SECOND EDITION

FUNDAMENTALS OF POLYMER SCIENCE
AN INTRODUCTORY TEXT

SECOND EDITION

FU DAMENT LS POLYMER SCIE CE
AN INTRODUCTORY TEXT

OF

PAUL C. PAINTER MICHAEL M. COLEMAN
The Pennsylvania State University

CRC PRESS
Boca Raton London New York Washingron, D.C.

o

~~ • Area S.B.A. Biblioteca Centrale d~Ingegneria

000298
··:"t .• ,~::i• ~~.

~S~

~,i

~;

',C

DEDICATION

AND ACKNOWLEDGEMENT

Library

of Congress Cataloging-in-PublicatiOli

Data

Main entry under title: Fundamentals of Polymer Science: An Introductory

Text -

Second Edition

Full Catalog record is available from the Library of Congress

This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are" listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use.

This is the book that we swore we would never write after we finished our last book. Bu~now that we have written it and the frustrations have passed, we gratefu~y dedicate the book to our fathers, Charles Painter and Ronald Coleman, whose lives were and much more difficult than ours and who sacrificed many things to give us a start.

Neither this book nor any part may be reproduced or transmitted in any form or by any means, "electronic or mechanical, including photocopying, microfilming, and recording, or by any information storage or retrieval system, without prior permission in writing from the publisher. The consent of CRC Press LLC does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing frpm CRC Press LLC for such copying. Direct all inquiries to CRC Press LLC, 2000 N.W. Corporate 33431. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe. Blvd., Boca Raton, Florida

Visit the CRC Press Web site at www.crcpress.com
© 1997 by CRC Press LLC Originally Published by Technornic Publishing No claim to original U.S. Government works International Standard Book Number 1-56676-559-5 Library of Congress Card Number 97-60515

POLITECNICO BIBL.CENTRALI INGEGNERIA

Printed in the United States of America 2 3 4 5 6 7 8 9 0 STUD. Printed on acid-free paper 0149

INS

298

TABLE OF CONTENTS

Preface to the Second Edition Preface to the First Edition 1. The Nature of Polymeric Materials
A. WHAT ARE POLYMERS-WHAT IS POLYMER SCIENCE? .•••••.•..••..•• B. SOME BASIC DEFINlTIONS- THE ELEMENTS OF POLYMER MICROSTRUCTURE •.•..•..•.•..•..••.•.....•.••..•..•....••.•..•.•.....•....• C. MOLECULAR WEIGHT ---SOME INITIAL OBSERVATIONS •....•..•......• D. STUDY QUESTIONS ..••.....••....•...•.......••......•......•..•..•....•.•.. E. SUGGESTIONS FOR FURTHER READING .•..•..•....••.•..•..•...••..•••.. F. APPENDIX: CHEMICAL STRUCTURES OF SOME COMMON POLYMERS •.•..•....•.........•.•..•..•....••.••..•....•••.••.•..

xi xiii

1 3

16 25 26 26

2. Polymer Synthesis
A. INTRODUCTION ...........•....•......•...•...•.............................. B. STEP-GROWTH POLYMERIZATION ......•.•..........•............•...•••.. C. CHAIN OR ADDmON POLYMERIZATION •....••.......•..•.•........•...•. D. POLYMERIZATION PROCESSES ..•.••.•••.•..•..•.......•..•..•.•..•..•.... E. STUDY QUESTIONS ....•..••.....•...•.••..•.••..•.••....•....••.•..•.••...•• F. SUGGESTIONS FOR FURTHER READING ...•..••...•..•.......•......•....

29
29
31

40
54

60 61

3. Kinetics of Step-Growth and Addition Polymerization
A. INTRODUCTION •.••.•..•.•••...•.••....••..•..•....••..•••....•..........•.•• B. THE KINETICS OF STEP-GROWTH POLYMERIZATION .••................ C. KINETICS OF FREE RADICAL POLYMERIZATION •....•.....•....•..•...• D. STUDY QUESTIONS •.•.......•..•.....•..••....•......•..•...........•......• E. SUGGESTIONS FOR FURTHER READING .•..•..•..........•.......•......•

63
63 64 68 79 81

4. Statistics of Step-Growth Polymerizations
A. INTRODUCTION .•..•.•......•..••.•...•.....•...••..••.•.•..•.•......•...•.•• B. THE STATISTICS OFLINEARPOLYCONDENSATION •.•.•.•..•...•.•.•...

83 83
84 vii

. D.•.•.••••.••••. SUGGESTIONS FOR FURTHER READING •••. STIJDY QUESTIONS ••..•.•.• H....•••..•..•••••..••••.•••••••...•..••••.•••.••••••••.••••.••• 381 G.••.•.••••••••••• E.•..viii Table 0/ Contents Table a/Contents CONSIDERATIONS . CRYSTALLIZATION AND MELTING: SOME EQUILffiRIUM 259 259 260 .••• G..••.••......•.•.••.••.• 394 11.. RANDOM FLIGHTS AND DISORDERED POLYMER CHAINS ••••...••.. Spectroscopy and the Characterization of Chain Structure A IN1RODUCTION .••..••.••.• B.•. SUGGESTIONS FOR FURTHER READING •. F.•.•.. G.•••.•• D..•.••.•.••.•.••..••.••••..•••.•..••• G..••.•••••.••.•.• C..•.••••.•.•••.•.•..•••.••••••.•..••••••••••..••.••••. EXCLUDED VOLUME AND THE THETA TEMPERATURE •••.••..469 471 7..•...•.••.•••••.••.•...•.•..•.••..••••.••••.•••.••. NON-LINEAR MECHANICAL AND RHEOLOGICAL BEHAVIOR . Crystallization..... THE FREE ENERGY OF MIXING •••. SEC AND THE DETERMINATION OF LONG CHAIN BRANCHING •.. STATES OF MATTER AND BONDING IN POLYMER MATERIALS C..•.••.•••.••.••••••.•..••..••••.••.•..••••.•..•.••.. A BRIEF REVIEW OF SOME FUNDAMENTALS ••••.•.••••.••.•.... 107 107 109 113 119 141 143 A IN1RODUCTION ••.•••••.•.•.• .•.••.•.••••••.•.•..••..••.•.••.•.•...••. STIJDY QUESTIONS ...••.•.••••• H..••••.• F...•. SIZE EXCLUSION (OR GEL PERMEATION) CHROMATOGRAPHY •.••.••.••••..•.•.•••••.••••••••.•••.•••. C.•.••••.••.••.••••••.••.•••.•••.•.••.•.•••..••••.•.••.••.•..•••••• D.•..•••••..••.•••.•.•••••.•. STIJDY QUESTIONS .•••••••.••• 208 218 224 237 255 256 257 Index 8. IN1RODUCTION TO VISCOELASTICITY •..•••....••••• D.• B.••.•..•.•••... POLYMER MORPHOLOGY •••••.••.•..• 390 H.•...•.•.••..••.••.••.•..••.•••••••••.••.•••.•.•••. H.• ix 273 278 283 290 303 305 C.•.•••••.•.•.••..••••..••. E.•••.•....•••.•.•••.•.. Molecular Weight and Branching A IN1RODUCTION •.•.•. 145 145 145 150 155 173 180 204 206 WEIGHT AVERAGE MOLECULAR WEIGHT •..•.•• D.....•••••.••..•. THEORY OF GELATION ••...•.•..• ~••••••.••.•..••..•••.. SUGGESTIONS FOR FURTHER READING •...•••••..• E.•.•• D.•... " ••••••. Melting and the Glass Transition A.•.. OSMOTIC PRESSURE AND THE DETERMINATION OF NUMBER AVERAGE MOLECULAR WEIGHT . CHARACTERIZATION OF POLYMERS BY NMR SPEClROSCOPY ..•..•.•••••...•. THE PHASE BEHAVIOR OF POLYMER SOLUTIONS AND BLENDS .•.••.••.••••.••.• B..••••••...••. ULTIMATE PROPERTIES •.••.•.••. Copolymerization AGENERALOVERVIEW •.••. LIGHT SCATTERING AND THE DETERMINATION OF 339 339 341 6.••••••..••.•. OVERVIEW AND GENERAL APPROACH .•••..•.•.•.•.•.•.•••. THE KINETICS OF POLYMER CRYSTALLIZATION ••..•.••.•.•••••••• B.•••.•.••.•.••. STIJDY QUESTIONS •.•.••.••.•. ••••••••••••.•.•.•. 330 335 337 10.•.•.•..•.•.•.••.•. C.••••..•.•.••••.•••••.•••.••..•...•.••••.•.••••.•••.••..••.••.•.••.••.••••••••.•.•.•.•..•.•. STIJDY QUESTIONS ••.•.••. STIJDY QUESTIONS •. DEVIA nONS !'ROM IDEAL BEHAVIOR ..•••.••. SUGGESTIONS FOR FURTHER READING •.••..••.•..••.••••.••.. F.••..•.•. C.. 321 D..••.•.••.•.•••••.••.•••••.•.•.••••.•. 1....•.•.•..••••••.••...•••.• G......••••.•.••.••..•.••• B..• H.•••••••..•.••• 308 C.•.••..•.•••.•.••.••••••..••••••• 207 207 •••...•..••••. REACTIVITY RATIOS AND COPOLYMER COMPOSmON ••.••...•.••• 307 B.••.•.••..••.•.•...••.•. THE CRYSTALLINE MELTING TEMPERATURE ...•.••.••.••.•.••.•.•..416 430 449 459 467 ..•.••.••.•.•.•••••••••••.••.••••..••..•.. BASIC INFRARED SPEClROSCOPY ••••....•••.. Structure A IN1RODUCTION .••. SUGGESTIONS FOR FURTHER READING ••••.•••••• H.•••••.•••..•••.•••••••••• C.•.••..•••.••...••.•.•.. STIJDY QUESTIONS •. SUGGESTIONS FOR FURTHER READING ..••.••.•••••..••••.••...••...•...•.••..••.••.....•••••••••••••..•.•.••••..•••••••.... THE GLASS lRANSmON TEMPERATURE •••••••.••.•.•..•••••.•...•.•.••• F.•..••••.••.•. RANDOM WALKS.••.. SOME FUNDAMENTALS •.•..•.••••••.404 ..••. THECOPOLYMEREQUA..••••...••.•••••••.••.....••.••..•..••.••. : ••.••.•...••.••••... E.•.•..•.•..•. MECHANICAL AND THEORETICAL MODELS OF VISCOELASTIC BEHAVIOR •.•• 352 D.•.•••.•••.•. E.•.•..••..•••.•. BASIC NMR SPEClROSCOPY ..•.•••.••...••.•••••.••••••.••.••.••••..•.•.••••••• F..•.•.•••••.•••.••.•..•.••.•••••..•.••••.••.• E.••••.••..•••••.•..••••.•.•.••.•.••... DILUTE SOLUTIONS.•••..••. Thermodynamics of Polymer Solutions and Blends 307 5.••.•.•••••..•••.•.•.••••••••....•.. SUGGESTIONS FOR FURTHER READING •.••.•.•.•.••.•.••..•.•••..••.•.•.•.••.•.••••.•.••••• E.••••...... 395 395 396 .•.••....••..••. 363 E.•••.••••••.•.••. MOLECULAR WEIGHT DISTRIBUTIONS IN LINEAR CONDENSATION POLYMERS ..•.•.•.•..•. RANDOM BRANCHING WITHOUT NETWORK FORMATION ••..•.••.••.••.•. CHARACTERIZATION OF POLYMERS BY INFRARED SPEClROSCOPY ..••.•.•.••.••.....•..••••...••.•.•.. COPOLYMER SEQUENCE DISTRIBUTION AND THE APPLICATION OF PROBABILITY THEORY •. B..•.•.•••.••....• G.••.••..•.••.••.. •.••. FINAL WORDS-A BRIEF COMMENT ON SIZE ••.•••.•••.•..••.•••••..•••••••.••.•••.•••.. THE CONFORMATIONS OR CONFIGURATIONS OF POLYMER CHAINS . SUGGESTIONS FOR FURTHER READING •..• 373 F.••••..•••• 89 93 96 103 105 106 9.••.••••••.••••••.••.••• ..••.•••••.••..•..•..••.•.•.•.•• F...••. MULTICHAIN CONDENSATION POLYMERS .. STIJDY QUESTIONS •.••••••..••.•.•.••••.••.••••..•••.. SOLUTION VISCOMElRY AND THE VISCOSITY AVERAGE MOLECULAR WEIGHT .•.••••. F.••.•..•..•••...••.••.••.TION .••.. Mechanical and Rheological Properties A IN1RODUCTION AND OVERVIEW .••.•••.••..•. FUNDAMENTALS OF SPEClROSCOPY •.•••••.

when you are confronted by the tedious tasks of preparing an index. can take a lot out of you: Writing a book is an adventure: it begins as an amusement. editing and proof reading. in passing and those who buy books for I" A. in the afternoon I put it back. There are those who notice mistakes the sole purpose of finding them. Dewdney It is particularly galling when your so-called friends participate in the criticism: xi . You don't realize this until later. each and every mistake will be found by your students and the number of revealed errors becomes daunting. particularly a textbook. the sun is shining and you would rather be playing golf. comma. then a master and finally a tyrant. when you start to find some minor typo's or garbled sentences. This is not too bad and you can always blame your publishers: It is with publishers as with wives: one always wants somebody else's. Norman Douglass I I But if your book is used as a text. Churchill Things get particularly hard near the end. especially when you look out the window. All morning I worked on the proof of one of my poems.PREFACE TO THE SECOND EDITION Writing a book. and I took Oscar out a Wilde So it is perhaps inevitable that your attention slips and errors creep in. Even worse. K. Winston S. these errors are often brought to your attention in a manner that you believe reflects a large measure of glee: . of course. then it becomes a mistress.

and so prepare themselves for the main body of the work. Anonymous But in the end a measure of objectivity returns and you start to make a list of corrections. dear reader. Frank Muir (An Irreverent and Thoroughly Incomplete Social History of Almost Everything) In addition. you are then faced with the oppressive but suddenly daunting task of writing a preface. thus enabling them to moderate their enthusiasm. thy wrath can send. You then also notice that there are certain topics that you really made a mess of: I know you believe you understand what you think I said. you are by now tired and fed up with the whole enterprise. modify a little here and there to improve clarity. but I am not sure you realize that what you heard is not what I meant. to one another Stephen Leacock xiii .xii Preface to the Second Edition Of all plagues. This is where you are supposed to explain what the book is about and why you wrote it in the first place. How unfit. But a preface has its uses. Save me. good heaven. an effort that can result in hundreds of pages of text and figures that perhaps (if you are lucky) only a few hundred people might ever read. humiliating than to see idiots succeed in enterprises Gustave we PREFACE TO THE FIRST EDITION Flaubert This base emotional response is quickly replaced by denial: If at first you don't succeed. George Canning Naturally. your first instinct is to attribute their comments to mere envy: Nothing is more have failed in. ever write a book. destroy all evidence that you have tried. you find that if your scribblings have been a collaborative effort you might no longer be on speaking terms with your co-author and one-time friend. you will eventually find that having labored long and hard to write down and give shape to your view of a subject. to undertake Sir Walter Raleigh a February 1997 PAUL PAIN1ER MIKE COLEMAN and you are starting to suspect that your explanation of what you know reveals even more about what you don't know. Vincent with Millay Even worse. lower their sights. I have known two professors of Greek who ceased speaking because of divergent views on the pluperfect subjunctive. will pay for the copies Clarence Darrow If you. a book willfully appears before Edna the populace St. A person his pants who publishes down. Anonymous So you end up writing a second edition. you just hope to get some minor satisfaction from your work: Some day I hope to write a book where the royalties I give away. add some study questions (for those idle teachers who cannot be bothered to make them up!) and end up totally rewriting and reorganizing two chapters. It can give readers a whiff of the author's prose style and an indication of his potential as an inducer of tedium. The problem is that you no longer remember. and how unworthy work of this mixture. But you no longer have the illusion that your efforts will be remotely error-free. oh save me from the candid friend. You hopefully fix all the errors in the first edition. a choice I have made of myself.

it has been our experience that the average student has only imperfectly digested certain fundamentals before starting their studies of polymers. The ion differs from the molecule. and with a wind of aether hlowing at no one knows how many miles a second through every interstice of my body. Don't quote Latin. seniors and some first-year graduate students that have drifted into polymers from other fields. the occurrence would be. indivisible unit of matter. Churchill but in the [mal analysis this is the way we like to write and if you don't like it.XlV Preface to the First Edition Preface to First Edition xv Fortunately. It is distinguished from the corpuscle. A fourth affirms. also the ultimate. if I make the venture one of the flies hits me and gives a boost up again. Of course. Sir Arthur Eddington (The Nature oj the Physical World) Keeping this warning in mind. so in our lectures and in this book we attempt to start at a simple level with brief reviews of what should be (but usually isn't) well-known material. its just that you wanted to emphasize or include subjects that perhaps others found uninteresting or irrelevant. the insides of organs. it is easier for a camel to pass through the eye of a needle that for a scientific man to pass through a door. It is a complicated business. Of course. but it is doubtful if they know any more about the matter than the others. So. not a violation of the laws of Nature. I ought really to look at the problem four-dimensionally as concerning the intersection of my world-line with that of the plank. As with all books of this type. write your own damned book! February 1994 PAUL PAINTER MIKE COLEMAN However. I fall again and am knocked upwards by another fly. Winston S. A fifth theory is held by idiots. Shall I not slip through? No. the plank would be miles away. Organic chemistry is the study of organs. say what you have to say. We naturally suppose that the student has digested some fundamentals of organic chemistry. over golf just because one could Stephen play Leacock What really happens is that if you teach. indivisible unit of matter. I've seen lifelong friends drift apart better. To step on it is like stepping on a swarm of flies. the corpuscular and the atomic. you then find that the books you were using were not so bad after all. Ambrose Bierce (The Devil's Dictionary) and in so doing we have deliberately adopted a conversational tone to our writing. Some might find this irritating and unrigorous. we have to guard against carrying this to extremes and we have tried to overcome a tendency to over-explain. the corpuscle and the atom in that it is an ion. and then sit down. indivisible unit of matter. The plank has no solidity of substance. with Haeckel. by a closer resemblance to the atom. but a rare coincidence. I may hope that the net result will be that I remain about steady. It is an introduction and survey of the subject based on a refinement of what we have taught over the last ten years. you realize that this too will pass and you must save your disagreements for more vital matters. Then again it is necessary to determine in which direction the entropy of the world is increasing in order to make sure that my passage over the threshold is an entrance not an exit.n. In the first place I must shove against an atmosphere pressing with a force of fourteen pounds on every square inch of my body. And that is what happened to us. I must do this whilst hanging from a round planet head outward into space. This is the sort of English up with which I will not put. . also the ultimate. MOLECULE. we have tried to say what we have to say simply. then at some point you become so frustrated with what you perceive as the shortcomings of the available texts that you decide to write your own. inorganic chemistry is the study of Max Shulman I am standing on the threshold about to enter a room. The ultimate. Verily. the condensation or precipitation of matter from ether-whose existence is proved by the condensation or precipitation. and so on. the essential parts of what we teach in two courses to juniors. I must make sure of landing on a plank travelling at twenty miles a second round the sun-a fraction of a second too early or too late. here is our effort. but the other counted better. These are some of the minor difficulties. Arthur Duke oj Wellesley Wellington and has enough physics and physical chemistry to be comfortable with basic molecular concepts. it is difficult to know what to assume as prior knowledge. The present trend of scientific thought is toward the theory of ions. but which to you are of consuming interest and central to the whole field. but if unfortunately I should slip through the floor or be boosted too violently up to the ceiling. Three great scientific theories of the structure of the universe are the molecular.

as we shall see.g. are simply a consequence of their size and chain-like structure.of.range. but it should be kept in mind that similar things were being said round about 1950. just before high density polyethylene and isotactic polypropylene made their debut (some of this terminology will be defined shortly). WHAT ARE POLYMERS-WHAT IS POLYMER SCIENCE? Simply stated. physical cheIIiistry. commercializing a new commodity plastic would probably cost more than $1 billion (The Economist. analytical chemistry. however. no one person has an in-depth knowledge of all these fields.e.CHAPTER 1 The Nature of Polymeric Materials "He fixed thee mid this dance of plastic circumstance" -Robert Browning A. second. b Some of these are unique to polymers (e. This book is a first step towards the former and to give a flavor for the diversity of this subject matter we will commence with an outline of some of the areas we will cover. achieve levels of production comparable to polyethylene or polystyrene). First. This is not promising in an industry that has become infected by MBA's with a 1 . Clearly. result in a class of materials that are characterized . for some special types of polymers. rubber elasticity) and. chemical and mechanical engineering and.. Today. polymers are very large molecules (macromolecules) that are comprised or built up of smaller units or monomers. pl!ysic&_(particularly theories of the solid state and solutions). all the monomers that can be readily polymerized already have been.jilectrical. there are two good reasons to think they may be right.. the various types of chains that can be synthesized and the shapes that these chains can bend themselves into. It involves asm:cts of organic chemislI¥.engineering. May 1980). The arrangements of these units. Polymer Synthesis Many polymer scientists think that it is unlikely that we will ever again see any new thermoplastic take the world by storm (i.properties. Most polymer scientists seek a broad overview of the subject that is then usually supplemented by a more detailed knowledge of a particular area.. Polymer science is also a relatively new discipline and one that is characterized by extraordinary breadth. y an enormouaand intriguing.

easily delineated. and scan the chapters on. or chains with the types of delocalized electronic structures that result in unusual electrical and optical properties. fatigue resistance. this does not mean that polymer synthetic chemists are out of business. In chapter 2 we simply cover the basics of polymer synthesis. SOME B_ASI_C_llEEl~JIIONS_:.ma.t:J:y_QLhm. however. These materials would be produced in smaller quantities than "bulk" plastics. but can be sold at a much higher price. Polymer Engineering Paul Flory was awarded the Nobel Prize in Chemistry for his work in this area and we will mention his name often in this book. still a classic after forty years. Once made of course.. there is enormous interest in producing ultra-high strength polymer fibers. Polymer Physics Polymer physics and polymer physical chemistry are overlapping disciplines that are not..LEMENTS OF ~ROSTRUCTURE It may seem a bit like putting the' cart before the horse. but by no means least.!. The last chapter in this book will deal with mechanical properties of polymers. there is an enormous field based on characterization. spectroscopic techniques are used to examine "local" chemical structure and interactions in polymer systems. but a. Our focus in this book will be on the basics. for example. but we will first describe~QI!le basi« molecular__ <. the theories of rubber elasticity. not an easy task. For example. (The areas of characterization and polymer physical chemistry overlap considerably and they are artificially separated here merely to illustrate the different types of things polymer scientists do. Our intention is only to give you a start in this area.uhe_se~.. kinetics of phase changes and !I electrical and piezoelectric properties. phenomena at interfaces.nsid_~~she_mis.number of leading_:physicists (notably de Gennes and Edwards) started to apply modem theories of statistical physics to . how chains fold into crystals. This involves more than stretching a fiber until it breaks. how components of a system phase separate into various types of structures. and there has been much recent emphasis on polymer blends or alloys and polymer liquid crystals. These novel analytical techniques are not only useful in studying new materials.By arranging the subject matter in this manner it immediately becomes clear why . or in producing better defined chain structures to give controlled properties. structural characterization and theory. solution properties and the determination of molecular weight. Last. often using the types of instruments mentioned above. viscoelastic and I: _relaxation properties.m:~ I!0lymer~. thermal properties (crystallization.rJ!. Some techniques are so expensive that national facilities are required. _----Polymer Physical Chemistry This subject is still attracting enormous interest. Electron and other microscopies and the scattering of electromagnetic radiation are used to characterize overall structure ~r morphology. Polymer physics is not confined to theory.:lH!.to _ characterize chain structure.. The result has been a revolution in polymer theory. and the shape of an individual' chain in a particular environment. solution thermodynamics. Fortunately.r_acterization What a chemist thinks he or she has made is not always the stuff that is lying around the bottom of his or her test tube. Polymer physical chemistry is a subject that demands a knowledge of theory and the ability to perform carefully controlled experiments.rhe descrip-tion of long chain molecules.) Thesimplest way to get a "feel" for this subject is get a copy of Flory's book Principles of Polymer Chemistry. synthesizing "specialty" polymers such as those with stiff chains and strong intermolecular attractions to give thermal resistance and high strength. one that is not easily assimilated by "traditional" polymer scientists and is still ongoing. however. synchotron radiation and neutron scattering. particularly those interfaced with highpowered yet small and relatively cheap dedicated computers.g. In our discussions of polymer physics and physical chemistry we will focus on chain conformation and morphology (structure).:_1'lI_E_E. but this is enough for you to grasp the essentials of how the majority of commercial polymers are produced. The mechanical properties of polymers are complicated by all sorts of factors (defects. but answering questions that have intrigued polymer scientists for decades. This is now a particularly exciting area because of recent advances in instrumentation. of -befor:~_o.e. There is considerable interest in using new catalysts (for example) to produce commodity plastics more cheaply. phase behavior. It turns out that even "common or garden" polymers like polyethylene or polypropylene can be processed to give "high-tech" properties.useofmolecular spectroscopy.terials~ate. etc. Historically. Accordingly.2 THE NATURE OF POLYMERIC MATERIALS Polymer Microstructure 3 six month time-horizon. the mechanical properties of these materials have to be determined. there is a vast area that i~hemical engineering (the processing of polymers) and mechanical engineering (studies of strength. B.r__C. Until then most theory was based on almost classical physical chemistry.) as applied to polymeric materials._I!mduc_ed. so you won't find any discussion of advanced synthetic methods in this book.. ~e. Experimental polymer physics continues to focus on areas such as chain conformation. The trick is to align the chains as perfectly as possible. e. For example. however. melting and the glass transition). etc. in many cases.1la.) and the field is an intriguing combination of mechanical measurements.f!~~Lqfth~£hID1L~tQ!c. etc. it is possible to point to an enormous impact by theoretical physicists starting in the late 1960's and early 1970's. relaxation processes. and in particular we will discuss the measurement of molecular weight and the . i.

There is one that is both descriptive and rem~ns widely used. It is usual to represent such a polymer in terms of its chemical repeat unit.. Polymers where there are different types of units in the same chain are called copolymers and we will describe these in more detail shortly. It turns out that when polyethylene was first synthesized on an industrial scale (in the 1930's) a high pressure process was used (at low pressures ethylene does not polymerize at all) and for various reasons (that we will consider when we discuss synthesis) branched polymers were produced. as follows: where n can be very large. A s th e names suggest. they flow and can be molded when initially constituted.1.1. If we could see such a chain under a microscope of some kind it would appear. more precisely. usually through the action of heat and (often) pressure. This difference in behavior is.4 THE NATURE OF POLYMERIC MATERIALS Polymer Microstructure 5 certain synthetic methods have become important and how polymer chemistry can now be used (in principle) to obtain designed materials. !IWO\V\1:IJQ. like a tangled up kite string or fishing line. relative to small roughly spherical molecules such as H20. We will discuss some of the history in chapter 2 and the interested reader should consult the references listed in that chapter. A part of a typical chain is shown in figure 1.. flow more easily. Reheating-a "cured" sample serves only to degrade and generally mess-up the article that has been made. Branched and Cros~s The simplest type of polymer is a linear homopolymer. The chemical structure of a branched polyethylene is also illustrated in figure 1. this process is often called "curing".Q!Ymers to t hermonlasti m ermop astlcs an d thermosets. usually at elevated temperatures. Thermosets are like concrete. of course. Many of these are quite arbitrary and the subject has outgrown them.1 only shows a small part of such a chain. but can be reheated and formed into different shapes. however. Normally. the former flow or. a direct consequence of the various types of arrangements of units in polymer chains (polymer microstructure) and we will now take a look at these structures.. with particular types of molecular groups arranged in a chosen pattern. The chemical repeat units of a few common polymers are listed at the end of this chapter.cllainJhatjs_mde up ofideoticallloits (except for the end groups) arranged in a linear s~uence. . by which we mean a .1 "Ball and stick" and space filling models of polyethylene. but then become set in their shape.em. It wasn't until various catalytic processes were discovered (in the early 1950's) that linear polyethylene was synthesized". pushed or stretched by a load. Figure 1. An example of a polymer that can be synthesized in the form of a linear chain is polyethylene (usually denoted PE). (A) Part of a Linear Polyethylene Molecule Linear. In many texts it is also customary to introduce various classification schemes at this point. commercial samples of this polymer are very long (have a high molecular weight). hold their shape at room temperature. so that figure 1. id I . (B) Part of a Branched Polyethylene Molecule * The history of these discoveries and the interplay of the personalities involved is both fascinating and intriguing.!h.~~\'\ an d thi s di VI es p. MQSLQf. when squeezed.

The types of random short chain branches described above are not the only type that occur. where every unit is interconnected through various tortuous pathways. long and star branching. Homopolymers and Copolymers: A Question of Semantics? Synthetic polymers have traditionally been named for their respective monomeric precursors.3 Schematic diagram depicting the development of branched chains and networks. At low degrees of branching. are pro.. The point that should be grasped here is the effect chain structure has on properties and hence the importance of having a firm understanding of these first principles and definitions..es.. strength. but it is lower thap that found in samples of linear nolyethylene. poly acrylonitrile [-CH?-CHCN-] from acrylonitrile CH2=CHCN.. crystallization in polymers is a broad and extensive subject in its own right and one we will return to later.2). CO+bU~A~E\G\ . Figure 1. there is more than one way to skin a cat.6 THE NATURE OF POLYMERIC MATERIALS Polymer Microstructure 7 The difference in chemical structure between linear and br!ill. then a socalled infinite network is formed._l'his is illustrated in figure 1. Because linear chains can pack in a regular three dimensional fashion. where Y or X shap~monomers are shown combining to give what usually becomes a very complicated three dimensional structur_e~s.from . they form a ca. Certain_types of monomers form highly branched structures at the_o_IlS_eu>f polymerization.undly_ affected by the changes in c~ that are a consequence of branching.however. etc. however. is only formed at a certain point in the reaction and always remains BRANCHING TYPES NETWORK FORMATION ~ x TRIFUNCTIONAL TETRAFUNCTIONAL Figure 1. (Highly branched and hence completely amorphous samples are not (presently) of any commercial significance. ethylene-co-vinyl acetate copolymers from ethylene CH2-CH2 SHORT LONG STAR Figure 1. Because of the regular arrangement and hence tighter packing of the chains in the crystalline regions. Such differences can have a profound affect on p-roJl-~. mixed with monomers and other small molecules. all the original functional groups are reacted). This process is called crosslinking (or vulcanization) and is crucial in forming various types of elastomeric materials.4 Schematic diagram depicting the cross-linking of linear polymer chains. optical clarity.a.tyrene CH2=CH(CJI5). there is a degree of crystallinity.fu. A network. for example. Obviously. as illustrated in figure 1.ched polyethylene app-ears to be gyite small and localized to those units that form the branch points. because the random location and varying lengths of the branches prevent regular packing. and networks can also be formed by taking linear polymer chains and chemically linking them together.4.stalline phase (while part of it stays tangled up in an amorphous phase)..n~..) Properties such as stiffn~ss. polystyrene [-CH2-CH~sBn . A highly branched polymer is incapable of doing this. This type of network is typical of various thermosets. If all the chains become interconnected by such reactions. The Y and X shaped molecules are trifunctional or tetrafunctional molecules (have 3 or 4 chemical groups that are capable of reacting with one another to form a covalent bond).2 Schematic diagram depicting shart. linear polyethylene is often called high density polyethylene (HDPE). unless the reaction is driven to ultimate completion (i. As the old saying has it.. while slightly branched material is called low density polyethylene (LDPE). Various reactions lead to long chain branches and it is also possible to synthesize comb-like or star-shaped polymers (illustrated in figure 1.3.e. poly(vinylidene chloride) [-CH2-CCI?-] n from vinylidene chloride CH2=CCI2.

CH2=CHX. This is somewhat confusing.micr. The term copolymer is usually used to describe a polymer derived from two or more monomers. only loosely described as homopolymers. poly(methyl methacrylate) and poly(vinyl chloride) are only a few examples of common polymers where addition is almost exclusively head-to-tail. synthesized at relatively low pressures using Ziegler-Natta catalysts..f\CA-\e) chains of regularly arranged units. e I :ve.iLusually_does-so-in-a 1?referred direction. givmg head to head and tail to tail placements: Head to Head I I I DA Isomerism in Polymers /V1ot.). However. : : (TH) Head to Tail Head to Tail ~" : • I 2 -CXY- (TH) (TH) (TH) S?dm)e p(X0lymers.placed.. low. poly(vinyl fluoride) (X = H. . as we shall see. Thus. The problem is that many polymers synthesized from a single monomer might be better described as "pseudo-copolymers".ety_of. for several specific polymers there are *- I. ~heree Q-\z.. numbers of structural units that are incorporated "backwards" into the cham leadin~ to hea:I-to-head and tail-to-tail placements.th .F) h ~v~ a sigm icant number of units incorporated iznifi "ba kw .e ther c b _ _:t:11J:1. free radical and catalytic methods (ZieglerNatta and supported metal oxides). or by the deliberate copolym~rization of a second comonomer. we mentioned above that polyethylene may be synthesized from ethylene via a number of techniques including high-pressure..es_cnnsis..' c ar s mto e cham. polyisoprene and polychloroprene. does not mean that the final product consists of ID01!.all. ___. at best.CH 2-CXY-:-CXY-CH2-:-CHiCXY-:-CH I ~" ~" ~" I I I I I I I I Tail to Tail Head to Tail 2 -CXY- ) (TH) (HT) (TH) (TH) Stereoisomerism Continuing ~th our discussion of polymers derived from a single monomer let us ~ow c~nslder another major facet of polymer microstructure nam 1 ' stereozsomensm. During the polymerization of ethylene. si~cant.described.In ~he general case of a vinyl monomer CH2=CXY.S :f=al1C-L . The main difference being that short chain branching units are incorporated either by polymerization conditions using a single monomer.~p.. are alkyl substituted groups (-CH2CHR.mme_~_aILQbtaiILa_Yari.6 thus: Y- Head to Tail -CH2-CXY-:-CH2-CXY-:-CH2-CXY-:-CH I I I ~"~.av. It is interesting to note that "actual" copolymers of ethylene and butylene.ti~ various geometric and stereo isomers.§ible_l:!1Tangements -t1 Sequence Isamexism When a monomeomiLaddUo_a__grmring_chaiIL.}L_Q_ er car on atom has two J]. The most important types are sequence isomerism. Cl~ssic examples include poly(vinyl fluoride). are commercially available.ths. Polystyrene. have been commercially introduced and called "linear low density polyethylenes" (LLDPE). . this term is not necessarily an adequate description of such polymers and may lead to serious misconceptions. There are large differences in the physical and mechanical properties of these polyethylenes which may be primarily attributed to variations in the polymer chain microstructure. medium and high density). the units are arrang ed almost exclusively in a head-to-tail fashion.I:. usually denoted in terms of density (i. A range of polyethylenes.ra .::CHX may be a h~oge~>-alkyl or other chernicl!l moj~~eA.0 ({\I\CUCQ.o§. Y = F) or poly(vinylidine e fl uon e d'" .Lb_yd1:_ogenJeads to. as it is analogous to low density polyethylene (LDPE).hloride) (X = H Y = Cl) pOly(me~ 1 methacrylate) (X = CH3. stereoisomerism and structural isomerism. it follows that polyethylenes can be.e.suhstituent.nn. if the term homopolymer implies a polymer chain containing identical structural units.iILlerms_of_.. which are the short-chain branches described above. ethyl etc. 5' or ' . where. In poly(vinyl . polymers synthesized from a single monomer are commonly referred to as homopolymers but. " T-h ::os!)) . as in the case of ethylene-co-vinyl acetate mentioned above. In fact.g. but there is a degree of logic in naming the polymer as such.. This is referred to as s~ru(e?celi~omen~m. We have already seen that branching can QCcur in a polymer such as polyethylene and it turns out that if we start with an Jl. . the head of the unit is ar~ltarily defined as the CXY end while the tail is defined as the CH end A umt may add onto the chain in two ways: 2 .B Just because a polymer is synthesized from a I1ure monomer. it might be better to describe polyethylene in terms of a copolymer of ethylene and other a-olefins.ostructur. Polymerization of a vinyl monomer. For example. Y =-COOCH3) or polystyren~ (X = H C H) / example .::sII" R -CHiCXY* R*+CH=CXY~ 2 ~ = (TH) T H ~ 1"2 = (HT) T R-CXY-CH* \f ~?r convenience we will define "normal" addition as (00 and "backw d" addwoD as (HI).8 THE NATURE OF POLYMERIC MATERIALS Polymer Microstructure 9 and vinyl acetate CH2=CH(OCOCH3) etc.Y . These structural irregularities or defects. To reiterate.lTI9lecules are initi~y identical. p y vmy e fluonde).s 120 ?ul2._sy. 6pol~ers WIth microstructures that are.zy. structural units other than methylene groups are incorporated into the polymer backbone. . where R = methyl.

.._QD. . Isoprene.J:here. or Plexiglas®.. at the other extreme. Commercia.ls__an_enoDllllus_nUmheLQ_f_p_ojym_er chain microstructures based on different distributions of meso and racemic QMIlf). At one extreme. may be synthesized to yield an essentially isotactic or syndiotactic polypropylene by varying the polymerization conditions. We will consider such conformations later). If there is no preferred direction to the ftrATh(.etw.A plastic beer glasses. which contains an asymmetric carbon atom~ into the polymer chain in a manner such that each X_grpJlI! is meso to the preceding X group (i. In common with our general theme. for example._an aiactic: ~AnSllcA polymer is fo~.e-X group is racemic to the preceding one (i.acb..6 Side ~iew ':ball and stick" and top view space filling models of extended chain tsotacttc and syndiotactic polypropylenes. randomly with respect to the stereochemistry of the preceding unit. i. This is called a syndiotactic polymer. for example...mes.5)..pQlystryene.e_en.. used for the manufacture of clear nD..C> addition of a monomer to the growing chain.10 Polymer Microstructure THE NATURE OF POLYMERIC MATERIALS 11 relative to the chain and the next X group along the chain..eseJ:YL<LeXtre.l. but would prefer to fold into a helix. addition of a m. B.9lY-m~in S'lhiclLe.successiv. the monomer unit. (A) Part of an Isotactic Polypropylene Molecule \ Structural lsomerism-The Microstructure of Polydienes Another outstanding example of the complex polymer chain microstructure that can result from the polymerization of a single monomer is seen in the synthesis of polydienes from conjugated dienes.th..9turalrepeating units. are examples SiO-)i\\J:-~f a commonly encountered atactic polymers.A.5 Schematic diagram depicting racemic and meso diads.e. This will prove a useful approach when we apply statistical methods to the description of microstructure in chapters 4 and 5. (Because of steric repulsions the isotactic chain does not like this zig-zag shape.e.e.see figure 1. Polypropylene (X=CH3). on the opposite side of the stretched out chain-again.QIIleI:Jmit the growing. to polymer chain may be directed to yield a p. it is also possible to conceive of vinyl homopolymers in terms of "copolymers" in which the comonomers are represented by the meso and racemic placements of the structural units of the polymer chain. the monomers are incorporated J~\C\1:.\A~ tllacem~lltS_9f st:nJ.. These structures are illustrated in figure 1.5). Figure 1. This polymer is described as isotactic. see figure 1. on the same side of the extended chain . may be (B) Part of a Syndiotactic Polypropylene Molecule racemic diad r meso diad m ~igur~ 1. and poly(methyl methacrylate).A.6 for an extended polypropylene chain. Conversely.

