction products, for example by conversion of the carbonyl compounds to carboxylates then esters followed by gas chromatography (GC), enables identification of these products.

336

Types of Unsaturation An example of the value of the application of this technique to a polymer structural problem is the distinction between polybutadiene made up of consecutive 1,4–1,4 butadiene sequences (I), and polybutadiene made up of alternating 1,4 and 1,2 butadiene sequences (II), i.e., 1,4–1,2–1,4:

Upon ozonolysis, followed by hydrolysis, these in the case of 1,4–1,4 sequences produce succinaldehyde (CHO–CH2–CH2CHO) and in the case of 1,4–1,2–1,4 sequences produce formyl 1,6-hexane-dial:

1,4 – 1,4 sequences

337

Introduction to Polymer Analysis 1,4–1,2–1,4 sequences:

Table 9.7 Micro-ozonoysis of polybutadiene Sample

1,4 Vinyl (cis + (1,2), trans) (%)


3-Formyl 1,6hexanedial

4-octene 1,8-dial

1, 2 units occurring in 1,4-1,2-1,4 sequences

1

98.0

2

50

1

49

0.5

2

89.1

10.9

30

10

60

5

3

89.0

11.0

43

7

50

3

4

81.0

19.0

34

14

52

6

5

76.2

23.8

36

25

39

11

6

71.8

28.2

33

27

40

11

7

69.7

30.3

48

26

26

10

8

67.7

32.3

36

26

38

9

9

64.2

35.8

38

31

31

10

10

62.8

37.2

45

27

28

8

11

50.5

49.5

26

41

33

12

12

56.0

44.0

30

39

31

11

13

26.0

74.0

33

64

3

5


Analysis of the reaction product for concentrations of succinaldehyde and 3-formyl1,6-hexane-dial can show whether the polymer is 1,4–1,4 or 1,4–1,2–1,4, or whether it contains both types of sequence.

338

Types of Unsaturation Various workers [152–158] have applied this technique to the elucidation of the microstructure of polybutadiene. They found that the 3 formyl-1,6-hexane-dial content was directly proportional to the 1,2 (vinyl) content of polymers containing 1,4–1,2–1,4 butadiene sequences. Polymers having 98% cis-1,4 structure, 98% trans-1,4 structure and a series of polymers containing from 11% to 75% 1,2 structure were ozonised (Table 9.7). Final products obtained from these polymers were succinaldehyde, 3-formyl-1,6 hexane-dial, and 4-octene-1, 8-dial. Model compounds were ozonised and products compared with those from the polymers.

A

C

B

0

2

4

6 8 Time (min)

10

12

Figure 9.7 Ozonolysis products from PBD containing 11% vinyl structure (a) succinaldehyde, (b) formyl-1,6-hexane dial, (c) 4-octene, 1,8 dial. Source: Author’s own files

Figures 9.7 and 9.8 show chromatographic separation of the ozonolysis products from polybutadienes having different amounts of 1,2 structure, as measured by infrared or NMR spectroscopy. Figure 9.9 shows the relationship of 1,2 content to the amount of 3-formyl-1,6 hexane-dial in the ozonolysis products.

339


A

B

C

0

2

4

6 8 Time (min)

10

12

Fraction 3 - formyl - 1, 6 - hexanedial

Figure 9.8 Ozonolysis products from PBD containing 37.2% vinyl structure (a) succinaldehyde, (b) 3-formyl-1,6 hexane dial, (c) 4-octene-1,8 dial. Source: Author’s own files

0.70 0.60 0.50 0.40 0.30 0.20 0.10 0

10

20

30

40

50

60

70

80

Figure 9.9 Relationship of yield of 3-formyl-1,6 hexane dial from ozonolysis to percentage of vinyl structure (from NMR and IR spectra) in PBD. Source: Author’s own files

340

Types of Unsaturation The amount of 1,4–1,2–1,4 sequences in polybutadienes can be estimated from the amounts of the different ozonolysis products if one considers the amount of 1,4 structure not detected. (Because ozonolysis cleaves the centre of a butadiene monomer unit, one-half of a 1,4 unit remains attached to each end of a block of 1,2 units after ozonolysis; these structures do not elute form the gas chromatographic column.) Using random copolymer probability theory, the maximum amounts of these undetected 1,4 structures can then be calculated. A further example concerns the application of the ozonolysis technique to butadiene– propylene copolymers [153, 158–160]. Samples of highly alternating copolymers of butadiene and propylene yielded large amounts of 3-methyl-1,6-hexane-dial when submitted to ozonolysis. The ozonolysis product from 4-methyl-cyclohexane-1 (i.e., 3-methyl-1,6 hexane-dial:

was used as a model compound for this structure. Ozonolysis of these polymers occurs as shown below:

The amount of alternation in these polymers can be determined if the amounts of 1,4 and 1,2 polybutadiene structure and total propylene have been determined by infrared or NMR spectroscopy. Table 9.8 shows results obtained for several butadiene–propylene copolymers having more or less alternating structure. Similar polymers have been analysed by Kawasaki [159] by use of conventional ozonolysis methods with esters as the final products. The technique has been applied to various other unsaturated polymers. Thus, polyisoprene, having nearly equal 1,4 and 3,4 structures, produced large amounts of laevulinaldehyde, succinaldehyde and 2,5 hexanedione, indicating blocks of 1,4 structures in head-tail, tail-tail and head-head configurations. Hill and co-workers [155] utilised ozonolysis in their investigation of a butadiene methyl methacrylate copolymer. The principal products were succinic acid,

341

Introduction to Polymer Analysis succindialdehyde and dicarboxylic acids containing several methyl-methacrylate residues. The percentage of butadiene (9.2%) recovered as succinic acid and succindialdehyde provided a measure of the 1,4 butadiene-1,4-butadiene linkages in the copolymers, and the percentage of methyl-methacrylate units (51%) recovered as trimethyl 2-methyl-butane-1,2, 4-tricarboxylate (4) n = 1, provided a measure of the methyl-methacrylate units in the middle of butadiene–methacrylate–butadiene triads.

