Modern Styrenic Polymers: Polystyrenes and Styrenic Copolymers
Wiley Series in Polymer Science Series Editor: Dr John Scheirs Excel Plas PO Box 2080 Edithvale VIC 3196 AUSTRALIA scheirs.john @ pacific.net.au Modern Fluoropolymers High Performance Polymers for Diverse Applications Polymer Recycling Science, Technology and Applications Metallocene-based Polyolefins Preparations, Properties and Technology Polymer-Clay Nanocomposites Dendrimers and Other Dendritic Polymers Forthcoming titles: Modern Polyesters Environmentally Degradable Polymers
Modern Styrenic Polymers: Polystyrenes and Styrenic Copolymers
Edited by
JOHN SCHEIRS ExcelPlas Australia, Edithvale, VIC, Australia and
DUANE B. PRIDDY 6004 Camelot Ct, Midland, Ml, USA
WILEY SERIES IN POLYMER SCIENCE
John Wiley & Sons, Ltd
Copyright © 2003 John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ, England Telephone (+44) 1243 779777 Email (for orders and customer service enquiries):
[email protected] Visit our Home Page on www.wileyeurope.com or www.wiley.com All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except under the terms of the Copyright, Designs and Patents Act 1988 or under the terms of a licence issued by the Copyright Licensing Agency Ltd, 90 Tottenham Court Road, London WIT 4LP, UK, without the permission in writing of the Publisher. Requests to the Publisher should be addressed to the Permissions Department, John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ, England, or emailed to
[email protected], or faxed to (+44) 1243 770620. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the Publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Other Wiley Editorial Offices John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA Jossey-Bass, 989 Market Street, San Francisco, CA 94103–1741, USA Wiley-VCH Verlag GmbH, Boschstr. 12, D-69469 Weinheim, Germany John Wiley & Sons Australia Ltd, 33 Park Road, Milton, Queensland 4064, Australia John Wiley & Sons (Asia) Pte Ltd, 2 Clementi Loop #02-01, Jin Xing Distripark, Singapore 129809 John Wiley & Sons Canada Ltd, 22 Worcester Road, Etobicoke, Ontario, Canada M9W 1L1
British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN 0 471 49752 5 Typeset in 10/12pt Times by Kolam Information Services Pvt Ltd, Pondicherry, India. Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham, Wiltshire. This book is printed on acid-free paper responsibly manufactured from sustainable forestry in which at least two trees are planted for each one used for paper production.
Contents Contributors Series Preface Preface About the Editors I
xxi xxvii xxix xxxiii
INTRODUCTION TO STYRENIC POLYMERS 1
Historical Overview of Styrenic Polymers John Scheirs
1 2 3 4 5 6 7 8 9 10 11
Introduction General-purpose Polystyrene Foamed Polystyrene Rubber-modified Polystyrene ABS ASA Early Styrene Copolymers Styrenic Block Copolymers Syndiotactic Polystyrene Modern Polystyrene Production The Future References
2 Polystyrene and Styrene Copolymers - An Overview Norbert Niessner and Hermann Gausepohl
1 2 3 4
Introduction Polymerization Processes Structure and Morphologies
3
3 4 13 18 18 20 21 21 22 22 22 23 25
25 27 29 29
vi
CONTENTS
5 Properties 6 Properties, Range and Applications of MABS Products References H
34 38 41
PREPARATION OF STYRENIC POLYMERS
43
3 Commercial Processes for the Manufacture of Polystyrene Bernard J. Meister and Clark J. Cummings
45
1 Introduction 2 Technical Constraints that Influence Reactor Selection 2.1 Temperature Control 2.2 Chemistry-related Constraints 2.3 Constraints Due to Reactor Mixing 2.4 Constraints Related to the Rubber Modification of Polystyrene 2.5 Reactor Requirements for Producing Copolymers 3 Polystyrene Devolatilization 3.1 Devolatilization Concepts 3.2 Devolatilization Equipment 3.3 Steam Stripping 4 Current Polystyrene Polymerization Processes 5 Process Simulation and Control References 4 Approaches to Low Residual Polystyrene Duane B. Priddy 1 Introduction 2 Summary of R&D Approaches 2.1 Devolatilizer Design 2.2 Assisted Devolatilization 2.3 Scavengers 2.4 Absorbers 2.5 High Monomer Conversion Polymerization 2.6 Solid Polymer Treatment 3 Friedel-Crafts Catalyst 3.1 Addition of Friedel-Crafts Catalyst to Monomer 4 Latent Acid Catalysts 4.1 Tosylates 5 Monomer Regeneration Upon Heating 6 Epilog References
45 46 46 48 52 54 57 59 59 60 65 66 69 71 73 73 75 75 76 78 80 80 82 83 85 85 86 88 91 91
CONTENTS
5 Process Modelling and Optimization of Styrene Polymerization J. Gao, K. D. Hungenberg and A. Penlidis 1 Introduction 1.1 General Kinetic Scheme of Styrene Homopolymerization 1.2 Treatment of Gel Effect 2 Process Simulation and Optimization of Styrene Homopolymerization 2.1 Using Initiator Combinations with Designed Temperature Profile 2.2 Using Bifunctional Initiators 2.3 Using Reactor Combinations 3 Conclusion 4 Symbols References 6 Living Free Radical Polymerization of Styrene Alessandro Butte, Giuseppe Storti and Massimo Morbidelli 1 Introduction 2 LRP Overview 2.1 Nitroxide-mediated Polymerization (NMP) 2.2 LRP by ATRP 2.3 LRP by Degenerative Transfer 2.4 LRP by RAFT 3 Kinetics of LRP 3.1 Main Features of the Different LRP Processes 3.2 Homogeneous vs Heterogeneous LRP Processes 4 Applications to Styrenic Polymers References
vii
93 93 94 98 100 100 101 105 107 107 108 111 111 112 115 116 118 118 120 120 122 125 127
7 Increasing Production Rates of High MW Polystyrene Bryan Matthews and Duane B. Priddy
129
\ Introduction 2 Speeding Up the Rate of Polystyrene Production Using Chemical Initiators 3 Speeding Up the Rate of Polystyrene Production Using Acid Mediation 4 Use of Acid to Tailor the Molecular Weight Distribution 5 Modeling Acid-mediated Styrene Polymerization
129 130 133 139 140
viii
CONTENTS
5.1 Styrene Auto-initiation Model 5.2 Acid Model Development 5.3 Model Results 6 Conclusions References
141 141 143 145 146
8 Preparation of Styrene Block Copolymers Using Nitroxide Mediated Polymerization Duane B. Priddy
147
1 Introduction 2 Mechanism and Limitations 3 How Living is NMRP? The Results of Model Studies 4 Block Copolymers via the Macroinitiator Approach 5 Preparation of Block Copolymers Using Alkoxyamines as Chain-stoppers in Step-growth Polymerization 6 Preparation of Block Copolymers via Sequential Addition of Monomers (SAM) 7 Preparation of Block Copolymers Using Multiple-headed Initiators References III MAJOR CLASSES OF STYRENIC POLYMERS
147 148 149 152 155 156 159 162 163
9 Particle Foam Based on Expandable Polystyrene (EPS) Rolf-Dieter Klodt and Brad Gougeon
165
1 Introduction 2 EPS Based on Suspension Polymerization 2.1 Production of EPS Raw Material 2.2 From Raw Material to Foam 2.3 Physical and Mechanical Properties 2.4 Applications 3 EPS Based on Extrusion Process 3.1 Extrusion 3.2 Post Extrusion 3.3 Steam Expansion of EPS Loose-fill Resin: Theory and Practice References
165 166 166 182 185 188 190 191 192 194 197
CONTENTS
ix
10 Rigid Polystyrene Foams and Alternative Blowing Agents Kyung Won Suh and Andrew N. Paquet
203
1 Introduction and General Description 2 Nomenclature 3 Theory of the Expansion Process 3.1 Bubble Initiation 3.2 Bubble Growth 3.3 Bubble Stabilization 4 Properties and Their Relation to Structure 4.1 Test Methods 4.2 Properties of Commercial Products 4.3 Cells 4.4 Gas Composition 4.5 Rigid Cellular Polymers 4.6 Creep 4.7 Structural Foams 5 Thermal Properties 5.1 Thermal Conductivity 5.2 Coefficient of Linear Thermal Expansion 5.3 Maximum Service Temperature 5.4 Moisture Resistance 5.5 Environmental Aging 5.6 Other Properties 6 Commercial Production and Processing 6.1 Manufacturing Process 6.2 Decompression Expansion Processes, Physical Stabilization 7 Applications 7.1 Thermal Insulation 7.2 Refrigeration 7.3 Construction 7.4 Structural Components 7.5 Marine Applications 7.6 Other Uses 7.7 Energy Considerations in Foam Insulation 8 Environmental, Health and Safety Considerations 8.1 Flammability 8.2 Blowing Agents and Environmental Issues References
203 204 205 205 206 207 207 207 207 209 210 210 211 212 213 213 216 216 216 217 217 218 218 218 221 221 223 223 223 224 224 224 225 225 226 228
CONTENTS