Jn terms of the concentration of A and B units incorporated into the polymer chain and how these units are distributed.. D. However.caJLb~~f!JLIllore. but they're often regarded with fear and loathing by new students of polymer science because a description of their chain stnlC_tw:e.6 86.e.. Technol. L. The geometric and stereoisomers that are found in polydienes are a result of the configuration of the double bond and whether the polymerization has proceeded by opening both monomer double bonds or just one. we must remember.1 1.7 (%) 2. _possible.9 1. Jr..4 (lIT) cis-1.miCI.4 11.4 1. Two monomers may be distributed in a polymer chain in a numbt..4 1. which may resemble (St)p-(Bd)g-(St) .7 (%) 13.5 0.7 (%) 5. G.6 0. there are still some of the other structural isomers incorporated.~c. are also used to describe polymers derived. of a copolyme:rt synthesized from say the two mOllQJl1ersAean_cLR.5 1. Brame. These have the same configurational properties as vinyl polymer chains and may occur in isotactic or syndiotactic sequences.tb. q and r represent blocks containing different numbers of St or Bd structural units. practically pure cis-1.cI X=H X = CH3 chloroprene We have numbered the carbon atoms for reasons that will soon be apparent.2 4.4 (%) 76. four or more comonomers.1 1.12 THE NA1URE OF POLYMERIC MATERIALS Polymer Microstructure Table 1. (using "living" anionic polymerization .7 Schematic representation of the structural isomers of polyisoprene. consider copolymers synthesized from the monomers styrene (St) and butadiene (Bd).2 1. we can synthesize a block copolymer. 50(1) 49 Figure 1. For example. The 1. (For completeness. the results shown in table 1. Three common examples of diene monomers can be represented by the general formula: Polym.5 1 CH2 where: - 2 3 CX . the terms terpolymer and tetrapolymer etc.1 79.:LoJ".1 0. 13 polymerized to polyisoprene by a variety of techni~ including free radical.I. At one extreme.6 (%) 1. Eour major structural isomers.7 1. for example. and so on. Consider. whether or not a catalyst has been used.. but many synthetic elastomers have various proportions of the isomers described above . Tabb and E.t:Illcture.are. Rubber Chern.1 Microstructure of Polychloroprenes. The average comp-osition of the chains varies witb the temperature. (1977).2 or 3. Even when the polymerization conditions dictate that a particular structural isomer is overwhelmingly favored. Such * M. Temp. a triblock copolymer where p. "real" copolymers are synthesized from two or more comonomers. Heavea (Natural Rubber) and Gutta Percha. Coleman.3 9..t:hat the. Using free radical methods of polymerization a wide distribution of all four structural isomers is obtained.eqIliresJeaming some nomenclature. respectively. '~~J!P_Qlymers-Additional Complexity or More of the Same? As mentioned above.. respectively. The arrangement of the substituent carbon atoms relative to the double bond in 1.6 8.7 0.7 0. ' .omplelLanlLalLoLthecmicrostruCWral variations that we have considered for homopolymers can be present in copolymers..2 3.2 iso-1.1 0.4-polyisoprene.4 polymers are formed by linking carbon atom 4 of one monomer to carbon atom 1 of the next.2 83. Mother Nature also synthesizes "natural" polyisoprenes._CJlpolymer.4 (TH) fd trans -1..4 polymers can be in either the cis or trans configuration. M. Sequence isomerism is also common.7 1... And structural isomerism is not the only complication occurring in the microstructure of polydienes.of.8 (%) 1. but uses trees instead of round bottom flasks as polymerization vessels. shown schematically in figure 1.1 which one of us (the better looking one) obtained from an NMR analysis (see Chapter 6) of polychloroprenes synthesized at various temperatures". The so-called 1. The way in which isoprene is incorporated into a growing polymer chain is dependent upon the method of polymerization and the precise experimental conditions..8 10.1 3.actual microstructure of.pclymerizatica.CH butadiene _isg_p~ = 4 CH2 +40 +20 0 -20 -40 _X_=_.4 87.4 polymers are formed when incorporation into the polymer chain occurs through either the first or second double bond.os. and so on.jy_en=defined ways. trans -1.see Chapter 2). from three. which may be varied significantly by changing the temperature of polymerization. Many common elastomers are polydienes and therefore important polymer materials..7. ~_oosideLthe.) In the simplest case. Natural rubber is high molecular weight. ionic and Ziegler-Natta catalysis.

mts and. 50% B units and the sequence of units followed the ?robabihtles of com toss (Bernoulian)statis_tk. As long as the monomers are incorporated into the gr~wmg cham entirely at random then the probability of adding either an A or B um~ solely. it is possible to synthesize an alternating copolymer of the two monomers which may be represented by. ' JlI\ Isr Qc.e_ss_aril~us. (Note also that it would be perfectly reasonable to consider this as a "homopolymer" of polyAB . Fine examples are the copolymers of styrene and maleic anhydride synthesized by free radical polymerization.. exotic arran~ments snch as star copolymers whose arms are blocks.e. one could build a hypothetical chain by assuming the probability that the first unit is a.75.necessary to . t~ some degree rmsleading and they are more accurately descnbed a~ statistical copolymers.anceiILproducing commercially useful materials. and fall into the general category of themwplastic elastomers. the ubiquitous soda bottle.w. (The chemistry involved in synthesizing such chains is very clever and polymer chemists get up to all sorts of weird and wonderful things these days!) Graft copolymers can also be regarded as a type of block cop-olym_gJilldJbs:y_bave. ~o~ymers ar~ often.a 50. however. Although . Polymer Blenas ::1)~lZlc SI (lOST) C~ '--D 0 f{J 9l:\De lOq Q::V1CO DA _. In this case.~ [... thud and up to the last unit). Certain types of mixtures can have enhanced impact resistance. and so on. This is not really an adequate description a~d ~t IS.the polymer chain certainly doesn't know the difference!) However.ay_oLdoing_thi.n A is equal to 0. the~ are ObVlOusl~ a. We will discuss this in more depth when we consider copolymerization but we must clearly be careful in our use of the word random.8. !he~eill~ce_of mosLp_oly. ..mers_to~ has.25 or that of a B is equal to 0. depends upon th~ relative concentration of each. At the other extreme. _A_A_A_A_A_A_A_A_B_B_B_B_B_B_B_B_B_B_A_A_A_A_A_A_ Block COp'olymers Q).we have def~ed three distinct types of copolymer microstructure. if the cham consisted of75% B uruts and 25% A units it can still be designated random ~s long as ~7 pla~ement probabilities were adjusted to account for the difference in comp.f.theI:. three (triblock).:r I A~ _A_B_A_B_A_B_A_B_A_B_A_B_A_B_A_B_A_B_A_B_A_B_ Alternating Copolymers . although this depen?s upon the character of the components and the size of the phase separated domains. In recent years an increasing number of miscible (single phase) systems ha:~ been reported and this is presently (1996) an area of intense research activity.Imponanr r~catl?ns m th7 plastics. we must recognize that random copolymers do notba~ haze.s to simply is IDlX ~e homop-olymers together to form blends: Most non-polar polymers do not m~x. If the copolymer chain consisted of exactly 50% ~ . There are even.ele_ss. Consider. tnads and sequences ofhigher order). which are essentially alternating (see later in chapter 5). Here _qlain_§__Qf we of uniL(sa~ one are ~graf~nto the backbone QfJu:.Jaken_oruncreasing_signifu:. -A-B-B-B-A-A-B-A-B-A-A-A-B-A-B-B-A-B-B-A-A-A-BRandom Copolymers _B_B_B_B_B_B_B_B_B_B_B_B_B_B_B_B_B_B_B_B_~_B_B_B_ « « « « « « « « I I -I I I I I I « « « « « « I I I I I I « « « « « « « « I I I I I I I Graft Copolymers 4= 4= I « Block copolymers can also be made with two blocks of units (diblock). . and so on for the second. -{AB)n::.Ano.reasons we will discuss in the chapter on solutions and blends.. altho~gh this IS.5~ % cO?Jpos~tion. supenor or Just different from those of the homopolymers that can b~ produced from the same monomers .bain Qf_a_differenL1ype-(B. for example.::>\O(I/'G1GI Iffii\h (j)1:t'tJ\'Il\ Betwee~ t~ese two extremes w~ can also define random copolymer~.\.)..e. For example. for example. . then the copolymer would mdeed be truly random.osItIOn (i..u.. illustrated in the as above schematic representation.. Illustrated schematically in figure 1. S~ch ImmlSClblil blends are not nec. each A unit incorporated into the polymer chain is immediately followed by a B unit and vice versa..descri~e the microstructure in terms of the sequence ~stnbutl~n ~usually diads. ~or . While some believe it is simply an mdustnal consprracy that prevents us from recycling all plastics materials it is actually a f~rmidable scientific problem.s. we see in the litera~e descriptive terms describing copolymer microstructure in terms of an al~e~a1!ng_or~bloc~endency.re_Q¥-~. synthesized in an attempt to obtain properties that are mtermedmte.. mynad of intermediate microstructures.14 THE NATURE OF POLYMERIC MATERIALS Polymer Blends 15 copolymers are commercially manufactured by the Shell Chemical Company under the trade name Kraton®. Thus.

--b---di-'---sm'g the measurement _---.1!S __. . every molecule is identical (ignoring isotope effects)]. e g.~lble ~th HD th' t are chemically quite similar. b:cause degree ojpolymerization (which is involve the relationship of mechanisms ~o . HDPE is immiscible . a .1 hi h are necessary ecaus nlastics matenais.0. It has been our experience that students are confused by these definitions the first few times they encounter them.. say. .. . We can obtain this result using equation 1.3.m PE For that matter.~ . which is ~~ed-bYSiffiI. We will first consider the ways 10 ~ 1 1 eight substance. We shall see that this results in diff<::!:entmolecular weight averages. and there is a spread around some mean value. or A key factor m deteI1!l!I!1:!!g_p~¥ID--P-will..!!. DuPont Polymers Figure 1. ible with one another? We will it that the vast majority of polymer are lmID1SCl attempt to answer that question in Chapter 9.. and is either a mosquito or an elephant".0041bs (1.!itthe individual chain molecules are not all of the same weight. A mildly vulgar English term for rear-end... defined as: Number average: (1. for a material of low molecular weight such as benzene or water. th USA but not before the differenlJllastics d Soda bottles are recycle in e the d then only the poly(ethylene materials are separated from one d~o h~\ :nsity polyethylene (HOPE) base terephthalate (pEn blown bottle ~ . . th remammg_plasn. d i di tribution and we e scus !!1()l~_!ll~\V~!ght. however.68.is completely lIDIDl.2% Polypropylene Label 1. Why is with polypropylene (PP).-0. ·tted.2}!. the type you encounter every time you go camping!).1) Ethylene-co-vinyl acetate . The polymer is said to have a molecular weight distribution. hich the molecular weight of a hig.8}!. N. w c. involving many diffe:ent ThIS IS quite a Q.7% important characteristic of the vast mruority of synthetic high_p. = x Mo)' ~ is the number (or number of moles) of such molecules and Nx and Wx are the total number and weight of the molecules of "length" x.toJandfilL Why led. ** If we were talking about chains. This might at first be confusing..IYadding-up-the weightsof3il the things present (elephants and mosquitoes) and dividing by the total number of these things (elephants and mosquitoes). Glue for Base Oo4}!. We need to consider some of molecular weight m some detai 10 P discussion of polymerizations will basic definitions here. would be the number of chains that had x repeating units.004Ibs.16 Molecular Weight ll-IE NATURE OF POLYMERIC MATERIALS 17 Polypropylene Cap 2. respectively. Two of the most commonly used averages are the number average molecular weight and the weight average molecular weight. so it is worth spending a little time discussing averages and distributions in more detail and to make a start we will consider an example worthy of Monty Python's Flying Circus. -r~ . two po ymers .91ymersis th.7g -204% Poly(ethylene terephalate) Stretch Blown Bottle 49. s diftierent parts of the bottle. 0 if a stan d ar d 2 litre sodabOllle. Since there are four mosquitoes their total weight is 4 x 1 = 4 lb. The total number of things present is: describing chain length).6g . al irements 0 e vanou physical and mechamc requ . Imagine an elephant with four mosquitoes perched upon its bum".000 lbs and the mosquitoes are the size of small birds weighing. Let's say the elephant weighs 10. Q("MOLECULAR /' WEIGHT-SOME IN IT IAL OBSERVATIONS where ~ is the molecular weight of a molecule corresponding to a degree of _P()!Y!!1erizationx (i.1). = 1O. . but the average molecular weight is not a unique _thing. which together with the 10.e. 1 er roperties is polymer chain len~th. . injection molding ~ac~m~ and . [In contrast.2404% (1. consisting of x monomer units of molecular weight Mo' thus M.2) Source: Phil Blatz.3) where x now represents a species.§_---:{.1: Total weight = .6% SBS Block Copolymer Gasket in Cap 0.e-c.e nd t the mixture into an extruder or can't we just ~d up th: whole b:es~me:'ng useful? It just doesn't work.Pmnu cup are actually recyc .s_ar. M. it can be defined in a number of different ways.QDlpJe p ec b e of the different chemical.000 lb weight of the one elephant yields a combined total weight of 10.£E_I. What is the average weight of the elephant and the four mosquitoes? The average that isfamilar to most people is the numberaverage (equation 1.Q!ymerdif_fers from that of a low-mo ecu ar-w .000) 0.c.Cha ter 10.8 Schematic representanon . f th . 1 lb each (you know. = (4 x 1) + (1 x 10.' x j e of-engineering.6% High Density Polyethylene Base Cap 1704g . The most :Q. e th chain and is another way of equal to the number of structural umts 10 e L N..

e. The effect of the mosquitoes sitting on the elephant's bum would be negligible compared to the stomping administered by the elephant.00015 0.J Wx == 10004 10 8 ' == 10. but takes into conslderiitionthelItotal weight. Table 1.7) where we use the symbol == to mean "approximately equal to" or "of the order of'. we obtain: Wi'L Wx' instead of a number fraction ' is therefore about 2. the average weight of each species is just the sum of the fractional contributions of each of the components in the mixture).2 Molecular Weight Data. (1.000 g/mole etc. suppose the elephant isst3rtled bya~.still don't get it then there is no hope for you when it comes to und~rstanding averages and you might as well give up and become a sociologist. but the calculation of M.00000 Weight of Fraction (NvMv) 0 3 781 37500 400000 1000000 600000 87500 3125 27 0 M..000 g/mole are so.2. There are 4 lbs of mosquitoes and 10. IS left as an exercise for the student (your answer should be of the order of 103.000 and 30.000) == 10 8 (1. by W x' the total weight of each species. the elephant. The s~ms necessary to calculate M" are given in table 1.6) Now. Clearly.000 or more than ~50. Mw. wx = however. so that in our absurd example one-fifth of our sample is elephant and four-fifth's mosquito. They are equivalent.. the third between 30. and then sum over all species (i. the number average is obtained by "sharing" the total weight of all the species presenteqilally amongst each of the spe:cies.000 lbs of elephant present.2. the number and weight averages can be calculat. Let us consider a sample where by careful fracn?nanon we can separate the molecules by size. We can determine the fractional contribution by number..) The number of moles of ~ach fracnon can ~e plotted as a function of the molecular weight of that fraction In the form of a hIstogram.-. Note also that equation 1.7 does not look like equation 1. so that the number average weight is: If you .000 g/mole.. ~. 1.000 Clearly. we obtain: L Wx Mx = (4 x 1) + (10..--- _ M.18 THE NATURE OF POLYMERIC MATERIALS Molecular WeigJu 19 Total number = The number average weight. = LWxMx '" £.7 leads to the same result as equation 1.5) = L Nx Mx '" £.004 :::::2000 5 -. the second contams those molecules with a molecular weight between 10 000 and 30 000 g/mole.IJllS_qf weight instead of numbers.4) If we use a weight fraction.000) == 2.00000 5.000 (1. A weight average does not "count" species just by their number.000 and 50.00015 0. (Note that the mole fraction of polymers with molecular weights less than about 50. it will become apparent to you once you have read ~ha~ter~ 2 through 4 that most methods of synthesis lead to a statistical disn:bun~n of molecular weights.) Using the data In table 1. as the total weight of all species x present is 'LW x = 'L NxMx and substitution into equation 1. so if w_l:_~c_Qunt'.J M.000 lbs. goes on a rampage and tramples the poor little beast.000 g/mole h~s an average value of 20.2.000 (1. = 10.2.000 ) (1. and is a much better measure of stomping ability than the number average.8) -.01953 0.J!LtS!.00000 0. If this idea of averaging by weight instead of number is still giving you concern. is given by:' >. small that they don't register on this plot.'-' . We can determine an average weight per species by multiQ!ying the fraction of e~hSPe~ies present. in order to obtain an average we must divide by the total weight of _l?ve!I~~ p!~~ent.000).3.2.62500 0.00000 10. The weight average is a much more realistic indicator of this gross mechanical property.S~7~N~~~---. M U'n L N. and so In this s~ple there are no molecules that have a molecular weight of greater than 200 000 and the number of moles ~f pol~mer in each fraction are/given in table (It is assumed that the fraction WIth a molecular weight between 10. as shown in figure 1.000 x 10. such that the first fraction con~s molecules with a molecular weight less than 10.62500 5.1 and 1. try one more way of looking at it. Thus we replace N.. . the weight average in this example is close to the weight of the largest species.000 g/mole.01953 0.2 andequations 1.--------------. however.9) Nx == 10. Molecular Weight Fraction 1 2 3 4 5 6 7 8 9 10 11 Mx 0 20000 40000 60000 80000 100000 120000 140000 160000 180000 200000 Number of Moles Nx 0.9. 0. in equation 1. Getting back to polymers. hlx.----- Here we use -"xto denote thenumber fraction (or mole fraction) which is defined - . = L~Mx = (~ Xl) + (tx 10.ed. Now. = (4 + 1) = 5 (1. soiliat our new average.00000 0.v = L Wx Mx = (10~04 Xl) + (~~:~~~ x 10. by the wcight of each species.

and the position of the peak maximum corresponds in this case to the number average molecular wei ~ and is equal to 100.11 Plot of the weight of polymer fractions versus molecular weight. by summing the values of NxMx for all the points shown in figure 1. for those of you interested in such things).000 10 10 8 8 6 4 6 Ii\• · •. • . • •• . This is because the distribution is symmetric about this maximum position (it is actually a Gaussian distribution.10 and then divided by "Q:j -= is 800000 600000 400000 200000 0 0 OJ) • : .. We calculated the value of M" as before. This fi~e plots the number of moles of a fraction with a particular molecular weight (Nx) as a function of molecular weight CMx).t'\.10 Plot of moles of polymer fractions versus molecular weight. • .9 Histogram of the distribution of molecular weights. or at least one where the separation is so fine that to all intent and purposes we can treat it as continuous.. · ••. '. ·•• ·•••• 100000 4 •• . 100000 Molecular Weight 200000 :E Nx' Figure 1. ~ 0 N 0 0 0 0 <:> <:> <:> <:> <:> N ~ 0 0 Molecular Weight Figure 1.· • • : : . Clearly.0:. At this juncture we wish to make some additional points about averages. as illustrated in figure 1.. .. . · · • • ·••• · ~ ':.000 g/mole. and to do this we will next consider the distribution of the same sample separated into much finer fractions. 1200000 1000000 I~ Mw = 104. 0 <:> <:> <:> <:> <:> N <:> <:> <:> <:> '" <:> <:> <:> <:> 00 <:> <:> <:> <:> <:> pol <:> <:> <:> <:> N pol <:> <:> <:> <:> '<t' pol <:> <:> <:> <:> '" pol <:> <:> <:> <:> 00 pol MOLECULAR WEIGHT Figure 1. It would obviously be much better if we could obtain the distribution as a continuous curve. 200000 2 N 0 0 0 0 10 C\ 2 0 .. .. ~ 0 <:> <:> <:> <:> '<t' 10 C\ . ~ Q ~ fIl I~ • • • • • o • • • • Mo = 100. :.10. however. •• • .500 . Later on in this book we will show how a technique called size exclusion (or gel permeation) chromatography (SEC/GPC) can be used to accomplish this.20 THE NATURE OF POLYMERIC MATERIALS Molecular Weight 21 12 o 12· . the answer you get may not be precise in that we have separated the sample rather crudely into fractions that each contain a significant range of molecular weights. • • • ·. • · •· .

we have seen that the number and weight averag~§__arenothing_!!lore' than the first moments of their resI1ective distributions. This peak position is the first moment of the weight distribution about the origin and is equal to Mw (i. But if there is a distribution of molecular weights.12.10) and 1. The position of the number average molecular weight. . for the weight distribution) no longer correspond to the peak positions. such as that illustrated in figure 1. adding up the values of NxMx for all the points shown in figure 1.500 glmole (the distribution for this example is still symmetric).12 Plot of moles of polymer fractions versus molecular weight for an asymmetric distribution of molecular weights. which is often referred to as a monodisperse sample.. number of moles versus moleCUlar weight) we plot a weight distribution (the weight of polymer in fraction x (W x = NxMxfversus the molecular weight of the polymers in that fraction.05). However. It is actually very difficult to get distributions this narrow. M.200 • ° 0 " 200000 300000 100000 Molecular Weight Figure 1.. ~ = 100.2 you should notice that Mw can be obtained from ~ by multiplying each term in the summations by Mx.eLth@_ th~!Do~~nts (although they are related) and this ratio is called the polydispersity of the sample. ~ 0 6 4 : 2 0 .11. L NxMx. ~. such as that shown in figure 1. We will see later when we discuss the kinetics and statistics of polymerizations that some quite unusual looking distributions can be obtained. We can also multiply each term in the expression for Mw by Mx to obtain a so-called z-average molecular weight. js shown as a vertical line which you will note does not go through the peak maximum. · .. are there more? If you go back and compare equations 1. If we take anyone of the points on the curve shown in figure 1. skewness or asymmetty (third moment) as etc. although plots of the distribution look quite broad. SUCB_ its breadth (second moment). and M. and M.1 * If you get into statistics in more depth you will find that these properties are related to moments about the mean.. In polymer science it is common to use the ratio of the weight average to the number average molecular weight as a measure of the bread~ of the distJjf~ rath. were not very different. but we have no need to worry about that here. rather than the origin.. · • •• • · · · · · · • • !i'C rr Mn == 108.".11) the polydispersity is actually very small (= 1. then the difference becomes larger and the average values (M" for the number distribution. More on this later! and so on to obtain higher order mokcular weight avera~ 12 10 8 ~ ~ <Il & For a homogeneous sample where the polymer chains are all the same length. Actually. First we have defined two ways of averaging molecular weights (M. ~ 1 (1. • . Mx) we obtain the distribution shown in figure 1.10. M. M... = ~ and the polydispersity is equal to 1. ~ = LNxMx LNx~ 3 (1. The second point we wish to make before concluding is that so far we have only considered symmetric distributions and in the examples we gave.e. is that there are experimental methods that give these averages directly and we will describe some of these in Chapter 10. This is of course the number average molecular weight.e. Polydispersity = M. Most commercial polymers have much broader distributions (polydispersities of the order of 5-10 are not uncommon). we are actually taking a moment about the origin (think back to your old high schoOEtUffclii-weights placed on a balanced beam or kids sitting on a see-saw)... but it can be achieved using some clever chemistty_(as we will see later in Chapter 2). 0 • · . L (NxMx)Mx / L NxMx). So. higher order moments (second moment. as soon as we consider asymmetric or skewed distributions. • · • • · . third moment and so on) can be usefuiastl1eydesci1. and normalizing by dividing by L Nx).e.500 Mw == 118..11) The importance of M. the peak position of the weight distribution is shifted to about 104. . You should also note that in-the simpl~ ex-irriPie (figure 1. There are two things that remain before we conclude this initial discussion of molecular weight. is always greater than ~.22 THE NATURE OF POLYMERIC MATERIALS Molecular WeigJu 23 We mentioned this to make an additional point concerning the moments of a distribution.10 and multiply the value Nx by Mx for this point. If instead of plotting a number distribution (i..000 g/mole.13 which shows both the number fraction and weight fraction distribution of chains found at a certain stage in a polycondensation reaction. M. and ~). It turns out that when describing distributions. . The first moment of the distribution as a whole is defined as the sum of all individual moments (i.be various properties of the distribution.

----.05 techniques. the simplest polymer would consist of an infinite chain of structurally and configurationally perfect repeating units. analytical problem and one that we address later in this book: We will SImp~y conclude this section by noting that various advances III spectroscopic v: (a) Calculate the number and weight average degree of polymerization and polydispersity of this collection of polymer chains. What is the difference between a block and a graft copolymer? 8. Conceptually. ~ (Weight Fraction) 3. describe the polymer cham microstructures that are theoretically possible. D.~ C.._"-----"i"T Study Questions 25 24 THE NATURE OF POLYMERIC MATERIALS 0.c U '0 0.--. A sample of polystyrene is described as having a polydispersity of 1.---. Also. STUDY QUESTIONS 1.I -: /' (Number Fraction) 2. the determination of the mo~ecular weight and molecular weight distribution would be sufficient to charactenze the structure of the chain. If polymer chains were made up of identical monomers.. have allowed a much more detailed analysis of chain structure than that available even 10 years ago. Polyethylenes are often characterized as being of high density (HDPE) or low density (LDPE). What configurational isomers are possible following the polymerization of butadiene. but fa~cIll~nng. We can approximate the ideal case in.. type a?d distnbunon ~f U?ItS present in the chains of a given sample. a polymer chain made up of a single well-defined type of unit is a rarer beast than one would first imagine. What is "linear low density" polyethylene (LLDPE) and how does it get its name? 4. In a systematic manner. Show an example of sequence isomerism. ~ven in this case the polymer by necessity has end groups (normally methyl or ~nyl group~) and inevitably there are chemical and structural defects present in the cham resulting from oxidation.. for the majority of synthetic polymers. However. Comment of the relative effect that this has on the calculated number and weight average molecular weights. Polymer Analysis (or How Do We Know What We've Got?) One of the more difficult and important problems in polymer science is the characterization of the structure of a macromolecule. Different polypropylenes does this mean ? may be described as atactic or isotactic..03 = .04 .---. A chemist synthesizes a copolymer containing 90 mole % isoprene and 10 ~ole.03. for exam~le..06 . (b) If these chains were poly(methyl methacrylate)s what would be the corresponding number and weight average molecular weights (ignore end group effects) ? (c) Add five more chains of degree of polymerization 10 to the above distribution.... What causes the difference in density between these polyethylenes? ~ I. . other competing reactions occur such as branching and the various types of isomerization. Degree of Polymerization Figure 6. but we have not ~e~tlo~ed how e would experimentally determine the number. 0. This is a ~fficult. particularly nuclear magnetic resonance spectroscopy. Consider the following distribution of polymer chains: 5 chains of degree of polymerization 25 chains of degree of polymerization 50 chains of degree of polymerization 30 chains of degree of polymerization 10 chains of degree of polymerization 5 chains of degree of polymerization 10 100 500 1000 5000 50000 a typical poly condensation reaction. CH2=CH-CH=CH2 ? 5. % methyl methacrylate.----. degradation and the incorporation of impurities. We have defined these and o~er aspects of polymer microstructure in this chapter. ~ ultra-high molecular weight linear polyethylene (PE).) Be careful! There are many possibilities. What does this tell you? 10. 9. because of the nature o~ the polymerization process.----'1 0.. What i13 Plot of the number and weight fraction distributions for polymer chains found in 7.. (Ignore limitations due to chemistry. What is the difference between a thermoplastic and a thermoset ? la . which we have not considered yet..--. However. but there are still many problems that remain to be solved.

"0 n H 0 H2-C /I /C ~ CH3-O 0 1~ H2-C I CH3 n H2-CH=CH 3 ---CH2 L CH t n +CH~5-~j-t . -i(CH2)5 Polycaprolactone (PCL) H2-t=CH---CH2 ~ t-CH21=CH-£H2i. Cornell University Press. Ithaca. APPENDIX: CHEMICAL STRUCTURES OF SOME COMMON POLYMERS Polymer Chemical Structure of the Polymer Repeat Poly(vinyl chloride) (PVC) tCH21HJ f-cH2 if5-'cH Polyethylene (PE) Polypropylene (PP) fm2~2t CHi CH2-tH CH3 I CH2-tH Poly(ethylene oxide) (PEO) Poly(methyl methacrylate) (PMMA) Polybutene-1 Polyisobutylene (PIB) Polybutadiene (PBD) cis-Polyisoprene (Natural Rubber) trans-Polychloroprene (Neoprene® Rubber) t-c t t CHi ~ ~4 In + -CH2 -o1 n n n Poly(methyl acrylate) (PMA) n Polytetrafluoroethylene (PTFE) Polycaprolactam (polyamide . Koenig Principles of Polymer Chemistry. (PVAc) l- I 0 I n 3 F. 1982. Flory. J. Wiley.Appendix 27 26 NATURE OF POLYMERS E. 1953. n . New York. SUGGESTIONS FOR FURTHER READING P.Nylon 6) FF F:t CH3-O 2-C tH21Hi /C:--. New York. J. L. Polymer Chemical Structure of the Polymer Repeat I-cH2-CHI- Poly(vinyl acetate) . -kH2-CH-~ l- Poly(ethylene terephthalate) (PET) Polystyrene (PS) © Poly(dimethyl siloxane) (PDMS) toJ-<Q)l-D-£HCCH2± H n ---<':! -ot n n F ~i-O 3 CH3 +. Chemical Microstructure ojPolymer Chains.

We will consider some of these shortly. Our aim will be to show how the mechanism of polymerization affects the molecular weight and microstructure of the polymer or copolymer that is being formed. INTRODUCTION In the next few chapters of this book we will be considering some of the . where spectroscopic methods for characterizing polymer chain microstructure will be described. for example. In the former the chemical repeat unit of the polymer has a different molecular formula to that of the monomers from which it is produced. as a result of the elimination or "condensation" of certain groups. In this case no further condensations are possible and polymers are not formed. followed by a discussion of the kinetics and statistics of some of these reactions. suggested that polymers could be broadly regarded as belonging to one of two types: condensation or addition. where it has had the biggest effect on everyday life" -Lord Todd. where a molecule of water is eliminated (hence the name condensation). This part of the book will culminate in Chapter 6. A typical example of this type of reaction is the formation of an ester from ethyl alcohol (the active component of beer) with acetic acid (the active component of vinegar). 1980 A.CHAPTER 2 Polymer Synthesis "I am inclined to think that the development of polymerization is perhaps the biggest thing that chemistry has done. Acetic Acid o II Ethanol + CH3-CH2-OH CH3-C-OH ~ Ethyl acetate Water + H20 CH3-C-O-CH2-CH3 • Wallace Home Carothers was a brilliant scientistwho made seminal contributions to polymer science. This will include a description of the basic mechanisms by which practically all synthetic polymers are formed.fundamentals of polymer synthetic chemistry. Carothers". 29 . It is convenient to classify polymerization reactions into two or three basic types.

but they form polymers that could also be synthesized by an appropriate condensation reaction. These conditions are characteristic of most laboratory studies of polymerization that we will be concerned with here.. and some polymerizations fit uneasily into these two categories. as these provide the fundamental insights into mechanism and its relationship to the nature of the product. although the arrangement of bonds is different. which then reacts to form polymer and is therefore used up proportionally. at the end of this chapter. In the following chapter we will discuss aspects of the kinetics of these reactions that will inherently assume that we are dealing with a batch process. initiator. even though water is not eliminated during the reaction: Each has 2 reactive end groups A can only react with B and vice versa Monomers Monomers ~ Dimers ~Trimers + Hexamethylene Diisocyanate HO-CH2-CH2-OH Ethylene Glycol Figure 2. as in the polymerization of caprolactam: Cr~O--. but we will only qualitatively discuss some of these large-scale systems.30 POLYMER SYNTHESIS Step-Growth Polymerization 31 The structural units of addition polymers.r N'u Caprolactam ·1CH2-CH2-CH2-CH2-CH2-l-& Nylon 6 (polycaprolactam) 1 n Accordingly. For example. which are formed by the reaction of isocyanates and alcohols. we prefer to adopt the classification given by Rempp and Merrill (Polymer Synthesis. have the same molecular formula as their monomers. as in ethylene and polyethylene: Cook at Elevated Temperature )Ira (Remove Water) B-B A-A B-B CH2=CH2 Ethylene Polyethylene There are exceptions.1 ~ Polymers Schematic representation of step-growth polymerization.. In certain cases we might also have solvent. In other words. are considered condensation polymers. On the industrial scale of production continuous processes are widely used. Furthermore. however. Monomers combine with one another to give dimers: MJ+MJ ~M2 fc >s-zJ CH2 i + Nylon 6 (Polycaprolactam) dimers and trimers can also combine with themselves or each other to give higher oligomers: . in contrast. One final thing needs to be kept in mind before we proceed to discuss the specifics of polymerization. ring opening polymerizations must be regarded as addition polymerizations. and perhaps one or two other additives present. 1986). STEP-GROWTH POLYMERIZATION The central feature of this type of polymerization is the slow building of chains in a systematic stepwise fashion as depicted schematically in figure 2. where polymerization reactions are considered to be either step-growth (this category includes polycondensations) or chain polymerizations (which include ring opening polymerizations). where the resulting polyamide could also be produced by polymerization appropriate amino acid: of the B.1. we will consider a reaction "pot" into which we place monomer. polyurethanes.

that these materials consisted of "giant" or macromolecules. melted • This view was challenged and defeated by Paul Flory. thus breaking up the chains.000. The established view had been that these were loosely bound associations or aggregates of small molecules. after this is complete no further chain forming reactions can occur). perhaps because of the controversy his views initially generated. to prove the existence of macromolecules by systematically building large molecules from smaller ones. but initially a molecular weight of 6. including the polyamides (nylons). At this point you are probably mystified as to why we are discussing the first polyesters in a section that is supposed to be about nylons. each has a single group capable of participating in a condensation. polyesters. high molecular weight polymer is only produced at the end of the polymerization. an observation of immense subsequent importance in the processing of fibers. for examples) if he were to obtain • Staudinger was not awarded a Nobel Prize until 1953. fundamental work on polymerizations. polymers (this seems simple and obvious with the benefit of hindsight.6. Hill noticed that the molten polymer could not only initially be "drawn". but is a great and ultimately tragic story. when many did not accept the existence of macromolecules at all. Stine believed that the success of his program would depend upon his ability to hire outstanding scientists. he decided to hire "men of exceptional scientific promise but no established reputation" (clearly a Politically Incorrect statement!).32 POLYMER SYNTHESIS Step-Growth Polymerization 33 M) +M2~M3 M2+M2~M4 M2+M3 ~M5 M3+M3~~ and so on. The reactants. The molecular weight of the product was found to be over 12. polycarbonates. The reason is that these polyesters.500-4. however. polyurethanes. so this result was perhaps not that surprising". but will start with a brief description of the discovery of nylon 6. He initially chose to examine the condensation of alcohols and acids to give esters. but must be placed in the context of the times. Because established people (predominantly academics) proved difficult to move. The champion of a different hypothesis. but further stretched or "cold drawn" after cooling to form strong fibers. At that time it was the consensus of opinion. producing long chain molecules: Near the end of 1926 Charles Stine. As a result of this step-growth mechanism. amongst other things. At that time (about 1928) there was still widespread disagreement concerning the nature of what we now call polymer materials. sometimes you have to outlive your enemies! X= The success of this approach helped to establish the macromolecular hypothesis. from a three carbon chain glycol and a 16 carbon chain diacid (m = 3 and n = 14 in the equation given above). With typical insight Carothers realized that he would require bifunctional monomers (see below. This is another story. constructed a molecular still that allowed the removal of water and polymerized a so-called 3-16 polyester. phenolic resins. ethanol and acetic acid are monofunctional. was Hermann Staudinger". There are a number of important types of polymers that are produced by this type of polymerization. whose units were linked by ordinary covalent bonds. who was described as brilliant but mercurial. although they appeared to have useful mechanical properties (they had an elasticity that at that time could only be matched by silk). from dicarboxylic acids and dialcohols (glycols) in the presence of an acid catalyst: o H0-C-(CH2)n-C-DH II 0 II + HO-(CH2)m-OH o HO-C-(CH2)n-C-D-(CH2)m-DH II 0 II + H20 Historical Background-The Discovery of Nylon Because functional groups remain at the end of the newly formed ester. apparently shared by Carothers. let alone designing reactions to produce them). This not only serves to introduce features that are characteristic of all polycondensations. and one to which we will return.000 appeared to be the upper limit for these reactions. Carothers accepted Staudinger'S view and his initial aim was a piece of pure science. using simple well-understood reactions. an example of which was given above. More on this later. that the reactivity of end group decreased as the chain length got larger. after some persuasion the committee agreed to this proposal which included. We will give examples of each of these. He would do this in such a way that the structure of the final product could not be questioned. submitted a short proposai to his Executive Committee to initiate a program of "pure" or fundamental scientific research. but this position was becoming untenable. etc. then an instructor at Harvard. One such was Wallace Hume Carothers. . By the end of 1929 Carothers' group had synthesized large molecules with molecular weights in the range 1. water could hydrolyze the ester group back to an acid and alcohol. further condensations can continue.e. Chemical Department director at DuPont. Julian Hill. Nevertheless. far higher than any previously prepared condensation polymer. however (i. Bitt Carothers also realized that because of the reversible nature of the reaction. Carothers and his colleague.000. In those days this was unusual (one could argue that the cycle has gone full circle and we are once again in the same unfortunate state).

The synthesis of other polyamides from diacids and diamines (e. Delaware only.6 H '-----y----' The synthesis of nylon 6. The monomers for forming nylon 6. m+2. nylon can be fashioned into filaments as strong as steel. o + HO-C-(CH2)m-C-OH II 0 II Diamine . synthesized from diacids and diamines and known to have higher melting points: substances. Accordingly. nylon 6. Carothers and Hill turned their attention to the chemically analogous polyamides. by then vice-president for research at DuPont. water and air.. the number of carbon atoms in the monomeric diamine and the number in the monomeric dicarboxylic acid.tW\III\IIMMNVI II N -C. First. as fine as a spider's web. or other Nylon Salt The second point we wish to make is that nylons can also be made using an acid chloride monomer. where the 1:1 complex is precipitated. This had enormous impact. were soluble or partially soluble in dry cleaning solvents and were sensitive to water. the first man-made organic textile fiber prepared wholly from materials from the mineral kingdom.6. The reason for this will become apparent later when we discuss the statistics of linear polycondensation. Polymerization then occurs upon heating to appropriate temperatures and removing water. 6 n In 1938 at the New York Herald Tribune's Eighth Annual Forum on Current Problems. Charles Stine.34 POLYMER SYNTHESIS Step-Growtb Polymerization 35 below 100°C. air and water. In 1939 nylon stockings were first sold to DuPont employees who were residents of Wilmington. derivable from coal. Four million pairs of nylon stockings were sold in New York City alone in the first few hours.tW\III\IIMMNVI I H o Diacid Amide Group The products were designated by two numbers. Though wholly fabricated from such common raw materials as coal. however.a protein chemical structure.g..)6-~j--(CR. In an interfacial polycondensation the acid chloride is dissolved in an organic solvent (such as chloroform) and the diamine is dissolved in water. proved to be a more convenient and economic choice. such as bristles and sheets .that eventually went into production. '" [Nylon] is the generic name for all materials defined scientifically as synthetic fiber-forming polymeric ami des having. made the following statement: I am making the first announcement of a brand new chemical textile fiber .6 from adipic acid and hexamethylene diamine has been described above. . A series of different nylons were produced and initially Carothers decided that nylon 5.)4Jf H 6 Nylon 6. if high molecular weight polymer is to be obtained then it is necessary to have precise equivalence of the reactants. 1940 there was sufficient polymer being produced that sales were extended to the whole country. it differs radically from all other synthetic fibers..10 was the most suitable for forming fibers. By 1934 he was suffering from increasing bouts of depression and on April 29.. so that the polymer represented schematically above would be called nylon n.. yet more elastic than any of the common natural fibers and possessing a beautiful luster. This stoichiometric equivalence can be achieved by first making a salt from adipic acid and hexamethylene diamine. and it is this polymer. There is not much of a market for polyester shirts made out of such materials. Some Examples Polyamides of Condensation Polymers-Linear Polymers o H2N -(CH2)6-NH2 Hexamethylene diamine + HO-C-(CH2)4-C-OH Adipic acid II 0 II t ~-(CR. On May 15. The tragedy is that Carothers never lived to see this success. and characterized by extreme toughness and strength and the peculiar property to be formed into fibers and into various shapes.10) should be obvious. convinced that as a scientist he had failed.. 1937 he locked himself in a Philadelphia hotel room and drank lemon juice laced with potassium cyanide. however. In its physical and chemical properties. There are three aspects of the synthesis of nylons that should be mentioned here. hexamethylene diamine and adipic acid. When added to one another the two solutions form separate phases and a polymer skin (polyamide) is formed at the interface.