Table 9.8 Microzonolysis of butadiene-propylene copolymers 1,4 (%)

1,2 (%)

Propylene (mole%)

A

45

5.7

B

47.8

C D


3-Methyl1,6hexanedial

3-Formyl 1, 6-hexanedial

4-Octene-1, 8-dial

Alternating BD/Pr (%)

49.3

5

92

1

2

77

2.2

50

11.5

85

0.5

3

71

53.1

3.2

43.7

25

61

6

8

48

-

-

30

49

38

1

12

33


References 1. D.W. Fraga, Private Communication, Emeryville Research Laboratories, Shell Chemical Co Ltd., Emeryville, CA, USA. 2. J.L. Binder, Journal of Polymer Science Part A: General Papers, 1963, 1, 1,47. 3. G. Natta in Proceedings of the 15th Annual Technical Conference of the Society of Plastic Engineers, New York, NY, USA, 1959. 4. H.C. Dinsmore and D.C. Smith, Rubber Chemistry and Technology, 1949, 22, 2,572. 5. D.L. Harms, Analytical Chemistry, 1953, 25, 8, 1140. 6. D. Hummel, Rubber Chemistry and Technology, 1959, 32, 3, 854. 7. M. Lerner and R.C. Gilbert, Analytical Chemistry, 1964, 36, 7, 1382.

342

Types of Unsaturation 8. M. Tyron, E. Horowicz and J.J. Mandel, Journal of Research of the National Bureau of Standards, 1955, 55, 219. 9. D.A. MacKillop, Analytical Chemistry, 1968, 40, 3, 607. 10. N.K. Albert, Private Communication Woodbury Research Laboratory, Shell Chemical Co, Woodbury, NY, USA., October 1966. 11. T.R. Crompton and V.W. Reid, Journal of Polymer Science Part A: General Papers, 1963, 1, 1, 347. 12. C.E. Miller, B.E. Eichinger, T.W. Gurley and J.G. Hermiller, Analytical Chemistry, 1990, 62, 17, 1778. 13. H. Edwards, A. Johnson and I. Lewis, Raman Spectroscopy, 1993, 24, 435. 14. R.R. Hampton, Rubber Chemistry and Technology, 1972, 45, 3, 546. 15. S.W. Cornell and J.L. Koenig, Macromolecules, 1969, 2,5, 540. 16. Y. Alaki, T. Yoshimoto, M. Imanari and M. Takeuchi, Rubber Chemistry and Technology, 1973, 46, 2, 350. 17. A.D. H. Clague, J.A.M. van Broekhoven and L.P. Blaauw, Macromolecules, 1974, 7, 3, 348. 18. P.T. Suman and D.D. Werstler, Journal of Polymer Science, Polymer Chemistry Edition, 1975, 13, 8, 1963. 19. S. Bywater, B. Firat and P.E. Black, Journal of Polymer Science, Polymer Chemistry Edition, 1984, 22, 3, 669. 20. S. Bywater, Polymer Communications, 1983, 24, 7, 203. 21. K.F. Elgert, G. Quack and B. Stuetzel, Polymer, 1975, 16, 3, 154. 22. H.J. Harwood, Rubber Chemistry and Technology, 1982, 55, 3, 769. 23. E.R. Santee, L.O. Malotky and M. Morton, Rubber Chemistry and Technology, 1973, 46, 5, 1156. 24. E.R. Santee, V.D. Mochel and M. Morton, Journal of Polymer Science, Polymer Letters Edition, 1973, 11, 7, 453. 25. G. Van de Velden, C. Didden, T. Veermans and J. Beulen, Macromolecules, 1987, 20, 6, 1252.

343

Introduction to Polymer Analysis 26. D.W. Carlson and A.G. Altenau, Analytical Chemistry, 1969, 41, 7, 969. 27. D.W. Carlson, H.C. Ransaw and A.G. Altenau, Analytical Chemistry, 1970, 42, 11, 1278. 28. J.L. Binder, Analytical Chemistry, 1954, 26, 12, 1877. 29. J.L. Binder, Journal of Polymer Science Part A: General Papers, 1963, 1, 1,47. 30. D. Braun and E. Canji, Angewandte Makromolekulare Chemie, 1973, 33, 143. 31. D. Braun and E. Canji, Angewandte Makromolekulare Chemie, 1974, 35, 27. 32. M. Host and D. Deur-Siftar, Chromatographia, 1972, 5, 9, 502. 33. R.S. Silas, J. Yates and V. Thornton, Analytical Chemistry, 1959, 31, 4, 529. 34. N. Neto and C. Di Lauro, European Polymer Journal, 1967, 3, 4, 645. 35. J.K. Clark and H.Y. Chen, Journal of Polymer Science, 1974, 12, 4, 925. 36. H.J. Harwood and W.M. Ritchey, Journal of Polymer Science, Polymer Letters Edition, 1964, 2, 6, 601. 37. J. Vodchnal and I. Kossler, Collection of Czechoslovak Chemical Communications, 1964, 29, 2428. 38. J. Vodchnal and I. Kossler, Collection of Czechoslovak Chemical Communications, 1964, 29, 2859. 39. D.W. Fraga and L.H. Benson, Private Communication Shell Chemical Co Ltd., Torrance, CA, USA. 40. J.L. Binder, Journal of Polymer Science Part A: General Papers, 1963, 1, 1, 37. 41. W.S. Richardson and A. Sacher, Journal of Polymer Science, 1953, 10, 4, 353. 42. E.I. Plkovskiu and M. Volkenstejn, Doklady Academii Nauk SSSR, 1954, 95, 301. 43. J.L. Binder and H.C. Ransaw, Analytical Chemistry, 1957, 29, 4, 503. 44. P.J. Corish, Spectrochimica Acta, 1959, 15, 598. 45. J.T. Maynard and W.E. Mochel, Journal of Polymer Science, 1954, 13, 69, 251.