11 Polystyrene Packaging Applications: Foam Sheet and Oriented Sheet Gary C. Welsh 1 Introduction 2 Oriented Polystyrene Sheet 3 Extruded Polystyrene Foam Sheet References 12 Preparation, Properties and Applications of High-impact Polystyrene M. F. Martin, J. P. Viola and J. R. Wuensch 1 Introduction 2 Properties 2.1 General Properties 2.2 Mechanical Properties 2.3 Impact Properties 2.4 Thermal Properties 2.5 Electrical Properties 2.6 Rheological Properties 2.7 Resistance to Solvents 3 Basic Chemistry 3.1 Matrix Molecular Weight 3.2 Elastomer Considerations 3.3 Environmental Stress Crack Resistance (ESCR) 3.4 Thermal and Oxidative Stability 4 Manufacture 4.1 Process Evolution 4.2 Modern Commercial Process 5 Fabrication 5.1 Fabrication Process and Part Properties 6 Application 7 Acknowledgements References 13 Key Structural Features Impacting SAN Copolymer Performance R. P. Dion and R. L. Sammler 1 Introduction 2 Characterization 2.1 Chromophores 2.2 Sequence Distributions
233 233 233 239 245 247 247 248 248 248 250 252 252 253 253 256 256 256 261 264 266 266 268 271 273 275 279 279 281 281 283 283 284
CONTENTS
3
4 5 6
xi
2.3 AN Levels 2.4 MWD 2.5 Composition Distribution 2.6 Multidimensional Analysis Fabrication Performance 3.1 Shear Flow 3.2 Entangled Chains 3.3 Time-Temperature Superposition 3.4 Cross Model 3.5 Nonlinear Shear Flows 3.6 Relaxation Spectra 3.7 Extensional Flow 3.8 Break Points 3.9 Brittle Breaks 3.10 Flow Birefringence Multiphase Systems Solid-phase Behavior Conclusion References
14 Rubber Particle Formation in Mass ABS Gilbert Bouquet 1
2 3 4 5
6
7 8 9
Manufacture of ABS 1.1 Emulsion Process 1.2 Mass Process Phase Separation Phase Inversion Phase Diagram Rubber Particle Sizing 5.1 Shear 5.2 Viscosity 5.3 Interfacial Tension Grafting 6.1 Graft Analysis 6.2 Effect of Process Parameters 6.3 Master Curve 6.4 Graft Model Crosslinking Sizing Window Rubber Particle Morphology References
285 285 285 286 287 287 287 288 289 289 290 291 293 293 293 294 296 297 298 305 305 305 306 306 307 307 308 308 308 310 311 311 311 313 313 314 316 317 318
xii
CONTENTS
15 High Heat Resistant ABS Technology Rony Vanspeybroeck, Robert P. Dion and Joseph M. Ceraso 1 2 3 4 5 6
Introduction Substituted Styrenes Imides Maleic Anhydride Modified Nitriles Various High Heat-resistant ABS Grades References
321 321 324 326 330 333 334 338
16 Synthesis, Properties and Applications of Acrylonitrile-Styrene-Acrylate Polymers 341 G. E. McKee, A. Kistenmacher, H. Goerrissen and M. Breulmann 1 Introduction 2 ASA Market 3 Production of ASA 3.1 Early Developments 3.2 Emulsion Polymerization Process 3.3 Bulk Polymerization Process 3.4 Microsuspension Polymerization Process 4 Properties of ASA 4.1 Ageing Properties 4.2 Impact Behaviour 5 Additional Areas of Investigation 6 ASA Blends 7 Applications of ASA 7.1 General 7.2 Solar Technology 7.3 Safety in the House and in the Office 7.4 ASA for Automotive Body Panels with PFM Technology 8 Future Perspectives References IV SYNDIOTACTIC POLYSTYRENE 17 Synthesis of Syndiotactic Polystyrene Norio Tomotsu, Michael Malanga and Juergen Schellenberg 1 Introduction 2 Catalytic Systems for SPS 2.1 Transition Metal Complexes
341 342 343 343 343 345 347 348 348 351 352 352 355 355 356 357 357 359 359 363 365 365 366 366
CONTENTS
xiii
2.2 Co-catalysts 3 Copolymerization 3.1 Polymerization of Substituted Styrenes 3.2 Copolymerization of Styrene and Ethylene 3.3 Copolymerization of Styrene and Dienes 4 Mechanisms of Polymerization of Styrene 4.1 Active Site Species 4.2 Kinetic Analysis of Styrene Polymerization 4.3 Effects of Hydrogen 5 Conclusion References
370 375 375 377 377 378 378 382 385 386 386
18 Characterization, Properties and Applications of Syndiotactic Polystyrene Komei Yamasaki, Norio Tomotsu and Michael Malanga
389
1 Introduction 2 Characterization 2.1 Structure 2.2 Crystal Form 3 Physical Properties 3.1 Thermal Properties 3.2 Crystallization Behavior of SPS 3.3 Comparison of Crystallization Properties of SPS with IPS 3.4 Solvent Resistance 3.5 Rheological Properties 3.6 Mechanical Properties of Neat SPS 4 Properties of Commercialized SPS and Its Applications 4.1 Mechanical and Flow Properties 4.2 Electrical Properties 4.3 Chemical Resistance 4.4 Improvement of Polystyrene by Blending SPS 5 Summary References
389 390 390 390 392 392 393 395 396 397 399 401 402 402 404 405 408 408
19 Rubber Modification of Syndiotactic Polystyrene G. E. McKee, F. Ramsteiner and W. Heckmann
411
1 Introduction 2 Energy Dissipation in Polystyrene Polymers 3 Impact Behaviour of Rubber-modified sPS 4 Rubber Modification
411 412 415 417
xiv
CONTENTS
4.1 Styrene Block Copolymers as Impact Modifiers 4.2 Core-Shell Impact Modifiers 4.3 Preparation of sPS in the Presence of a Rubber 5 Present Situation and Future Perspectives References 20 Polymeric Blends Based on Syndiotactic Polystyrene L. Abis, R. Braglia, G. Giannotta and R. Po 1 2 3 4
Introduction Overview of sPS Properties Patent Literature on sPS Blends Microscopic, Thermal and Mechanical Properties of sPS Blends 4.1 Miscible Blends 4.2 Immiscible Blends 5 Conclusions 6 List of Abbreviations References V
418 423 428 428 429 431 431 431 433 438 439 447 458 459 460
STYRENIC BLOCK COPOLYMERS
463
21 Styrenic Block Copolymer Elastomers R. C. Bering, W. H. Korcz and D. L. Handlin, Jr
465
1 2 3 4
Introduction Synthesis of Styrenic Block Copolymer Elastomers Properties of Styrenic Block Copolymer Elastomers Applications of Styrenic Block Copolymer Elastomers 4.1 Commercial Styrenic Block Copolymers 4.2 Adhesives and Sealants 4.3 Bitumen Modification 4.4 Footwear 4.5 Polymer Modification 4.6 Viscosity Index Improvers and Other Applications References
22 Preparation, Properties and Applications of High Styrene Content Styrene-Butadiene Copolymers David L. Hartsock and Nathan E. Stacy 1 History 2 SBC Synthesis and Manufacture 3 Key Features, Properties and Grades
465 465 474 487 487 489 492 493 493 496 497 501 501 502 504
CONTENTS
xv
4 Current Commercial Applications 4.1 Major Markets 4.2 Single Service 4.3 Rigid Packaging 4.4 Garment Hangers 4.5 Flexible Packaging 4.6 Medical 4.7 Consumer Goods 4.8 Toys 4.9 Displays 5 SBC Blends 5.1 Clear Blends 5.2 Opaque Blends 5.3 Others 6 Future Applications References
507 507 508 508 511 514 515 518 519 519 520 520 525 528 529 529
VI NOVEL POLYSTYRENES
531
23 Hydrogenated Polystyrene: Preparation and Properties Stephen F. Hahn 1 Introduction 2 Synthesis of Polycyclohexylethylene (PCHE) 3 Catalytic Hydrogenation 3.1 Catalysis and Conditions 3.2 Hydrogenation Mechanism 4 Polymerization of Vinylcyclohexane to PCHE 5 Characterization of PCHE 5.1 Atactic PCHE 5.2 Isotactic PCHE 5.3 Syndiotactic PCHE 6 Copolymers Containing PCHE 6.1 Random Copolymers 6.2 Block and Graft Copolymers 6.3 Graft Copolymers 7 Proposed Applications of PCHE-based Materials 8 Acknowledgment References 24 Branched Polystyrene Kurt A. Koppi and Duane B. Priddy 1
Introduction
533
,
533 533 534 534 536 539 539 539 545 546 547 547 547 551 551 553 553 557 557
xvi
CONTENTS
2 Preparation of Branched Polystyrene 2.1 Radical Polymerization 2.2 Anionic Polymerization 3 Rheology of Branched Polystyrenes 3.1 Star Branched Polymers 3.2 Comb Branched Polymers 3.3 Randomly Branched Polymers 3.4 Extensional Rheology 4 Conclusions References 25 'Super Polystyrene' - Sryrene-Diphenylethylene Copolymers G. E. McKee, F. Ramsteiner, W. Heckmann and H. Gausepohl 1 Introduction 2 Preparation of DPE Monomers and Polymers 2.1 1,1 -Diphenylethylene Monomer Synthesis 2.2 S/DPE Polymer Synthesis 3 Properties of Styrene-Diphenylethylene Polymers 4 Blends of S/DPE Polymers 5 Rubber Modification of S/DPE Polymers 5.1 Modified High-impact Polystyrene (HIPS) Process 5.2 Core-Shell Impact Modifiers 5.3 Tri-block Copolymer of StyreneHydrogenated Butadiene-Styrene [S-B(H)-S] 5.4 Tri-block Copolymers of S/DPEHydrogenated Butadiene-S/DPE 6 Thermoplastic Elastomers 7 Summary References 26 Ethylene-Styrene Copolymers Y. W. Cheung and M. J. Guest 1 Introduction to Ethylene-Styrene Copolymers 2 Copolymerizations of Ethylene and Vinyl Aromatic Monomers 3 Structure-Property Relationships for Ethylene-Styrene Interpolymers 3.1 Thermal Transitions/Viscoelastic Behavior 3.2 Mechanical Properties 3.3 Melt Rheology and Processability 4 Materials Engineering Aspects
557 557 564 565 566 569 571 573 577 577 581 581 582 582 582 583 585 586 587 588 596 599 600 601 603 605 605 606 608 609 613 614 616
CONTENTS