This is often demonstrated in the classroom and is referred to as the nylon rope trick. In this case hydrochloric acid (HCI) is eliminated rather than water and a base is normally added to the aqueous phase to neutralize the acid formed.2. which is used to produce the commercially important polymer. CH3-O_-C~-O-CH3 ~--I(5U~ ~ + 2HQ-CH2-CH2-OH Ethylene Glycol Dimethyl Terephthalate Figure 2. new polymer is immediately formed and the material can be removed in the form of a fiber or "rope".2 Schematic representation of interfacial polymerization: the "nylon rope trick". + HQ 1 and by transesterification. Polyesters can also be made from the acid chloride: o [ -QH + """"'C-Cl II . which we will describe later.x. ~~~ HOCH2CH20-C~C-OCH2CH20H + 2CH30H o II~II HOCH2CH20-C~-OCH2CH20H HOCH2CH20 0 o H2N-(CH2)6-NH2 Hexamethylene diamine + CI-C-(CH2)4-C-CI II 0 II Adipoyl chloride -c-IQ\--'o II~II 0 C70CH2CH. but by a ring-opening polymerization. poly(ethylene terephthalate) or PET (if you buy plastic coke bottles or clothing that contains "polyester". + HOCH2CH20H and so forth. This is illustrated in figure 2. it is usually PET). ~ 'Iransesterification o II 0 HOCH2CH20CC6H4COCH2CH20CC6~COCH2CH20H II II 0 II If the nylon is carefully lifted from the interface. ..oH 0 o H2N -(CH2)6-~ H -C-(CH2)4 II -C-Cl 0 II + HCI and so on. Finally.36 POLYMER SYNTHESIS Step-Growth Polymerization 37 Nylon "skin" carefully drawn off to form fiber or rope Polyesters General aspects of the synthesis of polyesters from diacids and dialcohols were mentioned above in our discussion of the work of Carothers. polyamides such as nylon 6 are not usually synthesized by a condensation reaction.

bedding.isomers) and hexamethylene diol is shown below: ---:l)o~ t By varying the chemical structure. a myriad of polyurethane foams with vastly different physical properties can be produced.o~--r-<£5(~H' O=C. BisphenolA + HO-(CH2)6 -OH Hexamethylene diol Toluene diisocyanate (TDI) 1 N-H I ° -(CH. lamentably (for nomenclature purists) usually called just that. temperature and amount of water present. 0-(CH2)6 . Formation of Networks by Polycondensation Polymers It should now be self-evident that if bifunctional monomers can be used to make linear polymers.6. A typical example of a polyurethane that would be formed from toluene diisocyanate (TDI-usually a mixture of 2. an innocuous gas which is used as an in situ blowing agent in the manufacture of polyurethane foams.). that is often used in the transparent covers seen in street lighting and the now ubiquitous compact discs. cushions. would be remiss if we did not mention that isocyanates react with water to form CO2. we Bisphenol A polycarbonate is an optically clear. such as football helmets. We will not dwell on the precise chemistry as it is beyond the scope of this book and covered in depth in many polymer chemistry texts. polycarbonate. POLYMER SYNTHESIS Step-Growth Polymerization 39 process of trans esterification is (in this case) the reaction of a terminal group with any ester group. packaging and lining for clothes etc. glassy polymer. stoichiometry. Polycarbonates Polycarbonates are formed by transesterification or by the reaction of diphenols and phosgene.) Where you are most likely to encounter polyurethanes is in the form of the structural hard foams that are used for thermal insulation or the soft foams that are ubiquitous in car seats. is derived from the diphenol. typically 1. However. lacquers etc. The resulting shuffling of the groups is a process which eventually leads to a most probable distribution of as described by Flory (who else). then tri or higher functional monomers can be used to make network or thermosetting polymers. It is a very tough material and has found many applications in products that have to withstand mechanical abuse. Two commercially important .4. on Bisphenol A Polycarbonate A typical aromatic / aliphatic polyurethane This particular polyurethane is a glassy polymer at room temperature and might find application in vamishes.38 The alcohol random species. bisphenol A. R-N=C=O + H20 Isocyanate R _:_NH2 + CO2 Amine Polyurethanes Polyurethanes are synthesized by reaction of a dialcohol with a diisocyanate. The most commonly encountered polycarbonate.and 2.000 glmole. Polyurethanes that are more flexible can be readily made by replacing all or some of the TDI with an aliphatic diisocyanate or by introducing ether linkages into the polymer backbone using hydroxy terminated ethylene oxide oligomers (molecules of moderate molecular weight.

Sometimes the active site can be transferred to a different part of the . allowing the synthesis of a range of polymers and copolymers.tI/VIIINIIIJWV M. a specially added reagent.3 Schematic representation oj chain or addition polymerization.). These chains grow very quickly. d) Chain transfer can occur. In spite of the differences in the nature of the active site. In contrast.M* 1 C. leads to complex mixtures of linear and branched molecules and finally various types of network structures. a general process illustrated schematically in Chapter 1 (figure 1. All of the species present are involved in this process throughout the course of the reaction. from monomers to oligomers to (eventually) polymers. one by one. c) The polymerization terminates due to destruction of the active site. Formaldehyde Typical Novolac Prepolymer A A A A A A AA Polymer Plus residual monomer A and very small amount of growing chains --A * One type of reactive monomer A + Initiator R* Monomers Figure 2. [ . A typical so-called novolac prepolymer that may be formed under acid catalysized conditions is shown below: R-A· R-A-A· R-A-A-A· R-A-A-A-A· R-A-A-A-A-A· R-A-A-A-A-A-A• (Q)-OH Phenol l~~yg~ + CH2=O . (ii) unreacted monomer and (iii) a very small number (relatively) of growing chains. as well as a variety of chain architectures or microstructures. In this process only a few species This active site can take a number of different forms. Reacting phenol with formaldehyde. in addition or chain polymerizations there is an active site at the en~ of a growing chain wh~re monomers are added sequentially.40 POLYMER SYNTHESIS Chain or Addition Polymerization 41 examples of these are phenol-formaldehyde and urea-formaldehy~e resins. CHAIN OR ADDITION POLYMERIZATION In the preceding sections we have described the key feature of condensation polymerizations: that chains are built step by step. The important point that should be grasped is the use of multifunctional monomers (where the number of functional groups >2) to make networks.tI/VIIINIIIJWV M* + M ---'))00'" . the relative proportions of reactants etc. for example. polyurethanes and epoxy resins. another polymer chain. The growing end (or sometimes ends) of the chain has an active site that obviously must be shifted to the new end after the addition of a monomer: Other examples of classes of polymers that can form thermosets by condensation reactions are polyesters.3. In this process the active site is transferred to another molecule (monomer. as depicted schematically in figure 2. are active and at any instant of time the distribution of species present usually consists of (i) fully formed and no longer reacting polymer chains. solvent. etc. but the details of the chemistry of these reactions is beyond the scope of the material we wish to cover. by which we mean an active site must be generated on a monomer. as we will see.3). depending upon the catalyst used. b) The chains propagate: monomers add to the active site and the active site is simultaneously transferred to the newly added monomer. addition polymerizations usually have most (but not always all) of the following features or steps in common: a) The polymerization must be initiated.

. as ther_ readily break down to grv. One of many experiments that were tested involved attempts to force ethylene and benzaldehyde to react...CH = CH2 CH=CH various olefms various dienes acetylene or they need to have a ring structure. with its unshared electron..'polymerization of ethylene at ordinary pressures the rate of _proeagation is much smaller than the rate of termiiliiIon. poly(vinyl chloride) (PVC) and poly(methyl methacrylate) (PMMA or Plexiglas®) had been reported. "chain transfer". This is not.42 POLYMER SYNTHESIS Chain or Addition Polymerization 43 same chain...9.~.. ..~.'so-cha-rnsao not get the chance to form._.>_ _ . the mechanism of such vinyl polymerizations was still not well understood and it is in this context that the initial work on polyethylene should be placed.._'< ..•. strictly speaking. As in our description of condensation polymerization.. At high pressures (and higher temperatures). . and polymer is produced.. Morawetz.._.tIIWVVIIWV\Iv'CH2-CH2 -CHi Propagation when two radicals meet through the process of teniiiiianonIater-fiiliTittirns'-outthai1i1'-ihifree niciicaJ.. however.~b"_ •• ~ ••_ •. usually containing at least one hetero-atom (e. This whole process is called initiation. At this point. __ ~ .. __~. John Wiley and Sons. Polyethylene (or polymethylene) was polymerized by the decomposition diazomethane around the turn of the century. It was immediately recognized that this was probably some form of polyethylene (this observation involved a good deal of insight and was not a trivial thing). but what was not realized for some time was that a peroxide impurity in the benzaldehyde had probably played a ke~ role.!¥iCiil~:_ R-Q-Q-R Alkyl Peroxide 2R-O· Peroxy Radical H2C-CH2 ethylene oxide C y~O H caprolactam N. we will accomplish this by picking a specific polymer that has an interesting history.. We-wilraescnoe >~ •• __ .Berthelot had actually described polymerizations some sixty years earlier. replacing the "sticks" representing bonds by dots representing electrons. to termination.__ . but Staudinger was apparently unaware of this (see H. but a white._ .eropagation is much. oxygen)._. -(CH2)nPolymethylene + n N2 where each pair of electrons.. We will discuss the key features of addition or chain polymerization by focusing our initial attention on so-called free radical polymerizations of olefin type monomers.. the rate of . but the unstable character of the monomer (it had a propensity to explode) did not make it suitable for industrial production. Initiation R-o-CH. waxy ~olid was found on the walls of the reactor.__ . On opening the pressure vessel after one such reaction it was observed that the benzaldehyde was apparently unchanged.dayroutin~ly used to initiate polymerizations.•.greater. This radical. 1985). As in step-growth polymerizations. when Staudinger published his paper "On Polymerization" that more serious attention was paid to the formation of polymer molecules by what was then described as an undefined process of "assembly" . This work was in any event ahead of its time (the concept of a polymer was not yet established) and it wasn't until 1920. represents a single covalent bond. New York.This reactio.research progr~ involving chemical reactions under high pressure was initiated m the laboratones of Imperial Chemical Industries (ICI). in this case polyethylene.~.~teITIl~?at~ . which usually proceeds raei_<!!x:._.1?. These monomers need to have either unsaturated bonds.rdmary random collisions.-.. By 1930 the synthesis of polystyrene... The Origins and Growth of a Science. The next step is~t?epropagation of the chain... but just ordinary common or garden transfer. chain polymerizations require monomers with a certain type of chemical structure.32 a.. . as a result of the introduction of previously unavailable high pressure equipment. Pero~des ~!-~<!. . and working our definitions into the narrative.. ----l)O~ .--~"--~---.~eroxy.-CHi 1 Free Radical Polymerization-The Development of Polyethylene of or._"_.g._. some examples being: CH2=CHX CH2 = CX . __ __ ~_.•__ relative ._~.. is highly reactive and can then add a monomer by pairing up with an electron "stolen" from the double bond of a vinyl molecule. [ R-O· + CH.=CH. :. Examples of this type of monomer are: I\ o ~ . Polymers..

The new reactor leaked and the presence of small amounts of oxygen were sufficient to initiate polymerization (there had presumably also been trace amounts of oxygen in the ethylene used in the experiment that had such explosive results. the discoverer of radar. however. Polyethylene CH2=CH2 CH3 R" + CH2=CH I Propylene etc. not all can be and for various reasons it is advantageous to use other meth~ds to obtain. but it is the short chain branches that play the major role in disrupting crystallinity. And so polythene played an indispensable part in the long series of victories in the air. but It was the advent of organometallic and metal oxide catalysts in the early 1950's that allowed the synthesis of linear or high density polyethylene (at low pressures). Up until the early 1950's polyethylene was made by the process. New equipment (and barricades) were built and in 1935 ICI again got lucky (if we define luck as the residue of hard work and intelligent persistence). Polyolefins The two most important polyolefins (in terms of commercial production) are polyethylene and polypropylene.!!la. and the point here is that at the time of these first experiments the process of initiation was not well understood".lUgh. We will come back to discuss the types of polymers produced by.. Because it has an extremely low dielectric loss it was used to insulate radar cables. but in ~ methyl groups know that this Because of the relative stability of a six-membered ring transition state.Y!!1Y!. the next set of experiments attempted the polymerization of ethylene alone. acrylic and diene monomers are polymerized free radically. U. R" + CH2=CH2 Ethylene ----J . This book is aimed at giving a broad overv~ew.hains. R-CH2-CHi . Examples of Polymers Synthesized by Free Radical Methods . radiation induced formation of a radical. or backbiting. A few long chain branches are also formed. hence resulting in a material that has a lower melting point and is weaker and softer than that made from "straighter" c. high pressure linear.. Sir Robert Watson Watt. certain polymers... As a result.. but it is appropriate to leave history hanging at this point. on the sea and land.Althc. for example..V. We now is due to intramolecular transfer of the radical. but the process was not easily contrclled and it depended on the weather. Sometimes nothing happened for hours. Nevertheless. involved placing monomer in big glass bottles and exposing them to sunlight. which were made possible by radar. however. pointed out that: The availability of polythene [polyethylene] transformed the design.I1Y. installation and maintenance of airborne radar from the almost insoluble to the comfortably manageable . • Early polymerizations of methyl rubber. tubes and gauges. production. the former can be polymerized free radically at high pressures and elevated temperatures. production was eventually achieved and the new material immediately found an important use.. sometimes the vessel exploded! Low molecular weight oligomers Propylene and other olefins with a hydrogen atom on the carbon adjacent to the double bond cannot be polymerized free radically. butyl branches can be formed.44 POLYMER SYNTHESIS Chain or Addition Polymerization 45 This is getting ahead of ourselves. As a result.... Branching could be reduced by going to still higher pressures. As we have seen. so we WIll neglect a detailed consideration of many specific systems and ~Imply conclude this section with some brief notes on the suitability of certain monomers or types of monomers for polymerization by a free radical process.. Ethyl branches may also be formed in a similar backbiting process.. because a . resulting in a "violent decomposition" (a chemist's way of saying it blew up) that burst pipes.. It was at first thought that the chains were reasonably 1940 it was found (by infrared spectroscopy) that there were more present than could be accounted for by chain ends alone. however. these catalysts in a while.) This took months of additional work to establish.

pipes. the nature of the active site was not well understood until the 1940's. Table 2.. I CI OCOCH3 CH2=CH Acrylonitrile I I CN CH2=CH I Acrylic acid COOH CH2=CH I Methyl methacrylate CH2 Methyl acrylate CoaCH3 C I Anionic Polymerization CH3 COOCH3 CH2-CH I The year 1956 was a very good one for anionic polymerizations. poly(vinylidine chloride) and poly(vinyl alcohol). Ionic and coordination polymerizations are now central to the synthesi~" of many commercially important polymers. poly(vinyl acetate) and its derivatives employed as adhesives. and their consequences in terms of the characteristics of the polymers produced.Poly(vinyl alcohol) is not synthesized directly from vinyl alcohol. Industrially significant polymeric materials with diverse physical properties such as polystyrene and poly(methyl methacrylate). Our purpose in this book is to establish general principles and many of the procedures and fundamental assumptions. which is produced by free radical polymerization. when we go on to consider the kinetics and statistics of polymerization. cyclics and symmetrically substituted monomers such as maleic anhydride. I . We will point out qualitatively where certain important differences occur.. so we will describe them here. I I CH26 Cl CH2=CH Vinyl chloride I Ionic and Coordination Polymerization Vinylidine chloride CH2-C Vinyl acetate Cl We have discussed aspects of free radical and condensation polymerizations in some detail and by doing so we have followed the general history of the development of polymer chemistry. "~yl-n1onOii1eis-"(6-ecaiise'ofcliafntransfer-and'-i11e stability of the radicals so generated). polychloroprene.46 POLYMER SYNTHESIS Chain or Addition Polymerization 47 hydrogen is transferred to propagating centers. poly(viIlyl-chioride) used as siding. however (see later). or polybutadiene.1 lists some common monomers that can be polymerized by free radical methods. polymerization. remain the same in dealing with these other types of polymerizations. which are elastomers. which are transparent glasses at ambient temperature. are all made by free radic3J.and copolymers produced from alkyl acrylates. Monomer Ethylene Tetrafluoroethylene Butadiene Isoprene Chemical Structure CH2=CH2 CF2=CF2 CH2=CH-CH=CH2 CH3 CH2=C-CH-CH2 Chloroprene Cl CH2=C-CH=CH2 Styrene Polystyrene. and polyacrylonitrile and acrylonitrile copolymers. In quick succession Szwarc announced that this mechanism can result in a "living" poly. alkyl methacrylates. Further examples inCludethe homo. such as the reactivity of the active site being independent of chain length. which are employed as fibers. Monomers that are not suitable for free radical polymerizations include those wi!h bulky su-6Stituents. Table 2. acrylic and methacrylic acids used in the surface coating industry. Although other methods. were actually used in some of the earliest reported polymerizations.however. -SoIlleof"iheseTatter monomers can be copolymeri:ze-d-freeracHcaIiy-Wfth"otller"monomers. Polyacrylics etc. however. Polydienes. but from the hydrolysis of poly(vinyl acetate). such as anionic polymerizations.1 Common Monomers that Polymerize Free Radically." used as barrier films in packaging. so that only low molecular weight oligomers are produced. and styrene / butadiene copolymers. fixtures and flooring for dwellings and numerous other applications. We will not consider these in as much ""detail.

also crucial in terms of creating the conditionS-necessary for continuous growth and we have to distinguish between "protic·i solvents~ which are those that can donate a proton to the growing chain -(Ceo chain transfer to solvent). at the start of the reaction.MNIMIII'CH2-CH-CH2 -. If all of these active centers are equally . Because the active site stays "alive". " . but also on the solvent. These developments had immediate consequences in producing almost monodisperse polymers. Clearly. those containing an "active" hydrogen). the number average degree of polymerization will depend upon the ratio of the rate of propagation relative to the rate of transfer. for example. In practice.g. which are "aprotic".. until this monomer is effecti. then it is possible to obtain a system where carbanion end groups are always present. . because of the absence of termination reactions. This depends not only upon the nature of the anion and counterion involved.. R . butadiene.. block copolymers and polymers that are more stereoregular than those obtained by free radical processes. Certain types of ring: _()pening polymerizations can also proceed by this mechanism: ... In "protic" solvents. and those that do not.. As an example we can consider a polymerization initiated by metallic sodium: • CH2-CH R Anion radical this gives an anion-radical that quickly forms a dimer 8 Na(±) + H2C\iCH2 I o Ethylene oxide ----'1. so that all active centers are effectively introduced at the same time.. while Stavely and co-workers used an alkyl lithium initiator to polymerize an isoprene containing greater than 90% cis-l. not all of the monomers that polymerize anionically can be used to make block copolymers and those students interested in this subject should consult the books listed at the end of this chapter that deal with polymer chemistry in more detail. . say. An example of the former is the polymerization of styrene in liquid ammonia using sodium amide as an initiator.. one in which there is no termination step. chain growth can be stopped by transfer to solvent... An anionic polymerization is one where the active site on the end of a and: groWfngchiilnIsnegatively charged:----------------------------------·---------- In these reactions there is no termination of the type that occurs when free radicals collide. The nature of the solvent is..4 units. If an inert solvent is used and there are no contaminants (e. there must be a counterion hanging around somewhere and it turns out that the separation of the end group and the counterion is a key factor in determining the stereochemistry of the propagation reaction.MNIMIII'CH2-CH8 + I R C1h=CH I ----'1.. however..tIIWWV'CH2-CH2-D-CH2-CH2-08 : The active site is in this case an oxanion.tIIWWV'CH2-CH2-08 As a result. R I R I .-elyu~~_c:l_lIPL!~_nrestart the polymerization using.1\ .-CH 8 The charged species is in this case a carbanion. Certain initiators are practically completely dissociated before the polymerization has time to start. One immediate consequence of the "living" nature of these polymerizations is that it allows the synthesis of block copolymers. As in free radical polymerization there are initiation and propagation steps: . : ! which can then propagate from both ends. A _ lhird_ block of styrene umiS-cansubsequentlY: be addecfon:--Such triblock copoiymeiii-canbe-usedas-ifierroopfasiIEefastomers:-Themorphologies of these block copolymers are both intriguing and beautiful and are something you should look forward to studying in more advanced treatments. it is possible to first polymerize styrene. 48 POLYMER SYNTHESIS Chain or Addition Polymerization 49 mer.

those monomers where R is an electron donating group are most appropriate for these types of polymerizations (e.2.) We won't go into the details of this. u Nx (u + 1)2 (2.weight distribution.2 Common Monomers that Polymerize Anionically. vinyl acetals. isobutylene. a carbonium ion): are those where the active site has apositive charge where u is the number of monomers reacted per initiator. recombination. accordingly. although these can be suppressed by lowering the temperature.. the types of vinyl monomers that can be polymerized anionically are usually those where there are electron withdrawing substituent groups. The distinguishing feature of cationic polymerization is that there is a tendency for the growing chains to deactivate. Termination can also occur by anion splitting: .1) Cationic Polymerization Cationic polymerizations (i. with the latter apparently requiring a co-catalyst such as water or methanol: Monomer Butadiene Styrene Chemical Structure Propagation occurs in the usual fashion: Acrylonitrile Methyl methacrylate Ethylene oxide Unlike anionic polymerizations. Table 2.e..e... alkyl vinyl ethers. because of various side reactions that can occur. and. Clearly as u gets larger the polydispersity approaches 1. Initiation can be achieved using protonic acids and Lewis acids. but the result is a Poisson distribution where the ratio of the weight to number average degree of polymerization is given by: Nw = 1 + . Typical examples are given in table 2.g.50 POLYMER SYNTHESIS Chain or Addition Polymerization 51 susceptible to monomer addition throughout the polymerization (i. and termination is absent. For the most part it is just easier to achieve these conditions with an anionic polymerization. if the reaction medium is kept well-stirred so that there are no large concentration fluctuations). gives this result. which act to stabilize the carbanion. as in: termination can occur by anion-cation Tetrahydrofuran Caprolactam where an ester group forms. Finally. then the result is a polymer with an extremely narrow molecular . para-substituted styrenes). (Any type of polymenzation where initiation occurs entIrely before propagation.

Typical examples are given in table 2. the use of various catalysts to produce high density polyethylene was reported by groups at Standard Oil of Indiana and Phillips Petroleum in the United States and a group directed by Ziegler in Germany. because of the expertise of Natta's group in x-ray diffraction methods. had also polymerized propylene with various catalysts. by writing the patent so narrowly that it only covered polyethylene.MMMIII'CH2-CHttl I AS + H20 R I ---. This coordination and orientation of the monomer relative to the growing chain end allows the insertion of the former in a specific manner. but that ended in disagreement and recrimination. At that time he did not believe that other olefins could be polymerized with his catalyst. The details of these discoveries make for fascinating reading.F. using what he called Ziegler catalysts. The nature of these catalysts appears to be still a matter of debate. M. Ziegler usually drafted his own license agreements and patents and in this instance made what the benefit of hindsight clearly shows to be a major error. which was demonstrated by Natta's systematic study of the structure of this and other stereoregular polymers synthesized during this period. This story is reconstructed in "The Chain Straighteners". in the course of a two to three year period. The greater strength and higher melting point of this polymer also led Ziegler to the conclusion that he had produced "a really new type of a really straight-chain high molecular weight polyethylene". but it is clear that Ziegler immediately recognized that the band corresponding to branches that is usually observed in the infrared spectrum of high pressure polyethylene could not be detected in his catalytically prepared material. which act to stabilize the cation. but Hogan and Banks of Phillips Petroleum were finally granted the patent for the production of crystalline polypropylene in 1983. Up until 1950 polyethylene could only be produced by a high pressure process. as mentioned above the types of vinyl monomers that can be polymerized cationically are usually those where there are electron donating substituent groups. but because of various delays their patent applications were ten days behind the Italians. By this time Ziegler's group had also succeeded in polymerizing propylene.S. Ziegler and Natta had been friends and collaborators.52 POLYMER SYNTHESIS Chain or Addition Polymerization 53 Coordination Polymerization or by reaction with trace amounts of water R . illustrated schematically below: . The Ziegler catalysts are heterogeneous and actually produce a mixture of atactic and isotactic polypropylene (from different sites on the catalyst). Monomer Isobutylene CH2=C Chemical Structure CH3 I CH3 Vinyl methyl ether CH2=CH I Styrene CH26 OCH3 I In tracing the history of the development of polyethylene we left the story at a most interesting point.3 Common Monomers that Polymerize Cationically. The latter was immediately identified. They were jointly awarded the 1963 Nobel Prize in Chemistry. history. McMillan. Table 2. In 1954.S.l»~ .MMMIII'CH2-CH-CH2-OH + AH Chain transfer to monomer can also occur by various mechanisms including: R I R I .MMMIII'CH2-CH ttl AS + CH2=CH R -~»~ ..3. but the mechanism is believed to involve complexes formed between a transition metal and the 1t electrons of the monomer. McMillan Press.. In 1952 Giulio Natta had heard a lecture by Ziegler on the polymerization of ethylene and on his return to Milano had started his own program along these lines. The Chain Straighteners. for those interested in how these things can occur. London. A number of other groups in the U. either by design or inadvertently. . Then. and the fundamental characteristics of its structure established. This resulted in patent litigation that has been amongst the longest and most complex in U. As we have noted earlier. polypropylene cannot be polymerized free radically. 1979. Natta's group polymerized propylene to give a product that could be fractionated into rubbery and crystalline components.NIIWW\I'CH=CH + I R CH3-CH I ttl AS Finally.

54

POLYMER SYNTHESIS

Polymerization Processes R Table 2.4 Common Monomers that Polymerize Using Ziegler-Natta Type Catalysts. Monomer Ethylene Propylene Chemical Structure CH2=CH2 CH3 CH2=CH l-Butene

55

.NVIIW\IV'CH2-CH IIIIII Catalyst

I

= ::

~

-=

It is beyond the scope of this book to go into the chemistry of these systems in more detail. Common monomers that can be polymerized using Ziegler-Natta catalysts are listed in table 2.4. All are hydrocarbons and do not contain strong polar groups. The crucial point is how the development of these catalysts allowed the synthesis of high density polyethylene, linear low density polyethylene, stereoregular polyolefins (polypropylene being the most important), diene polymers and ethylene/propylene/diene (EPDM) elastomers. One final comment before we close this section. The cost of research, development, scale up and introduction of a new polymer into the marketplace is enormous and it has been argued by many that essentially all the large scale economically viable (i.e. cheap) polymers have now been discovered. This, in large part, is why the commercial and academic interest in polymer blends is at an all time high and unlikely to abate in the foreseeable future. It follows logically that if desired properties can be attained by simply mixing two or more existing commercial polymers, surely this will be more cost effective than developing a new single polymeric material from scratch. While this is a persuasive argument, those of us who are approaching senior citizen status well remember similar arguments being expressed immediately prior to the discovery of Ziegler-Natta and metal oxide catalysts, which resulted in the introduction of isotactic polypropylene (PP), high density polyethylene (HDPE), ethylene-co-propylene elastomers (EPDM), etc.! And there appears to be no limit to the ingenuity of polymer chemists. Recently introduced metallocene and other catalysts promise "designer" homo- and copolymers, where desired structural properties of the polymeric material are, in principle, "dialed in". This part of the subject is still very much alive and we eagerly anticipate developments that in some ways will no doubt surprize us.

I

C2HS CH2=CH

I

Butadiene Isoprene

CH2=CH-CH=CH2 CH3 CH2=C-CH=CH2

I

Styrene

CH26

D. POLYMERIZATION

PROCESSES

In the preceding sections we have described the principal synthetic mechanisms used in the synthesis of polymers. Making these polymers in the laboratory can usually be carried out by straightforward methods using the established art and craft of the organic chemist. At this scale, polymerizations need not be driven to high conversions and subsequent separations etc., are not usually a big problem (you just tell your graduate student or postdoc to get on with it and try not to blow up the lab!). At the industrial scale, however, there are all sorts of difficulties. The viscosity of the polymerization mass can become unmanageably large, heat transfer (practically all polymerizations are exothermic) is often a problem, the kinetics of polymerization may effectively preclude one

polymerization method, but allow a different one, and so on. In addition, the process used to polymerize a monomer can often have a profound effect on factors such as molecular weight and microstructure. It is therefore important to have at least a passing familiarity with polymerization processes, even at this introductory level, so we will briefly describe the main features of methods used in large scale polymerizations. --The first distinction that is usually made by engineers is between batch and continuous processes. The "latter is preferred for large scale production, but CiiUiOt be applied to every type of polymerization. As we will see in the following chapter, step-growth poIYIi1erizations are generally slow· (hours), so that continuous reactors with inordinatel~g residence times would be requiTed~-For such polymerizations batch processes are employed~ -----rhe next distinction that IS generally made is between single phase E!Qcesses and multiphase processes. S"ingle phase processes include bulk (or mass), melt and so7utionj50lymenzatfons. In a bulk polymerization, as the name suggests, only reactants (and added catalyst, if necessary) are included in the reaction vesseLThis type of process is widely used for step-growth polymerizations., because as we will show in the following chapters, high molecular weight polymer is only produced in the last stages of the polymerization, so that the viscosity of the reaction medium stays low throughout most of the course of the
• Certain polyurethanes, such as those used in reaction injection molding (RIM), are formed very rapidly by step-growth polymerization and are exceptions to this general rule of thumb.

56

POLYMER SYNTHESIS

Polymerization Processes

57

Rapid Stirring

Rapid Stirring Suspended Beads of Monomer + Initiator

ED

ED@)

e
@
Polymerization

@

€V @}

Figure 2.4

Schematic representation of suspension polymerization.

reaction. _T~_prevent crystallization (hence the appearance of a solid phase in the reaction vesselj bulk polymerizations are carried out at temperatures higher than the melting point of the polymer (e.g., '" 250°C-[or nylon 6,6) and are often called_melt polymerizations. Chain polymerizations are less often performed in the bulk, because of problems with the control of the reaction .... There is a tendency for the reaction mass to form a gel (i.e., have an extraodinarily high viscosity) and "hot spots" can develop. At the extreme, the reaction rate can accelerate to "runaway" proportions (for reasons we will discuss when we consider kinetics) with potentially disastrous (explosive) consequences. Viscosity and heat control can be achieved, if necessary, by carrying out the polymerizations to a relatively low conversion, with the unreacted monomer being separated and recycled. Another way to control the viscosity and heat transfer problems of chain polymerizations is to perform the polymerization in solution. A major concern with this method is that "chain transfer" to solvent can occur (i.e., the solvent can take part in the reaction (see next chapter). In addition, the problem of removing and recycling the solvent is introduced, which is expensive and not a trivial factor in today's enviromentally conscious society. Even trace amounts of potentially toxic or carcinogenic solvents can pose a major predicament. _Q~"tQusly, one solvent, ideal from many points of view, is water. Nobody cares about trace amounts of water In polymer films tllat migfiil)e used to wrap food, for example. Trace amounts of benzene ( a grade A carcinogen) would be unacceptable, however. The problem is, of course, that most polymers (or monomers for that matter) do not dissolve in water. This brings us to the topic of mutiphase processes where polymerizations are performed in water with the
... An interesting exception, nonetheless, is poly(methyl methacrylate), a polymer that is soluble in its own monomer (not all polymers are), and which is synthesized commercially by chain (free radical) polymerization very slowly in bulk. The resulting polymer has outstanding optical properties (clarity) because there are very few impurities.

monomer or polymer suspended in the form of droplets or dispersed in the form of an emulsion. . Wit~ the excep?on of a few polar molecules, such as acrylic acid, ethylene OXIde, VInyl pyrrohdone and the like, most monomers are largely insoluble" in water. A mixture of the two will phase separate into layers, usually a less dense "oily" layer of monomer sitting on top of the water layer. Polymerization in this state :vourao6vi?usI~ n?t have much of an advantage over bulk polymerization. ContInuous rapId stImng can produce smaller globs or spherical beads of monomer suspended In the water, however, and each bead then becomes a miniature reactIon vessel, as depIctedScliematiCaIly In figure 2-:4-:---Thenitiator i used must also be essentIally water insoluble so that it prefers to reside in the monomer phase where it can imtiate polymerization. Problems of heat transfer and viscosity are overcome by this method. In addition, the final product is in ~e fo~ Of."bead~", which is convenient for subsequent processIng. There are difficultIes InvolVIng coalescence of particles, however, so a variety of additives (protective colloid.s etc.) are used to stabilize the droplets. The beads or monomer droplets In such a suspension polymerization (sometimes also called bead or pearl polymerization) are usually about 5 urn in diameter and require the mechanical energy of stirring to maintain their integrity. If the stirring is stopped, a gross phase separation into two layers occurs. There is a way to suspend even smaller monom~articles in water such that the mo?omer droplets are stable ~_c!_~?Ilot_~~~~_!_<? to form a separate layer. Essentially, ~rfactantjsoap) is used to form an emulsion. Surfactant molecules c?nsis~ of a polar h~ad (hydrophilic) group attached to a non-polar (hydrophobic) tall, such that It looks something like a tadpole, as depicted schematically in figure 2.5. . In water,. soap molecules arrange themselves so as to keep the polar groups .!!!.E~nta~wIth water molecules, but the non-polar tails as far from the water as possible (at concentrations above a: certain level, called the critical micelle conce.ntration). _O?e way of doing. this is to form a micelle, which usually has a sp~~nc.al or rOd-h~e shape, ~s al~o illustrated in figure 2.5. In this micelle, WhICh IS about 10- -10-4 urn In diameter, the polar groups are on the outside surface whi.le the ~on-polar tails are hidden away inside. Th~_~~Il:p~lar groups ~c:~!DpatIble WIth non-polar monomer, however, so that if monomer is now a~9:~_~<?_~~_water d dispersed-by stiiTIng,-tlie very small amounts of monomer ~ th.a_tdiss?I'{~il1_t!1e\yatercan-cliffuse to the micelles- and enter the interior nonpolar h~drocarbon part~tiieri more monomer enters-the aqueous phase to ~place that WhIChhas departed). In the same way, surfactants molecules can diffuse to ~e disp~~;e_~_~~~~~~ dr~}Jlets (whose size depends upon the stirring rate, but IS usually In the range of 1-10 urn), where they are absorbed onto their surface thus_~tabilizing them, again as illustrated in figUre-2.5. ._! Unlike suspension polymerization, where a water insoluble initiator is used ~n ~{'l~~.£I7~_.polymerization a water soluble initiator is added. The polymer~ ~atI0Il,~orth~_~<:>~t_E~-,--~~urs_In the swollen micelles, which can be thought
... It is important to. emp~asize the word largely, because even when you see a grossly phase separated system, like 011and water, a small amount of oil has actually dissolved in the water and vice versa. This will be important when we go on to consider emulsion polymerization.

58

POLYMER SYNTHESIS

Polymerization Processes Micelle Swollen with Monomer

59

Spherical Micelle

,

" ,, , " ,, " ,,
'" 1---0---1----,;---\

Hydrophilic (water loving) Head

~~--------------1'
Surfactant Molecule

,

Hydrophobic (water hating) Tail

Figure 2.5 Schematic representation of the initial stages of an emulsion polymerization. of as a meeting place for the water soluble initiator and the (largely) _\\,ll!~L insol uble-monomer*~-As'iile'p-oiymerizatlori-proceeas';-ihe micellesg£ow J2y_ addition of new monomer ({iffusing -i,aroiii-ineaqueous-plias-e~--At the same time, the size of the monomer dioplets snnnks, as monomer-dfffuses out into the aqueouspnase.-Therrn~wnere ilie polymerization is occunng gr?~ to about 0.5 urn in diameter and at thIS stage are called polymer particles. After a while ('" 15% conversion of monomer to polymer) all the micelles become polymer

* A small amount of polymerization sometimes occurs in the monomer droplets and almost certainly in solution, but the latter does not contribute significantly because of the low monomer concentration in the water phase.

particles and a while later (40-60% monomer conversion) the monomer droplet ~hase finally disappears. It is all quite remarkable. Termination occurs when a radical (usually from the initiator) diffuses in _from the aqueous phase. This is a major advantage of the emulsion technique for the polymerization of monomers such as butadiene, which cannot be polymerized easily by free radical means using homogeneous (single phase) polymerization, because it has a fairly high rate of termination. The termination step in an emulsion polymerization is controlled by the rate of arrival of radicals at the polymer particles, which depends only on the concentration.of initiator and the concentration of surfactant (hence initial micelle concentration). There are some obvious additional advantages to emulsion polymerization; as in suspension polymerization, the viscosity is always low and heat control is _~vely straightforward. Also, the final product is an emulsion of"" 0.1 Jlm diameter polymer particles that is typically 50% by volume polymer and 50% ~ater. This mak:es1t rumost immediately applica61e as a surface coating (paint). The major disadvantage 'iSine presence oftJieSUrractant, whlcliTsOifficult to completely remove even lIthe polymer product is precipitated and washed. There are variol.ls-types of muffiphase processes that are widely used in the mass production of polymers. The two phases can both be Iiquids, as in suspension and emulsion polymerization, or can be a gas/solid, gas/melt (liquid) or liquid/solid system. In the interfacial polymerization of nylon 6,6, for example, the two monomers are InItIally dissolved in different solvents, hexamethylene diamine in water and adipoyl chloride in chloroform, as we mentioned previously on page 36. In this system, the aqueous phase sits on the top of the chloroform solution and a solid polymer is formed at the interface. Polymerizations of an initial miscible mixture of, say, monomer + initiator or monomer + solvent + initiator also become multiphase if the polymer formed is insoluble in its own monomer (e.g. poly(vinyl chloride) or polyacrylonitrile), or the monomer solvent mixture (e.g. the cationic polymerization ofisobutylene in methylene chloride). . Finally, multiphase processes where the monomer is a gas are important methods for polymerizing monomers such as ethylene and propylene. Some low density polyethylenes, as well as several well-known ethylene copolymers such as ethylene-co-vinyl acetate copolymers, are still made by high pressure free radical polymerization methods. The process is usually continuous, with compressed ethylene, for example, at about 200°C flowing through a heavy walled tubular reactor. Initiator is injected into the ethylene stream as it enters the vessel. About 15% of the entering gas is converted to molten polymer which is carried out in the gas stream (there is often a polymer buildup on the lower surface of the reactor which must be periodically removed by a process called "blowing down"). These days, however, polymers like high density polyethylene are often produced in a gas/solid multiphase process. Catalyst can be suspended in an upward flowing stream of monomer gas in a fluidized bed reactor, for example, and crystalline polyethylene grows on the catalyst surface. Polyethylene and polypropylene are also polymerized by forming a slurry of monomer, insoluble catalyst and a solvent and again the polymer "grows" on the catalyst surface. But this is enough, you should have the general picture by

60

POLYMER SYNTHESIS

Suggestions/or Further Reading

61

now. There is a range of industrially important methods for synthesizing polymers and if your interest is in the engineering side of producing these materials you will have to study the methods we have mentioned (and others) in much greater depth.

10. Compare and contrast suspension and emulsion polymerization processes that have been used to produce polystyrene beads of very different dimensions.

F. E. STUDY QUESTIONS
(1) (2) (3) (4)

SUGGESTIONS

FOR FURTHER

READING

1. Write down a typical chemical structure for the following polymers together with the starting materials (monomers) involved in their synthesis: (a) Polyisobutylene (b) Poly(ethylene terephthalate) (c) Nylon 6,12 (d) Polyacrylonitrile (e) Poly(ethylene oxide) (f) Poly(vinyl alcohol) 2. Give a typical example of a commercial polymers listed above in question 1. 3. What are the major differences addition) polymerization? 4. Nylon commercially. differ? application for each of the six step growth and chain (or

K. J. Saunders, Organic Polymer Chemistry, Chapman and Hall, London, 1973. G. Odian, Principles of Polymerization, 3rd. Edition, Wiley, New York, 1991. H. R. Allcock and F. W. Lampe, Contemporary Polymer Chemistry Prentice Hall, New Jersey, 1981. ' P. Rempp and E. W. Merrill, Polymer Synthesis Heuthig and Wepf, Basel, 1986. '

between

6 and nylon 6,6 are two polyamides that are produced How are they synthesized and how do their chemical repeat units

~\

5. What is transesterification and why is it important in the production of high molecular weight poly (ethylene terephthalate) ? . 6. Describe the chemistry involved in the formation of a urethane group and briefly explain how flexible polyurethane foams are prepared. 7. Describe polymerizations. the four primary features that characterize most addition

8. Under conditions of high pressure and temperature and in the presence of a free radical initiator ethylene polymerizes to produce a high molecular weight polymer which is referred to as low density polyethylene (LDPE). (a) Why is LDPE formed and what chemical reactions are responsible for the formations of this particular polymer? (b) If a 50:50 mixture of ethylene and propylene is substituted for ethylene under the same reaction conditions no high molecular weight polymers are formed. Why do you think this is so ? (c) High density polyethylene (HDPE) also cannot be produced by the free radical reaction conditions mentioned above. How is HDPE synthesized and how does it differ in terms of polymer chain structure from that of LDPE ? 9. Describe the unique feature that distinguishes "living" polymerization from that of ordinary addition polymerization and show how the former can be used to synthesize styrene-butadiene-styrene block copolymers.

in that a different product is formed at a much faster rate.. where the composition and sequence distributions of the copolymer will depend upon the relative rates at which the monomers add to the growing chain end). This is the release of a lot of energy. CA. Copolymerization will be treated in a separate chapter. For example. the chain.7 kJ per mole. can have a profound effect on molecular weight and microstructure (e. it provides information on how long a reaction takes. have a lower free energy).. 1979. while the hills remain. Eisenburg and D. but also to physical processes such as crystallization (see Chapter 8). In chemical synthesis a particular product might be thermodynamically preferred (i. but nothing about its rate.. Physical Chemistry with Applications to the Life Sciences.e. We will see that step-growth polymerizations generally require long times. ordinary sugar (sucrose) can react with oxygen in the atmosphere with a free energy change of about . 63 . but some addition polymerizations occur so rapidly that they could be considered explosive. We will assume that you have a basic familiarity with the fundamentals of • We stole this analogy from D. Lord Tennyson A.CHAPTER 3 Kinetics of Step-Growth and Addition Polymerization "For some cry 'Quick' and some cry 'Slow' But.g.5. In this chapter we will discuss the kinetics of step-growth (condensation) and free radical polymerizations. The relative rates at which different reactions occur (e.. First. leaving the kinetics of other addition polymerizations to more advanced treatments." -Alfred.g. Menlo Park. but ·we have never seen the sugar bowl lying around on our kitchen table suddenly burst into flames". not only to an understanding of the fundamentals of polymerization. Up hill Too-slow' will need the whip. Crothers. but be kinetically unfavorable. the rate at which chains are terminated relative to the rate at which they propagate). Benjamin/Cummings Publishing Co. as this will serve to illustrate the general principles. INTRODUCTION Thermodynamics can tell us whether or not a process or reaction is capable of occurring. but also to its practical implementation by process engineers. The application of chemical kinetics is crucial. because the rate at which this reaction occurs is so slow. which is often all important. in copolymerization. We will find in our studies of polymers that such considerations apply not only to chemical synthesis. 'Down hill Too quick'.

and this probability depends only upon the number of groups remaining. As we will see. Flory's assumption means that the rate of reaction (esterification) in these two mixtures would be the same. in that for very small molecules the rate can be appreciably different. In other words.e. Principles of Polymer Chemistry. The meaning of the Flory assumption is that in this reaction mass the probability of reaction of a group attached to a dimer is exactly the same as that of a group attached to a 20-mer. p. kA' vs chain length. J. at first. say hexane (assuming all these molecules mix with one another). THE KINETICS OF STEP-GROWTH POLYMERIZATION The assumption that the reaction rate is independent of chain length can. at that time many found this assumption implausible. 1. Principles of Polymer Chemistry. so that the relationships we will now describe become invalid. F1ory. essentially the relating of reaction rates (i. after the reaction has proceeded a while there is a smaller number of unreacted functional groups per unit volume. 1953. Cornell University Press. Cornell University Press. but there is another way of looking at some of the same problems through the use of probability theory. we could take an acid with N=3 (i. Clearly. Flory. This is the approach we will take in this chapter. N for CH3CH20H + H(CH2JNCOOH. The rate constant quickly approaches an asymptotic value as chain length increases. First. Figure 3. We will consider this latter approach separately in the following chapter. The amount of diluent would be chosen so that the concentration of functional groups (acids and alcohols) present in this first mixture would be exactly the same as in a second mixture. 3CH2 groups) and react it with an' alcohol in the presence of a given amount of an inert (non-reacting) diluent. Flory investigated polyesterifications of the type: [ A-A + B-B A-AB-B 1 . however. say N=7. shown in figure 3.e. 1953. Carothers had mentioned that he believed that mathematics could be used extensively in the study of polymers and as early as 1935 he presented some of Flory's initial work on molecular weight distributions at a Faraday Society meeting. There are other aspects to the Flory assumption that should be considered.1. reversing the flow of historical events. Clearly. Redrawnfrom the data of P. which are assumed to be randomly distributed in the mixture. 25~----------------------------~ 20 -- 15 10 5 o+-~~~~~~~~~~~~~~~ 5 10 o Chain Length: N 15 20 B. as in the example of the reaction of ethanol with various carboxylic acids. 71. these groups are attached to chains of different length. as there are fewer reacting groups per unit volume. it is not exactly true.1 Plot of esterification rate constant. there may be a non-equilibrium distribution of reacting groups and the rate of reaction becomes diffusion controlled. • P..64 KINETICS OF POLYMERIZATION The Kinetics of Step-Growth Polymerization 65 kinetics. We will discuss these latter results first. be confusing. Even with these experimental results. Flory's assumption" is that the inherent reaction rate per functional group is independent of chain length in a situation where there are equal numbers of reacting groups per unit volume. this assumption does not mean that the overall reaction rate remains constant throughout the course of chain growth. but it is interesting to observe here that the first of many major contributions made by Paul Flory to polymer physical chemistry (which would culminate in the 1974 Nobel Prize for Chemistry) was the theoretical determination of the distribution of the molecular weights of chains formed in condensation polymerizations using probability arguments. Flory had been hired by DuPont in 1934 and assigned to Carothers' group. As we have already mentioned. but Flory subsequently proved his point with a detailed study of the kinetics of polyesterification. where the acid chain length is longer. We will simply make the additional observation that in polymerizations where the reaction rate is fast and the viscosity of the medium is high. the overall rate of reaction of acid groups with alcohol groups (in the example given above) will decrease as the length of the alkyl group attached to the acid functionality increases. It is important to note that if we now go on to consider polymerizations. The problem relates to how we define the concentration terms in the usual rate equations. In deriving his results Flory made the fundamental assumption that the reaction rate constants of the functional groups are independent of the length of the chain to which they are attached. the rate of disappearance of reactants or appearance of products) to the concentration of the various species present and a rate constant. but we will neglect them here in favor of proceeding to an experimental demonstration of the validity of this assumption and refer the interested student to Flory's book. many students still feel that there is something not quite right about Flory's assumption or that it is counter-intuitive.