344

Types of Unsaturation 46. R.C. Ferguson, Analytical Chemistry, 1964, 36, 11,2204. 47. I. Kössler and J. Vodchnal, Journal of Polymer Letters, 1963, 1, 8, 415. 48. J.L. Binder, Rubber Chemistry and Technology, 1962, 35, 1, 57. 49. J.L. Binder, Journal of Polymer Science, 1963, 1, 1, 37. 50. M.A. Golub, Journal of Polymer Science, 1959, 36, 130, 523. 51. M.A. Golub, Journal of Polymer Science, 1959, 36, 10. 52. Y. Tanaka, H. Sato and T. Seimiya, Polymer Journal, 1975, 7, 2, 264. 53. Y. Tanaka, H. Sato, A. Ogura and I. Nagoya, Journal of Polymer Science, Polymer Chemistry Edition, 1976, 14, 1, 73. 54. B. Marese-Seguela, M. St Jacques, J.M. Renaud and J. Prud’Homme, Macromolecules, 1977, 10, 2, 431. 55. W. Gronski, N. Murayama, H.J. Cantow and T. Miyamato, Polymer, 1976, 17, 4, 358. 56. D.H. Beebe, Polymer, 1978, 19, 2, 231. 57. E.R. Dolinskaya, A.S. Khachaturov, I.A. Poletayeva and V.A. Kormer, Makromolekulare Chemie, 1978, 179, 409. 58. M.W. Duch and D.M. Grant, Macromolecules, 1970, 3, 2, 165. 59. P.G.M. Van Stratum and J. Dvorák, Journal of Chromatography, 1972, 71, 1, 9. 60. J.P. Luongo and R.J. Salovay, Journal of Applied Polymer Science, 1963, 7, 6, 2307. 61. M.J. Barral, Journal of Polymer Science, Polymer Chemistry Edition, 1975, 13, 7, 1515. 62. D.R. Rueda, F.J. Balta-Calleja and A. Hidalgo, Spectrochimica Acta, 1974, 30A, 1545. 63. A. Dankovics, Munay Gumi, 1974, 11, 380. [Chemical Abstracts, 1975, 82, 86773g].

345

Introduction to Polymer Analysis 64. C.L. Hammond, P.J. Hendra, B.G. Lator, W.F. Maddams and H.A. Willis, Polymer, 1988, 29, 1, 49. 65. D. Kühnle and W. Fanke, Macromoleculare Chemie, 1972, 158, 135. 66. H. Fischer and K-H. Hellwege, Journal of Polymer Science, 1962, 56, 163, 33. 67. H. Staudinger and H. Heuer, Chemische Berichte, 1934, 67, 1174. 68. K. Dusek in Developments in Polymerisation, Volume 3, Ed., R.N. Howard, Applied Science Publishers, London, UK, 1982, p.143. 69. K. Dusek and J. Spevacek, Polymer, 1980, 21, 7, 750. 70. K. Dusek, H. Galina and J. Mikes, Polymer Bulletin, 1980, 3, 1-2, 19. 71. G. Hild and P. Rempp, Pure and Applied Chemistry, 1981, 53, 8, 1541. 72. G. Hild and R. Okasha, Die Makromolekulare Chemie, 1985, 186, 1, 93. 73. J. Malinsky, J. Klaban and K. Dusek, Journal of Macromolecular Science Chemistry A, 1971, 5, 6, 1071. 74. G.J. Howard and C.A. Midgley, Journal of Applied Polymer Science, 1981,26, 11, 3845. 75. G. Popov and G. Schwachula, Plaste und Kautschuk, 1981, 28, 6, 312. 76. A.S. Gozdz and B.N. Kolarz, Die Makromolekulare Chemie, 1979, 180, 10, 2473. 77. P. Weiss, G. Hild, J. Merz and P. Rempp, Die Makromolekulare Chemie, 1970, 135, 249. 78. B. Soper, R.N. Haward and E.F.T. White, Journal of Polymer Science A1, 1972, 10, 9, 2545. 79. R.N. Haward and W. Simpson, Journal of Polymer Science, 1955, 18, 89, 440. 80. M. Bartholin, G. Boissier and J. Dubois, Die Makromolekulare Chemie, 1981, 182, 7, 2075. 81. D-Y.D. Chung, M. Bartholin and A. Guyot, Angewandte Makromolekulare Chemie, 1982, 103, 109.

346

Types of Unsaturation 82. B. Walczynski, B.N. Kolarz and H. Galina, Polymer Communications, 1985, 26, 9, 276. 83. D. Kuhnle and W. Funke, Makromolekulare Chemie, 1971, 139, 255. 84. J. Stokr, B. Schneider, A. Frydrychova and J. Coupek, Journal of Applied Polymer Science, 1979, 23, 12, 3553. 85. J.K. Fink, Journal of Polymer Science Polymer Chemistry Edition, 1981, 19, 1, 195. 86. M. Periyasamy and W.T. Ford, Journal of Polymer Science: Polymer Chemistry Edition, 1989, 27, 7, 2357. 87. Y. Nagasaki, H. Ito and T. Tsuruta, Die Makromolekulare Chemie, 1986, 187, 1,23. 88. T. Brunelet, M. Bartholin and A. Guyot, Angewandte Makromolekulare Chemie, 1982, 106, 79. 89. J. Malinsky, J. Klaban and K. Dusek, Journal of Macromolecular Science A, 1971, 5, 6, 1071. 90. K. Kun and R. Kunin, Journal of Polymer Science: Polymer Chemistry Edition, 1968, 6, 10, 2689. 91. H. Stuurman, J. Köhler, S-O. Jansson and A. Litzén, Chromatographia, 1987, 23, 5, 341. 92. P. Wieczorek, B. Kolarz and H. Galina, Angewandte Makromolekulare Chemie, 1984, 126, 1, 39. 93. H. House, Modern Synthetic Reactions, 2nd Edition, W.A. Benjamin Inc, Reading, MA, USA, 1972, p.422. 94. T. Morikawa, Kagaku Kogaku, 1967, 41, 169. 95. R.G. Earnshaw, C.A. Price, J.H. O’Donnell and A.K. Whittaker, Journal of Applied Polymer Science, 1986, 32, 6, 5337. 96. P.E. Allen, G.P. Simon, D.R.G. Williams and E.H. Williams, European Polymer Journal, 1986, 22, 7, 549. 97. P.E.M. Allen, G.P. Simon, D.R.G. Williams and E.H. Williams, Macromolecules, 1989, 22, 2, 809.

