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

554 downloads 2584 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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Types and Properties of Polymers

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

Introduction to Polymer Analysis

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

Types and Properties of Polymers

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

Determination of Elements

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

Determination of Elements

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

Determination of Elements

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

Introduction to Polymer Analysis

(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

Determination of Elements

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

Introduction to Polymer Analysis

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

Determination of Elements

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

Introduction to Polymer Analysis

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

Determination of Elements

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 (<1 kW) plasma as an atomiser.

2.2.1.7 Inductively Coupled Plasma Optical Emission Spectrometry–Mass Spectrometry (ICP-MS) ICP-MS combines the established ICP to break the sample into a stream of positively charged ions which are subsequently analysed on the basis of their mass. ICP-MS does not depend on indirect measurements of the physical properties of the sample. The elemental concentrations are measured directly; individual atoms are counted, giving the key attribute of high sensitivity. The technique has the additional beneﬁt of unambiguous spectra and the ability to directly measure isotopes of the same element. The sample under investigation is introduced (usually in solution) into inductively coupled plasma at atmospheric pressure and a temperature of approximately 5700 oC. Sample components are rapidly dissociated and ionised, and the resulting atomic ions introduced via a carefully designed interface into a high-performance quadrupole mass spectrometer at high vacuum. A horizontally mounted ICP torch forms the basis of the ion source. Sample introduction is via a conventional nebuliser optimised for general-purpose solution analysis and suitable for use with aqueous and organic solvents. Nebulised samples enter the central channel of the plasma as a ﬁnely dispersed mist which is rapidly vaporised; dissociation is virtually complete during passage through the plasma core with most elements fully ionised. Ions are extracted from the plasma through a water-cooled sampling aperture. A molecular beam is formed in the ﬁrst vacuum stage and passes into the high-vacuum stage of the quadrupole mass analyser. In an ICP-MS system a compact quadrupole mass analyser selects ions on the basis 29

Introduction to Polymer Analysis of their mass-to-charge ratio (m/e). The quadrupole is a simple compact form of mass analyser that relies on a time-dependent electric ﬁeld to ﬁlter the ions according to their m/e. Ions are transmitted sequentially in order of their m/e with constant resolution across the entire mass range. This technique has been applied to the analysis of aqueous digests of polymers containing up to 2% solids. Dobney and co-workers [60] used laser ablation ICP-MS to determine various metals in polyoleﬁns.

2.2.1.8 Pre-concentration AAS Techniques Detection limits can be further improved in the case of all atomic absorption techniques by a pre-concentration technique [61]. One technique that has found great favour involves converting the metals to an organic chelate by reaction of a larger volume of sample with a relatively small volume of an organic solvent solution, commonly of diethyldithiocarbamates or ammonium pyrrolidone diethiocarbamates. The chelate dissolves in the organic phase and is then back-extracted into a small volume of aqueous acid for analysis by either of the previously mentioned techniques. If 0.1–1.0 litre of sample is originally taken and 20 ml of acid extract ﬁnally produced, then concentration factors of 25–50 are thereby achieved with consequent lowering of detection limits. This additional step considerably increases analysis time and necessitates extremely careful control of experimental conditions. Microscale solvent extractions involving extraction of 2.5 ml sample with 0.5 ml of an organic solvent solution of a chelate give detection limits for lead and cadmium by the Zeeman graphite furnace AAS method of 0.6 and 0.02 μg/l, respectively. This is equivalent to determining 1.2 ppm lead in polymers (assuming the digest of 10 mg of polymer is made up to 20 ml).

2.2.1.9 Applications of Techniques Detailed So Far Two types of catalysts used in polymer manufacture are metallic compounds such as aluminium alkyls and titanium halides used in low-pressure polyoleﬁn manufacture. Because residual catalysts can have important effects on polymer properties, determining trace elements which reﬂect these substances is important.

30

Determination of Elements

2.2.1.9.1 Elemental Analysis of Polymers Elements occurring in polymers and copolymers can be divided into three categories: •

Elements that are a constituent part of the monomers used in polymer manufacture, such as nitrogen in acrylonitrile used in the manufacture of, for example, acrylonitrile–butadiene–styrene terpolymers.

•

Elements that occur in substances deliberately included in polymer formulations, e.g., zinc stearate.

•

Elements that occur as adventitious impurities in polymers. For example, during the manufacture of PE by the low-pressure process, polymerisation catalysts such as titanium halides and organo-aluminium compounds are used, and the ﬁnal polymer contains traces of aluminium, titanium, and chlorine residues.

The classic destructive techniques are generally based on one of three approaches to the analysis: (i) dry ashing of the polymer with or without an ashing aid, followed by acid digestion of the residue, alternatively acid digestion of the polymer without prior ashing; (ii) fusion of the polymer with an inorganic compound to effect solution of the elements; and (iii) bomb or oxygen ﬂask digestion techniques. Another method for avoiding losses of metals during ashing is the low-temperature controlled decomposition method using active oxygen. This has been studied in connection with the determination of trace metals in polyvinyl chloride (PVC), polypropylene (PP), and polyethylene terephthalate [62].

2.2.1.9.2 Trace Metals in Polymers Sources of trace metals in polymers are neutralising chemicals added to the ﬁnal stages of manufacture to eliminate the effects of acidic catalyst remnants on polymer processing properties (e.g., hygroscopicity due to residual chloride ion). A case in point is high-density polyethylene (HDPE) and PP produced by the aluminium alkyl–titanium halide route which is treated with sodium hydroxide in the ﬁnal stages of manufacture. A technique that involves combustion of the polymer under controlled conditions in a platinum crucible, followed by dissolution of the residual ash in a suitable aqueous reagent before ﬁnal analysis by spectrophotometry, is of limited value. A quite complicated and lengthy ashing programme is necessary in this technique to avoid losses of alkali metal during ignition: 0–1 hour from start: heat to 200 °C; 1–2

31

Introduction to Polymer Analysis hours from start: hold at 200 °C; 3–5 hours from start: heat to 450 °C; 5–8 hours from start: hold at 450 °C. After ignition, the residue is dissolved in warm nitric acid and made up to a standard volume before evaluation by ﬂame photometry or AAS. Alternatively, the polymer is ashed overnight at 500 °C with sulfur and a magnesium salt of a long-chain fatty acid (Magnesium AC dope, Shell Chemical Company Limited) and the ash mixed with twice its weight of carbon powder containing 0.1% palladium before emission spectrographic evaluation of the sodium/palladium 330.3/276.31 line pair. Table 2.3 clearly shows that ﬂame photometry after dope ashing at 500 °C gives a quantitative recovery of sodium relative to results obtained by a non-destructive method of analysis, i.e., neutron activation analysis (NAA). Direct ashing without the magnesium ashing aid at 500 °C causes losses of r10% of the sodium, whereas direct ashing at 800 °C causes even greater losses. Dry ashing in platinum gives reasonably good results for the determination of low concentrations of vanadium in an ethylene–propylene copolymer. Ten grammes of polymer is ashed in platinum by charring on a hot plate followed by heating over a Meker burner. Dilute nitric acid is added to the residue and residue in the crucible dissolved by fusion with potassium persulfate. The vanadium is determined spectrophotometrically by the 3,3-diaminobenzene method. Results obtained by this method are in very good agreement with those obtained by NAA, which in this case can be considered to be an accurate reference method. Good agreement is obtained between the two methods for samples containing vanadium.

Table 2.3 The effects of modiﬁcation of ashing procedure on the ﬂame photometric determination of sodium in polyethylenes Sodium by ﬂame photometry (ppm) Sample

By neutron activation

By emission spectrography

Original (ashed between 650 oC and 800 oC)

Dope ash at 500 oC

Diect ash at 500 oC

1

99, 96, 99

95

60, 76, 55

100

75

2

256, 247, 259

258, 259

160, 178, 271

225

208

3

343, 321, 339

339, 287

250, 312

282

265

4

213, 210, 212

218, 212

140, 196

210

191

5

194, 189, 192

209, 198

80, 158, 229

196

169

6

186, 191, 198

191, 191

96, 173

193

173

Source: Author’s own ﬁles.

32

Determination of Elements It has been shown [63, 64] in studies using a radioactive copper isotope that, when organic materials containing copper are ashed, losses of up to some 10% of the copper occur due to retention in the silica crucible; this could not be removed by acid washing. Virtually no retention of copper in the silica crucible occurred when copper was ashed under the same conditions in the absence of added organic matter. This was attributed to reduction of copper to the metal by the organic matter present, followed by partial diffusion of the copper metal into the crucible wall. Distinctly higher copper determinations are obtained for polyoleﬁns by the procedure involving a magnesium oxide ashing aid than are obtained without an ashing aid, or by a molten potassium bisulfate fusion technique to take up polymer ash. Henn [65] reported on a ﬂameless atomic absorption technique with solid sampling for determining trace amounts of chromium, copper and iron in polymers such as polyacrylamide with a detection limit of approximately 0.01 ppm. AAS is a useful technique for the determination of traces of metals in polymers. In general, the polymer is ashed at a maximum temperature of 450 °C: 0.1 hour from start: heat to 200 °C; 1–3 hours from start: hold at 200 °C; 3–5 hours from start: heat to 450 °C; 5–8 hours from start: hold at 450 °C. Ash is digested with warm nitric acid before spectrometric analysis. Detection limits for metals in polymers achievable by this procedure range from 0.02 ppm (zinc) to 0.57 ppm (iron). Arsenic, antimony, mercury, selenium, and tin can, after producing the soluble digest of the polymer, be converted to gaseous metallic hydrides by reaction of the digest with stannous chloride or sodium borohydide:

As2O 3 + 3SnCl2 + 6HCl = 2AsH3 + 3H2O + 3SnCl4 NaBH4 + 2H2O = NaBO 2 + 4H2 6H2 + As2O 3 = 2AsH3 + 3H2O These hydrides can be determined by AAS. To illustrate, let us consider a method developed for the determination of trace amounts of arsenic in acrylic ﬁbres containing antimony oxide ﬁre-retardant additive [66]. Arsenic occurs as an impurity in the antimony oxide additive and its concentration must be controlled at a low level. In this method, a weighed amount of sample is digested with concentrated nitric acid and perchloric acid and digested until the sample is completely dissolved. Pentavalent arsenic in the sample is reduced to trivalent arsenic by the addition of titanium trichloride dissolved in concentrated hydrochloric acid:

33

Introduction to Polymer Analysis Trivalent arsenic is separated from antimony by extraction with benzene, leaving antimony in the acid layer. Trivalent arsenic is extracted with water from the benzene phase. The solution is extracted with a mixture of hydrochloric acid, potassium iodide, and stannous chloride to convert trivalent arsenic to arsine (AsH3), which is swept into the AAS. Arsenic is then determined at the 193.7 nm absorption line. Recoveries of 96–104% are obtained by this procedure in the 0.5–1.0 μg arsenic range, with a detection limit of 0.04 ppm.

2.2.1.10 Pressure Dissolution Technique Two main methods have been used to digest polymer samples before the determination of metals: (i) pressure dissolution in sealed Teﬂon lined steel bombs; or (ii) dissolution in sealed bombs in a microwave oven. In Teﬂon lined steel bomb pressure dissolution digestion with nitric acid, at 200 °C (80 °C over the atmospheric boiling point) and 0.7 MPa can be achieved in 12 minutes, and for hydrochloric acid at 135 °C (43 °C over the atmospheric boiling point) and 0.7 MPa can be obtained in 5 minutes. The aggressive digestion action produced at the higher temperatures and pressures generated in these bombs results in remarkably short digestion times, with many materials requiring less than one minute to obtain complete dissolution, i.e., considerably quicker than open-tube wet ashing or acid digestion procedures. Combustion with oxygen in a sealed Parr bomb has been accepted for many years as a standard method for converting solid combustible samples into soluble forms for chemical analysis. It is a reliable method whose effectiveness stems from its ability to treat samples quickly and conveniently within a closed system without losing any of the sample or its combustion products. Sulfur-containing polymers are converted to soluble forms and absorbed in a small amount of water placed in the bomb. Halogen-containing polymers are converted to hydrochloric acid or chlorides. Mineral constituents remain as ash but other elements such as arsenic, boron, mercury, nitrogen and phosphorus, and all of the halogens are recovered with the bomb washings. In recent years, the list of applications has been expanded to include metals such as beryllium, cadmium, chromium, copper, iron, lead, manganese, nickel, vanadium and zinc by using a quartz liner to eliminate interference from trace amounts of heavy metals leached from the bomb walls and electrodes [69–71]. Once the sample is in solution in the acid and the digest made up to a standard volume, the determination of metals is completed by standard procedures such as AAS and ICP-OES.

34

Determination of Elements Microwave ovens have also been used for polymer dissolution. The sample is sealed in a Teﬂon bottle or a specially designed microwave digestion vessel with a mixture of suitable acids such as nitric acid, and aqua regia and, occasionally, hydroﬂuoric acid. Perchloric acid must not be used in these bombs due to the risk of explosion. The high-frequency microwave temperature (~100–250 °C) and increased pressure have a role to play in the success of this technique. An added advantage is the signiﬁcant reduction in sample dissolution time [67, 68]. Table 2.4 shows results obtained for the microwave digestion in closed vessels of 1 g samples digested in (a) 20 ml of 1:1 nitric acid:water and (b) 5 ml of concentrated nitric acid and 3 ml of 30% hydrogen peroxide. In the former, at a power input of 450 W, the temperature and pressure rose to 180 °C and 0.7 MPa. At that point, microwave power was reduced to maintain the temperature and pressure at those values for an additional 50 minutes. In the latter case, 1 g samples were open-vessel digested in 1:1 nitric acid:water for 10 minutes at 180 W. After cooling to room temperature, 5 ml of concentrated nitric acid and 3 ml of 30% hydrogen peroxide were added to each. Vessels were then sealed and power applied for 15 minutes and at 152 °C at 0.3 MPa after the ﬁnal 15 minutes of heating. With both reagent systems, element recoveries are in good agreement with the values obtained using a hot plate total sample digestion technique, which typically requires 4–6 hours.

Table 2.4 Solid sample microwave digested in 1:1 HNO3:H2O Element

(a) in HNO3:H2O (1:1)

(b) in HNO3:H2O2 (5:3)

Certiﬁed value (%)

Amount recovered (%)

Amount recovered (%)

As

0.0060, 0.0060

0.0075, 0.0070

0.0066

Cd

0.0012, 0.0012

0.0011, 0.0012

0.0012 ± 0.00015

Cr

3.00, 2.98

3.04, 2.96

2.96 ± 0.28

Cu

0.0122, 0.0113

0.0118, 0.0119

0.0109 ± 0.0019

Mg

0.72, 0.72

0.70, 0.70

0.74 ± 0.02

Mn

0.0790, 0.0780

0.0720, 0.725

0.0785 ± 0.0097

Ni

0.0050, 0.0050

0.0044, 0.0044

0.00458 ± 0.00029

Pb

0.0736, 0.0737

0.0736, 0.0733

0.0714 ± 0.0028

Se

0.0001, 0.0001

0.0001, 0.0001

(0.00015)

Zn

0.170, 0.168

0.160, 0.160

0.172 ± 0.017

Source: Author’s own ﬁles

35

Introduction to Polymer Analysis Flame and GFAAS techniques have adequate sensitivity for the determination of metals in polymer samples. In this technique, up to 1 g of dry sample is digested in a microwave oven for a few minutes with 5 ml of aqua regia in a small PTFE-lined bomb, and then the bomb washings are transferred to a 50 ml volumetric ﬂask before analysis by ﬂame AAS. Detection limits (mg/kg) achieved by this technique were: 0.25 (cadmium, zinc); 0.5 (chromium, manganese); 1 (copper, nickel, iron); and 2.5 (lead). Application of this technique gave recoveries ranging between 85% (cadmium) and 101% (lead, nickel, iron) with an overall recovery of 95%.

2.2.1.11 Visible and UV Spectroscopy The theory of visible and UV spectroscopy is discussed in an publication from Her Majesty’s Stationery Ofﬁce (HMSO) [72]. Visible spectrophotometry is used extensively in the determination of some anions such as chloride, phosphate and sulfate formed by the decomposition of chlorine, phosphorus and sulfur in polymers. An extensive modern application of visible spectrophotometry is in the determination of organic substances, including non-ionic detergents, in polymer extracts. Some commercially available instruments, in addition to visible spectrophotometry, can also carry out measurements in the UV and near-infrared regions of the spectrum.

2.2.1.12 Polarography and Voltammetry A large proportion of trace metal analysis carried out in polymer laboratories is based on AAS and ICP-AES. Both of these methods give estimates of the total concentration of metal present and do not distinguish between different valency states of the same metal. For example, they could not distinguish between arsenic and antimony in the tri- or pentavalent state in water extracts of polymers. Polarographic techniques can make such distinctions. Three basic techniques of polarography are of interest, the basic principles of which are outlined next. Universal: differential pulse. A voltage pulse is superimposed on the voltage ramp during the last 40 ms of controlled drop growth with a standard dropping mercury electrode; the drop surface is then constant. The pulse amplitude can be pre-selected. The current is measured by integration over a 20 ms period immediately before the start of the pulse, and again for 20 ms as the pulse nears completion. The difference between the two current integrals (I2–I1) is recorded, and this gives a peak-shaped

36

Determination of Elements curve. If the pulse amplitude is increased, the peak current value is raised and the peak is simultaneously broadened. Classic: direct current (DCT). Integration is carried out over the last 20 ms of controlled drop growth (‘Tast procedure’). During this time, the drop surface is constant for a dropping mercury electrode. The resulting polarogram is step-shaped. Compared with classic DC polarography according to Heyrovsky, i.e., with a freedropping mercury electrode, the DCT method offers great advantages: considerably shorter analysis times, no disturbance due to current oscillations, simpler evaluation, and larger diffusion-controlled limiting current. Rapid: square wave (SQW). Five square-wave oscillations of frequency around 125 Hz are superimposed on the voltage ramp during the last 40 ms of controlled drop growth – with a dropping mercury electrode the drop surface is then constant. Oscillation amplitude can be pre-selected. Measurements are done in the second, third, and fourth square-wave oscillation; the current is integrated over 2 ms at the end of the ﬁrst and the end of the second half of each oscillation. The three differences of the six integrals (l1–l2–l3–l4–l5–l6) are averaged arithmetically and recorded as one current value. The resulting polarogram is peak-shaped. Polarography is an excellent method for trace and ultra-trace analysis of inorganic and organic substances and compounds. The basic process of electron transfer at an electrode is a fundamental electrochemical principle, so polarography can be used over a wide range of applications. After previous enrichment at a ranging mercury drop electrode, metals can be determined using differential pulse-stripping voltammetry. Detection limits are of the order 0.05 μg/l. Mal’kova and co-workers [73] described an AC polarographic method for the determination of cadmium, zinc, and barium stearates or laureates in PVC. Samples are prepared for analysis by being ashed in a mufﬂe furnace at 500 ºC, a solution of the ash in hydrochloric acid being made molar in lithium chloride and adjusted to pH 4.0 ± 0.2. The solution obtained is de-aerated by the passage of argon and the polarogram recorded. Barium, cadmium and zinc, give sharp peaks at –0.90, –0.65 and –0.01 V, respectively, against the mercury-pool anode.

2.2.1.13 Ion Chromatography Ion chromatography [74] has several advantages over AAS, ICA-AES, and polarography if it is necessary to determine several metals in a polymer. These include speciﬁcity, freedom from interference, speed of analysis, and sensitivity. The polymer

37

Introduction to Polymer Analysis must be digested using suitable reagent systems to produce an aqueous solution of the ions to be determined. Ion chromatography can complement atomic absorption and plasma methods as a back-up technique. At the heart of the ion chromatography system is an analytical column containing an ion exchange column on which various anions and/or cations are separated before being detected and quantiﬁed by various detection techniques such as spectrophotometry, AAS (metals), or conductivity (anions). Ion chromatography is not restricted to the separate analysis of only anions or only cations. With appropriate selection of the eluent and separator columns, it can be used for simultaneous analysis of anions and cations. The principles of ion chromatography are discussed in an HMSO publication [75]. Numerous manufacturers now supply instrumentation for ion chromatography. Dionex are leaders in the ﬁeld; they have been responsible for many of the innovations introduced into this technique and continue to make such developments. The micromembrane suppressor enables detection of non-UV-absorbing compounds such as inorganic anions and cations, surfactants, fatty acids, and amines in ion exchange and ion pair chromatography. Two variants exist: the anionic (AMMS) and the cationic (CMMS) suppressor. The micromembrane suppressor consists of a low dead volume eluent ﬂow path through alternating layers of high-capacity ion exchange screens and ultra-thin ion exchange membranes. Ion exchange sites in each screen provide a site-to-site pathway for eluent ions to transfer to the membrane for maximum chemical suppression. Dionex anion and cation micromembrane suppressors transform eluent ions into weaker conducting species without affecting sample ions under analysis. This improves conductivity detection, sensitivity, speciﬁcity, and baseline stability. It also dramatically increases the dynamic range of the system for inorganic and organic ion chromatography. The high ion exchange capacity of the micromembrane suppressor permits changes in eluent composition by orders of magnitude, making gradient ion chromatography possible. Because of increased detection speciﬁcity, preparation is dramatically reduced, making it possible to analyse most samples after simple ﬁltering and dilution. Various detectors are used in this technique such as those based on conductivity refractive index and UV/visible detectors.

38

Determination of Elements A typical system for the determination of metals is one in which a liquid sample is introduced at the top of the ion exchange analytical column (separator column). An eluent (containing a complexing agent in the case of metal determination) is pumped through the system. This causes the ionic species (metal ions) to move through the column at rates determined by their afﬁnity for the column resin. The differential migration of the ions allows them to separate into discrete bands. As these bands move through the column they are delivered, one at a time, into the detection system. For metals, this comprises a post-column reactor that combines a colouring reagent (pyridyl azorescorcinol; PAR) with the metal bands. The coloured bands can then be detected by the appropriate detection mode. In the case of metal– PAR complex detection, visible wavelength absorbance is employed. The detector is set to measure the complexed metal band at a pre-selected wavelength. The results appear in the form of a chromatogram, essentially a plot of the time the band was retained on the column versus the signal it produces in the detector. Each metal in the sample can be identiﬁed and quantiﬁed by comparing the chromatogram against that of a standard solution. Because only the metal ions of interest are detected, ion chromatography is less prone to interferences compared with other methods. Because individual metals and metal compounds form distinct ions with differing retention times, it is possible to analyse several of them in a single run – typically <20 minutes. By selecting the appropriate column for separating the ions of interest in a sample, it is possible to separate and analyse the oxidation state of many metals, and determine Group I and II metals, metal complexes, and a complete range of inorganic and organic ions in a sample with excellent speed and sensitivity. With sample pre-concentration techniques, the detection limits for ion chromatography can surpass those of GFAAS. High concentrations of acids or bases can limit the applicability of AAS, but ion chromatography allows direct injection of up to 10% concentrated acids or bases. This is extremely convenient in the direct analysis of acid-digested samples such as polymer digests (Figure 2.7). Utilising ion exchange pre-concentration methods, extremely low concentrations of metals in polymer digests can be measured with ion chromatography. These detection limits are typically in the sub-picogram range.

39

Introduction to Polymer Analysis

Fe3+

Recorder response

Co2+ Ni2+ Metal 2+

Cu

Zn2+ Cd2+ Fe2+

Pb2+

0

4

Conc. (mg/l) 2+

Mn2+

8 12 Minutes

Lead (Pb ) Iron III (Fe3+) Copper (Cu2+) Nickel (Ni2+) Zink (Zn2+) Cobalt (Co2+) Cadmium (Cd2+) Manganese (Mn2+) Iron II (Fe2+)

10 1 1 1 1 1 3 1 1

16

Figure 2.7 Ion chromatography: determination of nine transition metal ions. Source: Author’s own ﬁles.

2.2.2 Non-destructive Methods 2.2.2.1 XRFS The XRF technique has a true multi-element analysis capability and requires no foreknowledge of the elements present in the sample. It is very useful for the examination of many of the types of samples encountered in the plastics laboratory. This technique is very useful for solid samples, particularly if the main constituents (matrix) are made of low-atomic-weight elements and the sought impurities or constituents are of relatively high atomic weight. Samples are irradiated with high-energy radiation, usually X-rays, to produce secondary X-rays characteristic of the individual elements present. The X-ray intensity due to a particular element is proportional to the concentration of that element in the sample. There are two types of instrument in production: those in which the emitted radiation is separated by wavelength using crystals as gratings, i.e., total reﬂection XRF [wavelength dispersive XRF (WDXRF) or total reﬂection XRF (TRXRF)], and those in which the radiation is not separated but identiﬁed by energy dispersive electronic techniques using solid-state detectors and multi-channel analysers, i.e., energy dispersive XRF (EDXRF). Energy dispersive instruments rely on solid-state energy detectors coupled to energy discriminating circuitry to distinguish the radiation by its energy level and measure the

40

Determination of Elements amount at each level. Most X-ray detectors are solid-state devices that emit electrons when X-rays are absorbed, the energy of the electrons being proportional to that of the incident X-rays, and the quantity proportional to the intensity. Typically, instruments will determine from a few percent down to ppm in a solid sample. WDXRF tend to be most accurate and precise for trace element determinations. EDXRF instruments tend to lose precision for traces of light elements in heavy element matrices unless longer counting times are used. With short counting times, for example, the coefﬁcient of variation for a minor constituent element determination by an energy dispersive instrument should be >10%, but for a light trace element it may be only 50%. The advantage of the energy dispersive instrument is that it can be made so that almost all the radiation emitted hits the detector. Qualitative analysis is made by comparison with standard samples of known composition using total line energy. This is given by the total detector output of the line, or line peak area depending on the method of read-out used. Due to the simple spectra and the extensive element range (sodium upwards in the periodic table) that can be covered using an Si(Li) detector and a 50 kV X-ray tube, EDXRF spectrometry is perhaps unparalleled for its power of qualitative element analysis. Qualitative analysis is greatly simpliﬁed by a few peaks that occur in predictable positions, and by tabulated element/line markers routinely available from computerbased analysers. To date, the most successful method of combined background correction and peak deconvolution has been digital ﬁltering and least squares ﬁtting of reference peaks to the unknown spectrum [76]. This method is robust, simple to automate, and applicable to any sample type. The major disadvantage of conventional EDXRF has been poor elemental sensitivity, a consequence of high background noise levels resulting mainly from instrumental geometries and sample matrix effects. TRXRF is a relatively new multi-element technique with the potential to be an impressive analytical tool for trace element determinations for various sample types. The fundamental advantage of TRXRF is its ability to detect elements in the picogram range in comparison with the nanogram levels typically achieved by traditional EDXRF spectrometry. The principles of TRXRF were ﬁrst reported by Yoneda and Horiuchi [77] and further developed by Aiginger and Wodrauschek [78] and others [79–82]. In TRXRF, the exciting primary X-ray beam impinges upon the specimen prepared as a thin ﬁlm

41

Introduction to Polymer Analysis on an optically ﬂat support of synthetic quartz or Perspex at angles of incidence in the region of 2–5 minutes of arc below the critical angle. In practice, the primary radiation does not (effectively) enter the surface of the support but skims the surface, irradiating any sample placed on the support surface. The scattered radiation from the sample support is virtually eliminated, thereby drastically reducing background noise. A further advantage of the TRXRF system, resulting from the geometry used, is that the solid-state energy dispersive detector can be accommodated very close to the sample (0.3 mm), which allows a large solid angle of ﬂuorescent X-ray collection, thereby enhancing signal sensitivity and enabling the analysis to be carried out in air at atmospheric pressure. XRF can be used to conduct destructive or non-destructive analysis of polymers. XRF spectrometry has been used extensively for the determination of traces of metals and non-metals in polyoleﬁns and other polymers. The technique has also been used in the determination of major metallic constituents in polymers, such as cadmium selenide pigment in polyoleﬁns. Specimen preparation is simple, involving compressing a disc of the polymer sample for insertion in the instrument, measurement time is usually less than for other methods, and X-rays interact with elements as such, i.e., intensity measurement of a constituent element is independent of its state of chemical combination. The technique has some drawbacks, and these are evident in the measurement of cadmium and selenium. For example, absorption effects of other elements present, e.g., the carbon and hydrogen of a PE matrix, and excitation of one element by X-rays from another, e.g., cadmium and selenium affect one another. The technique has been applied to the determination of metals in polybutadiene, polyisoprene, and polyester resins [1]. The metals determined were cobalt, copper, iron, nickel and zinc. The samples were ashed, and the ash dissolved in nitric acid before X-ray analysis. Concentrations as low as 10 ppm can be determined without inter-element interference. Many investigators have found much higher recoveries using various ashing aids such as sulfuric acid [83], elemental sulfur [4, 84], magnesium nitrate [85], benzene sulfonic acid and xylene sulfonic acid [86]. Leyden and co-workers [87] used XRF spectrometry to determine metals in acid digest of polymers. The aqueous solutions were applied to ﬁlter paper discs. They found that recoveries of metals by the X-ray technique were 101–110% compared with 89–94% by chemical methods of analysis. XRF spectrometry has been applied very successfully in industry to routine determination in hot-pressed discs of PE and PP down to a few ppm of aluminium, bromine, calcium, chlorine, magnesium, potassium, sodium, titanium and vanadium.

42

Determination of Elements Dithiocarbamate precipitation methods have been used to determine between 2 mg/l and 2 μg/l of ﬁve elements. Excellent agreement is obtained between X-ray instrumental and AAS techniques in the analysis of pre-concentrates. Agreement does not extend over the whole concentration range examined for manganese. Some disparity also occurs in zinc determinations, and it is believed that the error is in the graphite furnace results. Wolska [88] reviewed recent advances in the application of XRF spectroscopy to the determination of antimony, bromine, copper, iron, phosphorus, titanium, and zinc, and various plastics. The new ED2000 high-performance EDXRF spectrometer manufactured by Oxford Instruments can determine up to 80 elements qualitatively and up to 50 elements quantitatively between sodium and uranium in various materials, including polymers [89].

2.2.2.2 NAA NAA is a very sensitive technique. Due to the complexity and cost of the technique, most laboratories do not have facilitates for carrying out NAA. Instead, samples are sent to one of the organisations that possess the facilities. An advantage of the technique is that a foreknowledge of the elements present is not essential. It can be used to indicate the presence and concentration of entirely unexpected elements, even when present at very low concentrations. In NAA, the sample in a suitable container (often a pure PE tube) is bombarded with slow neutrons for a ﬁxed time together with standards. Transmutations convert analyte elements into radioactive elements, which are different elements or isotopes of the original analyte. After removal from the reactor, the product is subjected to various counting techniques and various forms of spectrometry to identify the elements and their concentrations. This technique can determine a wide range of elements, e.g., chlorine in polyoleﬁns, metals in polymethylmethacrylate [90], total oxygen and polyethylene–ethylacrylate and polyethylene–vinylacetate copolymers [91] and total oxygen in polyoleﬁns. In many cases the results obtained by NAA can be considered as reference values and these data are of great value when these samples are analysed by alternative methods in the original laboratory. To illustrate this, some work is discussed on the determination of ppm of sodium in polyoleﬁns. It was found that replicate sodium contents determined on the same sample by a ﬂame photometric procedure were frequently widely divergent. NAA

43

Introduction to Polymer Analysis offers an independent non-destructive method of checking sodium contents that does not involve ashing. In the ﬂame photometric procedure, the sample is dry ashed at 650–800 °C in a nickel crucible and the residue dissolved in hot water before determining sodium by evaluating the intensity of the line emission at 589 nm. NAA (ﬂux of 1012 neutrons/cm/s) for sodium was carried out on PE- and PP-moulded discs containing up to about 550 ppm sodium that had been analysed by the ﬂame photometric method. Up to three-times higher sodium contents were obtained by NAA, suggesting that sodium is lost during the ashing stage of the ﬂame photometric method. Sodium can also be determined by a further independent method: emission spectrographic analysis. This involves ashing the sample at 500 ºC in an ashing aid consisting of sulfur and the magnesium salt of a long-chain fatty acid [55, 63–65]. The results by NAA and emission spectrography agree well with each other. The losses of sodium in the ﬂame photometric ashing procedure were probably caused by the maximum ashing temperature used, which exceeded that used in the emission spectrographic method by 150–300 °C. Flame photometry after dope ashing at 500 °C gives a quantitative recovery of sodium. Direct ashing without an ashing aid at 500 °C causes losses of r10% of sodium, whereas direct ashing at 800 °C causes even greater losses.

References 1 W.S. Cook, C.O. Jones and A.G. Altenau, Canadian Spectroscopy, 1968, 13, 64. 2 W.W. Houk and L. Silverman, Analytical Chemistry, 1959, 31, 6, 1069. 3 B.J. Mitchell and H.J. O’Hear, Analytical Chemistry, 1962, 34, 12, 1620. 4 J.S. Bergmann, C.H. Ekhart, L. Grantelli and J.L. Janik in Proceedings of the 153rd ACS Meeting, Miami Beach, FL, USA, 1967. 5 K. Tanaka and T. Morikawat, To Kogyo (Osaka), 1974, 48, 387. 6 J.Z. Falcon, J.L. Love, L.J. Gaeta and A.G. Altenau, Analytical Chemistry, 1975, 47, 1, 171. 7 C.A. Johnson and M.A. Leonard, Analyst, 1961, 86, 1019, 101. 8 W.D. Mitterberger and R. Gross, Kunststoff Technic, 1973, 12, 7, 176.

44

Determination of Elements 9 M.L. Bakroni, N.K. Chakavarty and S. Chopra, Indian Journal of Technology, 1975, 13, 576. 10 S.L. Manatt, D. Horowitz, R. Horowitz and R.P. Pinnell, Analytical Chemistry, 1980, 52, 9, 1529. 11 E.A.Williams, E.M.S. Frave, P.E. Donahue, N.A. Marotta and R.P. Kambour, Applied Spectroscopy, 1990, 44, 7, 1107. 12 Z. Qi and P.G. Pickup, Analytical Chemistry, 1993, 65, 6, 696. 13 A.F. Colson, Analyst, 1963, 88, 1042, 26. 14 A.F. Colson, Analyst, 1963, 88, 1051, 791. 15 H. Narasaki, V. Hijaji and U. Unno, Bunseki Kagaku, 1973, 22, 5, 541. 16 L.S. Kalinina, N.I. Nikitina, M.A. Matorina and I.V. Sedova, Plasticheskie Massy, 1976, 5, 66. 17 H.A. Hernandez, International Laboratory, 1981, 84. 18 A. Wirsen, Makromolekulare Chemie, 1988, 189, 4, 833. 19 F. Mocker, Kautschuk und Gummi Kunstoffe, 1964, 11, 1161. 20 T. Yoshizaki, Analytical Chemistry, 1963, 35, 13, 2177. 21 The Elemental Analysis of Various Classes of Chemical Compounds Used with Perkin Elmer PE2400 CHN Elemental Analyser, Perkin Elmer Elemental Newsletter EAN-5, Perkin Elmer, Norwalk, CT, USA. 22 Principles of Operation – the Perkin Elmer PE2400 CHN Elemental Analyser, Perkin Elmer Elemental Newsletter EAN-2, Perkin Elmer, Norwalk, CT, USA. 23 Principles of Operation – Oxygen Analysis Accessory for the PE2400 CHN Elemental Analyser, Perkin Elmer Elemental Newsletter EAN-4, Perkin Elmer, Norwalk, CT, USA. 24 Calculation of Weight Percentages for Organic Elemental Analysis, Perkin Elmer Elemental Newsletter EAN-6, Perkin Elmer, Norwalk, CT, USA. 25 Analysis of a Copolymer or Polymer Blend when One Component Contains a Heteroelement, Perkin Elmer Elemental Analysis Newsletter, EAN-13, Perkin Elmer, Norwalk, CT, USA.

45

Introduction to Polymer Analysis 26 Application of the PE2400 CHN Elemental Analyser for the Analysis of Plasticizers, Perkin Elmer Elemental Analysis Newsletter, EAN-23, Perkin Elmer, Norwalk, CT, USA. 27 H. Small, T.S. Stevens and W.C. Bauman, Analytical Chemistry, 1975, 47, 11, 1801. 28 T.L. Smith and B.N. Whelihan, Textile Chemist and Colorist, 1978, 10, 5, 35. 29 G.D.B. van Houwelingen, Analyst, 1981, 106, 1267, 1057. 30 G.D.B. van Houwelingen, M.W.M.G. Peters and W.G.B. Huysmans, Fresenius’ Zeitschrift für Analytische Chemie, 1978, 293, 5, 396. 31 J. Wolska, Plastics Additives and Compounding, 2003, 5, 3, 50. 32 F. Blockhuys, M. Claes, R. Van Grieken and H.J. Geise, Analytical Chemistry, 2000, 72, 14, 3366. 33 L.M. Sherman, Plastics Technology, 1998, 44, 56. 34 J.M. Bruna and S.A. Izasa, Revista de Plasticos Modernos, 1995, 69, 468, 550. 35 M. Niino and A. Yabe, Journal of Polymer Science Part A: Polymer Chemistry, 1998, 36, 14, 2483. 36 C. Cambo and B. Loiseau, Double Liaison, Chimie des Peintures, 1987, 34, 17. 37 R.J. Watling, Analytica Chimica Acta, 1977, 94, 1, 181. 38 C. Hallam and K.C. Thompson, Analyst, 1995, 110, 5, 497. 39 R.G. Godden and D.R. Thomerson, Analyst, 1980, 105, 1257, 1137. 40 B.W. Renoe, Journal of Automatic Chemistry, 1982, 4, 2, 61. 41 P.D. Goulden and P. Brooksbank, Analytical Chemistry, 1974, 46, 11, 1431. 42 J.K. Foremand and P.B. Stockwell, Topics in Automatic Chemical Analysis, Ellis Horwood, Chichester, UK, 1979. 43 A.L. Dennis and D.G. Porter, Journal of Automatic Chemistry, 1980, 2, 3, 134. 44 B. Pahlavanpour, M. Thompson and L. Thorne, Analyst, 1981, 106, 1261, 467. 45 R.G. Godden and D.R. Tromerson, The Analyst, 1980, 105, 1137.