1 k't = (1 _ p) .p)2 as a function of time should be linear. 61. which is equal to the fraction of functional groups that have reacted after a time t: Number of COOH groups reacted 100 I I II I I I I I I" I I I I I II ." o 1000 2000 Time (mins) Figure 3. but at higher conversions the 'predicted straight line relationship is obtained.d~~l = k2[A][B] (3.4) If Co is the initial concentration of acid and alcohol groups (i.12) . 3334 (1939).6) We will define p as the extent of reaction.-------. and in the absence of any added strong acid a second molecule of the carboxylic acid can act as a catalyst in each reaction step and the kinetics are now third order 1 (1 .10) (3. Flory. between monofunctional reactants).1 (3.. Redrawnfrom It can be seen from the insert in figure 3. however. then we can integrate as follows: (3. A plot of some of Flory's data is shown in figure 3.2 that data for the initial stages of the reaction are not linear. It is important to grasp that [A] and [B] are the concentrations of functional groups not molecules (there are two functional groups per molecule in the example we are considering here). In ion sensitive reactions of this type this behavior can be attributed to the change in the character of the medium as the reaction proceeds and the acids and alcohols are converted to esters.1) 300 where kz is the rate constant.2) If the concentrations of carboxylic acid and alcohol groups are exactly equal we can let: c = [A] = [B] (3. at t = 0).p) 2 '0 . II (3.3) hence..p) 2 _1 (3..for an uncatalysed polyesterification. JACS.66 KINETICS OF POLYMERIZATION The Kinetics of Step-Growth Polymerization 67 where A-A is a dicarboxylic acid or diacid and B-B a dialcohol or diol. Esterifications are catalyzed by acids.2 Plot of 11(1 . (3. = .8) Substituting in equation 3. Further evidence supporting Flory's assumption concerning the reaction rates of functional groups was obtained from polyesterifications catalyzed by the addition of small amounts of strong acid (p-toluene-sulfonic acid).5) to obtain: (3. 1.2. This typically also happens in simple esterifications (i.11) 1 (1 .. I I I I j) / 'I I I" I I 100 I" I I / 200 I I . The concentration of this added catalyst remains constant throughout the reaction and can therefore be included in the rate constant k' l= k2[Acid]}: _ d[A] = k'[A][B] dt (3..e.9) and: Co and plots of 1/(1 . the data of P.6: 2co k3t = 2 Again assuming equal amounts of reacting functional groups _ dc = k'c2 dt (3.pfZ vs time. The kinetic equations for such bimolecular reactions are often of the form: Reaction Rate 400~--------------------------------' 20.e.

and azobisisobutyronitrile A plot of 1/(1 .3 Plot of 1/(1 . Most often. so that polymer is only obtained at very high conversions near the end of the reaction. p ~ 1. lACS. 3334 (1939).68 KINETICS OF POLYMERIZATION Kinetics of Free Radical Polymerization 69 100~--------------------------------. again displays the predicted linear relationship after a certain degree of conversion has been reached. There are several groups of such initiators and two common examples are benzoyl peroxide: 20 o o+---~---r--~--~--~~--r---~~ o 200 400 600 800 Time (mins) Figure 3. so we will consider it at the beginning of the next chapter.p) vs time. free radicals are generated by adding initiators which form radicals when heated or irradiated.. 2R' (3.1 and 3. It can be seen from both figures 3. 80 a) b) c) d) Initiation Propagation Chain transfer Termination (1 .3. Redrawn from the :\. which we can represent as: ---. b and d first and for the moment defer our discussion of chain transfer. This brings us to the kinetics of free radical polymerization. however..13) where kd is the first order rate constant describing t!tis process.. because we think it makes for an easier introduction into the use of probability arguments for obtaining molecular weight distributions. C III N I :N N: C-(CH3h C III N I Azobisisobutyronitrile 2 (CH3 h-9' III N C radicals upon These types of initiators usually give two identical decomposition. shown in figure 3.e.. Flory..p) 40 1 60 It is convenient to consider a. data ofP. 61.. hence 11(1 .for a catalysed polyesterification. The crucial point is that this behavior stands in marked contrast to addition polymerizations. We will not derive this equation here. a topic which seems to give many students a hard time. (CH3h-C (AlliN). where some polymer is obtained almost immediately. KINETICS OF FREE RADICAL POLYMERIZATION We saw in Chapter 2 that the processes occurring in free radical polymerization can be broken down into a number of steps. This radical can now "attack" the double bond of a monomer: C. . Obviously.p) against t.p) ~ very large). This is a general characteristic of step-growth polymerizations or polycondensations and is important because it is relatively easy to show that the number average degree of polymerization is equal to 1/(1 . if the intrinsic reactivity of functional groups decreased with increasing chain length a linear relationship would not have been obtained. Initiation of free radical polymerization can be achieved by a number of methods.p). such as the irradiation of monomers with high energy radiation or thermal initiation without added initiator.. Styrene and methyl methacrylate can be polymerized in this way.)100 .2 that polyesterifications take a long time to reach high degrees of conversions (i. J.

19) Note that just as in polycondensation we have assumed that reactivity is independent of chain length by using the same rate constant.18) M Mi (3. obviously enough.21) Propagation now proceeds through the successive addition of monomers. in general: R+ O (3. The rate of propagation.70 KINEncs OF POLYMERIZATION Kinetics of Free Radical Polymerization 71 We will use the abbreviation M to represent monomer and let the rate constant for this process be k. These two reactions are together considered to be the process of initiation. which can be written as: kId ---J~~ M. combination. is given by . where a proton is transferred and a double bond is formed: ~~ <Q)-~-~ ~~ <QMQ) If we let f be the fraction of initially formed radicals that actually start chain growth. not all primary radicals formed by the decomposition of initiator will react with the first monomer. but this can go in one of two ways.. it is much slower than the addition to the first monomer). so that we can write: or. The first is the simple formation of a bond between two radicals: H ° C-CH2 I I oNV\I\I\I\MI' The factor 2 could be included in the rate constant. ri: (3..20) Termination always involves the reaction of two radicals.16) rp = -~ d[M] =~ [M ][M] ° (3. In the case of benzoyl peroxide. if one was of a mind to do so.e.17) Mx + M . for example. for each step. rp' or monomer removal. In our development of the kinetics of these polymerizations we can usually assume that the first of these two reactions (decomposition of initiator) is the rate determining step (i. then the rate of initiation is given by: r i o H ° C-CH2 I I X oNV\I\I\I\MI' H H-C-CH2 X These reactions can be represented schematically as: I I oNV\I\I\I\MI' = __ d[M°] dt 1 = 2fk [I] d (3.14) (3.22) . However.15) where the 2 is obtained because: d[I] 1 d[M~] -= --= k [I] dt 2dt d (3. disproportionation. so that we can write for the rate of initiation. . + My (disproportionation) (3. ~."'/ O O ---J~~ Mx+y (combination) (3.. several other reactions occur: X H I X I H I X I oNV\I\I\I\MI' . The second termination mechanism is called.t\I\I\IVWII'CH2 -C-C-CH2 <Q)-~_oo o IR\_II ~C-o o ° This is called.

because of its dependence on ~/ ~1/2.. The reason for this is that the viscosity of the medium increases as polymer is formed. = ~c + ktd and the factor 2 is included because two radicals are consumed in each termination reaction. we simply increase the monomer concentration and the rate of reaction should increase proportionally. but proportional to the square root of the concentration of initiator (i.72 KINETlCS OF POLYMERIZATION Kinetics of Free Radical Polymerization and after substituting in equation 3. The "reactivity" of the radicals is not altered. and then the degree of conversion as a function of polymerization time-things that are nice to know.27) ~1f2. Third.05. The rate of heat evolved from these exothermic reactions also increases and in situations where this is not quickly dissipated there can be catastrophic consequences. If we reexamine equation 3.. This is not always practical or desirable as this will reduce the molecular weight of the product. we see that the rate of polymerization is still proportional to [M][I]olf2. This phenomenon of where k.23) (3. Consequently. as we will show shortly. in other words. is reduced.20: (3. ·This has a dramatic effect on the rate of polymerization.23: (3.e. just monomer and initiator and maybe some inconsequential additives) would require significantly increasing the initiator concentration (because of the square root dependence). At 130aC and a pressure of 1 bar this ratio has a value of 0. At low conversions it is reasonable to assume that the concentration of initiator. because the concentration of radicals during a polymerization would be a difficult thing to measure experimentally. r.26 the rate of polymerization becomes: 73 Both these reactions involve two radicals and are kinetically identical. radicals must be generated at the same rate as they are consumed. so that the apparent value of k. First.28) (3. . For this to be so. however.25) This is necessary in all the equations that follow. We have already mentioned one consequence of this for the specific case of ethylene polymerization. [I]. ri' must equal the rate of termination.16: _ d[I] = k [I] dt d We can integrate to obtain: (3. More initiator would have to be added to keep·the polymerization going (or we start with the right amount for our purposes and recover polymer product before the reaction slows down too much).7 at 2500 bar. Termination involves the much slower diffusion of larger macromolecular species. To reiterate. ~. or the rate of initiation. so we can write for the rate of termination. We can now obtain expressions for the rate of polymerization. in the absence of catalyst the rate of propagation is too small relative to the rate of termination for ethylene to polymerize at ordinary pressures. From equations 3. Polymerizations in the bulk (i.17 and 3. The relative proportions of disproportionation and combination depend upon the type of monomer involved and the temperature. The second point is that the polymerization will slow down in an exponential fashion as the initiator is used up (final term).e. which depends upon the barely affected diffusion of small molecules. We can then immediately obtain an expression for the rate of polymerization (which is equal to the rate of propagation) by substituting in equation 3. although now we are talking about the initial initiator concentration. is constant.. one-half order).29) where we have broken the equation into three parts for the purpose of our discussion.24) which allows us to obtain an expression for MO: [Mo] = ( T)1f2 fk [I] (3. rt r t = _ d[Mo] = dt 21r[Mo{ '''1 (3. in this case MO. For methyl acrylate this can occur after a conversion of only 1 percent and can result in an explosion. but a more precise equation can be obtained by accounting for the consumption of initiator. but this increases to 0. but the ability of the radicals to find each other is. however. In concentrated solutions or in the polymerization of an undiluted monomer there is often a marked acceleration of the polymerization rate at some point in the reaction. we see that the rate of polymerization is proportional to ~/ ktll2 (first term in brackets). if we are dealing with a solution polymerization and we wish to speed up the rate of polymerization. Increasing the temperature to 200°C at this latter pressure increases the ratio to 3. and this increase in viscosity can result in a large decrease in the rate of termination.26) This tells us that the rate of polymerization is directly proportional to monomer concentration (or. This does not affect the rate of propagation. The first expression we will obtain is an approximation based on the often used assumption in kinetics of a steady state concentration of a transient species. There is another consequence of the dependence of the rate of polymerization on ~/ Various kinetic experiments performed upon polymers in solution have yielded the expected first order dependence on monomer concentration. is first order in monomer concentration).

Schematic plots of conversion (or % polymerization) vs time illustrating this are shown in figure 3. Finally. Chern. When the active site is transferred to monomer.[M]) where [M]o is the initial monomer concentration.(1 . Instead. conversion is never complete as it is always less than I by an amount that is given by the exponential term. monomer adds very quickly and a complete polymer is formed. . cited earlier).4. But unlike condensation polymers. so there will again be a distribution of chain lengths. however. We conclude here by noting that chain transfer to certain types of molecules called inhibitors can retard or prevent the onset of polymerization for a while. another radical is generated and the total number of active sites is not altered.e-kdt/2)} (3.2~ { (f)l12 kX [1]0 112( 1 . There is a continuous' distribution of species present. we obtain: Obviously. often in a matter of seconds.exp . • is the concentration of monomer that has reacted).[M]o (3.34) The degree of conversion is equal to ([M]o . In discussing condensation polymers we can use (next chapter) a statistical method to obtain expressions for the molecular weight distribution. 1106 (1967). that are being stored and then remove them (by distillation or by passing through an appropriate packed column) when they are to be polymerized.74 KINEllCS OF POLYMERIZATION Kinetics of Free Radical Polymerization or: Conversion 75 autoacceleration is often called the Trommsdorffeffect (although it was originally discovered by Norrish and Smith. polymer itself. so that ([M] 0 . etc. Macrornol.30) Substituting the expression for ~ given in equation 3.= [M]o .( ~ Y"[1]:"} (3. = 1 . In a condensation . for ~hefree ra~ical polymeri~atio~ of methyl methacrylate at 50"C using benzoyl peroxide at various concentrations In benzene. then the rate of polymerization is barely affected.33) d[M] dt (3. and at a given degree of conversion we have a mixture of polymer chains and monomers (although there will obviously be some oligomeric species from chains that by chance terminated quickly). there is no continuous distribution of species. because of the stepwise nature of the polymerization and the homogeneous nature of the reaction mass. Harborth. in this section we have so far ignored the effects of chain transfer upon the rate of polymerization. Obviously. it is a good idea to add such inhibitors to monomers such as methyl acrylate. etc. If we let t go to infinity we obtain an expression for the maximum conversion that is less than 1 by an amount that depends upon the initial initiator concentration: Maximum-conversion In [M] = 2~ [M]o ( kd~ f )112[1]0112.e. We can now obtain an expression for the degree of conversion as a function of time by simply noting that: R __ AI! - . if the radicals so generated are unable to react with monomer. in free radical polymerizations an active site is formed by an initiator.29 and integrating.31) = 1 . initiator. Average Degree of Polymerization in Free Radical Polymerizations o 500 1000 1500 2000 Time (mins) Figure 3.[M])/[M]o' the fraction of monomer that has been reacted (recall that [M] is the concentration that is left or unreacted .kdt 12) (3. ranging from monomer through polymer and all species take part in the reaction.v.32) If we had assumed a steady state concentration of initiator. and we will discuss that in the next section. solvent. see Morawetz. We can then obtain: Conversion . If these new sites add monomer.[M] [M]o [M] = 1 .4 Plot of conversion vs time. . We do not generally have a situation where shorter chains can latch on to one another to form longer chains. Redrawn from the data ofG. The length of this polymer chain depends on the random radical encounters that lead to termination. Schulz and G. The degree of polymerization is altered. then the maximum conversion would approach 1 after long periods of time. Chain polymerizations are entirely different. styrene. This figure also shows the autoacceleration effect that can occur at high monomer concentrations.exp -{ 2k...

e.e. It should be intuitively obvious that in a reaction vessel containing monomer and initiator (and perhaps some inert solvent) the average length of the chains we will obtain will be inversely proportional to the initiator concentration (or some power of the initiator concentration). Obviously. i. The kinetic chain length is clearly equal to the number average degree of polymerization of the radicals formed at the particular time in the polymerization that we have chosen..20 to obtain: v = _:. which is therefore all we will attempt./ 76 KINETICS OF POLYMERIZATION Kinetics of Free Radical Polymerization 77 polymerization the average degree of polymerization increases continuously during the reaction and when we calculate say ~. The rate of dead chain formation is given by: _d[M'] dt = 2k [M. providing that we have started with the correct proportions of monomer. the number average degree of polymerization. initiator.000). which is equal to the average number of dead chains formed per termination. For the moment we are neglecting what happens upon termination. We will start by neglecting transfer reactions. but it is a fairly straightforward procedure to calculate a number average chain length. then as a polymerization proceeds and initiator is used up the number of radicals present decreases and the average polymer chain length of the product will be greater. then we could simply obtain the number average degree of polymerization of these chains by dividing one by the other (i. This is something we want to know. This initially formed polymer then usually sits around inertly while other polymers are formed. as the rate of propagation divided by the rate of initiation at some instant of time: If we use the steady state assumption.~-:::. and also how many monomer units have been polymerized. We get a more accurate representation of what is occurring if we consider shorter and shorter time periods. These could have a different molecular weight. We can obtain this by the definition of a kinetic chain length. In contrast. If we generate a smaller number of initial radicals then on average the polymer chains will grow longer before their active ends meet and terminate (i..38) where the factor 2 is included because disproportionation results in two dead chains. Xn = v ~ = 2v (termination by disproportionation) (termination by combination) Since both mechanisms of termination can occur in certain polymerizations. chain transfer agents. it is useful to define a parameter ~. (3. This must also be equal to the number average degree of polymerization of the polymer. if 1. it reflects all the species present in the reaction vessel and is a function of the extent of reaction only.37) where we see that there is a first order dependence on monomer concentration. It also follows that if we do not have steady state conditions. if termination is by disproportionation. _ rp v =r. we can first substitute from equations 3.40) Accordingly. We can obtain an expression for ~ by again using a kinetic argument and noting that this must be equal to the rate of dead chain formation divided by the rate of termination. I (3.e.17 and 3.e.-2fk)I] then substituting for [M'] from equation 3. the longer the time period we choose. however. the more things will be different at the start than near the end. there is a smaller number of radicals per unit volume).41) ..000 monomers have been used up and 100 chains formed in this time period..]2 td + k [M. v. in a free radical polymerization we have polymer of the required molecular weight immediately. as we will shortly demonstrate more explicitly).. etc. Because of the wide variety of side reactions that can occur in free radical polymerizations the theoretical calculation of distributions is difficult. then subsequently modify our result to account for these. If for a given period of time we know how many growing polymer chains have been started. but that the average chain length varies as the inverse of the square root of the initiator concentration. have active sites on one end).000.25: ~[M][M'] (3. as this does not change the number of chains nor their length. i. etc. [M. as monomer is used up. the number of chains would halve and their length would (on average) double. and this is in effect what we do by defining a kinetic chain length.39) (3.]2 .e.]2 Hence: = (~e + k.d)[M.. (note that the degree of polymerization obtained in a free radical polymerization depends upon far more variables than a condensation polymerization. so that we can judge when we have polymer of the required molecular weight.]2 tc (3. then the number average length of each chain would be 10. If termination were exclusively by combination however. so we can consider these chains to be radicals (i.36) (3.35) Xn = (2k ~ [M][M'] td + ~)[M']2 = (~) ~(fkd~)lf2 ( [M] ) [I]If2 (3. In this situation what we want to know (and actually all that we can determine) is the degree of polymerization of the product. the instantaneous number average chain length can now be obtained by dividing the rate of addition of monomer units to the rate of "dead" polymer formation. The rate of termination reactions is simply given by: k. as initiator is depleted and monomer is used up. not the average of the polymer and monomer in the reaction vessel at a given time in the reaction.

where (xn)o is the average length of the chain in the absence of chain transfer agent (cf.4_I) and C is simply kull). There IS very good agreement between theory and experiment (in this case). is deliberately added to control molecular weight..44) The Effect of Chain Transfer As we noted at the start of our discussion of free radical polymerization. The effect of chain transfer on the instantaneous number ~verage ":.5 Effect of solvent on XII on polymerization of styrene atlOO°C. where a growing chain radical is terminated and a new one is initiated in its place. R. A plot of It'Xnagainst [T]/[M] should then b~ linear WIth a slope of ku/kp and an intercept 11("")0' This equation hasbe~n appl~ed to the polymerization of styrene in various solvents.. and so on.95 11.IN sodium hydroxide solution...328 (1947). STUDY QUESTIONS I... Mayo. We have to strike a balance for the product we are making. weighed.2978 5. quenched in a dri ice/acetone..3402 0.:. because on average shorter chains are produced.45) [Note: it is not always a proton that is transferred.82 16.92 11.78 KINETICS OF POLYMERIZATION Study Questions 79 It is important to note that the rate of polymerization is proportional to [I]lf2. Time (min) 25 50 100 .. equat. 2. In other words. dead or terminated polymer (producing long chain branches).62 10. such as a mercaptan.. Such reactions are common and generally take the form: M~ + R-H where [T] is the concentration that: of the chain transfer reagent.. l). Discussions..5.. they can). dissolved and titrated WIth a standard O.ion 3.28 11. monomer. A lot of components in a polymerization can act as chain transfer agents.09 o o 10 20 [S]/[M] Figure 3.42) o [M] (3.35. Redrawn from the data of R.9007 6. and the results are shown schematically in figure 3.300 460 500 600 700 Sample Weight (g) 0.. weight is easily accounted for by adding the effect of dead cham formation by this mechanism to equation 3..7 12. which act as cha~n transfer agents.p) = Co k't + constant 2.~--1 8.. Sometimes a chain transfer agent. introduced l?tO tared.5118 5. .2 + ~r[T][M'] (3. Then: d[M'] .150 200 1__ . At various intervals small samples are removed..6 5 1.. these reactions do not affect the rate of polymerization (although in certain circumstances.[M] ~ = -----~-----~ (fkd~[I])If2 +k)T] Rearranging: l=_I_+C[T] __ 'lIo). A. which is not always desirable.- R'+ M R-M' etc. ~ (~) (3.1387 Titration (cm3'O. Show that the kinetics of polyesterification catalysed by a strong acid may be represented by: 11(1 .-dt= 2kld[M] . there is also the possibility of chain transfer reactions..38.1377 4.1014 1.5113 2.. From the results given below determine whether this was a catalysed or non-catalysed reaction.58 30. the solvent.32 10.] In general.olecular. A large scale polycondensation reaction of a diol and a diacid is ~erfonned at ~OO°C. _ _ .2 + ~JM] .88 10.4931 0. if we try to speed up the polymerization by adding more initiator we end up getting shorter chains. comcal flasks.4501 0. while xn is proportional to V[I]If2.43) It follows directly (3. Gregg and F. but obviously affect the molecular weight.IN NaOH) 46. D. Faraday Soc.

' of the initial product were obtained as a function of initiator concentration. Given that for a certain monomer thermal generation of free 2 . it has been shown that in a number of systems the rate of polymerization. Rp. J.100 0. R. Organic Chemistry of Synthetic High Polymers. second order in monomer concentration) such that: M+M --4)1o~ R' + inert products (a) Show that the rate of polymerization varies as [M]2. New York.] 8.3 moles!l) at 70°C.20 1. which is a technique that measures volume changes. and the number average degrees of polymerization.20%. Ithaca.e. xn rn (g /1 oOml) 0. Cornell University Press. Given that there is an 18% volume contraction when styrene is completely (100%) polymerized. The following initial rates of polymerization.061 0. calculate: (a) the value of kt / kp (b) the efficiency of AIBN as an initiator (c) the value of C (= ktr / kp) State explicitly what additional assumptions you make in performing these calculations. initiated by AIBN (molecular weight = 164. measured in terms of % contraction per hour. Why is this? 6. Wiley & Sons. Flory. 1967.5 X 10"5 sec+). . (b) Obtain an expression for the kinetic chain length and comment on the dependence of n on [M].48 2. radicals involves two molecules of monomer (i.352 Contraction (% / h) 0. = 8.84 1. Ethylene does not polymerize at ordinary pressures under conventional free radical processes.09 Xn 860 625 500 363 Assume that termination occurs solely by disproportionation and that chain transfer takes place only to monomer. Assume that this holds for the bulk polymerization of styrene (initial concentration. In this chapter we only considered the initiation of polymerization by chemical methods. is approximately constant for conversions of up to 10 . [IJ [MJ 4. Interscience. New York. SUGGESTIONS (1) (2) FOR FURTHER READING P. Principles of Polymer Chemistry.185 0. J. ~enz. kI = 3. W. [M]. Furthermore. What effect would increasing the initiator concentration have on the degree of polymerization in the bulk free radical polymerization of styrene? 5. [Hint: first derive an expression for xn in terms of Rp and other constants.80 KINETICS OF POLYMERIZATION Suggestions for Further Reading 81 3. Describe the following giving appropriate examples: (a) Chain transfer (b) Retardation (c) The Trommsdorff effect (d) Inhibition (e) Termination by disproportionation 7. The rate of many free radical polymerizations can be followed fairly accurately by dilatometry. Define all symbols employed and show that the rate of polymerization for a free radical process is given by: fk ]1/2 1/2 Rp = kp [ k. x E. 1953.

Springer-Verlag. Galton found evidence to support this. 83 . you would calculate that the probability of getting a head in any coin toss is 50/100 or 0. because of fluctuations. (In truth. Watson and Galton (in 1874+) considered the problem of the extinction of particular family names. but we also suspect that they did not account for the unfortunate effects of in-breeding. In any event. Berlin. For example. if you toss an unbiased coin 100 times and get 50 heads. Here we will confine ourselves to very simple probability arguments and about the only tool you will need to start with. They came to the conclusion that all such family names would quickly die out. The Theory of Branching Processes. In a subsequent study of the reproductive rates of the British Aristocracy (whose records go back a long way and therefore provide a good data base). Pj E] is given by: prE} = !if NE (4. apart from a modicum of common sense. when we discuss copolymerization.CHAPTER 4 Statistics of Step-Growth Polymerization 'The true logic of this world is the calculus of probabilities" -James Clerk Maxwell A.5. which (in those days of moral and ~ intellectual darkness when women took the names of their husbands) was related to average fertility and the probability of having sons. Harris. For example. We will consider this + See T. having its origins in the desire to work out the odds in games of chance and thus get rich without having to work for a living. However.1) where N E is the number of E events occurring and N is the total number of events. Even the parts of it whose origins are not apparently connected to gambling can seem a bit weird. the result of their efforts was the development of the theory of branching processes (or Markov chains). if you actually tossed a coin 100 times you would be unlikely to get exactly 50 heads. INTRODUCTION Probability theory has a long and disreputable history (which is why it appeals to us). 1963. is the definition that the probability of an event E. We will use the simplest part of this theory in the next chapter. E. the original analysis contained an algebraic error.

This really involves practically no statistics at all and in most texts (e. the probability that anyone of these molecules has reacted after a certain time is simply equal to the fraction that have reacted in the same time period. we have found it useful pedagogically to start our discussion of statistics with this topic by noting that the extent of reaction. then the probability of the other event must equal 1 .) Also. the A group could be an acid. because if we make a mole of polymer we have billions and billions of chains. (Of course. we started with 20 monomers and ended up with 4 (short) chain molecules. if I go out on Friday night).. as these are particularly well-suited to a statistical treatment (because all the species present are . if there are only two possible outcomes. We will start by examining molecular weight averages and distributions in step-growth polymerizations. We could indeed have four chains of equal length. The same holds for polycondensations of Type II. II A-B II A-BA-B I IA-BI Figure 4. . defined in the previouschapter as equal to the fraction of functional groups of a particular type that have reacted (e. If. but it is a powerful tool in the study of such aspects of microstructure as copolymer sequence distributions and various types of isomerism. by titration or spectroscopic methods). I A-BA-BA-BA-BA-BA-B I I A-B I I A-B I I A-BA-BA-B II A-BA-BA-BA-BA-B I A-BA-BA-BA-BA-B I I A-BA-B I I. acid groups in polycondensation). this assumes that reactivity does not change with chain length.1 Schematic representation of n-mers. We will discuss these latter applications of probability theory in Chapters 5 and 6. This says nothing about the actual distribution of the chain lengths.1 and counting the number of molecules and the number of (say) A end groups. more or less.5 you would have to perform the 100 coin tosses a large number of times and average all of the results. the first involving bifunctional monomers that have different (but complementary) functional groups: [ A-B II A-B I I A-BA-B I I A-BA -BA -BA -B I I A-BA -BA -BA -B I ~======::::::_-__::::::. the average length of each chain would be 20/4 or 5. p. Over the years. is also equal to the probability that one such group.g. THE STATISTICS OF LINEAR POLYCONDENSATION [ A-A + B-B A-AB-B (TypeI~ In reactions of Type I we can "count" the number of molecules present if we can measure the concentration of one of the end groups (the acid group. Remember. -BA -BA -BA -BA -B I .) We can consider two types of polycondensations. equal to No.involved in the polymerization throughout the course of the reaction). This counting of the number of molecules present can be used to provide a measure of the number average degree of polymerization. A feel for the breadth of the distribution can be obtained by also We will start by considering the number average degree of polymerization in a linear step-growth reaction. The second type of reaction involves monomers that each contain only one type of functional group: I A-BA-BA-BA-B I . what is the probability that I will drink too much and do something stupid. for example. taken at random from the reaction vessel.A-BA-BA-B I A-B I + A-B A-BA-B (TypeI) 1 where. N: (4. A-BA I I. so the probabilities all work out. This is not always grasped immediately by students who think about probabilities in a sloppy manner (like. We don't have to worry about that here.84 STATISTICS OF POLYMERIZATION The Statistics of Linear Polycondensation 85 some more when we talk: about chain conformations. B. providing that we start with exactly equal equivalents of the two reactants. the B an alcohol and the BA product an ester. has reacted. This follows directly from the definition of a probability given in equation 4. Xn' which is simply defined as the total number of molecules originally present (monomers).P'[E]. for example. when we are talking about reactions in a vessel containing billions and billions of molecules. in the example we have used. but to get the expected result of 0. divided by the total number of molecules in the system after the polymerization has been stopped. Students whose brains have not yet engaged in forward gear at the time of reading this section can convince themselves of this by examining figure 4. like in a coin toss. The nature of addition polymerization is such that the statistical approach cannot be readily applied to the analysis of molecular weight distributions. Note that the probability of an event is equal to the number fraction of these events and is always between 0 (never occurs) and 1 (always occurs).g.2) This definition of an average degree of polymerization is easily grasped.1. but a distribution where one of the chains contained 17 units while the other 3 units remained monomers would also give the same average. however. Flory's) is included in the discussion of kinetics. A-BA-BA-B.

c.6 a "chain stopper" (a monofunctional reagent.2. then the total number of monomers initially present is equal to: f II a Z-(CH2)6-Z-C-(CH2)4 -c 0+ II n 1 (1 .the reaction has to be driven almost to completion (p = 0. the chemical repeat is: particularly useful material). This would give us a polymer with a number average degree of polymerization of only 10 (and this would not be a 0. ~. is related to the initial concentration of such groups by: (4.3) c = c (1.4 0. We will come back to this shortly. To obtain a degree of polymerization of the order of 200. we can use: _ _ ~=Moxn=(l_p) Mo (4.p) x.No) Co can be simply converted to c 1 1-p (4. This requires that we know two things.995).6 it is desirable from the point of view of subsequent processing to limit the molecular weight to the 10. p Figure 4.0 Conversion.co) so that: _ No x =-=-=-n N (N.6 In terms of a degree of polymerization as we have defined it here however. o b) the number of molecules concentrations (c. which would make our hypothetical chemist hop up and down with glee. .000 range ( in"" 100-150). If we wish to obtain a number average molecular weight. polymerization would recommence when the product is reheated during processing. instead of a degree of polymerization. We will start our discussion of this topic by considering the effect of a nonstoichiometric equivalence of the bifunctional monomers.9). Our intention is to obtain an expression for the number average degree of polymerization. Accordingly. is not much better (in = 20).Although this could be achieved by stopping the reaction at the appropriate degree of conversion by simply cooling.only achieved at high degree of conversions. First. = (1 _p) o~~~--~~~~~~~~~ It tells us that high molecular weight is . which displays a plot of in as a function of conversion (P).6 0.0 0. this chemical repeat unit contains two structural repeat units (it has been made by joining two monomers). present after a fraction p has been reacted. We will designate the groups that are present in excess by the letter B. but it is often advantageous to have a more precise knowledge of the actual distribution (many physical properties can be affected by this). but it is necessary to do this reproducibly on an industrial scale! These results are summarized in figure 4. so we will write it out again and put it in a box so you will hopefully remember it: 1 100 x. versus conversion.6. For nylon 6. In nylon 6. p. For condensations of Type II the chemical repeat unit contains bits of both monomers.4 is an important one. so that if N A is the number of A groups and Nn the number of B groups initially present. we must define Mo as the mean molecular weight of a structural unit (in the above example this can be obtained by dividing the molecular weight of the chemical repeat unit by 2). As it turns out. for nylon 6.p) 200~------------------------------e Chemical repeat of Nylon 6. xn . the number of monomers we started with (No) and the number of chains after a fraction of p groups has been reacted (N).8 1.2 0. The result obtained in equation 4.2 Plot oj the number average degree oj polymerization. we can express in terms of the conversion p by recalling that: a) the concentration of one of the functional groups..000-15.86 STATISTICS OF POLYMERIZATION The Statistics oj Linear Polycondensation 87 measuring the weight average molecular weight. In their laboratory experiments many organic chemists would be happy with a 90% conversion (p = 0. A conversion of 95%.4) where the conversion constants cancel.5) Be careful here. for example. acetic acid) is added to the polymerization and this brings us to the general subject of the control of molecular weight. Not only that.