347

Introduction to Polymer Analysis 98. J-E. Rosenberg and P. Flodin, Macromolecules, 1986, 19, 6, 1543. 99. T. Hjertberg, T. Hargitai and P. Reinholdsson, Macromolecules, 1990, 23, 12, 3080. 100. E. Verdu and J. Devasa, Revista de Plasticos Modernos, 1989, 58, 545. 101. H.L. Browning, H.D. Ackermann and H.W. Patton, Journal of Polymer Science, Polymer Chemistry Edition, 1966, 4, 6, 1433. 102. K. Tsuji, Journal of Polymer Science: Polymer Chemistry Edition, 1973, 11,6, 1407. 103. K. Tsuji, Journal of Polymer Science: Polymer Chemistry Edition, 1973, 11,2, 467. 104. D. Campbell, K. Araki and D.T. Turner, Journal of Polymer Science: Polymer Chemistry Edition, 1966, 4, 10, 2597. 105. D. Campbell and D.T. Turner, Journal of Polymer Science: Polymer Chemistry Edition, 1967, 5, 8, 2199. 106. Y. Hama, Y. Hosono, Y. Furui and K. Shinohara, Journal of Polymer Science: Polymer Chemistry Edition, 1971, 9, 5, 1411. 107. K. Tsuji, T. Seiki and T. Takeshita, Journal of Polymer Science: Polymer Chemistry Edition A-1, 1972, 10, 10, 3119. 108. Y. Hori, S. Shimada and H. Kashiwabara, Polymer, 1977, 18, 6, 567. 109. Y. Hori, S. Shimada and H. Kashiwabara, Polymer, 1977, 18, 11, 1143. 110. T. Kawashima, S. Shimada, H. Kashiwabara and J. Sohma, Polymer Journal, 1973, 5, 2, 135. 111. S.I. Ohnishi, S.I. Sugimoto and I. Nitta, Journal of Polymer Science Part A: General Papers, 1963, 1, 2, 605. 112. Y. Ono and T. Feih, Journal of Polymer Science, A-1, 1966, 4, 2429. 113. T. Seguchi and N. Tamura, Journal of Polymer Science, Polymer Chemistry Edition, 1974, 12, 8, 1671. 114. T. Seguchi and N. Tamura, Journal of Polymer Science, Polymer Chemistry Edition, 1974, 12, 9, 1953.

348

Types of Unsaturation 115. B.R. Loy, Journal of Polymer Science, A-1, 1963, 1, 255. 116. L.A. Wall, Journal of Polymer Science, 1955, 17, 83, 141. 117. T. Buch, Electron Spin Resonance Spectroscopy in Irradiated Polypropylenes, NorthWestern University, Ann Arbor, MI, USA, 1960. [PhD Thesis] 118. L.J. Forrestal and W.G. Hodgson, Journal of Polymer Science Part A: General Papers, 1964, 2, 3, 1275. 119. A. Ohnishi, S. Sugimoto and I. Nitta, Journal of Polymer Science Part A: General Papers, 1963, 1,2, 625. 120. N. Kusumoto, K. Matsumoto and M. Takayanagi, Journal of Polymer Science, Polymer Chemistry Edition, 1969, 7, 7, 1773. 121. T. Ooi, M. Shiotsubo, Y. Hama and K. Shinohara, Polymer, 1975, 16, 7, 510. 122. R.E. Florin, L.A. Wall and D.W. Brown, Journal of Polymer Science Part A: General Papers, 1963, 1, 5, 1521. 123. A.T. Bullock, G.G. Cameron and P.M. Smith, Polymer, 1973, 14, 10, 525. 124. R.E. Florin and L.A. Wall, Journal of Chemical Physics, 1972,57, 4, 1791. 125. I. Ouchi, Journal of Polymer Science Part A: General Papers, 1965, 3, 7, 2685. 126. S.A. Liebman, J.F. Renwer, K.A. Gollatz and C.D. Nauman, Journal of Polymer Science, Polymer Chemistry Edition, 1971, 9, 7, 1823. 127. J.N. Hay, Journal of Polymer Science, Polymer Chemistry Edition, 1970, 8, 5, 1201. 128. M.J. Bowden and J.H. O’Donnell, Journal of Polymer Science, Polymer Chemistry Edition, 1969, 7, 7, 1665. 129. Y. Sakai and M. Iwasaki, Journal of Polymer Science, Polymer Chemistry Edition, 1969, 7, 7, 1749. 130. J.A. Harris, O. Hinojosa and J.C. Arthur, Journal of Polymer Science, 1973, 11, 12, 3215. 131. G. Geuskens and C. David, Makromolekulare Chemie, 1973, 165, 273.