46

Determination of Elements 46 B. Pahlavanpour, M. Thompson and L. Thorne, Analyst, 1981, 106, 1261, 467. 47 S. Greenﬁeld, I.LI. Jones and C.T. Berry, Analyst, 1964, 89, 1064, 713. 48 R.H. Wendt and V.A. Fassel, Analytical Chemistry, 1965, 37, 7, 920. 49 R.H. Scott, V.A. Fassel, R.N. Kniseley and D.E. Nixon, Analytical Chemistry, 1974, 46, 1, 75. 50 M. Thompson and J.N. Walsh, A Handbook of Inductively Coupled Plasma Spectrometry, 1st Edition, Blackie, London, UK, 1983, p.55. 51 R.F. Suddendorf and K.W. Boyer, Analytical Chemistry, 1978, 50, 13, 1769. 52 B.L. Sharp, inventor; no assignee; GB 8,432,338, 1984. 53 A.M. Gunn, D.L. Millard and G.F. Kirkbright, Analyst, 1978, 103, 1231, 1066. 54 H. Matusiewicz and R.M. Barnes, Applied Spectroscopy, 1984, 38, 5, 745. 55 M.W. Tikkanen and K.M. Niemczyk, Analytical Chemistry, 1984, 56, 11, 1997. 56 E.D. Salin and G. Horlick, Analytical Chemistry, 1979, 51, 13, 2284. 57 E.D. Salin and R.L.A. Sing, Analytical Chemistry, 1984, 56, 13, 2596. 58 A.L. Briseno, A. Baca, Q. Zhou, R. Lai and F. Zhou, Analytical Chimica Acta, 2001, 441, 1, 123. 59 J. Ruzicka and E.H. Hansen, Analytica Chimica Acta, 1978, 99, 1, 37. 60 A.M. Dobney, A.J.G. Mank, K.H. Grobecker, P. Conneely and C.G. de Koster, Analytica Chimica Acta, 2000, 423, 1, 9. 61 A.M. Riquet and A. Feigenbaum, Food Additives and Contaminants, 1997, 14, 1, 53. 62 H. Narasaki and K. Umezawa, Kobunshi Kogaku, 1972, 29, 438. 63 T.T. Gorsuch, Analyst, 1962, 87, 1031, 112. 64 T.T. Gorsuch, Analyst, 1959, 84, 996, 135. 65 E.L. Henn, Analytica Chimica Acta, 1974, 73, 273. 66 T. Korenaga, Analyst, 1981, 106, 1256, 40. 47

Introduction to Polymer Analysis 67 R. Reverz and E. Hasty in Proceedings of the Pittsburgh Conference and Exposition of Analytical Chemistry and Applied Spectroscopy, 1987. 68 R.A. Nadkarni, Analytical Chemistry, 1984, 56, 12, 2233. 69 H.M. Kingston and L.B. Jassie, Analytical Chemistry, 1986, 58, 12, 2534. 70 Parr Manual for 207M, Parr Instruments Co., Moline, IL, USA. 71 R.A. Nadkarni, American Laboratory, 1981, 13, 22. 72 K.C. Thompson, Ultraviolet and Visible Solution Spectrophotometry and Colorimetry 1980 - An Essay Review, Her Majesty’s Stationery Ofﬁce, London, UK, 1981. 73 L.M. Mal’kova, A.I. Kalanin and E.M. Derepletchikova, Zhurnal Analiticheskoi Kimii, 1972, 27, 56. 74 H. Small, T.S. Stevens and W.C. Bauman, Analytical Chemistry, 1975, 47, 11, 1801. 75 High Performance Liquid Chromatography, Ion Chromatography, Thin-Layer and Column Chromatography of Water Samples, Standing Committee of Analysts, Her Majesty’s Stationery Ofﬁce, London, UK, 1984. 76 P.J. Statham, Analytical Chemistry, 1977, 49, 14, 2149. 77 Y. Yoneda and T. Horiuchi, Review Scientiﬁc Instruments, 1971, 42, 1069. 78 H. Aiginger and P. Wodrauschek, Nuclear Instruments and Methods, 1974, 114, 157. 79 J. Knoth and H. Schwenke, Fresenius’ Zeitschrift für Analytische Chemie, 1978, 291, 3, 200. 80 J. Knoth and H. Schwenke, Fresenius’ Zeitschrift für Analytische Chemie, 1980, 301, 1, 7. 81 H. Schwenke and J. Knoth, Nuclear Methods, 1982, 193, 239. 82 P.A. Pella and R.C. Dobbyn, Analytical Chemistry, 1988, 60, 7, 684. 83 A.A. Vasil’eva, Y.V. Vodzinskii and I.A. Korshunov, Zavodskaia Laboratoriia, 1968, 34, 1304. 84 W.A. Rowe and K.P. Yates, Analytical Chemistry, 1963, 35, 3, 368.

48

Determination of Elements 85 L.W. Gamble and W.H. Jones, Analytical Chemistry, 1955, 27, 9, 1456. 86 J.E. Shott, Jr., T.J. Garland and R.O. Clark, Analytical Chemistry, 1961, 33, 506. 87 D.E. Leyden, J.C. Lennox and C.U. Pittman, Jr., Analytica Chimica Acta, 1973, 64, 1, 143. 88 J. Wolska, Plastics, Additives and Compounding, 2003, 5, 3, 50. 89 Rubber World, 1999, 219, 4, 79. 90 M. Kabayashi, Journal of Polymer Science: Polymer Chemistry Edition, 1979, 17, 1, 293. 91 D. Hull and J. Gilmore in the Proceedings of the 141st ACS Division of Fuel Chemistry Meeting, Washington, DC, USA, 1962.

49

Introduction to Polymer Analysis

50

3

Functional Groups

Various instrumental techniques have been used to determine functional groups in polymers and to elucidate the detail of polymer structure. These include infrared (IR) spectroscopy, near-IR spectroscopy including Raman spectroscopy, Fourier transform infrared spectroscopy (FT-IR), nuclear magnetic resonance (NMR) and proton magnetic resonance (PMR) spectroscopy, chemical reaction gas chromatography, pyrolysis gas chromatography (Py-GC), pyrolysis gas chromatography–mass spectrometry, pyrolysis–NMR spectroscopy, and X-ray ﬂuorescence spectroscopy (XRFS); newer techniques such as time-of-ﬂight secondary ion mass spectrometry (ToF-SIMS), X-ray photoelectron spectroscopy (XPS), tandem mass spectrometry (MSMS), matrix-assisted laser desorption/ionisation mass spectrometry (MALDI-MS), microthermal analysis, atomic force microscopy (AFM); as well as X-ray methods, including scanning electron microscopy (SEM) and energy dispersive analysis using X-rays (EDAX). Unsaturation may need to be determined, as well as functional groups such as hydroxy, carbonyl, carboxyl, alkyl, aryl, alkoxy, oxyalkylene, nitrile, ester, amino, nitro, amide, amido, imino, epoxy cyano and organosilicon groups.

3.1 Hydroxy Groups 3.1.1 Chemical Methods Chemical methods for the determination of hydroxyl groups in polymers are based on acetylation [1], phthalation [2] and reaction with phenyl isocyanate [3] or, if two adjacent hydroxy groups are present in the polymers by reaction, with potassium periodate [1]. The reactions on which these methods are based are as follows:

51

Introduction to Polymer Analysis Acetylation:

Phthalation:

Reaction with phenyl isocyanate:

ROH +

NCO

NHCOOR (urethane)

Reaction with potassium periodate:

In all these methods, an excess of standardised reagent is added and, at the end of the reaction, unconsumed reagent is estimated. The concentration of hydroxyl groups can then be calculated from the amount of reagent consumed.

3.1.1.1 Acetylation and Phthalation Procedures (a) Determination of Hydroxyl Groups in Polyoxyalkylene Glycols In these procedures, the hydroxyl-containing polymer is reacted with an excess of a standard non-aqueous solution of acetic anhydride or phthalic anhydride, sometimes in the presence of a catalyst such as p-toluene sulfonic acid. After ester formation is complete, an excess of water is added to convert excess anhydride to the free carboxylic acid. The acid is titrated with aqueous or alcoholic standard potassium hydroxide to the phenol phthalein end-point to determine unconsumed acid. A blank run is carried out in which the sample is omitted. The hydroxyl

52

Functional Groups content of the polymer can then be calculated from the difference between the sample and blank titrations. Stetzler and Smullin [2] found that for polypropylene glycols the classical phthalation procedure gave consistently low results, as did the perchloric-catalysed acetylation procedure developed by Fritz and co-workers [4]. In addition to its greater intrinsic accuracy, the advantages claimed by Stetzler and Smullin [2] for the toluene p-sulfonic acid catalysed procedure over phthalation include a shorter reaction time and lower reaction temperature, as well as good reproducibility, including no reaction with ether groups. Table 3.1 compares hydroxyl values obtained by three methods. In every case the acid-catalysed acetylation method gives a higher result than that obtained by phthalation, the mean difference between the two methods being 2.5% of the determined value. Agreement between the p-toluene sulfonic acid method and the phenyl isocyanate method is good. (b) Reaction with Phenol or Toluene Diisocyanate As mentioned above, organic isocyanates react with hydroxyl groups to produce a urethame as follows:

In many practical situations, distinguishing between primary and secondary hydroxyl groups in polymers is necessary. The reaction product of a glycerol– propylene oxide condensate tipped with ethylene oxide would therefore contain both types of hydroxyl group:

53

Introduction to Polymer Analysis Differences in reaction rate between primary and secondary hydroxyl groups with phenyl isocyanate are the basis of methods for carrying out this determination (see next) [5, 6]. (c) Determination of Primary and Secondary Hydroxyl Groups in Ethylene Oxide Tipped Glycerol-Propylene Oxide Condensates Crompton [7] developed a kinetic method for the evaluation of the reactivity of polyols in the range 3000–5000. This procedure may be used to determine the degree of ethylene oxide ‘tipping’ produced by the addition of ethylene oxide to glycerol/propylene oxide condensates. The method was developed to measure the relative reactivity with isocyanates of such condensates as a function of their ethylene oxide tipping content. The method is based on the observation that primary hydroxyl groups react with phenyl isocyanate to form a urethane faster than do secondary hydroxyl groups. A ‘tipped’ polyol will therefore react more completely in a given time with an equivalent amount of phenyl isocyanate than an ‘untipped’ polyol reacting under identical conditions. The greater the amount of ethylene oxide tipping, the greater its rate of reaction with phenyl isocyanate. The reaction is carried out under standard conditions in which a calculated weight of the polyol (depending on its hydroxyl number) is reacted with an excess of a standard toluene solution of phenyl isocyanate in the presence of a basic catalyst. Unconsumed phenyl isocyanate is then reacted with excess standard potassium hydroxide:

PhCNO + 2KOH = K 2CO 3 + PhNH2 Excess sodium hydroxide is estimated by titration with standard acetic acid to the phenol phthalein end-point. A blank run is done in which the sample is omitted. From the difference between the sample and the blank titrations, it is possible to calculate the hydroxyl content of the original polymer: 1 mole hydroxyl groups y 1 mole phenyl isocyanate y 2000 ml N KOH

54

Functional Groups

Reactivity is calculated by a kinetic procedure in which the percentage of the original phenyl isocyanate addition, which reacts with the sample in a given time, is taken as an index of its reactivity. A calibration graph can be prepared in which the determined phenyl isocyanate reactivity is plotted against the ethylene oxide content, enabling a determination to be made of the ethylene oxide content of unknown samples from the calibration graph by interpolation. Crompton [7] applied this method to a range of glycerol/propylene oxide adducts containing various accurately known amounts of ethylene oxide tipping, up to 5.3 moles (Table 3.2). Figure 3.1 shows the reaction curves for each of the standard samples analysed. Increasing the ethylene oxide tip content of a polyol leads to a

55

Introduction to Polymer Analysis distinct increase in the reactivity of the polyol with phenyl isocyanate. As expected, untipped polyols, which are relatively free from primary hydroxyl groups, i.e., curves A and B, react comparatively slowly with phenyl isocyanate. Decinormal solutions of untipped glycerol/propylene oxide condensates of molecular weight 3000 and 5000 had an identical rate of reaction with phenyl isocyanate (Figure 3.1, curves A and B). Thus the rate of reaction with phenyl isocyanate of the terminal isopropanol end-groups in polyols is independent of molecular weight in the molecular weight range 3000 to 5000 and depends only on the proportions of primary and secondary hydroxyl end-groups present.

Table 3.1 Hydroxyl values obtained on polypropylene glycol by catalysed acetylation, phthalation and phenyl isocyanate reaction Molecular weight

p-Toluene sulfonic acidcatalysed acetylation method. Average hydroxyl value (mg KOH/g)

Phthalation method. Average hydroxyl value (mg KOH/g)

700

353

346

Reaction with phenyl isocyanate. Average hydroxyl value (mg KOH/g)

1260

276

265

5000

34.9

34.0

34.9

5000

34.3

33.8

35.1

2890

58.3

57.0

Reproduced with permission from R.S. Stetzler and C.F. Smullin, Analytical Chemistry, 1962, 34, 2, 194. © 1962, ACS

A calibration curve is prepared by plotting moles ethylene oxide per mole of glycerol for the range of standard tipped polyols of known ethylene oxide content against percentage of original phenyl isocyanate addition consumed after 60 and 100 minutes, i.e., P60% and P100%. This curve can be used to obtain from P60% and P100% data obtained for tipped glycerol for propylene oxide polyols of unknown composition their tipped ethylene oxide contents (in moles of ethylene oxide). The Stetzler and Smullin [2] method is applicable to polyoxyethylene-, polyoxypropylene- and ethylene oxide-tipped glycerol/propylene oxide condensates. Compounds of this type in the molecular weight range 500 to 5000 can be analysed by this procedure.

56

Functional Groups Van Houwelingen [8] described a procedure for the determination of hydroxy groups in polyethylene terephthalate and 4-polybutylene terephthalate, and ester interchange elastomers of 4-polybutylene terephthalate and polypropylene glycol. This method has been applied to the determination of polyamides. The hydroxyl groups in polyesters are determined by acetylation with an excess of dichloroacetic anhydride in dichloroacetic acid, and measurement of the amount of acetylation by chlorine determination. Owing to the low concentration of hydroxyl groups (especially in high relative molecular weight materials), determination of the excess is inaccurate, so determination of the amount of reagent incorporate is much more attractive.

Table 3.2 Application of reactivity method to standard ethylene oxide tipped polyols Sample identiﬁcation

Approximate molecular weight

Hydroxyl number (mg KOH/g polyol)

Ethylene oxide tipping moles ethylene oxide/ mole glycerol (by weight addition)

Reactivity, i.e., percentage of original phenyl isocynate addition consumed in: 60 min reaction

100 min reaction

A

3000

59.0

0.0

25.2

35.2

B

5000

34.9

0.0

27.1

36.8

C

5000

34.3

3.0

41.5

46.6

D

5000

35.9

3.5

44.4

51.0

E

5000

36.0

4.3

47.5

54.1

F

5000

33.3

5.3

51.9

57.4

Source: Author’s own ﬁles.

The derivatisation of the polyester is carried out in a 10% mmol solution of dichloroacetic anhydride in dichloroacetic acid at 60 ºC. A reaction time of 1 hour sufﬁces. After the reaction, the solution is poured into water and the precipitated polymer washed out. To remove the last traces of solvent and acetylation agent, reprecipitation of the derivatised polymer from a hexaﬂuoroisopropanol solution into water is carried out. For polymers with low hydroxyl content (<100 mmol/kg), reprecipitation is carried out from a solution in nitrobenzene into cold light petroleum to obtain more reproducible results.

57

Introduction to Polymer Analysis

Weight addition tipping moles ethylone oxide/ mole glycerol

% of original phenyl isocyoante addition consumed

Iden.

60

A B C D E F

Nil Nil 3 3.5 4.3 5.3

Molecular weight 3000 5000 5000 5000 5000 5000

F E D

50

C

40 A, B

30

20

10

0

10

10

10

10 10 10 70 80 90 Time from start of reaction (min)

100

Figure 3.1 Standard polyol samples: reaction curves. Source: Author’s own ﬁles.

The content of hydroxyl groups is subsequently determined by measurement of the chlorine content of the puriﬁed derivative. This is done by potentiometric titration with silver ions after combustion or by XRFS of a compressed disc of the polymer. The suitability of this method for several samples is demonstrated in Table 3.3. The standard deviation of the method for high relative molecular weight polymers is 0.7 mmol/kg, whereas it is about 10 mmol/kg for low relative molecular weight materials.

58

Functional Groups

3.1.2 Spectrophotometric methods A method for the determination of hydroxy group in polyethylene glycols has been described by Fritz and co-workers [4]. The method consists of substituting the hydroxyl group with a chromophoric siloxy group, puriﬁcation of the silylated polymer, and a photometric determination of the chromophor concentration. The silanisation of the primary hydroxyls with dimethylaminosilanes proceeds quantitatively under very mild conditions and elimination of the excess reagent by precipitation is easy. The precision of the method is ±4.5% (95% conﬁdence level) down to 5 × 10–4 mol/kg. The method has about the same precision as acylation methods, but yields a 100-fold gain in sensitivity.

Table 3.3 Hydroxyl end group content of several PET, PBTP and PBTP-PPG samples Sample

Viscosity ratioa

Hydroxyl end groups/mmol/kg

Remarks

PET A

1.80

30.9-29.8

Total end groupsb 76.6 mmol/kg

PET B

1.82

14.4-14.8

Total end groups 79.1 mmol/kg

PBTP A

2.07

70.7-70.6

Total end groups 95.2 mmol/kg

PBTP B

2.08

50.6-49.3

Total end groups 92.6 mmol/kg

PBTP C

1.24

664-653

PBTP-PPG A

1.25

742-729

PBTP-PPG B

1.34

329-345

PBTP-PPG C

1.5

229-212

Mn (calc) 9000; Mn (measured) 10100

PBTP-PPG D

2.38

51-47

Mn (calc) 35000; Mn (measured) 44800

Mn(calc)c 2700; Mn (measured) 2400d

a

Measured for a 1% mmol solution in a m-cresol at 25 oC (PET and PBTP), or for a 1% m/V solution in o-chlorophenol at 25 oC (PBTP-PPG).

b

Total end groups in sum of OH + COOH + methyl ester end groups.

c

Measured by gel permeation chromatography in m-cresol as a solvent.

d

Calculated from OH + COOH content.

Mn: Number average molecular weight PET: Polyethylene terephthalate PBTP: Polybutylene terephthalate PBTP-PPG: Ester interchange elastomer of 4-polybutylene terephthalate and polypropylene glycol Reproduced with permission from G.D.B. Van Houwelingen, Analyst, 1981, 106, 1267, 1057. © 1918, RSC

59

Introduction to Polymer Analysis In a further spectrophotometric method, the polyester is reacted with a dimethyl formamide monochlorbenzene solution of vanadium 8-hydroxyquinolinate (V8HQ). A coloured complex is formed according to Tanaka and Kojima [9]. After removal of the excess of reagent by extraction, the complex is acidiﬁed with dichloroacetic acid and the blue colour formed is measured at 620 nm. Calibration is carried out with an alcohol as internal standard. The application of this method to some esters of new types of acids showed that a precise determination is possible (Table 3.4). The standard deviation of this method is 1.5 mmol/kg, about ten times more precise than that of the classical acetylation procedure.

Table 3.4 Comparison of the V8HQ method with the acetylation method Sample

OH content (mmol/kg) V8HQ

Acetylation

I

47.9–47.6 43.1–46.8

50–70

II

35.5–34.9

35–50

III

130.1–126.7

170–120

IV

73.8–74.9

50–70

Reproduced with permission from M. Tanaka and I. Kojima, Analytica Chimica Acta, 1968, 41, 75. © 1968, Elsevier

3.1.3 Direct Injection Enthalpimetry Direct injection enthalpimetry involves the reaction of a small portion of sample with a large excess of acetic anhydride under conditions where the reaction is rapid (<1 s), and the change in temperature associated with the reaction (dT) is recorded using a thermistor bridge. Under conditions of constant heat capacity, dT should be directly proportional to the number of reactive groups per unit mass of the sample. The main advantages of the technique are that (i) it is relatively simple to operate and (ii) only a few minutes are required to carry out an analysis. Kaduji and Rees [10] and others [11] employed direct injection enthalpimetry to determine the hydroxy value of glycerol–alkylene oxide polyethers and butane-1,4diol-adipic acid polyesters.

60

Functional Groups Direct injection enthalpimetry has great potential for determining the hydroxyl values of polyethers and polyesters.

3.1.4 IR Spectroscopy 3.1.4.1 Determination of Hydroxy Groups in Dinitropropyl Acrylate Prepolymer This method [3, 10, 12, 13] utilises the strong IR absorption band at 2.90 μm. The hydroxyl concentration of ~30 mequiv/l is sufﬁciently low that the hydroxyl groups are completely associated with the tetrahydrofuran spectroscopic solvents, and there are no apparent free self-associated hydroxy peaks. The sample must be dry because water absorbs strongly in the 2.9 μm region of the spectrum. A further limitation of the method is that other functional group in the sample such as phenols, amides, amines, and sulfonic acid groups that absorb in the 2.94–2.86 μm region are likely to interfere in the determination of the hydroxyl groups. Kim and co-workers [13] compared hydroxyl-equivalent weights obtained by this method and chemical methods of analysis for a range of prepolymers. Table 3.5 shows that good agreement is obtained. Brako and Wexler [14] described a useful technique for differentiating the presence or absence of functional groups such as hydroxyl, carboxylic acid or ester in polymers containing small percentage components of such groups. Films of latexes or polymers are subjected to chemical treatment, which results in marked changes in the IR spectrum and which can be associated with the disappearance of a functional group. IR data may be readily interpreted negatively so that one may deﬁnitely preclude the presence of hydroxyl, carbonyl, amine, amide, nitrile, ester, carboxylic, aromatic, methylene, tertiary butyl, and terminal vinyl groups if the corresponding group vibrations are absent in the IR spectrogram. More difﬁcult is the assignment of functional groups if multiple or several alternative possibilities exist, as in the mixture of a carboxylic and keto group, or in the assignment of a band to an oleﬁnic group. Figure 3.2(a) shows the IR spectra of a sodium polyacrylate ﬁlm before and after exposure to hydrochloric acid vapour. Exposure to acid results in the disappearance of the broad, intense band associated with the carboxylate group in the polyacrylate ion at around 6.25 μm. A broad intense absorption at about 5.8 μm is associated with carbonyl of the carboxylic acid group in the polymer. Signiﬁcant changes are also observed in the 9.09–8.33 μm region. Heating the acidiﬁed ﬁlm resulted in minor changes in the spectrum. Figure 3.2(b) shows the changes in the IR spectrum resulting from the exposure of acrylic acid–vinylidene chloride copolymer ﬁlm to ammonia vapour. Bands associated with the carboxylic acid carbonyl stretching frequencies at

61

Introduction to Polymer Analysis 5.83–5.75 μm disappear on exposure to ammonia vapour. A well-deﬁned carboxylate band appears at 6.37 μm. This change is sufﬁcient to conﬁrm that the copolymer contains carboxylic acid groups.

Table 3.5 Hydroxy equivalent weights of prepolymers: comparison of IR method with methods Prepolymers

Vendor’s equivalent weight

IR method

Hydroxy terminated polybutadiene

1300

1280

Hydroxy terminated butadiene acrylonitrile copolymer

1820

1770

Hydroxy terminated polycaprolactone

980

Polyethylene glycol Hydroxy terminated polytetrahydrofuran

Chemical methods PA 614*

AA 615 616*

1350

1440

970

970

1000

1660

1630a

1670

1740

500

480

500

850

a

Alkoxyethanols used for calibration

* Reaction with excess acid anhydride in presence of base catalyst PA - phthalic anhydride, AA - acetic anhydride. Excess anhydride hydrolysed at end of reaction and carboxylate groups titrated with standard potassium hydroxide Reprinted with permission from C.S.Y. Kim and co-workers, Analytical Chemistry, 1982, 54, 232. © 1982, ACS

3.1.5 NMR A further method for distinguishing between different types of hydroxy groups in polymers is the NMR method described by Li-Ho [12]. This procedure is based on the reaction of the hydroxy compound with hexaﬂuoroacetone to form an adduct which is amenable to 19F-NMR spectroscopy:

There are two interesting features in the spectrum of the hexaﬂuoroacetone adducts. First it illustrates clearly, the high resolution and information on structural aspects

62

Functional Groups of the molecules one can obtain by this method. The hydroxyl adduct from each alcohol gives a sharp resonance, with the tertiary adducts at high ﬁeld, followed by a secondary and then primary. The chemical shift of the hexaﬂuoroacetone alcohol adduct is determined by the structural environment of the hydroxyl. For example, the gradual upﬁeld shift observed in the series methanol, ethanol, isopropanol and tertbutanol results from the increase in shielding caused by replacing a hydrogen atom with a methyl group. A high degree of resolution was also observed for polymeric materials. Therefore, not only can the total hydroxyl concentration be determined but also the type(s) present in the polymer. (a) 100 90 80 70 60 50 40 30 20 10 0 2000

1800

1600

1400 1200 1000 Frequency (cm–1)

800

600

400

1800

1600

1400 1200 1000 Frequency (cm–1)

800

600

400

(b) 100 90 80 70 60 50 40 30 20 10 0 2000

Figure 3.2 (a) Spectra of polybutadiene and of the ﬁlm after exposure to bromine: --------, polybutadiene; --------, polybutadiene after exposure to bromine vapour. (b) Spectra of a ﬁlm of butadiene-styrene copolymer and of the ﬁlm after exposure to bromine vapour: -------; copolymer of butadiene and styrene; ------ copolymer of butadiene and styrene after exposure to bromine vapour. Reproduced with permission from F.D. Brako and A.S. Wexler, Analytical Chemistry, 1963, 35, 1944. © 1963, ACS

63

Introduction to Polymer Analysis

3.2 Carboxyl Groups 3.2.1 NMR Johnson and co-workers [15] investigated the application of proton NMR to the determination of carboxy groups in copolymers of methacrylate and methacrylic acid. Using the integral of the ester methoxy protons and combining this result with the total integral for CH2 and CH3 protons (overlap between CH2 and CH3 resonances was sufﬁcient at 100 MHz to prevent separate determination of these integrals), copolymer composition could be ascertained. It was necessary to carry out the determinations at r100 °C to obtain resolution sufﬁcient for reliable integrals. An additional problem was that the reaction solvents (toluene and hexane) and comonomers had resonances that overlapped those of the CH2 and CH3 of the copolymers, introducing considerable inaccuracy in the total CH2–CH3 integral. They therefore investigated the applicability of 13C-NMR. Because of the greater spectral dispersion and narrower resonance lines obtained with 13C-NMR relative to proton NMR, problems associated with resonance overlap can be resolved. Excellent agreement was obtained between carboxyl values obtained by this procedure and conventional titration in 1:1 ethanol:water with standard potassium hydroxide to the phenolphthalein end-point over the acid content range 13–100%.

3.2.2 Titration Procedures Johnson and co-workers [15] found that, in titrating copolymers of methylacrylate and methacrylic acid with standard base to determine carboxyl groups, several deﬁciencies were encountered. The presence of up to 5% water and unreacted monomer both led to underestimations of acid content. Titration was not applicable to polymers of molecular weight more than one million (or high acid content) because of a tendency to reprecipitate during titration. Titration procedures should therefore be used with caution. Most methods for the determination of carboxyl groups in polymers are based on titration techniques including, for example, the following copolymers: acrylic acid–itaconic acid [16], acrylic acid–ethyl acrylate [17] and maleic acid–styrene [18]. High-frequency titration has been applied [19] to the analysis of itaconic acid–styrene and maleic acid–styrene copolymers. The method can also be used to detect traces of acidic impurities in polymers, and in the identiﬁcation of mixtures of similar acidic copolymers. Titration indicates that the acid segments in the copolymers of itaconic acid–styrene, and maleic acid–styrene, and the homopolymer polyitaconic acid, act as dibasic acids. The method has a sensitivity that permits identiﬁcation and approximate resolution of two carboxylate species in the same polymer, for example:

64

Functional Groups

High-frequency titration gives a precise location of the inﬂection points related to the polymer carboxyl groups, and is a sensitive method for the determination of the freedom of the copolymer samples from monobasic acid impurities (comonomer acids). This is because mixtures of copolymer acids with monobasic and dibasic acids show deﬁnite inﬂection points that can be related to the individual carboxylate species present. A titration curve (Figure 3.3) is shown for a monomethyl ester of an itaconic acid– styrene copolymer. Potentiometric titration provides a method for investigating changes in conformation undergone by polyelectrolytes in solution because the environment of the dissociating groups is dependent on the conformation of the polymer chain helix-coil transitions of polyacids. Thus, precise potentiometric titration of solutions of high molecular weight polyacrylic acid at constant ionic strength indicates such conformational transition [20–22]. Figure 3.4 shows titration results for polyacrylic acid plotted as pH + log (1 – A)/A versus A (degree of dissociation), as points connected by full curves. The four curves at the different ionic strengths all show the same features. The ﬁrst short region, labelled A in the ﬁgure, is probably due to instability in the solution, such as aggregation preceding precipitation. This region extends to higher values of A at the higher ionic strengths. The second region, B, represents the ionisation of the ﬁrst conformation of the polymer; the third, C, the conformational transition; and the fourth, D, the ionisation of the second conformation. The ﬁrst conformation, which is at the lower degree of dissociation, has presumably the more tightly coiled structure, and is denoted PAA-polyacrylic acid (PAA(a)). The second conformation, stable at high degree of

65

Introduction to Polymer Analysis dissociation, and less tightly coiled, is denoted PAA(b). The four curves of PAA(a) in Figure 3.4 have been extrapolated (dashed curves) semi-empirically to zero A, and meet there at a value of pH + log ((1 – A)/ A) of 4.58, which is the value of pK´o, the intrinsic dissociation constant of the polyacid for ionic strengths of 0.06 and 0.11 found previously. This extrapolation is made with the help of plots pf pH versus log ((1 – A)/A) shown in Figure 3.5. These plots, though almost linear over the entire range, as previously reported by Mandel and Leyte [23, 24], are not quite so, and regions A, B, C and D can be also be distinguished. Linear extrapolation of region B, which represents PAA(a), to higher values of log ((1 – A)/A) was used to obtain the extrapolations of the PAA(a) curves of Figure 3.5 to zero A.

820 12

Dial reading

10

8

810

pH 6 A

B 4

800 2

1

5

9

13 HCl (ml)

17

21

0 25

Figure 3.3 High-frequency (A) and potentiometric (B) displacement titration of the monosodium salts of the monomethyl esters of poly(itaconic acid-co-styrene). Titration of 0.2345 g 57:43 anhydride-styrene copolymer + MeOH (heat) + excess NaOH with 0.1286 N HCl. Reproduced with permission from J. Douglas, A. Timnick and R.L. Guile, Journal of Polymer Science Part A: General Papers, 1963, 1, 5, 1609. © 1963, Wiley

66

Functional Groups

C

6.0

D

B A

pH Log 1

6.0

6.0

6.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Figure 3.4 Dependence of the function pH + log ((1- A)/A) on A for poly(acrylic acid) at various ionic strengths u: (O) u = 0.02; (x) u = 0.065; (⁪) u = 0.11; ( ) u = 0.20. Source: Author’s own ﬁles.

U

3.2.3 IR Spectroscopy IR spectroscopy has very limited application to the determination of hydroxy groups in polyacrylamides [25, 26]; carboxy terminated butadienes [27] and polyethylene.

67

Introduction to Polymer Analysis

8.0

D

8.0

8.0 C pH B 8.0

A

8.0

+1

0

1

Log 1–

Figure 3.5 pH versus log ((1– A)/A) for poly(acrylic acid) at various ionic strengths u: (O) u = 0.02; (x) u = 0.065; (⁪) u = 0.11; ( ) u = 0.20. Source: Author’s own ﬁles.

U

3.3 Ester Groups Most methods for the determination of ester groups in polymers are based on the following procedures: (a) Saponiﬁcation.

68

Functional Groups (b) Zeisel procedures (based on hydriodic acid). (c) Pyrolysis–gas chromatography. (d) Physical methods (e.g., IR and NMR).

3.3.1 Saponiﬁcation Ester groups occur in a wide range of polymers, e.g., polyethylene terephthalate and in copolymers such as, for example, ethylene–vinyl acetate, acrylic acid–vinyl ester, methyl acrylate–vinyl ester and polymethylacrylate. The classical chemical method for the determination of ester groups, saponiﬁcation, can be applied to some types of polymers. For example, copolymers of vinyl esters and esters of acrylic acid can be saponiﬁed in a sealed tube with 2 M sodium hydroxide. The free acids from the vinyl esters were determined by potentiometric titration or gas chromatography. The alcohols formed by the hydrolysis of the acrylate esters were determined by gas chromatography. Vinyl acetate–ethylene copolymers can be determined by saponiﬁcation with 1 N ethanolic potassium hydroxide at 80 °C for 3 hours [28, 29]. Polymethyl acrylate can be hydrolysed rapidly and completely under alkaline conditions; the monomer units in polymethylmethacrylate prepared and treated similarly are resistant to hydrolysis [30], although benzoate end-groups react readily [31]. Only about 9% of the ester groups in polymethylmethacrylate reacted even during prolonged hydrolysis; hydrolysis of polymethylacrylate was complete in 30 minutes. Although only about 9% of the ester groups in methylmethacrylate homopolymers are hydrolysed by alcoholic sodium hydroxide, this proportion is increased by introduction of comonomer units into the polymer chain. Thus, saponiﬁcation techniques should be applied with caution to polymeric materials. Saponiﬁcation procedures can be applied to the determination of ester groups in polymers. A copolymer of ethylene and vinyl acetate has the following structure which, upon hydrolysis in the presence of potassium hydroxide/p-toluene sulfonic acid catalyst [32, 33] reacts as follows:

69

Introduction to Polymer Analysis Excess potassium hydroxide is then determined by titration with standard acetic acid, and the vinyl acetate content of the polymer calculated from the amount of potassium hydroxide consumed. Esposito and Swann [34] described a technique involving methanolysis of a polyester resin with lithium methoxide as a catalyst: the methyl esters formed were separated from polyols and identiﬁed by gas chromatography:

This method was recently improved by Percival [35] using sodium methoxide as a catalyst and injecting the reaction mixture of the transesteriﬁcation directly into the gas chromatograph (without preliminary separation).

3.3.2 Hydriodic Acid Reduction–Gas Chromatography Alkoxy and ester groups have been determined in polymers and co-polymers by the Ziesel procedure involving reaction with anhydrous hydrogen iodide at 100 °C. Alkoxy groups: for example

Ester groups: for example

Hydrolysis using hydriodic acid has been used for the determination of the methyl, ethyl, propyl, and butyl esters of acrylates, methacrylates, or maleates [33] and the determination of polyethyl esters in methyl-methacrylate copolymers [36, 37]. First the total alcohol content is determined using a modiﬁed Zeisel hydriodic acid hydrolysis [38]. Secondly, various alcohols, after being converted to the corresponding alkyl iodides, are collected in a cold trap and separated by gas chromatography. Owing to the low volatility of the higher alkyl iodides, the hydriodic acid hydrolysis technique is not suitable for the determination of alcohol groups higher than butyl alcohol. This technique has also been applied to the determination of alkoxy groups in acrylate esters.

70

Functional Groups

3.3.3 IR Spectroscopy 3.3.3.1 Determination of Free and Combined Vinyl Acetate Groups in Vinyl Chloride-Vinyl Acetate Copolymers IR spectroscopy has been applied to the determination of free and combined vinyl acetate in vinyl chloride–vinyl acetate copolymers [39]. This method is based upon the quantitative measurement of the intensity of absorption bands in the near-IR spectral region arising from vinyl acetate. A band at 1.63 μm due to vinyl groups enables the free vinyl acetate content of the sample to be determined. A band at 2.15 μm is characteristic for the acetate group and arises from free and combined vinyl acetate. Thus, the free vinyl acetate content may be determined by difference at 2.15 μm. Polymerised vinyl chloride does not inﬂuence either measurement. The vinyl acetate content of ﬁlms of ethylene–vinyl acetate copolymers can be determined by methods based on the measurement of absorbances at 16.1 and 1.39 μm [40] and at 8.03 μm and 5.73 μm [41]. The acrylate salt in acrylate salt–ethylene ionomers has been determined from the ratio absorbances at 6.41 μm (asymmetric vibration of the carboxylate ion) and 7.25 μm [42].

3.3.3.2 Determination of Bound Vinyl Acetate in Ethylene-Vinyl Acetate Copolymers There are two methods for this determination which are dependent on the concentration of vinyl acetate. At levels <10%, a band at 2.89 μm is used. This band is not suitable for higher concentrations because the necessary ﬁlm thickness is <0.1 mm. For higher concentrations, the carbonyl overtone band can be used because much thicker ﬁlms are needed to give suitable absorbance levels. Here the carbonyl overtone band was used for a series of standards with vinyl acetate concentrations up to 35%, the nominal thickness being about 0.5 mm. A combination of CDS and QUANT software on a Perkin Elmer model 683 infrared spectrometer has been used to establish the calibration for this analysis. Once the calibration has been carried out, the simplest way to measure an unknown sample would be with an OBEY routine in the CDS II software. This would incorporate calibration data so that the single routine would measure the spectrum and calculate the vinyl acetate concentration with an error of approximately 5%.

71

Introduction to Polymer Analysis

3.3.4 NMR NMR spectroscopy has been used for the determination of isophthalate in polyethylene terephthalate isophthalate dissolved in 5% trichloroacetic acid. The NMR spectra of these polymers were measured on a high-resolution NMR spectrometer at 80 ºC. A singlet at 7.74 ppm is due to the four equivalent protons attached to the nucleus of the terephthalate unit. The complicated signals that appear at 8.21, 7.90, 7.80, 7.35, 7.22 and 7.10 T are due to the four protons attached to the nucleus of the isophthalate unit. The content of the isophthalate unit can be calculated from the integrated intensities of these peaks. NMR has also been used to determine ethyl acrylate in ethyl acrylate–ethylene and vinyl acetate–ethylene copolymers [43]. Measurements were made on 10% solutions in diphenyl ether at elevated temperature. Resolution improved with increasing temperature and lower polymer concentration in the solvent. NMR spectra for an ethylene–ethyl acrylate copolymer clearly indicate copolymer identiﬁcation and monomer ratio. A distinct ethyl group pattern (quartet, triplet), with the methylene quartet shifted downﬁeld by the adjacent oxygen, is observed. The oxygen effect carries over to the methyl triplet, which merges with the aliphatic methylene peak. No other ester group would give this characteristic pattern. The area of the quartet is a direct and quantitative measure of the ester content. All these features are consistent with identiﬁcation of an ethylene-rich copolymer with ethyl acrylate. Ethyl acrylate contents obtained by NMR (6.0%) agreed well with those obtained by PMR (6.2%) and neutron activation analysis for oxygen (6.1%). NMR spectroscopy has been applied to the determination of ester groups in polyethylene terephthalate (PET) [44]. A blend of a protic solvent (dimethyl sulfoxide), sodium hydroxide, and methanol hydrolyses ester groups in this polymer much more rapidly than previously used hydrolysis reagents. NMR spectroscopy has been used for the determination of isophthalate in polyethylene terephthalate–isophthalate dissolved in 5% trichloroacetic acid. NMR spectra of these polymers were measured on a high-resolution NMR spectrometer at 80 ºC. A singlet at 7.74 ppm is due to the four equivalent protons attached to the nucleus of the terephthalate unit. The complicated signals which appear at 8.21, 7.90, 7.80, 7.35, 7.22 and 7.10 ppm are due to the four protons attached to the nucleus of the isophthalate unit. The content of the isophthalate unit can be calculated from the integrated intensities of these peaks. Aydin and co-workers [45] hydrolysed polyesters and converted the product acids and glycols, including 1,4-cyclohexane di-methanol and isophthalic acid, to the corresponding trimethylsilyl esters and ethers, which were then analysed by gas chromatography.

72

Functional Groups

3.3.5 Pyrolysis–Gas Chromatography Barrall and co-workers [46] described a pyrolysis–gas chromatographic procedure for the analysis of polyethylene–ethyl acrylate and polyethylene–vinyl acetate copolymers and physical mixtures thereof. They used a specially constructed pyrolysis chamber as described by Porter and co-workers [47]. Less than 30 seconds is required for the sample chamber to assume block temperature. This system has the advantages of speed of sample introduction, controlled pyrolysis temperature, and complete exclusion of air from the pyrolysis chamber. The pyrolysis chromatograph of poly(ethylene–vinyl acetate) contains two principal peaks: the ﬁrst is methane and the second is acetic acid:

Variations from 350 °C to 490 °C in pyrolysis temperature produced no change in the area of the acetic acid peak, but did cause area variation in the methane peak. The pyrolysis chromatogram of poly(ethylene–ethyl acrylate) at 475 °C shows one principal peak due to ethanol. No variation in peak areas was noted in the temperature range 300 °C to 480 °C. Table 3.6 shows the analysis of 0.05 g samples of poly(ethylene–ethyl acrylate (PEEA) and poly (ethylene–vinyl acrylate) (PEVA) obtained at a pyrolysis temperature of 475 °C.