4.1 Interpolymer Blends 4.2 Blends of Ethylene-Styrene Interpolymers: Miscibility Considerations 4.3 Filler Composites 4.4 Terpolymers 5 Attributes and Applications 6 Summary 7 Acknowledgments References VII PROPERTIES OF STYRENIC POLYMERS 27 Fracture Behaviour of High-impact Polystyrene and Acrylonitrile-Butadiene-Styrene T. Vu-Khanh 1 Introduction 2 Quantitative Characterization of Fracture 2.1 Brittle Fracture 2.2 Semi-ductile Fracture 2.3 Ductile Fracture 3 High-impact Polystyrene 3.1 Effect of Temperature 3.2 Effects of Loading Rate 3.3 Dynamic Effect and Adiabatic Heating 4 Acrylonitrile-Butadiene-Styrene 5 Conclusion References 28 Dynamic Mechanical Behaviour of Atactic Polystyrene, High-impact Polystyrene and Other Styrenic Polymers S. N. Goyanes and G. H. Rubiolo 1 Introduction 2 Polystyrene 2.1 Effect of Polymer Structure and Additives on the Dynamic Mechanical Spectroscopy of Polystyrene 3 Copolymers of Styrene 4 Rubber-modified Polystyrene (HIPS) and SAN Copolymers (ABS) References
xvii
617 617 620 623 625 626 627 627 631 633 633 635 635 637 639 643 645 648 653 654 661 662
665 665 666 667 676 678 681
xviii
CONTENTS
29 Flame-retardant Polystyrene: Theory and Practice Bruce King 1 Introduction 2 Applications of Flame-retardant Styrenic Polymers 3 Flammability Requirements and Tests 3.1 Regulatory Test Methods 3.2 Research Methods 4 Mechanisms of Flame Retardation 4.1 Vapor-phase Mechanisms 4.2 Condensed-phase Mechanisms 5 Halogen-based Flame Retardants for Styrenics 6 Styrenic Blends 7 Environmental Concerns 8 Summary References 30 Photochemical Degradation of Styrenic Polymers B. Mailhot, A. Rivaton and J. L. Gardette 1 Introduction 2 Photooxidation of the Homopolymer Polystyrene (PS) Under Irradiation at A > 300 nm 2.1 Experimental Results 2.2 Discussion 3 Photooxidation of Poly(styrene-coacrylonitrile) (SAN) 3.1 Experimental Results 3.2 Discussion 4 Photooxidation of Acrylonitrile—ButadieneStyrene (ABS) 4.1 Analysis of the Photooxidation 4.2 Photooxidation Rate 4.3 Discussion 5 Photooxidation of a Blend of SAN and EPDM (AES) 5.1 FTIR Analysis of AES Films During the First Stages of Photooxidation 5.2 FTIR Analysis of AES Films for Longer Irradiation Periods 5.3 Discussion 6 Photooxidation of Blends of Polystyrene and Poly(vinyl methyl ether) (PVME-PS) 6.1 Introduction
685 685 686 687 687 689 690 690 692 692 699 699 700 701 703 703 704 704 707 709 709 710 712 712 713 715 716 717 718 718 720 720
CONTENTS
xix
6.2 Experimental Results 6.3 Surface Analysis 6.4 Discussion 7 Conclusion References 31 Analysis and Levels of Styrene Dimers and Trimers in Polystyrene Food Containers Hiromi Sakamoto 1 2 3 4 5 6 Index
Introduction Structure and Analysis of Styrene Dimers and Trimers Content of SDs and STs in PS Food Containers Migration of SDs and STs from PS Food Containers Biological Evaluation of SDs and STs Conclusion References
720 721 722 723 725 727 727 728 730 731 737 742 743 745
This page intentionally left blank
Contributors L. Abis Polimeri Europa Centre Ricerche Novara 'Istituto G. Donegani' Via G. Fauser 4 1-28100 Novara Italy
R. Braglia Polimeri Europa Centre Recerche Novara 'Istituto G. Donegani' Via G. Fauser 4 1-28100 Novara Italy
R. C. Bening Kraton Polymers LLC Westhollow Technology Center 3333 Highway 6 South Houston, TX 77082 USA
A. Butte Federal Institute of Technology Zurich Laboratorium fur Technische Chemie ETH Honggerberg, HCI F125 CH-8093 Zurich Switzerland
M. Beulmann BASF AG D-67056 Ludwigshafen Germany
J. M. Ceraso 438 Bldg The Dow Chemical Company Midland, MI 48667, USA
G. Bouquet The Dow Chemical Company 200 Larkin Building 1605 Joseph Drive Midland, MI 48674 USA
Y. W. Cheung INSITE, Technology R&D The Dow Chemical Company Dow Texas Operations Freeport, TX 77541-3257 USA
XXII
CONTRIBUTORS
C. J. Cummings Polystyrene Research and Development 438 Bldg, Dow Chemical Company Midland, MI 48667, USA
H. Goerrissen BASF AG D-67056 Ludwigshafen Germany
R. P. Dion Materials Sciences The Dow Chemical Company 1707 Building Midland, MI 48674 USA
B. Gougeon The Dow Chemical Company 200 Larkin Building 1605 Joseph Drive Midland, MI 48674 USA
J. Gao Polymer Technology and Process Development BASF AG D-67056 Ludwigshafen Germany
S. N. Goyanes Departmento de Fisica Facultad de Ciencias Exactas y Naturales Universidad Nacional de Buenos Aires Ciudad Universitaria, Pabellon I 1428 Buenos Aires, Argentina
J. L. Gardette Laboratoire de Photochimie Moleculaire et Macromoleculaire UMR CNRS 6505 Universite Blaise Pascal (Clermont-Ferrand) F-63177 Aubiere Cedex, France
M. J. Guest Polythylene and INSITE Technology R&D The Dow Chemical Company 1707 Building Midland, MI 48674 USA
H. Gausepohl BASF AG D-67056 Ludwigshafen Germany
S. F. Hahn Polymer Chemistry Discipline, Corporate Research and Development The Dow Chemical Company 1707 Building, Midland, MI 48667 USA
G. Giannotta Polimeri Europa Centro Ricerche Novara 'Istituto G. Donegani' Via G. Fauser 4 1-28100 Novara, Italy
D. L. Handlin, Jr Kraton Polymers LLC Westhollow Technology Center 3333 Highway 6 South Houston, TX 77082 USA
CONTRIBUTORS
XXIII
D. L. Hartsock Chevron Philips Chemical Co. 201A ARE Bartlesville, OK 74004 USA
W. H. Korcz Kraton Polymers LLC Westhollow Technology Center 3333 Highway 6 South Houston, TX 77082, USA
W. Heckmann BASF AG ZK/Z B001 D-67056 Ludwigshafen Germany
B. Mailhot Laboratoire de Photochimie Moleculaire et Macromoleculaire UMR CNRS 6505 Universite Blaise Pascal (Clermont-Ferrand) F-63177 Aubiere Cedex, France
K.-D. Hungenberg Polymer Technology and Process Development BASF AG D-67056 Ludwigshafen, Germany
M. Malanga R&D Engineering Plastics The Dow Chemical Company Midland, MI 48667 USA
B. King The Dow Chemical Company Midland, MI 48667 USA
M. F. Martin BASF AG D-67056 Ludwigshafen Germany
A. Kistenmacher BASF AG D-67056 Ludwigshafen Germany
B. Matthews Dow Polystyrene R&D Midland, MI 48667 USA
R.-D. Klodt Dow Central Germany Building H 108 D-06258 Schkopau Germany
G. E. McKee BASF AG ZK/Z B001 D-67056 Ludwigshafen Germany
K. Koppi Dow Polystyrene R&D Midland, MI 48667 USA
B. J. Meister 2925 Chippewa Ln. Midland, MI 48640 USA
xxiv
CONTRIBUTORS
M. Morbidelli Swiss Federal Institute of Technology Zurich Laboratorium fur Technische Chemie, ETH Honggerberg, HCI F125 CH-8093 Zurich, Switzerland
F. Ramsteiner BASF AG ZK/Z B001 D-67056 Ludwigshafen Germany
N. Niessner BASF AG Styrene Copolymers and Ultraform ZKT/C-B1 D-67056 Ludwigshafen Germany
A. Rivaton Laboratoire de Photchemie Moleculaire et Macromoleculaire UMR CNRS 6505 Universite Blaise Pascal (Clermont-Ferrand) F-63177 Aubiere Cedex, France
A. N. Paquet The Dow Chemical Company 200 Larkin Building 1605 Joseph Drive Midland MI 48674 USA
G. H. Rubiolo Departmento de Fisica Facultad de Ciencias Exactas y Naturales Universidad Nacional de Buenos Aires Ciudad Universitaria, Pabellon I 1428 Buenos Aires, Argentina
A. Penlidis Chemical Engineering University of Waterloo Waterloo Ontario Canada, N2L 3G1
H. Sakamoto Kanagawa Environmental Research Center 1-3-39 Yonnomiya (Shinomiya), Hiratsuka Kanagawa, 2540014, Japan
R. P6 Polimeri Europa Centro Recerche Novara 'Istituto G. Donegani' Via G. Fauser 4, 1–28100 Novara Italy
R. L. Sammler Materials Sciences The Dow Chemical Company 1702 Building Midland, MI 48672 USA
D. B. Priddy Priddy & Associates LLC 6004 Camelot Ct. Midland, MI 48640, USA
J. Scheirs ExcelPlas PO Box 2080, Edithvale, VIC 3196 Australia
CONTRIBUTORS
XXV
J. Schellenberg R&D Engineering Plastics Dow Central Germany D-06258 Schkopau, Germany
J. P. Viola BASF AG D-67056 Ludwigshafen Germany
N. E. Stacy Chevron Philips Chemical Co. 201A ARB Bartlesville OK 740004 USA
T. Vu-Khanh Universite de Sherbrooke Faculte de Genie/Department de Genie Mecanique 2500 boul. De 1'Universite Sherbrooke, Quebec Canada, J1K 2R1
G. Storti Federal Institute of Technology Zurich Laboratorium fur Technische Chemie, ETH Hoggerberg, HCI F125 CH-8093 Zurich Switzerland
G. C. Welsh The Dow Chemical Company 200 Larkin Center 1605 Joseph Drive Midland MI 48674 USA
K.W. Suh The Dow Chemical Company 200 Larkin Building 1605 Joseph Drive Midland, MI 48674, USA
J. R. Wuensch BASF AG D-67056 Ludwigshafen Germany
N. Tomotsu Polymer Research Laboratory Idemitsu Petrochemical Co., Ltd Anesaki-kaigan, Ichihara Chiba, 229-0193 Japan
K.Yamasaki Plastics Technical Center Idemitsu Petrochemical Co. Ltd Anesaki-kaigan, Ichihara Chiba, 229-0193 Japan
R. Vanspeybroeck 438 Building Dow Chemical Company Midland, MI 48667, USA
This page intentionally left blank