349

Introduction to Polymer Analysis 132. H. Yoshioha, H. Matsumoto, S. Uno and F. Higashide, Journal of Polymer Science, Polymer Chemistry Edition, 1976, 14, 6, 1331. 133. M. Sakaguchi, S. Kodama, O. Edlund and J. Sohma, Journal of Polymer Science, Polymer Letters Edition, 1974, 12, 11,609. 134. A.T. Bullock, G.G. Cameron and J.M. Elsom, Polymer, 1974, 15, 2, 74. 135. R.E. Michel, F.W. Chapman and T.J. Mao, Journal of Polymer Science, A-1, 1967, 5, 1077. 136. Y. Hama and K. Shinohara, Journal of Polymer Science, Polymer Chemistry Edition, 1970, 8, 3, 651. 137. M.R. Clay and A. Charlesby, European Polymer Journal, 1975, 11, 2, 187. 138. J. Placek, F. Szocs and E. Borsig, Journal of Polymer Science, Polymer Chemistry Edition, 1976, 14, 6, 1549. 139. Y. Shioji, S.I. Ohnishi and I. Nitta, Journal of Polymer Science Part A: General Papers, 1963, 1, 11, 3373. 140. Y. Hajimoto, N. Tamura and S. Okamoto, Journal of Polymer Science Part A: General Papers, 1965, 3, 1,255. 141. M. Iwasaki and Y. Sakai, Journal of Polymer Science, Polymer Chemistry Edition, 1969, 7, 6, 1537. 142. F.C. Thryion and M.D. Baijal, Journal of Polymer Science, Polymer Chemistry Edition, 1968, 6, 3, 505. 143. L.R. Lerner, Journal of Polymer Science, Polymer Chemistry Edition, 1974, 12, 11, 2477. 144. K. Hiraki, T. Inoue and H. Hirai, Journal of Polymer Science, Polymer Chemistry Edition, 1970, 8, 9, 2543. 145. H. Hirai, K. Hiraki, I. Noguchi, T. Inoui and S. Makishima, Journal of Polymer Science, Polymer Chemistry Edition, 1970, 8, 9, 2393. 146. K. Takeda, H. Yoshida, K. Hayashi and S. Okamura, Journal of Polymer Science, Polymer Chemistry Edition, 1966, 4, 10, 2710. 147. L. Wiechec, Analytical Chemistry (Warsaw), 1973, 18, 853.

350

Types of Unsaturation 148. J.M. Vigneron and A.R. Deschreider, Lebensmittel-Wissenschaft und Technologie, 1972, 5, 198. 149. B. Eda, K. Numone and M. Iwasaki, Journal of Polymer Science, Polymer Chemistry Edition, 1970, 8, 7, 1831. 150. O. Tanaka, Journal of Polymer Science, A-1, 1973, 11, 2069. 151. J. Pilar, L. Toman and M. Marek, Journal of Polymer Science, Polymer Chemistry Edition, 1976, 14, 10, 2399. 152. R.V. Albarino, E.P. Otocka and J.P. Luongo, Journal of Polymer Science, Polymer Chemistry Edition, 1977, 9, 6, 1517. 153. M.J. Hackathorne and M.J. Brock, Journal of Polymer Science, Polymer Chemistry Edition, 1975, 13, 4, 945. 154. J. Furukawa, K. Haga, E. Kobayashi, Y. Iseda, T. Yoshimoto and K. Sakamato, Polymer Journal, 1971, 2, 3, 371. 155. R. Hill, J.R. Lewis and J.L. Simonsen, Transactions of the Faraday Society, 1939, 35, 1067. 156. E.N. Alekseeva, Journal of the General Chemistry of the USSR, 1941, 11, 353. 157. N. Rabjohn, C.E. Bryan, G.E. Inskeep, H.W. Johnson and J.K. Lawson, Journal of American Chemical Society, 1947, 69, 2, 314. 158. A.I. Yakubehik, A.J. Spaaskova, A.G. Zak and I.D. Shotatskaya, Zhurnal Obshchei Khimii, 1958, 28, 3080. 159. A. Kawasaki in Proceedings of the 27th Autumn Meeting of the Japan Chemical Society, 1972, Pre-prints Volume 2, p.20. 160. M.J. Hackathorn and M.K. Brock, Rubber Chemistry and Technology, 1972, 45, 15, 295.

351


352

10

Polymer Branching

Branching is another aspect of polymer microstructure which is of great interest. Polyethylene, for example, can contain side-chain alkyl groups ranging from methyl to octyl or even higher, which can be identified and determined by techniques such as nuclear magnetic resonance (NMR) and infrared (IR) spectroscopy. Such groups often have a profound effect on the physical properties of polymers, and different types of alkyl side-groups account for the fact that many different grades of this polymer are available, each with its own particular physical properties. Along with chemical composition, molecular weight, molecular weight distribution and type and amount of gel, branching is considered to be one of the fundamental parameters needed to characterise polymers fully, and this latter property, which is a microstructural feature of the polymer, has very important effects on polymer properties. Changes in branching of a given polymer such as polypropylene lead to changes in its stereochemical configuration and this, in turn, is a fundamental polymer property to formulating polymer physical characteristics and mechanical behaviour. Identifying the type and amount of side-group branching in polymers is important. Although molecular symmetry is well understood, until the development of proton NMR, and later 13C-NMR, a study of this aspect of polymer structure presented many problems. Low-density polyethylene, for example, has a very complex molecular structure despite the fact that it consists of only one monomeric unit, i.e., ethane. Low-density polyethylene exhibits a broad molar mass distribution. As a result of intra- and inter-molecular chain-transfer reactions, short- and long-chain branching exists [1–4]. Revelation of the chain structure is increasingly recognised as a prerequisite because reaction kinetics determine the structure, which in turn determines the properties. Thus, knowledge of the structure supplies not only information about the reaction kinetics, but also about the relation between chain structure and properties. In the case of low-density polyethylene, the most pronounced chain feature is the long-chain branching which can occur in various chain architectures, i.e., comb or random [1, 2, 5, 6]. Revelation of the type as well as the number of long-chain branching is still difficult.

353

Introduction to Polymer Analysis The quantitative analysis of branching in polyethylene has been the subject of much investigation [7–11]. Several techniques have been applied to a study of polymer chain branching, predominantly NMR spectroscopy and to a lesser extent, IR spectroscopy. Some applications of each of these techniques are reviewed below.