Table 3.6 Pyrolysis results on physical mixtures of polyethylene-ethyl acrylate and polyethylene-vinyl acetate Mixture

Acetic acid (wt%)

Ethylene (wt%)

Oxygen (wt%)

Found

Calculated

Found

Calculated*

Found

Calculated

50% PEEA-1 and 50% PEVA-2

9.10

9.05

2.65

2.62

7.88

8.25

33.3% PEEA-2 and 66.6% PEEA-3

12.15

12.33

0.75

0.70

7.33

7.49

*Calculated from results for acetic acid and ethylene content for individual samples on weight percent basis. Reproduced with permission from E.M. Barrall II, R.S. Porter and J.F. Johnson, Analytical Chemistry, 1963, 35, 1, 73. © 1963, ACS

73

Introduction to Polymer Analysis Haslam and co-workers [36] employed a procedure based on pyrolysis for the determination of polyethyl esters in methacrylate copolymers. The alkoxy groups in the polymers were reacted with hydrogen iodide and pyrolysed to their corresponding alkyl iodides, which were then determined by chromatography on a dinonyl sebacate column at 75 °C. Similarly, Miller and co-workers [48] determined acrylate ester impurities in polymers by converting the alkoxy groups to alkyl iodides, which were gas chromatographed on a di-2-ethyl hexyl sebacate column at 70 °C.

3.4 Carbonyl Groups Various physical techniques such as IR and NMR spectroscopy have been described for the determination of very low concentrations of carbonyl groups in polymers.

3.4.1 IR Spectroscopy The polypropylene carbonyl band after 335 hours exposure is broad, with few discernible features except for the vinyl alkene band at 6.08 μm. The broadness of the carbonyl band indicates various functional groups, and makes accurate quantitative analysis difﬁcult. The large vinyl alkene at 6.08 μm stands out clearly, and distinct carboxylic acid 5.83 μm and G-lactone 5.58 μm spikes can be readily identiﬁed. The carbonyl band for polypropylene decreases after volatile products are removed by the vacuum oven. Isopropanol extraction removes about 40% of the polypropylene carbonyl. The carbonyl band is then narrow and appears to centre at the ester absorption at 5.75 μm. Treatment with base converts lactones, esters and acids to carboxylates 6.33 μm, leaving only a small band at 5.81 μm, which is due to aldehyde and ketone. Upon re-acidiﬁcation of the polypropylene, some of the original esters at 5.75 μm do not re-form, but become carboxylic acids and G-lactones. Curiously, the vinyl alkene band becomes less intense with each step and broader, shifting down to 6.10–6.25 μm, the vinyl groups may be isomerised into internal alkenes or become conjugated during various treatments, although no such change occurs with the polyethylene vinyl alkene or with the process-degraded polypropylene vinyl alkene. Wood and Statton [49] developed a new technique to study the molecular mechanics of orientated polypropylene during creep and stress relaxation based on use of the stress-sensitive 10.25 μm band and the orientation-sensitive 11.12 μm band. The far IR spectrum of isotactic polypropylene was obtained from 300 cm–1 to 10 cm–1, and several band assignments made [50]. Isotacticity of polypropylene has been measured from IR spectra and pyrolysis–gas chromatography after calibration from standard mixtures

74

Functional Groups of isotactic and atactic polypropylene. The IR spectrum of oxidised polypropylene indicated small amounts of OOH groups plus larger concentrations of stable cyclic peroxides or epoxides in the polypropylene chain [50]. Grassie and Weir [51] described an apparatus for the measurement of the uptake of small amounts of oxygen by polystyrene with a high degree of precision. Grassie and Weir [52] investigated the application of ultraviolet (UV) and IR spectroscopy to the assessment of polystyrene ﬁlms after vacuum photolysis in the presence of 253.7 nm radiation using the apparatus mentioned above. During irradiation, there is a general increase in absorption in the region 230–350 nm. Rates of increase are relatively much greater in the 240 nm and 290–300 nm regions. Absorption in the 240 nm region is characteristic of compounds having a carbon–carbon dioxide double bond in conjunction with a benzene ring. Styrene, for example, has an absorption band at 244 nm. Schole and co-workers [53] applied an oxidative degradation technique to the study of polystyrene. The polystyrene sample is mixed with a support in a precolumn which is mounted at the inlet to a gas chromatographic column. Shaw and Marshall [54] carried out an IR spectroscopic examination of emulsiﬁerfree polystyrene which had been oxidised during polymerisation. Evidence was found for surface carboxyl groups bound to the polymer chains, presumably formed by oxidation during polymerisation. The band at 5.86 μm was assigned in part to the carbonyl stretching mode of dimeric carboxylic acid, formed by oxidation, in the polystyrene chains. Absorption at 5.65 μm, which was very weak, was tentatively attributed to the carbonyl stretching mode of the monomeric form of this acid. The structure of the acid end-group was not established, but the results suggest that it was possibly a phenylacetic acid residue or a residue of standard (unoxidised) and of oxidised emulsion polymerised polystyrene in the region 1.25–25.0 μm. The spectrum of latex B1 contains weak bands at 5.65 μm and 5.86 μm.

3.4.2 Derivatisation Methods Derivatisation methods have been applied to the determination of low concentrations of carbonyl groups in polyvinyl chloride and vinyl chloride–vinyl acetate copolymers [55]. Carbonyl groups in the polymer are reacted with 2,4-dinitrophenylhydrazine to produce the corresponding phenylhydrazone. Excess reagent is washed away from the polymer, which is then digested with concentrated sulfuric acid to convert the bound hydrazone to ammonium sulfate, which is then estimated using Nessler’s reagent.

75

Introduction to Polymer Analysis

3.4.2.1 Spectrophotometric methods Direct spectrophotometry in the visible and UV region has been used to determine low concentrations of functional groups in the surface of polymer ﬁlms. Kato [56] followed the regeneration of carbonyl groups from 2,4-dinitrophenylhydrazones formed on the surface of irradiated polystyrene ﬁlms by absorption measurements at 378 nm.

3.5 Ether Groups 3.5.1 Cleavage Gas Chromatography Anderson and co-workers [57] studied the application of an alcohol exchange–gas chromatography method to the determination of etheriﬁcation levels in acrylamide interpolymers of the type:

where: R = H or CH3 R1 = H, CH3, C2H5 C4H9 or C8H17 R11 = H, CH2OC4H9 or CH2OH In this method, a cast ﬁlm of the acrylamide interpolymer is reacted with an alcohol, e.g., octyl alcohol, to exchange with etherifying alcohols present in the polymer backbone as follows:

The etherifying alcohol content of the digest, in this case butanol, is determined by gas chromatography. The method is calibrated against standard solutions of the etherifying alcohol and an internal standard.

76

Functional Groups The level of alcohol obtained during the exchange reaction of acrylamide is correlated with data obtained using Zeisel cleavage of the alkoxy groups. These data are summarised in Table 3.7. Comparable results are obtained using both procedures, but the Zeisel cleavage reaction will also cleave ester linkages as well as ether functionalities. This presents no problem with polymers which do not contain ester groups. In systems employing butyl esters and butyl ethers, the Zeisel cleavage reaction gives the total O-C4H9 content in the sample. Alcohol exchange will cleave only the butyl ether groups in the sample. By subtracting the alcohol exchange data from that obtained from Zeisel cleavage, one can assess the relative amounts of alkyl ester and alkyl ether in the sample. Smith and Dawson [3] determined traces of ether linkages in polyethylene glycol and polypropylene glycol by reacting the sample with hydrobromic acid then determining the dibromoethane and dibromopropane ﬁssion products by gas chromatography. The Zeisel method has been used to determine ether groups in cellulose and polyvinyl ethers [58].

Table 3.7 Correlation of alcohol exchange and Zeisel cleavage Percentage butylated acrylamidea

Sample

Alcohol exchange

Zeisel cleavage

14% Acrylamide

100

101

23% Acrylamide

101

100

a

Based on experimentally determined butoxy and nitrogen content

Reproduced with permission from D.G. Anderson, K.E. Isakson, J.T. Vandeberg, M.Y.T. Jao, D.J. Tessari and L.C. Afremow, Analytical Chemistry, 1975, 47, 7, 1008. © 1975, ACS

3.6 Alkoxy Groups 3.6.1 IR Spectroscopy Sworej and Banby [59] used IR spectroscopy to determine epoxy groups in methylmethacrylate–glycidyl metacrylate copolymers. The IR analysis was done on dried potassium bromide pellets containing 0.5 mg sample in 200 mg potassium bromide. The peaks at the wave numbers 11.02 μm and 5.82 μm are the most suitable ones for analysis of epoxy and carbonyl groups, respectively.

77

Introduction to Polymer Analysis Using the ‘base line density’ method, values of the absorbances at 11.02 μm and 5.82 μm have been determined in triplicate. The average values of the absorbances, their ratio and the glycidylmethylacrylate mole fraction determined chemically are presented in Table 3.8. The absorbance ratio at 11.02 μm versus 5.82 μm is linearly related to the glycidyl methacrylate content in the copolymer and can be expressed by the following equation: R = 0.250 XG = 0.033 where R is the absorbance ratio at 11.02 μm and 5.82 μm, and XG is the mole fraction of glycidylmethacrylate containing the epoxy group in the copolymer. The term 0.033 is considered to be a correction factor arising from the very weak absorption due to poly(methyl methacrylate) at wave numbers near 11.02 μm.

Table 3.8 Analytically determined mole fraction of glycidyl methylacrylate in the methyl methylacrylate–glycidyl methylacrylate copolymers and the infrared absorbance at 5.82 μm and 11.02 μm. Experiment number

Mole fraction of glycidyl methylacrylate in the copolymer mined chemically

A, 5.82 μm

A, 11.02 μm

Absorbance ratio A @ 11.02 μm A @ 5.82 μm

R50

0.218

1.366

0.128

0.093

R51

0.394

0.911

0.120

0.131

R52

0.584

0.629

0.115

0.182

R61

0.623

0.921

0.177

0.192

R53

0.706

0.955

0.204

0.213

Reproduced with permission from S. Paul and B. Ranby, Analytical Chemistry, 1975, 47, 8, 1428. © 1975, ACS

Hard epoxy resins of the diglycidylether-bisphenol A type (e.g., Epikote 1004 (Shell Chemicals) and Bakelite epoxy resin ERL 1774) are manufactured by the reaction of bisphenol and epichlorohydrin. Peltonen and co-workers [60] described an IR method for determining epoxy resins and their thermal degradation products in workspace air samples collected by

78

Functional Groups ﬁltration on glass-ﬁbre ﬁlters. Epoxy residues were extracted from the ﬁlters with chloroform, the extract evaporated to dryness, and the residue dissolved in 0.5 ml deuterochloroform (CDCl3). The IR spectrum is shown in Figure 3.6. If the thermal degradation products of epoxy resins were present, then the initial chloroform extract was washed with 1 N sodium hydroxide, then water to remove phenolic products, then evaporated and the residue dissolved in deuterochloroform before IR spectroscopy. Deuterochloroform was used instead of chloroform for preparing the extract for IR spectroscopy because chloroform has a strong absorption at 6.62 μm which precludes the use of this aromatic absorption for the determination of epoxy resin in the extract. The absorbance at 6.62 μm is linear with concentration in the concentration range 570–740 μg epoxy resin per 0.5 ml with a detection limit of 50 μg per 0.5 ml and recoveries ranging from 70% at the 72 μg per 0.5 ml level to 97% at the 288 μg per 0.5 ml level. The high relative molecular mass fraction, derived from the cracked epoxy network, contains the same aromatic substances as the intact epoxy resin. This enabled Peltonen and co-workers [60] to use pure epoxy resin as a standard for quantiﬁcation of the high relative molecular mass fraction of the thermal degradation products. High-performance liquid chromatography (HPLC) revealed phenolic compounds. These were washed off because they also absorb strongly at 6.62 μm. The removal efﬁciency of phenols (100%) was conﬁrmed by HPLC. The removal of phenols was also observed in the IR spectrum as reduced absorption by hydroxy groups. The remaining absorption was probably due to the secondary aliphatic alcohol groups in the chains. Secondary aliphatic alcohols are not acidic and therefore not removed by sodium hydroxide extraction. Miyanchi and co-workers [62] determined methoxy groups in methylated melamine– formaldehyde resins by establishing the IR spectroscopic correlation between area ratio 12.27 μm (triazine) and 10.95 μm (methoxy and ethoxy contents) as established by gas chromatography.

3.6.2 NMR Spectroscopy Hammerich and Willeboordor [63] compared repeat unit (n) values obtained by PMR spectroscopy and those obtained by chemical methods of analysis. In the PMR method, areas were excluded whose integrals were overlapped. NMR determinations are usually consistently lower than the corresponding titrimetric determinations.

79

Introduction to Polymer Analysis

(a) OH

CH3, CH3

Ar

CH3, CH2

Ar Absorbance

OH

4000

Ar

Ar

OH

(b)

3000

2000

1600

1200

800

400

Wavenumber, cm−1

Figure 3.6 Infrared spectrum of Epicote 004 showing a strong absorption peak at 510 cm–1 to epoxy groups. Reproduced with permission from K. Peltonen, P. Pfäfﬂi and A. Itkonen, Analyst, 1985, 110, 1173. © 1985, Royal Society of Chemistry

3.6.3 Pyrolysis-Based Method 3.6.3.1 Alkoxy Groups in Ethylene Oxide-Propylene Oxide Condensates Upon pyrolysis at 360–410 °C in an evacuated vial, these polymers produce a mixture of ethylene and propylene proportional in amount to the concentrations of ethoxy and propoxy groups in the original polymers [61, 64]:

Figure 3.7 is a plot of the percentage ethylene as a function of the ethylene oxide content of ethylene oxide–propylene oxide condensate. As the ethoxy character of the condensate increases, so does the amount of ethylene produced. The curve is linear up to about 50% ethylene oxide, then turns sharply upward to an ethylene content of 38.6% for pure polyethylene glycol.

80

Functional Groups The relative contents of ethylene oxide and propylene oxide in polyethylene– polypropylene glycols has been determined using combined pyrolysis–gas chromatography calibrated with polyethylene glycol and polypropylene glycol standards [65].

3.6.3.2 Miscellaneous Methods Epoxy groups have also been determined spectrophotometrically using 2,4 dinitrobenzene sulfonic acid [66] and by the Zeisel method [67–69].

48 44 40 36

% C 2H 4

32 28 24 20 16 12 8 4 0

10

20

30 40 50 60 70 80 wt. % ethoxy in “Pluronics”

90

100

Figure 3.7 Percentage C2H4 as function of ethoxy content of Pluronics. Source: Author’s own ﬁles.

3.7 Oxyalkylene Groups 3.7.1 Cleavage–Gas Chromatography Several mixed anhydrides of carboxylic and sulfonic acids, as proposed by Karger and Mazur [70], act as reagents for the cleavage of ether linkages, particularly that of acetic anhydride toluene-p-sulfonic acids, which is not only a powerful reagent for the cleavage of ether linkages, but is also an active acetylating agent. For example, when the propylene oxide adduct of glycerol is treated with this reagent, the polyether is split, thus giving glycerol triacetate and propylene glycol diacetates, which are readily identiﬁed by gas chromatography. Tsuji and Kounishi [71] extended this method to

81

Introduction to Polymer Analysis the identiﬁcation of base compounds and the determination of their oxyethylene and oxypropylene group contents:

In this method, the cleavage reagent comprises a solution of acetic anhydride (84 g) in 20 g of p-toluene sulfonic acid. This mixture is reﬂuxed at 120 °C for 30 minutes. The product obtained is used as the reagent without removal of the acetic acid produced and the excess of acetic anhydride. The gas chromatogram of the acetate peaks are satisfactorily separated from each other and the peak of propylene glycol diacetate produced by the cleavage of the polyoxypropylene groups do not overlap those of the derivatives from polybase compounds except that for the polyether based on propylene glycol. These base compounds could therefore be easily distinguished and identiﬁed. By decreasing the temperature of the gas chromatographic column, the peak for propylene glycol diacetate can be accurately identiﬁed. Cervenka and Merrall [72] investigated the application of acidic dehydration of ethylene oxide–propylene oxide condensates in bromonaphthalene in the presence of p-toluene sulfonic acid to the elucidation of the molecular structure and monomer sequence of these polymers. Gas chromatography was used to determine dehydration products. Studies on poly(ethylene glycol) and poly(propylene glycol) homopolymers showed that dioxane and its derivatives (e.g., methyl dioxine 1.4) are not the only reaction products. Dehydration of poly(ethylene glycol) gave three products; while dehydration of poly(propylene glycol) six products. The majority of them were identiﬁed. Cervenka and Merrall [72] conclude that results on homopolymers, their blends and model copolymers of different chain architectures demonstrate that acidic dehydration

82

Functional Groups can distinguish ethylene oxide–propylene oxide copolymers of different structures, giving correct absolute values of overall monomer content, and also rank polyols according to their degrees of randomness. Ethylene oxide and propylene oxide adducts of polyhydric alcohols and amines are widely used as polyethers in the production of polyurethane foams by reaction with diisocyanate. The physical properties of the foams depend to a certain extent on the chemical structure of these polyethers, so it is very important to establish a method for the identiﬁcation of the base compounds and for the determination of the proportions of their oxyethylene and oxypropylene groups. Mattias and Mellor [73] split the polyethers with hydrobromic acid–acetic acid to give bromo compounds, which were analysed by gas chromatography: –CH2 – CH2O + 2 HBr = BrCH2 CH2Br + H2O Ch CH3 – Ch2O + 2 HBr = BrCH(CH3) – CH2Br + H2O In this way, the content of oxyethylene groups (and therefore the original polyhydric alcohols) can be determined. Stead and Hindley [74] modiﬁed this method and obtained good results for the determination of the oxyethylene group contents of ethylene oxide–propylene oxide copolymers.

3.7.2 Pyrolysis–Gas Chromatography Various workers have described methods for determining alkoxy groups based on thermal degradation of oxyalkylene groups to the corresponding oleﬁn, which is determined by gas chromatography [66–68]. Newmann and Nadeau [66] and Swan and Dux [67] studied the pyrolysis of ethylene oxide, propylene oxide condensates containing various proportions of ethylene oxide. There were no signiﬁcant differences in the pyrolysis chromatograms of samples heated for 0.5 hours to 2 hours. The only effect of time of pyrolysis is the total amount of gas produced. The relative concentrations of components are not signiﬁcantly changed. The temperature of pyrolysis plays a large part in the amount and type of components in the volatile gases. Between 390 °C and 410 °C, there was no noticeable change in products. As the ethylene oxide content of the copolymer increases, so does the amount of ethylene produced. The curve is linear up to about 50% ethylene oxide, then turns sharply upward to an ethylene content of 38.6% for pure polyethylene glycol. The relative contents of ethylene oxide and propylene oxide in polyethylene–polypropylene glycols has been determined using combined

83

Introduction to Polymer Analysis pyrolysis–gas chromatography calibrated with polyethylene glycol and polypropylene glycol standards [68].

3.7.3 IR Spectroscopy Simak [75] investigated absorptions of the C–O–C group in polyethylene terephthalate in the IR region at 1.58, 1.56 and 26.31 μm. Oxymethylene groups have also been determined [76, 77].

3.7.4 NMR Spectroscopy Determination of the content of oxyethylene groups in ethylene oxide–propylene oxide can also easily be carried out by NMR spectrometry without chemical splitting of the ether linkage, identifying the base compounds by this method is difﬁcult. Several methods for the cleavage of ethers have been studied, but few applied to the identiﬁcation of the base compounds of the polyurethane polyethers. Chu and co-workers[78] showed, using 1H- and 13C-NMR, that chlorobutyl rubber and bromobutyl rubber contained oxymethylene groups.

3.8 Anhydride Groups Van Houwelingen [8] discussed the determination of anhydride groups in a resin derived from octadecene-1 and maleic anhydride. The common method for anhydride groups involving reaction with an excess of aniline and subsequent back titration of the excess is unsuitable because the reactivity of the anhydride group is low. Even after hydrolysis with aqueous pyridine (containing 40% v/v of water) in a Parr bomb at 150 °C for 4 hours, anhydride groups are still seen in IR spectra. A suitable method for determining the anhydride group is titration with aqueous potassium hydroxide in pyridine after previous esteriﬁcation of the carboxyl group with diazomethane. This esteriﬁcation is carried out in diethyl ether methanol (9 + 1). After methylation, which takes about 10 minutes for 0.5 g of sample, solvents are removed by evaporation and a portion of the derivatised polymer is dissolved in pyridine and titrated. In the IR spectra of the resin before and after methylation, the absorption band of the acid group at 1710 cm–1 (5.84 μm) disappears and a carbonyl band of the ester at 7104 cm–1 (5.74 μm) is formed. The acid content of the sample is found from the difference in titres of an unmethylated and a methylated product.

84

Functional Groups

3.9 Total Unsaturation The methods described in this section cover only the determination of total unsaturation in polymers. As in the case of elemental analysis, functional groups can occur in polymers over a wide range of concentration, ranging from a few parts per million (ppm), as occurs for example in the case of end-groups, or micro-unsaturation to the percentage range. The occurrence of two or three double bonds per thousand carbon atoms in, for example, polyethylene, can affect intrinsic polymer properties and can help to distinguish between polyethylene manufactured by different manufacturers using different processes. As such, this type of determination falls within the province of microstructure, which is discussed in Chapter 9. At the other end of the concentration scale, a copolymer of, for example, ethylene and vinyl acetate, will contain between 1% and 90% of ether monomer. Analysis of polymers and copolymers in this concentration range is considered next. A wide range of physical and chemical techniques have been employed in such analyses. The application of these techniques to the determination of particular functional groups is discussed under two main headings.

3.9.1 Hydrogenation Methods Unsaturation in polymers is usually measured by physical techniques, as discussed later. This is especially so in the case of low levels of unsaturation, or in instances where a distinction must be made between types of unsaturation. Hydrogenation techniques have been used to measure higher levels (0.5–5.0 mole%) of total unsaturation in polymers, a good example of which is the determination of terminal unsaturation in polystyrene a oligomers [79–81] (low molecular weight polymers), e.g., polystyrene dimer:

3.9.2 Halogenation Methods In acidic medium, potassium bromide and potassium bromate produce bromine stoichiometrically, and this is the basis of a titration method [82], which has been used to determine double bonds in polymethylacrylate. After bromination, excess bromine is estimated by the addition of potassium iodide and the iodine produced by the reaction is titrated with standard sodium thiosulfate to the starch end-point.

85

Introduction to Polymer Analysis

Hensen and Eatough [83] described a direct injection enthalpimetric determination of residual total unsaturation in polymers. This is based on the heat produced during the bromination of double bonds. Kolthoff and Mitarb [84] and Boyer [85] discuss the application of bromination to the determination of total unsaturation in di and polyoleﬁns. McNeil [86, 87] discussed methods for the determination of total unsaturation based on chlorination with radiolabelled chlorine (‘radiochlorine’, 35Cl). This method, with slight modiﬁcation, can determine unsaturation at the level of 0.01–0.1 mole%. If the speciﬁc activity of radiochlorine in the gas phase is known, the weight of chlorine in the polymer can be found by counting. The mole% unsaturation of the polymer is calculated from the weight per cent unsaturation by assuming that one atom of chlorine enters the polymer per double bond, producing a monochloro compound. With this assumption it was found that polyisobutene consumes a mean of 1.2 chlorine atoms per double bond producing a monochloro compound. This is probably an indication that side reactions such as addition of chlorine are occurring to a small extent. Because the true unsaturation is close to one double bond per chain, it follows that the main chain-breaking reaction during the cationic polymerisation of isobutene is proton transfer. An elegant method for determining unsaturated compounds is the in situ generation of a halogenating agent (e.g., bromine) by constant-current coulometry. The amount of reagent consumed is proportional to the amount of electricity used to generate the reagent. The reaction between the double bond and bromine is catalysed by mercury(II)chloride. The advantages of a coulometric method are that standard reagent is not needed; minimum side reactions (substitution) occur because the bromine concentration is kept low; and the method is precise, suitable for low levels, and can be readily automated. Von Houwelinger [8] used coulometric bromination to determine vinyl ester endgroups in polyethylene terephthalate formed by thermal chain scission with the subsequent liberation of acetaldehyde.

86

Functional Groups The constant-current generation of bromine is carried out in a medium of dichloroacetic acid, hexaﬂuoroisopropanol, 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 μl and 1.0 μmol of methyl vinyl terephthalate (corresponding to 30 μmol and 2 μmol of vinyl ester end groups per kilogram of polymer) were recovered quantitatively (recoveries of 99.8% and 98.5%, respectively). The coulometric analysis must be completed within 30 minutes because hydrolysis of the vinyl ester end group is no longer negligible at longer times.

x

(a)

20

Br2 consumption/ mmol kg-1 of PET

Br2 consumption/μmol

The vinyl ester end group is not the only reactive moiety in polyethylene terephthalate that consumes bromine because impurities also do so. To determine this background in a second sample, the vinyl ester end group is previously hydrolysed at 80 °C in a dichloroacetic acid–water medium. Figure 3.8(a) shows the relationship between bromine consumption and hydrolysis time for methyl vinyl terephthalate. Figure 3.8(b) represents the same relationship for two polymers containing a high and negligible amount of vinyl ester end groups.

16 12 8 4

(b)

20 16 12 8 4

A

x

x x x

4

8

4 12 16 20 24 28 Hydrolysis time/h

8

x

12 16 20 24 28

Figure 3.8 Coulometric determination of vinyl ester end groups in polyethylene terephthalate. Effect of hydrogenation. Reproduced with permission from G.D.B. van Houwelingen, Analyst, 1981, 106, 1057. © 1987, RSC It can be seen that the vinyl ester end group in methyl vinyl terephthalate is hydrolysed completely after 4–6 hours at 80 °C. For polymer A, a sharp decrease occurs, which can mainly be attributed to the hydrolysis of the vinyl ester end group in the polymer. A much smaller effect is observed due to the hydrolysis of the bromine-consuming impurities (see relationship for polymer B, Figure 3.8(b)). This small effect is corrected by extrapolating this relationship from time 6 hours to time zero (i.e., 0.7 mmol/ kg).

87

Introduction to Polymer Analysis The vinyl ester end-group content is calculated by subtracting the background (i.e., value measured after 6 hours hydrolysis + 0.7 mmol/kg) from the content originally measured. The standard deviation of the method is 0.2 mmol/kg.

3.9.3 Iodine Monochloride Procedure Styrene butadiene copolymers contain residual double bonds which enable the butadiene content of the copolymer to be determined:

In this procedure, the polymer is reacted with an excess of standard iodine monochloride dissolved in glacial acetic acid (‘Wijs reagent’):

After reaction completion, excess iodine monochloride is reacted with potassium iodide and the liberated iodine estimated by titration with standard sodium thiosulfate:

88

Functional Groups ICl + KI = KCl + I2 The double bond content of the original polymer can be calculated from the measured consumption of iodine monochloride. Crompton and Reid [89] described procedures for the separation of high-impact polystyrene into the free rubber plus rubber grafted polystyrene plus copolymerised rubber and gel fraction; and for using the iodine monochloride procedure to estimate total unsaturation in the two separated fractions. To separate a sample into gel and soluble fractions, it is ﬁrst dissolved in toluene. Only gel remains undissolved. Methanol is then added, which precipitates the polystyrene–rubber graft, ungrafted rubber and polystyrene. Any styrene monomer, soap or lubricant remains in the liquid phase, which is separated from the solids and rejected. The toluene solubles are separated from the solid gel by centrifugation and made up to a standard volume with toluene. The gel is then dried in vacuo and weighed. Gel and toluene-soluble fractions are reserved for determination of unsaturation. To determine unsaturation in styrene–butadiene rubbers with good accuracy using the iodine monochloride procedure, it was necessary to contact the sample with chloroform for 15 hours before reaction with iodine monochloride. Figure 3.9 shows a plot of sample size against the determined iodine value; even with a 30 hour reaction period, a constant iodine value (approximately 320) is obtained only when the sample size is b0.05 g, i.e., a ﬁve-fold excess of iodine monochloride reagent. The solid gel, separated from high-impact polystyrene by solvent extraction procedures, is completely insoluble in chloroform and in the iodine monochloride reagent solution. A contact time with chloroform of 90 hours and a reaction time of 75 minutes with the reagent is required. Crompton and Reid [89] used these procedures to study the distribution of rubber added in several laboratory preparations of high-impact polystyrene containing 6 wt% of a styrene–butadiene rubber and 94% styrene, i.e., theoretical 4.1% butadiene. The results in Table 3.9 show the way in which the added unsaturation of 4.1% butadiene distributes between the gel and soluble fractions. The butadiene content of the separated gel remains fairly constant, in the 20–25% region, regardless of the quantity of gel in the sample. As the gel content increases, therefore, more of the rubber becomes incorporated into the gel and less remains as free rubber or soluble graft. The recovered unsaturation lies mainly in the 90–95% region, indicating that loss of unsaturation due to grafting or crosslinking reactions occurs only to the extent of 5–10%.

89

Introduction to Polymer Analysis

Determined iodine value

330 x

320

x

310

x

x

30 hours reaction with 50 ml 0.2N WIJS reagent

300 290 x

280 7 hours reaction with 50 ml 0.2N WIJS reagent

270 260 250

0.02

0.04

0.06 0.08 0.10 0.12 Weight rubber for analysis (g)

0.14

0.16

Figure 3.9 Inﬂuence of excess iodine monochloride reagent and reaction time on determination of the iodine value of rubber. Source: Author’s own ﬁles.

Albert [90] compared determinations of butadiene in high-impact polystyrene by an IR method and by the iodine monochlorine method described by Crompton and Reid [91]. The IR method is based on a characteristic absorbance in the IR spectrum associated with the transconﬁguration in polybutadiene: Trans-polybutadiene units:

Cis-polybutadiene units:

90

Functional Groups

Table 3.9 Distribution of butadiene between soluble and gel fractions obtained from polystyrene containing different amounts of gel Gel content of sample (wt%)

Butadiene content isolated gel (wt%)

Soluble graft content A (calculated on original sample) (wt%)

Gel butadiene content B (calculated on original sample) (wt%)

Total butadiene content (A+B) (calculated on original sample) (wt%)

Amount of original rubber unsaturation in the sample C = (A+B) × 100%

-

-

3.5

-

3.5

85

4.7

19.5

2.8

0.9

3.7

90

5.6

16.2

2.9

0.9

3.8

93

8.9

23.3

1.5

2.1

3.6

88

11.8

20.0

1.5

2.4

3.9

95

Reproduced with permission from T.R. Crompton and V.W. Reid, Journal of Polymer Science Part A: General Papers, 1963, 1, 1, 347. © 1963, Wiley

Because different grades of high-impact polystyrene may contain elastomers with different transbutadiene 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 3.10. 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 in interpolymerised polymers because of crosslinking, which reduces rubber unsaturation. 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 to the 9.0% found by the titration method. The iodine monochloride method has been used for various polymers. These polymers include those which are highly unsaturated, such as polybutadiene and polyisoprene [91–94] and polymers having low unsaturation such as butyl rubber [95], and ethylene–propylene–diene terpolymers. Considerable work has been done investigating the side reactions of iodine monochloride with different polymers. [95]. These side reactions are substitution and splitting out rather than the desired addition reaction.

91

Introduction to Polymer Analysis

Table 3.10 Rubber content of high-impact polystyrenes (based on polybutadiene; PBD) Sample

PBD, iodine monochloride method (wt%)

PBD, IR method (wt%)

Standard: 6.0 wt% diene 55

6.2

-

Standard: 12 wt% diene 55

12.2

-

Standard 15 wt% diene 55

14.8

-

High-impact polystyrene 1

8.6

9.7

High-impact polystyrene 2

5.6

5.8

High-impact polystyrene 3

9.0

1.2

High-impact polystyrene 4

11.2

11.4

High-impact polystyrene 5

5.8

5.9

Reproduced from W.K. Albert, private communication, 1966.

3.9.4 IR Spectroscopy Fraga [96] described an IR thin-ﬁlm area method for the analysis of styrene–butadiene copolymers. The integrated absorption area between 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 of butadiene structure present. This method can be calibrated for bound styrene contents ranging from 25% to 100%. IR spectroscopy and pyrolysis gas chromatography have been used for the determination of unsaturation in ethylene–propylene–diene terpolymers [88]. Determination of extinction coefﬁcients for the various terpolymers is required for quantitative work. Brako and Wexler [97] described a useful technique for testing unsaturation in polymer ﬁlms such as polybutadiene and styrene–butadiene. They expose the ﬁlm to bromine vapour and record its spectrum before and after exposure. This results in marked changes in the IR spectrum. Noteworthy is the almost complete disappearance of bands at 13.691, 10.99, 10.36 and 6.10 μm (i.e., 730, 910, 965 and 1640 cm–1) associated with unsaturation. A pronounced band, possibly associated with a C–Br vibration appears at 550 cm–1 (12.73, 8.73 and 8.00 μm), which is due to exposure to bromine vapour. Exposure of butadiene–styrene copolymer to bromine vapour results in the disappearance of bands at 910 cm–1 and 965 cm–1 (10.99 μm and 10.36 μm) associated with unsaturation in the butene component of the copolymer. Some

92

Functional Groups alteration of the phenyl bands at 700 cm–1 and 765 cm–1 (14.28 μm and 10.37 μm) is evident. The loss of a band at 1550 cm–1 (6.45 μm) and appearance of a band at 1700 cm–1 (5.88 μm) are probably due to the action of acidic vapours on the carboxylate purifactant of the latex. Panyszack and Kovar [98] showed that results accurate to within 2.5% can be obtained in determinations by FT-IR of the butadiene content of styrene–butadiene copolymers. IR methods are reported to have serious disadvantages when applied to the determination of unsaturation in vulcanisation [99–103]. IR spectroscopy has been used for the determination of unsaturation in ethylene– propylene–diene terpolymers [104]. Determination of extinction coefﬁcients for the various terpolymers is required if quantitative work is to be done. Raman spectroscopy has been applied to the determination of unsaturation in styrene–butadiene–methyl methacrylate terpolymers [96, 105, 106].

3.9.5 NMR Spectroscopy An advantage of NMR is that it can distinguish between the different types of unsaturation that can occur in a polymer. NMR spectroscopy has been used to determine unsaturation in acrylonitrile– butadiene–styrene terpolymers [107], ethylene–propylene–diene terpolymers [108] and 1,2-polybutadiene [105]. Regarding acrylonitrile–butadiene–styrene terpolymers [105], NMR can determine ungrafted butadiene rubber in solvent extracts of these polymers. About 60–80% of the butadiene in the entire sample was present as ungrafted rubber. Using the compositional analysis of the grafted and ungrafted rubber, and the amount of ungrafted rubber extracted, one can calculate the composition of the graft. Figure 3.10 shows a typical NMR spectrum of the grafted material. No aromatic protons of styrene or acrylonitrile protons are seen in the NMR spectra. The vinyl content of the polybutadiene is about 20%. Polymerisation of ethylene and propylene results in a saturated copolymer. To vulcanise this rubber, some unsaturation must be introduced. This is commonly done by adding a few percent of non-conjugated diene (termonomer) such as dicyclopentadiene, 1,4-hexadiene, or ethylidene norbornene, during polymerisation. Because only one of the double bonds of the diene reacts during polymerisation, the other is free for

93

Introduction to Polymer Analysis vulcanisation. The amount of unsaturation left in the ethylene propylene diene terpolymers is of great interest because vulcanisation properties are affected.

8

7

6

5

4 ppm ()

3

2

1

0

Figure 3.10 Typical NMR spectrum of extracted ungrafted rubber from samples 1 through 6. Signal at 5.4 ppm: oleﬁnic protons of 1,4-polybutadiene. Signal at 5.1 ppm: oleﬁnic protons 1,2-polybutadiene. Signal at 3.5 ppm: dioxane (internal standard). Signal at 2.0 ppm: aliphatic protons of 1,4-polybutadiene. Signal at 1.2 ppm: methylene protons of the soap. Signals between 6 ppm and 7 ppm: impurities in hexachlorobutadiene. Source: Author’s own ﬁles. Sewell and Skidmore [108] used time-averaged NMR spectroscopy at 60 megacycles/s to identify low concentrations of non-conjugated dienes introduced into ethylene– propylene copolymers to permit vulcanisation. Although IR spectroscopy [109] and iodine monochloride unsaturation methods [110] have been used to determine or detect such dienes, these two methods can present difﬁculties. Identiﬁcation of the incorporated third monomer by IR spectroscopy at the low concentrations involved is not always practical; unsaturation is not usually detected in the high-resolution NMR spectra of these terpolymers because the signals from the oleﬁnic protons are of such low intensity that they are lost in background noise. The spectra obtained by time-averaged NMR are usually sufﬁciently characteristic to allow identiﬁcation of the particular third monomer incorporated in the terpolymers. Because the third monomer initially contains two double bonds, differing in structure and reactivity, the one consumed in copolymerisation may be distinguished from the one remaining for subsequent use in vulcanisation. Information concerning the structure of the remaining unsaturated entity may therefore be obtained. Table 3.11 shows the chemical shifts of oleﬁnic protons of several third monomers in the copolymers. Cyclooctadiene and dicyclopentadiene terpolymers have oleﬁnic protons with the same chemical shift (4.55 T), so these cannot be differentiated by this technique, but they may be distinguished by iodine monochloride. The hexadiene type of terpolymers may be identiﬁed by its oleﬁnic resonance at 4.7 T. 94

Functional Groups These three monomers have what appears as a single oleﬁnic resonance in the terpolymers, but, the two norbornadiene types of monomer each show two characteristic resonances. In the methylene noroborene terpolymer, the oleﬁnic resonances arise from two protons, each giving a separate signal; whereas in the ethylidene norbonene terpolymers this is only one proton, the signal of which appears as a doublet. In view of these considerations, it is more difﬁcult to detect the oleﬁnic resonance in the latter instance.

Table 3.11 Chemical shifts of oleﬁnic protons of third monomer in ethylene – propylene – diene terpolymers Third monomer

Chemical shift, T

Cyclooctadiene-1,5

4.55

Dicyclopentadiene

4.55

1,4-Hexadiene

4.7

Methylene norbornene

5.25 and 5.5

Ethylidene norbornene

4.8 and 4.9

Reproduced with permission from P.R. Sewell and D.W. Skidmore, Journal of Polymer Science, Polymer Chemistry Edition, 1968, 6, 8, 2425. © 1968, Wiley

Altenau and co-workers [111] applied time-averaging NMR to the determination of low percentages of termonomers such as 1,4-hexadiene, dicyclopentadiene and ethylidene norbornene in ethylene–propylene termonomers. They compared results obtained by NMR and the iodine monochloride procedure of Lee and co-workers [81]. The chemical shifts and splitting pattern of the oleﬁnic response were used to identify the termonomer. Figure 3.11 shows the time-averaged NMR spectra of ethylene–propylene terpolymers containing various dienes. Table 3.12 compares the amount of termonomers found by the NMR method of Altenau and co-workers [81] and by iodine monochloride procedures [113]. The termonomers were identiﬁed by NMR and IR spectroscopy. Table 3.12 shows that the data obtained by the NMR method agree more closely with the Lee, Kolthoff and Johnson iodine monochloride method [81] than with the iodine monochloride method of Kemp and Peters [113]. The difference between the latter two methods is best explained on the basis of side reactions occurring between the iodine monochloride and polymer because of branching near the double bond [114]. The reason for the difference between the NMR and Lee, Kolthoff and Johnson methods [114] is not clear. The reproducibility of the NMR method was ±10–15%.