Series Preface The Wiley Series in Polymer Science aims to cover topics in polymer science where significant advances have been made over the past decade. Key features of the series will be developing areas and new frontiers in polymer science and technology. Emerging fields with strong growth potential for the twenty-first century such as nanotechnology, photopolymers, electro-optic polymers, etc., will be covered. Additionally, those polymer classes in which important new members have appeared in recent years will be revisited to provide a comprehensive update. Written by foremost experts in the field from industry and academia, these books have particular emphasis on structure-property relationships of polymers and manufacturing technologies as well as their practical and novel applications. The aim of each book in the series is to provide readers with an in-depth treatment of the state-of-the-art in that field of polymer technology. Collectively, the series will provide a definitive library of the latest advances in the major polymer families as well as significant new fields of development in polymer science. This approach will lead to a better understanding and improve the cross fertilization of ideas between scientists and engineers of many disciplines. The series will be of interest to all polymer scientists and engineers, providing excellent up-to-date coverage of diverse topics in polymer science, and thus will serve as an invaluable ongoing reference collection for any technical library. John Scheirs June 1997
This page intentionally left blank
Polystyrene was the first synthetic polymer to be prepared. In fact there are reports of its existence as early as 1839 (see Chapter 1). Polystyrene was first produced for commercial sale in 1931 by BASF and in the US by Dow in 1938. It is well known that polystyrene is a glassy, amorphous polymer with outstanding clarity, gloss and processability. Unfortunately it is also inherently brittle and suffers from poor chemical resistance (see Chapter 2). These deficiencies were remedied early on by the development of high-impact polystyrene (see Chapter 12 on HIPS) and styrene-acrylonitrile copolymers (see Chapter 13 on SAN copolymers and Chapters 14 and 15 on ABS terpolymers,* as well as Chapter 16 on ASA terpolymers). Styrenic block copolymers have also considerably widened the scope of these polymers from elastomeric materials (Chapter 21) to high-clarity, high impact strength resins (Chapter 22). These latter durable copolymers offer a balance of performance and economics that bridges the gap between high cost, clear engineering polymers and low cost, brittle plastics like general purpose polystyrene. Since polystyrene is one of the oldest commercial polymers with over 9 million tonnes/yr of sales, there have been thousands of patents issued covering all aspects of its manufacture and property enhancement. The styrene monomer readily polymerizes to polystyrene either thermally or with free-radical initiators (see Chapter 6 on free-radical polymerization and Chapter 8 on nitroxidemediated polymerization). Commercial processes for the manufacture of polystyrene are described in Chapter 3 while process modelling and optimization of styrene polymerization is examined in Chapter 5. Styrene also can be polymerized via anionic and Ziegler-Natta chemistries using organometallic initiators. Using free radical and anionic polymerization chemistries, the Strictly speaking, ABS is a rubber-tonghened SAN copolymer.
xxx
PREFACE
specific position of the benzene ring in the monomer units of regular polystyrene is somewhat random and hence inhibits crystallization. Advances in the development of new metallocene polymerization catalyst technology however has enabled the development of syndiotactic polystyrene which is semicrystalline, has a melting point of 270°C and has excellent environmental stress crack resistance (see Chapters 17–20). New metallocene catalyst technology has also enabled the development of novel ethylene-styrene interpolymers (see Chapter 26). Modified variants of polystyrene have also been developed with the advent of hydrogenated PS (Chapter 23), branched PS (Chapter 24) and 'super' PS (Chapter 25). Since the strength, flammability and photodegradation of styrenic polymers have major end-use implications, these properties are covered in detail in Chapters 27, 28, 29 and 30 respectively. The high melt strength of polystyrene enables it to be easily foamed (see Chapters 9 and 10 on PS foam), blown into films, and drawn into sheets (see Chapter 11 on OPS). Polystyrene foams find a variety of uses including insulation and packaging. The family of styrenic polymers now span the breath from commodity plastics to high-grade engineering polymers. Ongoing advances in new catalyst technology and 'controlled radical polymerisation' will undoubtedly yield new styrenic polymers with well-defined architecture (as we have recently seen with the introduction of syndiotactic PS and ethylene-styrene interpolymers). Advances in the synthesis of dendritic and hyperbranched styrenic polymers will also contribute to the state of new polystyrenic products. The key attribute of polystyrene that has led to its huge commercial success is its low cost. Resistance of polystyrene fabricators to pay extra for improved performance and intense competition of polystyrene producers for increased market share have led to highly optimized and huge polystyrene production facilities (a typical 'world-scale' polystyrene plant produces about 230000 tonnes/yr of product). The costs associated with the introduction of new and improved polystyrene products must be low enough that profit can be realized by the manufacturer without raising the sales price. This limitation, and the ongoing effort of the chemical industry to scrutinize/justify R&D budgets, places an intense challenge on industrial polystyrene researchers. Other pressures on the polystyrene industry include environmental and regulatory issues (i.e., litter, migration of residual small molecules into food products, evolution of volatile organics during manufacture and processing, etc.) - see Chapters 4 and 31. These issues will undoubtedly dominate much of the research efforts devoted to polystyrene. Academic researchers are not under such focused cost constraints and therefore they will likely continue working on the development of new chemistries for making new styrenic polymers having novel controlled architectures.
PREFACE
xxxi
The future direction of polystyrene R&D efforts is uncertain but it is likely that it will continue focusing on lowering manufacturing costs, improving product performance/properties (especially flow/strength balance), reducing the level of residual small molecules left in the product, and developing new applications. This book provides the reader with comprehensive information about polystyrene, and a historical overview of its development, as well as reviews describing the latest new technological developments. J. Scheirs and D. B. Priddy June 2002
This page intentionally left blank
About the Editors JOHN SCHEIRS (PhD) John Scheirs is a polymer research specialist with broad interests in polystyrenes and styrenic copolymers. He is the principal consultant with ExcelPlas, a polymer consulting company. John was born in 1965 in Melbourne and studied applied chemistry at the University of Melbourne and obtained a PhD in polymer science. He has worked on projects concerning the fracture, stress cracking, processing, characterization and recycling of styrenic polymers. John has authored over 50 scientific papers, including eight encyclopedia chapters, and a number of books on polymer analysis and polymer recycling.
DUANE PRIDDY (PhD) Duane B. Priddy has worked for the Dow Chemical Company for 33 years, having retired at the end of 2001 from his role as Research Scientist in Polystyrene R&D. Duane began his career in Dow in the Benzene Research Laboratory in 1966. In 1968 he took a two-year leave of absence to attend Michigan State University where he obtained a PhD in Organic Chemistry. In 1970, he returned to Dow. In 1972, Duane joined Polystyrene R&D where he developed the initiator DP275 for polystyrene manufacture. DP275 is currently utilized globally in all Dow polystyrene production plants and is recognized as the industry standard. Duane holds more than 65 patents and has published more than 100 technical papers outside of Dow. He is an adjunct professor at Central Michigan University and at Michigan Technological University. In 2001, Duane was named a Fellow of the American Chemical Society and was awarded the Lifetime Achievement Award by Dow's Polystyrene Business. Most recently he was awarded the Excellence in Science Award by the Midland Section of the American Chemical Society during their annual Fall Scientific Meeting.