10.1 IR Spectroscopy 10.1.1 Methyl Branching in Polyethylene A regression analysis of IR, differential thermal analysis and X-ray diffraction data by Laiber and co-workers [12] for high-density polyethylene showed that, as synthesis conditions varied, the number of methyl groups varied from 0 to 15 per 1000 carbon atoms and the degree of crystallinity varied from 84% to 61%. Saturated hydrocarbons evolved during electron irradiation of polyethylene are characteristic of short side-chains in the polymer. Salovay and Pascale [13] showed that a convenient analysis is enabled by programmed temperature gas chromatography. To minimise the relative concentrations of extraneous hydrocarbons, i.e., those not arising from selective scission of complete side-chains, it is necessary to irradiate at low temperatures and doses. Such analyses of low-density polyethylene indicate that the 2–3 methyls per 100 carbon atoms detected in IR absorption (high-density polyethylenes at least one order of magnitude lower) are probably equal amounts of ethyl and butyl branches. These arise by intramolecular chain transfer during polymerisation. At a dose of 10 Mrad, about 1–4% of the alkyl group are removed. Methane is the only hydrocarbon detected on irradiation of polypropylene, indicating little combination of methyl radicals to form ethane during irradiation. Measurement of methyl absorption at 1378 cm–1 (7.26 μm) in polyethylene can serve as a good estimation of branching. Interference from the methylene absorption at 1368 cm–1 (7.31 μm) makes it difficult to measure the 1368 cm–1 (7.31 μm) band, especially in the case of relatively low methyl contents. A method has been developed which utilises a suggestion by Neilson and Holland [14]. They associated the amorphous phase absorption of polyethylene at 1368 cm–1 (7.31 μm) and 1304 cm–1 (7.69 μm) with the trans–trans conformation of the polymer chain about the methylene group. Therefore the intensities of these two absorptions are proportional to one another. By placing an annealed film (approximately 254-381 μm) of high-density polyethylene in the reference beam of a double beam spectrometer and a thin, quenched film of the sample in the sample beam, most of the interference at 1368 cm–1 (7.31 μm) can be removed. The method has the advantage that it is not necessary to have complete compensation for the 1368 cm–1 (7.31 μm) band because

354

Polymer Branching a correction for non-compensation at 1378 cm–1 (7.25 μm) can be applied based on the intensity of the 1368 cm–1 (7.31 μm) absorption. Calibration for the methyl absorption based on mass spectrometric studies of such gaseous products produced during electron bombardment of polyethylene has demonstrated irradiation-induced detachment of complete alkyl units [15]. In addition to saturated alkanes characteristic of the branches, small quantities of methane, other paraffins, and olefins were simultaneously evolved. It was suggested that ‘extraneous’ paraffins result from cleavage of the main chain [15, 16]. Nerheim [17] has described a circular calibrated polymethylene wedge for the compensation of CH2 interferences in the determination of methyl groups in polyethylene by IR spectroscopy. Methyl-group content of low-density polyethylene has been determined with a standard deviation of 0.8% provided methylene group absorptions were compensated by polyethylene of similar structure [18]. De Pooter and co-workers [19] describe an IR method which utilises the absorbance of the methyl group at about 1380 cm–1 (7.25 μm) for the determination. This method suffers from limitations, namely, that the absorbance must be corrected due to interferences of the methylenes and other bands. The absorbance frequency and absorptivity of the methyl groups are also dependent upon the type of branch and upon crystalinity [20]. This presents a problem for the quantitative analysis of branching in ethylene copolymers of two or more comonomers. The IR method has some distinct advantages over the 13C-NMR method, including precision and analysis time. Therefore, there is a need to provide well-defined and accepted standards for this analysis. NMR is an absolute method, not requiring standards, and specificity, because the location of the resonance identifies it as being from a given type of branch. Branches shorter than six carbons in length can be unambiguously assigned from their 13CNMR spectrum. Branches longer than five carbons in length cannot be differentiated from long-chain branches [21]. Therefore, the advantages of NMR, accuracy and specificity, can be utilised to define standard materials which can then be used to standardise the IR method.

10.2 NMR Spectroscopy The advantages of 13C-NMR in measurements of polymer stereochemical configuration arise primarily from a useful chemical shift range which is approximately 20 times that obtained by proton NMR. Structural sensitivity is enhanced through well-separated resonances for different types of carbon atoms. Overlap is generally not a limiting problem. The low natural abundance (–1%) of 13C nuclei is another favourable

355

Introduction to Polymer Analysis contributing factor. Spin–spin interactions among 13C nuclei can be safely neglected, and proton interactions can be eliminated entirely through heteronuclear decoupling. Thus each resonance in a 13C-NMR spectrum represents the carbon chemical shift of a particular polymer moiety. In this respect, 13C-NMR resembles mass spectrometry because each signal represents some fragment of the whole polymer molecule. Carbon chemical shifts are ‘well behaved’ from an analytical viewpoint because each can be dissected, in a strictly additive manner, into contributions from neighbouring carbon atoms and constituents. This additive behaviour led to the Grant and Paul rules [22], which have been carefully applied in polymer analyses for predicting alkane carbon chemical shifts. The advantages so clearly evident when applying 13C-NMR to polymer configurational analyses are not devoid of difficulties. The sensitivity of 13C-NMR to subtle changes in molecular structure crates a wealth of chemical shift-structural information which must be ‘sorted out’. Extensive assignments are required because the chemical shifts relate to sequences from three to seven units in length. Model compounds, which are often used in 13C-NMR analyses, must be very close structurally to the polymer moiety reproduced. For this reason, appropriate model compounds are difficult to obtain. A model compound found useful in polypropylene configurational assignments with a heptamethylheptadecane, where the relative configurations were known [23]. To be completely accurate, the model compounds should reproduce the conformational as well as the configurational polymer structure. Thus, reference polymers such as predominantly isotactic and syndiotactic polymers form the best model systems. Even when available, only two assignments are obtained from these particular polymers. Pure reference polymers can be used to generate other assignments [24]. To obtain good quantitative 13C-NMR data, one must understand the dynamic characteristics of the polymer under study. Fourier transform techniques, combined with signal averaging, are normally used to obtain 13C-NMR spectra. Equilibrium conditions must be established during signal averaging to ensure that the experimental conditions have not led to distorted spectral information. The nuclear Overhauser effect (NOE), which arises from 1H, 13C heteronuclear decoupling during data acquisition, must also be considered. Energy transfer, occurring between the 1H and 13C nuclear energy levels during spin decoupling, can lead to enhancements of the 13C resonances by factors between 1 and 3. Thus, the spectral relative intensities will reflect only the polymer’s moiety concentrations if the differentiated NOE are equal or else taken into consideration. Experience has shown that polymer NOE are generally maximal, and consequently equal, because of a polymer’s restricted mobility [25, 26]. To be sure, one should examine the polymer NOE through gated decoupling or paramagnetic quenching, and thereby avoid any misinterpretation of the spectral intensity data.