95

Introduction to Polymer Analysis

Table 3.12 Determination of termonomer in ethylene–propylene diene terpolymers: comparison of methods (data shown as wt% termonomer) NMR, Altenau [111]

NMR, Lee, Koltoff and Johnson [81]

Iodine monochloride, Kemp and Peters [113]

7.3

5.0

Dicyclopentadiene

1.1

1.6

1,4-hexadiene

5.7

5.9

9.0

Ethylidene norbornene

2.8

3.6

4.5

Ethylidene norbornene

1.7

2.3

4.8

Ethylidene norbornene

4.6

5.4

6.0

Ethylidene norbornene

Termonomer

Reproduced with permission from A.G. Altenau, L.M. Headley, C.O. Jones and H.C. Ransaw, Analytical Chemistry, 1970, 42, 11, 1280. © 1970, ACS

6

5 ppm ()

4

Figure 3.11 Time-averaged NMR spectra of EPDM containing one of the following termonoments: upper spectrum: 1,4-hexadiene; middle spectrum: dicyclopentadiene; lower spectrum: ethylidene norbornene. Reproduced with permission from A.G. Altenau, L.M. Headley, C.O. Jones and H.C. Ransaw, Analytical Chemistry, 1970, 42, 1280. © 1970, ACS

96

Functional Groups

3.9.6 Pyrolysis–Gas Chromatography Van Schooten and Evenhuis [114, 115] applied their pyrolysis (500 °C)– hydrogenation–gas chromatographic technique to unsaturated ethylene–propylene copolymers, i.e., ethylene–propylene–dicyclopentadiene and ethylene–propylene– norbornene terpolymers. The programs show that very large cyclic peaks are obtained from an unsaturated ring: methyl cyclopentane is found when methyl norbornadiene is incorporated; cyclopentane when dicyclopentadiene is incorporated; methylcyclohexane and 1,2-methylcyclohexane when the addition compounds of norbornadiene with, respectively, isoprene and dimethylbutadiene are incorporated; and methylcyclopentane when the dimer of methylcyclopentadiene is incorporated. The saturated cyclopentane rings in the same ring system in equal concentrations give rise to peaks that are an order of magnitude smaller. Peaks which stem from the termonomers could therefore be used to determine its content if a suitable calibration procedure could be found. Van Schooten and Evenhuis [114, 115] subjected several terpolymers containing dicyclopentadiene, and having different amounts of unsaturation, to pyrolysis–gas chromatographic analysis, and plotted the height of the characteristic peaks (or ratio of the heights of these peaks to the height of the n-C peak) against unsaturation measured by ozone absorption [116]. A linear relationship was found between peak height ratio and ozone unsaturation up to about 16 double bonds per 1000 atoms. Similar curves were found for the methylcyclopentane or ethylcyclopentane peaks. McKillop [117] found that the application of pyrolysis–gas chromatography to the determination of unsaturation in vulcanisation had serious disadvantages. Vinyl groups in styrene–divinyl benzene copolymer have been determined by pyrolysis–mass spectrometry.

3.9.7 Derivitivisation–Gas Chromatography Ethylene glycol, 1,4-butane diol, terephthalic acid and isophthalic acid repeat units in terylene. Terylene is manufactured from terephthalic acid and ethylene glycol. Such polymers contain repeat units based on ethylene glycol (and indeed other glycols such as 1, 4-butane diol) and terephthalic acid and proportions of isophthalic acid. Allen and co-workers [118] developed a precise quantitative method for determining such units in terylene. The sample is subject to an alkaline hydrolysis and glycol and acidic products after conversion to the trimethylsilyl derivatives are analysed by gas chromatography.

97

Introduction to Polymer Analysis In this method, the sample is weighed into a 100 ml ﬂask and 50 ml of 1 N potassium hydroxide in 2-ethyoxyethanol is added. A condenser cooled by chilled water is attached and the reaction mixture protected from carbon dioxide by a tube packed with Ascarite absorbent and Drierite desicant. Flask contents are heated and maintained at reﬂux temperature for 10 minutes with constant stirring. The ﬂask is allowed to cool to room temperature and the hydrolysate adjusted to a pH of 1 using concentrated hydrochloric acid (5 ml). An internal standard is added to the ﬂash. Pyridine (25 ml) is added to dissolve the acids present and an aliquot sample centrifuged to remove potassium chloride. Approximately 50 μl of hydrolysed sample is allowed to react at room temperature for 5–10 minutes with 500 μl of N,Obis(trimethylsilyl)triﬂuoracetamide to form silyl ethers and esters of the glycols and acids, respectively. The silyated hydrolysate is chromatographed by injecting 0.1 μl of sample into the gas chromatograph. The silyl derivative of diethylene glycol is separated from the silyl derivative of ethylene glycol using a 3 m x 3.2 mm stainless-steel column packed with 100–200 mesh Chromosorb G-HP solid support containing 3% by weight of Versilube F-50 liquid phase (for a mixture of trichlorophenyl silicone (10%) and methyl silicone manufactured by General Electric). Figure 3.12 is a chromatogram obtained for the silyl derivatives of the hydrolysate of a polyethylene terephthalate. Dodecane is used as an internal standard and the column is operated isothermally at 127 °C with a nitrogen carrier gas ﬂow rate of 10 ml/min. Figure 3.12(c) is a chromatogram of the silyl derivatives of ethylene glycol, 1,4-butanediol and the cis and trans isomers of 1,4-cyclohexanedimethanol from the hydrolysate of an experimental polyester. A 1.8 mm × 6.4 mm glass column packed with 100–200 mesh Chromsorb W-HP solid support containing 10% by weight of Versilube F-50 liquid phase is used for this separation. Nonyl alcohol is used as an internal standard, the column operated at 120 °C for 8 minutes and programmed to 210 °C at 4 °C/min with a nitrogen carrier gas ﬂow rate of 20 ml/min. High boiling point acids, such as terephthalic and isophthalic acids, are separated using a 1.8 m × 3.2 mm stainless-steel column packed with 100–200 mesh Chromosorb W-HP solid support containing 10% by weight Versilube F-50 liquid phase. Figure 3.12(c) is the chromatogram obtained for the separation of the silyl derivatives of isophthalic and terephthalic acid from an experimental polyester. The column is operated isothermally at 183 ºC with a nitrogen carrier gas ﬂow rate of 33 ml/min. N-heptadecane was added as an internal standard to permit calculations of the weight per cent acids. In the preparation of polyester for compositional analysis by NMR spectroscopy, much higher rates of hydrolysis are achieved by reﬂuxing polyethylene terephthalate with a

98

Functional Groups mixture of sodium hydroxide, an alcohol (methanol) and a protic solvent (dimethyl sulfoxide) than is achieved with the usual alcoholic alkali reagents [120]. (a)

20

Diethylene glycol, TMS ether Internal standard

16 Liquid phase: 'Versilube' F-5O Column temp: 127°C N2 flow rate: 10 ml/min

Time, min

12

Ethylene glycol, TMS ether

8 4

0

(b)

36 trans 1,4-cyclohexanedimethanol, TMS ether cis 1,4 cyclohexanedimethanol, TMS ether Liquid phase: 'Versilube' F-50 20 Column temp: 120°C Hold 8 min Program to 190°C 1,4 butanediol, TMS ether at 8°C/min Internal standard

32 28 Time, min

20 16 12

Ethylene glycol, TMS ether

8 4 0

(c)

28

Time, min

24

Terephthalic acid, TMS ester

20

Isophthalic acid, TMS ester

16

Internal standard

12 8

Liquid phase: 'Versilube' F-5O Column temp: 183°C N2 flow rate: 33 ml/min

4 0

Figure 3.12 Gas chromatograms of hydrolysates of (a) silyl derivative of polyethylene terephthalate, (b) silyl derivative of ethylene glycol 1,4 butane diol and trans isomers of 1,4 cyclohexane diol dimethanol and (c) silyl derivative of iso- phthalic and terephthalic acids derived from a polyester. Reproduced with permission from B.J. Allen, G.M. Elser, K.P. Keller and H.D. Kinder, Analytical Chemistry, 1977, 49, 741. © 1977, ACS

99

Introduction to Polymer Analysis

3.9.8 Radiochemical Methods 3.9.8.1 Determination of unsaturation in butyl rubber Radiochemical methods involving 36Cl have been used for the measurement of higher levels of unsaturation (0.5–2 mole%) in butyl rubber [121, 122]. This method, with slight modiﬁcation, can also determine unsaturation at the 0.01–0.1 mole% level [87]. If the speciﬁc activity of the radiochlorine in the gas phase is known, the weight of chlorine in the polymer can be found by counting. The mole% unsaturation of the polymer was calculated from the weight per cent unsaturation by assuming that one atom of chlorine enters the polymer per double bond producing a monochloro compound. With this assumption, it was found that polyisobutene consumers a mean of 1.2 chlorine atoms per double bond, i.e., per molecule. This is probably an indication that side reactions such as addition of chlorine are occurring to a small extent. Because the true unsaturation is close to one double bond per chain, it follows that the main chain-breaking reaction during the cationic polymerisation of isobutene is proton transfer.

3.10 Alkyl and Aryl Groups 3.10.1 Alkali Fusion Reaction–Gas Chromatography This technique has been applied to the quantitative determination of alkyl and aryl groups in polysiloxanes [123] and of imides in aromatic polyamides and poly(amide imides) [124]. The method involves fusion of the polymer with powered potassium hydroxide, which converts alkyl and aryl groups in siloxanes into the corresponding hydrocarbons and amino and imino group to the corresponding amino or diamine:

100

Functional Groups The resulting volatile products are cold-trapped and subsequently determined by gas chromatography. A suitable temperature–time proﬁle when heating the samples must be established to obtain reliable results. Thus, when siloxanes are fused with alkali in a platinum boat, the reaction mixture is ﬁrst heated to 100 ºC then gradually raised to 300 ºC. The nitrogen compounds referred to above reacted very smoothly at 250 ºC. This avoided spattering of the reaction mixture and achieved quantitative clearage of methyl and phenyl groups in 5 minutes. Methyl contents in the range 6% to 36%, and phenyl contents in the range 10% to 81%, can be determined by this method giving results that are in good agreement with those obtained by NMR spectroscopy. Regarding nitrogen-containing polymers, the following polymer, upon alkali fusion, produced m-phenylene diamine in 100.7% yield.

3.11 Oxirane Rings Oxirane rings in epoxy resins have been determined by ring cleavage with pyridine– hydrochloric acid [125]. Cleavage with pyridine–hydrochloric acid, glacial acetic acid, hydrobromio acid or dioxon–hydrochloride has been employed as methods of cleavage of oxirane rings before their determination [126].

3.12 Amino Groups Amino groups have been determined by acetylation with acetic anhydride in dimethylacetamide. Diethylamine was added and the excess amine was titrated potentiometrically [127]. Heterogeneous derivitivisation with 1-ﬂuoro-2,4-dinitrobenzene has been used to determine amino groups in polyethylene terephthalate [8]: NO 2

NH2 + F

NO 2

NO 2 + HF

101

Introduction to Polymer Analysis The polymer is treated for 4 hours at 80 °C with 1-ﬂuoro-2,4-dinitrobenzene in an ethanol hydrogen carbonate medium. After washing out the excess of reagent, the dinitrophenyl group introduced is measured spectrophotometrically at 430 nm after dissolution of the derivatised polymer in methane sulfonic acid. The spectrophotometric method was calibrated with the 1-ﬂuoro-2,4-dinitrobenzene derivative of the model compound N,N´-bis(p-amino-phenyl) terephthalamide. The correctness of this procedure was conﬁrmed using 14C-labelled 1-ﬂuoro-2,4 dinotrobenzene and measurement of incorporated radioactivity. Other spectroscopic procedures have been used to determine primary amino groups include those based on p-amino-benzaldehyde [128], ninhydrin [129], dimethoxy trityl chloride [129]. A spectroﬂuorimetric method has also been described [130].

3.13. Amino and Imido Groups 3.13.1 Alkali Fusion–Gas Chromatography Schleuter and Siggia [131–133] and Frankowski and Siggia [134] used alkali fusion reaction gas chromatography for the analysis of imide monomers and aromatic polyimides, polyamides, and poly(amide–imdes). Samples are hydrolysed with a molten potassium hydroxide reagent at elevated temperatures in a ﬂowing inert atmosphere:

Volatile reaction products are concentrated in a cold trap before separation by gas chromatography. The identiﬁcation of the amine and/or diamine products aids in the

102

Functional Groups characterisation of the monomer or polymer; the amount of each compound generated is used as the basis for quantitative analysis. The average relative standard deviation of the method is ±1.0%.

Table 3.13 Structure, water content and decomposition temperature of the polymers studied Designation

Structure of repeat unit

Water (wt%)

Decomposition temperature (oC)

PI-1

6.5

385

PI-2

2.2

410

PI-3

3.4

410

PI-4

0.6

310

PA-1

5.8

315

PA-2

9.7

330

PAI-1

6.2

340

PAI-2

12.6

395

PAI-3

8.9

340

Reroduced with permission from D.D. Schlueter and S. Siggia, 1977, 49, 14, 2349. © 1977, ACS

103

Introduction to Polymer Analysis Table 3.13 summarises the chemical structures and sample designations of the polymers studied. Table 3.14 summarises the diamine recoveries obtained. These data represent the mole percentage of theoretical diamine based on the dry weight of sample and the idealised linear polymer repeat units depicted in Table 3.13. Recoveries were, in all cases, between 90% and 101% of theoretical values.

Table 3.14 Analysis of polyimides, polyamides and poly(amide-imides) by alkali fusion reaction gas chromatography Sample

Diamine produced

Mol% of theoreticala ± RSDb 5 min at 380 oC

30 min from 100 to 390 oC

PI-1

4,4a-Methylenedianiline

98.3 ± 0.8

97.5 ± 0.9

PI-2

m-Phenylenediamine

89.1 ± 0.7 91.2 ± 0.9 91.6 ± 1.3

91.0 ± 0.8 91.0 ± 0.7

PI-3

2,4-Toluenediamine 4,4a-Methylenedianiline

73.1 ± 0.2 97.8 ± 0.5 24.8 ± 0.3

73.2 ± 1.0 99.1 ± 1.2 25.9 ± 0.2

PA-1

m-Phenylenediamine

100.7 ± 0.6

97.0 ± 1.7 97.7 ± 1.7 97.9 ± 1.0 99.6 ± 2.6

PA-2

m-Phenylenediamine

100.7 ± 0.6 96.4 ± 1.5

92.9 ± 1.0 93.3 ± 0.6

PAI-1

m-Phenylenediamine

93.5 ± 0.9

93.1 ± 0.7 95.3 ± 0.5

PAI-2

m-Phenylenediamine

93.5 ± 1.1 94.1 ± 0.6 95.1 ± 1.2

97.4 ± 0.9

PAI-3

4,4a-Methylenedianiline

98.0 ± 1.1

98.8 ± 1.1

a

These recovery values are based on the structure shown in Table 3.13, which assume that one mole of diamine is produced for each mole of repeat unit b

The relative standard deviation is based on ﬁve or more determinations.

Reproduced with permission from D.D. Schlueter and S. Siggia, 1977, 49, 14, 2349. © 1977, ACS

104

Functional Groups In an attempt to correlate alkali fusion recoveries (expressed as weight percent nitrogen) with the elemental nitrogen analyses of the polymers, it was noted that the elemental nitrogen analyses of the polymers were higher. Analysis of the fusion residue indicated that no detectable amounts of nitrogen were present. This led to the discovery that ammonia was also produced from the samples. When the nitrogen contribution from ammonia was added to that of the diamine, the total was in good agreement with the elemental value. Table 3.15 lists the results obtained from the analysis of mixtures of polymers. The amount of m-phenylenediamine produced was, in all cases, within 1.2% relative of the theoretical value.

Table 3.15 Analysis of polymer mixtures by alkali fusion reaction gas chromatography Polymer mixture

Micromoles of m-phenylenediamine a

Taken PI-2

2.24

PAI-2

9.26

PA-1

7.23

PA-2

7.43

PAI-1

4.99

PI-2

7.67

PAI-1

5.28

PAI-2

7.09

PA-1

5.51

PAI-2

5.81

PI-2

2.89

PA-2

8.79

PAI-1

3.25

PA-1

2.76

Recovery (%)

Found 11.50

11.36

98.8

14.66

14.71

100.3

12.66

12.58

99.4

12.37

12.48

100.9

14.21

14.33

100.8

14.80

14.98

101.2

a

These values were calculated from the weight of polymer taken and were corrected for the weight of absorbed water (Table 3.13) and the average experimental recovery Reproduced with permission from S.P. Frankoski and S. Siggia, Analytical Chemistry, 1972, 44, 3, 507. © 1972, ACS

3.14 Nitrile Groups An IR method has been described [135] for the compositional analysis of styrene– acrylonitrile copolymers. In this method, relative absorbance between a nitrile N(CN) mode at 2272 cm–1 (4.4 μm) and a phenyl N(CC) mode at 1613 cm–1 (6.2 μm) is used.

105

Introduction to Polymer Analysis A near-IR method using combination and overtone bands has been used [136] for carrying out the same analyses. The bands which occur in the region 6250–4545 cm–1 (1.6–2.2 μm) result from overtones and combination tones which occur, respectively, in the regions 6250–5555 cm–1 (1.6–1.8 μm) and 5263–4545 cm–1 (1.9–2.2 μm). The band near 5952 cm–1 (1.68 μm) is assigned as an overtone of phenyl N(CH) mode near 3000 cm–1 (3.33 μm) (3000 × 2 = 6000 cm–1 = 1.67 μm) and its absorbance is directly proportional to the styrene content of the copolymer and the ﬁlm thickness. The band near 5714 cm–1 (1.75 μm) is assigned as an overtone of the aliphatic N(CH) mode near 2900 cm–1 (3.45 μm) (2900 × 2 = 5800 cm–1 = 1724 μm) and acrylonitrile and styrene absorb at this wavelength. Bands at 5235 cm–1 and 5123 cm–1 (1.910 μm and 1.952 μm) are combination tones of acrylonitrile and are assigned as, N(CN) + N asym (CH3) (2237 + 2940 = 5177 cm–1 = 1.932 μm) and N(CN) + N asym (CH3) (2237 + 2870 = 5107 cm–1 = 1.958 μm), respectively. For calculation of the absorbance of these four characteristic bands, two baseline methods were used. The method using the extrapolated baseline from 2 introduces less deviation than the other baseline method. The absorbance ratio A1.675/A1.910 proved to be the best one for analytical measurements.

3.15 Silicon Functions Total silanol (SiOH) silane hydrogen (SiH) groups, and tetrapropyoxysilane and diphenylmethyl silanol crosslinking agents Dubiel and co-workers [137] described methods for determining these reactive components in room temperature-vulcanised silicone foams. Total SiOH and SiOH are determined by FT-IR spectrometry, the SiOH peak at 2.71 μm and the SiH peak at 4.61 μm were used for quantiﬁcation. Tetrapropoxysilane content was determined by gas chromatography using a solid capillary open tubular (SCOT) column and linear programmed temperature control. The diphenylmethylsilanol content was determined by gel permeation chromatography using the tetrahydrofuran solvent. The preferred approach is to use secondary standards, i.e., well-characterised compounds or polymers that quantitatively react with the reagent to give the desired product. For example, when only methane is determined, GE Viscasil 60,000, a high molecular weight polydimethylsiloxane, is used. DC 704, a polymethylphenylsiloxane, is used as the secondary standard when both methane and benzene are determined. In all cases, standards are trapped and chromatographed in exactly the same manner as reaction products. The best straight-line calibration curves are determined by a least-square regression curve ﬁtting computer program. Various other reagents have been used for the determination of organic substituents bonded to silicon in organosilicon polymers (Table 3.16). 106

Bromine in glacial acetic acid

Phosphorus pentoxide and water

Powered potassium hydroxide

Sulfuric acid

Ethylbromide in the presence of aluminium chloride

Phosphorus pentoxide and water

Phosphorus pentoxide

90% sulfuric scid

Sodium hydroxide pellets

Potassium hydroxide pellets

Ethyl and phenyl

Methyl and ethyl

Methyl

Phenyl

Vinyl

Vinyl

Vinyl

Vinyl

Vinyl

TC: Thermal conductivity detector Source: Author’s own ﬁles

60% aqueous KOH in DMSO

Phenyl

Reagent

Phenyl

Group determined

Methane and ethane Methane

2 h at 250-270 oC 20 min at 280-300 oC

Ethylene Ethylene

75-250 oC at 10 oC/min and 1 h at 250 oC 300 oC for 15 min

Ethylene

Ethylene

Ambient to 500 oC

Heat with Meker burner

Ethylene

80-600 oC over 40 min

Hexaethyl benzene

Ethane and benzene

30-580 oC over 45 min

-

Bromobenzene

Benzene

Product

Boiling solution

2 h at 120 C

o

Reaction conditions

GC-FID

Colorimetric

GC-TC

GC-FID

GC-FID

Gravimetric

Gas burette

Gas burette

GC-FID

Titration of excess bromine

GC

Analysis

[149]

[148]

[147]

[146]

[145]

[144]

[142, 143]

[141]

[140]

[139]

[138]

Reference

Table 3.16 Reaction methods for the quantitative determination of organic substitutes bonded to silicon

Functional Groups

107

Introduction to Polymer Analysis

References 1 J. Majewski, Polimery, 1973, 18, 3, 142. 2 R.S. Stetzler and C.F. Smullin, Analytical Chemistry, 1962, 34, 2, 194. 3 L.S. Bark and J.K. Grime, Analyst, 1972, 97, 911. 4 D.F. Fritz, A. Sahil, H.P. Keller and E. Kovats, Analytical Chemistry, 1979, 51, 1, 7. 5 L. Yu Spirin and T.A. Yatismirskaya, Vysokomolekulyarnye Soedineniya Seriya A, 1973, 15, 11, 2595. 6 J.S. Fritz and G.H. Schenk, Analytical Chemistry, 1959, 31, 11, 1808. 7 T.R. Crompton, Unpublished work. 8 G.D.B. Van Houwelingen, Analyst, 1981, 106, 1267, 1057. 9 M. Tanaka and I. Kojima, Analytica Chimica Acta, 1968, 41, 75. 10 I.I. Kaduji and J.H. Rees, Analyst, 1974, 99, 1180, 435. 11 R.M. Smith and M. Dawson, Analyst, 1980, 105, 1246, 85. 12 F.F. Li-Ho, Analytical Chemistry, 1973, 45, 3, 603. 13 C.S.Y. Kim, A.L. Dodge, S.F. Lau and A. Kawsaki, Analytical Chemistry, 1982, 54, 2, 232. 14 F.D. Brako and A.S. Wexler, Analytical Chemistry, 1963, 35, 12, 1944. 15 D.E. Johnson, J.R. Lyerla, Jr., T.T. Horikawa and L.A. Pederson, Analytical Chemistry, 1977, 49, 1, 77. 16 S. Crisp, B.G. Lewis and A.D.J. Wilson, Journal of Dental Research, 1975, 54, 6, 1238. 17 J.S. Tan and S.P. Gasper, Macromolecules, 1973, 6, 5, 741. 18 E.R. Garret and R.L. Guile, Journal of American Chemical Society, 1951, 73. 19 J. Douglas, A. Timnick and R.L. Guile, Journal of Polymer Science Part A: General Papers, 1963, 1, 5, 1609.

108

Functional Groups 20 A.R. Mathieson and J.V. McLaren, Journal of Polymer Science A, 1965, 3, 2555. 21 A. Katchalsky, N. Shavit and H.J. Eisenberg, Journal of Polymer Science, 1954, 13, 68, 69. 22 A.R. Mathieson and J.V. McLaren, Journal of the Chemical Society, 1960, 3581. 23 J.C. Leyte and M. Mandel, Journal of Polymer Science, Polymer Physics Edition, 1964, 2, 1879. 24 M. Mandel and J.C. Leyte, Journal of Polymer Science, 1962, 56, 23. 25 J.C. Oxley and W.D. Perkins, Analysis, 1987, 15, 61. 26 J. Tackett, Applied Spectroscopy, 1990, 44, 9, 1581. 27 R.D. Law, Journal of Polymer Science, 1971, 9, 589. 28 J. Majer and J. Sodomka, Chemicky Prumsyl, 1975, 25, 11, 601. [Chemical Abstracts, 1976, 85, 34101] 29 D. Munteanu and N. Savu, Revista de Chimie (Bucharest), 1976, 27, 10, 902. [Chemical Abstracts, 1977, 86, 121973D] 30 J.C. Bevington, D.E. Eaves and R.L. Vale, Journal of Polymer Science, 1958, 32, 317. 31 J.C. Bevington, Transactions of the Faraday Society, 1960, 56, 1762. 32 W. Aydin, B.U. Kaezmar and R.C. Schulz, Angewandte Makromolekulare Chemie, 1972, 24, 171. 33 G. Leukrath, Gummi Asbest und Kunststoffe, 1976, 29, 9, 585. [Chemical Abstracts, 1977, 86, 44183S] 34 G.G. Esposito and M.H. Swann, Analytical Chemistry, 1962, 34, 9, 1048. 35 D.F. Percival, Analytical Chemistry, 1963, 35, 2, 236. 36 J. Haslam, J.B. Hamilton and A.R. Jeffs, Analyst, 1958, 83, 983, 66. 37 J. Haslam and J. Jeffs, Journal of Applied Chemistry, 1957, 7, 24.

109

Introduction to Polymer Analysis 38 E.P. Samsel and J.A. McHard, Industrial & Engineering Chemistry, Analytical Edition, 1942, 14, 9, 750. 39 J. Helmroth, Polyvehromarium Plast, 1973, 3, 7. 40 J. Majer and J. Sodenka, Chemicky Prumsyl, 1975, 25, 11, 601. 41 A.G. Siryuk and R.A. Bulgakova, Vysokomolekulyarnye Soedineniya Series B, 1977, 19, 2, 152. [Chemical Abstracts, 1977, 86, 140597a.] 42 K. Czaja, M. Nowadowska and J. Zubek, Polimery Tworzywa Wielkoczasteckowe, 1976, 21, 4, 158. [Chemical Abstracts, 1976, 85, 124396h] 43 R.S. Porter, S.W. Nicksic and J.F. Johnson, Analytical Chemistry, 1961, 35, 12, 1948. 44 G.W. Tindall, R.L. Perry and J.L. Little, Analytical Chemistry, 1991, 63, 13, 1251. 45 V. Aydin, B.U. Kaczmar and R.C. Schulz, Angewandte Makromolekulare Chemie, 1972, 24, 171. 46 E.M. Barrall II, R.S. Porter and J.F. Johnson, Analytical Chemistry, 1963, 35, 1, 73. 47 R.S. Porter, A.S. Hoffman and J.F. Johnson, Analytical Chemistry, 1962, 34, 9, 1179. 48 D.L. Miller, E.P. Samsel and J.G. Cobler, Analytical Chemistry, 1961, 33, 6, 677. 49 R.P. Wool and W.O. Statton, Journal of Polymer Science, Polymer Physics Edition, 1974, 12, 8, 1575. 50 M. Goldstein, M.E. Seeley, H.A. Willis and V.J.I. Zichy, Polymer, 1973, 14, 11, 530. 51 N. Grassie and N.A. Weir, Journal of Applied Polymer Science, 1965, 9, 963. 52 N. Grassie and N.A. Weir, Journal of Applied Polymer Science, 1965, 9, 975. 53 R.G. Schole, J. Bedworszyk and E. Tamano, Analytical Chemistry, 1966, 38, 2, 331.

110

Functional Groups 54 J.N. Shaw and M.C. Marshall, Journal of Polymer Science, Part A1, 1968, 6, 449. 55 E.N. Getmanenko and E.M. Perepletchikova, Zhurnal Analiticheskoi Khimii, 1974, 29, 4, 830. 56 K. Kato, Journal of Applied Polymer Science, 1973, 17, 1, 105. 57 D.G. Anderson, K.E. Isakson, J.T. Vandeberg, M.Y.T. Jao, D.J. Tessari and L.C. Afremow, Analytical Chemistry, 1975, 47, 7, 1008. 58 F. Viebok and C. Brechner, Berichte der Deutschen Chemischen Gesellschaft, 1930, 63, 3207. 59 S. Paul and B. Ranby, Analytical Chemistry, 1975, 47, 8, 1428. 60 K. Peltonen, P. Pfäfﬂi and A. Itkonen, Analyst, 1985, 110, 1173. 61 E.W. Neumann and H.G. Nadeau, Analytical Chemistry, 1963, 35, 10, 1454. 62 V. Miyauchi, T. Takeshita, M. Akhashi and R. Machida, Journal of Applied Polymer Science, 1987, 34, 7, 2601. 63 A.D. Hammerich and F.G. Willeboordse, Analytical Chemistry, 1973, 45, 9, 1696. 64 W.B. Swann and J.P. Dux, Analytical Chemistry, 1961, 33, 4, 654. 65 I. Zenan, L. Novak, L.Mitter, J. Stekla and O. Holendova, Journal of Chromatography, 1976, 119, 581. 66 J. Urbanski, Plaste und Kautschuk, 1968, 15, 260. 67 R. Kretz, Fresenius’ Zeitschrift für Analytische Chemie, 1960, 176, 421. 68 A. Steyermark, Journal of the Association of Ofﬁcial Analytical Chemists, 1944, 38, 367. 69 S. Ehrlich-Rogozinski and A. Patchornik, Analytical Chemistry, 1964, 36, 4, 840. 70 M.H. Karger and Y. Mazur, Journal of the American Chemical Society, 1968, 90, 3878. 71 K. Tsuji and K. Kounishi, Analyst, 1974, 99, 1174, 54.

111

Introduction to Polymer Analysis 72 A. Cervenka and C.H. Merrall, Private Communication, 1975. 73 A. Mathias and N. Mellor, Analytical Chemistry, 1966, 38, 3, 472. 74 J.B. Stead and A.H. Hindley, Journal of Chromatography, 1969, 42, 470. 75 P. Simak, Die Makeomolekulare Chemie, Makromolecular Symposia, 1986, 5, 61. 76 G. Meszlenyi, E. Juhasz and M. Lelkes, Tenside Surfactants Detergents, 1994, 31, 2, 83. 77 K. Konische and Y. Konoh, Japan Analyst, 1966, 15, 1110. 78 C.Y. Chu, K.N. Watson and R. Vukov, Rubber Chemistry and Technology, 1987, 60, 636. 79 D.C. Pepper and P.H. Reilly, Proceedings of the Chemical Society, 1961, 1, 460. 80 S.G. Gallo, H.K. Wiese and J.F. Nelson, Industrial & Engineering Chemistry, 1948, 40, 7, 1277. 81 T.S. Lee, I.M. Kolthoff and E. Johnson, Analytical Chemistry, 1950, 22, 8, 995. 82 E.C. Kuryatnikov, R.V. Vigert and A.A. Berlin, Vijn L’viv Politekhinst, 1971, 57, 36. 83 L.D. Hansen and D.J. Eatough, Thermochimica Acta, 1987, 117, 37. 84 J.M. Kolthoff and J. Mitarb, Journal of Polymer Science, 1947, 2, 199. 85 R.F. Boyer, Journal of Physical and Colloid Chemistry, 1947, 51, 1, 80. 86 I.C. McNeill, Polymer, 1963, 4, 15. 87 R. McGuchan and I.C. McNeill, Journal of Polymer Science: Polymer Chemistry Edition, 1966, 4, 9, 2051. 88 R. Hank, Rubber Chemistry and Technology, 1967, 40, 3, 936. 89 T.R. Crompton and V.W. Reid, Journal of Polymer Science Part A: General Papers, 1963, 1, 1, 347. 90 N.K. Albert, Woodbury Research Laboratory, Shell Chemical Co Ltd, Woodbury, USA, Private Communication, October 1966.

112

Functional Groups 91 A. Basch and M. Lewin, Journal of Polymer Science: Polymer Chemistry Edition, 1973, 11, 12, 3071. 92 D. Dollimore and B. Holt, Journal of Polymer Science, Polymer Physics Edition, 1973, 11, 9, 1703. 93 D.S. Varma and V. Narasimhan, Journal of Applied Polymer Science, 1972, 16, 12, 3325. 94 M. Funt and J.H. Magill, Journal of Polymer Science, Polymer Physics Edition, 1974, 12, 1, 217. 95 B.V. Kokta, J.L. Valade and W.N. Martin, Journal of Applied Polymer Science, 1973, 17, 1, 1. 96 D.W. Fraga, Shell Chemical Co Ltd, Emeryville Research Center, Emeryville, CA, USA, Private Communication. 97 F.D. Brako and A.S. Wexler, Analytical Chemistry, 1963, 35, 12, 1944. 98 M. Panyszack and J. Kovar, Canadian Journal of Spectroscopy, 1986, 31, 130. 99 H.C. Dinsmore and D.C. Smith, Rubber Chemistry and Technology, 1949, 22, 2, 572. 100 D.L. Harms, Analytical Chemistry, 1953, 25, 8, 1140. 101 D. Hummel, Rubber Chemistry and Technology, 1959, 32, 3, 854. 102 M. Lerner and R.C. Gilbert, Analytical Chemistry, 1964, 36, 7, 1382. 103 M. Tyron, E. Horowicz and J.J. Mandel, Journal of Research of the National Bureau of Standards, 1955, 55, 219. 104 R. Hank, Rubber Chemistry and Technology, 1967, 40, 3, 936. 105 C. Shibata, M. Yamazaki and T. Tabeuchi, Bulletin of the Chemical Society of Japan, 1977, 50, 1, 311. 106 H.J. Sloane and R. Bramston-Cooke, Applied Spectroscopy, 1973, 27, 3, 217. 107 R.R. Turner, D.W. Carlson and A.G. Altenau, unpublished work. 108 P.R. Sewell and D.W. Skidmore, Journal of Polymer Science, Polymer Chemistry Edition, 1968, 6, 8, 2425.

113

Introduction to Polymer Analysis 109 W. Cooper, D.E. Eaves, M.E. Tunnicliffe and G. Vaughan, European Polymer Journal, 1965, 1, 2, 121. 110 M.E. Tunnicliffe, D.A. Mackillop and R. Hank, European Polymer Journal, 1965, 1, 4, 259. 111 A.G. Altenau, L.M. Headley, C.O. Jones and H.C. Ransaw, Analytical Chemistry, 1970, 42, 11, 1280. 112 D. Yang, S.D. Li, W.W. Fu, J.P. Zhong and D.M. Jia, Journal of Applied Polymer Science, 2003, 87, 2, 199. 113 A.R. Kemp and H. Peters, Industrial & Engineering Chemistry, Analytical Edition, 1943, 15, 7, 453. 115 J. Van Schooten and J.K. Evenhuis, Polymer, 1965, 6, 11, 561. 114 J. Van Schooten and J.K. Evenhuis, Polymer, 1965, 6, 7, 343. 116 B. Hoer and E.C. Kooyman, Analytica Chimica Acta, 1951, 5, 550. 117 D.A. MacKillop, Analytical Chemistry, 1968, 40, 3, 607. 118 B.J. Allen, G.M. Elser, K.P. Keller and H.D. Kinder, Analytical Chemistry, 1977, 49, 6, 741. 119 Pyrolysis GC/MS/IR Analysis of Kraton 1107, Hewlett Packard Application Note 228-100, 1989. 120 G.W. Tindall, R.L. Perry, J.L. Little and A.T. Spaugh, Jr., Analytical Chemistry, 1991, 63, 13, 1251. 121 X.H. Li, Y.Z. Meng, Q. Zhu and S.C. Tjong, Polymer Degradation and Stability, 2003, 81, 1, 157. 122 L.E. McNeill, Polymer, 1963, 4, 15. 123 P.G.M. van Statum and J. Dvorák, Journal of Chromatography, 1972, 71, 9. 124 W.S. Richardson and A. Sacher, Journal of Polymer Science, 1963, 10, 4, 353. 125 G.A. Stenmark, Analytical Chemistry, 1957, 29, 9, 1367. 126 A.J. Burbetaki, Analytical Chemistry, 1956, 28, 12, 2000.

114

Functional Groups 127 E.A. Emelin, V. Savinov and L.B. Sakolov, Journal of Analytical Chemistry, USSR, 1973, 28, 1188. [Chemical Abstracts, 1973, 79, 146959] 128 V. Gerschchenko, B.F. Blinov and B. Zinin, Plasticheskie Massy, 1975, 12, 12. 129 R.K. Gaur, P. Sharma and K.C. Gupta, Analyst, 1989, 114, 1147. 130 Y. Eckstein and P. Dreyfuss, Analytical Chemistry, 1980, 53, 3, 537. 131 D.D. Schleuter and S. Siggia, Analytical Chemistry, 1977, 49, 12, 2343. 132 D.D. Schleuter and S. Siggia, Analytical Chemistry, 1977, 49, 12, 2349. 133 D.D. Schleuter, University of Massachusetts, 1976. [PhD Thesis] 134 S.P. Frankoski and S. Siggia, Analytical Chemistry, 1972, 44, 3, 507. 135 G.A. McCrory and R.T. Scheddel, Analytical Chemistry, 1958, 30, 7, 1303. 136 T. Takeuchi, S. Tauge and Y. Sugimura, Journal of Polymer Science, Polymer Chemistry Edition, 1968, 6, 12, 3415. 137 S.V. Dubiel, G.W. Grifﬁth, C.L. Long, G.K. Baker and R.E. Smith, Analytical Chemistry, 1983, 55, 9, 1533. 138. R.D. Parker, Dow Corning Corporation, Barry, Wales, unpublished procedure 139. G. Gritz and H. Hurcht, Zeitschrift für Anorganische und Allgemeine Chemie, 1962, 317, 35. 140. V.M. Krasikova, A.N. Kaganova and V.D. Lobtov, Journal of Analytical Chemistry USSR, 1971, 28, 1458. 141. M.G. Voronkov and V.T. Shemyatenkova, Bulletin of the Academy of Science USSR, Division of Chemical Science, 1961, 178. [Chemical Abstracts, 1961, 55, 16285b] 142. J. Franc and K. Placek, Collection of Czechoslovak Chemical Communications, 1973, 38, 513. 143. J. Franc, Chemical Abstracts, 1975, 82, 67923q. 144. A.P. Kreshkov, V.T. Shemyatenkova, S.V. Syavtsillo and N.A. Palamarchuck, Journal of Analytical Chemistry USSR, 1960, 15, 727. [English Translation]

115

Introduction to Polymer Analysis 145. G.W. Heymun, R.L. Bujalski and H.B. Bradley, Journal of Gas Chromatography, 1964, 2, 300. 146. V.M. Krasikova and A.N. Kaganova, Journal of Analytical Chemistry USSR, 1970, 25, 1212. 147. E.R. Bissell and D.B. Fields, Journal of Chromatographic Science, 1972, 10, 164. 148. J. Franc and K. Placek, Mikrochimica Acta, 1975, 64, 1, 31. 149. C.L. Hanson and R.C. Smith, Analytical Chemistry, 1972, 44, 9, 1571.