This page intentionally left blank
PART I
This page intentionally left blank
1 Historical Overview of Styrenic Polymers JOHN SCHEIRS ExcelPlas, Edithvale, VIC, Australia
1 INTRODUCTION Styrene has been known from the mid-nineteenth century as a clear organic liquid of characteristic pungent sweet odour. It was also known to have the ability to convert itself under certain conditions into a clear resinous solid that is almost odour-free, this resin then being referred to as 'metastyrol'. Styrene readily polymerizes in air and it is therefore not surprising that there are a number of early obscure reports referring to its 'polymerization' predating 1900. However, because the concept of polymerization had not yet been proposed (until Staudinger in 1920), many of these early reports referred to the 'oxidation' or 'hardening' of the styrene monomer. In 1839, Eduard Simon, an apothecary in Berlin, distilled storax resin obtained from the 'Tree of Turkey', (liquid ambar orientalis) with a sodium carbonate solution and obtained an oil which he analysed and named styrol (what we now call styrene) [1]. He recorded the following observation: 'that with old oil the residue which cannot be vaporised without decomposition is greater than with fresh oil, undoubtedly due to a steady conversion of the oil by air, light and heat to a rubberlike substance'. Simon believed he had oxidised the material and called the product styrol oxide. Later, when he realised that it contained no oxygen, the product became known as metastyrol. This puzzled the early chemists as there was no change in empirical formula despite the very pronounced alteration in chemical and physical properties. Unknowingly, this was the first recorded instance of polymerization.
Modern Styrenic Polymers: Polystyrene and Styrenic Copolymers. Edited by J. Scheirs and D. B. Priddy CO 2003 John Wiley & Sons Ltd
4
J. SCHEIRS
A few years later, in 1845, Blyth and Hofmann [2] observed that 'metastyroF was formed when styrene was exposed to sunlight, while it remained unchanged in the dark. This is the first report of photopolymerization. It was established by Blyth and Hofmann in 1845 [2] that 'styrol and its conversion product 'metastyroF had the same elemental composition with an equal number of carbon and hydrogen atoms. After nitration, the N:C ratio was 1:8 for styrol and 1:7 for metastyrol, leading to the conclusion that C8H8 had been converted to C 7 H 7 . However, the conclusion that styrol had polymerized was not reached until 75 years later. The first samples of polystyrene were characterized by the German organic chemist Staudinger [3]. It was observed that the polystyrene could be fractionated into samples with different solution viscosities and this observation was incompatible with the notion that the substance was a colloidal aggregate. Staudinger challenged the notion that polymeric substances are held together by 'association forces'. It was Staudinger who first realised that the solid that Simon had isolated from natural resin was in fact composed of long chains of styrene molecules. Staudinger postulated that polystyrene was a high molecular weight polymer. His critics argued that it could not be a high molecular weight polymer because of its solubility in common solvents. He introduced the term 'macromolecules' to describe these long-chain compounds. Fierce controversy with his colleagues caused Staudinger to move from the Swiss Federal Institute of Technology in Zurich (ETH) to the University of Freiburg. In 1929, Staudinger and co-workers also synthesized hexahydropolystyrene by the nickel-catalysed hydrogenation of polystyrene [4,5]. The hydrogenated polystyrene, also known as poly(cyclohexylethylene), had improved oxidative and radiation stability relative to conventional polystyrene. It was also Staudinger in 1932 who first proposed that the inability of polystyrene to crystallize was due to its lack of stereoregularity which rendered it amorphous. It is its amorphous nature that is responsible for its solubility though others claimed that polymer solubility was incompatible with very high molecular weight [3].
2 GENERAL-PURPOSE POLYSTYRENE (GPPS) Styrene readily polymerizes to polystyrene (PS) either thermally or with freeradical initiators. A limiting factor in the commercial exploitation of polystyrene was the high reactivity and considerable heat of polymerization of styrene. The polymerization rate of styrene is exceedingly fast and considerable heat is generated. This was an intimidating obstacle to commercial production of PS since many in the industry were concerned that the large-scale polymerization of styrene may result in a dangerous uncontrolled reaction. The process
HISTORICAL OVERVIEW OF STYRENIC POLYMERS
5
involved heating styrene monomer in bulk containers. A major limitation of this approach was the need for heat removal from the highly viscous melt. High temperatures can be reached in large mass reactors (>300°C) and thermal degradation of the resultant PS can occur. This problem was later solved by installing heat exchanger tubes in the reaction medium. The first commercial production of PS was in 1931 by BASF. To prevent premature polymerization of styrene monomer, special inhibitors had to be added so that it could be stored until needed. Polymerization inhibitors were also required to prevent polymer formation during distillation of styrene monomer from ethylbenzene. Before the large-scale production of PS could occur, a consistent supply of styrene monomer was required. In 1930, Dow started to produce styrene monomer by cracking its ethylbenezene precursor. In 1938, Dow began manufacturing commercial quantities of PS. The propensity for styrene monomer to polymerize allowed extremely simple and crude polymerization techniques. The first technique used by Dow was known as the 'can process' since it basically involved filling 10 gallon metal cans (Figure 1.1) with styrene monomer followed by heating the cans in a heating bath at progressively higher temperatures for a number of days. After this time the polystyrene (polymerized to approximately 99 % conversion) was removed from the can and crushed to a free-flowing powder. The development of styrene and PS manufacturing technology was spurred on by the advent of World War II. During this time the supply of natural
Figure 1.1 Early photograph of the 'can' process for the commercial production of polystyrene. This simple process involved filling 10 gallon metal cans with styrene monomer, thermally polymerizing it in heated baths and then grinding the polystyrene cylinders that formed, (courtesy of Dow Chemical Company)
6
J. SCHEIRS
rubber from the Far East was terminated. The acute rubber shortage accelerated the development of styrene-based synthetic rubber. With the outbreak of the war, the United States embarked on a scientific programme that rivalled the Manhattan Project in its scope and significance. Nearly a billion dollars was spent on research and development of synthetic rubber needed to keep the Allied war effort in motion. The key players such as Dow, Monsanto and Koppers Chemical cooperatively produced record quantities of styrene monomer for the preparation of styrene-butadiene rubber. A staggering 180 0001 of styrene monomer were produced per year towards the end of the War with most being used for the production of the synthetic rubber Buna-S (also known as GRS rubber, where GR stands for government rubber and S for styrene). There was also cooperation between the main rubber-producing companies, Goodyear, B. F. Goodrich, Standard Oil, Firestone and US Rubber. In early 1942, the American Synthetic Rubber Research Program commenced. Along with the major rubber-producing companies, 11 university research groups, including Carl 'Speed' Marvel at the University of Illinois, Izaak 'Piet' Kolthoff at the University of Minnesota and W. D. Harkins and Morris Kharasch of the University of Chicago, joined the effort to make synthetic rubber work. Their objective was to set up four plants that would produce 30 000t each of synthetic rubber per year. By the end of 1942, four plants were established but their output was under the target. By the end of 1943, 15 plants were in operation, and supply had begun to meet demand. The research focus during the War was on refinement, enhancement and incremental improvement of existing processes. For example, if the rubber is allowed to polymerize until no monomer is left then long, branched molecules are produced, which gel and make the rubber difficult to process. To solve this problem, the reaction is only allowed to proceed to 72 % conversion and a thiol modifier, a chain-transfer agent, was used to control molecular weight. It was also observed that the polymerizations have an 'induction period' which varied from batch to batch. During the induction period nothing seems to be happening, then, all of the sudden, the reaction takes off. The researchers at the University of Illinois found that this is due to different fatty acids present in the different soaps needed for the emulsion process. These soaps also cause the solution to foam during the recovery of the remaining monomer. This problem leads to the development of silicone defoamers. The properties of the Buna-S type rubber are highly dependent on the amount of styrene in the rubber. To determine properties, it is important to know how much styrene had been incorporated. William O. Baker of Bell Telephone Laboratories solved this problem by developing a procedure for determining the amount of styrene using the refractive index of a solution of the rubber. It was not until after World War II, when styrene monomer capacity could be diverted from its essential wartime use for styrene-butadiene synthetic rubber, that polystyrene became an important commercial plastic. When the War fin-
HISTORICAL OVERVIEW OF STYRENIC POLYMERS
7
ished the supply route for natural rubber was re-established and there was an oversupply of styrene monomer. The extensive infrastructure for styrene production and the enormous body of process and technical knowledge laid the foundation for the post-War development of polystyrene and styrenic copolymers. Prior to 1941, Germany had a major technical and industrial lead over the USA, having already established an industrial styrene monomer production process, a styrene-butadiene elastomer process and a mass styrene polymerization process [6]. Figure 1.2 shows the polymerization vessels at I. G. Farben in 1940. Figure 1.3 shows a bank of polymerization kettles. The Germans began the first technical production of polystyrene in 1930 while the first production of polystyrene in the USA was some 8 years later by Dow in 1938. Interestingly, at the inception Dow did not have a strategic objective to enter into the polystyrene business. Rather, Dow believed that ethylcellulose and poly (vinylene chloride) (Saran) were the commodity polymers of the future. Dow was producing ethylbenzene as a solvent and electrical fluid. However, when the markets for ethylbenzene did not develop it decided to crack the ethylbenzene and produce styrene. After it had stockpiled large quantities of unstable styrene, Dow initiated a 'crash' programme to develop polystyrene. Thus even though Dow did not initially intend to produce polystyrene commercially when its petrochemical programme was initiated it became a logical business decision to do so. Early on there were numerous technical barriers that made polystyrene difficult to produce and to process. For example, it was made by an extremely slow production process and its high average molecular weight and broad molecular weight distribution made it difficult to injection mould [6]. Dow researchers ultimately developed ways to lower the average molecular weight and added certain lubricants to improve processability, thus making general-purpose polystyrene which fast acquired the reputation of being the easiest thermoplastic to mould. Other technical barriers were the need to control the exotherm of polymerization and to produce colour-free polystyrene. While the manufacture of styrene seems simple and straightforward, in the early days at Dow there were three major impurities in the styrene monomer apart from residual ethylbenzene. These were phenylacetylene (which acted as an inhibitor for styrene polymerization), divinylbenzene (which caused plugging and fouling of the distillation column for separating styrene from its precursor, ethyl benzene) and sulphur (which caused discoloration of the polystyrene). Finally, in 1938, the 'crash' programme resulted in the first saleable polystyrene batches. This was produced in metal cans lined with tin to yield high-purity polystyrene. These cans were filled with styrene and immersed in heated waterbaths where the styrene would thermally polymerize. The process was very slow and labour intensive. Further, the exotherm of polymerization was greatest in the centre of each can, which led to a core of lower than average molecular weight. After polymerization was complete, the polystyrene billet was ground up and mixed to distribute the different molecular weight regions [6].