. It is sufficient for our discussion of general principles to limit our attention to polymerizations of A-B type molecules where.. is an x-mer (i.9) N No 1 +r = 1 + r ... MOLECULAR WEIGHT DISTRIBUTIONS CONDENSATION POLYMERS IN LINEAR (1 .. Therefore the number of B units unreacted (equal to the number of chain ends that are B units) is given by the second term...e.p) p x-I (4.. p ..10) In this section we really start using probabilities and we will derive expressions for the molecular weight distributions in Type I condensation polymerizations....p) + (1 ~rp) ] (4. where the extent of reaction is equal to p. we determine the probability that a molecule selected at random from a mixture of polymerizing chains... Flory... and so on: A-BA -BA -BA -BA p p p ... Equivalent expressions can be obtained for Type II. N= N. Cornell University Press.... X.. the number of B units that have reacted must be equal to pNA... so that we can write: * P. (\\ Following Flory.1) AB linkages in an x-rner..14) .. Principles of Polymer Chemistry. = (1 ... equal to px-I.. there is an excess of just 1% of B units at the start of the reaction (by which we mean NA/NB = 99/100). Px' is therefore: If.... Because NA molecule reacts with exactly one Nn unit.. The factor 2 in equation 4... not reacted) is equal to: (4.p) p x-I (4. by evaporation) in preference to the other. .e. do not have monofunctional contaminants) and that one of these monomers does not get lost or escape (e.. if it is desired to obtain high molecular weight polymer great care must be taken to make sure that there are equal amounts of monomer at the start of the reaction. which is equal to). complete reaction) to show that there is a theoretical upper limit to the molecular weight that cannot be exceeded if there is an excess of one component: _ [1""l:-r xn -) .e. We focus our attention on the first B unit in the chain and ask "What is the probability that this has reacted?" This must..11) There are (x . providing that r is now defined by the following equation (see Flory*): P.--B We can now let p = 1 (i. a monofunctional reactant such as acetic acid (CH3COOH)].p. The probability that the second group in the chain has reacted is also equal to p. I-p +r ] as p -) 1 (4...8) where the origin of the first term should be self evident.. for example. be equal to the mole fraction of x-mers present...88 STATISTICS OF POLYMERIZATION Molecular Weighl Dislribulions NA r = ----'"'--M N +2N B 89 (4. be equal to p (to reiterate.13) The probability of finding an x-rner at random must. of course. 1953.... The same equation can be applied to a polymerization involving equal numbers of bifunctional reagents and a small amount of a "chain stopper" [e.. the probability that a group has reacted is just equal to the fraction that have reacted).. = (1 . ..... The second term is equally straightforward.. that these monomers are pure (i.. so the probability that all these groups have reacted is just the product of each term.. has x units in its chain).g. but the algebra can get much more complicated when the number of A-A molecules does not equal the number of B-B molecules...2rp (4. We must also include a term that gives the probability that the end B group has not reacted... The probability that the molecule has exactly x units.. C..e. of course. the number of A functional groups is always equal to the number of B groups.. then the upper limit for xn is 199.5 is there because the number of monomers is equal to half the number of (bi)functional groups. Substituting r = NA/Nn and dividing by two to obtain the number of molecules: N = t[NA(I-P)+Nn(1-rp)] = Substituting and rearranging: xn = ~A [ where N: is the number of monofunctional groups (note that N A= N B)' Clearly.7) B NB and can only take fractional values between 0 and 1. by definition.. After a fraction of p groups has been reacted the number of chain ends (not molecules) remaining (i.12) where: NA r=(4. J.g.

the weight fraction of x-mers is then: wx=x(1-p) 2 p x. Rather than use the mole fraction of x-mers. is the mole fraction of x-mers. Figure 4.l (4. where N 'is the number of x-mers present in our polymerization pot.90 0.14 to calculate the mole fraction of x-mers present for various values of p. By weight.3.14 and 4. ThIS can be s~ow. which can be obtained from equation 4.3 Mole fraction distribution of x-mers.) Substituting from equation 4. 0.06 Xx -: p = 0. the amount of monomer can be very s~all (obviously a single 100 mer is much heavier than 5 or 10 monomers). some authors prefer the number distribution Nx. N is the total number of molecules present when the extent of reaction is equal to p and X.14 using: .17) We can now use equation 4.10 .03 Wx 0.4 Weight fraction distribution of x-mers.90 0. An unusually sharp fractionation was obtained many years ago by Taylor.(1 -p) hence: and: (4 15) .17.08 P = 0.4 for the same values of p that were used to calculate the mole fraction distribution.!< 90 STATISTICS OF POLYMERIZATION Molecular Weight Distributions \ 91 0.p) (4..n more explicitly by calculating the weight fraction of x-mers present. The distribution described by equations 4.19 is called the most probable or Schulz-Flory distribution. The validity of this assumption has never been contradicted by experiment. but until the advent of gel permeation chromatography (Ope see later) there were few studies of the molecular weight distribution. It can be seen that there is always a larger number of monomers present than any other species and this is true of all stages of the reaction (i. OH groups etc..N .04 1 x Figure 4. This is because of the difficulty of cleanly separating a polymer into its fractions by methods such as precipitation from solution (you tend to get fractions that also contain a distribution of molecular weights). The distribution also broadens with increasing extent of reaction. where we have assumed that the averJge molecular weight of the structural repeat unit is the same as the monomer aV6rage (this introduces a small error through neglect of end protons.04~-----------------------------------' 0.e.----------------. wx' WhIChIS given by: wx = = weight of all x-mers weight of all units (4.18) Values of Wx as a function of x are plotted in figure 4. .02 0.16) (4._ No __ 1_ "n . The results are shown graphically in figure 4.5. It can be seen that there is a maximum in this distribution that shifts to higher values as p increases.19) N = N o (1. It rests on the fundamental assumption that the reactivity of a functional group is independent of the length of a chain to which it is attached. however.. all values of_p). It can be seen that the agreement of theory with experiment is within error. however. and these results are shown in figure 4.

and commonly find.25) 0. MULTICHAIN x. lACS. 69. Substituting from equations 4. 0. As the polymerization reaction approaches completion.003 ( • Expt Data ) 1 x" = (1 . may be determined-as we will see later in Chapter 10 ).9900 and p = 0.·I 92 STATISTICS OF POL YMERIZA nON Multichain Condensation Polymers ~ x-I k.24) 0.22) X'n Because p =LxpX-I(1-p) (4. The polymer formed upon complete reaction will consist of f chains attached to a central unit R. ~. and informative The number equations for the most probable distribution given above can be expressions for the number and weight average degree of The former has been given already. the ratio of the two is called the polydispersity and is a measure of the breadth of the distribution. Redrawn from the data of G.28: ~ =x" =~1-+ p) ~ (4. but it is usually interesting to obtain fundamental results by different methods. B.001 we obtain: _ (1 + p) x w =--.004.e. we would expect.Jxp = to obtain: 1 (1 .------------------. to be roughly twice that of M. if we let X.6 polymer for molecular weight analysis by say both osmometry and light scattering (from which number and weight average molecular weights.. results showing M. linear species will disappear in favor of multichain polymer molecules. however.20) We have seen that for an essentially complete reaction (i. =2 (4. p= l) the polydispersity of polymers synthesized by linear polycondensation is predicted to be 2. from chemical side reactions. the polydispersity generally broadens. (1 _p) (4. R-Ar.26) ~ 2 x-I (1+p) k. Thus.23) < 1 we can use the following equation for a series convergence: .14: 2N2 3N3 + -+ -+ --------------LNx LNx (4. respectively. Consider the condensation polymerization of A-B with a small amount of a multifunctional monomer.27) (4.29) Finally.p) The weight average degree of polymerization is given by: (4.21) As above.5 Comparison of calculated weight fraction distributions (p = 0.6 data.. be the mole fraction of x-rner (NxlLNx)' it follows (4.p) 2 93 (4.Jxp = 3 (1 .28) 1 100 200 300 X 400 500 600 Figure 4. 638 (1947). average degree of polymerization is defined as: LxNx Xn = ~ and asp ~ 1 =x" D. As mentioned in Chapter 1. if we were to submit a commercial nylon 6. If branching reactions occur. where the polydispersity narrows.9925) to experimental nylon 6. The weight average molecular weight (degree of polymerization) is always larger than the number average (they are only equal in a monodisperse sample. the used to obtain polymerization.. impurities or by design using multifunctional monomers.002 Xw = LXwx Using: = Lx2px-I(1-p{ (4. There is an interesting polycondensation case.25· and 4. Taylor.p) 0. where all thechains have exactly the same length).30) CONDENSATION POLYMERS s: Nx (4. which in its expanded form is: _ NI Xn = -LNx that: substituting from equation 4.

large and small sizes. Say we initially perform a linear polycondensation to essentially complete conversion.13.P 1 x-I P (4. we must take a step up in "degree of difficulty". because of attachment to the RAe molecule.1 bonds. To link them together would require f .p) If you think about it.36) * W. Unlike the linear case..32) For example: e-aminocaproic yield schematically: I A-R-A A acid (A-B) plus a tetrabasic acid (RA4)would then the probability of obtaining a f-anned x-mer may be expressed as: x-I e Px... In our reaction pot is a distribution of linear chains with a polydispersity of 2.13 describing the probability of reacting x-I groups to form a linear chain. where P.1) + (f . this is difficult to grasp at first.1) things taken at a time..e..31) ppp p (1 . It is exactly the same as equation 4. while the total number of units in the entire molecule. different x-mers can have their units distributed amongst the f arms in many different ways (e.94 STATISTICS OF POLYMERIZATION Multichain Condensation Polymers Now. In other words. however.I)!] [ fp + 1 . the number of combinations of (x . Y3.I)! (x ..2)! [ (f-I)! (x-I)! The mole fraction ofx-mers thus becomes: (x + f .. such cooperation among statistically uncoordinated compounds will be comparatively rare and it is much more likely that the collection of molecules we select at random will contain molecules of average. but note that one more A group has reacted in this case.whatsoever is therefore equal to equation 4. The probability that the molecule contains exactly x units distributed over f chains in any manner. 125 (1944) . (This is similar to equation 4.34) If you are not used to combinatorics. could be equated directly to mole fraction of x-mer species. but instead of a final term of (1 . R..35) ] (4. only small (or large) molecules.) The probability that the f chains have lengths YI' Yz. As before. the polydispersity decreases.e = [ (f . As the selection of the chain lengths was performed randomly. the distribution will be narrower than conventional linear polycondensation (i. is counted as one of the total x: YI + y/+ Y3+ ---------.p) we have (1 . Phys . depends upon the sum of the y-values of the individual chains. holding an RAe molecule in one hand. Stockmayer. < 2).32. Accordingly. YI' yz. Chern.e. if the central unit. Now. The problem is therefore reduced to the number of ways of distributing f .1 bonds amongst the x-I units.1) really short chains and one really long one. This can be worked out by imagining the f blocks arranged in a line (of x-I units) instead of in a star-like fashion. Now. A size.33 multiplied by the total number of combinations of y's that fulfills the condition expressed by equation 4. Y3..Ye. much larger or smaller than average is only likely to occur if several (or all) of the chains in the molecule are abnormally large or small. i... it is very unlikely that we would select. x(x+f-2)! (l-p) e+ I wx.+ Ye = (x . here we must also include the condition that the lengths of the individual chains. because there are f end groups in the f-anned star.p). If this is not immediately obvious. (f .33) A I A-B + Large excess Small amount Tetrafunctional monomer ---:~~ A -(BA) -BA Y4 ?-(AB)~A -R-AB-(AB)-A AB-(AB)y-A Complete reaction I Y1 2 Tetrafunctional star The total number of molecules will equal the number of tetrafunctional units in the system.p) (4.2)! ] x -I e Xx. We don't care how the units are distributed. x.e = [ (f _ I)! (x _ I)! P (1 .p) (4. we are just counting those star molecules whose total number of units is equal to x. From simple combinatorics this is equal to: (x + f . We now repeat this until all the A groups of the RAe molecule have attached chains.g. and so on). this result is straightforward.• 12. J.f = p (1 . In fact. So far we have been able to derive all our results from the simplest type of probability argument.1) 95 (4. look at it this way. all the same length. While the average molecular weight of the fully reacted star shaped molecules increases vis-a-vis the original linear molecules. The probability that a particular chain contains y units is pY(I . we will define p as the probability that an A group has reacted. can vary from x-mer to xmer. x. so we won't discuss the derivation of the corresponding equation for the weight fraction of x-mers and we will just present the result derived by Stockmayer". we reach into the pot with our molecular tweezers and withdraw a molecule at random and attach it to one of the A groups of the RAe molecule. ~. ye' respectively is: Y1 Y2 Y3 Yf e (4.p)e. for example..

has a somewhat different reactivity than the primary OH groups. have the same reactivity. are 1. however. After some time.~ A-(BA)n-BA-R-AB-(AB)m-A the polydispersity is predicted to be 1. Flory. For f = 2. respectively. we now need to define some terms.Bmolecules to this system. using a number of simplifying approximations it was shown by Stockmayer that: B-BA >-AB-BA A B >-AB-BA-A Z~(l++) (4. this would lead to typical branched structures like: It may be helpful for some students to visualize a reaction pot in which we put known amounts of A-A. Our purpose will be to define the conditions under which an infinitely large chemical structure or infinite network will occur. but is actually a simple concept as we illustrate below: E. in his analysis. The latter assumption is not as reasonable. but fortunately. 1. which lead us to our next major topic. Flory defined a parameter. For example. and we will see that intramolecular cyclization does lead to significant errors. Consider a polycondensation reaction involving two difunctional molecules and one trifunctional molecule. . H04:::H{-CH(OH)-CH2-OH.20 and 1. which he termed o. The former is a reasonable assumption. Examples include phenolic resins. "What is the probability that the molecule selected at random has the structure shown above?" This may seem a rather strange question to ask. We again impose the restriction that reaction can only occur between A and B.:. which is equivalent to the case of combining two independent chains into one molecule. via a chain of bifunctional units..5.. as illustrated below: A A-A + A >-A + B-B A ---lJDooIID- A A>-A [B-BA--A}-BA >-A l=Otooo Definition of ex. In this section we will be concerned with defining the critical conditions for the formation of infinite networks and molecular weight distributions for non-linear polymers. If we were to add B. Enough philosophy. where it is assumed that all functional groups. THEORY OF GELATION The presence of polyfunctional units nearly always presents the possibility of forming chemical structures of macroscopic dimensions or infinite networks. B-B and RA3• After a certain reaction time let us pretend that we have a pair of molecular tweezers and can randomly pull out a molecule. This may appear to be a mouthful.) / It should be emphasized that in this section we have considered a rather special case of the polycondensation of A-B molecules in the presence of multifunctional molecules that contain only A functional groups.25. polyurethanes. (Incidentally. regardless of where they are situated. and second. A-(BA)n-B A I A I B A B ~ A >-AB-BA A >-A A I B B + A-R-A + B-(AB)rn-A ----:l. which is the probability that a given functional group of a branch unit leads. that intramolecular condensation does not occur.37) This simple equation illustrates nicely the point we wish to make. We now ask the question. gel formation in diene polymers etc. to another branch unit. invoked two major assumptions: first the principle of equal reactivity. network structures are possible. the polydispersities predicted for A-B polycondensation systems containing small amounts of RAf molecules with functionalities of 4. 5 and 10.96 STATISTICS OF POLYMERIZATION Theory of Gelation 97 This looks truly awful. but bear with us as this is an example of the type of conceptual leap made by Flory that separates the extraordinary scientist who shakes up the field from the mundane scientist who merely extrapolates. that as f increases the distribution sharpens. an analogous situation is present in free radical polymerization when chain termination is exclusively by combination.10. although it is known that the secondary hydroxyl group in glycerol.

] Hence: • For those students who have never observed the onset of gelation.44) jPBP The probability that an A group is connected depicted above is thus PA[PB(1.p). Hence: The probability that a B group has reacted with a branch unit is PBP. or: rpAp [ 1.) As in the case of linear polycondensation B groups which leads to: [Note that if there are more A groups than B.42) . I • . . when the bubbles stop rising and appear frozen in place. when there are no A-A groups present. as both r and P are determined by the concentrations of the initial ingredients and the unreacted end groups. ~ . P be the ratio of A's (reacted and unreacted) belonging to branch units to the total number of A's.43) Second.=0 ~x=~x i 1 =-I-x we will let r equal the ratio of A to (4.four chains are produced and then eight and so forth. may be determined analytically by a variety of techniques at various stages of the reaction. There are a two special cases that deserve mention as they further simplify equations 4. = (4.. PB be the probability that a B group has reacted.rp!(1 .P)PA]iPBP. when A and B groups are present in equivalent amounts. (4. • I I It is important to recognize that it is a relatively straightforward task to calculate a. . One can observe the nitrogen bubbles rising monotonously for a long period of time (typically 2 hours) and then there is an remarkably abrupt loss of fluidity at the incipient gel point. as a function of conversion. PA J_. is given by: (4.40) when x < 1. then it follows that the probability that a B group has reacted is greater than that of the A group. Schematically: (For those of you who wish to work this out. - [t -p APB(1. note that. Now examine the chain linking two branch units: a. adipic acid and a trifunctional acid is carried out in a glass reaction pot containing a fine capillary tube that bleeds nitrogen bubbles into the material.=0 [p APB(1. [ PB(1_p)PA]i •• I I t . differing reactivities such as the secondary hydroxyl in glycerol. is given by: = (4. which is less reactive than the primary hydroxyl groups. r = 1 and PA = PB = p. . = t . Flory reasoned that at some critical value of a.41 and 4. J . and a.41) PA be the probability that an A grouphas reacted. a. First. Similarly. we recommend that a rising bubble experiment be performed. Let us first consider a trifunctional branching unit. is therefore given by: to a sequence of units as Note that the above treatment is not completely rigorous.38) This is another of those summations that converge to give: _ The Critical Value of a. The parameter a.42. are not taken into account. Each chain which reacts with this unit is succeeded by two more chains. incipient formation of an infinite network" will occur and this will depend upon the functionality of the branching unit.98 Let: STATISTICS OF POLYMERIZATION Theory of Gelation 99 (4. The precision and reproducibility of this measurement is quite amazing.p) ]ip APBP (4.p) ] PBP [r-p!(1P)] 2 2 a. A simple polycondensation involving say ethylene glycol. The probability that a B group is connected to an A-A unit is PB(1 .=1 i-I PAPBP a. P 1 and a. . which is a truly spectacular phenomenon. In addition. .39) Now we come to the clever bit. For example.p) ] ~ . the polycondensation of R-Af and R'-Bg cannot be handled. A or B. if both of these react with the branching unit.

. 481 (1941). Stockmayer and L.5. if f = the functionality of the branching unit. Principles of Polymer Chemistry.02 were obtained which is again significantly higher than the theoretical value of <Xc = 0. Thus we reach the important conclusion that when <X < 0. We have intimated that the discrepancy between the theoretical prediction and experimental observation of <X at the gel point is primarily due to the fact that intramolecular condensation was not taken into consideration in the development of the theory. in other words. ibid.46) Kienle and coworkers" studied the reaction between glycerol and equivalent amounts of a number of dibasic acids. etc. <X = 0.58... but only a fractional amount of indefinitely large structures exist beyond the gel point. The value of <X at the gel point was then calculated using equation 4. Note also the very rapid rise of the viscosity as the gel point is approached. but not all of it.1) exceeds unity.2258 ibid.6 we show a schematic diagram depicting the variation of PA'.5 represents the critic. p = 0. In fact. H..5 ± 1% esterification.100 STATISTICS OF POLYMERIZATION Theory of Gelation 101 higher than the theoretical <Xc 0.60 ± 0. An accurate experimental value of p. In other words. ibid.x. 63. 1. On the other hand. L. Flory" studied a different system. I Accordingly.1 condition for the incipient formation of an infinitely large network in a trifunctionally branched system. For this case p = 1 and <X = p2.5 this does not mean that all the material is combined in an infinite structure. there is a less than evenchance that each chain will lead to a branch unit and thus to two more chains. B. (1939).. is not very large.50. Chapter 6 in Advancing Fronts in Chemistry. For the four systems studied. Weil. gelation will occur when <x(f. 61.. S. values of <X = 0. 1053 (1940). It is very important to recognize that when <X > 0. Both gel (infinite network) and sol (molecules of finite size) exist in varying amounts.5. at the gel point was obtained by extrapolation. This remarkably simple expression illustrates that <Xc is dependent solely on the functionality of the branched molecule. 1953. Ed. 1945. Molecular stuctures of an infinite size (networks) are now possible. This discrepancy is primarily caused by the failure of the theory to account for intramolecular condensation. HOOC-R Now if <X < 0. Flory. allowing for a sharp transition which is experimentally easy to observe. all molecular structures must be limited. if <X > 0. Reinhold. In figure 4. Substituting for p yields an experimentally determined value for (X of 0.P+ f(I-P++)+2P Experimental Testing of the Gelation Theory <X +- 2p A) + 2p (4. there is a greater than even chance that it will end at an unreacted functional group. HO-CH2-CH2-OH Ethylene glycol '-COOH = Adipic or succinic acid + HOOC-R-COOH COOH Tricarballylic acid I . In another form: (4.5.45) where <Xc is the critical value of <X. Flory also calculated the number average degree of polymerization from: ~ = ~~----~~---f ( 1. termination of chains must outweigh continuation of the network through branching. The gel point was determined experimentally by the loss of fluidity and the absence of rising bubbles.5. H. . W. . the reaction between diethylene glycol with adipic or succinic acids (bifunctional) and varying amounts of a tricarballylic acid (trlfunctional). all the molecules in the pot are of finite size.765. each chain has a better than even chance of reproducing two more chains. 62.____/ Recall that r and p are calculated from the relative amount of each component used initially and PA can be determined by withdrawing small samples and titrating to obtain the extent of esterification of the acid groups. 61. Note that x. HO-CH2-CH-CH2-OH OH Glycerol Dibasic acid I + HOOC-R -COOH The incipient gel point was observed to occur at 76. nor is it increasing rapidly at the gel point. JACS. Some of the discrepancy was no doubt due to the different reactivity of the secondary OH group. An elegant set of experiments by Stockmayer and Weil" showed • P." and viscosity (11) as a function of reaction time. Eventually..41. Cornell University Press. In general. This is significantly • R. This means that at the gel point many molecules are present. 2268 (1939). Kienle et al. New York. Twiss..

if A-A. 0.4 § c. Interestingly enough. '" 0 24 16 0. These authors studied the reaction pentaerythritol (tetrafunctional) and adipic acid (bifunctional). i. After some reaction has occurred typical structures such as that shown below (a 7-mer) will be present: A cursory look at this structure might lead you to the conclusion that an infinite network is inevitable. or A-A and B-B molecules are added to the A-R-Br_1 system. the polycondensation of the monomer A-R-Br_l. it was found that = the gel point occurred when p = 0. Again. The extrapolated value of p at (c-I)c=oo was determined to be 0.. This is again in line with the previous results. This is not the case. Cornell University Press..e.578 ± 0.. It is included because it is an interesting special case of branching which currently (1996) happens to be undergoing a revival of sorts with the renewed interest in so-called hyperbranched polymers.6 0.. in general. 1953..e. J.8 tI '0 Ix" '0 32 § 0 I1.-CH2-DH T Pentaerythritol As pentaerythritol is tetrafunctional the theoretical critical value of alpha is <Xc 0. Redrawn from the data of P. This corresponds to a Pc = 0. Principles of Polymer Chemistry.6 Schematic diagram of a typical polyesterification involving a trifunctional monomer.63. in a classic experiment the authors reasoned that the probability of intramolecular condensation would increase if the system was diluted with an inert solvent.. if bifunctional A-B monomers are added to the A-R-Br_1 system the essential character of the molecular structure remains identical.0 0. It is important to note that now stuctures are formed which have more than one A group and this leads to the possibility of networks. Experimentally. Consider. '" o '" :> 0 .2)x + 1 unreacted B groups. analogous polymers formed from A-R-Br_1 will always only contain one unreacted A and (f .. A___/ ~A-A + Different structures . at (c-I)c=oo. however. this indeed to be the case.... A series of experiments were performed in the presence of different amounts of such a solvent and p was obtained at the gel point in each case. if you will. we immediately introduce the possibility of forming infinite structures... Essentially the same structure However..2 0 0 50 100) 150 200 250 Time (mins) Figure 4. .l '"""' . a result higher than predicted by theory. RANDOM BRANCHING WITHOUT NETWORK FORMATION 40 1.102 STA nSTICS OF POL YMERIZA nON Branching Without Gellation 103 F. However.. Flory. A may only condense with B and vice versa. i. Note that each x-meric species contains (x + 1) unreacted B groups (8 for the 7-mer shown above) and only one A group.eo Gel Point t This is the final system that we will consider in this chapter.005 which is in remarkable agreement with theory.333.where intramolecular condensation would be expected to be completely eliminated. CH2-DH CH2-DH between In fact.577. HOOC-(CH2)4-COOH Adipic acid I + HO-H2C. The results were then plotted against inverse concentration of the inert solvent in order to obtain an extrapolated value of p at infinite concentration. A )3 + ~ A-A --l .

a. but finite molecules will be present at high extents of reaction.a. but wltho~t the derivation which is outside the scope of this book. The extent of reaction for the linear polycondensation of A-B is defined as p. to another branch umt-and here there are no bifunctional units and a.95 and 0. Finally. Recalculate the number average molecular weights as a function of conversion and display your results graphically.48 into 4. Write down the probability Px that a molecule selected at random contains exactly x BA units.52 moles of terephthalic acid are reacted with 0. 1 x" = [1 .45) that the critical value of a. The gel point was by the rising bubble technique.49. This broadening of the molecular weight distribution is thus in stark contrast to the narrowing discussed previously for the system consisting of A-B and a small amount of RAr.c = (f . (c) Comment on the significance of your results. 6. -the probability that a given functional ~oup of a branch unit leads. The interested reader IS referred to Flory's book for the details -.51) Define the symbols used in the above equations porydispersity of this system approaches two at high conversions.1) (4. cannot equal a.95 and 0.1) more B groups than A groups. A polycondensation reaction is performed using 46g of glycerol and 109.88 ml of the standard NaOH was reqired to neutralize this sample.(f -1) ] (4. that. The extent of reaction can approach 1 but not attain it.a.a.90.requation 4. polydispersity values of approximately 6.(f . Calculate the weight and number average molecular weights as a function of conversion for the linear polycondensation of the aminoacid: NH2 .1) We have previously seen.1) ] Xw = 2 (4. is calculated from equation 4. = PB (4. as there are (f .104 STATISTICS OF POLYMERIZATION Study Questions 105 Let us now see if we can use the principles that we have developed in this chapter to explain why the A-R-Br_1 system does not form ~~finite netw~rks.1) (This is self-evident.r and weight average degree of polymerization for the A-R-Br_1 s~stem. . 2. 11 and 51 are calculated for PA values of 0.. is: 1 a. for this reaction and compare your result to the theoretical critical value of a. is thus simply PB') Now: P A = PB (f .52) [ 1 . = (f . (4.COOH (a) Select appropriate values for the extent of reaction and display your results graphically.99. 5. (f . Comment on the significance of your result. a 20 ml was removed and 5.1) ] The polydispersity is thus given by: Kw ~ 1 . Calculate the value of the parameter a.47) G.1) 2 = 1 .(CH2)7 .99.50 moles of ethylene glycol.5g of adipic acid.48) w x = x (1.5 ml of the solution was required to neutralize the sample. (b) l g of acetic acid is added per 100g of the above aminoacid.1) (4. 12. (c) Describe the significance of the graphs you have drawn. 0.c' as this requires p A to equal 1. Just before the gel point. via a chain of bifunctional units. thus the formation of network structures (gelation) is not possible! Very large.47 leads to: PA a. respectively.2(f .53) For the polycondensation reaction of A-R-B2' where f = 3 and a.a.50) (4. for completeness we will present equations for the numbe. what is the number average degree of polymerization at 90 and 100% reaction? [ 1. 3. STUDY QUESTIONS 1.a. For the polycondensation of A-R-Br_1 we know from the definition of a. A 10 ml aliquot is removed at the beginning of the and titrated with a standard aqueous NaOH solution.) Substitution of equation 4.49) (a) Define all the symbols/used in these equations / (b) Sketch two curves-showing the relationship between Nx and Wx as a function of x using two values of p = 0.p) p 2 (x .(f . The number and weight distribution functions for linear polycondensation are given below: (Remember the definition of a. If 0. The number and weight average degrees of polymerization polycondensation are given respectively by: for linear xn = L x Xx and Xw = L x wx and show that the The magnitude of a.1).

One application of these materials was found to be as "cut proof' outer covers for golf balls. Ithaca. [IV] and [V] (d) Compare the expected polydispersities of [III]. R(NH2b. Ethylene-co-methacrylic acid (containing » 5% methacrylic acid) is an example. say A and B. block. statistical) copolymer of the two. were prepared by condensation polymerizations that were taken to high conversion. Principles of Polymer Chemistry. according to the arrangements of units in the chain. J. RCOOH. the property obtained being dependent upon the exact arrangement and the relative length of the blocks. thus making them one of the most momentous discoveries of this century! Clearly.aliqu. all based upon the bifunctional monomer e-aminocaproic acid [NH2 (CH2)6 COOH]. Five different polymers. but a random (or more precisely. The feed composition. 8. or a coordination complex (and could be a B rather than 10. R(COO~)5. is an elastomer. Ionomers consist of statistical copolymers of a nonpolar monomer with (usually) a small proportion of ionizable units. at what extent of reaction will the mixture th~re?cally gel ? (c) How would you explain experimental results that indicate a gr~ater extent of reaction is required to reach the gel point than that predicted theoretically? 9. and we will do so here. Copolymerization allows various monomers to be combined in such a way so as to provide materials with useful and sometimes unique properties. is added [N] A small amount of a pentafunctional acid. 0.0 moles of phthalic acid.NH2 groups that is incorporated in a polycondensatIon reaction with bifunctional adipic acid. Calculate the extent of reaction at the gel point for a mixture of ~oles of pentaerythritol.ot is remo:ed and it is determined that the extent of reaction of carboxylhc acid groups IS 0. is added [III] A small amount of a trifunctional amine. denoted * in the schematic overleaf. The single atoms to each other tend. by including two monomers. SUGGESTIONS (1) FOR FURTHER READING P. [IV] and [V] (e) If polymers [I] and [II] are mixed together what would be the effect on the polydispersity ? H. 1953. The protons on the acid are exchanged to form salts and these ionic species phase separate into domains that act to cross-link and toughen the polymer.106 S"fATISTICS OF POLYMERIZATION 7. Copolymers come in various forms. and graft.func~ional . Chain growth polymerization can be used to make statistical copolymers. A polycondensation reaction involving an aminodicarboxyllic acid (A-R-B2) is performed. differed in the folllowing manner: [I] No additional reagents [II] A small amount of a monofunctional acid. IS added U sing reasonable assumptions and with appropriate justification: (a) Compare the relative molecular weights of [I] and [II] (b) Compare the expected polydispersities of [I] and [II] (c) Describe the type of polymers that are formed in [ill]. alternating. After a certain period of time an . GENERAL OVERVIEW At the beginning of this text we defined copolymers as those containing more than one type of chemical unit in the same chain. Calculate the polydispersity of the polymer at this stage and comment upon the value you obtain. Cornell University Press. (NH2hRCOOH. is added [V] A small amount of a diaminoacid. Block copolymers of styrene and butadiene can be used to make thermoplastic rubbers or alternatively. and we defined four types in Chapter 1: statistical (random). for example. attracted to. A chemist synthesizes a cyclic molecule .43. Flory. the next in place' Form'd and impell'd its neighbour 10 embrace. denoted [I] through [V]. Attract. it is useful to spend some time discussing copolymerization. an impact resistant rigid material. We will initially consider some examples of the types of copolymers that can be produced by various chain and step-growth polymerizations and then examine the kinetics of free radical polymerization in more detail.5 moles of ethylene glycol and 2. ?8 CHAPTER 5 Copolymerization "See plastic Nature working 10 this end. in the appropriate proportions. The active site. ." -Alexander Pope A. Linear ~ polyethylene and isotactic polypropylene homopolymers are both semi-crystalline plastic materials (we will discuss exactly what we mean by semi-crystalline later).containing six . in the same reaction vessel. 107 . Copolymers are synthesized using the same types of chemistry as described in Chapter 2. New York. could be a or ionic species. (a) What is the critical value of exfor the onset of gelation ? (b) If equivalent amounts of acid and amine function~l groups are originally present.

scrambling any initial distribution of species imposed by kinetic factors. into a chain and then grafting onto these sites a second polymer with an end group Y that reacts with X. which for generality we will label M. In principle. For example.e. making sure that monomer is used up in each stage before adding a different batch.. The macromers can then be reacted to form chams (in this example. but at the expense of a significant increase in algebra. This can either add another unit of type 1 with a rate constant given by kll (i. initiation and termination are much the same as in homopolymerization. Similar equations can be written for the addition to a type 2 radical: .e. Then we will turn our attention to the powerful application of simple probability theory to sequence distribution and the characterization of copolymer chains. linked by ester groups). Blocks. The first two apply to the reactions that can occur if the active site (radical) at the end of a growing chain happens to be monomer of type 1. have been extensively studied and we will now examine this in more detail. for example can be made by first separately polymerizing short polymer chains (macromers) with one type of functional group at each end (sayan acid) and a different type of macromer with a distinct but complementary functional group at each of its ends (sayan alcohol). polymers with different sequence distributions could be produced at different catalyst sites.108 COPOl YMERIZA TION The Copolymer Equation 109 B.e. It is the general principles of the relationship of copolymer composition to the initial concentration of monomers and the kinetics of chain addition that we are seeking to get across here. are produced in a similar fashion. particularly free radical. and it is the propagation step that gives copolymerization its special character. but there may be no way to distinguish and separate chains produced from one site from those produced at another for the purpose of analysis. Step copolymerization of different types of units in the same vessel usually leads to truly random copolymers which usually have the same composition as the feed (i. we add them sequentially. irreversible. We will examine this in more detail shortly. The kinetics of chain growth copolymerization. we can make block copolymers. Usually this results in a statistical distribution of units in a chain that it is possible to describe in terms of rate constants (actually. in a polymerization that uses a heterogeneous catalyst. because even here we will see that the composition of copolymer chains can vary dramatically throughout the course of a polymerization (i. we will confine our discussion to free radical copolymerization. for example. as in chain growth polymerizations. Alternatively. A sequential procedure can also be used to make graft copolymers by first copolymerizing a unit with an appropriate functional group. Free radical copolymerization involves the same steps as ordinary free radical polymerization. These copolymers are usually truly random because in nylons and polyesters. propagation and termination. For the polymerization of a binary copolymer we can distinguish four possible propagation reactions. step-growth reaction. and here the kinetics of the polymerization will be the deciding factor. it should be possible to apply these or similarly derived equations to other types of polymerizations. as a function of conversion). X. The full scope of this difficulty should become more apparent as we discuss simple free radical copolymerization. We will focus our attention on the types of copolymers that can be produced by copolymerizing two comonomers.. so that a truly random product is formed. free radical sites on polymer chains swollen in monomer can be generated by irradiation. if we start with 50% A and 50% B units in a pot (the feed). then the resulting copolymer has 50% A and 50% B units). The copolymer equation should be applicable to each type of site. THE COPOLYMER EQUATION The distribution of species in the chain will depend upon the rate at which one monomer adds to an active site relative to the other. In the final product all the different types of chains are jumbled up together. their ratios). The free radical sites then add monomer. We will see that a knowledge of the kinetics allows the prediction of the instantaneous composition of the resulting copolymers. and Mz• It is possible to extend the treatment to copolymers with three or more units. providing we use a living polymerization technique (usually anionic). reflecting the rate of adding a type 1 to a type 1 radical) (I) (II) where the rate constant k1Z is a measure of the rate which a type 2 unit is added to a type 1 radical. This brings us full circle. This may not hold true when chain growth is a rapid. in this case taking advantage of the facile urethane chemical reaction. but in doing so there are often significant complications that are beyond the scope of ~at we want to cover. so we will neglect these complications.. initiation. Thermoplastic elastomers. transamidation and transesterification reactions occur. With one or two exceptions. If instead of including the monomers in the reaction vessel at the same time. as in the in-situ formation of polyurethanes that are used in reaction injection molding (commonly referred to as RIM). /Step or condensation polymers can also be used to make copolymers. which amongst other things are used extensively in the automobile industry. In the same vein.

ill .2) We can now write equations for the rate of consumption of monomer 1 and monomer 2: --cit --cit d[~] d[MI] 0 0 = kl1[MI][MI] + kzl[M2][MI] = ~[~][M2] (5. nth A I' ' .respectIvely. 2) The terminal model is adequate in describing most copolymerizations. If we accept the general validity of the terminal model. we can immediately describe some interesting limiting conditions: 1) If kll » kl2 and k22» k21' then a chain whose active site happens to be a type 1 monomer unit always prefers to add another 1. of type 2 radicals.1) This gets algebraically complicated. defined below) with initial monomer composition. means that the radical M. that the rate of addition of monomers depends only upon the nature of the radical species at the end of the growing chain.2 and then defining reactivity ratios: . preferential solution of polymer chain by one of the two monomers). k21 ~ o/VVIMIW'M i 1 1 (III) v'IIV\MN'Mi + M2 k22 ~ v'IIV\MN'Mi (IV) Here we have made a fundamental assumption.3) + kI2[M. an VVVWVI/' 2 are formed and removed at equal rates.5) where the top and bott?m parts of the equation have each been divided by [~]. so that: species VWVVVIIIi'M' d of t?e radical .110 COPOLYMERIZATION The Copolymer Equation 111 [-M. while the concentration of monomers is usually known (at least at the begl~mng of t~e reactio~). so we obtain an equation for their ratio: ~ (5. alternate explanation to penultimate effects can often (but not always) be invoked (e.. in turn. A reru:angement Into a more useful form can be obtained by substituting from equation 5. Here we have more than one radical species.re~ctions I and IVa type 1 radical is converted to another type 1 radical and a similar thing happens to type 2 radicals Th . In other words.][~] (5. I o ] .4) The ratio of these two equations is then: (5. the concentration It w. we get the same equation by considering the rate of formation and. based on the same type of kinetic argument. while the ends whose active sites are 2's always prefer to add other 2's. ~e radicals of one type generated from the other (i.) Note that we have let the concentration of c[Mh~]n radical. We can obtain an equation for the instantaneous copolymer composition by applying our old friend. This. In . ere IS no e c ange. to the radical species "IIWWVII'Mi and v'IIV\MN'Mi present in the polymerization. 2) If kI2» kll and k21 » k22 then there is a large tendency for the copolymers to be alternating. the ability to calculate copolymer composition and sequence distributions is a valuable thing (as is the ability to measure them.. 3) If kl2 = kll and kzl = k22 then the copolymers will be truly random. the steady state assumption. to make sure the model is applicable to the system you happen to be interested in). This states that in the (short) time frame that we will be considering (remember that a free radical polymerization can initiate. whether it is a "IIWWVII'Mi a VVWWV'Mi or This is the so-called terminal model.other). disap?earance. We will ignore the penultimate model for now. or even homopolymers (if the tendency is so great that one type of monomer unit almost never adds a unit of the other type).][M2] = kzl [M~][MI] (5. Thus there is a tendency to form block copolymers. as we now have to consider eight rate constants instead of four. /' kI2[M. pp ymg the steady state assumption we can therefore say that the rate of genera~on of new species of type I (reaction III) is equal to their rate of removal (reaction IT).eneral we do no~ know the former and they are very difficult to me~sw:e. [ + M. forming one and removing the. propagate and terminate all in a matter of seconds). If we exarrune reactI?ns I through IV we immediately see that in only two of these.g. It has been proposed that penultimate effects may be important in certain systems and a penultimate model has been constructed (see later) where there is now a dependence on the character of the final two units in the growing chain: v'IIV\MN'Ml-Mi or v'IIV\MN'M2-Mi etc. 3) In situations where it is apparently inadequate.1 (<?bVlOusly. II and III. where there appears to be a variation in the values of the ratios of the rate constants (the reactivity ratios.(hopefully) be recalled that in our discussion of free radical polymenz~tIon we used the steady state assumption to obtain an equation for the concentr~tlOn of radical species in terms of the concentration of monomer units because in g. '. species is constant. for three reasons: 1) The terminal model illustrates the general principles that most concern us.specles VWVVVIIIi'Mi VlNlilNV'Mi be represented by [M and and 2 .e. Clearly.

the ratio of monomers found in the polymer chains (or segments of those chams) that are being produced in this very small time period and is therefore called the instantaneous copolymer composition equation. Equation 5. the copolymer composition usually varies throughout the polymerization (except for a special case we will consider below) and differs from the monomer "feed" concentration.9 and 5. In this situation a perfectly alternating copolymer is produced until all of the monomers are used up (if [Mtlo [M:Jo where the subscript 0 indicates the initial monomer concentration) or until one of the monomers is used up. monomer 1 will only add to MI radicals and monomer 2 will only add to M2 radicals..e. The relative proportions of unreacted monomer ajter. It also means that we can not employ equations 5.10) preference of a radical species for the monomers.7 or 5.10 we can obtain: rl~ + fl2 FI = -. C. For example.e.7. experimental methods for the measurements of rl and r2 have been established.J. a measure of the amount of monomers that at that time has not yet reacted).' e The reactivity ratios that we have employed in the copolymer equations extremely important quantities.Y0lymerization has proceededfor a while may be very different to the proportions at the very start of the polymerization (e.6) to obtain: (5. Hence two homopolymers are produced. k21 = O. rl is a measure of the relative preference of [M.FI and 1 . so this equation must be equal to .] for MI relative to M2 (= kll / kl2 ' the rate of adding 1 divided by the rate of adding 2). If FI is the mole fraction of monomer 1 that is being incorporated into the copolymer at some instant of time.g.7. Similarly.9) from the definition of a mole fraction.7 expressed simply as: (5.8) where x is the monomer feed ratio: [MI] x = [~] These monomers are not just disappearing into thin air. so that in principle we should be able to calculate copolymer composition. This means that we have composition drift. as these provide a good "feel" for the significance of the reactivity ratios' and the results we might expect from using the copolymer equation.e.11) This means that a radical of type 1 never wants to add to itself (i. of course. a radical of type two never wants to add to itself either (k22 = 0). rl and r2.=--=-~----O-=---. but are being incorporated into the polymer. at which point polymerization stops (i.fl' respectively. Special Case I: r1 = rz = 0 Combining equations 5.J. rather than the four individual rate constants kll etc. If in~tead of the ratio of the monomers in the polymer.-.7) This equation is the amount of monomer 1 that is being polymerized relative to monomer 2 in a time dt. First. one monomer could react far more quickly than the other and is therefore used up faster). To reiterate. 5.112 COPOl YMERIZA TION Reactivity Ratios 113 (5.:.7 tells us the ratio of monomers in the polymer at some instant relative to the ratio of the concentration of monomers in the reaction mass (i.[M:Jo) = Special Case II: r1 = rz = 00 The quantities F2 and f2 are. . Accordingly. This is where we come to the next problem. the effect of the reactivity ratios on the types of copolymer produced can be calculated and we will now consider this in more detail. kll = 0). Similarly: [MI] (5. Second. they are a measure of the relative :rr If rl = 00 then kl2 must be equal to zero. We also sometimes see equation 5.11 by simply treating [Mtl / [M2] as a constant over the course of the polymerization. it is useful to consider some limiting cases. REACTIVITY RATIOS AND COPOLYMER COMPOSITION Before getting down to the nitty gritty of calculating copolymer composition for some real monomer pairs.represent the only two independent rate variables that we need to know. . we wish to know the mole 'fraction of each. k21 are also not equal to zero). Finally. but each can add to the other (if k12. we can use a rearrangement of equation 5. the copolymer equations relate the proportions of monomer incorporated into the copolymer at some point in time during the polymerization to the monomer concentrations present at the same time. these two quantities.jwhile fl is the mole fraction of monomer that is left in the reaction mass at that same instant. Nevertheless. if [Mtlo i:. then: (5. simply given by 1 . for r2 = 00.rltl + 2flf2 + r2t2 (5..

unless we have the special limiting case that rl = r2 = 1. The resulting copolymer not only has a truly random distribution of monomers. In practice.. Furthermore.e m?nomer IS Incorporated into the polymer exactly in proportion to ItS composition In the feed. The II_lost re~uent1y used depend upon algebraic rearrangement of the copolymer f eq~a~on to give a f?rm which allows linear plots. This happens when rl. We will present an example of this below..by analogy to the situation where a vapour has the same CO~pOSIt10n as a liquid with which it is in equilibrium. means that the copolymer composition changes during the polymerization). The point where the curve intersects the diagonal is termed ~h.e.114 COPOLYMERIZATION Reactivity Ratios 115 Special Case HI: r1 = rz = 1 This is a very special case. Of course. and means that monomers 1 and 2 add with equal facility to either type of radical active site. b. against fl (or. then the curve that crosses the diagonal can be obtained.13) 5. The special case rl = r2 = 1 ~ves the straight line along the diagonal of this plot. This does no~ tell us ex~ct1y how the copolymer composition varies as a function of conversion for a grven starting monomer composition (we will consider this later). equivalently. A more precise description of the effect of the values of the reactivity ratios can be obtained by making plots of the instantaneous copolymer composition. The curve that IS above the diagonal over the whole composition range represents a situation where the copolymer formed instantaneously is always richer in units of type 1 than the monomer mixture from which it was polymerized. FI "# fl). There might be only 10% A units in the chain. MI and M2 radicals. of course. then again. but the copolymer composition is exactly the same as that of the initial monomer concentration and stays so throughout the course of the polymerization (i. but the copolymer composition is not usually the same as the composition of the monomers in the reaction mass (i. > 1 and r2 < 1. there is no composition drift.e. how many people really do?). because they don't have a good grasp of probability theory (but. In certain systems ideal copolymerization (rlr2 = 1) has been approached. so that d[~tl/d[M2] = MliM2• ~~der this condition.ut does tell us the copolymer composition we will obtain for some given proportion of ~onomers ". Examples are shown in figure 5. but the sequence distribution in the polymer is nevertheless random. To see how this comes about. (i. Iri other wor1~' no matter what the radical active site happens to be. Here the copolymer being formed has the same composition as the monomer feed. This is confusing to many students.e. A random placement does not necessarily mean there has to be 50% A units and 50% B units placed according to coin-toss statistics along a chain. kll k . F2 against f2). then there will also be composition drift (which. special case II (rl = r2 = 00) has not been obtained.eazeotro~e. we first write out the condition rlr2 = 1 in terms of the rate constants and rearrange the terms to obtain: 21 k=k 12 22 Figure 5.1.1 Schematic diagram of instantaneous compositions of copolymers. . In an ideal copolymerization the distribution of units in the copolymer formed at some instant is usually different than the distribution of monomers in the reaction vessel at that time. as will soon be apparent. you will recall. the rate of adding MI relative to an M2 is always the same (it could be 100:10 for kll/k22' for example).1. but if we accept this then we can start by first (5. but in most cases rlr2 is less than 1. but providing that the probability that an A is located at any point along that chain is in this case 0.12) This means that each radical displays the same preference for one of the monomers over the other. where both rl and r2 are very small). beca~s.o)' Special Case IV: rIrz =1 This situation is called an ideal copolymerization. If both rl and r2 are less than 1. This results in a random placement of the units. ?i The Determination of Reactivity Ratios defining: Reactivity ratios have been determined by a number of different methods. the? the copolymer equation has the simple form shown previously in equation . but there are circumstances in which special case I (r. . By this we mean a plot of F.8..e. FI = fl = (fl). (Ideal copolymerization) (5.. this assumes the valI~ty of the terminal model. the ratio of the rate of adding an MI to an MI radical to the rate of adding an M2 to an MI radical is exactly equal to the equivalent ratio of adding an MI to an M2 radical relative to an M2 to an M2 radical). The distribution of monomers in the chain at any point in the polymerization is truly random. then the distribution of units is random. = r2 = 0) has been approached (i.