116

4

Determination of Monomer Ratios in Copolymers

Quantitative determination of the weight percentage of the various bound monomer units of copolymers and terpolymers is a very important aspect of polymer analysis. Such data can have very important implications for various polymer characteristics: oxidative stability, thermal stability, mechanical properties and ﬂexibility, and elasticity and tensile properties. Various experimental techniques have come to the fore in copolymer composition studies. The two major techniques employed are infrared (IR) spectroscopy and nuclear magnetic resonance (NMR) spectroscopy. Controlled pyrolysis linked to gas chromatography (Py-GC) and IR spectroscopy is being employed in a growing number of applications. The applications of each of these techniques to typical copolymer composition studies are illustrated in the following sections by a series of selected examples.

4.1 IR Spectroscopy 4.1.1 Ethylene Propylene Copolymers Tosi and Simonazzi [1] described an IR method for the evaluation of the propylene content of ethylene-rich ethylene–propylene copolymers. This is based on the ratio between the absorbance of the 7.25 μm band and the product of the absorbances by the half width of the 6.85 μm band obtained on diecast polymer ﬁlm at 160 °C (Table 4.1). The calibration curve, based on a series of standard copolymers prepared with 14Clabelled ethylene or propylene, is obtained by plotting the 7.25 absorbance:6.85 absorbance ratio against the C3 weight fraction. The basis for the calibration of many methods for the analysis of ethylene–propylene copolymers is the work published by Natta and co-workers [6], which involves measuring the IR absorption of polymer solutions at 7.25 μm (presumably due to methyl vibrations related to the propylene concentration in the copolymer). In some cases, the dissolution of copolymers with

117

Introduction to Polymer Analysis low propylene content or some particular structures is difﬁcult [7, 8]. Natta’s solution method was calibrated against his radiochemical method [6], for which the precision of the method was not stated, and a considerable amount of scatter is evident in the data presented. Typical methods that have used Natta’s solution procedure [6] for calibration are described in publications by Wei [9] and Gössl [10]. These IR methods avoid the solution problems by employing intensity measurements made on pressed ﬁlms. The ratio of the absorption at 13.95 μm to that at 8.70 μm is related to the propylene content of the copolymer. Some objections [11] to the use of solid ﬁlms have been raised because of the effect of crystallinity on the absorption spectra in copolymers with low propylene content. These ﬁlm methods are reliable only over the range 30–50 mol% propylene.

Table 4.1 Comparison of calibrations based on the A7.25/A6.85 ratio C3 (wt%)

A7.25/A6.85 Reference [2]

Reference [3]

Reference [4]

Reference [5]

This Work

10

0.21

0.12

-

0.15

0.264

20

0.35

0.24

-

0.28

0.395

30

0.47

0.38

-

0.41

0.521

40

0.60

0.52

0.44

0.53

0.647

50

0.70

0.68

0.53

0.65

-

Source: Author’s own ﬁles

Lomonte and Tirpak [12] developed a method for the determination of the percentage of ethylene incorporated in ethylene–propylene block copolymers. Standardisation is done from mixtures of the homopolymers. Standards and samples are scanned at 180 °C in a spring-loaded demountable cell. The standardisation is conﬁrmed by the analysis of copolymers of known ethylene content prepared with 14C-labelled ethylene. By comparison of the IR results from the analyses at 180 °C and also at room temperature, ethylene homopolymer can be detected. These workers derived an equation for the quantitative estimation of the percentage of ethylene present as copolymer blocks. The method distinguishes between true copolymers and physical mixtures of copolymers. The method makes use of a characteristic IR rocking vibration due to 118

Determination of Monomer Ratios in Copolymers sequences of consecutive methylene groups. Such sequences are found in polyethylene (PE) and in the segments of ethylene blocks in ethylene–propylene copolymers. This makes it possible to detect them at 13.70 μm and 13.89 μm. There are bands at both these locations in the IR spectrum of the crystalline phase, but only at 13.89 μm in the amorphous phase. The ratio of these two bands in the IR spectrum of a polymer ﬁlm at room temperature is an approximated measure of crystallinity. As seen by this ratio, the IR spectra of the copolymers show varying degrees of PE-type crystallinity, dependent on the ethylene concentration and method of incorporation. It is this varying degree of crystallinity that allows qualitative detection of ethylene homopolymer in these materials. A calibration curve of absorbance at 13.89 μm versus ethylene is made from known mixtures for hot and cold runs. Both plots result in straight lines from which the following equations are calculated: % Ethylene at 180 °C = A/(0.55b) % Ethylene at room temperature = A/(3.0b) where A is the absorbance measured at 13.89 μm and b the thickness of the specimen in centimetres. A series of ethylene–propylene block copolymers prepared with 14C-labelled ethylene was analysed for percentage of ethylene incorporation by radiochemical methods. These samples when scanned at 180 °C gave IR results that agreed reasonably well with the radiochemical assay. When cooled samples were scanned, the results from the cold calibration were low in comparison with the known ethylene content. These data are shown in Table 4.2.

Table 4.2 14C-labelled ethylene-propylene copolymers ethylene (%) Sample

Radiochemistry

Hot infrared scan

Cold infrared scan

3401

2.4

2.9

0.9

3402

4.0

3.65

1.3

3403A

22.4

20.7

14.7

3403B

24.5

22.2

15.3

3404

12.4

13.0

7.1

3405

14.0

14.1

7.5

Reproduced with permission from J.N. Lomonte and J.A. Tirpak, Polymer Science, 1964, A2, 705. © 1964, Wiley.

119

Introduction to Polymer Analysis A pair of samples was prepared in which the active sites on the growing propylene polymer were eliminated by hydrogen before addition of ethylene. Practically identical values for percentage of ethylene incorporation were calculated for the hot and cold scans (Table 4.3). Paxton and Randall [13] used Fourier transform infrared spectroscopy (FT-IR) to measure the concentration of bound ethylene in ethylene propylene copolymers in amounts down to 0.1%. These polymers contained >95% propylene, with the ethylene units present as isolated entitles between two head-to-tail propylene units. These workers point out that most IR bands used for determining copolymer compositions are sensitive to sequences of both monomers. This IR method for compositional analysis can be calibrated if: (a) known standards of similar constitution to the copolymers being analysed are available and (b) assignments and behaviour of the calibration bands are well established; preferably the absorptivities of these bands should be relatively independent of the position of monomer units in the chain. Thus, quantitative IR analysis of copolymers depends primarily on the standards employed whose composition can be determined directly and reliably. Paxson and Randall [13] used 13C-NMR to provide such reference standards for the less time-consuming IR measurements because it is relatively inexpensive and easy to operate for copolymer analysis. They showed that an excellent correlation is obtained between 13C-NMR and IR results on a series of ethylene–propylene copolymers containing >95% wt% propylene.

Table 4.3 Samples with hydrogen-reduced active sites Sample no.

Ethylene (%) Hot infrared scan

Cold infrared scan

1487

5.0

5.3

1553

7.1

6.9

Reproduced with permission from J.N. Lomonte and J.A. Tirpak, Polymer Science, 1964, A2, 705. © 1964, Wiley.

Determination of IR spectra of moulded ﬁlms of the polymer at 13.66 μm enables the propylene unit content of the polymer to be determined. The absorbance of the 13.66 μm infrared band is attributed to gamma (CH2)3 and is characteristic of an ethylene unit isolated between two head to tail propylene units. The method is calibrated against

120

Determination of Monomer Ratios in Copolymers known copolymers of similar constitution to the copolymers being analysed:

i.e.,

The equation relating the IR absorbance to the ethylene content is: Y = 2.465 X + 0.451 where Y is the IR absorbance at 13.66 μm divided by the ﬁlm thickness in centimetres and X is the wt% ethylene. The standard errors are 0.110 and the intercept 0.051 for the slope. IR absorbance at 13.66 μm is sufﬁciently sensitive to the ethylene incorporation to determine the wt% ethylene within 0.1–0.2% at the 95% conﬁdence level. From 13 C-NMR data, it can be concluded that the propylene units occur in predominantly isotactic, heat-to-tail sequences, and that the ethylene units are incorporated as isolated units only. Thus, this structural prerequisite is needed for application of this method because it has not been tested on copolymers containing propylene conﬁgurational irregularities or ethylene sequences two units and longer. Ciampelli and co-workers [14] developed two methods based on IR spectroscopy of carbon tetrachloride solutions of polymers at 7.25, 8.65 and 2.32 μm for the analysis of ethylene–propylene copolymers containing >30% propylene. One method can be applied to copolymers soluble in solvents for IR analysis, the other can be applied to solvent-insoluble polymer ﬁlms. The absorption band at 7.25 μm due to methyl groups is used in the former case, whereas the ratio of the band at 8.6 μm to the band at 2.32 μm is used in the latter. IR spectra of polymers containing 55.5 and 85.5% ethylene are shown in Figure 4.1. Johnson-Plaumann and co-workers [15] used IR and NMR spectroscopy to quantify the composition of various ethylene–propylene copolymers. Absolute concentrations were measured by IR spectroscopy. An IR calibration curve was obtained by plotting the absorbance ratio 6.82 μm/7.26 μm versus concentration of monomer unit. Partnov and co-workers [16] measured the ratio of monomer units in ethylene–propylene copolymers by IR measurements at 14.49, 13.85, 13.66, 13.29 and 11.75 μm.

121

Introduction to Polymer Analysis

Figure 4.1 Infrared spectrum of ethylene-propylene copolymer showing bands at 7.25, 8.6 and 2.32 μm. Reproduced with permission from F. Ciampelli, G. Bucci, A. Simonazzi and A. Santambreglio, La Chimica e l’Industria, 1962, 44, 489. © 1962, Promedia Publishing e l’Industria

4.1.2 Ethylene–Vinyl Acetate There are two methods for this determination, depending on the concentration of vinyl acetate. At levels <10%, a band at 2.89 μm is used. This band is not suitable for higher concentrations because the necessary ﬁlm thickness is <0.1 mm. For higher concentrations, the carbonyl overtone band at 16.39 μm can be used because much thicker ﬁlms are needed to give suitable absorbance levels. Here the carbonyl overtone band was used for a series of standards with vinyl acetate concentrations up to 35%, the nominal thickness being about 0.5 mm. A combination of CDS and QUANT software on a Perkin Elmer model 683 infrared spectrometer was used to establish the calibration for this analysis. Once the calibration has been carried out, the simplest way to measure an unknown sample would be with an OBEY routine in the CDS II software. This would incorporate calibration data so that the single routine would measure the spectrum and calculate the vinyl acetate concentration with an error of approximately 5%.

122

Determination of Monomer Ratios in Copolymers Pallacini and co-workers [17] and Jones and McClelland [18] also used IR spectroscopy to analyse ethylene–vinyl acetate copolymers.

4.1.3 Styrene-based Copolymers Kandil and El-Gamal [19] used IR spectroscopy at 5.78 μm to study the composition of methylacrylate–styrene copolymers. The carbonyl bond intensity at 5.78 μm was correlated with copolymer composition styrene acrylic acid.

4.1.3.1 Styrene Acrylic Acid Urban and co-workers [20] and Kondil and El Gamal [19] have shown that IR spectroscopy can be used to yield information on monomer composition and triad sequence distribution in these polymers.

4.1.3.2 Styrene–acrylate and styrene methylacrylate copolymers Anderson and co-workers [21] described an IR method for the determination of down to 1% of bound styrene units in styrenated acrylate resins. Styrene units were determined by IR spectroscopy at 14.28 μm, which is the phenyl ring out of mode. This frequency provides speciﬁcity, freedom from interferences and an absorption that is directly proportional to styrene content. The accuracy of this method on typical acid-modiﬁed styrenated acrylic polymers is shown in Table 4.4. In the 5% to 70% styrene range, determined values are better than 95% relative.

4.1.4 Vinyl Chloride–Vinyl Acetate Copolymers IR spectroscopy has been applied to the determination of free and combined vinyl acetate in vinyl chloride–vinyl acetate copolymers [23]. This method is based upon the quantitative measurement of the intensity of absorption bands in the near-IR spectral region arising from vinyl acetate. A band at 1.63 μm due to vinyl groups enables the free vinyl acetate content of the sample to be determined. A band at 2.15 μm is characteristic for the acetate group, and arises from free and combined vinyl acetate. Thus, the free vinyl acetate content may be determined by difference at 2.15 μm. Polymerised vinyl chloride does not inﬂuence either measurement.

123

Introduction to Polymer Analysis

Table 4.4 Infrared analysis for bound styrene in styrenated acrylic polymers Sample

Styrene Calculated (%)

Observed (%)

1

5.0

5.0, 4.9

2

7.5

7.1, 6.9

3

10.0

19.6, 19.5

4

24.3

24.4, 24.3

5

24.3

25.6, 24.6

6

25.0

25.1, 25.8, 24.2a

7

35.0

36.3, 35.6

8

35.0

34.7, 35.2

9

40.0

38.9, 40.1

10

60.7

59.1, 60.3, 59.7a

11

70.0

69.3, 70.3, 69.5a

a

Percentage styrene values, determined in tetrahydrofuran. All remaining values were determined using acetone as the solvent. Reproduced with permission from D.G. Anderson, K.E. Isakson, D.L. Snow, D.J. Tessari and J.T. Vandebery, Analytical Chemistry, 1971, 43, 7, 894. © 1971, ACS

Van Doremaele and co-workers [22] also studied the composition of styrene–acrylate copolymers.

4.2 NMR Spectroscopy NMR spectroscopy seems to be overtaking IR spectroscopy as a means of ascertaining the ratio of monomer units in copolymers. Thus, most references to IR spectroscopy quoted in Section 4.4.1 are dated pre-1980, whereas most of the references to NMR spectroscopy are dated after 1985.

4.2.1 Ethylene Propylene Copolymers Paxson and Randall [9] described an NMR method for determining band ethylene units in ethylene–propylene copolymers containing >95% propylene. In this method, a 1,2,4 trichlorobenzene–perdeuterobenzene solution of the polymer is examined on a 13 C-NMR spectrometer to evaluate methane resonances. The method is calibrated against known copolymers of similar constitution to the copolymers being analysed. 124

Determination of Monomer Ratios in Copolymers The best results are obtained using methane resonances 4 and 5 (Table 4.5) to determine copolymer composition. This is because the methine carbon resonance is the least sensitive towards conﬁgurational differences and also the least affected by overlap from neighbouring resonances. Similar results are obtained peak heights or peak areas are used. From the NMR data it can be concluded that the propylene units occur in predominantly isotactic head-to-tail sequences, and that the ethylene units are incorporated as isolated units only. The NMR method can be used to provide reference standards for the less time-consuming IR method. Provided the IR method is calibrated in this way, excellent agreement is obtained between the IR and NMR methods for copolymers containing >95% propylene. Cheng and Kakugo [24] and Zhang and co-workers [25] discussed the application of 13C-NMR to the measurement of monomer ratios in ethylene–propylene copolymers. Figures 4.2 to 4.6 show NMR spectra of ethylene copolymers with propylene, butene-1, hexane-1, octane-1 and 4-methyl pentene-1. These spectra show chemical shift assignment of T values of the resonances of the copolymers. The molar composition of these copolymers could be determined with a relative precision at about 6% and μm value measurements at 10 ppm and 50 ppm.

30.000 Isolated CH2 30.384

br CH2 CH2 CH2 CH CH2 CH2 CH2

27.455

37.557

1 CH3

37.933 45.795

38.401

1 20.008

27.663 28.921

32.296 30.951

30.762

br 33.259

20.721 24.895 22.855 17.683 14.060 H

50

A

40 B

C

30 C+D+E

F

20 G

10

0 PPM

Figure 4.2 13C-NMR spectrum of ethylene-propylene copolymer showing chemical shift assignments. Reproduced with permission from H.N. Cheng and M. Kakugo, Macromolecules, 1991, 24, 1724. © 1991, ACS

125

Introduction to Polymer Analysis

Table 4.5 Observed and reference 13C-NMR chemical shifts in ppm for ethylene–propylene copolymers and reference polypropylenes as measured with respect to an internal trimethylsilane standard Resonance line

Carbon

E/P 3/97

E/P

1

AA-CH2

46.4

2

AA-CH2

3

E/P 97/3

Sequence assignment

Reference crystalline polypropylene

Amorphous polypropylene

46.3

PPPP

46.5

47.0–47.5 r 46.5 m

46.0

45.8

PPE

AG-CH2

37.8

37.8

PPEP

4

CH

30.9

30.7

PPE

5

CH

28.8

28.7

PPP

28.5

28.2 mmm 28.6 mmmr 28.5 rmmr 28.4 mr+rr

6

BB-CH2

24.5

24.4

PPEPP

7

CH3

21.8

21.6

PPPPP

21.8

21.3–21.8 mm 20.6–21.0 mr 19.9–20.3 rr

8

CH3

21.6

21.4

PPPE

9

CH3 CH3 CH A-CH2 B-CH2 (CH2)n

20.9

20.7 19.8 33.1 EPE 29.8

19.8 33.1 37.4 27.3 29.8

PPPEP EPE EPE EPE EEE

Reproduced with permission from J.R. Paxson and J.C. Randall, Analytical Chemistry, 1978, 50, 13, 1777. © 1978, ACS

126

Determination of Monomer Ratios in Copolymers 30.000 Isolated CH2

br CH2 CH2 CH2 CH CH2 CH2 CH2 2 CH2

30.500

30.948

br 39.747 34.627

1 CH 3

27.352

34.126

2 26.789

1 11.209

32.180

29.136

37.418

24.687

39.197

14.057

22.849 50

40 A

B

30

C

20

D+E

10X 10

0 PPM

F

Figure 4.3 13C-NMR spectrum of ethylene-o-butene-1 copolymer showing chemical shift assignments. Reproduced with permission from H.N. Cheng and M. Kakugo, Macromolecules, 1991, 24, 1724. © 1991, ACS

30.000 Isolated CH 2

br CH 2 CH 2CH 2 CH CH 2 CH 2CH 2

4

CH 2

3

CH 2

2

CH 2

1

CH 3

30.495

34.602

27.329

4 35.075 34.202 br 38.190

2 23.380

50

40

30 A B C D+E

1 14.114

27.696

35.934 60

37.569

40.272

3 29.370

32.298

24.614

20 F+G

22.860

10

0 PPM

G

Figure 4.4 13C-NMR spectrum of ethylene-hexene-1 copolymer showing chemical shift assignments. Reproduced with permission from H.N. Cheng and M. Kakugo, Macromolecules, 1991, 24, 1724. © 1991, ACS

127

Introduction to Polymer Analysis 30.000 Isolated CH2 4 6, 34.612 30.497

br CH2 CH2 CH2 CH CH2 CH2 CH2 6 CH2 5, 5 CH2 27.331 4 CH2 3 CH2 2 CH2 1 CH3

3 32.210 35.092

2 22.885

24.638

27.920

1 14.069

35.994

br 38.242

D 60

30 40 D+E F+G+H AB C H

50

I P

10

20

0 PPM

Figure 4.5 13C-NMR spectrum of ethylene-octene-1 copolymer showing chemical shift assignments. Reproduced with permission from H.N. Cheng and M. Kakugo, Macromolecules, 1991, 24, 1724. © 1991, ACS

30.000 Isolated CH2 br CH2 CH2 CH2 CH CH2 CH2 CH2

30.507

45.442

50 A

40 B C

D

CH 2 CH3 1

H

E 60

1 CH3

24.259

32.187

br 36.064

35.392 33.837

3 44.856

1 23.330 27.352

2 26.086

34.918

CH2 3

30 F+G

20

10

0 PPM

G

Figure 4.6 13C-NMR spectrum of ethylene-methyl-pentene-1 copolymer showing chemical shift assignments. Reproduced with permission from H.N. Cheng and M. Kakugo, Macromolecules, 1991, 24, 1724. © 1991, ACS

128

Determination of Monomer Ratios in Copolymers

4.2.2 Ethylene–hexane-1 Kissin and Brandolini [26] reported on the 13C-NMR spectra of ethylene-hexane-1 copolymers and assigned chemical shifts for these copolymers. This enabled them to determine the ratio of the two comonomers. They consist of the 29.4 ppm and 29.3 ppm peaks to the hexane-1 B-CH2 carbon atoms. Cheng [27] also used this technique to analyse ethylene-hexane-1 copolymers.

4.2.3 Styrene-Based Copolymers 4.2.3.1 Styrene Methacrylate Evans and co-workers [28] described techniques employing Py-GC, proton NMR and carbon analysis for the determination of styrene and methacrylate units in styrene–methylmethacrylate and styrene–n–butyl methacrylate copolymers with an accuracy of ±2%. Agreement between the three independent methods is excellent. The comparison stresses the complementary nature of all three methods. The Py-GC method possesses advantages such as simplicity and rapidity. In the NMR method, a deuterated chloroform solution of the polymer containing trimethylsilane internal standard is run on the proton NMR spectrometer. Integration of peak areas at delta 8 and 1–4 delta, respectively, gives a measure of signal strengths from aliphatic protons (methacrylate units) and aromatic protons (styrene). These areas can be used as a basis for calculating the methacrylate and styrene unit contents of the polymer. The analysis of comonomer composition does not appear to be signiﬁcantly affected by the extent of randomness observed for these materials. The difference between NMR and Py-GC results are in the range of 0–4% and 0–4.8% for styrene–n–butyl methacrylate and styrene–methyl methacrylate, respectively. The difference in carbon analysis and Py-GC results is also in the same range for both copolymers. Standard deviation for Py-GC ranges from 1.2% to 2.1%. Precision for NMR analyses is >1%. Pyrolysis of methyl methacrylate–ethylene dimethacrylate copolymer gives only one major peak (methyl methacrylate) using a hot ﬁlament detector. The composition of methyl methacrylate–ethylene dimethacrylate copolymer can be determined by pyrolysing a weighed sample and using the ratio of sample weight to area of the methyl methacrylate peak for obtaining a standard analysis curve. Under favourable conditions and with careful control of the pyrolysis column and detector, variable constituents can be determined within ±0.5%.

129

Introduction to Polymer Analysis Differences in the pyrograms of block and random copolymers allow estimation of comonomer distribution. Random copolymers of ethylene with methyl acrylate or methyl methacrylate yield on pyrolysis a lower ratio of methanol/methyl acrylate or methanol/methyl methacrylate, respectively, than block polymers of the same composition. Differential thermal analysis measurements give a ﬁrst-order transition for block polymers only, and by measuring the area under the transition, an indication of the minimum chain length between acrylate units can be obtained.

4.2.3.2 Styrene methyl acrylate Van Doremaele and co-workers [22] applied 1H-NMR spectroscopy to the determination of monomer ratios in styrene methyl acrylate copolymers; 400-MHz 1 H-NMR spectra were obtained in CDCl3 solutions at 25 °C. Expansions of the methoxy region display additional ﬁne splitting due to combined conﬁgurational (i.e., tacticity) and compositional sequence effects. Mean copolymer composition (mole fraction styrene, FS) can be readily obtained by using absorbances, which represent the total peak areas of the aromatic and methoxy proton resonances, respectively. The initial feed (qo = [S]/[M]), the average copolymer composition, and the conversion are summarised in Table 4.6, [S] = concentrated styrene in feed, [M] = concentration of methyl acrylate in feed.

Table 4.6 Observed cumulative average copolymer composition (FS = mole fraction of styrene) and ﬁnal conversions of some low-conversion styrene– methylacrylate q0a Fs 0 0 0.038 0.12 0.11 0.33 0.36 0.46 0.85 0.57 1.98 0.67 2.97 0.77 4.78 0.81 a Initial monomer feed ratio q0 = [S]/[M]

Conversion (mol%) 20 7 5 12.6 14.3 16.1 10.2 9.7

Reproduced with permission from G.H.J. van Doremaele, A.L. German, N.K. de Vries and G.P.M. van der Velden, Macromolecules, 1990, 23, 19, 4206. © 1990, ACS

130

Determination of Monomer Ratios in Copolymers

4.2.3.3 Styrene–acrylic acid Shouting and Poehlein [29] showed that 1H-NMR is a valuable tool for the quantitative determination of monomer ratios in styrene–acrylic acid copolymers. In this work, results from 1H-NMR and 13C-NMR were compared. The composition results obtained by the two methods are compared in Table 4.7.

Table 4.7 Fraction of styrene in initial monomer feed (fs) and in bulk styreneacrylic acid copolymers (Fs) obtained by 1H- and 13C-NMR Sample code

1

Fs Fs-1

13

H-NMR

C-NMR

Fs-2

Fs-3

Fs-4

SA-5

0.929

0.8473

0.8651

SA-10

0.863

0.7686

0.7647

0.7832

SA-20

0.735

0.6865

0.6568

0.6742

SA-30

0.618

0.6598

0.6254

0.6378

SA-40

0.509

0.5312

0.5797

0.5715

0.6080

SA-50

0.410

0.5384

0.5483

0.5401

0.5780

SA-60

0.316

0.4629

0.5311

0.4973

0.5329

SA-80

0.148

0.3342

0.3331

0.3479

0.3515

Reproduced with permission from W. Shouting and G.W. Poehlein, Journal of Applied Polymer Science, 1993, 49, 6, 991. © 1993, Wiley

The measurement of NMR spectra of S-AA copolymers was complicated because of the difﬁculties in dissolving the samples. Copolymers with high acrylic acid content cannot be dissolved in CDCl3. If styrene content is high, the copolymer cannot be dissolved by DMSO-D6. Hence the NMR spectra were recorded using DMSO-D6 or CDCl3-DMSO-D6 solvent mixtures. 1H-NMR spectra were obtained using a Varian XL-400 spectrometer operating at 400 MHz at 50 °C. The recording conditions for 1H-NMR were: sample concentration: 1% (g/ml); spectral width: 4798.5 Hz; acquisition time: 1.67 s; pulse delay: 5 s; and number of scans: 16. 13C-NMR spectra were obtained with the same spectrometer at 100 MHz at 90 °C. The conditions for 13 C-NMR measurements were: sample concentration: 10% (g/ml); spectral width: 20,000 Hz; acquisition time: 0.4 s; ﬂip angle: 45º; pulse delay: 1.6 s; and number of scans: 512. DMSO was used as the locking agent for all NMR spectra.

131

Introduction to Polymer Analysis

4.2.4 Benenyl Acrylate–Vinyl Acetate [27] Subrahmanyam and co-workers [30] used 1H-NMR spectroscopy to determine the composition of benenyl acrylate–vinyl acetate copolymers. The reactivity ratios were evaluated by different methods, and found to be 0.021 for vinyl acetate and 1.76 for benenyl acrylate. The Q and e values for benenyl acrylate were calculated as 0.25 and 0.94, respectively. The experimental copolymer composition was found to be in close agreement with calculated values.

4.2.5 Vinyl Acetate–Methylacrylate Brar and Charan established the vinyl acetate: methyl acrylate monomer ratio in copolymers by 1H-NMR [31]. Peaks were shown at approximately 0.7, 1.2, 1.7, 3.3 and 4.8 ppm, for CH(V + M), CH3(V), CH(M), CH3 (M) and CH(V), respectively, where V is:

4.2.6 Hexaﬂuoropropylene–Vinylidene Fluoride Copolymer Two methods have been described for determining the compositional analysis of these copolymers: one based on high-resolution continuous and Fourier transform 19 F-NMR, and the other on Py-GC [32, 33].

4.2.6.1 19F-NMR When the composition of a single component in a mixture is required, it is necessary to relate the component resonance to that of another compound, an internal standard,

132

Determination of Monomer Ratios in Copolymers which is of known chemical composition and has been added in known weight to a known weight of unknown. Brame and Yeager [33] used dichloro-benzotriﬂuoride as an internal standard in the continuous wave method for determining the compositional analysis of both repeat units in hexaﬂuoro–propylene–vinylidene ﬂuoride copolymers. This work demonstrated the utility of the Fourier transform NMR method in quantitative analysis of the copolymer in relation to results obtained by continuous wave 19F-NMR and proton NMR. The lines observed are attributed to the following: CF3 group (D is –70 to –75), CF2 groups (D is –90 to –120) and CF group (D is –180 to –185). The value obtained is in excellent agreement with those obtained by mass balance.

4.2.6.2 Pyrolysis–Gas Chromatography Blackwell [32] used a Curie point pyrolyser to carry out quantitative analysis of monomer units in polyhexafluoropropylene–vinylidene fluoride. The polymer composition is calculated from the relative amounts of monomer regenerated and the triﬂuoromethane (CHF3) produced during pyrolysis. The exact mechanism by which triﬂuoromethane is produced during pyrolysis is not known, but it is presumed that the free triﬂuoromethyl group is cleaved from the polymer backbone. The triﬂuoromethyl group then extracts a proton from the polymer chain to form triﬂuoromethane. The monomer composition of the series of copolymers was calculated from pyrolysis data and compared with that calculated from NMR data. These data are summarised in Table 4.8. The difference in weight per cent calculated for the two techniques, averaged ±6.0% for hexaﬂuoropropylene and ±0.85% for vinylidene ﬂuoride.

4.2.7 Acrylamide–Methacryloyl Oxy-ammonium Chloride Vu and Cabestany [34] determined the composition of water-soluble polyacrylamides (acrylamide–methacryloyl ethyl trimethyl ammonium chloride and acrylamide– acryloloxyethyl trimethyl ammonium chloride) by 1H- and 13C-NMR of deuterium oxide solutions of these copolymers. Good agreement was obtained between these results and those obtained by Kjedahl nitrogen determination. This 1H-NMR method involves integration of the N(CH3)3, the CH2, the carbonyl and the CH signals in a deuterium oxide solution of either polymer. The results in Table 4.9 show reasonably good agreement of both NMR methods with theoretical values.

133

Introduction to Polymer Analysis

Table 4.8 Comparison of the monomer composition of HFP/VF2 copolymer as calculated by pyrolysis data and NMR data HFP (wt%) 19

F-NMR 28.6

VF2 (wt%) 19

Pyrolysis 28.8

29.0 28.8 34.79 36.04 36.46 36.73 37.24 37.33 39.04 39.07 39.04 38.35 39.82 38.46 40.10 39.45 42.4 43.2 44.8 46.2 44.8 45.4 47.3 47.0 49.95 49.45 SD: Standard deviation

F-NMR

SD for Curie point pyrolyser

Pyrolysis

0.2

71.4

70.9

0.5

1.04

0.2 1.25 0.27 0.11 0.03 0.69 1.36 0.65 0.8 0.2 1.4 0.3 0.5

71.0 65.21 63.96 62.76 60.96 60.96 60.18 59.9 57.6 55.2 55.2 52.7 50.05

71.0 62.63 62.91 62.61 60.88 61.55 61.89 61.49 56.7 53.8 54.3 52.8 50.55

0 2.58 1.05 0.15 0.08 0.59 1.71 0.57 0.9 1.4 0.9 0.1 0.50

0.34 0.52 0.82 0.39 0.67 0.63 0.30 0.10 0.39 0.48 0.59 0.92 0.30

Reproduced with permission from J.T. Blackwell, Analytical Chemistry, 1976, 48, 13, 1883. © 1976, ACS

Table 4.9 Comparative chemical compositions resulting from different analytical methods % CMA or CMM CMM 15

CMM 30

CMA 15

CMA 30

15

30

15

30

C1- /N

16.2

32

15.2

31.2

By 1H-NMR

18.5

33

20

31.7

15.5

31.5

13

27

15.6

32.9

-

-

Theoretical

By

13

C-NMR (CO)

13

By C-NMR (OCH2, CBH2)

CMM: methacryloyl oxyethyltrimethyl ammonium chloride CMA: acryloyl oxyethyltrimethyl ammonium chloride Reproduced with permission from C. Vu and J. Cabestany, Journal of Applied Polymer Science, 1991, 42, 11, 2857. © 1991, Wiley

134

Determination of Monomer Ratios in Copolymers

4.3 Pyrolysis–Gas Chromatography The consensus is that, certainly in the cases of those polymers such as the polyoleﬁns where complex pyrograms are produced, ﬁlament pyrolysis is the preferred method. For the purposes of fundamental studies, pyrolysis at various temperatures and heating rates is preferred. Smaller sample weights of occurrence of secondary side reactions in the pyrolysis which may confuse the interpretation of the pyrogram when carrying out polymer structural studies. Sample weights >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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Sequencing of Monomer Unit in Polymers

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

Sequencing of Monomer Unit in Polymers

¥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

Sequencing of Monomer Unit in Polymers

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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Source: Author’s own ﬁles

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

Introduction to Polymer Analysis

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

Sequencing of Monomer Unit in Polymers

(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

Introduction to Polymer Analysis

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

Sequencing of Monomer Unit in Polymers

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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Sequencing of Monomer Unit in Polymers

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

Introduction to Polymer Analysis

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

Sequencing of Monomer Unit in Polymers

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

Sequencing of Monomer Unit in Polymers

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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

(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%)

Source: Author’s own ﬁles

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 (%)

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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 (<1%) of 13C nuclei is another favourable 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 entire polymer molecule. Finally, 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 [3], which have been 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

215

Introduction to Polymer Analysis in molecular structure creates 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 PP conﬁgurational assignments was heptamethylheptadecane, in which the relative conﬁgurations were known [4]. To be completely accurate, 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 [5]. 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 used to obtain 13C-NMR spectra. Equilibrium conditions must be established during signal averaging to ensure that 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 between the 1H and 13C nuclear energy levels during spin decoupling can lead to enhancements of 13C resonances by factors between 1 and 3. Thus, the spectral relative intensities will only reﬂect the polymer’s moiety concentrations if the differentiated NOE are equal or else taken into consideration. Polymer NOE are generally maximal, and consequentially equal, because of a polymer’s restricted mobility [6, 7]. One should examine the polymers NOE through gated decoupling or paramagnetic quenching, thereby avoiding misinterpretation of spectral intensity data. The 13C conﬁgurational sensitivity falls within a range from triad to pentad for most vinyl polymers. In non-crystalline PP, three distinct regions corresponding to methylene (-46 ppm), methine (-28 ppm) and methyl (-20 ppm) carbons are observed in the 13 C-NMR spectrum. (The chemical shifts are reported with respect to an internal tetramethylsilane (TMS) standard). The 13C spectrum of a 1,2,4-trichlorbenzene solution at 125 °C of a typical amorphous PP is shown in Figure 6.2(a). 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, rrr, rrrm and mrrm, from low to high ﬁeld [3, 4, 8].

216

Stereoisomerism and Tacticity mm

21.82

a

rr

CH3 rm

20.30

21.03 CH

28.47

28.92

Recorder response

CH2

46.50

47.16

CH3 CH

CH2

b

HMDS

50

40

30

20

10

0

TMS (ppm)

Figure 6.2 (a) 13C-NMR spectrum of 1,2,4-trichlorobenzene solution of an amorphous polypropylene (b) crystalline polypropylene. Source: Author’s own ﬁles.

The 13C-NMR spectrum of crystalline PP (Figure 6.2(b)) contains only three lines, which can be identiﬁed as methylene, methine and methyl from low to high ﬁeld by off-resonance decoupling. An amorphous PP exhibits a 13C spectrum which

217

Introduction to Polymer Analysis contains not only these three lines, but additional resonances in each of the methyl, methane and methylene regions. The crystalline PP must therefore be characterised by a single type of conﬁgurational structure. In this case, the crystalline PP structure is predominantly isotactic, thus the three lines in Figure 6.2(b) 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 [9, 10] carried out a detailed study of the PP methyl group in triad and pentad conﬁgurational environments. Figure 6.3 shows the proton magnetic resonance (PMR) spectrum of hot n-heptane solution of atactic PP containing up to 40% syndiotactic placement, and which by Natta’s deﬁnition may be stereoblock [11]. The spectrum is inherently complex, as a ﬁrst-order theoretical calculation, and this indicated the possibility of at least 15 peaks with considerable overlap between peaks because differences in chemical shifts are about the same magnitude as the splitting due to spin-spin coupling. The largest peak, at high ﬁeld represents pendant methyls in propylene units. It is characteristically split by the tertiary hydrogen. By area integration, about 20% of the nominal methyl proton peak is due to overlap of absorption from chain methylenes. This overlap is consistent with a reported syndiotactic triplet [6]. The absence of a strong singlet peak in the methylene range indicates the virtual absence of ‘amorphous’ polymer in the atactic PP, which could possibly be due to the head-to-head and tailto-tail units. The low ﬁeld peak represents the partial resolution of tertiary protons which are opposite the methyls on the hydrocarbon chain. Various workers have developed analyses for physical mixtures and block copolymers based on the ratio of the incremental methylene area to the total polymer proton absorption. This concept has been tested by Barrall and co-workers [12, 13] using PMR analyses on a series of physical mixtures and block copolymers synthesised with 14 C-labelled propylene and others with 14C-labelled ethylene. A most important feature of this analysis is that the methylene peaks A and B have virtually the same relative heights in PP with a variety of tacticities (Figure 6.3(a)). This is also true for PMR spectra given by Satoh and others for a tactic series of PP [14, 15]. This suggests that PMR analyses for ethylene are independent of tacticity because the area increment of peak A above peak B has been used for analysis. Quantitative PP tacticities can be estimated by PMR not only for homopolymers (Figure 6.3(a)) but also in the presence of polyethylene (PE) and ethylene copolymer blocks. The relative heights of the peaks for secondary and for tertiary hydrogen in Figure 6.3(b) indicate that the PP in the copolymer and the physical mixture is dominantly isotactic.

218

Stereoisomerism and Tacticity

a

“Atactic” Recorder response

“Isotatic”

1

1

0

0

Arbitrary (ppm)

Recorder response

b

Sensitivity change

7

Homopolymer mixture

Block copolymer

Solvent 20

6

A

B A B

1

0

1

0

Arbitrary (ppm)

Figure 6.3 13C-NMR spectrum of (a) crystalline polypropylene (predominantly isotactic) and (b) ethylene – propylene copolymers. Source: Author’s own ﬁles.

219

Introduction to Polymer Analysis Barrall and co-workers [13] attempted to determine the syndiotactic content of some experimental PP. NMR spectra or samples dissolved in o-dichlorobenzene were obtained at 170 °C at 100. Spectra of syndiotactic and isotactic samples gave the NMR parameters in Table 6.1.

Table 6.1 Analytical Data for Physical Mixtures and Ethylene-Propylene Block Copolymers Sample

Ethylene (wt%)

Peak Height Ratio

Type of Sample

Freezing Curve Exotherms (oC)

3-14

0.0

0.

Hompolymer propylene

103

3-2

3.0 ± 0.5

0.172

Multisegment block

102

3-3

7±1

0.206

Multisegment block

102

3-6

18 ± 1

0.325

Multisegment block

140, 111

3-13

49 ± 2

0.675

Multisegment block

140, 110

3-30

6.0 ± 0.2

0.125

3-31

9.1 ± 0.2

0.170

3-32

14.9 ± 0.3

0.376

3-40

18.1 ± 0.5

3-41

18.6 ± 0.4

3-33

Multisegment block with randomness Multisegment block with randomness

100 100

Multisegment block with randomness

139, 110

0.315

Bisegment block

141, 111

0.328

Molecular mixtureb

100

23.3 ± 0.1

0.410

Multisegment block with randomness

140, 110

3-35

25.3 ± 0.2

0.390

Bisegment block

140, 110

3-34

36.0 ± 0.4

0.570

Multisegment block with randomness

140, 111

1

9.35

0.217

Physical mixture of polymers

110

2

5.0

0.147

110

3

26.0

0.406

110

4

49.0

0.672

110

5

17.0

0.308

110

6

5.0

0.150

110

7

9.0

0.238

110

8

34.0

0.512

9

100.0

0.0

110 Homopolmer ethylene

a

Radiometric analysis. Physical mixtures were made up by weight from 100 mesh polymer ﬂour

b

Made by polymerising ethylene in the presence of ﬁnely divided polypropylene.