J. SCHEIRS
Figure 1.2 Photograph taken in 1940 of a styrene polymerization vessel inside the I. G. Farben plant in Ludwigshafen, Germany (courtesy of BASF, Ludwigshafen)
Figure 1.3 Reaction kettles in the BASF polystyrene production plant (courtesy of BASF, Ludwigshafen)
10
J. SCHEIRS
Although the can process was very slow, it lent itself to easy expansion of production output by simply adding more cans and heating baths. It was also discovered that by adding some peroxide catalyst to the styrene monomer the production throughput could be increased significantly. In fact, this innovation led to a doubling of Dow's plant capacity since faster polymerization rates could be achieved while still controlling the exotherm [6]. While the USA was progressing with the can process, Germany had already developed a continuous process for the mass polymerization of styrene. After World War II, researchers from Dow visited the German polystyrene plants and were surprised to learn of their scale and sophistication. One of the key people on the investigating team that went to Germany was Dr Goggin, founder of Dow's Plastics Technical Service Department. The American teams that visited I. G. Farben after the War recorded their findings in a historic report [7]. This report clearly showed the advantages of a continuous production process for polystyrene. Further, the first industrial production of SAN was in 1936 also by I. G. Farben in Ludwigshafen. Central to Germany's development of polystyrene technology was Herman F. Mark (Figure 1.4). Mark worked at I. G. Farben Industrie for 6 years from 1927 to 1932, first as a research chemist (1927-28), then as Group Leader (1928–30) and finally as Assistant Research Director (1930-32). Because of the changing political climate, Mark moved to the University of Vienna, where he became Professor of Chemistry and Director of the First Chemical Institute (1932-38). While at I. G. Farben Industrie, Mark played a major role in the development of styrene monomer and PS. Mark patented a process in 1929 for the production of styrene from ethylbenzene via catalytic dehydrogenation [8]. The German Chemical giant I. G. Farben developed the continuous tower process for PS in the 1930s. The German PS polymerization plant shown in Figure 1.5 overcame the problem of the polymerization exotherm and thermal runaway by using a tank reactor with heat-transfer tubes criss-crossed through it. The reaction temperature was gradually increased and controlled, and polystyrene was removed via an auger. This design was later improved by prepolymerizing in stirred kettles prior to the tower process (Figure 1.6). After the War, Dow began to focus on constructing its own continuous mass polymerization plants for PS. Known as the tube tank process, it consisted of two nonagitated horizontal tube tanks containing arrays of tubes through which a heat-transfer fluid (Dowtherm™) flowed in order to control the exotherm of polymerization (Figure 1.7). Each tank had a capacity of 18000kg of styrene monomer and represented a batch process, but when alternately sequenced the process was continuous. When the styrene in tank 1 had reached high conversions, a special polymer pump pumped the molten polystyrene at 220240 °C to the bottom receiving tank. Polymerization was then started in tank 2. The bottom receiving tank was under vacuum to remove volatiles such as unreacted monomer and also dimers, trimers and other oligomers. There was
HISTORICAL OVERVIEW OF STYRENIC POLYMERS
11
always polystyrene in the receiving tank so that the extruder and pelletization process could operate on a continuous basis [6]. The exotherm in the Dow process supplied most of the heat needed to produce a molten PS ready for pelletization. These units were extremely successful because of the very large heat transfer surface and the efficiency of the Dowtherm™ heat transfer fluid. Twelve such tube tank plants were installed and operated at the Dow site in Midland, Michigan and they produced prodigious quantities of polystyrene over many years [6].
Figure 1.4 Photograph of Herman F. Mark taken in 1936. Mark worked at I. G. Farben Industries in Germany for 6 years, from 1927 to 1932, and played a major role in the industrial development of styrene monomer and polystyrene (courtesy of BASF, Ludwigshafen)
12
J. SCHEIRS Styrene Reaction Temperature
!00°C
!50°C
BASF's Continous Styrene Polymerization Plant - circa 1932
Figure 1.5 Schematic of BASF's early tower process for the continuous polymerization of styrene. This configuration was designed by C. Wulff and E. Dorrer in the early 1930s. Polymerization was thermally initiated and the exotherm controlled by heat transfer tubes (courtesy of BASF, Ludwigshafen)
In the following years, other methods of polymerizing styrene were developed such as suspension polymerization by Koppers Chemical, which was first introduced in the 1940s and which showed rapid development in the 1950s. The suspension polymerization process is still in use for the production of PS [9], although it has been largely replaced by more economical techniques such as continuous mass polymerization. It is interesting that the polystyrene produced by suspension polymerization, particularly the Koppers material, had a heat distortion temperature superior to that of the Dow polystyrene [6]. This was attributed to the measurable levels of residual dimers and trimers in the Dow product due to its thermal initiation and which were absent in the peroxide-initiated suspension process. The suspension polymerization process has a number of distinct advantages over competitive processes. It allows excellent control over the polymerization temperature and a lower viscosity reaction medium. Furthermore,
HISTORICAL OVERVIEW OF STYRENIC POLYMERS styrene
13
styrene reaction temperature
prepolymerization
polymerization
extrusion BASF's Continuous Styrene Polymerization Plant - circa 1936 Figure 1.6 Schematic of BASF's improved tower process for the continuous polymerization of styrene. In this design (dated 1936) the styrene was first polymerized up to 30–35% conversion in a stirred kettle and then transferred to the tower reactor for polymerization up to 97% completion (courtesy of BASF, Ludwigshafen)
expandable polystyrene and high-impact polystyrene are also produced by this technique.
3
FOAMED POLYSTYRENE
The concept of cellular polystyrene was first reported in 1935 by the Swedish inventors Munters and Tandberg [10], who filed a patent entitled 'Foamed Polystyrene'. Ray Mclntire, a young researcher at Dow Chemical, is credited with inventing Styrofoam. Mclntire said his invention of foamed polystyrene was accidental. His invention came as he was trying to find a flexible electrical insulator around the time of World War II. He worked at developing a rubberlike substance that could serve as an electrical insulator. Although polystyrene was a good insulator it was far too brittle. Mclntire tried to make a new rubberlike polymer by combining styrene with isobutene, a volatile liquid, under pressure. He tried combining styrene with isobutene, but he accidentally
14
J. SCHEIRS MONOMER TUBE TANK I o oo o o ooo o o o o oo o oo o oo o oooooo 0 0 0 0 0 0 0
o ooo o ooo oooo ooo oo o
o o o o oo oO oo
o o o o
o o o o
o o o o
o o o o
oo oo o o
O O O 0
o oo
MOLTEN POLYMER PUMPS
TUBE TANK 2 o oo o o o o o oo 0000 o o o o o oooo o o o o o ooooo o o o o o ooooo o o o o o ooooo o o o o o ooooo o o o o o oooo o o o o o oooo o o o o o ooo O O O 0 o o
o o o o o o o o
o o o o
X^
VACUUM
MOLTEN V. ••'/ POLYMER ••// • 's
COLORANTS I I
z
1
t
EXTRUDER
Figure 1.7 Schematic of Dow's 'tube tank' process which represents the first commercial continuous polymerization process for polystyrene in the USA. The figure shows a cross-section through the centre of three longitudinal unagitated tanks. Styrene was thermally polymerized in tube tanks 1 and 2 and then devolatilized in the bottom receiving tank, which was always about half full and under vacuum [adapted from Boyer, R. F., J. Macromol. Sci. Chem., A15, 1411 (1981)] added too much of the latter - and was surprised to see that the isobutene formed tiny bubbles. The result was a foamed polystyrene with a cellular microstructure, 30 times lighter than regular polystyrene. Mclntire stayed with Dow Chemical until his retirement in 1981. The word Styrofoam is still trademarked by Dow, and it technically only applies to a kind of insulation
HISTORICAL OVERVIEW OF STYRENIC POLYMERS
15
used for building materials. Today, however, many companies produce products made of PS foam, and the brand name Styrofoam is commonly used to describe them all. It was only in the early 1940s that commercial production of cellular polystyrene began. In 1942, Dow began research on an extrusion process for the production of PS foam using a low-boiling chlorocarbon (methylene chloride) as the blowing agent. The product was extruded into large foam logs, which were then cut into boards and planks. This material was given the trademark Styrofoam™ in 1943 [11]. This foam was rapidly adopted by the US Coast Guard and US Navy as a buoyancy medium and insulation material. BASF developed its own process for foaming polystyrene in the early 1940s (Figure 1.8). This process was later refined by the improved suspension polymerization process which produces foamable polystyrene beads. A blowing agent (typically pentane) can be introduced during the polymerization of styrene or introduced later in a separate impregnation step under pressure and heat [11]. All major PS foam bead producers took out a license from BASF for this patented technology [12–14]. Figure 1.9 shows a promotional illustration from 1952. The single most important factor responsible for the rapid commercial growth of expandable PS is its ability to be steam-moulded into lightweight, closed-cell, low-cost foams suitable for beverage cups, packages, ice buckets, picnic chests, insulation board, etc [11]. (Figure 1.10).