356

Polymer Branching The 13C configurational sensitivity falls within a range from triad to pentad for most vinyl polymers. In non-crystalline polypropylenes, three distinct regions corresponding to methylene (–46 ppm), methine (–28 ppm) and methyl (–20 ppm) carbons are observed in the 13C-NMR spectrum. (Chemical shifts are reported with respect to an internal tetramethylsilane (TMS) standard.) The 13C spectrum of a 1,2,4trichlorobenzene solution at 125 °C of a typical amorphous polypropylene is shown in Figure 10.1. Although configurational sensitivity is shown by all three spectral regions, the methyl region exhibits by far the greatest sensitivity and is consequently of the most value. At least ten resonances, assigned to the unique pentad sequences, are observed in the order, mmmm, mmmr, rmmr, mmrr, mmrm, rmrr, mrmr, rrrr, rrrm and mrrm from low to high field [22–27]. mm

CH3 rm

21.82

21.03

rr

20.30

CH

28.92 28.47 CH2

47.16

46.50

Figure 10.1 Methyl, methine and methylene regions of the 13C-NMR spectrum of a non-crystalline polypropylene. Source: Author’s own files

10.2.1 Ethyl and Higher Alkyl Groups Branching in Polyethylene High-density (low-pressure) polyethylenes are usually linear, although the physical and rheological properties of some high-density polyethylenes have suggested longchain branching (butyl and higher groups) at a level one to two orders of magnitude below that found for low-density polyethylenes prepared by a high-pressure process. A measurement of long-chain branching in high-density polyethylenes has been elusive because of the low concentrations involved [28] and can only be directly provided by high-field, high-sensitivity NMR spectrometers.

357

Introduction to Polymer Analysis High-density polyethylenes prepared with a Ziegler type, titanium-based catalyst have predominantly n-alkyl or saturated end groups. Those prepared with chromium-based catalysts have a propensity toward more olefinic end groups. The ratio of olefinic to saturated end groups for polyethylenes prepared with chromium-based catalysts is approximately unity. The end-group distribution is therefore another structural feature of interest in low-pressure polyethylenes because it can be related to the catalyst employed, and possibly to the extent of long-chain branching. It is possible not only to measure by 13C-NMR concentrations of saturated end groups, but also the olefinic end-groups and, subsequently, an end-group distribution (see Chapter 8). Nishoika and co-workers [29] determined the degree of chain branching in low-density polyethylene using proton Fourier transform NMR at 100 MHz and 13C Fourier transform NMR at 25 MHz with concentrated solutions at approximately 100 °C. Methyl concentrations obtained agreed well with those obtained by IR based on the absorbance at 1378 cm–1 (7.25 μm). Bugada and Rudin [30] combined exclusion chromatography with determine long-chain branching in low-density polyethylene.

13

C-NMR to

The 13C-NMR spectrum of a crystalline polypropylene shown in Figure 10.2 contains only three lines which can be identified as methylene, methine and methyl from low to high field by off-resonance decoupling. An amorphous polypropylene exhibits a 13 C spectrum which contains not only these three lines, but additional resonances in each of the methyl, methine and methylene regions (Figure 10.2). The crystalline polypropylene must therefore be characterised by a single type of configurational structure. In this case, the crystalline polypropylene structure is predominantly isotactic, thus the three lines in Figure 10.2 must result from some particular length of meso sequences. This sequence length information is not available from the spectrum of the crystalline polymer, but can be determined from a corresponding spectrum of the amorphous polymer. To do so one must examine the structural symmetry of each carbon atom to the various possible monomer sequences. Randall [31, 34] carried out a detailed study of the polypropylene methyl group in triad and pentad configurational environments. Stehling [33] also studied polypropylene. The study of stereochemical configuration by 13C-NMR has not been limited to the polyolefins.

10.2.2 Branching in Ethylene–propylene Copolymers In this section, we discuss the occurrence of side groups in ethylene copolymers ranging from ethylene–propylene to ethylene–octane. In the case of ethylene–propylene copolymers it is possible by 13C-NMR to determine the methyl side-groups due to propylene units.

358

Polymer Branching

CH3 CH

CH2

HMDS

50

40

30 20 TMS (ppm)

10

0

Figure 10.2 13C-NMR spectrum at 25.2 MHz of crystalline polypropylene. Reproduced with permission from D.C. Bugada and A. Rudin, European Polymer Journal, 1987, 847. © 1987, Elsevier

5 7

CH2 1

CH3

CH2 3

2

50

40

CH 4

CH2 6

8

HMDS

9

30 20 , ppm (TMS)

10

0

Figure 10.3 13C-NMR spectrum of ethylene-polypropylene copolymer containing 97% propylene in isotactic sequences. Reproduced with permission from J.R. Paxson and J.C. Randall, Analytical Chemistry, 1978, 50, 1777. © 1978, ACS