113

The reference for the chemical shift measurement was hexamethyl disiloxane. The methylene hydrogens in the isotactic material are non-equivalent – as expected on geometrical grounds. Spectra calculated with the above parameters agreed reasonably well with the observed spectra. High-resolution PMR spectroscopy has been applied to an examination of very highly isotactic, very highly syndiotactic, and stereoblock PP in o-dichlorobenzene solutions. Barrall and co-workers [13] discuss 220

Stereoisomerism and Tacticity the effects of stereoregulation on proton shielding and some of the complexities of the methylene proton resonances and determine tactic placement contents for several polymers by a method based on the methylene proton resonances. Tactic pair contents were determined for two stereoblock fractions by a method based on the methyl proton resonances. Their results revealed much higher stereoblock characters than those determined for the same fractions from melting data. All PMR results were in very good accord with the results obtained on several polymers by infrared (IR), X-ray diffraction, and differential thermal analysis. Stehling and Knox [5, 16] deﬁned the stereochemical structure of PP from the PMR spectrum of normal deuterated and epimerised polypropylene. To determine the isotactic content of PP, Peraldo [17] carried out a normal vibrational IR analysis. He considered the primary unit as an isolated three-fold helix. From this work and several subsequent publications [18–20], it was suggested that the absorptions at 8.57, 10.02 and 11.90 μm were indicative of the helical conformation of the isotactic form. Measurement of the isotactic contents of a series of PP fractions based on these bands were made [20] and compared with results from Flory’s melting point theory. Melting points were determined as the point of disappearance of the birefringence on highly annealed samples, all three bands give qualitative agreement with the melting point data, only the method based on the 8.57 μm band gives quantitative agreement. Therefore, the method based on this band appears to gives a good measure of the isotactic content in PP, at least in the 60–100% range. Burﬁeld and Lo [21] pointed out that although various methods have been developed over the years to measure the stereoregularity of PP [22–24] including: solvent extraction, IR spectrometry, X-ray diffraction, calorimetry and density measurements, all these techniques provide only an indirect assay of stereoregularity because the parameters are measures relating to the crystallinity or conformational arrangement of the PP macromolecule and not tacticity per se. A major breakthrough in PP characterisation was made possible with the application of high-resolution 1H and 13C-NMR, which allows absolute determination of tacticity. These methods, which have been comprehensively reviewed [8, 9, 25], are of paramount importance because they permit not only quantiﬁcation of tacticity, but also elucidation of stereo-sequences. Although NMR characterisation provides the most fundamental information about stereostructure, secondary methods based on IR spectroscopy or solvent extraction are very widely employed. The continuing use of these secondary methods is a direct consequence of simplicity, speed, and low equipment cost which are particularly important for routine or screening analysis.

221

Introduction to Polymer Analysis Measurement by IR spectroscopy is potentially promising and semi-quantitative evaluations are possible. IR methods suffer from two main drawbacks: absence of suitable calibration, and thermal history effects. The second limitation is the change induced in the infrared spectrum by thermal history effects. Thus, the IR bonds used to measure isotacticity are bonds sensitive to the formation of regular isotactic helices rather than isotacticity. As such, IR absorbance will be dependent on pretreatment and annealing, which determine the conformation and morphology of polymer chains. IR spectroscopy has been widely used in the elucidation of PP stereostructure since Natta and co-workers [26–29] ﬁrst reported the spectrum of the crystalline polymer. Most studies have concentrated on identifying suitable bands to measure tacticity and to correlate with other indices of stereoregularity. The bands apparently associated with isotactic helices absorb at 8.19, 8.56, 10.02, 11.11, 11.89 and 12.36 μm, and of these the bands of 11.89 μm and 10.02 μm have been principally used for the derivation of calibration curves (Figure 6.4). Because it is difﬁcult to prepare ﬁlms of standard thickness, it is customary to use an internal reference bond, and the absorptions at 6.85, 8.56 and 10.28 μm are used for this purpose. The origin of the reference band at 10.28 μm, which is observable in the melt of isotactic samples [19, 30] as well as in purely atactic material, is disputed [30]. Thus the band, which has been attributed [30, 31] speciﬁcally to the PP head-to-tail sequence of repeating units, has also been associated with short isotactic helices apparently still present in the melt or atactic material [25]. The earliest study seeking to provide an IR calibration for isotacticity determination appears to be the work of Luongo [19], who indexed the ratio A10.26/A10.05 against physical mixtures of supposedly atactic and isotactic polymers. This calibration was later modiﬁed by an annealing procedure, initially in air [32] and subsequently under argon [33], the purpose of which was to ensure that all of the isotactic segments were converted to IR-active helices. Hughes [33] suggested that for annealed highly stereoregular PP, A10.25/A10.2 was very close to the isotactic fraction (I). Subsequently, the most extensive studies relating to IR characterisation was carried out by Kissin and co-workers, and much of this work has been summarised [24, 34, 35]. Essentially, three indices for measurement of isotacticity were proposed. These included the ratio A10.02/A10.28 earlier proposed by Luongo [19], which was now designated as ‘macrotacticity’ (M). Two further ratios: A10.28/A14.63 (known as the spectral degree of isotacticity (A)) and A11.89/A10.28 were also proposed. The signiﬁcance of utilising three distinct indices lies in the observation that the appearance of these helix bands is sensitive to the length of the isotactic sequences [24]. Thus, the critical sequence lengths for the appearance of these bands are: 10.01 μm (11–12 units); 10.28 μm (5 units); and 11.89 μm (13–15 units). The advantage claimed for using the 10.28 μm band is its sensitivity to even low levels of isotactic units and its apparent insensitivity to thermal history effects. The use of this band is complicated

222

Stereoisomerism and Tacticity by the uncertainty of the assignment and the reduced sensitivity for highly isotactic polymers. Quantitative measurements proposed by Kissin and co-workers [34] are based on calibration with supposedly completely isotactic samples, and are based on the assumption that there is a linear relationship between the absorption ratios and isotacticity [36]. Thus, for example, the degree of isotacticity A is given by the relationship [34]: A = A10.28/A6.85: N N is a normalising factor obtained by calibration with a highly isotactic sample and is sensitive to instrument type. Values of N = 0.265 and 0.300 were observed for two dispersive instruments. 100

Transmittance (%)

80

60

40 841 20

998 1167 973

0 1600

1400

1200 Wavenumber (cm-1)

1000

800

Figure 6.4 Infrared spectrum of polypropylene showing bonds at 11.89 cm–1 and 9.98 cm–1 associated with isotactic helices. Source: Author’s own ﬁles It is clear from the above discussion that none of the numerical values derived from the above methods can be easily equated to numerical tacticity values determined by more fundamental NMR measurements. Furthermore, there are considerable uncertainties with respect to: sample pretreatment, instrument effects, and calibration standards.

223

Introduction to Polymer Analysis In light of these deﬁciencies, Burﬁeld and Lo [21] attempted to establish an empirical method based on an absolute NMR calibration, with standardised sample preparation methods and which was applicable to a wide range of instrumentation. Four possible calibration curves of IR absorption ratio versus NMR tacticity were investigated for PP ﬁlms hot pressed at 120–200 °C: absorption at 10.02 μm referenced to bands 10.28 μm or 8.57 μm, and the absorption at 11.89 μm with the same reference bands i.e., ratios A10.02/A10.28, A10.02/A8.57, A11.89/A10.28 and A11.89/A8.57, respectively. Burﬁeld and Lo [21] use the 10.28 μm band as reference to provide linear calibration curves for both ratios. Calibrations based on the 8.57 μm reference are curves that provide rather low sensitivity in the important high isotacticity (mm >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

Introduction to Polymer Analysis

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

Stereoisomerism and Tacticity

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

Introduction to Polymer Analysis

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

Stereoisomerism and Tacticity

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

239

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Stereoisomerism and Tacticity

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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

*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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Determination of End groups

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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Determination of End groups

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

Introduction to Polymer Analysis

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

Determination of End groups

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

Introduction to Polymer Analysis

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

Determination of End groups

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

Introduction to Polymer Analysis

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

Determination of End groups

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 <3 × 105 even by the 500 MHz 1H-NMR used in this work. To estimate the higher Mn values by NMR, a much higher number of scans should be needed to enhance the S/N ratio. Because the T1 values for the end groups are generally larger than those for the main chains such as O — CH3, a much longer pulse delay time may be required to fulﬁl the sufﬁciently quantitative conditions for the measurement. Figure 8.5 shows a typical 1H-NMR spectrum for a PMMA sample (S–1; Mn = 20,220). On the expended partial spectrum (× 100) around 7.1–7.4 ppm, three phenyl-proton peaks reﬂecting the cumyl chain ends are observed. The intensity ratio of these three peaks is about 2:2:1, which corresponds to the proton number of o–, m– and p– positions in the phenyl ring, respectively. The methoxy proton peak at about 3.6 ppm can be used as the key peak reﬂecting the backbone of the polymer chain. Using these peak intensities, Mn, of a given PMMA sample can be estimated.

296

Determination of End groups

Table 8.5 Assignment of peaks concerned with the end groups observed in the pyrograms along with that of MMA monomer Peak

Molecular weight

Chemical structure

Effective carbon number

MNA monomer

100

3.65

A

120

9.0

B

118

8.9

C

134

10.0

D

132

9.9

E

174

12.9

F

206

11.75

G

218

12.65

Reproduced with permission from H. Ohtani, Y. Takehana and S. Tsuge, Macromolecules, 1997, 30, 9, 2542. © 1997, ACS

297

Introduction to Polymer Analysis

H2O (impurity) CH2

OCH3

× 100 - CH3

7

6

5

3

4

2

1 ppm

Figure 8.5 1H-NMR spectrum of PMMA sample (Mn = 20,200) showing 3 phenyl - proton peaks (inset) reﬂecting the cumyl chain ends around 7.1–7.4 ppm. Reproduced with permission from H. Ohtani, Y. Takehana and S. Tsuge, Macromolecules, 1997, 30, 2542. © 1997, ACS

Table 8.6 Number-average molecular weight of poly(methylmethacrylate)s calculated from end-group analysis by Py-GC and 1H-NMR Sample

Number-average molecular weight (Mn) Reference valuea

Py-GC

1

S-I

20,200

21,400

27 000

S-II

33,900

35,500

35 700

S-III

66,200

70,200

67 200

S-IV

91,400

103,000 b

S-V

296,000

317,000

S-VI

797,000

731,000c

S-VII

1,330,000

a

H-NMR

98 700 (213 000)

1,160,000 b

Reference Mn speciﬁed by the manufacturer. Coefﬁcient of variance (CV) = 1.4% for ﬁve measurements. c CV = 4.1% for ﬁve measurements.

Reproduced with permission from H. Ohtani, Y. Takehana and S. Tsuge, Macromolecules, 1997, 30, 9, 2542. © 1997, ACS

298

Determination of End groups This method can precisely quantify cumyl end groups in anionically polymerised PMMA even for samples with Mn = 106 (which were difﬁcult to analyse even by NMR). For these PMMA, it was not necessary to take into account contribution of the side reaction products. Mn values for PMMA samples can be directly estimated by the relative peak intensities between MMA formed from the polymer main chains through almost quantitative depolymerisation and the characteristic products of the end groups on the observed pyrograms after making molar sensitivity corrections for those products. Consequently, this technique is one of the absolute methods for determining Mn without using standard polymer samples with known Mn, provided that all the characteristic products of the end groups are quantitatively assigned against those from the main chain for given polymeric systems.

8.7.2 MALDI-ToF-MS MALDI-ToF (collision-induced dissociation) MS has been used to obtain information on the end groups and other structural features in polyalkylmethacrylates [61].

8.7.3 Dye Partition Methods This technique is best illustrated by an example concerning the determination of end groups in persulfate-initiated PMMA [42, 62, 63]. Sulfate and other anionic sulfoxy end groups were determined by shaking a chloroform solution of the polymer with aqueous methylene blue reagent. The greater the anionic content, the more the methylene blue phase is decolourised. The blue colour is evaluated spectrophotometrically at 660 nm. The quantity of anionic sulfoxy end group present in the polymer is obtained by comparing the experimental optical density values with a calibration curve of pure sodium lauryl sulfate. Results obtained by Ghosh and co-workers [42, 64] in end-group analyses of PMMA indicate that all the polymer samples exhibit a positive response to methylene blue reagent in the dye partition test, indicating at least some sulfate (OSO3–) end groups. Maiti and Saha [65] described a dye partition technique [66–68] utilising disulﬁne blue for the qualitative detection and, in some cases, the determination of amino end groups in the free radical polymerisation of PMMA. They found only 0.01–0.62 amino end groups per chain in PMMA made by amino-azo-bisbutyronitrile system, whereas in polymer made by the titanous chloride and acidic hydroxylamine systems they found 1.10–1.90 amino end groups per chain. Ghosh and co-workers [42, 69] also described a dye partition method for the determination of hydroxyl end groups in PMMA samples prepared in aqueous

299

Introduction to Polymer Analysis media with hydrogen peroxide as the photo-initiator. Dried polymers were treated with chlorosulfonic acid under suitable conditions whereby the hydroxyl end groups present in them were transformed to sulfate end groups. Spectrophotometric analyses of sulfate end groups in the treated polymers were carried out by application of the dye-partition technique, and thus a measure of hydroxyl end groups in the original polymers was obtained (average 1 hydroxy end group per polymer chain). Saha and co-workers [70] and Palit [66, 67] developed a dye-partition method for the determination of halogen atoms in copolymers of styrene, methylmethacrylate, methylacrylate, or vinyl acetate with a chlorine-bearing monomer such as allyl chloride and tetrachloroethylene. Copolymers were quaternised with pyridine, then precipitated with petroleum ether or alcohol, and further puriﬁed by repeated precipitation from their benzene solutions with a mixture of alcohol and petroleum ether as the non-solvent. The precipitated polymers were then washed with petroleum ether and dried in air. The test for quaternary halide groups in polymers was carried out with a reagent consisting of disulﬁne blue dissolved in 0.01 M hydrochloric acid and the colour evaluated spectrophotometrically at 630 nm. Saha and co-workers [70] and Palit [66, 67] found that there may be some uncertainty in the quantitative aspects of this method. Banthia and Palit [38] applied dye partition end-group analysis procedures to the examination of sulfate end groups in 13C-labelled PS.

8.8 Terminal Epoxides 8.8.1 IR spectroscopy Goddu and Delker [71] reported that terminal epoxides exhibit sharp absorbances relatively free of spectral interferences in the near IR at 2.2 μm and 1.65 μm (4532 cm–1 and 6060 cm–1). These absorptions result from overtones and/or combinations of fundamental vibrations found in the mid-IR. Using a dispersive infrared spectrometer, Dannenburg [72] conducted a study of epoxides in solution. Sensitivity was restricted by the capabilities of instrumentation available at that time. These investigations were limited to epoxide resins with an equivalent weight <1000 g resin/g-eq epoxy. Concentration levels for these resins were >1.0 eq/l. Peck and co-workers [73] used near-FT-IR spectrometry to achieve improved sensitivity over previous near-IR techniques. A mercury-cadmium-telluride detector has sufﬁcient sensitivity in the 4600–4500 cm–1 region to monitor the epoxide response at the 4532 cm–1 combination with an adequate S/N ratio. Co-addition of the interferograms can further diminish the inherent detector noise. Data manipulation routines can

300

Determination of End groups isolate the epoxide band from neighbouring absorption bands to facilitate direct numeric integration. With these methods they studied solutions of resins with epoxide equivalent weights >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

Introduction to Polymer Analysis

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

Introduction to Polymer Analysis

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

Infrared spectroscopy

Acrylonitrile determined

Terminal hydroxy

Infrared spectroscopy

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

Determination of End groups

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

-

Source: Author’s own ﬁles

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

Introduction to Polymer Analysis

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

Types of Unsaturation

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

-

Standard: 12.0 wt% diene 55

12.2

-

Standard: 15.0 wt% diene 55

14.8

-

High-impact polystyrene 1

8.6

9.7

High-impact polystyrene 2

5.6

5.8

High-impact polystyrene 3

9.0

1.2

High-impact polystyrene 4

11.2

11.4

High-impact polystyrene 5

5.8

5.9

Source: Author’s own ﬁles.

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

Introduction to Polymer Analysis

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

Source: Author’s own ﬁles

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

Introduction to Polymer Analysis

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

Types of Unsaturation

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

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

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

-

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

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

Introduction to Polymer Analysis

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 RR´C = 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

Types of Unsaturation

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

Types of Unsaturation

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) (%)

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

Source: Author’s own ﬁles.

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

Introduction to Polymer Analysis

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

Area from GC (%) Succinaldehyde

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

Source: Author’s own ﬁles.

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

Introduction to Polymer Analysis

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 (<0.5 per 1000 carbons) that characterisation through size exclusion chromatography in conjunction with low-angle laser light scattering or intrinsic viscosity measurements becomes feasible [41–44]. When 13C-NMR measurements have been compared with results from polymer solution property measurements, good agreement has been obtained between long-chain branching from solution properties with the concentration of branches six carbons long and longer [41, 42]. Unfortunately, these techniques utilising solution properties do not possess sufﬁcient sensitivity to detect long-chain branching in a range of 1 in 10,000 carbons, the level suspected in high-density polyethylenes. The availability of superconducting magnet systems has made measurements of long-chain branching by 13C-NMR a reality because of a greatly improved sensitivity. An enhancement by factors between 20 and 30 over conventional NMR spectrometers has been achieved through a combination of higher ﬁeld strengths, 20 mm probes, and the ability to examine polymer samples in essentially a melt state. The data discussed by Randall [40, 45] have been obtained from conventional iron magnet spectrometers with ﬁeld strengths of 23.5 kG and superconducting magnets operating at 47 kG. 13

C-NMR spectra from a homologous series of six linear ethylene 1-oleﬁn copolymers beginning with 1-propene and ending with 1-octene are reproduced in Figure 10.4. The side-chain branches are therefore linear, and progress from one to six carbons in length. Also, the respective 1-oleﬁn concentrations are <3%, thus only isolated branches are produced. Unique spectral ﬁngerprints are observed for each branch length. The chemical shifts, which can be predicted with the Grant and Paul parameters [40, 46] are given in Table 10.2 for this series of model ethylene-1-oleﬁn copolymers. The nomenclature used to designate those polymer backbone and side-chain carbons discriminated by 13C-NMR, is as follows:

362

Polymer Branching

(a) 3 CH2 37.8

CH2 30.4

(b) CH2 27.3

CH 33.3

(c)

CH2 27.3

CH2 2 CH2 27.3

CH2 CH 2 34.6 30.4

CH 38.2

CH2 27.3

CH2 34.2

CH2 30.4

+5 CH2 34.6

3 CH2 32.8

CH3CH2CH2CHCH2CH2CH2

CH2 27.3

CH2 34.6

CH2 CH2 CH2 CH3 2 4 CH2 22.8 CH2 26.9

CH3 14.1

CH 38.2

CH2 30.4

3 CH2 32.2

CH3 11.2

CH2CH2CH2CHCH2CH2CH2 CH2 4 CH2 3 CH2 2 CH3 2 CH2 22.8

CH3 14.6

(f)

(e)

CH3 2 CH2 26.7

CH 39.7

CH2CH2CH2CHCH2CH2CH2 CH2 CH2 CH3

2 CH2 20.3

CH 37.8

CH 38.2

CH2 34.1

(d)

CH2 30.4

3 CH2 CH2 34.4 36.8

CH2 30.4

CH2

CH2 20.0

CH2CH2CH2CHCH2CH2CH2

CH2CH2CH2CHCH2CH2CH2

CH3 14.1

CH2CH2CH2CHCH2CH2CH2 CH2 27.3 2 CH2 22.8

CH2 CH2 CH2 CH2 CH3 CH3 14.1

Figure 10.4 13C-NMR at 25.2 MHz of (a) O-ethylene-1-propene copolymer, (b) an ethylene-butene copolymer, (c) an ethylene-1-pentene copolymer, (d) an ethylene1-hexene copolymer, (e) an ethylene-1-heptene copolymer and (f) an ethylene1-octene copolymer. Reproduced with permission from J.C. Randall, Journal of Polymer Science: Polymer Physics Edition, 1973, 11, 275. © 1973, Wiley

363

Introduction to Polymer Analysis The distinguishable backbone carbons are designated by Greek symbols, while the side-chain carbons are numbered consecutively starting with the methyl group and ending with the methylene carbon bonded to the polymer backbone [40]. The identity of each resonance is indicated in Figure 10.4. In Figure 10.4, the ‘6’ carbon resonance for the hexyl branch is the same as A, the ‘5’ carbon resonance is the same as B and the ‘4’ carbon resonance is the same as G. Resonances 1, 2 and 3, likewise, are the same as the end-group resonances observed for linear polyethylene. Thus a six-carbon branch produces the same 13C spectral pattern as any subsequent branch of greater length. 13C-NMR, alone cannot therefore be used to distinguish a linear six-carbon from a branch of some intermediate length or a true long-chain branch. The capability for discerning the length of short chain branches has made 13C-NMR a powerful tool for characterising low-density polyethylenes produced by free-radical, high-pressure processes. Others are also present, and Axelson and co-workers in a comprehensive study [47] concluded that no unique structure can be used to characterise low-density polyethylenes. They have found non-linear short chain branches as well as 1,3 paired ethyl branches. Bovey and co-workers [41] compared the content of branches six carbons and longer in low-density polyethylenes with the long-chain branching results obtained through a combination of gel permeation chromatography and intrinsic viscosity. An observed good agreement led to the conclusion that the principal short chain branches contained fewer than six carbons and the six and longer branching content could be related entirely to long-chain branching. Others have now reported similar observations in studies where solution methods are combined with 13C-NMR [42]. As a result of the possible uncertainty of the branch lengths, associated with the resonances for branches six carbons and longer, 13C-NMR should be used in conjunction with independent methods to establish true long-chain branching. From these results it is easy to predict the 13C-NMR spectrum anticipated for essentially linear polyethylenes containing a small degree of long-chain branching. An examination of a 13C-NMR spectrum from a completely linear polyethylene, containing terminal oleﬁnic and saturated end groups, shows that only ﬁve resonances are produced. A major resonance at 30 ppm arises from equivalent, recurring methylene carbons, designated as ‘m’, which are four or more removed from an end group or a branch. Resonances at 14.1, 22.9 and 32.3 ppm are from carbons 1, 2 and 3, respectively, from the saturated, linear end group. A ﬁnal resonance, which is observed at 33.9 ppm, arises from an allylic carbon, designated as ‘a’, from a terminal oleﬁnic end-group. These resonances, depicted structurally next, are fundamental to the spectra of all polyethylenes:

364

Polymer Branching

Table 10.2 Polyethylene backbone and side-chain 13C chemical shifts in ppm from trimethylsilane (+0.1) as a function of branch length (carbon chemical shifts, which occur near 30.4 ppm, are not given because they are often obscured by the major 30 ppm resonance for the ‘n’ equivalent, recurring methylene carbons) (solvent: 1,2,4-trichlorobenzene; temperature: 125 oC) Branch length

Methine

6

5

4

3

2

1

33.3

37.6

27.5

20.0

2

39.7

34.1

27.3

11.2

26.7

3

37.8

34.4

27.3

14.6

20.3

36.8

4

38.2

34.6

27.3

14.1

23.4

-

34.2

5

38.2

34.6

27.3

14.1

22.8

32.8

26.9

34.6

6

38.2

34.6

27.3

14.1

22.8

32.2

30.4

27.3

1

34.6

Reproduced with permission from F.A. Bovey, F.G. Schilling, F.L. McCrackin and H.L. Wagner, Macromolecules, 1976, 9, 1, 76. © 1976, ACS

An introduction of branching, either long or short, will create additional resonances to those described above. From the observed 13C-NMR spectrum of the ethylene-1-octene copolymer Randall [40, 45] found that the A B and methane resonances associated with branches six carbons and longer occur at 34.56, 27.32 and 38.17 ppm, respectively. Thus, in high-density polyethylenes, where long-chain branching is essentially the only type present, 13C-NMR can be used to establish unequivocally the presence of branches six carbons long and longer. If no comonomer has been used during polymerisation, it is very likely that such resonances will be indicative of true long-chain branching. In any event, 13C-NMR can be used to pinpoint the absence of long-chain branching and place an upper limit upon the long-chain branch concentration whenever branches six carbons and longer are detected. It can be seen from the above considerations that 13C-NMR is a highly attractive method for characterising polyethylenes. A serious drawback is not encountered even though branches six carbons in length and longer are measured collectively. The short branches are generally less than six carbons in length and truly long-chain branches tend to predominate. On occasions there may be special exceptions for ‘intermediate’ branch lengths, so independent rheological measurements should be sought as a matter of course. Nevertheless 13C-NMR is a direct method, which possesses the required sensitivity to determine long-chain branching in high-density polyethylenes, and provide much information of microstructural interest.

365

Introduction to Polymer Analysis De Pooter and co-workers [19] applied 13C-NMR spectroscopy at sample temperatures of 130 °C to branching studies of copolymers ethylene with 1–10 mole% of propylene, butene-, hexene-1, octene-1 and 4 methyl pentene-1. 13C-NMR spectra were recorded with proton noise decoupling to remove unwanted 13C-1H scalar couplings. It was shown that the NOE is nearly full and constant for all carbons in these copolymers. Delay times between pulses in the pulse FT-NMR method must be ﬁve times the longest spin lattice relaxation time (T1) value for 99% relaxation. Unless ﬁve T1 values are allowed between pulses, saturation will occur and the area of the resonance will be attenuated. The methyl carbons in polyethylene generally have the longest T1 values, some of which are as long as 8–10 seconds. Therefore, to be quantitative, a 50 second delay time between pulses would be necessary. Because several thousand scans must be accumulated to achieve adequate signal to noise ratio, this delay time is not reasonable. An alternate approach for this analysis would be to ignore the methyls and other carbon resonances which have very long relaxation times. The T1 values for most of the carbons in the samples are <2.0 seconds and, therefore, a delay time of only 10 seconds would be necessary. Because there are several other carbons in the structure to use for quantitative purposes, the methyls and some branch carbons can be neglected without interfering with the quantitative aspects. This has the advantage of shortening the delay time between pulses from 50 seconds to 10 seconds with equivalent accuracy. The terminology used by De Pooter and co-workers [19] is that commonly accepted, using an octane-1 as example:

This investigation includes the chemical shift assignments (relative to that of the isolated methylene carbons at 30.0 ppm) and a set of experimental conditions for the quantitative analysis of branching in the commercial types of linear low-density polyethylenes. The integration limits given in Table 10.3 not only take into account the isolated branches, but also branches which are separated by one ethylene unit and branches next to each other. These latter two structures occur very infrequently in most commercial products but are signiﬁcant to the calculations.

366

Polymer Branching

Table 10.3 Integration limits for ethylene copolymersa Copolymer

Area

Region (ppm)

Ethene-propene

A B C C+D+E F G H

47.5 to 44.5 39.8 to 36.8 35.5 to 32.5 35.5 to 25.8 25.8 to 23.8 22.5 to 18.5 Peak at 21.6

Ethene-butane-1

A A´ B C D+E F

41.5 to 38.5 Peak at 39.4 37.8 to 36.8 36.0 to 33.2 33.2 to 25.5 25.2 to 24.0

Ethene-hexene-1

A B C D D+E F+G G H

41.5 to 40.5 40.5 to 39.5 39.5 to 37.0 Peak at 35.8 36.8 to 33.2 33.2 to 25.5 28.5 to 26.5 24.9 to 24.1

Ethene-octene-1

A B C D D+E F+G+H H I P

41.5 to 40.5 40.5 to 39.5 39.5 to 37.0 Peak at 35.8 36.8 to 33.2 33.2 to 25.5 28.5 to 26.5 25.0 to 24.0 24.0 to 22.0

Ethene-4-methylpentene

A B C D E F+G G H

46.5 to 43.5 43.0 to 41.8 41.8 to 40.5 37.5 to 34.2 Peak at 33.7 33.2 to 25.2 28.0 to 25.2 Peak at 24.1

a

Isolated methylene carbons at 30.0 ppm. Reproduced with permission from M. De Pooter, P.B. Smith, K.K. Dohrer, K.F. Bennett, M.D. Meadows, C.G. Smith, H.P. Schouwenaars and D.A. Geerards, Journal of Applied Polymer Science, 1991, 42, 2, 399. © 1991, John Wiley

367

Introduction to Polymer Analysis Good precision is obtained in this technique (Table 10.4).

Table 10.4 Precision in 13C-NMR determination of ethylene copolymers Comonomer incorporation level (mol%) EB

E4MP

EH

EO a

Known values

4.86 4.86

3.67 3.79

4.21 4.05a

4.92 4.93

Average operator 1

4.86 5.13 5.14 4.95b

3.73 3.77 3.71 3.54b

4.13 4.76 4.56 4.43b

4.93 5.15 4.94 5.04 5.05 5.14 5.00 5.14 5.12 5.16 5.15 5.07b 5.07b 5.11b 5.17b

Average operator 2

5.07 4.60 4.90 4.92 4.95 4.96 4.81 4.74 4.61 5.10 4.86

3.67 3.82 4.08

4.58 4.43 4.75

5.09c 5.18 5.20

Average

4.85d

3.95

4.59

5.19

a Contained approximately 0.8 ethyl branches. This value is not included in comonomer incorporation. b Operator 2 c Average value for the 14 determinations is 5.09 with a standard deviation of 0.07 (relative precision at 2S of 2.7%) d Average value for the 10 determinations is 4.85 with a standard deviation of 0.16 (relative precision at 2S of 6.8%) EB = ethylene-butene E4MP = ethylene-4-methylpentene-1 EH = ethylene-1-hexene EO = ethylene-1-octene Reproduced with permission from M. De Pooter, P.B. Smith, K.K. Dohrer, K.F. Bennett, M.D. Meadows, C.G. Smith, H.P. Schouwenaars and D.A. Geerards, Journal of Applied Polymer Science, 1991, 42, 2, 399. © 1991, Wiley

368

Polymer Branching

10.2.4 Polystyrene Randall [48] and other workers [49, 50] studied amorphous polystyrene. Randall [48] concluded that nine methylene resonances could be resolved. A high-ﬁeld strong methane resonance is also present, but shows no apparent conﬁgurational splitting. The observation of nine resonances is interesting because a 13C-NMR sensitivity to just tetrad sequences would have produced six resonances, whereas a complete hexad sensitivity would have produced 20 resonances.

10.2.5 Polyvinyl Chloride It has been recognised since the work of Cotman [51] that by studying the polyoleﬁn obtained from the reduction of PVC with lithium aluminium hydride, valuable structural information can be gained concerning the starting molecule. This reduction reaction has been investigated and reﬁned [52] to the point where conditions have been established so that the chlorine may be efﬁciently removed from the polymer without degradation. The reduced polymer is similar to high-density polyethylene in almost all respects [53]. Thus, studies that have been applied to polyethylene may also be applied to reduced PVC. Better qualitative agreement has resulted when G-ray radiolysis followed by identiﬁcation of the gaseous hydrocarbons by mass spectrometry used in conjunction with IR measurements is applied to reduced PVC than when this technique is applied to conventional polyethylene. It was concluded from the large yields of methane relative to butane and ethane that the predominant side-chains along the PVC backbone are mainly one carbon long. It has been demonstrated by 13CNMR that most of the short branches in PVC are pendent chloromethyl groups [54]. This information was obtained from PVC samples reduced with lithium aluminium hydride and lithium aluminium deuteride, respectively.

10.2.6 Polyvinyl Fluoride Investigations of the microstructure of poly(vinyl ﬂuoride) by ﬂuorine NMR have been reported by Weigert [55] and Bruch and co-workers [56]. Weigart studied a laboratory sample at 100 °C in dimethyl-d6 sulfoxide. The broadband proton-decoupled ﬂuorine NMR spectrum at 94.1 MHz showed two groups of peaks. The major group, in the region –178 to –182 ppm, was assigned to head-to-tail monomer units, while the minor group, in the region –189 to –197 ppm, was assigned to tail-to-tail monomer units. Peaks within each group were assigned to ﬂuorine atoms from monomer units in different tactic sequences. Bruch and co-workers [56] studied a commercial sample and also a laboratoryprepared sample containing no monomer reversals. Broadband proton-decoupled 369

Introduction to Polymer Analysis ﬂuorine NMR spectra of solutions in N,N-dimethylformamide at 130 °C were obtained at 188.2 MHz. A two-dimensional J-correlated NMR experiment enabled deﬁnitive assignments of the individual peaks to different tactic sequences to be made. Ovenall and Uschold [57] examined samples of poly(vinyl ﬂuoride) by 19F-NMR at 376.5 MHz and proton NMR at 400 MHz as solutions in N-methylpyrrolidone and as swollen gels in dimethyl-d6-sulfoxide. Weak peaks in the ﬂuorine NMR spectra have been assigned to CH2CH2F end groups and to tertiary ﬂuorine atoms at branch points. A differential decoupling experiment, in which the proton spectrum was observed with and without selective irradiation of the weak ﬂuorine peaks, permitted resonances of protons in the CH2CH2F end groups to be observed selectively. Pressure and temperature conditions used for polymer synthesis were found to inﬂuence branching but not the number of monomer reversals in the chain leading to head-tohead units. Monomer reversals account for about 13% of the vinyl ﬂuoride in the polymer. Chain branching varies from about one branch every 80 monomer units to one branch every 200 monomer units depending on reactor conditions. Low polymerisation temperature and high pressure yield the most linear polymers. Melting point and heat of fusion measured by differential scanning calorimetry (DSC) increase with the linearity of the product. The 400 MHz proton NMR spectrum of a poly(vinyl ﬂuoride) sample swollen in Me2SO-d6 showed features around 4.8 ppm which are assigned to CHF protons and the features around 2.0 ppm to CH2 protons. The feature at 2.9 ppm is from adventitious water. It was hoped it would be possible to identify in the proton spectrum weak points from the branched structures detected by ﬂuorine NMR. The proton spectrum was too broad for this to be done directly. Attempts at improving the resolution by Gaussian to Lorentzian resolution enhancement were unsuccessful. A weak peak visible at about 1.2 ppm may be due to methyl protons in end groups with structures CHFCH3, originating from hydrogen abstraction by growing polymer chains ending in reversed monomer units. Irradiation of a particular ﬂuorine peak should result in narrowing of the resonances of any protons coupled to the corresponding ﬂuorine atom. These will then appear in a differential mode as sharpened features with the original broad peaks subtracted. The experiment was tested by irradiating one of the major ﬂuorine resonances and gave a positive response. When both ﬂuorine irradiation frequencies were in an empty part of the spectrum, little or no response was given in the proton spectrum, even at the positions of the major peaks. The differential proton spectrum obtained when the ﬂuorine NMR spectrum was irradiated at –220 ppm corresponded to weak CH2F resonances. A positive response from the CH2F protons occurs at 4.5 ppm. No distinct peak is seen at this point in the normal proton spectrum. A weaker

370

Polymer Branching response from the CH2 protons in the CH2CH2F end groups is seen at 2.0 ppm. No responses were obtained in the proton spectrum when the weak features at –147 ppm were irradiated. These are assigned to tertiary ﬂuorine atoms which have the closest hydrogen atoms three bonds away, and the H–F couplings are presumably too weak to give a detectable response. For detecting the resonances of hydrogen atoms in chain ends and near branch points, the differential selective ﬂuorine irradiation experiment was used to link the two spectra via H–F couplings:

10.3 Vacuum Radiolysis 10.3.1 Ethylene Copolymer A study of the products produced upon vacuum radiolysis of ethylene homopolymers and copolymers is another means of obtaining information on branching in these polymers [58]. If a correction is applied, to take into account the fragments arising from scission at chain ends, the remaining products can be quantitatively accounted

371

Introduction to Polymer Analysis for as entirely due to scission of side-branches introduced onto the backbone chain by the A-oleﬁn comonomer. The cleavage of branches takes place, for all practical purposes, exclusively at the branch points at which the branches are attached to the backbone chain. The same data, together with similar radiolysis data of poly(3-methyl pentene-1) and poly(4-methyl pentene-1), further showed that all branches cleave with equal efﬁciency, regardless of their length. Radiolysis does, therefore, provide a reliable and convenient tool for the quantitative characterisation of high-pressure polyethylene with regard to the unique short chain branching distribution that is characteristic of each. The results obtained in a series of irradiations of ethylene-A-oleﬁn copolymers containing about 4 mole% comonomer are shown in Table 10.5. During radiolysis of polyethylene, there also takes place a certain amount of random scission at chain ends in addition to the cleavage of branches. The observed extraneous hydrocarbons are the products derived from this fragmentation at the chain ends. It is quite probable that a portion of the extraneous hydrocarbons is derived from scission of stray branches that may have been introduced on the chains by stray impurities during polymerisation, but the fact that one also observes a consistent decrease in the total amount of these extraneous hydrocarbons derived from the homopolymer with an increase in its molecular weight leads to the conclusion that the random scission at chain ends is their main cause. If one makes an appropriate allowance for these radiolysis fragments derived from chain ends, then the only signiﬁcant parafﬁn left in the radiolysis products of each copolymer is that corresponding to the branch introduced on the polyethylene backbone by the comonomer. Because it is generally agreed that high-pressure polyethylene contains various short branches, and not just one type of branch, if one wants to translate the hydrocarbon analysis of its radiolysis products into quantitative branch-type analysis, one will also need accurate information on the relative efﬁciency of the scission of different branch types. To obtain this information, Kamath and Barlow [59] irradiated some additional ethylene A-oleﬁn copolymers, differing signiﬁcantly in their comonomer content, and the hydrocarbons in their radiolysis products were analysed as before. Table 10.6 lists these results. The efﬁciency of scission is calculated from the known comonomer content (methyl group analysis) and the observed G values of the principal hydrocarbons after application of the appropriate correction against the small chainend fragmentation. All branches of up to six carbon atoms or more break off with equal efﬁciency, and that the branch length per se exerts little or no effect on the ease of scission.