Figure 1.8 Early photograph (ca 1948) showing some of the earliest polystyrene foam (Styropor™). Foamed polystyrene has unrivalled low-density and thermal insulating properties (courtesy of BASF, Ludwigshafen)
16
J. SCHEIRS ,,Das fekhteste Schiff der Welt ist das STYROPOR-Schiffchen
Figure 1.9 A promotional photograph highlighting showcasing polystyrene foam for the BASF stand at the 1952 Kunststoffemesse show in Dusseldorf, Germany (courtesy of BASF, Ludwigshafen)
HISTORICAL OVERVIEW OF STYRENIC POLYMERS
17
Figure 1.10 Polystyrene foam rapidly became the packaging material of choice for everything from medical instruments to engines (courtesy of BASF, Ludwigshafen)
In the late 1960s, demand for foamed PS increased dramatically owing to its increasing use in meat trays, fruit boxes and egg cartons. The expansion was stimulated by the design of extruders with provision for introducing the blowing agent into the barrel, thereby obviating the need to buy the more expensive expandable PS pellets as a raw material [11]. In 1969, the market for PS foam grew enormously especially for self-extinguishing grades. The main applications were for insulation board in cold-storage refrigeration, housing insulation and cut ceiling tiles.
18
4
J. SCHEIRS
RUBBER-MODIFIED POLYSTYRENE
Rubber-modified polystyrene was the next logical evolution after generalpurpose polystyrene. Very early on it was apparent that the Achilles heel of polystyrene was its inherent brittleness. Rubber-modified polystyrene is a twophase system consisting of a dispersed rubber phase and a continuous polystyrene phase (or matrix). Impact-modified polystyrene was invented as early as 1927 by Ostromislensky [15] by addition of natural rubber either polymerized with styrene or blended in polystyrene. In the early 1940s, researchers at Dow produced interpolymer blends of styrene and butadiene by an emulsion process. The polymer, called Styralloy™ 22, was used as insulation for radar cables until it was displaced by low-density polyethylene produced by ICI. Later, Dow experimented with soluble GRS copolymerized with styrene to make high-impact polystyrene. In 1954, Haward of Shell obtained a patent for rubber-modified PS made by suspension polymerization [16]. This early product, however, contained 'fisheyes' - small crosslinked gel particles - since each suspension particle was crosslinked by styrene-butadiene rubber. Researchers at Monsanto overcame this problem [17] by including a prepolymerization step with shearing agitation. In 1954, Dow finally perfected a 'can' process to make high-impact polystyrene (HIPS). The secret was that the traditional 'can' process could not simply be used since the product would be full of gel particles of rubber ('fish-eyes'); instead, the styrene-rubber mixture was first carried out to 30% conversion with shearing agitation. Then the mixture was transferred to 10 gallon cans where the reaction was completed. This process was documented in the now famous Amos patent [18]. In the late 1960s, Dow initiated patent infringement suits against its major competitors (Monsanto, Standard Oil, Amoco Chemical, Dart Industries) over the patent for high-impact polystyrene by Amos [6].
5 ABS While Dow had experimented extensively with ABS-type polymers and even produced ABS in their commercial HIPS plant, they lost the lead in the development of commercial ABS resins. Dow sued Monsanto in 1969 for infringement of the ABS claim in the patent by Amos. The judge ruled in Monsanto's favour on the basis that Monsanto's product was outside the claimed composition. In another case, where Dow sued Dart Industries, the judge ruled that the Amos patent was invalid for reasons of obviousness [6].
HISTORICAL OVERVIEW OF STYRENIC POLYMERS
19
As early as 1948 it was known that poly(styrene-co-acrylonitrile) or SAN could be blended with Buna-N (a copolymer of butadiene and acrylonitrile) or Buna-S (a copolymer of butadiene and styrene) to produce useful thermoplastics [19]. The commercial introduction of these polymers, however, was restricted by their poor low-temperature impact properties. Researchers at Marbon (a division of Marsene Corp., later renamed Borg Warner) knew that polybutadiene remained rubbery at temperatures lower than the earlier cited copolymers; however, polybutadiene and SAN were incompatible [20]. The trick was first to produce polybutadiene by emulsion polymerization and then to use this latex as the medium for the emulsion polymerization of styrene and acrylonitrile (Figure 1.11). In 1959, Borg Warner patented this ABS produced by grafting styrene plus acrylonitrile into polybutadiene. The Cyclolac™ brand of grafted ABS produced by Borg Warner rapidly became the market leader in ABS. In 1988, GE Plastics acquired Borg Warner Chemicals and its ABS technology. The chief researcher responsible for this discovery was Calvert [21,22]. The material he produced from the emulsion copolymerization had uniformly dispersed domains of rubber in a continuous phase of SAN. The rubber dispersion was stabilized by the SAN that was grafted to the polybutadiene emulsion particles [20]. It is interesting to note that Borg Warner conducted more than 10000 laboratory trials before the optimum ABS composition was commercially produced in Styrene Acrylonitrile
Emulsion Polymerization of Styrene and Acrylonitrile Figure 1.11 Schematic of BASF's stirred tank emulsion polymerization reactors (dated 1940) for the production of styrenic copolymers (courtesy of BASF, Ludwigshafen)
20
J. SCHEIRS
1954. In 1957, Cyclolac™ T was commercialized. The T designated toughness and later it was thought to stand for telephone as this grade of ABS became the industry standard for making telephone housings. The ability of ABS sheets to be thermoformed opened the door to a range of markets such as luggage (Samsonite™), machine housings, refrigerator liners and boat hulls. When ABS was first commercialized, there was much confusion in the plastics industry referring to it as a terpolymer. The system is not a terpolymer as butadiene is added to the reactor as a polymer along with styrene and acrylonitrile monomers. Polymerization causes SAN to be grafted to the rubber to produce a dispersible domain. It is indeed a requirement that the polybutadiene regions exist as a separate phase of a specified size. Since the domain size is critical to its impact properties, it is important that it is stable through compounding and processing steps [20]. High-heat versions of ABS were subsequently introduced by using a-methylstyrene as a partial replacement of styrene, in both the graft and the matrix. This resulted in an ABS polymer with a higher heat deflection temperature. The heat deflection temperature could be tailored by varying the level of a-methylstyrene. These new high-heat versions of Cyclolac™ ABS were introduced to the automotive market in 1958 and won widespread acceptance [20]. The high heat resistance grades, however, had higher melt viscosities, making them more difficult to process owing to the stiffer molecule that results from the addition of methyl groups to the backbone. The next evolution in ABS technology was the need to produce a transparent ABS. Existing ABS was opaque owing to the scattering of light by the rubber domains. While producing smaller domains would make the system clear, it led to a loss of impact strength. The answer was to modify the refractive index of the components so that the various phases were less optically different. A fourth 'monomer', methyl methacrylate, was used to minimize the refractive index variation in the ABS and a clear impact-resistant thermoplastic named Cyclolac™ CIT was achieved [20].
6 ASA
Acrylonitrile-styrene-acrylate (ASA) polymers share obvious similarities with ABS but ASA was only developed in the 1960s. ASA polymers are essentially SAN polymers impact modified with an acrylate rubber. The earliest attempt to make ASA was by Herbig and Salyer of Monsanto [23] using butyl acrylate as the rubber phase. This work was then refined by Otto [24] and Siebel [25], both of BASF, who copolymerized butyl acrylate with butadiene to prepare the rubber phase.