359

Introduction to Polymer Analysis As shown by a typical example in Figure 10.3, each 13C-NMR spectrum was recorded with proton noise-decoupling to remove unwanted 13C–1H scalar couplings. No corrections were made for differential NOE because constant NOE were assumed [25, 34], in agreement with previous workers [35–37]. Constant NOE for all major resonances in low ethylene content ethylene–propylene copolymers have been reported [36]. The 13C-NMR spectrum of an ethylene–propylene copolymer, containing approximately 97% propylene in primarily isotactic sequences, is shown in Figure 10.3. Major resonances are numbered consecutively from low to high field. Chemical shift data and assignments are listed in Table 10.1. Greek letters are used to distinguish the various methylene carbons and designate the location of the nearest methine carbons. Paxson and Randall [38] in their method use the reference chemical shift data obtained on a predominantly isotactic polypropylene and on an ethylene–propylene copolymer (97% ethylene). They concluded that the three ethylene–propylene copolymers used in their study (97–99% propylene) contained principally isolated ethylene–ethylene linkages. Knowing the structure of their three ethylene–propylene copolymers, they used the 13C-NMR relative intensities to determine ethylene–propylene contents and thereby establish reference copolymers for the faster IR method involving measurements at 732 cm–1 (13.66 μm). After a detailed analysis of resonances Paxson and Randall [38] concluded that methine resonances 4 and 5 (Table 10.1) gave the best quantitative results to determine the comonomers composition. The composition of the ethylene–propylene copolymers was determined by peak heights using the methine resonances only. In no instance was there any evidence for an inclusion of consecutive ethylene units. Thus, composition data from 13C-NMR could now be used to establish an IR method based on a correlation with the 732 cm–1 (13.66 μm) band which is attributed to a rocking mode, r, of the methylene trimer, –(CH2)3–. Randall [39] developed a 13C-NMR quantitative method for measuring ethylene– propylene mole fractions and methylene number-average sequence lengths in ethylene–propylene copolymers. He views the polymers as a succession of methylene and methyl-branched methine carbons, as opposed to a succession of ethylene and propylene units. This avoids problems associated with propylene inversion and comonomers sequence assignment. He gives methylene sequence distributions from one to six and larger consecutive methylene carbons for a range of ethylene–propylene copolymers, and uses this to distinguish copolymers which have random, blocked, or alternating comonomers sequences.

360

Polymer Branching

Table 10.1 Observed and reference 13C-NMR chemical shifts in ppm for ethylene-propylene copolymers and reference polypropylenes with respect to an internal trimethylsilane standard Line

Carbon

E/P (3/97)

E/P [37]

1

AA-CH2

46.4

2

AA-CH2

3

Sequence assignment

Reference crystal [25]

PP amorphous [25]

46.3

PPPP

46.5

47.0–47.5 r 46.5

46.0

45.8

PPPE

AA-CH2

37.8

37.8

PPEP

4

CH

30.9

30.7

PPE

5

CH

28.8

28.7

PPP

28.5

28.8mmmm 28.6 mmmr 28.5 rmmr 28.4 mr + rr

6

BBCH2

24.5

24.4

PPEPP

7

CH3

21.8

21.6

21.8

21.3–21.8 mm 20.6–210 mr 19.9–20.3 rr

8

CH3

21.6

21.4

9

CH3 CH3 CH ACH2 BCH2 ─(CH2)n-

20.9

20.7 19.8 33.1

29.8

E/P (97/3

P

PPPPP

PPPE 19.8 33.1 37.4 27.3 29.8

PPPEP EPE EPE EPE EPE EEE

E/P: ethylene/propylene Reproduced with permission from J.R. Paxson and J.C. Randall, Analytical Chemistry, 1978, 50, 13, 1777. © 1978, ACS

10.2.3 Branching Ethylene–Higher Olefin Copolymers Short-chain branches can be introduced in a controlled manner into polyethylenes by copolymerising ethylene with a 1-olefin. The introduction of 1-olefins allows the density to be controlled, and butane-1 and hexane-1 are commonly used for this purpose. As in the case of high-pressure process low-density polyethylenes, 13CNMR can be used to measure ethyl and butyl branch concentrations independently of the saturated end groups. This result gives 13C-NMR a distinct advantage over corresponding IR measurements because the latter technique can only detect methyl groups irrespective of whether the methyl group belongs to a butyl branch or a chain

361

Introduction to Polymer Analysis end. 13C-NMR also has a disadvantage in branching measurements because only branches five carbons in length and shorter can be discriminated independently of longer-chain branches [40, 41]. Branches six carbons in length and longer give rise to the same 13C-NMR spectral pattern independently of the chain length. This lack of discrimination among the longer side-chain branches is not a deterring factor in the usefulness of 13C-NMR in a determination of long-chain branching. By far the most difficult structural measurement is long-chain branching. In lowdensity polyethylenes, the concentration of long-chain branches in such (

Introduction to Polymer Analysis

Introduction to Polymer Viscoelasticity

Introduction to polymer viscoelasticity

Introduction to polymer dynamics

Introduction to polymer spectroscopy

Introduction to polymer viscoelasticity

Polymer Analysis Polymer Theory

Polymer Analysis Polymer Theory

Polymer Analysis Polymer Physics

Polymer Analysis

Introduction to Physical Polymer Science

An Introduction to Polymer Physics

An Introduction To Polymer Physics

An Introduction to Polymer Physics

Introduction to Physical Polymer Science

Introduction to physical polymer science

An Introduction to Polymer Physics

Introduction To Physical Polymer Science

An Introduction to Polymer Physics

Introduction to environmental analysis

Introduction to real analysis

Introduction to Modern Analysis

Introduction to Matrix Analysis

Introduction to Functional Analysis

Introduction to numerical analysis

Introduction to analysis

Introduction to Numerical Analysis

Introduction to real analysis

Introduction to numerical analysis

Introduction to Real Analysis

Introduction to Numerical Analysis

Introduction to Polymer Analysis

Introduction to Polymer Viscoelasticity

Introduction to polymer viscoelasticity

Introduction to polymer dynamics

Introduction to polymer spectroscopy

Introduction to polymer viscoelasticity

Polymer Analysis Polymer Theory

Polymer Analysis Polymer Theory

Polymer Analysis Polymer Physics

Polymer Analysis

Introduction to Physical Polymer Science

An Introduction to Polymer Physics

An Introduction To Polymer Physics

An Introduction to Polymer Physics

Introduction to Physical Polymer Science

Introduction to physical polymer science

An Introduction to Polymer Physics

Introduction To Physical Polymer Science

An Introduction to Polymer Physics

Introduction to environmental analysis

Introduction to real analysis

Introduction to Modern Analysis

Introduction to Matrix Analysis

Introduction to Functional Analysis

Introduction to numerical analysis

Introduction to analysis

Introduction to Numerical Analysis

Introduction to real analysis

Introduction to numerical analysis

Introduction to Real Analysis

Introduction to Numerical Analysis

Recommend Documents