372

Polymer Branching

Table 10.5 Radiolysis products of ethylene-A-oleﬁn copolymers (235U radiation source) G Value × 102*

Copolymers CH4

C2H6

C3H8

i-C4H10

n-C4H10

n-C5H12

n-C6H14

Ethylene-propylene

1.7

0.1

0.1

─

0.03

─

─

Ethylene-butene-1

0.2

1.5

0.1

0.1

0.2

0.03

─

Ethylene-pentene-1

0.2

0.3

2.1

0.02

0.05

─

─

Ethylene-hexane-1

0.4

0.3

0.2

0.03

1.2

0.02

─

Ethylene-octene-1

0.2

0.3

0.1

─

0.1

0.1

1.2

Linear polyethylene

0.2

0.3

0.2

0.03

0.06

0.05

─

*G value deﬁned as the number of molecules of the particular product per gramme of sample per 100 eV of incident radiation dose. This is calculated with 10% accuracy, from the gas chromatographic analysis data of the products, weight of the polymer sample irradiated, and radiation dose to which the sample has been subjected. Reproduced with permission from J. van Schooten and J.K. Evenhuis, Polymer, 1965, 6, 7, 343. © 1965, Elsevier

Table 10.6 Efﬁciency of radiation scission of short branches (G value × 102) Copolymers

CH3 per 1000 CH2

CH4

Ethylenepropylene

41.9 26.4

3.4 1.7

Ethylenebutane-1

28.7 21.6 4.1

Ethylenepentene-1

25.6 17.1

Ethylenehexene-1

23.5 7.8

Ethyleneoctene-1

19.1 15.1 11.6

C2H6

C3H8

n-C4H10

n-C6H14

Average G × 102 CH3 per 1000 CH4 0.074

2.0 1.5 1.1

0.072 2.1 1.1

0.072 1.2 0.7

0.073 1.3 1.2 0.9

0.074

Reproduced with permission from P.M. Kamath and A. Barlow, Journal of Polymer Science: Polymer Chemistry Edition, 1967, 5, 2023. © 1967, Wiley

373

Introduction to Polymer Analysis Table 10.7 presents results obtained in the radiolysis of high-pressure polyethylenes of various densities. Noteworthy is the disparity between the total number of all branches, as determined from the radiolysis data, and the methyl group content, as derived by the IR method. This may be attributed to the fact that the usual IR method of determining the methyl group content, based on the absorbance of 1379 cm–1 (7.25 μm), does not necessarily count all methyl groups. For example, if two methyl groups are attached to one carbon atom, as in polyisobutylene, the characteristic methyl absorption at 13.79 cm–1 (7.25 μm) splits, giving two bands at 1389 cm–1 (7.20 μm) and 1351 cm–1 (7.40 μm), respectively, [60], which are consequently lost in the methyl group determination procedure.

Table 10.7 Distribution of short-chain branches in high-pressure polyethylene (branches per 1000 methylene units)* Resin Density ─CH3 ─C2H5 ─C3H7- i-C4H9 ─N-C4H9 ─C5H11 Total No (g/cm3) branches per 1000 methylene**

Methyl group content per 1000 methylene†

A

0.934

2.9

11.8

1.7

0.3

5.9

1.6

24.2

18.8

B

0.929

2.5

15.2

2.1

0.3

8.5

2.0

30.6

24.0

C

0.924

3.8

16.7

2.2

─

9.9

1.8

34.4

30.9

* Calculated from G values of isolated hydrocarbons assuming scission efﬁciency of 0.073 × 102 per 1000 carbon atoms. Because short-chain branches outnumber chain ends by 30:1, no correction for the fragmentation products at chain ends was made. †

Determined by infrared analysis.

** Determined by radiolysis. Reproduced with permission from P.M. Kamath and A. Barlow, Journal of Polymer Science: Polymer Chemistry Edition, 1967, 5, 2023. © 1967, Wiley

In line with the observations reported by earlier workers, the two most populous branches, according to the radiolysis method, are ethyl and n-butyl, which occur in the ratio 2:1, as observed by others [16].

374

Polymer Branching

10.4 Pyrolysis-based Techniques 10.4.1 Elucidation of Short Chain Branching in Polyethylene Van Schooten and Evenhuis [58, 61] applied their pyrolysis (at 500 °C) hydrogenation– gas chromatographic technique to the measurement of short chain branching and structural details of three commercial polyethylene samples, a linear polyethylene, a Ziegler high-density polyethylene of density 0.945 and a high-pressure low-density polyethylene of density 0.92. Details of the pyrograms are given in Tables 10.8 and 10.9. The sizes of the iso-alkane peaks increase strongly with increasing amount of short chain branching, i.e., increase as we go from linear polyethylene to low-density polyethylene. None of the n-alkane peaks in the Ziegler high-density polyethylene program is signiﬁcantly greater than the corresponding peak in the pyrogram of a linear polyethylene. The pattern of the increased iso-alkane peaks in the Zeigler polyethylene pyrogram strongly suggests that these peaks are mainly due to ethyl side-groups. This is in good agreement with the results of electron irradiation experiments. For low-density polyethylene, van Schooten and Evenhuis [58, 61] found that the n-butane peak of the pyrogram showed a clear increase in size, and the n-pentane peak a smaller, although probably signiﬁcant, increase. The increases in the n-butane and n-pentane peaks are probably due to n-butyl and n-pentyl side-groups, respectively, pentyl groups being much less numerous than butyl groups. From the pyrograms of the ethylene–butene, ethylene–hexene-I and ethylene–octene-I copolymers, it is known that n-butyl side groups give a 2-methyl C6 peak which is at least equal to the 3-methyl C7 peak. In the low-density polyethylene pyrogram, the 3-methyl C7 peak is more than three-times as large. The main part of the 3-methyl C7 peak, therefore, is probably due to ethyl side-groups. This is in agreement with the sizes of the other 3-methylalkane peaks. It may be concluded that in high-density polyethylene, the short-chain branches are mainly ethyl and, for a smaller part, n-butyl groups, while some n-pentyl groups may also be present. Van Schooten and Evenhuis concluded that the results obtained by pyrolysis– hydrogenation–gas chromatography appear to be in good agreement with those obtained by IR and electron irradiation studies. They showed that the pyrograms of a linear polyethylene contained only very small peaks for branched and cyclic alkanes and very large n-alkane peaks. The largest peaks in the pyrogram are those for propane, n-hexane, n-heptane, n-decane and n-undecane, indicating important hydrogen exchange reactions followed by scission and the ﬁfth (C3 and n-C6), ninth (n-C7 and n-C10) and thirteenth (n-C11 and n-C14) carbon atoms. Hydrogen transfer with the sixth carbon atom would account for the rather large n-C4 peak (n-C4 and n-C7), but this peak could also be due to intermolecular chain transfer reactions.

375

Introduction to Polymer Analysis

Table 10.8 Relative sizes of n-alkane peaks in pyrograms of various polyethylenes Peak ratio

Linear polyethylene

Ziegler polyethylene

High-pressure polyethylene

n-C4-n-C7

0.71

0.74

1.38

n-C5-n-C7

0.59

0.57

0.75

n-C6-n-C7

1.52

1.25

1.28

n-C8-n-C7

0.36

0.65

0.68

n-C9-n-C7

0.67

0.64

0.72

n-C10-n-C7

1.02

0.91

0.88

n-C11-n-C7

1.00

0.71

0.75

Reproduced with permission from J. van Schooten and J.K. Evenhuis, Polymer, 1965, 6, 11, 561. © 1965, Elsevier

Table 10.9 Iso-alkane peaks in polyethylene pyrograms Sample

Iso-alkane peak

Ziegler polyethylene High-pressure polyethylene

i-C4, i-C5, 2MC5, 3MC5, 2MC6, 3MC6, 2MC7-4MC7, 3MC7, 2MC8-4MC8, 3MC8, 4MC9-5MC9, 2MC9, 3MC9, 4MC10-5MC10, 2MC10-4MC10, 3MC10

Reproduced with permission from J. van Schooten and J.K. Evenhuis, Polymer, 1965, 6, 11, 561. © 1965, Elsevier

Pyrograms were also prepared for a range of polyethylene containing different amounts of short chain branching (see Table 10.10 for details of samples). Previous work by high-energy electron irradiation and mass spectrometry has shown that the short branches in low-density polyethylenes are mainly ethyl and n-butyl groups, but other short branches have also been supposed to be present. The pyrograms of various polymers show marked differences which can be attributed to differences in short chain branching. The small amount of branching in Marlex 5003 and Ziegler polyethylenes is reﬂected only in the somewhat larger iso-alkane peaks, whereas the n-alkane pattern is practically the same as found for Marlex 50 (low branching). The Alkathene 2 and Lupolen H high-pressure polyethylenes show larger n-butane and n-pentane peaks (Figure 10.5). The iso-alkane peaks that show 376

Polymer Branching the largest increase (for Alkathane 2 and Lupolen H) are the iso-pentane and 3-methyl alkane peaks (Figure 10.6). The results in Figure 10.6 clearly show that the highest amount of branching is present in Lupolen H, the lowest in Marlex 50. Assuming arbitrarily that these polymers contain 24 and 1 short side chains/1000 carbon atoms, respectively, and that the relative increase of the n-butane and the iso-alkane peaks is linearly related with the amount of branching, then the branching frequency of the other three samples can be obtained by interpolation. These values are in good agreement with those found in the literature. From the large increase in the n-butane peak and the relatively small increase in the ethane peak, it is concluded that the two high-pressure polyethylenes (Alkathene 2 and Lupolen H) contain mainly n-butyl side-chains.

Table 10.10 Branching frequency of polyethylenes estimated from program (branch/1000 carbon atoms, from peak surface area ratios) Sample

n-Butane

Isopentane

3-Methyl pentane

3-Methyl heptane

Marlex 5003 - Phillips (polyethylene containing a little copolymerised butane-1)

0.5

4

2

2

Ziegler low-pressure polyethylene (about 3–6 branches/1000 carbon atoms)

0.3

7

5

5

Alkathene 2 - ICI (high-pressure polyethylene 20–30 branches/1000 carbon atoms)

21

19

17

19

Lupolen H - BASF (high-pressure polyethylene 20–30 branches/1000 carbon atoms)

24

24

24

24

Reproduced with permission from J. van Schooten and J.K. Evenhuis, Polymer, 1965, 6, 7, 343. © 1965, Elsevier

Michailov and co-workers [62] identiﬁed some of the isoalkane peaks on the hydrogenated pyrograms of low-density polyethylenes and pointed out that ethyl and butyl branches predominate. Seeger and Barrall [18], using ethylene-1-butene and ethylene-1-hexene copolymers as standards, detected about 20 ethyl and 10 n-butyl branches per 1000 carbons from the high yields of 3- and 5-methylalkanes. The resolution of the pyrograms was insufﬁcient because of the use of packed columns. Ahlstrom and Liebman [63] demonstrated ethyl and butyl branches from increases in the size of the 3-methylalkanes and n-butane peaks, in the size of the 3-methylalkanes and n-butane peaks, respectively, although 5-methyl aklanes were not resolved. 377

Introduction to Polymer Analysis 1.1 1.0

Relative peak surface area

0.9 0.8 0.7 0.6 0.5 0.4 0.3

Lupolen H Alkathene - 2 Ziegler Marlex 5003 Marlex 50

0.2 0.1 0 C2

C3

C 4 C5 C 6 C 7 C 8 C 9 Relative peak surface area of n -- alkanes (reference nC6) for various polyethylenes

C10

C11

Figure 10.5 Pyrolysis-gas chromatography carbon number distribution (relative peak surface areas of n-alkanes, reference n-C6) of various polyethylenes. Reproduced with permission from J. van Schooten and J.K. Evenhuis, Polymer, 1965, 6, 7, 343. © 1965, Elsevier Later, Mlejnek [64], employing a more effective open tubular column and Curiepoint pyrolyser, obtained high-resolution pyrograms of polyethylenes. On the basis of the relative peak area of iso-alkanes, he concluded that the presence of methyl, ethyl, and butyl groups was equally probable in low-density polyethylenes. In recent work by Sugimura and co-workers [65-67], glass capillary pyrolysis–hydrogenation gas chromatography (PHGC) was applied to the quantitative analysis of the short chain branches. Relating peak intensities and characteristics of iso-alkanes for low-density polyethylene to those of reference model copolymers, they determined methyl, ethyl and butyl branch contents in low-density polyethylenes. Liebman and co-workers [68] reported a comparable study on short chain branches in polyethylenes by fused-silica capillary PHGC and 13C-NMR spectroscopy. To extend the interpretative capabilities of PHGC, a computer simulation method was applied to reproduce the fragmentation pattern of the pyrogram of low-density polyethylenes using the data of known references. They suggested that PHGC can

378

Polymer Branching estimate contents of short chain branching as low as one branch per 10,000 CH2. Haney and co-workers [69] proposed a new PHGC method at relatively low pyrolysis temperature (360 °C). By this method, enhanced yields of the products pertaining to the branch points were observed on the resulting pyrograms of polyethylenes. The observed excess amount of 3-methylpentane from polyethylenes was qualitatively attributed to Willbourne-type branches such as 2-ethylhexyl (branched branch) and 1,3-paired ethyl (pair branch). 0.22 0.20

Relative peak surface area

0.18 0.16

Lupolen H Alkathese - 2 Ziegler Marlex 5003 Marlex 50

0.14 0.12 0.10 0.08 0.06 0.04 0.02 0 iC4 iC5

2MC5 3MC5 2MC6 3MC6 2MC7 3MC7 3MC8 3MC9 Relative peak surface areas of iso - alkanes (reference nC6) for various polyethylenes

Figure 10.6 Pyrolysis-gas chromatography carbon number distribution (relative peak surface areas of iso-alkanes, references n-C6) of various polyethylenes. Reproduced with permission from J. van Schooten and J.K. Evenhuis, Polymer, 1965, 6, 7, 343. © 1965, Elsevier Existence of amyl branches and branches longer than hexyl, which have been conﬁrmed by 13C NMR spectroscopy [32, 41, 45, 47, 70], has not been clearly characterised by PHGC, mainly because of the lack of well-deﬁned model polymers and of insufﬁcient resolution for the associated iso-alkanes on the pyrograms. PHGC methods for determining the short chain branching in low-density polyethylene (LDPE) were extended up to hexyl branches using a high-resolution fused-silica capillary column and well-characterised model copolymers. Branch content thus estimated was compared with that obtained by 13C-NMR spectroscopy. The possible existence of

379

Introduction to Polymer Analysis pair and branched branches is also discussed on the basis of theoretical and observed yields for 5-ethylnonane, which is characteristic of the ethyl branch. Ohtani and co-workers [71] extended the hydrogenation technique by using fused silica columns as compared with the previous pyrograms for LDPE [67]. Drastic improvement in resolution was attained for the iso-alkanes, which are closely related to short chain branching. Although the peak resolution between 2-methyldecane (2M) and 3-ethylnonane (£3) in the C11 region was not sufﬁcient with an ordinary glass capillary column, almost complete separation was attained by the fused-silica column. Thus, the determination of the short chain branching was carried out on the basis of characteristic peak intensities in the C11 region according to the same approach reported previously [67]. Table 10.11 summarises the possible fragmentation products of C11 formed from the short-branch structures in the polymer chain provided that side-chain cleavages do not occur. Each compartment divided by oblique dotted lines depicts the fragments corresponding to each type of branching from C1 to C10. Branches up to hexyl were taken into consideration except for propyl branches, whose existence in LDPE, if any, is known to be negligibly small.

Table 10.11 Possible fragmentation products of C11 from the site of branching in PE branch type scission point

C9

C8

C7

C6

C5

C4

C3

C2

n

n

n

n

n

n

n

n

n

n

n

2M

3M

4M

5M

5M

4M

3M

n

3M

3E

4E

5E

4E

3E

3M

n

4M

4E

4P

4P

4E

4M

n

5M

5E

4P

5E

5M

n

5M

4E

4E

5M

n

4M

3E

4M

n

3M

3M

n

2M

n

n : n-undecane 2M : 2-methyldecane 3M : 3-methyldecane 4M : 4-methyldecane 5M : 5-methyldecane 3E : 3-ethylnonane 4E : 4-ethylnonane 5E : 5-ethylnonane 4P : 4-propyloctane

C

C10

C

C

C

C

C

C

C

C

C

C R

C

C

C

2M

C

C1

C

C

C

C

C

C

R: short branches

Reproduced with permission from H. Ohtani, S. Tsuge and T. Usami, Macromolecules, 1984, 17, 12, 2557. © 1984, ACS

380

1-hexene 1-heptene 1-octene

EHP

EO

0.15

hexyl

amyl

butyl

ethyl

methyl

Branch type

0.025

0.070

0.33

20

12

18

24

20

Branch contents

0.01

0.16

5E

0.2

0.05

4P

Key peak intensities relative to peak intensities of I(n-C11)obsd + 1000. b Calculated correlation factor c the underlined values are for key peaks. Source: Author’s own ﬁles

a

0.20

0.15

0.032

1.00

0.21

EHX

390

0.031

0.18

0.30

2.52

1.00

0.31

1-butene

51.4 (4M)

590

0.095

0.035

0.17

1.00

0.24

EB

20 hexyl

EO

20.5 (4E)

410

0.061

0.36

0.33

0.093

1.00

0.10

propylene

12 amyl

EHP

43.5 (5M)

430

0.38

3E

0.44

4M

0.46

4E

1.00

5M

3M

2M

Relative peak intensities of isoalkanes to key peakc

E/P

18 butyl

EHX

55.3 (3M)

260

f

A-Oleﬁn comonomer

24 ethyl

EB

77.3 (2M)

Iobsda

Sample

20 methyl

Branch content per 1000 C

E/P

Reference polymerb

Table 10.12 Relative peak intensities of isoalkanes to key peak and observed correlation factors for reference copolymer

Polymer Branching

381

Introduction to Polymer Analysis The most intense peak for amyl was 5-methyldecane (5M), which was also the most intense for butyl, so that the second most intense peak, 4-ethylnonane (4E) was allocated for the key peak for amyl instead of 5M. Thus, the correlation factors (f) for the content of branches deﬁned previously [67] were ﬁrst calculated from the corresponding data for the model co-polymers with known branch content: I(key)obsdf = branch number/1000 C Where I(key)obsd was the relative value when the observed peak intensity of n-C11 was regarded as 1000(I(n-C11)obsd = 1000). Table 10.12 summarises the observed I(key)obsd and the calculated f values together with the relative peak intensities of the iso-alkanes to the corresponding key peaks. Table 10.13 summarises the short chain branching content obtained by this method for the four low-density polyethylenes along with that found by 13C-NMR spectroscopy. As a whole, the estimated individual short-branch content and the total values are in fairly good agreement with those obtained by 13C-NMR spectroscopy.

Table 10.13 Estimated short chain branch (SCB) content in LDPE by PHGC and 13C-NMR spectroscopy Estimated SCB content/1000 Ca

Sample

LDPE-A LDPE-B LDPE-C LDPE-D a

Methyl

Ethyl

Butyl

Amyl

Hexyl

1.3 (0.4)

4.9 (6.4)

8.3 (7.4)

2.3 (2.6)

0.2

1.5 (0.5)

7.2 (7.2)

11.2 (8.5)

2.2 (2.7)

0.3

1.3 (0.5)

4.8 (5.4)

8.6 (6.4)

1.9 (2.2)

0.4

1.0 (0.1)

2.1 (2.3)

4.5 (2.4)

0.6 (0.7)

0.3

Longerb

Totalc

(2.8)

17. (19.9)

(3.5)

22.4 (24.9)

(2.4)

17.0 (17.3)

(1.0)

8.5 (6.7)

Estimated SCB content by PHGC are given ﬁrst, and estimated SCB content by C-NMR spectroscopy are given in parentheses.

13 b

The longer-chain branches are not taken into consideration in the case of PHGC.

c

Content by NMR also involves propyl branch content between 0.2 and 0.5.

Reproduced with permission from H. Ohtani, S. Tsuge and T. Usami, Macromolecules, 1984, 17, 12, 2557. © 1984, ACS

382

Polymer Branching

10.4.2 Short Chain Branching in Ethylene–Higher Oleﬁn Copolymers Van Schooten and Evenhuis [58] reported on the application of their pyrolysis (at 500 °C)–hydrogenation–gas chromatographic technique to the measurement of short chain branching and structural details in modiﬁed polyethylene containing small amounts of comonomer (about 10% by weight of butene-1, hexene-1 and octene-1). A survey of the iso-alkane peaks of the pyrograms of these polymers with their probable assignment is given in Table 10.14. The effect of the comonomer on the relative sizes of the n-alkane peaks is given in Table 10.15. The peaks stemming from the total side group appear to be increased in intensity (n-C4 peak in the polyethylene–hexane pyrogram, n-C6 peak in polyethylene–octene pyrogram). The polyethylene–octene pyrogram also shows that the peak stemming from the n-alkane with one carbon atom less is enlarged. This suggests that chain scission may occur at the A and at the B C–C bond. Only the latter type of scission will produce methyl-substituted iso-alkanes, e.g., after intramolecular hydrogen transfer. A list of iso-alkanes that may be expected to be formed in this way is compared in Table 10.16 with a list of those that have been found to be signiﬁcantly increased in size in the copolymer pyrograms.

(a) Polyethylene, Ziegler (linear) 4 - P?

2 - M?

(b) Polyethylene low density (0.918 g/cc) 5-M

3-M

4-E

2-M

3M=3 methyl isomer 3E=3 ethyl isomer 2M=2 methyl isomer 4P=4 propyl isomer 4E=4 ethyl isomer 5M=5 methyl isomer

(c) Copolymer ethylene with 1 - hexene (0.5%) 5-M

4-E

Retention time

Figure 10.7 Single branched fragments (C10) from ethylene-hexene-1 copolymer. Reproduced with permission from M. Seeger and E.H. Barrall, Journal of Polymer Science, Polymer Chemistry Edition, 1975, 13, 7, 1515. © 1975, Wiley

383

Introduction to Polymer Analysis

Table 10.14 Areas of the main iso-alkane peaks in the pyrograms of linear polyethylene, ethylene - butene-1, ethylene - hexene-1 and ethylene octene-1 copolymers. Component

Peak area in arbitrary units Linear PE

Ethylene butene-1

Ethylene hexene-1

Ethylene octene-1

i-C4

10

20

37

30

i-C5

18

47

62

46

2MC

9

14

18

20

5

3MC ─CyC5

31

54

46

76

2MC6

14

13

74

22

3MC6

13

33

24

26

2MC7─4MC7

16

17

25

27

3MC7─3EC6

16

67

64

32

2MC8─4MC8─ECyC6

50

44

38

119

3MC8

9

39

12

17

4MC9─5MC9─4EC8

11

21

51

36

94

i─PCyC6─BuCyC5 2MC9─n-PCyC6

40

35

40

65

3MC9

8

27

6

45

4MC10─5MC10─sec BuCyC6

25

23

67

32

2MC10─4MC10─n-BuCyC6

33

30

47

102

3MC10

15

32

14

34

Reproduced with permission from J. van Schooten and J.K. Evenhuis, Polymer, 1965, 6, 7, 343. © 1965, Elsevier

Seeger and Barrall [18] investigated the applicability of PHGC to the elucidation of side-chain branching in ethylene-hexane-1 copolymers. Figure 10.7 shows the single-branched fragment pattern up to ten carbon atoms obtained for this copolymer compared with those obtained for Ziegler (linear) polyethylene and low-density (0.918 g/cm3) polyethylene. It is possible not only to measure the kind of branching present, but also to identify the polymer with this simple series. The distribution of isomers varies signiﬁcantly with the type of branching. In the copolymer with 1-hexene (butyl branches), the

384

Polymer Branching 5-methyl and 4-methyl isomers are dominant. Low-density polyethylene (0.918 g/cm3) shows a high 3-methyl peak as well as a high 5-methyl and 4-methyl peak yield. This indicates that ethyl and butyl branches are present in the polyethylene material. Such evidence conﬁrms former suggestions, mainly on the basis of IR measurements, that low-density polyethylene has branches which are the result of certain intramolecular transfer reactions during high-pressure polymerisation [16].

Table 10.15 Relative sizes of n-alkane peaks for polyethylene and for ethylene butene-1, ethylene - hexene-1 and ethylene - octene-1 copolymers. Peak ratio

Marlex 50

Ethylene─butene-1

Ethylene─hexene-1

Ethylene─octene-1

n-C4─n-C7

0.71

0.86

1.15

0.86

n-C6─n-C7

0.59

0.69

0.61

0.91

n-C6─n-C7

0.52

1.66

1.62

1.88

n-C8─n-C7

0.63

0.58

0.66

0.79

n-C9─n-C7

0.67

0.63

0.62

0.75

n-C10─n-C7

1.02

0.93

0.99

1.05

n-C11─n-C7

1.00

0.82

0.90

0.92

Reproduced with permission from J. van Schooten and J.K. Evenhuis, Polymer, 1965, 6, 7, 343. © 1965, Elsevier

Table 10.16 Expected and observe iso-alkane peaks in copolymer pyrograms Ethylene - butene-1

Ethylene - hexene-1

Ethylene - octene

Expected

Observed

Expected

Observed

Expected

Observed

Number of backbone C atoms in fragment

─

─

─

i-C4

─

i-C4

─

─

─

─

i-C5

─

i-C5

─

i-C5

i-C5

2MC6

2MC6

2MC8

2MC8

3

3MC5

3MC5

3MC7

3MC7

3MC9

3MC9

4

3MC6

3MC6

4MC8

─

4MC10

4MC10

5

3MC7

3MC7

5MC9

5MC9

5MC11

─

6

3MC8

3MC8

5MC10

5MC10

6MC12

─

7

3MC9

3MC9

5MC11

─

7MC13

─

8

Reproduced with permission from J. van Schooten and J.K. Evenhuis, Polymer, 1965, 6, 7, 343. © 1965, Elsevier

385

Introduction to Polymer Analysis

10.4.2 Branching in Ethylene–propylene Copolymers 10.4.2.1 Microstructure of Ethylene-Propylene Copolymers In a study of the microstructure of ethylene-propylene copolymers, van Schooten and Evenhuis [72] carried out a high-vacuum pyrolysis at 40 °C, followed by trapping of released volatiles in dry ice–acetone, to produce a mixture of oleﬁns and A-oleﬁns. The most volatile fractions, collected in dry ice–acetone, were hydrogenated to saturated hydrocarbons, which were analysed by gas–liquid chromatography. For samples of copolymer prepared using titanium trichloride or vanadium oxychloride catalysts the chromatogram of this fraction showed peaks of 2,4-dimethylheptane, 2-methylheptane, 4-methylheptane, 2,4-dimethylhexane, 3-methylhexane and 2-methylhexane, but only in the chromatogram of the volatile fraction from the copolymer produced using vanadium was a peak of 2,5-dimethylhexane found. This is an indication that polymers prepared with a catalyst containing vanadium oxychloride contain methylene sequences of two units between branches. van Schooten and co-workers [72] conclude that ethylene–propylene copolymers prepared with vanadium-containing catalysts, especially those with VOCl3 or VO(OR)3, have methylene sequences of two and four units.

10.4.3 Branching in PVC Ahlstrom and co-workers [63] applied the techniques of Py-GC and PHGC to determination of short-chain branches in PVC and reduced PVC. Their attempts to determine the shortchain branches in PVC by Py-GC were complicated by an inability to separate all the parameters affecting the degradation of the polymer. Not only does degree of branching change the pyrolysis pattern, but so do tacticity and crosslinking [73]. For the pyrolysis of PVC, a ribbon probe was used. Online hydrogenation of the pyrolysis products was accomplished using hydrogen as the carrier gas with 1% palladium on Chromasorb-P catalyst inserted in the injection port liner. Maximum triplet formation occurred at C14 for LDPE for reduced PVC and at C15 for HDPE. The occurrence of the peak maxima at C14 for reduced PVC indicates that the total branch content is higher than that of HDPE. However, aside from the C14-C15 peak maxima difference, the pyrolysis pattern for even the most highly branched PVC resembles HDPE more than LDPE. These data indicate that the type of short-chain branch in PVC is qualitatively more like that of HDPE, but that the sequence length between branch sites is shorter in LDPE and PVC. Because LDPE contains a large amount of ethyl and butyl branches, and PVC and HDPE contain mainly methyl and some ethyl branches, this qualitative resemblance would be expected. A relative measure of the total amount of short-chain branches for these polymers can be obtained by calculating the percentage of branched products formed (Table 10.17). 386

Polymer Branching

Table 10.17 Relative total branch content of high-density polyethylene, lithium aluminium hydride (LAH)-reduced PVC and low-density polyethylene Sample

Branch products (%)

High-density polyethylene

12.0

Reduced PVC*

19.0

Low-density polyethylene

26.0

* Average value Reproduced with permission from D.H. Ahlstrom, S.A. Liebman and K.B. Abbas, Journal of Polymer Science, Polymer Chemistry Edition, 1976, 14, 10, 2479. © 1976, Wiley

More information about the speciﬁc type of short-chain branch in PVC can be found from an examination of the C1–C11 hydrocarbons. Here quantitative differences between the reduced PVC become more apparent. The most obvious differences occur in the amounts of iso-C7 and iso-C8 products formed, which indicate differences in the total branch content. As the amount of short-chain branching (C10) in the reduced PVC increases, there is a decrease in the amount of iso-alkanes formed (Table 10.18). The data in Table 10.18 show small (but distinguishable) differences in the shortchain branch content of the reduced PVC.

10.5 Size-Exclusion Chromatography (SEC) 10.5.1 Polyethylene Application of size-exclusion chromatography ofﬂine low-angle and multi-angle light scattering (SEC–MALLS) [74]. This technique has been used to determine long-chain branching in polyethylenes. An example of the latter is the work of Tackx and Tacx [75]. These workers pointed out that the separation of low-density polyethylene exclusively according to one molecular parameter (e.g., molar mass, long-chain branching or short chain branching), without interference of the other parameters, with subsequent characterisation of the fractions according to exclusively another parameter, is nearly impossible. Separation according to short chain branching is inﬂuenced by molar mass and type of short chain branching. Application of liquid–liquid phase separation results in a separation mainly according to long-chain branching. However, an inﬂuence of molar mass cannot be excluded.

387

Introduction to Polymer Analysis Long-chain branching causes a decrease in molecular size and hence in the radius of hydration and the hydrodynamic radius, as compared to polymers having a linear structure and having the same molar mass. In principle there is a simple relation between the ratio of the radii of gyration and the hydrodynamic radii: g´ = gb. At constant temperature and one solvent/polymer combination, the b value is mainly dependent on the chain architecture. In the classical approach, determination of b as a function of molar mass requires time consuming fractionation with subsequent characterisation of the fractions to determine the hydrodynamic radius and the radius of gyration. By application of SEC in combination with only multi-angle laser light scattering and using the universal calibration principle based on viscometry, it appeared possible to determine the g and g’ as a function of the molar mass. Utilisation of this method revealed that low-density polyethylene exhibits a continuous decrease with increasing molar mass of the b-value from 1.8 and 1.2 for a tubular product, and from 1.5 to 1 for an autoclave product. This was also expected from the polymerisation conditions.

10.5.2 Other Polymers SEC with viscometric detection has been used in branching studies on polystyrene [76]. Paparagopoulos and Dondos characterised branching in polystyrene star polymers by gel permeation chromatography and light scattering [77].

Table 10.18 Short-chain branch content of lithium aluminium hydridereduced PVC Sample

iso C8/n-C8

C10 (%)

High-density polyethylene

0.1

8.6

Pevikon R341

0.52

7.7

Nordforsk E-80

0.71

7.2

Nordforsk S-80

1.20

7.0

Nordforsk S-54

1.59

6.8

Nordforsk E-54

1.63

6.5

Ravinil R100/650

1.63

6.5

Shinttsu TK1000

1.87

6.3

Low-density polyethylene

1.83*

5.2

* This ratio does not include C4 branch content Reproduced with permission from D.H. Ahlstrom, S.A. Liebman and K.B. Abbas, Journal of Polymer Science, Polymer Chemistry Edition, 1976, 14, 10, 2479. © 1976, Wiley

388

Polymer Branching

References 1. R. Kuhn, H. Kroemer and G. Rossmanith, Angewandte Makromolekulare Chemie, 1974, 40-41, 361. 2. R. Kuhn and H. Kroemer, Colloid and Polymer Science, 1982, 260, 12, 1083. 3. G. Luft, R. Kampf and H. Seidl, Angewandte Makromolekulare Chemie, 1982, 108, 203. 4. G. Luft, R. Kampf and H. Seidl, Angewandte Makromolekulare Chemie, 1983, 111, 133. 5. E. Nordmeier, U. Lanver and M. Lechner, Macromolecules, 1990, 23, 4, 1072. 6. E. Nordmeier, U. Lanver and M. Lechner, Macromolecules, 1990, 23, 4, 1077. 7. H.N. Cheng, Polymer Bulletin, 1986, 16, 5, 445. 8. M.E.A. Cudby and A. Bunn, Polymer, 1976, 17, 4, 345. 9. C. Baker and W.F. Maddams, Makromolekulare Chemie, 1976, 177, 2, 437. 10. M.A. McRae and W.F. Maddams, Makromolekulare Chemie, 1976, 177, 2, 449. 11. ASTM D2238-92, Standard Methods for Absorbance of Polyethylene Due to Methyl Groups at 1378 cm, 2004. 12. M. Laiber, H. Opera, M. Toader and V. Laiber, Revista De Chimie (Bucharest), 1977, 28, 9, 881. [Chemical Abstracts, 1978, 88, 23539p]. 13. R. Salovey and J.V. Pascale, Journal of Polymer Science Part A: General Papers, 1964, 2, 5, 2041. 14. J.R. Neilson and R.F. Holland, Journal of Molecular Spectroscopy, 1960, 4, 1-6, 488. 15. D.A. Boyle, W. Simpson and J.D. Waldron, Polymer, 1961, 2, 323. 16. A.H. Willbourn, Journal of Polymer Science, 1959, 34, 127, 569. 17. A.G. Nerheim, Analytical Chemistry, 1975, 47, 7, 1128. 18. M. Seeger and E.M. Barrall, Journal of Polymer Science: Polymer Chemistry Edition, 1975, 13, 7, 1515.

389

Introduction to Polymer Analysis 19. M. De Pooter, P.B. Smith, K.K. Dohrer, K.F. Bennett, M.D. Meadows, C.G. Smith, H.P. Schouwenaars and D.A. Geerards, Journal of Applied Polymer Science, 1991, 42, 2, 399. 20. F.C. Stehling and J.R. Knox, Macromolecules, 1975, 8, 5, 595. 21. C.J. Carman, R.A. Harrington and C.E. Wilkes, Macromolecules, 1973, 10, 3, 536. 22. J.C. Randall, Journal of Polymer Science, Polymer Physics Editon, 1974, 12, 4, 703. 23. A. Zambelli, P. Locatelli, G. Bajo and F.A. Bovey, Macromolecules, 1975, 8, 5, 687. 24. F.C. Stehling and J.R. Knox, Macromolecules, 1975, 8, 595. 25. J. Schaefer and D.F.S. Natusch, Macromolecules, 1972, 5, 4, 416. 26. D.E. Axelson, L. Mandelkern and G.C. Levy, Macromolecules, 1977, 10, 3, 557. 27. A. Provasoli and D.R. Ferro, Macromolecules, 1977, 10, 4, 874. 28. G.K. Kraus and C.J. Stacy, Journal of Polymer Science, Polymer Symposium, 1973, 43, 329. 29. A. Nishoika, I. Ando and J. Matsimoto, Bunseki Kagaku, 1977, 20, 5, 308. 30. D.C. Bugada and A. Rudin, European Polymer Journal, 1987, 23, 11, 847. 31. J.C. Randall in 13C NMR Polymer Analysis, Ed., M.W. Pasika, ACS Symposium Series No.103, ACS, Washington, DC, USA, 1979. 32. J.C. Randall, Journal of Polymer Science, Polymer Physics Edition, 1976, 14, 11, 2083. 33. F.C. Stehling, Journal of Polymer Science Part A: General Papers, 1964, 2, 4, 1815. 34. Y. Inoue, A. Nishioka and R. Chûjô, Journal of Polymer Science: Polymer Physics Edition, 1973, 11, 1, 2237. 35. R.S. Porter, S.W. Nicksic and J.F. Johnson, Analytical Chemistry, 1963, 35, 12, 1948.

390

Polymer Branching 36. C.E. Wilkes, C.J. Carmen and R.A. Harrington, Journal of Polymer Science: Polymer Symposium, 1973, 43, 237. 37. G.J. Rais, P.E. Johnson and J.R. Knox, Macromolecules, 1977, 10, 4, 773. 38. J.R. Paxson and J.C. Randall, Analytical Chemistry, 1978, 50, 13, 1777. 39. J.C. Randall, Macromolecules, 1978, 11, 1,33. 40. J.C. Randall, Journal of Polymer Science, Polymer Physics Edition, 1973, 11, 2, 275. 41. F.A. Bovey, F.G. Schilling, F.L. McCrackin and H.L. Wagner, Macromolecules, 1976, 9, 1, 76. 42. G.N. Foster, Polymer Preprints, 1970, 20, 2,463. 43. A. Barlow, L. Wild and R. Ranganath, Journal of Applied Polymer Science, 1977, 21, 12, 3319. 44. L. Wild, R. Ranganath and A. Barlow, Journal of Applied Polymer Science, 1977, 21, 12, 3331. 45. J.C. Randall, Journal of Applied Polymer Science, 1978, 22, 2,585. 46. D.M. Grant and E.G., Paul, Journal of the American Chemical Society, 1964, 86, 15, 2984. 47. D.E. Axelson, G.C. Levy and L. Mandelkern, Macromolecules, 1979, 12, 1, 41. 48. J.C. Randall, Journal of Polymer Science, 1975, 13, 5, 889. 49. Y. Inoue, A. Nishioka and R. Chuyo, Makromolekulare Chemie, 1972, 156, 207. 50. K. Matsuzaki, T. Uryu, K. Osada and T. Kawamura, Macromolecules, 1972,5, 6, 816. 51. J.D. Cotman, Journal of the American Chemical Society, 1955, 77, 10, 2790. 52. M. Carrega, C. Bonnebat and G. Zednik, Analytical Chemistry, 1970, 42, 14, 1807. 53. A.J. de Vries, C. Bonnebat and M. Carrega, Pure & Applied Chemistry, 1971, 26, 2, 209.

391

Introduction to Polymer Analysis 54. F.A. Bovey, K.B. Abbas, F.C. Schilling and W.H. Starnes, Macromolecules, 1975, 8, 4, 437. 55. F. Weigert, Organic Magnetic Resonance, 1971, 3, 3, 373. 56. M.D. Bruch, F.A. Boven and R.E. Cais, Macromolecules, 1984, 17, 12, 2547. 57. D.W. Ovenall and R.E. Uschold, Macromolecules, 1991, 24, 11,3235. 58. J. van Schooten and J.K. Evenhuis, Polymer, 1965, 6, 7, 353. 59. P.M. Kamath and A. Barlow, Journal of Polymer Science: Polymer Chemistry Edition, 1967, 5, 2023. 60. L.J. Ballamy, Infrared Spectra of Complex Molecules, Wiley, New York, NY, USA, 1959. 61. J. van Schooten and J.K. Evenhuis, Polymer, 1965, 6, 11, 561. 62. L. Michajlov, P. Zugenmaier and H-J. Cantow, Polymer, 1968, 9, 325. 63. D.H. Ahlstrom, S.A. Liebman and K.B. Abbas, Journal of Polymer Science, Polymer Chemistry Edition, 1976, 14, 10, 2479. 64. O. Mlejnek, Journal of Chromatography, 1980, 191, 181. 65. T. Sugimura and S. Tsuge, Macromolecules, 1979, 12, 3, 512. 66. S. Tsuge, Y. Sugimura and T. Nagaya, Journal of Analytical Applied Pyrolysis, 1980, 1, 3, 221. 67. Y. Sugimura, T. Usami, T. Nagaya and S. Tsuge, Macromolecules, 1981, 14, 6, 1787. 68. S.A. Liebman, D.H. Ahlstrom, W.H. Starnes, Jr., and F.C. Schilling, Journal of Macromolecular Science Chemistry, 1982, A17, 6, 935. 69. M.A. Haney, D.W. Johnston and B.H. Clampitt, Macromolecules, 1983, 16, 11, 1775. 70. T. Usami and S. Takayama, Macromolecules, 1984, 17, 9, 1756. 71. H. Ohtani, S. Tsuge and T. Usami, Macromolecules, 1984, 17, 12, 2557. 72. J. van Schooten, E.W. Duch and R. Berkenbosch, Polymer, 1961, 2, 357.