3 Schematic diagram of a Fineman-Ross plot. the smaller the degree of initial conversion. and spectroscopic characterization. x (5. This can be confusing at first sight. . J. 1 (1975).. A schematic example of such a plot is shown in figure 5.2 Schematic diagram of a Mayo-Lewis plot. • T. Up to now we have said nothing about the experimental quantities x and y. while y is given by the average or the integral value of the copolymer composition over the range of conversion chosen and is usually measured by spectroscopic methods. This is repeated for various experimental values of x and y and as a result a set of straight lines is obtained. The quantity x is then assumed to be accurately given by the ratio of the initial monomer concentrations. Obviously. the ratio of the monomer concentrations incorporated into the copolymer relative to the ratios of the monomer concentrations in the reaction mass at some instant of time. Furthermore. Kelen and F.. because we don't know rl and rz! However.these shortcomings.· has gained populanty 10 recent years.2. as illustrated in figure 5.r 1 +-1. .116 COPOLYMERIZATION Reactivity Ratios 1 + rlx y =--. Each experiment.nis to ~etermine rl and rz with a reasonable degree of accuracy is a time con~um1Og. grven above. it is usual practice to limit the conversion to a small fraction of the monomer mixture. Because in most cases both compositions vary as the reaction proceeds. the more ac\:urately y becomes a representation of the instantaneous copolymer composition. The copolymer equation written in terms of the quantities x and y.15) (intercept) The graph obtained using this equation is called the Mayo-Lewis plot. Accordingly. but ~\ one WhIChdepends upon performing the same experimental measurements is the Finem~-Ross plot.15 to calculate rz. In certain systems penultimate effects may playa role. the determination of sufficient data pOi. An . Tudes. Both the Mayo-Lewis and Fineman-Ross methods depend upon use of the mstantaneous copolymer equation and as such are subject to various errors.business. if (in principle) we pick any old value of rl. The intersection point defines the values of r I and rz.alternative graphical procedure for determining the reactivity ratios. A9(1). In practice. An alternative method designed to overco~e . Macromol.17) where: Figure 5. then we can use equation 5. attributed to Kelen and Tiid6s. but such experiments have nevertheless been performed on a WIde range of monomer pairs. Sci-Chem. the lines intersect over an area and some method for picking the best point within this area has to be applied. the validity of the terminal model is assumed. requires a separate synthesis r1 (- slope) Figure 5. Their equation is written in the form: (5. can be rearranged into the following form: (5.3.14) 117 Solving for rz: rz = x(rlx + 1) -x y (5. where y is determined for a polymer obtained with a chosen value of x.16) A plot of the quantities in brackets then gives a straight line whose slope is rand whose intercept is -rz. First of all. and the equation can be solved by an intersection method by plotting values ofrz vs rl for given values of x and y. it has been shown that these linearization methods transform the error structure in such a way that linear least squares procedures are inappropriate.

5 1.2. Coleman.118 COPOLYMERIZATION Copolymer Sequence Distribution x(y .5 r = -0. P.0 D. These were determined relative to a chosen monomer standard (styrene) and allowed the calculation of reactivity ratios.2x R = 0. monomers with the common structure CH2=CR2 (R being the bulky group) will polymerize. these and higher sequences can now be routinely measured by nmr spectroscopy. Finally. +2L y A plot ofTl vs ~ calculated from the experimental data (x and y) should yield a straight line with extrapolated intercepts at ~ = 0 corresponding to -r-jo.19) where the subscripts low and high represent the lowest and highest values of x2/y calculated from the experimental data.Ttidos plot that we prepared to determine the reactivity ratios of n-decyl methacrylate and p(t-butyldimethylsilyloxy jstyrene". The interested reader is referred to standard polymer chemistry books for further details as we will not cover this here. particularly nuclear magnetic resonance (nmr) spectroscopy.= There are a number of factors that affect the ability of a monomer to react in a copolymerization and hence the reactivity ratios. The theoretical treatment of copolymerization can be extended to the prediction of the frequency of occurrence of various sequences (e.68. where chain propagation can proceed in various ways (1. usually accomplished by elemental analysis. have their origin in a time when even the determination of the composition of a copolymer was a difficult task.0 The methods described above to determine reactivity ratios. r = 0. Painter and M. 0. Although this scheme ignores steric effects it nevertheless allowed reactivity ratios to be at least semi-quantitatively estimated at a time when their determination was more difficult. 1. however). Xu.g. and is generally determined from: 0.3010 (1993). but monomers of the type CHR=CHR will not (they can add to radicals of another kind. Polymer. resonance stabilization and polarity.36 DMA 0. '" Y. C.52 tBSOS + 1. M.. This is why styrene polymerizes (free radically) in a head-to-tail fashion. although still used. but in poly(vinylidenlJluoride). but allowed the measurement of sequence distributions. serves to distribute the experimental data uniformly and symmetrically between the limits of 0 and 1.18) Affecting Monomer Reactivity a.. (5. with parameters labeled Q and e. . or. The obvious ones include: a) Steric factors b) Resonance stabilization of the radical site c) Polarity of the double bond Monomers with very bulky groups could clearly have trouble polymerizing. and at ~ = 1 corresponding to rl. 34. where several mesomeric resonance forms can be written). MlMlMI "triads"). it should be mentioned that problems can also arise in the determination of reactivity ratios when the monomers are capable of association. Obviously. As we will see later. The advent of spectroscopic methods.4. etc). -(CH2-CF2)n-' the smaller fluorine groups allow the formation of some head-to-head and tail-to-tail sequences. COPOLYMER SEQUENCE DISTRIBUTION AND THE APPLICATION OF PROBABILITY THEORY -0. The parameter a. a monomer of one type would preferentially add a monomer of the other and in certain circumstances the effect can be so marked that charge transfer complexes are produced.5 0.4 Kelen-Tudos plot for u-decyl methacrylate-co-p(l-butyldimethylsilyloxy)styrene copolymers. Alfrey and Price attempted to account for two of these factors.0 Figure 5.g. Figure 5.98 = 0.4 shows a typical Kelen. then in general there is a decrease in the reactivity of this monomer. for example. styrene. in the polymerization of dienes. not only facilitated the determination of overall copolymer composition. Electron donating and electron withdrawing groups can affect the polarity of the double bond. In general.1) 119 TI = --=--:2:- y Factors and (5. \ y 0. If the monomer has a structure that allows a radical site to be resonance stabilized (e. relative to one where there is no resonance stabilization.

. practice a little and learn a new language.28) as the conditional number fraction of units of a particular polymer chain with a sequence Xlx2. for the diads In terms of triads: (5. The concentrations of the lower order placements can be expressed as sums of the two appropriate higher order sequences. which is a far more convenient way to describe sequence distributions. then PtfA} and PI {B} are the mole fractions of A and B in the polymer. the probability is equal to the number fraction of desirable events occurring and the terms number fraction and probability will often be used interchangeably. which represent impossibility and certainty.25) In other words. A or B. where NE is the number of E events occurring and N is the total number of events... in truth.26) predecessor General Statistical Relationships Probability theory is designed to tackle situations in which there is more than one possible outcome and attemp~~ to answer questions concerning the likelihood of the various possible results. P6{ABABBA} must equal P6{ABBABA}. the number fraction of AB sequences must equal the number fracti?n . Regarding the latter. It should be self-evident that if there are only two events. the probability of an event E is defined as: prE} = P3{AAB} = P3{BAA} P3{BBA} = P3{ABB} (5.30) . We would do this deliberately.22) and: (5. P{AIB} is the number fraction of A units given that the previous unit was a B.. This really shouldn't worry even those who are not mathematically inclined. ~ and which has the particular unit ~+I as the next unit. Hence. For example. PI {A} = P2{AA} + P2{A]!} = P2{AA} + P2{]!A} PtfB} =P2{BB} +P2{BA} =P2{]!B} +P2{AB} successor predecessor This leads to the important conclusion: (5. First we will be considering general statistical relationships between the various experimental sequence probabilities that do not depend upon polymerization mechanism.23) (5. Let us define: Conditional Probability We will require a more specific way of defining a particular sequence for our purposes. to one based upon probability theory. rather than MI and M2 favored in kinetic treatments. which is referred to as conditional probability.of sequen~es. To remind you. but when introduced to students is often met with vacant stares. we will immediately start by changing the nomenclature used so far in this chapter..27) N NE (5. For example.120 COPOLYMERIZATION Copolymer Sequence Distribution 121 Brace yourself. as the concepts that we will discuss are truly simple (and elegant) and just require us to think a little differently.20) ~o~e the reversibility of these sequences. .24) Bf P2{AA} P2{AB} P2{BA} P2{BB} which requires that: = P3{AAA} + P3{AA]!} =P3{ABA} +P3{AB]!} = P3{BAA} + P3{BA]!} =P3{BBA} +P3{BBB} successor = P3{AAA} =P3{AAB} = P3{ABA} =P3{ABB} + P3{RAA} +P3{RAB} + P3{BBA} +P3{RBB} (5.21) as the measured number fraction of a particular sequence of units XI X2X3. But first a review of some definitions. Then we will consider specific relationships characteristic of specific polymerization mechanisms. even if we didn't have to. just to satisfy the sadistic streak within us! But. P{AABA} is the probability of the sequence AABA in a copolymer chain consisting of only two monomers A and B. We define: (5. and: PtfA} + PtfB} = 1 Similarly. xn• For example.. respectively.. when considering pairs of units or diads: (5. It follows therefore that: P2{AA} =PtfA} P{AIA} P2{BA} = PtfB} P{AIB} P3{AAA} = PI {A} P{NA} P{NAA} P3{BAA} =PI{B} P{AIB} P{AlBA} (5. Similarly. Probabilities are restricted to a range of 0 to 1. since any sequence has a successor or predecessor which can only be either an A or a B. respectively.29) (5. for now we are going to change from the kinetic description that we have used so far in this chapter and that chemists usually like. it is conventional and more convenient when applying probability theory to use A and B to define the two different comonomers. In fact this is a general principle and It IS always true that the probability or number fraction of a sequence is always equal to that of its mirror image.

Remember that nmr now provides a way to measure sequences such as this.31) and P(A/A} P(A/B} P(NAA} P(AfBA} P(A/AB} P(AfBB} (5. respectively.32) terms the probability may be expressed as: Ps(ABABA} = PtfA} PtfB} PtfA} PtfB} PtfA} or (5. or BA.37) + P(B/A} = 1 + P(B/B} = 1 + P(B/AA} = 1 + P(B/BA} = 1 + P(B/AB} = 1 + P(B/BB} = 1 When k = 2.38) As an example. then gather all like terms. Unfortunately. you just write down all the conditional probabilities required to describe the sequence. its a bit like having your teeth cleaned. For example. B. we have second order Markovian statistics which correspond to the penultimate model.41) . and: (/ (5. and: (5. For the general case of kth order Markovian statistics we have: (P(A/B}) 2 (5. represent the conditional number fraction that the next unit is A given that the preceding placement (or placement pair) was A. 6ut the equations have a simple structure and once you have used them a couple of times they become familiar. Rearranging these equations.36) When k = 1. but it's necessary!) Conditional Probability of Different Orders In considering compound events like P3(AAA}' it is perfectly legitimate to break them down into conditional probabilities of different orders: For example: P3 (AAA} = PI (A} P(NA}P(NAA} (5. and: (5. You don't like it. "Why are these guys giving us all this stuff?" They are tools y06 will need to analyze microstructure. first order Markovian statistics (the so-called terminal model) are applicable.34) However. try not to miss the wood for the trees. we will be considering polymerization systems where only one type of conditional probability is applicable.122 COPOLYMERIZATION Copolymer Sequence Distribution k n Pn(A } = (PI(A } n-k P(NA}k) statistics.39) In 1st order Markovian statistical terms: Ps(ABABA} =PtfA} P(B/A} P(AfB} P(B/A} P(A/B} or Ps(ABABA} = PI (A} (P(B/A}) 2 At this point you may be wondering. we are not finished yet. P(AfB}. we have zero order Markovian Bernoullian statistics.35) also known as When k = 0. AA. in free radical polymerization where commonly two reactivity ratios are used to describe the instantaneous composition of a copolymer chain formed from two different monomers. in this case all A's. so these equations have direct practical importance. But if you are having trouble with this stuff. we have first order Markovian statistics which correspond to the terminal model. In Bernoullian statistical (5. consider the sequence ABABA. P(A/AA}. (In a sense. In 2nd order Markovian statistical terms: Ps(ABABA} = P2(AB} P(NAB} or Ps(ABABA} = P2(AB} (P(A/AB}) 2 P(B/BA} P(A/AB} P(B/BA} (5. The order referred to here depends upon how many preceding units are included in the conditional probability.33) The general equation relating the number fraction or probability of a sequence of n consecutive units.40) You can see that this is quite easy. is: n n-I Pn(A } = PI (A} P(NA} P(NAA}-------P(A/A} (5. 123 where P(NA}. P(AfBA}. the conditional probabilities can be computed as ratios of the measured probabilities: P(NA} = P2(AA} I PI (A} P(A/B} = P2(BA} I PI (B} P(NAA} =P3(AAA} IP2(AA} P(AfBA} = P3 (BAA} I P2(BA} P(NAB} =P3(ABA} IP2(AB} P(AfBB} = P3(BBA} I P2(BB} P(B/A} = P2(AB} I PtfA} P(B/B} =P2(BB} IPI(B} P(B/AA} =P3(AAB} IP2(AA} P(B/BA} =P3(BAB} IP2(BA} P(B/AB} = P3(ABB} IP2(AB} P(B/BB} =P3(BBB} IP2(BB} (5.

NA(3) = Ps {BAAAB} P {AB} 2 (5. ISmcluded.. We therefore have: The Run Fraction or Number (5.A .42) (5.e. which is equivalent to: LnP n+ 2{B~B) I then: = P {A} I (5.49) Remember P2{AB} = P2{BA}.Ps 00 {BAAAB })+.50) Now: = P1{A} . We will be using IA and ~ information later 10 our diSCUSSIOnsf copolymer microstructure.45) Number Average Length of A or BRuns The number average length of A or B runs. .52) and substituting equation 5.. It is defined as the fra~tlon of A and B sequences (runs) occurring in a polymer chain.24): PdB} =P2{BA} +P2{BB} (5. Similarly. We will assume that the chain is sufficiently long so that end effects can be Ignored: The chain above contains 20 units arranged in 12 alternating runs (u?de~lmed): The run number is thus 12/20 = 0.43) . i.. AA . Since.PdB} so the number average length of A or B runs may be readily calculated if one has molar com~osition . is equal to the number fraction of BAAAB sequences divided by the number fraction of AB diads..etc.because It has been commonly used in the nmr literature.B.B. and since: Number Fraction of Sequences of A Units The number fraction of sequences of A units of length n. respectively. Consider..44) For example. and PdA} = 1 . Although we wi~l not use it much in this book.B. Note that every run of A ~mts IS term mated by an A-B link. have the definition: denoted by the R = Fraction of (AB + BA) links =P2{AB} + P2{BA} = 2P2{AB} And since (equation 5. the portion of a polymer chain depicted below: A.47) There are a number of parameters that will be of particular use to us in the forthcoming discussion. .44: P2{BB} = P1{B}- . for example.48) N (n) = A _oo. N A (n).46) we arrive at the rather useful relationships: P2{AA} (5. P {B~B} - (5.~ ~ (5.:n::.B. R. the run fraction or number. L P n+ I 00 2 {B~B} = P2{AB} then: (5. is defined as the number of such sequences divided by the number of all possible sequences of length n: P 2 t n Pn+2{B~B} 00 = (loP3 {BAB})+(2oP4{BAAB})+(3.:+~_ _ t. o n+2 {B~B} Obviously a sequence of n A units must be preceded and succeeded by a B unit.. L Pn+2 00 {B~B} I = P {BAB}+ P4{BAAB}+ 3 Ps{BAAAB} etc.AAA BB A BBBB AA .124 COPOLYMERIZATION Copolymer Sequence Distribution 125 Some Useful Parameters (5. the number fraction of say a sequence of three A's.6. AAA. symbols IA and ~. B runs are terminated by B-A hnks. all the sequences end in AB.and ~ad sequence data.51) (5..

the fewer AB links.60) The Terminal Model We will now revisit the copolymer equation (terminal model) that we discussed earlier in this chapter having obtained some necessary background in probability theory. the deviation from randomness. On the other hand. the number and weight fractions of A or B se~uences.56) where x is once again the monomer feed ratio = [A]/[B]. P(A/B} and P(B/A}. For real block copolymers the value of X will approach 0.54) P2(AB} and we know that for a completely random process: P2(AB} =PdA}~' PI(B} and using the same definition of reactivity ratios (equation 5.kAA[A*][A] + kAB[A*][B] (5..AA7[:-A-:*]_[A-:]. random (Bernoullian) statistics are obeyed when the number fraction of AB (or BA) diads equals the mole fraction of A's multiplied by the mole fraction of B's in the copolymer. These are readily obtained from: P(A/B} P(B/A} = 1 . in general. Recall: .--~k:::. and the closer X will get to O. and defined as: may be In terms of conditional probabilities we can write.55) Rate of reaction producing AA * Sum of all reactions involving A * (5.57) that we shall ~here are two further independent conditional probabilities ~\require.5. Thus the value of X is 2.PI(B} _ P(A/B} . PI (A} = PI (B} = 0. or when X = 1. there are two independent conditional probabilities. for example: P (N A} = Therefore: (5. (equation 5. In other words. These two conditional probabilities can be derived in terms of the kinetics of polymerization using the following approach: Terminal GrOUU ---A* ---B* ---A* ---B* Added Grouu A A B B Rate kAA[A*] [A] kBA[B*] [A] kAB[A*] [B] kBB[B*] [B] Final Product ---AA* ---BA* ---AB* ---BB* x This is the now familiar instantaneous copolymer composition first derived by Mayo and Lewis. ~i~ilarly.6) and dividing through by [A*][B]: Accordingly. and as there are no AA or BB sequences. When the rate of addition of monomer to a growing chain depends upon the nature of the terminal end group. In summary: X= 1 X> 1 X= 2 X< 1 X ""0 (/ P(AlA} = = (5. P2(AB} = P2(BA} = 0. Similarly: P(B/B} r =-_ B rB + x (5. the number average length of A or B runs.62) = P2(BA} (5.P(B/A} = l+rAx rB 1+(5..58 and 5.-_ .-. can be expressed in terms of conditional probabilities or reactivity ratios.5.61) Rearranging and substituting for equations 5.126 Copolymer Sequence Distribution COPOLYMERIZATION 127 A Measure of the Departure from Randomness A convenient measure of the deviation from random statistics obtained from a parameter denoted by the symbol.58) (5. X. If you think about it the limit in this case is two homopolymers of A and B where there are no AB links and the value of X is O.59) Completely random copolymer Copolymer with an alternating tendency Completely alternating copolymer Copolymer with a "blocky" tendency Block copolymer 1 rAx+ 1 Recall that. the run number.25): P2(AB} For the terminal model we can write: (5.P(AlA} = _x_ = I rB + x (5. For the limiting case of a completely alternating copolymer.59 we obtain: _ PI(A} y .P(B/B} = 1. the number of AB diads is greater than that calculated for a truly random case. the greater the molecular weight of the blocks. if the number fraction of AB diads is less than that calculated from PdA} x PdB} in the copolymer. If the number fraction of AB diads exceeds PI (A} x PI (B} in the copolymer. then we have an alternating tendency.53) P (AlA} _ -. then we have a "blocking" tendency..

IN ~. The computer calculates and prints out the instantaneous copolymer composition.Chem. Fortunately. We have developed such a program" to calculate not only the composition of copolymers but also the microstructure as a function of conversion based upon the equations presented in the previous section above. The input necessary is the initial monomer concentrations [A] and [B] and the reactivity ratios rA and rB.en) = C ::AX ( rB ~.62 and assume that [A] and [B] are constants over the entire course of the polymerization. Instead an interval in which a given amount of total monomer is used up (i.67) (5. W A(n). Copolymer Composition as a Function of Conversion For a batch system (i.01 gives one hundred points for complete reaction. First. (1982).59 we get" R = _--=2. let us now apply them to the problem of compositional drift. Ed. the 'amount of A plus the amount of B) is defined. An interval of 0. Obviously. we cannot simply use equation 5.---::rB rAx+2+xSimilarly for X: (5.. NA(n). NB(n).68) = N.. b) Starting with the initial known concentrations. c) Calculate the amount of each type of monomer that remains after this period of time. It then calculates the new monomer concentrations and repeats the process for the next interval etc.e.70) 1+: and finally for the weight fraction of A (or B) sequences. denoted STEP. J.71) =n( ~r)n-l(l_ 1+: ~r)2 1+: (5. the errors in assuming constant monomer concentrations get smaller as we choose shorter and shorter intervals. wA(n) and wB(n): (5. then we defme the conversion as: Conversion (5. We'll leave the student to prove this (hint: start with equation Substituting equations 5.66) ( and for NA(n) and NB(n): I= A lB 1 P(B/A} 1 P(AIB} = = 1 + rAx 1+~ r x (5.64) 5.72) Obviously. by which we mean the fraction of total monomer that has been used up.e. one where we are not continuously introducing monomer) we can calculate the composition of the copolymer we obtain as a function of conversion.25). . D.847. This can make for a lot of calculations if we wish to know how the copolymer composition varies over the entire course of a polymerization. Coleman and W. 59. Varnell.73) (5.69) NB(n) = )n_l( 1-~ (5.63) As first order Markovian statistics are applicable to the terminal model. If we start with monomer compositions [A]o and [B]o and at some point the monomer composition remaining (i. and WB(n) for n = 1 to 10 at the given interval. d) Repeat these steps for each subsequent time interval starting with the newly calculated monomer concentrations.128 COPOLYMERIZATION Copolymer Sequence Distribution 129 (5. * M. A further input. is the interval between successive calculations.. [A]o and [B]o at the very beginning of the reaction calculate how much monomer of each type is incorporated into the polymer in this chosen time period. it is usually not convenient to pick a "time period" as such.65) I. M. this is the type of problem that is ideally suited to solution by construction of a simple computer program.58 and 5. = (rA]o + [B]o) .e.. X. There are two points concerning these calculations that need to be mentioned. unpolymerized) is [A] and [B]. this is equivalent to: R = 2P1(B}P(AIB} 2P(B/A} = P(AIB} +P(B/A} P(AIB} Having derived all of these equations for the terminal model. Calculations must be performed using some variation or form of the following steps: a) Choose a short period of time over which [A] and [B] are assumed to be constant. 1+: r( 1 -1 ::AX) rB) (5.(rA] + [BJ) ( ) [A]o + [B]o (5.

2.05 9 0.79 0.17 0.0 and rvc 0. QI '" c.02 0. with approximately 50% as single units (i. [VDCJ = [AJ = 0.4 VC = 0.. AA or BB) and 12% as trimers .00 0. Yinylidene Chloride (VDC) and Vinyl Chloride (VC) For this calculation equimolar initial concentration of the monomers was chosen (i.14 0. and (VC).00 0.1 and lB of 7.19 0. the polymer composition is nearly equimolar and the distribution of the number fraction of VDC and VC is almost identical.0 1. The copolymer composition is initially rich in VDC.00 0.00 8 0..04 1.52 0..00 0.46 0.13 0. This graph convincingly demonstrates the large variation of copolymer composition with degree of conversion and illustrates that a wide compositional distribution of polymer chains will be present for a copolymer synthesized to moderately high conversions.00 0.13 0.10 om om om 0.51 om 0.06 om 0. (i.00 0.13 VC = 0.04 20 1.10 0.00 0.4 1.03 0.5 1.06 :. The compositional variation as a function of conversion is illustrated in figure 5.49 80 0.19 n= 1 0..03 0.19 0. [VCJ = [BJ = 0.00 0.5.00 0.00 0.32 60 40 % Conversion Initial Polymer Composition 0.3 3.2 ~ I:: '0 vc 0 20 40 60 80 100 0.00 0.10 0. but it is made up of a wide distribution of chains of different composition.34 0. indicate that the polymerization of VDC and VC is nearly random. Table 5.00 0.5 Compositional variation for vinylidine chloride/vinyl chloride copolymers. after which chains rich in VC are formed.130 COPOL YMERIZA TION Copolymer Sequence Distribution Table 5.21 0. ~ is only 1.19 1.00 0.05 1. seq_uences.03 0.0 while the corresponding number averag~ length of Vc.04 1.00 0.00 0.81 0. if the polymerization was taken to completion. The calculated values of x" which are close to unity.9 1. the average composition of the copolymer is still 50:50.02 0.12 3 0.87 of VDC/VC.08 6 0.00 0.04 0. one hundred such intervals are determined).11 5 0.8 ::: 0 VDC=0.79 0.5. 0.51 0.03 Case I.25 0. = r / 1.e.6 0.83 0.00 0.0 = .20 0. 0.54 0.00 0. All reactivity ratios were taken from the published data listed in the polymer handbook and represent typical values.1 Copolymerization Monomer Concentration VDC= VC= VDC= VC= VDC= VC= VDC= VC= VDC= VC= VDC=' VC= 0. and the reactivity ratios employed were rVDC= 4. lA' ~ values 5.04 10 0.00 0.e. 100% conversion).20 0.5 X..00 7 0.5. Throughout the polymerization.0 1.00 0. At 60% conversion.08 0.05 0.0 Degree of Conversion Figure 5. lAdecreases while lB increases.1 7.2.9 1.00 0. The !_lumberaverage lengths of a girren sequence IA or ~ vary considerably.01 0. 131 To illustrate the type of information gained from these computer calculations several examples are presented. it is apparent that initially the vast majority of VC units exist as isolated units while there is a large distribution of sequence lengths of VDC units.00 0..02 Table 5. No attempt was made to ensure the validity of any specific set of reactivity ratios.03 0.e.40 0.04 0.07 0.e. polymer chains are being formed with lA values of 1.16 0. A or B).1 shows a summary of the results obtained at six specific degrees of conversion (in our calculations.24 0.68 0.81 VC = 0. Initially lAthe number average length 9f VDC seque~ces is 5.03 0.76 0.76 VDC VC= 0.08 0.06 0. From the calculated number of A and B sequences shown in table 5.12 4 0.-. but this decreases as the polymerization proceeds.12 0.09 I:: 0 u r.05 0.0 1.~ 0.. At ~O% conversion. value 1.25 0.50 0.2 4.2 4.13 2 0.50 0.2 Number Fraction of (VDC).25 0. Number Fraction of (A}n and (B)n Sequences Polymer Composition VDC= 0.07 0.14 0.32 0.2.00 0..24 VDC = 0.06 0.04 0.42 0.48 0.5). In other words. 25% as dimers (i.87 om 0.08 0.00 0. Polymer chains containing equimolar concentrations of VDC and VC are formed at approximately 63% conversion. Sequences.89 0.02 0.49 VDC= 0.

02 0. isolated VDC units are present in greater amounts than any other species (i. this is an excellent example of a predominantly alternating copolymer.08 0.53 0.47 10 0.05 9 0.89 The copolymerization of St and MAH is a classic case of a highly alternating copolymer. trimers.5).12 0.06 0.06 0.0 1.40 0.09 0.).01 0.00 0.07 0.11 0.07 0.03 3 0.03 0.36 60 0.75 0.00 0.1 1. Styrene (St) and Maleic Anbydride (MAH) % IA. III values 1.08 0. Table 5.10 6 0.81 VC = 0.79 0.11 0.00 0.69 0. dimers.13 VC = 0.2 0.02 2 0.64 0.05 7 0.6 e ~ "" 0.00 0. on a weight basis. At the start of the polymerization the polymer composition is calculated to be 53% St and 47% MAH.27 0.02 0. on the other hand.06 0.23 0.0 1.08 0.132 COPOLYMERIZATION Copolymer Sequence Distribution 1.3 shojs the corresponding weight fraction of A and B sequences. chains are formed.09 0.00 0.04 0.76 VC =0. At 80% conversion polymer chains are being formed with number fractions of A and B sequences that are almost the converse of that calculated at the start of the polymerization.06 0. increases while X decreases.45 50 0.19 0.05 0. However.08 0..1 1.6 = shows the compositional variation as a function of conversion. formed.0 0 om 0. As the polymerization proceeds toward approximately 50% conversion there is a tendency toward increasingly rich St polymer chains being Concentration ST= MAR = ST= MAR = ST= MAR = ST= MAR = ST= MAR = 0.00 0.07 0.87 n= 1 0.0 1.11 0. for example.05 8 0.e. (i. Monomer Polymer Composition 0.47 30 0.4 Copolymerization of St/MAH.00 0. However.07 0.48 0.88 1.8 St MAH ( C Q.08 0.~ .Q1 0.00 0.19 0. The number fraction of A and B sequences is also informative (see table 5.04 and rMAH 0. Weight Fraction of (A}n and (B}n Sequences Polymer Composition VDC=0.00 0.55 0.20 0. Although we have started with a 3: 1 molar feed ratio of St to MAH.Q1 0.25 for MAH.26 0.06 0.82 1.26 0. Note that for MAH the fraction of isolated B groups is unity regardless of the extent of reaction.e.19 VDC = 0.00 0. In essence.4 shows a summary of the calculated results for several specific degrees of conversion.05 0.00 0.0 X value 1.8 1.00 0.00 0..96 0.49 VDC=O.51 VC = 0.24 VDC=0.04 4 0.0 25 1.05 e =. and eventually almost pure polystyrene.2 1.. Consider.27 0.05 = .07 0. the III value is invariably unity" which demonstrates that the MAH units are exclusively one unit in length.001 Conversion Initial 1. Until about 50% conversion.. which together with the calculated value of 1.01 0.3 Weight Fraction of (VDC)n and (VC)n Sequences.::: <n 0.89 for X demonstrates that the polymer chains formed have a predominantly alternating microstructure. sequences containing four and five VDC units predominate.00 0.05 10 0.04 0.59 0.53 0.11 0..05 5 0. the number fraction of isolated St sequences is also dominated by isolated units.00 0.2 and 5. This is analogous to the number and weight average molecular weight distributions found for linear polycondensation. and reactivity ratios of rSt = 0. the number and weight fraction of VDC sequences at the start of the polymerization reaction (initial).Q1 0. IS :.04 0.00 0. This illustrates that the MAH-radical cannot add to another MAH-monomer. U IS 0.4 0. the composition of the polymer chains formed is approximately equimolar up to about 50% conversion at which time St-rich.. AAA or BBB sequences).02 0.25 0. The IA value.015 were used. In this example initial concentrations of 0.00 0.03 0. Figure 5.04 Case II.00 0.3 that the number and weight fraction distributions vary significantly.04 .62 0.00 0.29 0.17 0.08 0. Table 5. The IA and III values are both approximately unity. Table 5.57 1. 20 40 60 80 100 Degree of Conversion Figure 5.70 0.6 Compositional variation for styrene/maleic anhydride copolymers. On a number basis. It is apparent from a comparison of tables 5.0 133 Table 5.00 0. etc.75 M for St and 0.

~ ~ 0.00 0. reactivity ratios chosen from the literature were rAN= 9.00 0.00 0.19 80 55 20 % Conversion Initial Polymer Composition 0.25 0.2 3.0 2.50 1.53 MAH=0. 0.57 1.25 0.00 0.50 0. For this example we chose equimolar monomer concentra- 0.82 0.00 6 0.96 MAH = 0.6 1.03 0.01 0. [MA] = [B] = 0.87 0.50 n= 1 0.0 2.55 MAH = 0.00 3 0.00 0.7 Compositional variation for acrylonitrllelaliyl methacrylate copolymers.7 Number Fraction of (VDC)n and (MA)n Sequences..00 0.00 0.00 0.50 0.03 0. The number fractions of A and B sequences are also identical (table 5.00 0.0 0 20 40 60 80 100 Degree of Conversion Figure 5.5 Number Fraction of (St).00 0. Monomer Number Fraction of (A)n and (B)o Sequences Polymer Composition VDC = 0.03 0.50 0.18 0.00 0.50 0.00 0.04 0.7 1.04 0.50 0.06 0.01 8 0.00 0.00 0.00 0.50 0.6 0 u .6 summarizes the computed data.50 2 0.05 0.5.03 0.7) and reflect a random distribution of the monomers incorporated into the polymer chain.47 0. 1.00 S 0.00 0.00 9 0.62 Case III.00 0.5 7.00 10 0.77 .00 0.04 1.01 0.00 0.50 1.00 0. Acrylonitrile (AN) and Allyl Methacrylate (AM) 99 0.15 0.00 0..00 0.2 0.03 0.8 8 0.00 0. = .5 11. Over the entire course of the polymerization the lAand lBvalues equal 2. Table 5.82 1. Table 5.00 0.0 2.13 0..00 9 0.0 X value 1.00 0.00 0.0 Number Fraction of (A)n and (B)n Sequences Polymer Composition St = 0.37 0.00 0.00 0.01 0.00 0.02 0.55 and rAM = 0.03 0.45 St = 0.00 7 0.7 3.00 2 0..25 0.76 0..00 0. Monomer Concentration VDC= MA= VDC= MA= VDC= MA= 0.50 0.02 7 0.04 0.04 n= 1 0.02 0.00 0. Vinylidine Chloride (VDC) and Methyl Acrylate (MA) AM= AN= AM= This is an interesting copolymer where the reactivity ratios for both monomers equal unity.4 QJ '0 p.50 IA.00 0.00 0.03 0.00 0. The composition of the copolymer is also equimolar and does not vary throughout the polymerizarlon reaction. S .09 0.13 0.00 0.71 0. i.00 5 0..05 ) % Conversion Initial Polymer Composition 0.00 AN= AM= AN= AM= AN= Concentration 0.89 1.50 MA = 0. [VDC] = [A] = 0. 0.03 0.00 '<ij 0 CI.5.00 0.00 0.05 0.00 0.00 Case IV.00 0.00 In this case.00 0.134 COPOLYMERIZATION Copolymer Sequence Distribution 135 Table 5.00 0.8 Copolymerization of AN/AM. indicating a completely random polymerization process.03 6 0. and (MAH)n Sequences.89 lA' fB values 10.00 0.25 0.06 5 0.3 X value 0.10 0..0 2.50 0.00 0.00 0.00 4 0.00 0. Table 5.2 1.00 0.12 0.01 0.13 4 0.25 3 0.50 0.6 Copolymerization of VDC/MA. Table 5.64 MAH = 0.0 2.00 tions.IB values 2..00 0.e.47 St = 0.0 and the X value is unity.515 and represent a copolymerization of two monomers where the product rArB> 1.10 0.00 0.00 0.33 0.00 10 0.50 0. 50 0.36 St = 0.00 0.00 0.

D7 6 0.15 N(n) n-4 0.06 0. = are skewed toward those characteristic of isolated units.06 0.12 AM = 0.89 1.02 0.00 0.00 0.00 0.06 1.50 AN= 0.21 0.10 0.89 and the lA and In values both equal 1. In contrast.12 0. For the VDC /MA copolymer lAand In are equal to 2. Note also that the values of NA(n) and Nn(n) are identical and the distribution of the sequence lengths is entirely consistent with random placement of the two monomers.00 0.01 0.D7 4 0. which is indicative of a completely random copolymer. i.. which also reflects a predominantly random copolymer. The Penultimate Model Summary It is convenient to compare X.04 VDC= 0.08 0. The X value of 0.0 1.04 0.0 and X is unity.00 0.08 1.2 suggests a copolymer with a "blocky" tendency. There are.00 0.07 0.05 0. Number Fraction of (A).47 AN=0.00 0. Initially the copolymer chains are rich in AN but the amount of AM increases progressively' as the reaction proceeds.10.06 0.9 Number Fraction of (AN)" and (AM)" Sequences.22 0. In the former copolymer the X value is 1.18 AN=0.e.04 0.9). This is indicative of a predominantly alternating copolymer.06 0.05 9 0.15 0.08 0. [AM] [B] = 0.03 0.00 0. A X value of 1.10 0.2 3.03 0.15 0.05 0.00 0.10 N(n) n-5 0.21 0.1 1.9 1.03 0.07 Valles 2.03 0.89 0.10 0.04 10 0.53 0. iA' in' NA(n) and NB(n) for polymer chains that are formed from approximately equimolar concentrations of the monomers for the four different copolymers considered.06 together with iA = in = 1.02 0. In the case of the AN/AM copolymer the opposite effect is observed.06 0.05 0.25 0. Furthermore. however.05 8 0. We will briefly show how it is possible to differentiate between terminal (first order Markovian) and penultimate (second order Markovian) polymerization from a knowledge of the sequence distribution.01 0. which also denote an altemating copolymer.13 0. The NA(n) and Nn(n) values are close to those seen for the VDC/MA copolymer.50 AM=0.02 0.00 0.13 AN=0.5.53 MAH=0. Terminal .53 0. The X values obtained are less than unity indicating a definite tendency toward a "blocky" microstructure.03 0. the NA(n) and Nn(n) values So far we have considered the terminal model.01 0. the St/MAH and AN/AM copolymers are significantly different.07 0.136 COPOLYMERIZATION Copolymer Sequence Distribution Table 5.01 0.00 0.04 0.02 0.03 0.50 MA = 0.0 2.03 0.31 An equimolar monomer feed ratio was employed.02 0. In addition. sequences (table 5.22 0. well documented cases where the specific character of the last two monomer units of the growing chain affects the addition of the incoming monomer unit and this is called the penultimate model.31 0. This is also reflected in the iA and in values and the number fraction of (A)n and (B).07 5 0.02 ' 0.00 0.10 Comparison of Copolymer Microstructure.31 0.9 1.10 0.9 suggests largely random placement of the two monomers with a very slight tendency toward alternation.00 0.89 n= 1 0.09 0.00 0.05 0.50 VC = 0. --A (B) -: <= --AA (AB) (BA) (BB) Penultimate etc.00 0. At approximately 55% conversion.04 0.00 0.00 0.8 shows the computed results for polymer chains formed at four specific degrees of conversion.00 0.13 0. Table 5.06 0.68 0.2 0.06 0. This is displayed in table 5.00 0.31 0.24 0.07 0.62 together with iAand ~ values of 3.05 0. as described by the reactivity ratios.12 0.50 VDC=0.0 3.87 AM=0.58 0.22 N(n) n=3 0.50 St = 0.66 0.62 0.25 0. [AN] = [A] = 0.1.D7 0.00 0. Polymer Composition 2 3 0.25 0.06 0.09 lA' i n- X value 1.00 0.10 0.09 0.10 0.04 0.11 0. Figure 5.02 0.07 0. where the end group of the growing chain affects the addition of the incoming monomer unit. the distribution of NA(n) and Nn(n) is skewed towards sequences of greater length.00 N(n) n=1 0.50 N(n) n=2 0. which again implies a "blocky" character.15 0.50 AM=0.5.7 illustrates the wide compositional variation of the copolymer during the ~olymerization reaction. the polymer chains formed contain about the same molar concentration of AN and AM.D7 7 0.82 AM=0.10 0. The equimolar VDC / VC copolymer chain formed at approximately 60% conversion is also close to random.50 0.02 0. and (B)n Sequences Polymer Composition AN=0.25 0.06 0.22 0.13 0.50 55 Initial 61 % Conversion All 137 Table 5.

etc. G..J!. in order to distinguish between the terminal and penultimate models. we start with a reversibility relationship. so we use: P2{AA)P{B/AA) = P2{BA)P{A/BA) (5.76) 1+ = r~x(1+rAx) (1 + r~x) (5.87) r~ ( P{A/BA) 1 +r~x r~ x \ where: r P{B/AB) = --r-' 1 +.P{B/A) It follows .78) ABA Testing the Models There have been a number of systems where penultimate effects have been established".86) 1 +rAx r~x P{A/AA) 1. Principles of Polymerization. four independent conditional probabilities can be written with four monomer reactivity ratios. See. 1991. P{A/AB) = 1+ x 1+- rll) x (1 + ~) x P{A/BB) = _1 r- P{B/BB) = I-P{A/BB) (5. experimental accuracy is paramount and is the crucial factor in determining whether or not the effect is real. Wiley. It is convenient to write the copolymer composition equation in the general form: Again in a similar manner to the methodology used to derive the terminal copolymer equation. Odian.74) P{B/AB) + P{A/BB) (5.P{A/BA) (5. kAAA r • A A =-- ~M ~AB r II =-- ~llll ~llA r~ = kAi3ll k (5.85) Substituting equations 5. However.80) Recall that P 2 (BA) = P 2 (AB) and therefore: PI{A) P{A/A) P{B/AA) And since: = PI{A) P{B/A) P{A/BA) (5. x =-kAAll which is the copolymer composition equation for the penultimate model. (5. Thus in general.82) (5. the form of the equation is quite simple and easily derived using conditional probability. While we would admit that this looks awful.: P{B/AA) = P{B/AA) + P{A/BA) Similarly starting with P3 (BBA) = P3 (ABB): P{A/BB) P{A/B) = P{A/BB) +P{B/AB) (5.84 leads to: P{A/BA) 1 + P{B/AA) 1 (5. .84) Added Group A A A A B B B B 1 Final Product -AAA* -BAA* -~BBA* -ABA* -AAB* -~BAB* ~~~BBB* -~ABB* Recall from equation 5. Penultimate Group -AA* -BA* ~-BB* -AB* -AA* -BA* ~-BB* -AB* And: P(B/ AA) = -----'=---- IP{B/A) . for example.J!. styrene and fumaronitrile.83) Conditional Probabilities for tire Penultimate Model P{A/A) = 1 .that: When the nature of the penultimate unit has a significant effect upon the absolute rate constant in copolymerization. Very precise measurements of copolymer composition are required. including the free radical polymerization of MMA and 4-vinyl pyridine. 3rd Edition.77) 1 +.83 and 5.81) .75) PI{A) y = PI{B) 1 .62 that: PI (A) PI{B) P{A/B) = P{B/A») (5.138 COPOLYMERIZATION Copolymer Sequence Distribution 139 (5. in a manner similar to the terminal model. especially at low and high monomer feed ratios.79) For the penultimate model we must employ 2nd order Markovian statistics.P{B/AB) (5.P{B/AA) Finally substituting for the reactivity ratios and molar feed ratio yields: P{B/BA) 1.