392

Polymer Branching 73. M. Suzuki, S. Tsuge and T. Takeuchi, Journal of Polymer Science, 1972, 10, 4, 1051. 74. S. Pang and A. Rudin, Polymeric Materials Science and Engineering, 1991, 65, 95. 75. P. Tackx and J.C.J.F. Tacx, Polymer, 1998, 39, 14, 3109. 76. G. Kuo, T. Provder and M.E. Koehler, Polymeric Materials Science and Engineering, 1991, 65, 142. 77. D. Paparagapoulos and A. Dondos, European Polymer Journal, 2004, 40, 10, 2305.

393

Introduction to Polymer Analysis

394

11

Block Copolymers

Water-soluble triblock polymers of poly(oxypropylene) and poly(oxyethylene), the so-called ‘pluronics’, ﬁnd widespread application as non-ionic surface-active agents. Poly(oxyalkylene) block copolymers, having surfactant properties, have been utilised as lubricants, dispersants, antistatic agents, foam control agents, and solubilisers; and in numerous other applications in the areas of pharmaceuticals, cleaning agents, foods, and personal care products [1, 2]. This diversity in applications results from the various possible structural arrangements of poly(oxyalkylene) block copolymers. In these materials, ethylene oxide (EOx) blocks provide hydrophilicity, and propylene oxide (PO) blocks provide the hydrophobicity necessary for surfactant properties. The distribution of chemical composition can be modiﬁed to achieve the desired surfactant performance. Within this context, it is necessary to obtain a complete description of the molecular composition distribution of individual blocks in the copolymer to understand the structure–function relationships of surfactants with respect to physical, rheological, and mechanical properties. It is important to have detailed knowledge about the sequence and distributions of EOx and PO blocks, the degree of polymerisation and block size, the identity and structure of particular end groups (initiator and terminator type), and impurities. Detailed reviews of analytical approaches for studying poly(oxyalkylene) non-ionic surfactants have been compiled by Kalinoski [3], Chu and Zhou [4, 5] and Schmitt [6]. These methods include liquid chromatography, infrared (IR) spectroscopy, Raman spectroscopy, viscosimetry, calorimetry, and nuclear magnetic resonance (NMR) spectroscopy. In general, these methods yield only an average distribution of speciﬁc structural features, and are not suitable for characterisation of the individual components of molecular weight distributions. In recent years, it has been demonstrated that mass spectrometric methods in combination with soft ionisation techniques, which minimise fragmentation during ionisation and thus produce (pseudo)molecular ions, can provide this information [7]. Within this range of techniques, especially matrix-assisted laser desorption/ ionisation (MALDI) mass spectrometry has proved its value in providing accurate and detailed molecular weight data on high molecular weight materials [8, 9]. Since its introduction by Tanaka and co-workers [10] and by Karas and Hillenkamp [11], the technique has evolved as one of the most successful volatilisation and ionisation methods for a wide variety of molecules, such as peptides, proteins, oligosaccharides,

395

Introduction to Polymer Analysis and synthetic polymers. Several reviews illustrate the continuously expanding scope of MALDI applications [8, 12–14]. In many cases, MALDI is coupled to relatively inexpensive time-of-ﬂight (ToF) mass spectrometers. These systems provide a large mass range over which samples can be analysed, with high sensitivity, a mass accuracy of typically 0.1%, and oligomeric mass resolution [15]. For example, MALDI-ToF has been successfully used for obtaining molecular weights of segments of synthetic copolymers and weight averages of the molecular weight distributions [16], and for obtaining the detailed description of the individual components of triblock copolymers of polystyrene-block-poly(Amethylstyrene) by Wilczek-Vera and co-workers [17]. Recent introduction of delayed extraction has extended the performance of MALDIToF systems to isotopic resolved measurements for masses up to ~4000 amu with mass accuracies of ~15 ppm [18]. For copolymers exhibiting frequent overlap of peaks corresponding to different copolymer compositions, even higher mass resolution is required to differentiate between them. In these cases, the mass resolution of MALDITOF systems (even with delayed extraction) is insufﬁcient, and Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometry becomes imperative. Using the standard resolution in broadband mode of this instrument, isotopically resolved, singly charged polymers in the mass range up to m/z ~4500 with mass accuracies >20 ppm are measured routinely [19–23]. Changing to the heterodyne (high-resolution) mode increases performance easily to mass resolutions >1,000,000 and mass accuracies <1 ppm [24]. Such mass accuracies are frequently sufﬁcient for exact elemental determination of the repeating units, terminal groups, and adducts [23]. Determination of accurate molecular weight distributions by MALDI-FT-ICR-MS is complicated by effects leading to a mass-dependent distortion of the measured distributions. The MALDI process leads to ions with a broad, mass-independent velocity distribution [25]. Consequently, the various components in a polymer sample enter the trap of the FT-ICR-MS with different mass-dependent kinetic energies. Because the trapping efﬁciency is highly dependent on the kinetic energy of the ions, this will lead to mass discriminations in MALDI experiments [26, 27]. If MALDI experiments are carried out in an internal source geometry, trapping of higher kinetic energy ions is provided by gated deceleration potentials [28–30]. Because MALDI ions with higher masses have higher kinetic energies than lighter ones, it takes longer gated trapping times to slow down these ions. Duration of the gated deceleration potentials therefore determines the mass range over which ions will be efﬁciently trapped. For external-source MALDI, the efﬁciently trapped mass range is also gate time-dependent. Here, the post-source acceleration during the transport of the ions to the ion cyclotron resonance (ICR) cell induces a mass-dependent ﬂight time between ion formation and trapping and results in a ﬂight-time-induced discrimination [31, 32].

396

Block Copolymers Van Rooij and co-workers [33] demonstrated the utility of MALDI carried out on an external source. Fourier transform ion cyclotron resonance mass spectrometer (FT-ICR-MS) is a method for the treatment of molecular weight distributions measured by MALDI-FT-ICR-MS to yield the individual block length distributions in a complex pluronic sample. The constituent monomer units in this block copolymer are small relative to the overall size of the copolymer, which means that overlap between different components of the molecular weight distribution is likely, and high-resolution mass spectrometry is imperative. Pluronic sample of the type (C2H4O)x – (OH3H6)y – (C2H4)3–H offers a unique possibility to test the corrections on distortions induced by the experimental technique. The block lengths of both constituents are expected to be uncorrelated (random coupling hypothesis). An equal distribution in the block length of one of the constituents for different block lengths of the other constituent would conﬁrm the validity of the corrections. van Rooij and co-workers [33] described a method for the treatment of molecular weight distributions measured by MALDI on an external ion source FT-ICR-MS to yield the actual molecular weight distribution and, from that, the individual block length distributions. Detailed and accurate molecular weight data were obtained on a complex sample using this methodology, which independently validates the data provided by the manufacturer. The experimentally veriﬁed random coupling hypothesis proves the validity of the methodology. The weight average number of propylene oxide units for the propylene oxide units (i.e., Mw-PO) obtained by this method was found to be 16.55, which was in good agreement with value of 16.4 supplied by the polymer manufacturer. Frisch and Xu [34] used 1H-NMR and 13C-NMR to study the effect of solution radical polymerisation of styrene with methacrylic acid (MA) in the presence of a homogeneous, high molecular weight (Mw = 2 × 105), poly(2-vinylpyridine) template. The presence of the template had little, if any, effect on the molecular weight, composition, or glass transition temperatures of the copolymers but produced copolymers with signiﬁcantly longer block lengths of styrene and methacrylic acid repeat units. Pyrograms will differentiate between random copolymers and block polymers or polymer mixtures [35–38]. Presence of foreign monomer may interrupt chain-transfer processes involved in the degradation. Similar products result, but their quantities as determined from peak heights are different. Voigt demonstrated that even for closely related polyoleﬁns, the programs will distinguish between poly(ethylene-propylene) block and random copolymers of the same composition [36]. Monge and Haddleston [39] reported on online NMR in the analysis of poly(nhydroxy-succinamide methacrylate)-6 poly(methyl methacrylate AB) type block copolymers.

397

Introduction to Polymer Analysis

References 1. R. Prior in Surfactants in Consumer Products: Theory, Technology and Application, Ed., J. Falbe, Springer-Verlag, Heidelberg, 1987, p.5. 2. A.J. Ryan and J.L. Stanford in Comprehensive Polymer Science, Eds., G. Allen and J.C. Bevington, Pergamon Press, Oxford, UK, 1989, p.427. 3. H.T. Kalinoski in Nonionic Surfactants Polyoxyalkylene Block Copolymers, Ed., V.M. Nace, Marcel Dekker, New York, NY, USA, 1996, p.31. 4. B. Chu, Langmuir, 1995, 11, 414. 5. B. Chu and Z. Zhou in Nonionic Surfactants Polyoxyalkylene Block Copolymers, Ed., V.M. Nace, Marcel Dekker, New York, NY, USA, 1996, p.67. 6. Nonionic Surfactants, Ed., T.M. Schmitt, Surfactant Science Series 40, Marcel Dekker, New York, NY, USA, 1992. 7. L.M. Nuwaysir, C.L. Wilkins and W.J. Simonsick, Jr., Journal of the American Society for Mass Spectrometry, 1990, 1, 1, 66. 8. A.M. Belu, J.M. DeSimone, R.M. Linton, G.W. Lange and R.M. Friedman, Journal of the American Society for Mass Spectrometry, 1995, 7, 1, 11. 9. U. Bahr, A. Deppe, M. Karas, F. Hillenkamp and U. Giessman, Analytical Chemistry, 1992, 64, 22, 2866. 10. K. Tanaka, H. Waki, Y. Ido, S. Akita, Y. Yoshida, T. Yoshida and T. Matsuo, Rapid Communications in Mass Spectrometry, 1988, 2, 8, 151. 11. M. Karas and F. Hillenkamp, Analytical Chemistry, 1988, 60, 20, 2299. 12. F. Hillenkamp, M. Karas, R.C. Beavis and B.T. Chait, Analytical Chemistry, 1991, 63, 24, 1193A. 13. M. Karas, U. Bahr and U. Giessmann, Mass Spectrometry Reviews, 1991, 10, 5, 335. 14. M.V. Buchanan and R.L. Hettich, Analytical Chemistry, 1993, 65, 5, 245A. 15. D.C. Schriemer and L. Li, Analytical Chemistry, 1996, 68, 17, 2721. 16. P.O. Danis and F.J. Huby, Journal of the American Society of Mass Spectrometry, 1995, 6, 11, 1112. 398

Block Copolymers 17. G. Wilczek-Vera, P.O. Danis and A. Eisenberg, Macromolecules, 1996, 29, 11, 4036. 18. R.D. Edmondson and D.H. Russell, Journal of the American Society of Mass Spectrometry, 1996, 7, 10, 995. 19. M.L. Gross and D.L. Rempel, Science, 1984, 226, 261. 20. F.W. McLafferty, Accounts of Chemical Research, 1994, 27, 11, 379. 21. P.B. O’Connor and F.W. McLafferty, Journal of the American Chemical Society, 1996, 117, 51, 12826. 22. G.J. van Rooij, M.C. Duursma, R.M.A. Heeren, J.J. Boon and C.G. de Koster, Journal of the American Society of Mass Spectrometry, 1996, 117, 440. 23. E.R.E. van der Hage, M.C. Duursma, R.M.A. Heeren, J.J. Boon, M.W.F. Nielen, A.J.M Weber, C.G. de Koster and N.K. de Vries, Macromolecules, 1997, 30, 15, 4302. 24. G.M. Alber, A.G. Marshall, N.C. Hill, L. Schweikhard and T. Ricca, Review of Scientiﬁc Instruments, 1993, 64, 7, 1845. 25. R.C. Beavis and B.T. Chait, Chemical Physics Letters, 1991, 181, 5, 479. 26. J.D. Hogan and D.A. Laude, Jr., Analytical Chemistry, 1992, 64, 7, 763. 27. S.A. Hofstadler, S.C. Beu and D.A. Laude, Jr., Analytical Chemistry, 1993, 65, 3, 312. 28. M. Dey, J.A. Castoro and C.L. Wilkins, Analytical Chemistry, 1995, 67, 9, 1575. 29. M.L. Easterling, C.C. Pitsenberger, S.S. Kulkarni, P.K. Taylor and I.J. Amster, International Journal of Mass Spectrometry and Ion Processes, 1996, 157158, 97. 30. S.J. Pastor and C.L. Wilkins, Journal of the American Society for Mass Spectrometry, 1997, 8, 3, 225. 31. F.M. White, J.A. Marto and A.G. Marshall, Rapid Communications in Mass Spectrometry, 1996, 10, 14, 1845. 32. P.B. O’Connor, M.C. Duursma, G.J. van Rooij, R.M.A. Heeren and J.J. Boon, Analytical Chemistry, 1997, 69, 14, 2751.

399

Introduction to Polymer Analysis 33. G.J. van Rooij, M.C. Duursma, C.G. de Koster, R.M.A. Heeren, J.J. Boon, P.J.W. Schuyl and E.R.E. van der Hage, Analytical Chemistry, 1998, 70, 5, 843. 34. H.L. Frisch and X. Qihua, Macromolecules, 1992, 25, 20, 5145. 35. A.K. Barlow, R.S. Lehrler and J.C. Robb, Polymer, 1961, 2, 27. 36. J. Voigt, Kunststoffe, 1964, 54, 2. 37. J. Strassburger, G.M. Brauer, M. Tryon and A.F. Forziati, Analytical Chemistry, 1960, 32, 4, 454. 38. K.J. Bombaugh, C.E. Cook and B.H. Clampitt, Analytical Chemistry, 1963, 35, 12, 1834. 39. S. Monge and D.M. Haddleton, European Polymer Journal, 2004, 40, 1, 37.

400

A

bbreviations

13

Carbon-13 nuclear magnetic resonance

1D

One-dimensional

1D-NMR

One dimensional-NMR

2D

Two-dimensional

2D-NMR

Two-dimensional-nuclear magnetic resonance

2M

2-Methyldecane

3D

Three-dimensional

3D-NMR

Three-dimensional-nuclear magnetic resonance

3E

3-Ethylnonane

4E

4-Ethylnonane

5M

5-Methyldecane

A/B

Acrylonitrile–butyl acrylate

AA

Acetic anhydride

AAS

Atomic absorption spectrometry

ABTMP

2,2´-Azobis(2,4,4-trimethylpentane)

AC

Alternating current

ACN

Acrylonitrile

AFM

Atomic force microscopy

AFS

Atomic ﬂuorescence spectrometry

AIBN

Azo bis(isobutyronitrile)

AM

Alfrey–Mayo

AMMS

Anodic micromembrane suppression

amu

Atomic mass units

AN

Acrylonitrile

aPS

Atactic polystyrene

BA

N-butyl acrylate

BD

Butadiene

C-NMR

401

Introduction to Polymer Analysis bp

Boiling point

BPO

Benzoyl peroxide

BZ

Benzene

CCl4

Carbon tetrachloride

CDCl3

Deuterated chloroform

CI

Chemical ionisation

CID

Collision-induced dissociation

CMA

Acryloyl oxyethyltrimethyl ammonium chloride

CMM

Methacryloyl oxyethyltrimethyl ammonium chloride

CMMS

Cathodic micromembrane suppression

CP-MAS

Cross polarisation – magic angle spinning

CV

Coefﬁcient of variance

CVAAS

Cold-vapour atomic absorption spectrometry

DCT

Direct current transducer

DMSO

Dimethyl sulfoxide

DMSO-D6

Deuterated dimethyl sulfoxide

DN

Dinitriles

DO

Polyoxymethylene–1,3-dioxolane

DSC

Differential scanning calorimetry

DVB

Divinyl benzene

E/P

Ethylene/propylene

E4MP

Ethylene-4-methylpentene-1

EB

Ethylene-butene

ECH

Epichlorohydrin

ECN

Effective carbon number

EDAX

Energy dispersive analysis using X-rays

EDXRF

Energy dispersive XRF

EH

Ethylene-1-hexene

EHP

Ethylene heptene

EHX

Ethylene hexene

EI

Electron ionisation

ENB

Ethylidene norbornene

EO

Ethylene-1-octene

EOx

Ethylene oxide

402

Abbreviations EPDM

Ethylene-propylene-diene terpolymer

EPE

Ethyelene-propylene-ethylene

EPM

Ethylene-propylene monomer

ESCA

Electron spectroscopy for chemical analysis

ESR

Electron spin resonance

FASA

Mole fraction of acrylate–styrene–acrylate units

FID

Flame ionisation detection

FPD

Flame photometric detector

FSSS

Fractions of styrene triplets

FT

Fourier transform

FT-ICR

Fourier transform - ion cyclotron resonance

FT-IR

Fourier-transform - infrared spectroscopy

FT-NMR

Fourier-transform - nuclear magnetic resonance

GC

Gas chromatography

GC-FID

Gas chromatography with ﬂame ionisation detection

GC-MS

Gas chromatography – mass spectrometry

GFAAS

Graphite furnace atomic absorption spectrometry

GPC

Gel permeation chromatography

HC

Hydrocarbons

HDPE

High-density polyethylene

HFP

Hexaﬂuropropylene

HIPS

High-impact polystyrene

HPLC

High-pressure liquid chromatography

ICP-AES

Inductively coupled plasma atomic emission spectrometry

ICP-MS

Inductively coupled plasma mass spectrometer

ICP-OES

Inductively coupled plasma optical emission spectrometer

ICR

Ion cyclotron resonance

iPP

Isotactic polypropylene

iPS

Isotactic polystyrene

IR

Infrared

KBr

Potassium bromide

KOH

Potassium hydroxide

LAH

Lithium aluminium anhydride

LDPE

Low-density polyethylene(s)

403

Introduction to Polymer Analysis LE

Linear oleﬁn peaks

M

Methyl methacrylate

MA

Methacrylic acid

MALDI

Matrix-assisted laser desorption/ionisation

MALDI-MS

Matrix-assisted laser desorption/ionisation - mass spectrometry

MALDI-ToF

Matrix-assisted laser desorption/ionisation - time-of-ﬂight

MCA

Mercaptoacetic acid

MMA

Methyl methacrylate

MN

Mononitriles

Mn

Number average molecular weight

MPA

Mercaptopropionic acid

MS

Mass spectroscopy

MS-MS

Tandem mass spectrometry

Mw

Molecular weight

NAA

Neutron activation analysis

NBR

Acrylonitrile–butadiene copolymers

NMR

Nuclear magnetic resonance

NO

Nitric oxide

NO2

Nitrogen dioxide

NOE

Nuclear Overhauser effect(s)

NPD

Nitrogen-phosphorus detector

PA

Phthalic anhydride

PAA

Polyacrylic acid

PAR

Pyridyl azorescorcinol

PBA

Polybutylacrylate

PBD

Polybutadiene

PBMA

Poly(n-butyl methacrylate)

PBTP

Polybutylene terephalate

PBTP-PPG

Polybutylene terephalate - polypropylene glycol

PCFE

Poly(1-chloro-1-ﬂuoro-ethylene)

PDMS

Polydimethylsiloxane

PE

Polyethylene

PECH

Polyepichlorohydrin

PEE

Propylene-ethylene-ethylene

404

Abbreviations

PEEA

Poly(ethylene–ethyl acrylate)

PEEP

Propylene-ethylene-ethylene-propylene

PEG

Polyethylene glycol(s)

PET

Polyethylene terephthalate

PEVA

Poly(ethylene–vinyl acrylate)

PGC

Pyrolysis gas chromatography

PHGC

Pyrolysis–hydrogenation gas chromatography

PIB

Polyisobutylene

PMMA

Polymethylmethacrylate

PMR

Proton magnetic resonance

PO

Propylene oxide

POM-EO

Polyoxymethylene - ethylene oxide

PP

Polypropylene

ppm

Parts per million

PPO

Polypropylene oxide

Pr

Propyl

PS

Polystyrene(s)

PTFE

Polytetraﬂuoroethylene

PU

Polyurethane

PVC

Polyvinyl chloride

Py-GC

Pyrolysis-gas chromatography

Py-GC-MS

Pyrolysis-gas chromatography – mass spectrometry

S

Styrene

S/N

Signal-to-noise

S-AA

Styrene-acrylic acid

SCB

Short chain branch

SCOT

Solid capillary open tubular

SD

Standard deviation

SEC

Size exclusion chromatography

SEC-MALLS

Size exclusion chromatography off line low angle and multi angle light scattering

SEM

Scanning electron microscopy

SIMS

Secondary ion mass spectrometry

sPP

Syndiotactic polypropylene

405

Introduction to Polymer Analysis sPS

Syndiotactic polystyrene

SQW

Square wave

STY

Styrene

T

Toluene

TC

Thermal conductivity detector

TFA

Triﬂuoroacetic acid

TFAA

Triﬂuoroacetic anhydride

TGA

Thermogravimetric analysis

THCSCH

Homonuclear spectroscopy

TMS

Tetramethylsilane

TOC

Total organic carbon

ToF

Time-of-ﬂight

ToF-SIMS

Time-of-ﬂight secondary ion mass spectrometry

TPI

Trans-1,4-polyisoprene

TRIM

Poly(trimethylolpropane trimethacrylate)

TRXRF

Total reﬂection X-ray ﬂuorescence

UV

Ultraviolet

V8HQ

Vanadium 8-hydroxyquinolinate

VC

Vinyl chloride

VCH

Vinyl cyclohexane

VDC

Vinylidene chloride

VF2

Vinyldene diﬂuoride

WDXRF

Wavelength dispersive nuclear magnetic resonance

XPS

X-ray photoelectron spectroscopy

XRF

X-ray ﬂuorescence

XRFS

X-ray ﬂuorescence spectroscopy

ZAAS

Zeeman atomic absorption spectrometry

406

I

ndex

A Abstraction reaction 277 Acetylation 51-52, 56, 60 Acid-catalysed 53 Acrylamide-methacryloyl oxy-ammonium chloride 133 Acrylonitrile-butadiene rubbers 180 Acrylonitrile-butadiene copolymers 186 Acrylonitrile-butyl acrylate copolymer 177 Acrylonitrile monomer 181, 186 Alcohol exchange - gas chromatography 76 Alfrey-Mayo statistics - kinetics model 174-175 A-oleﬁn copolymers 176 Analyser, halogen-sulfur 14 Analyser, trace sulfur 10 Annealing 227-228 Ashing procedure 32 Atomic absorption spectrometer, single-beam 20 Atomic absorption spectrometer, double-beam 20 Atomic force microscopy 51 Atomisation 19, 21 Average sequence length method 148 Azobis(2,4,4-trimethylpentane) initiator 294

B Balata 324 Beer’s law 29 Benzol peroxide polymerisation initiator 287 Bernoullian propagation statistics 232-233 Block copolymers 395 Bromination methods 333 Coulometric 86 Butadiene monomer 186 Butadiene-acrylonitrile-methacrylic acid-terpolymer 187 Butadiene-propylene copolymers 203-204, 342

407

Introduction to Polymer Analysis Butyl rubber 100

C C2U symmetry 230 Calorimetry 395 Carbon cross polarisation – magic angle spinning - nuclear magnetic resonance 336 13 Carbon-nuclear magnetic resonance spectroscopy 10, 129, 177, 216, 253, 317320, 325, 366, 378-379, 382 Chain transfer reagents 275 Chemical ionisation 137, 237, 241, 244 Chemiluminescence detection 12 Chicle 324 Chlorobutyl rubber 8 Collision induced dissociation 278, 284 Combustion microcoulometric technology 14 Copolymerisation 157, 282 Copolymers, unsaturation of 332 Copper calibration curves 24 Coulometry, constant-current 86 Cross polarisation – magic angle spinning 13C-Nuclear magnetic resonance 333, 335 Crystallisation 170, 227-228 Curve resolution method 232

D Derivatisation methods 75 Diastereoisomeric tetramers 237 Differential pulse-stripping voltammetry 37 Differential scanning calorimetry 370 Differential thermal analysis 130, 221 Digestion technique 35 Dionex instrument 15-16 Direct current integration 37 Direct ﬂame atomisation 21 Direct injection enthalpimetry 60-61 Direct injection techniques 20 Dithiocarbamate precipitation methods 43 Dodecane thiol chain transfer agent 287 Double-beam instrument 19 Dye interaction technique 281 Dye partition method 281-282, 299-300

408

Index

E Energy dispersive analysis using X-rays 51 Energy dispersive X-ray ﬂuorescence 40-41, 43 Electron ionisation 137, 237, 244 Electron spectroscopy for chemical analysis 200 Electron spin resonance spectra 331-332, 336 Epoxides, terminal 300 Esteriﬁcation 84 Ethylene-A-oleﬁn 372-373 Ethylene copolymer 371, 367 Ethylene, short chain branching of 361 Ethylene-propylene copolymers, branching of 358, 386 Ethylene-propylene copolymer 142, 214, 219, 258, 359-361, 386 Ethylene-propylene block copolymers 220 Ethylene-propylene diene terpolymer 96, 181-185 Ethylene-propylene-ethylidene-norbornene rubber 182 Ethylene-propylene monomer 182

F Flame ionisation detection 155, 192, 237, 281, 289-290 Flame photometric ashing procedure 44 Flame photometric detection 287, 289 Flame photometry 32 Flash combustion 14 Fourier-transform mass spectrometer 278 Fourier-transform infrared spectroscopy 51, 120, 106, 171 Fourier-transform ion cyclotron resonance mass spectrometry 304, 396-397 Fourier-transform raman spectroscopy 317 Fourier-transform nuclear magnetic resonance method 366 Fragmentation mechanism 200 Furnace atomisation 21 Furnace combustion system 8

G Gas chromatography 69-70, 72, 76, 81-83, 101-102, 106, 117, 135-136, 139, 268, 276, 336, 375 Chemical reaction 51 Cleavage 76, 81 Derivatisation 97 Gas chromatography – mass spectrometry 276, 289-290, 294, 296 Gas-ﬂow geometry 24 Gas-liquid chromatography 386 Gel permeation chromatography 106, 159, 203, 205, 388

409

Introduction to Polymer Analysis Curve 164-165 Matrix-assisted laser desorption/ionisation - time-of-ﬂight mass spectrometry 284

Gelation 332 Glass capillary pyrolysis - hydrogenation gas chromatography 378 Graft copolymers 215 Gutta percha 324

H Halogens, determination of 8 Alkali fusion methods 9 Combustion methods 8 Oxygen ﬂask combustion 8, 11 Physical methods for determining halogen 10 Hexaﬂuoropropylene-vinylidene ﬂuoride copolymer 132 High-density polyethylene 282, 386 Higher oleﬁn copolymers, short chain branching of 361 1 H-nuclear magnetic resonance spectroscopy 130, 279, 284, 334 1 H-nuclear magnetic resonance - Fourier-transform infrared spectroscopy 272 Homonuclear spectroscopy 271-272 Homopolymers, unsaturation of 313 Hydride generator system, analytical 21 Hydriodic acid hydrolysis technique 70 Hydriodic acid reduction - gas chromatography 70 Hydroboration 286 Hydrogenated natural rubber 254, 258 Hydrogenation - gas chromatographic technique 97 Hydrogenation, in-line 135 Hydrogeneated acrylonitrile-butadiene copolymers 185 Hydrolysis 69, 84, 87, 98, 170, 280 Hydrolysis - gas chromatography 198

I Inductively coupled plasma spectrometer 21, 25 Infrared calibration curve 121 Infrared spectrometer 71, 122 Infrared spectroscopy 51, 61, 67, 71, 74-75, 78-79, 84, 90-95, 105, 117, 120-121, 123-125, 135, 141,168, 180-181, 185, 221-222, 230, 258, 300, 314, 321322, 324-325, 329-330, 335-336, 340, 353-355, 374, 395 Non-dispersive 14 Styrene-divinyl benzene 332 Initiators 292 Carbon-labeled 278 Ion cyclotron resonance 396 Isobutyronitrile 333

410

Index Isomerisation 241, 277, 279 Isomerism, geometrical 253 Isotacticity 74, 211, 213, 222-223, 225

K Kjeldahl digestion 12 Micro techniques 7

L Linear oleﬁns 183 Liquid chromatography 395 High-pressure 16, 79, 205 Liquid-liquid phase separation 387 Lithium catalysts 324 Low-angle laser light scattering measurements 362 Low-density polyethylene 382, 386

M Macrotacticity 222 Mass spectrometry, secondary ion 147, 198 Mass spectroscopy, tandem 51 Matrix-assisted laser desorption/ionisation 278, 304, 396-397 Mass spectrometry 51, 395 Time-of-ﬂight 284, 396 Time-of-ﬂight - mass spectrometry 299 Methyl branching in polyethylene 354 Microcombustion techniques 12 Micro-ozonolysis 338 Micro-Parr bomb 9 Microthermal analysis 51 Microzonolysis 204, 342

N Natural rubber 138, 142, 324 Near-Fourier-transform infrared spectroscopy 300 Near-infrared spectroscopy 13, 106, 300-301 Nessler’s reagent 75 Nitrogen 11 Combustion methods 11 Nitrogen analyser 12 Nitrogen-phosphorus detector 287 Non-linear least square errors-in-variables model 177 Nuclear magnetic resonance, ﬂuorine 62,132-133, 272, 369-370 Nuclear magnetic resonance spectroscopy 74, 79, 84, 93-94, 98, 101, 117, 121,

411

Introduction to Polymer Analysis 124, 147, 168, 174, 180, 199, 278, 301, 317, 319, 325, 339, 341, 354-355, 395 Wavelength dispersive 41 Nuclear Overhauser effect 216, 317, 356, 360, 366

O Oleﬁn copolymers 141 Oleﬁnic resonance 94 Oleﬁn-terminated polyisobutylene 286 Oligomer 154, 171, 192 Optical emission spectroscopy 7 Oxidative degradation technique 75 Oxirane rings 101 Oxonolysis-gel permeation chromatography 166 Oxygen ﬂask combustion - ion chromatography 15 Ozonisation 161, 163, 168 Ozonisation - gel permeation chromatography 170 Ozonisation - high performance liquid chromatography 205 Ozonolysis 162, 164-165, 167-168, 170, 203, 205, 268, 337, 339-341 Gel permeation chromatography method 163

P Parr bomb 13, 34 Phenyl isocyanate method 53, 56 Phosphorus 11 Acid digestion 11 Phthalation 51-53, 56 Pluronics 395 Proton magnetic resonance spectroscopy 79, 220, 303 Poly(amide-imides) 104 Poly(N-butyl methacrylate), tacticity 233 Poly(1-chloro-ﬂuoroethylene), tacticity 242 Poly(2,6-dimethyl 1,4 phenylene oxide) 301 Poly(methylmethacrylate)s 298 Poly(oxyalkylene) block copolymers 395 Poly(trimethylolpropane trimethacrylate) 335 Polyamides 2-3, 104, 159 Polybutadiene 161-163, 166-168, 190, 260 Micro-ozonolysis of 169 Unsaturation in 313, 319 Polybutene-2-ethylene copolymer 259 Polychromator system 27 Polycondensation 2, 4 Polydimethylsiloxane homopolymer 200-201

412

Index Polydimethylsiloxane - urethane-segmented copolymers 199 Polyepichlorohydrin 269-272 Polyester 2, 16, 57, 60, 98 Polyethylene 282-283, 326, 354-355, 362, 364-366, 369, 372, 376-377, 385, 387 Unsaturation in 325 Polyethylene glycols 278 Polyethylene matrix 42 Polyethylene terephthalate 283-284 Polyimides 104 Polyisobutylene 284 Hydroxyl-terminated 286 Tert-chlorine terminated 285 Polyisoprene 153, 161, 168, 262, 323 Unsaturation in 321 Polymer additive system 7 Polymer, amorphous 218 Polymer analysis 7, 117 Polymer, crystalline 218 Polymer pyrolysates 142 Polymers, isotactic 211, 223 Polymerisation 2, 4-5, 8, 75, 93, 147-148, 157, 162, 187, 189, 211, 232, 269-270, 275, 279, 287, 294, 336, 385, 388 Anionic 276 Catalysts 31 Cationic 278 Degree of 395 Free radical 299 Radical 287 Ring-opening 270 Polymethacrylonitrile polymer 172 Polymethylmethacrylate 241, 289-290, 292-296, 298-299 Polyoleﬁns 8, 11, 16, 33, 42-43, 135, 152 Polyols 164-166 Polyoxymethylene-1,3,6-trioxocane copolymer 196, 197 Polyoxymethylene-1,3-dioxolane copolymer 196, 197 Polypropylene 253, 256, 353, 356, 361 Unsaturation in 329 Polypropylene, amorphous 254, 358 Polypropylene, crystalline 358-359 Polypropylene, isotactic 213 Polypropylenes, non-crystalline 357 Polypropylene oxide resin 302-303 Polypropylene, syndiotactic 213 Polypropylene, tacticity 212

413

Introduction to Polymer Analysis Polystyrene 160, 229, 278, 280-281, 303 Syndiotactic 227 Syndiotactic, tacticity of 226 Polytetraﬂuoroethylene 9 Polyurethane-polydimethylsiloxane copolymer 200-202 Polyvinyl chloride 8, 154, 277, 369 Tacticity of 230 Polyvinylidene chloride 272 Polyvinylidene ﬂuoride 272 Potassium bromide disc 173-174 Potentiometric titration 69 Propylene-1-ethylene copolymer 257 Proton abstraction reaction 276 Proton magnetic resonance spectroscopy 51, 79, 220, 303 Proton nuclear magnetic resonance spectroscopy 215, 295, 317, 353 Pyrogram 184-185, 237, 241, 281, 292, 297 Pyrolysis 12, 73, 83, 117, 130, 133-140, 142, 152, 154-156, 189, 191-192, 194, 196, 198, 240 Pyrolysis, high-vacuum 386 Pyrolysis - gas chromatography 51 73-74, 81, 83-84, 92, 97, 129, 133, 135, 137, 139, 147, 151-153, 156, 158, 180-181, 185-186, 190, 194-196, 198-199, 276, 280, 287, 294, 296, 298, 336, 386 Carbon number distribution 378-379 Mass spectrometry 51, 97, 230, 238, 294 Pyrolysis - hydrogenation - gas chromatography 135, 375, 378, 382, 379, 384 Pyrolysis infrared spectroscopy 141 Pyrolysis-nuclear magnetic resonance spectroscopy 51

Q Quadrupole mass spectrometer 29

R Radiochemical method 100, 118 Radiolysis 372-374 Raman spectroscopy 93, 335, 395 Reaction gas chromatography, alkali fusion 100, 102, 104-105 Regioisomerism 213, 261-264, 272 Regression method 304 Resins, synthetic 1 Thermoplastic 1-2 Thermosetting 1-2 Rubber, synthetic 5

414

Index

S Saponiﬁcation 68-69 Scalar coupling 237 Scanning electron microscopy 51 Schoniger oxygen ﬂask combustion technique 8 Silanisation 59 Silica-nuclear magnetic resonance 272 Single-beam ﬂame spectrometer 19 Size exclusion chromatography 13, 287, 294, 304, 362, 387-388 Low-angle light scattering 13 Multi-angle light scattering 387 Sodium tetrahydroborate system 21 Spectroﬂuorimetric method 102 Spectrophotometer, grating 9 Spin-spin coupling 218 Stereoisomerism 253 Styrene butadiene copolymer 5, 203 Styrene methacrylonitrile copolymer 172 Styrene-butadiene rubber 89, 142 Styrene-butylacrylate copolymer 194 Styrene-methyl methacrylate copolymer 137 Sulfur 10 Combustion methods 10 Sodium peroxide fusion 10 Polymers, synthetic 304

T Tacticity 147, 162, 212, 221, 223, 226, 230, 233, 242 Thermal annealing 226 Thermal dehydrochlorination 155 Thermogravimetric analysis 18 Thermometric techniques 272 Three-dimensional nuclear magnetic resonance 281 Time-of-ﬂight mass spectrometers 278, 396 Time-of-ﬂight - secondary ion mass spectrometry 51, 159-161, 199, 201-202 Titration 64, 85, 88 Curve 65 P-Toluene sulfonic acid method 52 Total organic carbon 8, 13 Trace metal analysis 18, 31 Destructive techniques 18 Anodic scanning voltammetric techniques 18 Atomic absorption spectrometry 18, 21, 23, 32-34, 36-37, 39, 43

415

Introduction to Polymer Analysis Atom trapping technique 21 Cold-vapour atomic absorption spectrometry 18 Graphite furnace atomic absorption spectrometry 18-20, 22, 36, 39 Ion chromatography 9, 16, 37-38, 40 Inductively coupled plasma atomic emission spectrometer 18, 21, 27-29, 34, 36-37 Inductively coupled plasma optical emission spectrometry - mass spectrometry 21, 29-30 Monochromator system 19, 25-27 Pre-concentration atomic absorption spectrometry techniques 30 Pressure dissolution technique 34 Polarography 36-37 Ultraviolet spectroscopic method 36, 284 Vaccum monochromator design 25 Vapour generation atomic absorption spectrometry 21 Visible spectrometry 18 Voltammetry 36 Zeeman AAS 18, 22 Non-destructive methods 18, 40 X-ray ﬂuorescence spectrometry 7, 15-16, 18, 40, 42-43, 51, 58, Neutron activation analysis 32, 43-44

Nuclear magnetic resonance, two-dimensional method 271

U Rubber, ungrafted 93

V Vacuum photolysis 75 Vacuum radiolysis 371 Vacuum ultraviolet monochromator 26 Vanadium 8-hydroxyquinolinate method 60 Vinyl acetate-methylacrylate 132 Vinyl chloride-vinyl acetate copolymer 123 Vinylidene chloride/vinyl chloride copolymer 138, 140, 156-158 Vinylidene chloride-methacrylonitrile 177 Vinylidene chloride-vinyl chloride-copolymer 154 Vinylidene-cyanovinyl acetate copolymers 177 Viscometry 388, 395 Viscosity measurements, intrinsic 362 Visible spectrophotometry 36 Volatilisation 36, 97, 395

X X-ray photoelectron spectroscopy 51

416

Index X-ray diffraction 221, 354 X-ray emission analysis 10 X-ray ﬂuorescence, total reﬂection 41-42 X-ray methods 17, 42, 51

Z Zeeman patterns 23 Zeeman spectrometer 24 Zeeman techniques 22 Zeeman effect 23 Zeigler catalysts 324 Zeigler polyethylene pyrogram 375 Zeisel method 77, 81 Ziegler-Natta catalysts 148 Copolymers 332

417

Introduction to Polymer Analysis

418

Published by Smithers Rapra, 2009

The aim of this book is to familiarise the reader with all aspects of plastic analysis, and it covers the analysis of the main types of plastics now in use commercially. Introduction to Polymer Analysis gives an up-to-date and thorough exposition of the present state of the art of polymer analysis and, as such, should be of great interest to all those engaged in this subject in industry, university research establishment 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, the engineering plastics industry, plastic components manufacturers, and those concerned with pharmaceuticals and cosmetics.

Shawbury, Shrewsbury, Shropshire, SY4 4NR, UK Telephone: +44 (0)1939 250383 Fax: +44 (0)1939 251118 Web: www.rapra.net