STUDY QUESTIONS 1. The following data was obtained from a series of free radical c?polymerization experiments employing the monomers styrene (M1) and diethylfumarate (M2).= (rA) .96) (rB) = r~-+ 1 +. that errors can lead to false conclusions testing for the penultimate effect requires rigorous experimental accuracy. Ml (weight %) ml (weight %) 13.88) For the terminal model: (rA) = rA and (rB) = rB which are independent of x. would be logical to develop the relev~nt statIs~cal relationships . But we think that most students will have had enough of probability theory by now and need a break. however. r~ and r~.68): As we pointed out in chapter 1.7 47. we can use the number average length of A or B runs.L(TA -1) 2 rA x (5.J!.1 28.6 40.l)/x versus y/x2 linearizes the equation and permits a determination of the reactivity ratios.93) (5." . 1 P(A/B} We will leave the student to derive the following for the penultimate model: TA = and: l+(r~~+r~x)x +rAx (5. and m2 for styrene and diethylfumarate. and Isomerism in POlymers-A Special Case of Copolymerization? x (5.5 38. Again.placements of the monomers in the chain of polybutadiene or "norm~l.90) where a plot of (y .5 48. Recall that we derived the relationship between fA or IB and the reactivity ratios for the terminal model (equations 5. (rA) and (rB) are d~pendent upon x and given by: r 1 +.89) If we let y = P{A}IP{B}.4. after we savored the dehghts of spectroscopy. which implies a terminal model. To ~araphrase General Douglas MacArthur. a concave or convex curve (also with an intercept of unity) should be observed-the shape of which depends upon whether rA (or rIl) is greater or less than r~ (or r~) . The symbols A and B could equally be used to denote. ~os~ible to directly determine the four reactivity ratios by rearranging ( linearizing ) the ab(ove. Finally. If not. X (5.(rB) 2" . - IA fB = = P(B/A} 1 = = 1 +rAx r 1 +. or the insertion of a meso or racemic placement in the cham of pol'y(~ethyl ~etha~rylat~). For the penultimate case.7 37.91) (5.and trans-l.J!.5 70. the greatest sensitivity occurs at high and low values of x.92) E.4. "we will return".(wlth appropnate changes in nomenclature!) to d~scnbe the microstructure of this class of "copolymers" in this chapter. Thus it. the penultimate effect only produces a slight deviation from a straight line at the high and low values of x. Another approach is to rearrange the penultimate copolymer equation into a Fineman-Ross type of equation: y-1 Y -x. it is theoretically possible to test whether or not the experimental curve of y versus x can be fitted adequately over the whole composition range by only two reactivity ratios.equations: TA-2 x and: =r A' _.6 58. " .67 and 5.5 45.140 COPOLYMERIZATION Study Questions 141 (5. x (5.6 54. however.95) (5.. cis-l. It is. respectively.0 .J!. where the inherent errors are the greatest. while for the penultimate model. Starting from the steady state assumption: kI2[M.a~d "bac~wards" placements of the monomers in the chai~ of pOI~(vlOyhdlOe fluonde). Unfortunately. can be estimated by an iterative process.8 30.Always remem~er. in the next chapter. 2. it is perfectly legitimate to consider structural sequence and stereo isomerism as special cases of copolymerization.HM2] = k21[M~HMl] Derive the copolymer equation for the terminal model.94) If fAis plotted against x (or ~ against l/x) we should obtain a straight line with an intercept of unity and a slope ofrA (or rB) for the terminal model. The copolymers were isolated at low conversions and the copoly~er composition determined (m. respectively). two further reactivity ratios.

Write a computer program to calculate the copolymer composition of monomers A and B as a function of conversion. Given that P{A/B} = x / (ra + x) and P{B/A} = 1 / (1 + rAx). You will be expected to hand in a listing of your program and the output.4 0. (b) The method of Fineman and Ross.e.16.8 1. respectively. J. From the literature the appropriate reactivity ratios are 0.22 (B)3 = 0. . Koenig.020 and 0. Assume an initial equimolar feed ratio and reactivity ratios of rA = 4. R.8 2. express the triad probability P3 {ABB} in terms of the sum of two tetrad probabilities. calculate the instantaneous copolymer composition for a monomer feed ratio of 40:60 parts by weight of methacrylic acid (MAA) and styrene (St) at 70°C. Wiley & Sons. 6. derive the Mayo-Lewis copolymerization equation starting from P2{AB} = P2{BA}.00 1. (a) Derive equations relating IAto the reactivity ratio(s) for the terminal and penultimate models.34 3. The following data was calculated for a copolymer synthesized from A and B assuming the terminal model. From nmr spectroscopy the number fraction of BAAAB pentad and AB diad sequences was determined to be 0. Number Ave Length of Runs: IA= 10. Principles of Polymerization. (b) The following experimental data was obtained from copolymerization studies at very low degrees of conversion involving monomers A and B: x (= AlB mole ratio) 2. New York. Sequences: (B)I = 0.4 1.44 / and: R = 2P}{B}P{AIB} Does the data fit the terminal or penultimate model? Estimate the values ofrA and rAt. What is the number fraction NA(n) of sequences of three A's (i.76 Number Fraction of (A).0 0'. Chemical Microstructure of Polymer Chains. 8. Copolymer?mposition: A =0.15 for MAA and St.08 What do these data tell you ? 10. (e) What do the reactivity ratios tell you about the microstructure of the resultant copolymers? 3. Sequences: (A)I = 0.10 (A)z = 0. Odian. ra X = P{A/B} +P{B/A} = F. G.06. (b) Starting from PdA} show that P2{AB} = P2{BA}. Comment upon the results you obtain. Assuming that the terminal model applies.09 (A)3 = 0. Wiley & Sons. You will not receive extra credit for beautiful graphics and tables. In addition. the deviation from randomness parameter. IAis defined as the number average length of A runs.2 1. 3rd Edition. Answer the following questions which pertain to the section on general statistical relationships: (a) Assuming the presence of only two monomers A and B. show that: 2 = ---=-ra rAx+2+xrAx+2+x-~-___!!'--- 9.6 0.74 2. respectively.76 1. SUGGESTIONS (1) (2) FOR FURTHER READING ra rAx + 1 +rAra+x- 7. 1991. 1982. L. but you may impress your professor! J.5 Deviation from Randomness: X = 0. 4.2 JA 3.142 COPOLYMERIZATION Suggestions/or Further Reading 143 Determine the reactivity ratios rl and r2 by: (a) The method of Mayo and Lewis. (c) Assuming only two monomers A and B. J. (d) Express the pentad Ps{BABAB} in terms of simple (first order) Markovian statistics.0 and ra = 0. calculate the run number.05 2. calculate the values of IA' Ia and X for each interval of the degree of conversion.6 ~ = 1.08 Number Fraction of (B). X. (c) The method of Kelen and Tiid6s. and the number average lengths of MAA and St runs.13. give an example of the equality of a reversible sequence using a pentad sequence.80 3. if applicable. AAA) ? 5.40 2. (d) Comment on the accuracy of the results obtained from the three different methods. In addition.7 and 0.66 (B)z = 0.87 B =0. Using the Mayo-Lewis copolymerization equation for the terminal model.

We will commence this discussion with a review of the general fundamentals of spectroscopy. Students that are not familiar with some of these fundamentals may wish to jump ahead and read Chapter 7 first. but do so in rather different ways. When a beam of light is focused on a sample a number of things can happen. INTRODUCTION Spectroscopy. but also those involving molecular vibrations.CHAPTER 6 Spectroscopy and the Characterization of Chain Structure "I'm picking up good vibrations. She's giving me excitations" -The Beach Boys A. or if the sample is 145 . It will involve some mention of conformational order. where we will also briefly mention other spectroscopic methods. It can be reflected. probes electronic transitions and provides direct. in all its forms. for example. characterizing microstructure. the reorientation of nuclei in magnetic fields. methods for identification. if not indispensable. is a powerful tool for the study of structure. Atomic spectroscopy. so our discussion of this technique will involve some of these specifics. infrared and nuclear magnetic resonance (nmr) spectroscopy. B. these techniques provide important. this is a specialization that includes an extensive body of knowledge. Our concern here is molecular spectroscopy. Our discussion of infrared spectroscopy will be more general. crystallinity and intermolecular interactions. Clearly. intermolecular interactions and other wonders too numerous to mention. because it does not provide this level of detailed analysis. topics that are introduced in the following chapter. so all we can describe in this text is an outline of the basics and an indication of how some of these tools are used in the most important tasks of polymer characterization. In polymer science. and other phenomena. compelling evidence of the quantization of energy. orientation. which encompasses not only electronic transitions. We shall focus our attention on what we believe are the two most significant techniques. The results of nmr analysis can be related directly to sequence distributions and the use of probability theory discussed in the last chapter. FUNDAMENTALS OF SPECTROSCOPY ! Spectroscopy is the study of the interaction of light (in the general sense of electro-magnetic radiation) with matter.

the les~ strongly bound electrons found in the 1t orbitals of conjugated system absorbing In the visible range of light. breaking bonds. in the absence of an external field.000 1.r . '" There is such a thing as emission spectroscopy. simply transmitted with no change in energy. If radiowaves having a frequency corresponding to the energy difference between the aligned and unaligned states (i. find th~t y-rays cause transitions between energy states within nuclei (the baSIS for Mossbauer spectroscopy). Hz). In conventional light scattering the scattered radiation that IS measured has the same frequency or energy as the incident beam. A.1. ------- = hu (6.trum. We will discuss the elements of light scattering later on in Chapte.10. Be that as it may. ++ Radiation in this energy range can also beat the living hell OUl of a polymer. defined as: (6. in principle. in terms of c u= ~ = / i (6.000 10. This latter effect provides the baSIS for Raman J~o ] Television Wavenumber 10 100 1. This is described by the Bohr frequency condition: L1E = E . i. but that will not concern us here.----j-------.00025 cm.. for example.1 to 0.e. there are transitions associated with the rotational energy levels of small molecules (not polymers!). This is because specific nuclei (protons.g. electrons ~f the atoms of a sample'". called modes.-----------------~~~~---------------------------] Visible ] Ultraviolet spectroscopy. But.. effects. . The electromagnetic spec~m is a continuum but humankind has given various parts of it names. are excited by infrared radiation. . knocking out atoms and causing other.) have magnetic dipole moments by virtue of their "spin" (but. A): lX-RaYS t 'Y-Rays A. as the scattering of electromagnetic radiation is a powerful tool for characterizing polymer structure at the level of overall chain conformation. although a small portion of the light can also exchange energy ~ith the samp~e and scatter at a different frequency..00001 0. there's strong coupling or resonance '"However.000 10. we.0001 0. crrr").001 0. a~d samp~e morphology.1 Frequency ranges for different types of spectroscopy. !2C and !60. and provide the basis for uv-visible spectroscopy. while absorptions of x-rays involve th~ inner shell. Visible and ultra-violet (uv) hght also excite electronic transitions. etc. u (in units of inverse length. . usually in the wavelength range of 0. !5N. em. The energy of the dipolar nucleus depends on whether or not it is aligned with an external magnetic field. deuterons. for example. also. Some of the light may also be absorbed or scattered. Molecules have vibrational energy levels.E 2 I . so that specific types of vibrational motion of the chemical bonds of a molecule. L1E= hu) now impinge on the sample. However.the incident light. The light is only absorbed if its energy and hence frequency correspond to the energy difference betwee~ ~wo quantum levels in the sample. or associated parts of it with various applications.000001 0.e. usually undesirabl~.lhls can be useful if applied cleverly. as this IS the basis for absorption spectroscopy".. however.000 (crrr+) --------~.. 0. find a wealth of absorptions in this radiowave region of the spectrum.. or (in infrared spectroscopy) wavenumber.-----------~--. 10 to 4000 crrr! (wavenumbers).1 om 0. Continuing into the microwave and radiofrequency range.2) Figure 6. Starting at the high energy or short wavelength region ~f ~he spec. the lagging of molecular dipole moments behind the oscillations of the electric field of light leads to some dissipation of energy and this is the basis for dielectric relaxation experiments. e. As frequency is decreased (or wavelength increased) the energy of the radiation is no longer sufficient to excite electrons.. as illustrated in figure 6.. or as it is more commonly expressed.000 100 10 147 transparent to the frequency or frequencies of . however.00000001 J. we would now. (in units of length. if we now trundle up a bloody great magnet and apply a large magnetic field to the sample. This relationship. as in the use of 'Y-radiation to cross-link polyethylene and Increase its environmental stability.3) (Molecular Processes and the Absorption of Radiation There are various types of energy changes that are excited in molecules as a result of an interaction with light and the transitions that occur depend upon the energy or wavelength of the incident radiation.1) where h is Planck's constant u and is the frequency of the light in cycles per second (Hertz.146 SPECTROSCOPY Fundamentals of Spectroscopy Wa~elength (ern) )-00.0000001 0. and many spectroscopic measurements are often expressed in terms of wavelength. !3C. Our interest is the light that is absorbed by the sample. = ~ where is the velocity of light. while higher energy (frequency) uv light is required to excite more strongly bound electrons. !9F.. have zero spin). there are no other absorptions associated with molecular processes in this frequency range".

esr.2 Schematic diagram of a spectrometer. Finally. Second. then you don't need a dispersion device (i. Such devices were indeed used in the first infrared instruments. and the key factors are interactions between the components of the . so that as the prism is rotated different frequencies are detected. and absorption occurs. assuming there is no interaction between the i components.'-e +~~I ~rn~ 1 Sample () Spectrum Dispersive Element Figure 6. Assuming you can obtain such a source (and a lot of applied physics has gone into developing good sources and detectors) you need to arrange the optics (for infrared or uv-visible spectroscopy) so that the light is focused onto the sample. electron spin resonance.2) at all. We use a picture of a prism which focuses light through a slit onto a detector to represent this in figure 6. we will see that the vibrations of certain types of bonds inherently absorb more radiation than others. or absorbed by it. First. replaced by gratings ~d more recently by interferometers (which don't require slits at all).5) Frequency or wavelength Figure 6. as a function of frequency or wavelength. the bands have a finite width. the intensities of the bands are all different.e. but were subsequently The spectrum of a material is a plot of the intensity of light transmitted through the sample. for example. and these various factors is given by the Beer-Lambert law: A = abc (6. the light or radiation consists of a range of frequencies. The relationship between the absorption (defined as the logarithm of the ratio of the intensity of transmitted to incident light). but the units change accordingly. The spectra of polymers are usually considerably broader than their low molecular weight counterparts.. For multicomponent systems. The first thing you need is a source of radiation in the frequency range you're interested in. Accordingly. In infrared spectroscopy.4) where A is the intensity (in terms of the amount of radiation absorbed. as the spin "flips". it must obviously depend upon the number of molecules or functional groups or nuclei giving rise to that particular absorption. 1= 1 1 11 n (6. First. such as a tunable laser. only some of which are absorbed by . however. occurs in the higher frequency microwave region and is a useful tool for studying unpaired electrons (free radicals). so you need to separate the light according to wavelength and measure the intensity of light at that wavelength. We'll discuss this initially. It must also depend upon the chemical character of the group or nature of the species. the absorption of any particular band can be written: Light Absorbed A= L a. the prism in figure 6. but it is important to have a feel for the nature of the experiment. whose basic elements are illustrated schematically in figure 6. and c is the concentration of the component of interest. integrated areas or peak heights can be used.have been absorbed.2. it must depend upon the concentration of that species and the total amount (thickness) of the sample in the spectrophotometer beam. * A close cousin.148 SPECTROSCOPY Fundamentals of Spectroscopy 149 so~___"'. Note that for various reasons this law does not hold for very thick or concentrated samples (in infrared and uv-visible spectroscopy) and sample preparation and technique are crucial to good quantitative work. and so on. the amount of absorption must also depend upon the intensity of the incident beam. Usually. We'll get back to this in a while. a is the absorptivity or extinction coefficient which depends on the nature of the absorbing species. We have drawn a schematic representation of an absorption spectrum in figure 6. This is nuclear magnetic resonance or nmr". There are instrumental effects.2.b. Now it is necessary to determine how much of which frequencies . not transmitted) of the band of interest. L_ Intensities and Band Shapes between the radiation and the nuclear spins. This is essentially the way nmr experiments are performed.. Turning our attention to bandwidth. b is the sample thickness.the sample.c. If you have a source that emits radiation of a given frequency and if this frequency can be changed so as to "sweep" through a given range. there are numerous reasons why bands are not infinitely narrow.3 Schematic diagram of a spectrum. factors associated with the finite lifetimes of excited states. The intensity of a spectral line or band depends on a number of things. Spectrometers The technicalities of spectroscopic instrumentation and the fundamentals of the spectroscopists' art belong in a book devoted specifically to the subject.3 and there are two things you should notice.

in the anti symmetric stretching vibration. providing a routine tool for the identification of materials and the characterization of microstructure. but this condition is not by itself sufficient. it is interesting to note that it also involves molecular vibrations. The symmetry of the molecule is maintained in this vibration and there is no net change in dipole moment. vibrational modes active in the infrared will be Raman inactive and vice versa.C .. this is not to say that scientists have hurled their infrared spectrometers out of their laboratory windows. The selection rules for the symmetric and anti symmetric stretching modes of CO2 were determined just by inspection. you could apparently excite all sorts of different vibrations by walking up to it and giving it a good whack with different degrees of force from different angles. The way we will approach it here is to describe the characteristics of the infrared spectra of various polymers. In fact.. In the totally symmetric vibration. But all good things must end and for many of these applications infrared spectroscopy has been superseded by the greater detail offered by nrnr spectroscopy. It provided numerous insights in the early days of polymer science (we mentioned the detection of short chain branches in polyethylene in Chapter 2.. The rules determining optical activity are known as selection rules.. because otherwise the authors of this text would be looking for work!) and the technique remains a fundamental characterization tool. Specifically. for a vibrational mode to be Raman active there must be a change of polarizability with molecular vibration. A discussion of this requires a more specialized treatment. and we will not discuss the equivalent quantum mechanical description in this text. which slates that if the molecule has a center of symmetry.150 SPECTROSCOPY Basic Infrared Spectroscopy 151 system coupled to the fact that even ordered polymers are only semi-crystalline (see Chapter 7). if you made a model of the CO2 molecule by suspending in the air three balls of appropriate weight attached by springs. hence no interaction with infrared radiation occurs for this motion. We will see an example of this later. Now let us look at the nature of molecular vibrations in a little more detail. Sl:--0 O~~C Infrared Active Raman Inactive C.. We start by considering some of the principles of infrared absorption. Symmetric Stretching Infrared Inactive Raman Active Antisymmetric Stretching Conditions for Infrared Absorption We mentioned in our discussion of basic spectroscopy that a fundamental and inherent condition for infrared absorption is that the frequency of the absorbed radiation must correspond to the frequency of what is called a normal mode of vibration and hence.. It is easier to obtain a physical picture of these interactions by considering the classical interpretation. it is easy to see that the symmetric stretch does not result in a change in dipole moment. a transition between vibrational energy levels. giving rise to slightly different frequencies of absorption and hence broader bands.. but said virtually nothing about the nature of a normal mode or how we determine what it is. it will only be absorbed under certain conditions. for example). with molecular vibration is fundamental. which have their origin in quantum mechanics. For more complex molecules. Certain normal modes do not result in such a change. so infrared spectroscopists still find things to do (which is just as well. For this simple molecule. and the important point you should grasp here is that infrared absorption involves a change in dipole moment with molecular vibration". We will consider molecular vibrations in a little more detail shortly. and an additional fundamental requirement is that there must be some mode of interaction between the impinging radiation and the molecule. Nevertheless. . The requirement of a change in dipole moment The net dipole moment of the symmetrical unperturbed molecule is zero.. Conversely. • Although we have only made a passing reference to Raman scattering. >-0 Infrared spectroscopy is an old and familar friend to the organic chemist. In the amorphous state the elements of the chains are in slightly different local environments. and for many years was the workhorse of all polymer synthesis laboratories. They are still ideally suited for specific tasks and provide certain types of information (including measurements of orientation and strong specific interactions). the two oxygen atoms move in phase sucessively from and towards the carbon atom. as illus&ated by the two in-plane stretching vibrations of CO2 depicted below: 1387 cm! . the symmetry of the molecule is perturbed and there is a change in the net dipole moment with consequent infrared absorption. Of particular interest is the rule of mutual exclusion. however. to give you a feel for the type of information that can be gained from infrared studies. Normal Modes of Vibration We presented above a schematic picture of two normal modes of vibration of the CO2 molecule. a tool which is also important in other characterization techniques (such as x-ray diffraction). Consequently. that cannot be obtained by nrnr. Infrared absorption is simply described by classical electromagnetic theory: an oscillating dipole is an emitter or absorber of radiation.. Even if infrared radiation with the same frequency as a fundamental normal vibration is incident on the sample. the periodic variation of the dipole moment of a vibrating molecule results in the absorption or emission of radiation of the same frequency as that of the oscillation of the dipole moment. however. BASIC INFRARED SPECTROSCOPY o ----il>-. but the selection rules are different from infrared. the number and activity of the normal modes can actually be predicted from symmetry considerations alone using the methods of group theory.

3-dimethylbutadiene). Note that these do not include translational or rotational movement of the molecule as a whole. L. P. can be resolved into components parallel to the x. Koenig. however. the intensity of an infrared band depends upon the extent of the displacement of atoms in a particular vibration.4-poly(2. For a theoretically infinite polymer chain. consider the displacement of the atoms. m. In the early days of infrared spectroscopy. . Figure 6. The nuclei are considered to be point masses and the forces acting between them springs that obey Hooke's Law.4 Selected normal modes of trans-I . the amplitude may be different from atom to atom.4 degrees of vibrational freedom. C.4-poly(2. and m2. although not necessarily significant displacements. before all this was worked out. The motion of any atom in a molecule. called the normal modes of vibration. However. only the three translations and one rotation have zero frequency and there are 3N . As an example. in general. New York 1982.Wiley & Sons. different types of vibrations will have different energies and so absorb or inelastically scatter radiation at different frequencies. which are greatly exaggerated in this figure. A system of N nuclei therefore has 3N degrees of freedom. The number of normal modes available to a molecule depends upon the number of atoms it contains and hence its degree of freedom. however. be it the result of an internal vibration or rotations and translations. rotation about the molecular axis does not change the position of the atoms and only two degrees of freedom are required to describe any rotation. the z axis is coincident with the molecular axis and movements of the atoms in the x and y directions are not allowed.6 vibrational degrees of freedom or normal modes of vibration. which represent the normal modes (and hence absorptions) occurring at 1163. Coleman and I.4. so there are 3N . meaning that the force on the atoms is proportional to their displacement from an equilibrium position. each particle carries out a simple harmonic motion of the same frequency and. but just motions of the balls (or atoms) relative to one another. that all the motions that you excite by whacking can be broken down into the sum (or difference) of just a few types of motion. Consequently. Painter. The details of normal mode calculations are beyond the scope of this book". Naturally. as can be seen from figure 6. so we will consider a simple example of two point masses. All the atoms may be vibrating with the same frequency. '__/ . For strictly linear molecules. M. linear molecules have 3N . 1074.. but the largest displacements from an equilibrium position can be localized in a small group of atoms. such as carbon dioxide. connected by a Hookean spring as shown diagrammatically below.5 normal vibrations. M. the vibrations can be calculated in a conceptually simple manner using the methods of classical mechanics. it was empirically determined that certain functional groups absorb infrared radiation at characteristic frequencies (the basis of the so-called group frequency approach widely used for identification of unknown samples). 1031 and 925 cm-1 in the infrared spectrum of trans-1. shown diagrammatically in figure 6.4. these oscillations are in phase. In a normal vibration. For this one-dimensional model.152 SPECTROSCOPY Basic Infrared Spectroscopy A Simple Model 153 ) It turns out. I. 1163 The normal modes of a polymer can appear to be complicated and you may be wondering how we determine the form of the vibration. Since normal vibrations involve displacements of all atoms it might be expected that the constitution of the rest of the molecule might have a more profound effect than the relatively small frequency shifts often observed for these groups. z )-. as larger displacements would lead to a bigger change in dipole moment. These normal modes of vibration are excited upon infrared absorption (or Raman scattering). but it is important to obtain a feel for the factors that affect the frequencies of a band. As it turns out. of all the atoms in the system. However. y and z axis of a Cartesian system and the atom is described as having three degrees of freedom.. The displacements of the atoms in the course of the vibration are proportional to the length of the arrows. The motion of each atom is assumed to be simple harmonic.Those with truly masochistic tendencies might wish to consult another of our books for further details. for a non-linear molecule' six' of these degrees of freedom correspond to translations and rotations of the molecule as a whole and therefore have zero vibrational frequency. . Let us define the displacement of the two atoms relative to each other by z = [ (z'. The Theory of Vibrational Spectroscopy and Its Application to Polymeric Materials. Each normal mode consists of vibrations.3-dimethylbutadiene).

: 0 we obtain what should be the familiar equation for harmonic motion: u=l!I_ 21t -. gross identification of polymeric materials has become a rather trivial task using modem computer assisted instruments. As most commonly used. Symmetry plays a major role in the observed spectra of ordered systems. no frictional forces) the following equation holds: . for a conservative field (i. atactic polystyrene (a-PS). The C-O stretching vibrations are found near 1100 crrr! while C=O frequencies are characteristically observed at about 1700 cm+.7) This is a 2nd order differential equation which has a periodic solution of the form: (6. say by formation of a double bond. infrared spectroscopy may be employed to characterize the structure of polymeric materials.8) z = A cos(2m:>t + E) where A is an amplitude. however. i.fz = fI\.e. leads to a consideration of the spectroscopic features observed when a polymer orients or crystallizes. Finally. For more Amorphous Polymers with Weak Intermolecular Interactions What we are considering here are non-crystalline polymeric materials in which van der Waals forces predominate". . Conversely.. or oxygen. One is initially struck by the large number of infrared bands present and the apparent complexity of the spectrum. D. Differentiating twice with respect to time leads to: -2 iz dt = . Examples would include atactic polystyrene. Infrared spectroscopy may also be employed to study the changes occurring upon chemical modification. CHARACTERIZATION SPECTROSCOPY OF POLYMERS BY INFRARED iz (6. This. By this we mean not only the overall chemical composition of the polymer chain but also how individual units are distributed.e. the vibrational frequencies naturally depend upon the type of motion and the geometry.~) ] where Zl . Thus.. the reduced mass. the exerted force is given by -fz where f is the force constant or ("stiffness") of the spring. jump ahead to the beginning of the next chapter where we review this material. so that the spectrum is sensitive to the conformation of polymer chains.. where X can be substituents such as hydrogen. one can also take the opposite view • We assume at this point that you have a basic knowledge of the various types of inter." At the next level of sophistication. This brings us to polymers.10) Assuming z . at the highest level of sophistication. It is also convenient at this stage to define m. because the voluminous number of standard spectra available in the literature demanded an excellent memory and the dedication of a Sherlock Holmes. as the mass of the substituent is increased there should be a decrease in frequency (or wavenumber). Although the model considered is extremely simple. degradation and oxidation of polymers. Libraries of standard spectra. At one time this was an art.. increases the frequency. Let us commence by considering the spectrum.-2 dt complex vibrating systems.6) You could obtain this equation by considering the individual displacements of the atoms. Upon further thought. molecular interactions. in turn.~ is the equilibrium separation between ml and m2 and (z'. Today. of a film of an amorphous polymer glass. Rearranging: Il\ i -T + fz dt =0 (6.(Zl . In fact. styrene-co-butadiene and ethylene-co-propylene copolymers etc. increasing the force constant between atoms. stored in the computer memory and used in conjunction with search routines. have reduced the problem to that of applying a "black box. it is now a rapid and easy method for the qualitative identification of major components through the use of group frequencies and distinctive patterns in the "fingerprint" region of the spectrum. the "spring". but when you combine everything together you end up with this expression in terms of the reduced mass.11) Infrared spectroscopy can be applied to the characterization of polymeric materials at various levels of sophistication. shown in figure 6. Accordingly.41t 22 Uz (6.9) Substituting we obtain: (6. the C-H stretching modes absorb near 2900 crrr! while C-CI stretching vibrations occur near'600' cm+.Jll\ (6. it is feasible to obtain information concerning the nature and concentration of structural and conformational units present in a particular sample.154 SPECTROSCOPY Characterization of Polymers by Infrared Spectroscopy 155 Z'2) . Assuming the spring obeys Hooke's Law. The chemical bond can be assumed to be the focus of the forces acting between the atoms. chlorine. in addition to the mass of the atoms and the forces acting between them. infrared spectroscopy is used in conjunction with computer calculations of normal modes in order to understand the fundamental features of the vibrational motion of polymers. consider as an example the isolated stretching vibration of the bond C-X. If you don't. To see how this helps us. equal to (mlm0/(ml + m2)' Since force is defined as mass times acceleration. it provides us with the important result that the vibrational frequency depends inversely on mass and directly on the force constant. so we will start with a simple polymer system.5.z'2) is the distance after a finite extension or compression. or have forgotten. E is a phase angle and u is the frequency of vibration. . As mentioned above our approach here is more of a survey.

In the amorphous state the polymer resembles a "bowl of spaghetti". that the adjacent unit exists.. Amorphous Polymers with Strong Intermolecular Interactions = In this case.e. we will assume that the a-PS sample is chemically "pure" and that there are no additional bands in the spectrum attributable to oxidation.4 44 fundamental vibrations.2). the gross features of the spectrum will reflect those of a low molecular weight analogue of the polymer repeat unit. 3800-2800 and 2000-450 cm+.. The figure is split into two frequency ranges. Examples include atactic poly(4-vinyl phenol) (PVPh). we can assume that a given repeat unit does not know. carbonyls etc. As there is no translational symmetry between the units of the chain or inherent symmetry in the structure of the repeat unit. PYPh is structurally similar to a-PS. B = 50aC. amorphous polyamides and polyurethanes . in infrared terms. symmetry may dictate that in ordered materials particular normal modes are inactive in the infrared.. we must not oversimplify the situation. although in certain cases.A word of caution: the C-H stretching region is often complicated by something called Fermi resonance. In simple terms. the higher frequency region has been expanded on the absorbance scale by a factor of about four times that of the lower = . bands characteristic of certain types of local conformations may be resolved.. n Poly(4-vinyl phenol) Infrared spectra of PYPh. bands that contain significant contributions from vibrations that involve the polymer backbone. the above is a reasonable "rule of thumb" and the corollary is that the relatively broad bands in the spectrum are conformationally sensitive. amides. because as we have mentioned. poly(vinyl alcohol). which arises from the interaction of a fundamenlal and an overtone occurring at about the same frequency. for example-see later) and. however. are unaffected by the shape of the chain). Let us return to the spectrum of a-PS' and ask the question. from group frequency correlations we can determine that the sample contains aliphatic and aromatic groups from the bands observed in the 2800 to 3200 crrr! region of the spectrum. There is an element of truth in both viewpoints. Be that as it may. all these normal modes are predicted to be infrared active. In infrared spectroscopic terms. The polymer chain of a-PS is composed of chemical repeat units (-CH2CHC6Hs-) that contain 16 atoms (see table 1. Third. The chemical repeat unit is relatively large (compared to that of polyethylene. We must not be dogmatic. Secondly. but we neglect these here.will' reflect the distribution of conformational states by broadening. to a first approximation. we can initially eliminate such groups as hydroxyls. nitriles..156 SPECTROSCOPY Characterization of Polymers by Infrared Spectroscopy 157 3800 3000 2000 1000 crrr ' Figure 6. amines. epoxy resins. Thus.. degradation. conformational states and sequences of stereoisomers. and consider the spectrum to be remarkably simple. except that there is a hydroxyl group present in the para position of the aromatic ring. It is also readily apparent that the spectrum of a-PS is characterized by infrared bands that vary quite substantially in breadth. as normal coordinate calculations indicate a degree of mixing of specific ring modes with backbone vibrations. the presence of a group of distinctive and relatively sharp bands that are characteristic of monosubstituted aromatic rings readily leads one to the conclusion that the spectrum resembles that of a styrenic polymer.. are displayed in figure 6. There are also bands due to overtones and combinations of the fundamental modes. we consider amorphous polymers in which there are hydrogen bonds as these are the most commonly found strong intermolecular interactions in polymers. In addition. For clarity of presentation. the relatively narrow bands can be attributed to localized normal modes that are conformationally insensitive (i. These would include such modes as the C-H stretching and the aromatic ring breathing vibrations. Polymer chains essentially obey random flight statistics (see next chapter) and there is a myriad of conformations taken up by the individual units in the chain. which all have distinctive group frequencies.6. incorporation of impurities or the presence of chemically distinct end groups etc.. recorded as a function of temperature (A 30aC. given that we are dealing with an amorphous material where the average molecular weight might be in excess of 105 g/mole and where there exist broad distributions of molecular weight. C = lOoac. we might anticipate 3N . . Accordingly. D = 150aC. Nonetheless. "What can we immediately gain from just a cursory glance?" First of all. E = 2ooaC.5 The infrared spectrum of atatic polystyrene. F = 250aC).

such as isotactic polystyrene (i-PS).. attributable to normal vibrations that are conformationally insensitive. do not have strong intermolecular interactions such as hydrogen bonds) and which contain large chemical repeating units. In the vast majority of cases infrared spectroscopy cannot be used as an absolute method to determine three dimensional crystallinity. Accordingly. but they are truly subtle. assigned to a broad distribution of hydrogen bonded hydroxyl groups (hydrogen bonding between chains and functional groups of the same type is often called "self-association") and a much narrower band at 3525 crrr! which is attributed to "free" non-hydrogen bonded hydroxyl groups. and at the same time. one may confidently assign the bands occurring at 825. In fact.158 SPECTROSCOPY Characterization of Polymers by Infrared Spectroscopy 159 frequency region. the relatively broad bands in the 1200-1400 crrr! region of the spectrum change markedly as a function of temperature. as illustrated in figure 6. however. These bands can be assigned to mixed vibrations containing contributions from O-H deformation and c-o stretching vibrations. In contrast to a-PS. an ordered polymer containing a relatively small chemical repeating unit. it is not as dramatic as it might appear at first glance from figure 6. This latter band appears to increase at the expense of that attributed to the hydrogen bonded groups as the temperature is increased.5 and 6. as it is consistent with our preconceived notions of temperature effects upon equilibrium. mixing polymers to form blends).g.7. This structural regularity results in the possibility of crystalline order and i-PS prefers to fold into a regular shape called a 31 helical chain conformation (more on this in Chapter 7) which has a crystalline melting point (Tm) of 230°C.6. 1445 and 1595/1610 cm+. This is very useful in a number of areas (e. which are essentially temperature independent. 1100. This region is characterized at room temperature by a very broad band centered at 3360 cm+. We will initially consider the general features that are seen in the spectra of ordered polymers that do not "selfassociate" (i. trans-l. As the temperature and hence molecular motion increase the average strength of the hydrogen bonds decreases. above the glass transition temperature (T g) but below the Tm. This spectrum may be considered to .6 The infrared spectrum of atatic poly(4-vinyl phenol) as a function of temperature: ~~~~~~~~~~~~~~~~~~~ Let us proceed with an examination of the infrared spectrum of i-PS. there are some very subtle differences in a few infrared bands which may be ascribed to weak vibrational coupling between different stereochemical repeating units. While this is intuitively pleasing. Preferred Conformation: Isotactic Polystyrene 3800 em-I (a) 3000 1800 em-I (b) 500 Figure 6. 1170. which is reflected by a shift to higher frequency. leading to a concomitant decrease in area of the hydrogen bonded band.7(B) leaves little doubt that one would be hard pressed to differentiate between them.e. and to all intents and purposes. In addition. And it certainly does. temperature may be attributed to two main factors: conformational sensitivity and intermolecular interactions. where the chemical repeat units are arranged in a random sequence of stereochemical placements. If one rapidly quenches a sample of i-PS from above the Tm to room temperature. Ordered Polymers with Weak Intermolecular Interactions The effect of order or "crystallinity" on the infrared spectrum of a polymer is one of the most interesting. Using the group frequency approach. A comparison of the spectra presented in figures 6. The changes observed with. Crystallinity vs. In contrast. in pure i-PS every adjacent unit is isotactic to the first.7(A). we now know that the absorption coefficient is a strong function of the strength of the hydrogen bond. This causes greater overlap with the "free" band at 3525 cm+. and broader bands that are sensitive to the distribution of conformations. to the aromatic ring. there is a wealth of information available from infrared temperature studies of what we call strongly self-associated polymers that provides direct evidence of the number and strength of the hydrogen bonds present. Although this may appear complicated. If we now anneal the quenched i-PS. More on this later. The spectrum of the annealed (semicrystalline) i-PS is shown in figure 6.4-polyisoprene and trans-l..3-dimethyl butadiene). where the O-H stretching frequency appears. Then we will turn our attention to polyethylene. We might therefore anticipate that the spectrum of quenched i-PS should closely resemble that of a-PS. crystallization is prevented and the material becomes an amorphous glass. most generally misunderstood topics in polymer vibrational spectroscopy.4-poly(2. can be ignored. In common with the spectrum of a-PS. there are infrared bands that are relatively narrow. the absorption coefficient decreases. Perhaps the most interesting region of the spectrum of PVPh is that occurring between 3100 and 3600 cm+. crystallization occurs and the sample transforms into a semi-crystalline material. as the strength of the hydrogen bond diminishes.

The P form contains one monomer per translational repeat unit with torsional angles of 1800 between adjacent CHz groups and +1050 between the CHz and -C=C groups'".. There are obvious and distinct differences between the two spectra (examine.. The difference spectrum is obtained by subtracting the spectrum of the quenched (amorphous) i-PS sample from that of the annealed (semi-crystalline).8 The infrared spectrum of the 1. form contains two monomers per translational repeat unit and the preferred conformation is reported to be trans-CfS-trans-CfS. crystallinity results in the . structurally regular repeat unit. 878 be composed of two major "non-interacting" components. for example. To be pedantic.. this spectrum has little to do with three dimensional order as such. contains more than one chemical repeat unit. infrared spectroscopy is not sensitive to polymorphism if there is not a parallel change in polymer chain conformation. (A) Annealed.. but to the different preferred chain conformations present in the two forms. These are spectra that can be directly assigned to two dissimilar preferred polymer chain conformations. these two components do not recognize. it suggests that there must also be significant differences in the chain conformation of the polymorphs. Since the intermolecular interactions are relatively weak. This spectrum.. 796 1450 1448 1382 1207 882 862 800 1672 Figure 6. amorphous and "crystalline". and p crystalline forms are shown in figure 6. in infrared spectroscopic terms. the ex. for all intents and purposes. presence of sharp bands in the infrared spectrum which can be attributed to a narrowdistribution of preferred conformations. its just that in crystalline domains structural units are in the preferred conformation. because extended ordered sequences..4-polyisoprene (TPI) is an excellent example of a polymer that exists in polymorphic forms which are easy to prepare.L__!-_l_-L I_-L _ _l__l____l_-'- _ _L___l'-. Polymorphism. the existence of one another in the semi-crystalline material. To reiterate.. intermolecular vibrational coupling between the units in the polymer chain. In contrast. if any. as the two components can be separated without significant distortion to the band shapes. and others like it. will only be found in crystalline domains). If we therefore observe significant differences in the infrared spectra of two polymorphic" forms of such a polymer.B). Difference spectra corresponding to the ex. a (bottom] and f3 (top) crystalline forms of trans- Polymorphism: Trans-l.8. In other words. are commonly referred to as the "crystalline" infrared spectrum.__ 3200 2600 2000 600 crrr ' Figure 6. A translational repeat unit often.4-Polyisoprene We have seen that in the case of a weakly interacting polymer containing a relatively large. but not always. so that the amorphous bands are removed (this takes a degree of spectroscopic judgement)..7 The infrared spectrum of isotactic polystyrene. Accordingly. This is somewhat misleading. the sharp bands seen in the "crystalline" difference spectrum actually reflect an individual repeat unit in "preferred" conformations. Trans-I. and (C) difference spectrum (A . ust as a chemical repeat unit can be taken as the basic element of the chain in terms of J chemical structure.160 SPECTROSCOPY Characterization of Polymers by Infrared Spectroscopy 161 A B c .4. . is when a polymer can have more than one different crystalline form. With very few exceptions.polyisoprene. (B) quenched from the melt. the transitional repeat unit describes the chain in terms of translational symmetry. . the region between 750 and 900 crrr+). and is also shown in figure 6. this is an ideal situation for what has become known as difference spectroscopy. the main point is that the infrared spectrum of TPI is sensitive not to crystalline form per se.7(C). in this context. This is the reason why infrared spectroscopy is not truly a direct measure of three dimensional crystallinity in such cases (although some useful empirical correlations have been established. in the case of a weakly self-associated polymer where there is little.

On the other hand. PE. rapidly and extensively. For example. this implies mutual exclusion of infrared bands and Raman lines and there are only five fundamental vibrations predicted to be active in the infrared spectrum. Under normal conditions PE crystallizes in an orthorhombic unit cell containing two chains. A combination of infrared and Raman studies is necessary to obtain information on the conformation of this type of polymer chain. 726 and 722 cm+.3-dimethylbutadiene) (TPDMB). PE. This is in marked contrast to those polymers that contain large chemical repeat units. It is perhaps ironic that one of the simplest polymers in terms of the structure of the chemical repeat unit.. especially the more structurally pure linear polymer. Nearly all infrared spectra were reported this way until about 20 years ago. (A total of 3(6) .9(A)] is dominated by a line at 1665 cm+. TPI and TPDMB. we still have to account for more than five obvious infrared bands. Transmission spectra are just absorption spectra plotted upside down on a log scale. Accordingly.e. beware of absent bands arising from the dictates of selection rules.4-poly(2. An Exception: 1463 1376 1331 1377 146 A Polyethylene 2BOO 2400 2000 1600 ern"! 1200 BOO 400 Figure 6. Even if we allow for the presence of overtone and combination bands and bands characteristic of amorphous conformations.3-dimethylbutadiene) Although we are not going to discuss Raman spectroscopy in this book. 8 are Raman active and 1 is inactive in both the infrared and Raman. oxidation and degradation etc. Above n = 5 the difference in frequency becomes insignificant. is at the same time. There is certainly no convincing evidence for a band in the 1660-1670 crrr! region of the spectrum. as n progresses from 1 to 5. however. you will find that this means that there are a number of normal vibrations in the spectrum of paraffins that are sensitive to the number and conformation of CH2 groups in a sequence. Nonetheless. If you study polymer vibrational spectroscopy in more detail. The preferred chain conformation of TPDMB is similar to that of the P form of TPI. and vice versa.) One glance at an infrared spectrum of a highly crystalline spectrum of PE (figure 6.4 = 14 normal modes are predicted. the infrared spectrum of PE is considerably different from that of a low molecular weight analogue. and vice versa (this is also what occurs in CO2 considered earlier). Unlike the polymers mentioned above. polymethylene. In the preferred chain conformation (planar zig-zag. end groups. We are seeing an excellent example of "mutual exclusion" where bands observed in the infrared spectrum are not observed in the Raman spectrum. if one suspects high symmetry and only infrared data are available. The additional methyl group on the C=C double bond.3-dimethylbutadiene) At first glance. crystallizes readily. see next chapter) there are two chemical repeats per translational repeat unit. respectively. The symmetry of the planar zig-zag polymer chain contains a center of inversion.10) will convince the reader that there are considerably more than five bands present! And we don't even show the region below 600 cm+. one of the most complex and interesting from an infrared spectroscopic point of view.9(B)*. this information is particularly useful in the characterization of. It should come as no surprise that PE and the low molecular weight paraffin analogues have been the subject of voluminous vibrational spectroscopic studies. there exists significant vibrational coupling between the units in the polymer chain. polymers containing sequences of CH2 units and has been successfully employed in the analysis of the sequence length distribution in ethylene/propylene copolymers. This dictates what is called mutual exclusion. 733. trans-l. the spectrum resembles that of a saturated hydrocarbon. There are now two ethylene units . The infrared spectrum is shown in transmission in figure 6. the frequency of the CH2 rocking mode in compounds represented by C(X)-(CH2)n-C(X) occurs at 815. 752.3-dimethyl butadiene). Suffice it to say.162 SPECTROSCOPY Characterization of Polymers by Infrared Spectroscopy 163 B 1665 Bands that appear in the infrared are absent in the Raman. i.9 The infrared (top) and Raman (bot/om) spectra of trans-l.4-Poly(2. it is interesting to consider the infrared and Raman spectrum of a highly c~stalline sample of trans-l. i-PS. The Effect of Symmetry: Poly(2. or as some people call it. contains a simple and relatively small CH2 chemical repeat. all we wish to do is highlight the differences between the spectral features observed in PE compared to those polymers discussed above. As we have mentioned above. leads to a center of inversion or symmetry. . Here. the Raman spectrum of the same polymer [figure 6. polyethylene (PE).4-poly-(2. where C=C double bonds absorb. however.. impurities..

164 SPECTROSCOPY Characterization of Polymers by Infrared Spectroscopy 165 intermolecular interactions. We will briefly discuss two regions of the infrared spectrum of nylon 11. The second is a much broader band (w1/2 = 38 ± 4 cm+) which systematically increases in intensity and frequency (from 1645 to 1654 crn+) as the temperature . large perturbations of certain vibrational modes of a single ·chain are found and we must consider both conformational and interactional effects. skewed to the high frequency side. between.e observe is the presence of additional bands and the phenomenon of what is termed crystal field splitting.. Actually. they are fascinating. this band can no longer be seen. It is important to recognize the distinction between crystal field splitting. Symmetry again rears its head and will dictate which of these normal modes is active in the infrared. The upshot is that the frequency of the Amide I mode for ordered structures is significantly different from that of disordered amorphous structures. we will not make serious errors if we consider the Amide I mode to be equivalent to the carbonyl stretching vibration. acid) .11.16oo and 1700 crrrI. we have a doubling of the number of the normal modes as one interacts with and perturbs the other (actually. In other words. With strong Extensive intermolecular hydrogen bonding occurs at ambient temperatures between the N-H group of one amide unit and the C=O of another. The polymer has a T g of 45°C and a T m of 196°C. a given ethylene unit of one chain "knows" that it is in close proximity to a similar group on an adjacent chain (there is an interaction between them). 2600 1600 Figure 6. Infrared spectra of nylon 11. recorded as a: function of temperature. these units do not "see" each other spectroscopically.5 700 Bonding: Nylon 11 nylon 11. that the development of order is reflected primarily in the infrared spectrum by changes in the distribution of conformations. The classic 733n21 and 1460/1475 em'! doublets in PE are examples of crystal field splitting. however. In any event. in highly ordered domains. at the same frequency). but this is enough for now). there remains much to be wrested from the infrared spectra of nylons and polyurethanes. there are three spectral contributions to this region which may be resolved by a curve fitting. The Amide I mode at 30°C is characterized by a relatively sharp band. which is greater than a one dimensional effect and categorically requires two or three dimensional order. Hydrogen 1460 1475 1300 1400 1000 900 733 721. Above the Tm (196°C). structure with a chemical repeat of -{CH2)w-NH-C0-. what w. Poly(aminoundecanoic that have to be considered in the calculation of the number of active infrared bands.nylon 11 Strongly Interacting Ordered Polymers The outstanding examples in this category are the structurally regular polyarnides (nylons) and polyurethanes. This is assigned to hydrogen bonded carbonyl groups in ordered domains. No longer can we simply assume. as in the majority of cases for polymers that do not selfassociate. the C-N stretching and C-C-N deformation vibrations. and between 3150 and 3500 cm+. but this conformational sensitivity does not arise through mechanical coupling to the main chain. and the bands due to a single chain preferred conformation. where the C=O groups are spatially arranged in a specific array. and their modes are superimposed (i. has a regular 2917 3200 2848 Poly(aminoundecanoic acid). are displayed in the range between 1600-1800 crrr! in figure 6. In essence. Instead it is due to differences in the pattern of hydrogen bonds that determine the relative arrangements of C=O groups and the degree of dipole-dipole interactions.e. a given C=O group "knows" that others exist within a sphere of influence through dipoledipole interactions.10 The infrared spectrum of polyethylene. or more commonly. The first is a narrow band (width at half-height. In brief. where we find the N-H stretching vibration. In most polymers. For our purposes. The Amide I mode is a mixed· mode comprised of contributions from the C=O stretching.. as PE has a small repeat unit that packs very efficiently in a crystalline lattice. Spectroscopically. At this time. Extensive interchain hydrogen bonding exists in these materials and the ability to form ordered ("crystalline") structures is generally accepted as being responsible for the useful physical properties of these materials. it is a bit more complicated than this. which simply depend upon interactions between adjacent groups in the one dimension along the chain. where the so-called Amide I mode absorbs. wl{2= 18 cm+) which systematically decreases in intensity and increases in frequency from 1636 to 1641 crrr! when the temperature is increased to 190°C. Significant changes in the spectra are observed with temperature. However. 12 infrared bands are predicted because the infrared spectrum depends upon the number of units in the crystal's translational repeat unit. and centered at 1638 crrr! . It is sensitive to local order.

The spectra are characterized by a very weak band occurring at 3447 cm+. in large part. Between those polymers that strongly self-associate and those that do not are a large number of polymers that may be categorized as moderately selfassociated. It is important to recognize that the pigeonholes we have chosen to categorize the major features of the infrared spectra of polymers are somewhat arbitrary and are only useful as "rules of thumb. In fact. which is consistent with a reduction of the average strength of the hydrogen bonds and a broadening of the distribution. vibration. is due to the "free" (non-hydrogen bonded) carbonyl groups.0 « 3500 3400 3300 em-I 3200 Figure 6.0 VI c: co « I I 1800 1700 em"! 1600 Figure 6. The answer lies in the differing sensitivities of the two normal vibrations." Many of the effects we have described so far can be assumed to be present in the spectra of moderately self-associated polymers. a direct consequence of a reduction of the value of the absorptivity coefficient with the reduction in the strength of the hydrogen bond. the loss of area is. It is interesting to note. One might ask the question. extraordinarily sensitive to hydrogen bonding. Finally. the N-H band envelope reflects the distribution of the strengths of the hydrogen bonded N-H groups regardless of whether they are in ordered or disordered domains. "Why do we see distinct bands associated with ordered and disordered structures in the Amide I region of the spectrum but not in the N-H stretching region?" There is. a one to one correspondence between hydrogen bonded N-H and C=O groups.25 o VI U c: Shades of Grey: Polymers with "Moderate" Interactions Intermolecular . Poly(vinyl chloride) (PVC) and simple aliphatic or aromatic polyesters are typical examples. essentially isolated. after all. attributed to "free" N-H groups. QJ . Figure 6. and a relatively broad dominant band which varies in frequency from 3300 to 3332 crrr! as the temperature is raised from 30 to 210°C. For .12 The infrared spectrum of nylon 11 in the N-H stretching region as afunction of temperature. that the breadth of the hydrogen bonded N-H band correlates nicely with the degree of order in the material. It is.166 SPECTROSCOPY Characterization of Polymers by Infrared Spectroscopy 167 1638 QJ U ~ 0. This . the third band which occurs between 1679 and 1683 crrr! (wl!2 = 26 ± 2 crrrt). It can be seen that there is a wealth of information to be gained from studies of the Amide I mode of polyamides.12 shows the infrared spectral region from 3150 to 3500 crrrJ of the nylon 11 sample recorded as a function of temperature. This latter band is assigned to hydrogen bonded N-H groups.11 The infrared spectrum of nylon 11 in the Amide I region as a function of temperature.band is attributed to hydrogen bonded carbonylgroups in disordered amorphous structures. The N-H stretching mode is a conformationally insensitive. however.35 .5 3300 is increased from 30 to 220°C. 0.e 0. Unlike the Amide I. it is just a question of which dominate and which can be ignored. Rather. however. There is also a marked reduction in absolute intensity (area) of the hydrogen bonded N-H band with increasing temperature which cannot be solely ascribed to the transformation of hydrogen bonded N-H groups to "free" groups. These would include polymers containing functional groups that can interact through weak hydrogen bonds and/or dipolar interactions. As the temperature is raised the band broadens and shifts to higher frequency. the hydrogen bonded N-H stretching vibration does not appear to be composed of two obvious contributions attributable to ordered and disordered hydrogen bonded amide groups.

4. these units efficiently find one another. 3. evidence for "free" MAA units at ambient temperatures. as infrared spectroscopists we should ask the question "What differences should we anticipate when we compare the spectra of homoand copolymers?" If we have two or more relatively large. From what we have seen so far we might expect a rather complex spectrum composed of features characteristic of an unassociated aliphatic chain containing a relatively small chemical repeat unit (like PE).13 shows the difference spectrum representative of the structural "irregularities" cis-1. but it is not ordinarily sensitive to sequence distributions. mentioned above.168 SPECTROSCOPY Characterization of Polymers by Infrared Spectroscopy 169 example.4. At the other extreme. The infrared spectra of EP copolymers are complex and contain information concerning sequence distribution and degree of order. In other words. Thus. In addition. In any event. Sequence and ordering effects pertaining to the ethylene portion of the copolymer are indeed observed. (C) Difference spectrum (B . polychloroprene (TPC) can also be considered a rather complex copolymer. So even though we have a random copolymer containing only 4 mole % of MAA units. the carbonyl stretching vibration of polyte-caprolactone) (PCL). 1652 I The spectrum is dominated by the characteristic bands of the individual chemical repeat units. This is not it bad approximation. It would be difficult to tell the difference between a random and a block SBR copolymer. It also nicely leads us into our next topic. in the free radical polymerization of chloroprene the predominant structural repeat unit is head-to-tail trans-1. could be described as an alternating copolymer of pentamethylene units and ester moieties. . Let us now mix things up a bit and consider the infrared spectrum of a random copolymer of ethylene and methacrylic acid. the homopolyester PCL. A similar argument can be made for TPC.13 The infrared spectrum of polychlaroprene polymerized at (A) -40"C and (8) -20"C. This frequency corresponds to the formation of an intermolecular hydrogen bonded dimer. The spectroscopic contribution from the methacrylic acid (MAA) portion is also very interesting. CH3 -(CH2)-CH2-C-n I rY hydrogen bonded dimer :: :: C I '0 I IJ 1105 I c Ethy lene-co-methacry lie acid 1305 I B A 3200 2600 2000 Wave number. Obviously. which absorb at a much higher frequency of 1750 cm+. there are many polymers synthesized from a single monomer that might better be characterized as copolymers.2. we have copolymers synthesized from relatively small monomers where one might expect significant vibrational coupling between the structural units in the polymer chain. is sensitive to order in much the same way as the nylons mentioned above. infrared spectroscopy can provide a useful measure of overall composition.4 structural units have been subtracted out of the spectrum. For example.2 and 3. Characteristic bands are readily identified and the reason the subtraction works so well is that the individual structural units act essentially independently. as in the case of styrene-co-butadiene (SBR). Copolymers and Structural Irregularities "True" copolymers are derived from two or more co-monomers.4 and head-to-head trans:-1. 1. as in ethylene-copropylene copolymers (EP). as we mentioned in chapter 1. ethylene-co-propylene (EP) or_~~~~~~methacrylic aClo(EMAAl copolymers. 1.4 umts present. However. a polymer containing the chemical repeat unit -(CH2kCOO-.4. together with features attributable to a strongly hydrogen bonded polymer. for polymers and copolymers such as this. Figure 6. This illustrates the strong driving force for association that is a property of such strongly interacting groups. The carbonyl stretching frequency of a random EMAA copolymer containing 4 mole % MAA is observed at 1700 cm-) [see Figure 6. but there ~e significant concentrations of cis-l.A). if any.4 units present in polychloroprene after the spectral contribution from the predominant trans-1. There is scant. cm-1 500 Figure 6.14(C)]. essentially uncoupled repeat units' present in the polymer chain. the sequence distribution will make little di~ference to the infrared spectroscopic features that are observed.

. Very Strong Intermolecular Interactions: Ionomers ~~rtunately. it is known that the nature. and (C) nonionized acid copolymer. as we have seen. The "rules of thumb" implied above also pertain to the spectroscopic changes seen upon the chemical reaction of polymers. (B) sodium ionomer. (AJ 0. and there are obvious differences We will not dwell on the interpretation of the spectral differences.J'" B 1900 B I 154B~ A 1200 3700 3150 2600 Wavenumber 2050 1500 Figure 6. you shou~d rec. (B) 15. For now. It is the case of miscible. but rather just state that we can infer the local structure of the ionic domains from. As we have . a consideration of the known coordinating tendencies of different cations together with a symmetry analysis of models of most probable structures. m~ny reactions result in the formation of carbonyl containing moeines and the VIbrational modes of carbonyl groups. Dramatic effects are observed in the infrared spectra upon ionization. the most important being the reduction or loss of the band at 1700 cm+. are some of the most intense and useful in infrared spectroscopy. More on this in Chapter 9. Thus. (-CHz-CHz-CHz-CHZ-O-)n.14 The infrared spectrum of ethylene-co-methacrylic acid containing 4 mole % MAA. the greater the confidence in assuming that the spectroscopic changes simply reflect isolated structural defects. attributed to hydrogen bonded acid dimers. one can assume that in infrared spectroscopic terms. the vast majority of studies involving chemical modification rely upon the simple group frequency approach. To simplify matters. At ~e e~tr~mes. and (CJ35 h. the spectrum of the blend would reflect the simple addition of the spectra of the two individual components. recorded at 150°C as a function of time. Degradation etc. This is in itself interesting and useful. one polymer does not recognize the existence of the other.15 shows infrared spectra of a film of pure polytetrahydrofuran. One example should suffice. The oxidation of the polymer may be readily followed by monitoring the carbonyl band at 1737 cm+. (A) Calcium ionomer. In other words. relative strength and number of intermolecular interactions occurring between the polymeric components of the blend playa key role in determining miscibility. respectively. Figure 6. Ionization of some or all of the acid groups present in the EMAA copolymers leads to a class of materials referred to as ionomers. Multicomponent Systems: Polymer Mixtures and Complexes Oxidation. For a grossly phase separated system. and vice versa. where the two polymers exist in essentially separate and distinct phases. Infrared spectroscopy is one of many techniques that have been applied to unravel the complexities of the interactions occurring in polymer mixtures. the larger the unit representing chemical modification and the smaller the effects of vibrational coupling.14 the spectra denoted A and B are the completely ionized calcium and sodium salts. or immiscible polymer blends where the ~eparate phases are both mixtures of the two polymers. but there are subtle effects occurring in the carboxylate stretching region that reveal information concerning the local structure of ionic domains so-called (multiplets). Accordingly. and the appearance of an intense band at about 1550 crrr! which is characteristic of carboxylate groups. however. that is much more mteresting.ognize that in multiphased systems the two phases do not necessarily consist of Just the pure components of the mixture. In figure 6.170 SPECTROSCOPY Characterization of Polymers by Infrared Spectroscopy 171 c -:.15 The infrared spectrum of polytetrahydrofuran in the carbonyl stretching region recorded at 1500C as afunction of time. Figure 6. pol~mer mixtures may be categorized as miscible (single phase) or immiscible (multiphased).

If we introduce a magnetic nucleus into a uniform external magnetic field. (C) 20:80. which is related to the mass number and the atomic number. These parameters can then be used to model phase behavior".. paying particular attention to those aspects that can be used to obtain information about polymer composition and microstructure. Rather.. we will emphasize the straightforward applications of the method to the chemical microstructure of polymer chains and copolymer compositional analysis. "magic angle spinning". The spinning nucleus gives rise to a magnetic field and the nucleus .!Ell_~be viewed as a small magnet of magnenc moment.1. Significantly. we will only be considering the !H. 0 1 2 3. which. two dimensional (2D) techniques. and (D) 0:100. But first we present a brief summary of the basic fundamentals of the technique.. you get the picture! even even . because it makes things simple. same lousy jokes. infrared spectroscopy is sensitive to strong intermolecular interactions especially hydrogen bonding. M. . and is loaded with jargon. Fortunately. or spin number.e. This is depicted in figure 6. suggests that the best is yet to come.17. nmr imaging etc. lH.e. fortuitously. Tfius. 1/2. if you must. can conjure fear into the hearts of students approaching this subject for the first time. Atomic Number even or odd even odd . O-H"""O=C/ -CH2-CH- CH3 '0 I CH2 I Phenolic hydroxyl. • See. F.. Phrases and acronyms like. the most commonisotopes-ofcarbofi. but it is possible to obtain a quantitative measure of the number and relative strength of the intermolecular interaction as a function of temperature.. Graf and P. a book entitled Specific Interactions and the Miscibility of Polymer Blends.16Q. quantized) set of (21 + 1) of orientations. it assume'slidiscrete (i. same publisher. C. and the like. (B) 50:50. An excellent example is shown in figure 6. as shown in table 6. I. however. The low-energy orientation corresponds to that Table 6.. Instruments have improved enormously and the emergence of solid state nrnr. During the past three decades there has been a veritable plethora of diverse (and very clever) experiments reported in the literature. which is described by the nuclear spin. A given nucleus has a specific spin number.-and oxygen. 3/2. and 19F nuclei (I = 1/2) will assume -only one of two possIble orientations that correspond to energy levels of ± ~Ho in an applied magnetic field (Ho is the strength of the external magnetic field). "CRAMPS".1 Spin Numbers of Isotopes. "cross polarization". 1=0.. by M. Coleman. as depicted below: What Is NMR? Nuclei of certain isotopes possess what can be thought of as a mechanical spin. all have spin numbers of 1/2. are nonmagnetic (I = 0) and do not exhibit nmr spectra. "dipolar decoupling". The band at 1739 crrr! is attributed to the PVAc carbonyl groups that are "free" (non-hydrogen bonded) while that at 1714 crrr! is representative of PVAc acetoxy carbonyl groups hydrogen bonded to the phenolic hydroxyl group of PVPh. "spin-spin relaxation".!2C. and !9F nuclei. For our purposes here. which hopefully will whet your appetite and send you scurrying to the li-brary for more. Nrnr spectroscopy occupies a special place in the arsenal of experimental methods that have been employed to study polymeric materials. \ seen. i. (A) 80:20.172 SPECTROSCOPY 1714 Basic NMR Spectroscopy 173 1739 I I I W! BASIC / NMR SPECTROSCOPY 3SOO 3000 1800 1650 Figure 6. Spin Number I 1/2 3/2 5/2 . This is a vast subject in its own right and we can only present what we consider to be representative and informative exam-ples.acetoxy carbonyl hydrogen bond Not only do the spectra indicate mixing at the molecular level. "COSY". Here we see the spectra in the carbonyl stretching region of poly(vinyl acetate) (PVAc) and miscible blends of PVAc and PVPh. etc. !3C. Same authors. in this introductory text we do not have to be concerned with such advanced applications of nmr spectroscopy. ~.. Painter. Mass Number odd CH I I -<Q)- . . I.16. 1.. l3C...16 Infrared spectra of poly(4-vinyl phenol)-poly(vinyl acetate) blends. or angular momentum.. The subject has become increasingly mathematical.

The extent of electronic shielding is directly proportional to the strength of the applied field. Pergamon Press.!II 174 SPECTROSCOPY Basic NMR Spectroscopy 175 Table 6. has been introduced. which is defined as 10 .B. such that E = hu = 21. Typically. state in which the nuclear magnetic moment is aligned parallel to the external magnetic field.18 shows a low resolution IH nrnr spectrum of ethanol. '" 600 ppm (see below). In order to express chemical shifts in a form independent of the applied field (or oscillator frequency). The corresponding range for !3C nuclei is much greater. however.18 Low resolution IH nmr spectrum of ethanol (reproduced with permission from L. and H. H ° J~ 1. are the field strengths corresponding to resonance for a particular nucleus in the sample (Hs) and a reference (Hr). A comparison of the nrnr frequencies of nuclei that are of major interest _for polymer spectroscopy tsglVen in ta6Ie (l. To be able to resolve these subtle effects requires an nrnr spectrometer that consists of a very strong magnet that produces a stable and homogeneous field.12) ) t The Chemical Shift If the resonance frequencies for all nuclei of the same type in a molecule were identical.5 10. = _.016 1. I:I! . chemical structure. The transition of a nucleus from one possible orientat ion to the other is a result of the absorption or emission of a discrete amount of e nergy. Figure 6. a radio-frequency oscillator (i. and -CH3 groups.Hs)/Hr.37 100 b 42. 1969). ppm). and the effective magnetic field felt by a nucleus is not identical to that of the applied field.1 . defined as B' = (H. tetramethylsilane (TMS)] is termed the chemical shift.0 00 gauss. the oscillator frequency). :2 . 1 I I I I I 2. Jackman and S.0010 0. . for lH nuclei a range of 10 ppm covers most organic molecules. -CH2-. where U is the frequency onhe electromagnet" lC radiation that is absorbed or such emitted.2. For equal numbers of nuclei at co nstant Ho. a chemical shift parameter. would resonate at the same frequency as those in an aromatic ring.2). Figure 6.7 3. II. B'. This is not so. .01 1/2 1/2 1/2 1/2 In 10 kG field. where H.0010 0. . the frequencyof 60 megacycles per second (60 energy is in the radio-frequency region-about MHz). depending upon where (uTMS .64 0.. There are three absorption peaks that have an area ratio of 1:2:3..17 The two orientations of a IH nucleus. Isotope Abundance (%) nmr Frequency (MHz) 11 Relative Sensitivit 1.1. and the high-energy orientation c orresponds to thitStirteil1Which the nuclear magnetic moment is aligned antipar allel (opposed) to this field. . a radio-frequency receiver (the detector) and a device that can vary the magnetic field or frequency over a relatively narrow range (analogous to the prism or grating in figure 6.11 99.-+~ I ----4 1..0159 0.6 6. For IH in a magnetic field of 14.01 4. =+- . respectively. and subtle differences in nrnr frequencies are observed. B (in units of parts per million.2 Characteristics of Certain Isotopes. This separation of resonance frequencies of nuclei in different structural environments from some arbitrarily chosen standard [e.0096 0. only one peak would be observed. .Ho. These are a consequence of the different molec ular environments of the nuclei. . . for example.Us ) • 6 B (6. .98 0. corresponding to the protons in -OH.000 0.us) is the difference in absorption frequencies of the sample and the reference in cps. Stern hell.e. Nuclear Magnetic Resonance Spectroscopy in Organic Chemistry. • f o j_-~ 2 o Field __ Figure 6. so that chemical shift values are proportional to field strength (or equivalently.2). is obtained: 10 (ppm) = spectrometer frequency (cps) ( UTMS .1. the source of radiation-see figure 6. M.g. Surrounding electrons shield the nuclei to di fferent extents. 't. If TMS is used as the internal standard.' .3 40. We should also mention another parameter commonly encountered in the 'H nrnr spectroscopic literature. the common equation for the chemical shift. an d protons of a methyl group. It is now appropriate to look at a very simple nrnr spectrum that illustrates the principles outlined above.1.834 b Spin Number IH 2H(D) 13C 14N 15N 19F a 99..

This spacing is called the spin-spin coupling constant. In simple cases of interacting nuclei.19 High resolution 1H nmr spectrum of ethanol (reproduced with permission from L.60 Figure 6. the relative intensities of a multiplet are symmetric about the mid-point and are approximately numerically proportional to the coefficients seen in Pascal's triangle (i. where the absorption coefficients for the bands are all different.00 2. 1969). All well and good. 1:2:1. You should also be aware that the spacing (in cps) of the three components of the methyl group triplet is equal to the spacing of the four components of the methylene group quartet and is independent of the strength of the applied field. and the methylene absorption is split into a quartet (relative areas > 1:3:3: 1). etc. which leads us to the fascinating subject of spin-spin interactions.70 2. . in lH nmr the intensities of the bands give a direct measure of the number of nuclei that they represent.20 Space filling model of (±)-2-isopropylma/ic acid.176 SPECTROSCOPY Basic NMR Spectroscopy 177 -OH Figure 6. This nicely illustrates first. The observed multiplicity of a given group of equivalent protons depends on the number of protons on adjacent atoms and is equal to n + 1. These splitting patterns are caused by the magnetic field of the protons of one group being influenced by the spin arrangements of the protons on the adjacent group. 1:3:3:1 etc. and -CH3 groups. The simple rules for determining the multiplicities for spin-spin interactions of adjacent groups hold only for cases in which the separation of resonance lines of interacting groups (the symbol ~'U) is much larger than the coupling constant J of the groups (~'U»J). the different chemical shifts for protons in the three different groups and second. Things are not quite so easy in 13C nrnr (see later). The methyl absorption is split into a triplet (relative areas r= 1:2:1).90 1H PPM 2. Figure 6. Stern hell. -CH2. but there is even more useful information to be gained if the nmr spectrum is recorded at a higher resolution. that a quantitative measure of the number of protons in each group is readily obtained". 1:1. a triplet. Pergamon Press. but the total relative area of each group is again > 1:2:3. for a doublet. where n is the number of protons on adjacent atoms. Thus the two CH2 protons in the ethyl group split the CH3 protons into a triplet.21 nmr spectrum of (±)-2-isopropylma/ic acid.e.. the two protons in the CH2 group) are not observed in simple systems like ethanol.80 2. then the peaks due to the methylene and methyl protons appear as multiplets. the three CH3 protons of the ethyl group split the CH2 protons into a quartet. and is denoted by the symbol 1.) • Note that unlike infrared spectroscopy. Nuclear Magnetic Resonance Spectroscopy in Organic Chemistry. Jackman and S. respectively. In systems of interact- 3. M. Interactions between magnetically equivalent protons in the same group (for example.19). a quartet. corresponding to the -OH. Spin-Spin Interactions H the nmr spectrum of ethyl alcohol is obtained at a higher resolution (figure 6.

20. there are cases (which are particularly relevant to our forthcoming discussions of polymer tacticity) where two magnetically non-equivalent protons on the same carbon can give rise to a maximum of four lines. the chemical shift difference is of the same order of magnitude as the coupling constants. Incidentally. if the two methylene protons were equivalent. but it is their dissimilarities that are more interesting. This molecule has two nonequivalent protons on a single carbon.2). nitriles) and (3) the direct observation of carbon reaction sites of interest. You will observe there are four lines there. carbonyls. As we have just observed.g. the probability of two adjacent 13Cnuclei is '" 0. more lines appear and simple patterns of spacings and intensities are no longer found. This system is designated an AB pattern. Of course. One cannot just simply relate the relative intensities to the number of equivalent carbons in the molecule. While this appears to be a daunting practical limitation. the early IH nrnr spectroscopists would have had serious problems in interpreting IH nrnr spectra. are shown directly below the methylene group itself. (3) the enhanced sensitivity.178 SPECTROSCOPY Basic NMR Spectroscopy 179 ing nuclei in which the coupling constant is of the. however. because of the complication of 13C-IH spin-spin coupling. which has a natural abundance of> 99. 13Cis only present to the extent of 1. but in fact has a number of competing advantages relative to 13C n~ including: (1) the ease of quantitative analysis. if 13Cdid exist in greater natural abundance. 13C NMR There are some features of IH and 13Cnrnr spectroscopy that are common. in that there are two protons involved in a slightly different chemical shift. Thus it cannot be observed in nmr experiments. It is situations similar to these we shall look for in tacticity measurements on polymers. (4) the direct observation of OH and NH groups (undetectable by 13C nmr) and (5) the separation of olefinic and aromatic protons. 13C nrnr spectroscopy has the following additional advantages over IH nmr for the analysis of organic polymers. In this case. One major consequence of this low natural abundance is that 13C_13C spin-spin coupling interactions are unlikely in unenriched compounds (assuming random placements of the 13C nuclei. .22 which shows the proton decoupled 13C nmr spectrum of 2-ethoxyethanol obtained in our laboratory. figure 6. 13C resonances of organic compounds are found over an enormous chemical shift range of 600 ppm.1 % (table 6. The two central lines are more intense than the satellite lines. but unlike IH. instrument manufacturers have effectively solved the problem and obtaining high quality 13C nrnr spectra of organic materials is now routine (see.same order of magnitude as the separation of resonance lines (LlU== J). Consider. Thus the relative sensitivity of a 13C nrnr experiment is some 6000 times less than that of a IH nrnr experiment. the two CH2 group protons in ethanol) are not ordinarily observed.5 to 3. 13Chas the same nuclear spin (I = 1/2) as IH. On the other hand.9%. and Hb. has no nuclear spin (I = 0). spin-spin interactions between protons within a group of magnetically equivalent protons (for example. CH3CH2-O-CH2CH20H 4 3 2 1 2 3 4 70 60 50 PPM 40 30 20 Figure 6. (2) the direct observation of carbon-containing functional groups that have no attached protons (e. To compound the problem further. the simple multiplicity rules no longer hold. but suffice it to say that quantitative analysis is much more difficult than in IH nrnr. however. as in the spectrum of 2ethoxyethanol which shows well resolved nmr lines attributable to each individual carbon. Frequently. which appear in different regions of the IH nrnr spectrum while olefinic and aromatic carbons overlap one another in the 13Cnrnr spectrum. that the lines are not of equal intensity. is the enhanced effective resolution possible in 13C nrnr. IH nrnr spectroscopy has not passed into oblivion. even though they are each assigned to a single carbon. (2) the rapidity of analysis time. 13C nuclei only produce 1/64 of the signal that IH nuclei yield on excitation. (±)-2-isopropylmalic acid illustrated as a space filling model in figure 6. The lines arising from the two protons.. one can identify individual resonances for each carbon in a compound. However. Note. for example. Additional information pertaining to relaxation mechanisms and the so-called nuclear Overhauser effect is required. (1) the direct observation of molecular backbones. for example.5 ppm. It is not possible to discuss here the many subtleties of 13Cnrnr.22 Be nmr spectrum of 2-ethoxyethanol. The interested reader is refered to the texts on the subject mentioned at the end of this chapter. H. However. This ~s s~~e SO times greater than that for IH nuclei.) Perhaps the most striking difference between 13C and IH nrnr spectroscopy. The effect of this coupling in 13C nrnr spectra may now be readily eliminated by a technique called proton decoupling. a single line would be seen. The inherently low natural abundance of 13Cobviously lowers the effective sensitivity of 13C vis-a-vis IH nmr experiments. 12C. In figure 6.21 a scale expanded IH nrnr spectrum is displayed in the region from 2. The most abundant isotope of carbon.0001).

We recently synthesized a series of methyl methacrylate-co-hexyl methacrylate copolymers (MMA-coHMA) for the purpose of testing a model that we have developed to predict the phase behavior of polymer blends. Thus. M. as depicted above.' and hexyl Figure 6. respectively.180 SPECTROSCOPY Characterization of Polymers by NMR Spectroscopy 181 F.3457 (1993). The rather complicated collection of nmr lines occurring . nmr spectroscopy is one of the major instrumental methods used routinely to identify and analyze the composition of copolymers. nmr spectrum of a copolymer of methyl methacrylate Copolymer Analysis As we mentioned above. nmr spectroscopy is used to probe the dynamics of polymer systems through the study of relaxation phenomena. Macio and P.9 ppm. structural isomerism and copolymer composition. S. taken from our own recent studies.. In Chapter 1.9 ppm /2 x 100 (6. respectively.6 ppm /3+A 3. including those involving stereoisomerism.. C. nmr spectroscopy has been extensively employed to characterize the microstructure of polymeric materials. quantitative compositional analysis becomes remarkably straightforward.5 ppm are assigned to the alkyl methylene and methyl protons of the copolymer. CHARACTERIZATION SPECTROSCOPY OF POLYMERS BY NMR In common with infrared spectroscopy.' . IH nmr was employed to determine the copolymer composition". Our approach here will be to start by presenting a few representative examples of the simple use of nmr spectroscopy in polymer analysis. but this latter topic must be left for more specialized texts. spending some time on the classic IH nmr spectroscopic studies of the tacticity of poly(methyl methacrylate). nmr spectroscopy may be applied to the characterization of polymeric materials at various levels of complexity. sequence isomerism. Finally. We will then tum our attention to the meat of this section and describe the application of nmr spectroscopy to the characterization of polymer sequence distributions. can been used to describe the sequence distribution in terms of the different tactic placements in the polymer chain. I. we will close by describing a few examples of analogous studies pertaining to sequence and structural isomerism. at the highest level of sophistication. Xu. Coleman. i. will be used to illustrate the utility of the nmr technique. They serve as routine analytical techniques for the screening.methylene group.13) . Here nmr excels and the technique has dominated the field of polymer sequence distribution in all its myriad forms. Three representative examples. R.5 and 2. 26. and again at the end of Chapter 5. Modem chemists involved in polymer synthesis would be lost without ready access to nmr spectrometers. Painter. for the purposes of measuring copolymer composition.6 and 3. M.e. Figure 6... we can ignore these nmr lines and focus our attention solely on the two isolated nmr lines occurring at'" 3. (co)polymers and reaction -produc~ At the next level of sophistication. Y. We will develop this theme. two employing IH and one using i3C nmr spectroscopy. Macromolecules. 1H '. if we divide the relative areas of these two nmr lines by 3 and 2.23 shows a typical spectrum of a MMA-colIMA copolymer.: A3•6ppm/3 %MMA =A 3. . Finally.. Fortunately. between 0. These are assigned to the three protons of the -OCR3 methoxy substituent of MMA and the two alkoxy protons of the -OCRz.23 methacrylate. Methyl Methacrylate-co-Hexyl Methacrylate Copolymers Let us start with a very straightforward example. we mentioned that isomerism in homopolymers can be viewed as a special case of copolymerization. and show how probability theory. identification and analysis of monomers.

95 ppm line 9 (6. Painter and M. also for our ongoing research on polymer blends. the nmr line at 0.16 ppm corresponds to the 6 protons present in the two methyl substituents of the t-BSOS. • Y. C.e.2 and 7. protected monomers are commonly employed. But there no isolated nmr line in the spectrum that is solely characteristic of styrene repeat unit. subtract out the contribution from the t-BSOS repeat (i. as summarized below".182 SPECTROSCOPY Characterization of Polymers by NMR Spectroscopy 183 Styrene-co-Vinyl Phenol Copolymers Another set of copolymers that we have synthesized.2 ppm in spectrum A correspond to the aromatic protons that occur in both St and t-BSOS. 32. Polymer. F.95 ppm corresponds to the 9 protons present in the three methyl substituents of the t-butyl group in tBSOS. ) The copolymer composition of the various styrene-co-t-BSOS copolymers synthesized was also determined from IH nmr spectra similar to that shown in figure 6. because we can obtain the needed information in an indirect manner. We should not despair. J.16 ppm in spectrum (A) are assigned to the methyl and t-butyl substituents in the tbutyldimethylsilyl monomer.14) IH nmr spectroscopy was used to identify the copolymers and follow the deprotection step. Xu. Because side reactions involving the phenolic hydroxyl group occur during the direct polymerization of 4-vinyl phenol (VPh). The nmr lines at 8 = 0.95 and 0. The relatively broad lines appearing between 6. Graf. Figure 6. M. and their absence in the nmr spectrum of the deprotected copolymer (B) clearly indicates the elimination of the t-butyldimethylsilyl group. and you just have to keep track of the number and type of protons that contribute to a given nmr line or group of lines.24(A).. It is a little more complicated than the methyl methacrylate-cohexyl methacrylate copolymers that we just discussed.3103 (1991). P. A B styrene t-BSOS ppm Figure 6. however.24 1H nmr spectrum of (A) a styrene-co-t-BSOS copolymer and (B) the corresponding styrene-co-vinyl phenol copolymer after deprotection. So if we measure this total area (which reflects the contributions from the 5 aromatic protons of the styrene repeat and the 4 aromatic protons of the t-BSOS repeat). 4 times the normalized area per proton . Similarly. the nmr line at 0.16 ppm line 6 Total area of 0. As we inferred above. Thus the normalized area per proton corresponding to the t-BSOS chemical repeat may be determined from either: styrene-co-vinyl phenol Total area of 0. In our studies we used VPh protected with the r-butyldimethylsilyl group (t-BSOS) to prepare copolymers of styrene and t-BSOS. STVPh (B). but the principle is the same. Coleman. Now we need the normalized area per proton corresponding to the styrene chemical repeat. These were subsequently hydrolyzed to produce styrene-co-vinyl phenol (STVPh) copolymers.24 shows a comparison of the IH nmr spectra of a parent styrene-co-t-BSOS copolymer (denoted A) and the product after the desilylation process. are those containing styrene and 4-vinyl phenol.

Master your semester with Scribd & The New York Times

Special offer for students: Only $4.99/month.

Master your semester with Scribd & The New York Times

Cancel anytime.