HISTORICAL OVERVIEW OF STYRENIC POLYMERS
7
21
EARLY STYRENE COPOLYMERS
The first styrene copolymer was reported in 1930 by Wagner-Juaregg and was a copolymer of styrene and maleic anhydride [26]. This copolymer (SMA), which was called a heteropolymer by its inventor, has excellent resistance to continuous exposure in boiling water. Koppers produced SMA moulding powders under the tradename Dylark™. Arco has since acquired this business and continues to produce these SMA resins today under the Dylark tradename. Another styrene copolymer with better heat resistance than regular polystyrene is the copolymer of styrene and fumaronitrile which was reported in 1948 [27]. Both of these styrene copolymers are based on nonpolymerizable monomers - that is, fumaronitrile, like its corresponding anhydride (maleic anhydride), does not form homopolymers but readily copolymerizes with styrene at levels of up to 40%. Monsanto attempted to commercialize the styrene-fumaronitrile copolymer under the tradename Cerex™, but residual fumaronitrile was a powerful vesicant (an irritant which causes blisters) and the project was shelved [28].
8
STYRENIC BLOCK COPOLYMERS
The advent of alkyllithium-initiated anionic polymerization based on the fundamental work by Szwarc in 1956 [29,30] opened the door to the commercial development of styrene-butadiene copolymers. Styrenic block copolymers were first produced in the late 1950s after the discovery by Szwarc of living anionic polymerization [31]. In the late 1950s, Shell, Phillips and Firestone manufactured styrene-butadiene copolymers produced by anionic polymerization also referred to as 'living polymerization'. The styrene-butadiene copolymer named K-resin with a high styrene content was invented in the early 1960s at Phillips Petroleum by Alonzo Kitchen and is purportedly named after his surname initial. The alkyllithium initiators such as n-butyllithium allowed the production of copolymers with tightly defined and controlled polymer microstructures with highly regular block structures. To this day, living anionic polymerization continues to be the main commercial route for the production of styrenic block copolymers [31]. Most of these polymers are made using butyllithium catalysts. Today the following styrene–butadiene copolymers are well known under the tradenames Kraton™, K-Resin™, Styroflex™ and Styrolux™.
22
J. SCHEIRS
9 SYNDIOTACTIC POLYSTYRENE Syndiotactic polystyrene (SPS) with its melting point of 270 °C has been claimed as the first styrenic engineering plastic. Syndiotactic polystyrene was first synthesized in 1985 by Ishihara of Idemitsu Kosan by using titanium complexes/methylaluminoxane catalyst [32–34]. SPS displays entirely different properties to conventional polystyrene such as high chemical resistance and excellent environmental stress crack resistance. SPS does share one major property with conventional PS, however, namely inherent brittleness. For this reason, SPS is generally modified with rubber tougheners or glass fibre reinforcement. When SPS is reinforced with glass fibres it has comparable toughness to glass-filled PBT and glass-filled nylon 66. Blending of SPS with other polymers is another strategy for improving performance properties (see Chapter 20). Dow and the Idemitsu Petrochemical entered into a cooperation in 1988 and are now commercially producing SPS under the tradenames Questra™ and Xarec™, respectively.
10 MODERN POLYSTYRENE PRODUCTION In the late 1970s, Dow undertook a major programme to change the way polystyrene was commercially produced. The conversion involved moving from thermally initiated polystyrene polymerization to polymerization initiated by a bifunctional initiator (Dow calls the initiator DP275, named after its inventor Duane Priddy). Today the majority of general-purpose polystyrene is produced by solution polymerization in a continuous process. The solution process allows for the easy removal of heat from the polymerization medium. This technique, however, necessitates the use of exhaustive post-polymerization devolatilization equipment employing high temperatures and high vacuum to ensure the removal of volatiles and oligomers.
11 THE FUTURE Figure 1.12 shows the timeline of discovery of various styrenic polymers and copolymers. It would be naive to suggest that the rate of invention and innovation will level off in this century. Rather, the pace of discovery of new styrenic polymers will probably increase. Advances in new catalyst technology and 'controlled radical polymerisation' technology will undoubtedly yield new styrenic polymers with well-defined architecture, as we have recently seen with the introduction of Syndiotactic PS and ethylene-styrene interpolymers.
23
HISTORICAL OVERVIEW OF STYRENIC POLYMERS DOW 1 ES
IDEMITSU
SYNDIOTACTIC PS
BASF
LOCK COPOLYMERS ASA ABS
HIGH IMPACT PS
BASF STYRENE^MONOMER
BASF I 1930
1940
1950
1960
1970
1980
1990
2000
Figure 1.12 Timeline of the development of styrenic polymers (adapted from a BASF document by Franz Haaf, entitled '50 Jahre Polystyrol - Entwicklung', BASF, Ludwigshafen)
Advances in the synthesis of dendritic and hyperbranched styrenic polymers will also contribute to the slate of new products. Hitherto uncommerciallized are a range of polymers in which styrene is copolymerized with three or more other monomers. This is because styrene readily copolymerizes with other monomers, even nonpolymerizable monomers such as anhydrides and nitriles. As far as process technology is concerned, all commercial polystyrene has until now been produced using free radical chemistry. BASF and Asahi have recently been aggressively developing 'retarded' anionic polymerization process technology which will allow them to manufacture polystyrene (both general-purpose polystyrene and high-impact polystyrene) in continuous bulk polymerization processes. The advantage of these new anionic processes will be the production of highly pure PS virtually free of traces of volatile impurities such as unreacted styrene monomer. Also, anionic polymerization processes allow the production of PS having a more narrow polydispersity, resulting in an improved flow/ strength property balance.
REFERENCES 1. Simon, E., Ann. Chem., 31, 265 (1839). 2. Blyth, J. and Hofmann, A. W., Ann. Chem., 53, 289 (1845).
24
J. SCHEIRS
3. Staudinger, H., Die Hochmolekularen Organischen Verbindungen, Kautschuk und Cellulose, Springer, Berlin, 1932. 4. Staudinger, H., Geiger, E. and Huber, E., Chem. Ber., 62, 263 (1929). 5. Staudinger, H. and Wiedersham, V., Chem. Ber., 62, 2406 (1929). 6. Boyer, R. F., J. Macromol. Sci. Chem., A15, 1411 (1981). 7. DeBell, J. M., Goggin, W. C. and Gloor, W. E., German Plastics Practice, Debell and Richardson, Springfield, MA, 1946. 8. Mark, H. and Wulff, C., German Patent DRP 550055 (to I. G. Farben Industrie) (1929). 9. Meister, B. J. and Malanga, M. T., 'Styrene Polymers', in Encyclopedia of Polymer Science and Engineering, ed. Moore, E. R.), Wiley, New York, vol. 16, p. 21 (1989). 10. Munters, C. G. and Tandberg, J. G., US Patent 2023204 (1935). 11. Frisch, K. C., 'History of Science and Technology of Polymeric Foams', in 'History of Polymer Science and Technology', eds. Seymour, R. B., Marcel Dekker, New York (1982). 12. Stastney, F. and Goeth, R., US Patent 2681 321 (to BASF) (1956). 13. Stastney, F. and Buchholz, K., US Patent 2744291 (to BASF) (1956). 14. Stastney, F., US Patent 2787809 (to BASF) (1957). 15. Ostromislensky, J. J., US Patent 1 613673 (1927). 16. Haward, R. N. and Elly, J., US Patent 2668806 (1954). 17. Stein, A. and Walter, R. L., US Patent 2862909 (to Monsanto) (1958). 18. Amos, J. L., Mclntire, O. R. and McCurdy, J. L., US Patent 2694692 (to Dow) (1954); Amos, J. L., Polym. Eng. Sci., 14, 1 (1974). 19. Daly, L. E., US Patent 2435202 (to US Rubber Co.) (1942). 20. Pavelich, W. A., 'A Path to ABS Thermoplastics', in High Performance Polymers: Their Origin and Development, ed. Seymour, R. B. and Kirshenbaum, G. S., Elsevier, New York, p. 125 (1986). 21. Calvert, W. C., US Patent 2908661 (to Borg-Warner) (1959). 22. Calvert, W. C., US Patent 3238275 (to Borg-Warner) (1966). 23. Herbig, J. A. and Salyer, I. O., US Patent 3 118855 (to Monsanto) (1964). 24. Otto, H.-W., German Patent DE 1 182811 (to BASF) (1965). 25. Siebel, H. P. and Otto, H.-W., German Patent DE 1 238207 (to BASF) (1967). 26. Wagner-Juaregg, T., Chem. Ber., 63, 3213 (1930). 27. Fordyce, R. G., Chapin, E. C. and Ham, G. E., J. Am. Chem. Soc., 10, 2489 (1948). 28. Seymour, R. B., 'Styrene-Maleic Anhydride-Vinyl Monomer Terpolymers and Blends', in High Performance Polymers: Their Origin and Development, ed. Seymour, R. B. and Kirshenbaum, G. S., Elsevier, New York, p. 125 (1986). 29. Swarc, M., Levy, M. and Milkovich, R., J. Am. Chem. Soc., 78, 2656 (1956). 30. Swarc, M., Nature (London), 178, 1168 (1956). 31. Szwarc, M.,J. Polym. Sci., Part A, Polym. Chem., 36, ix (1998). 32. Ishihara, N., Kuramoto, M. and Uoi, M., Japanese Patent JP 62 187708 (to Idemitsu Kosan Company) (1985). 33. Ishihara, N., Kuramoto, M. and Uoi, M., European Patent JP 210615 (to Idemitsu Kosan Company) (1986). 34. Ishihara, N., Kuramoto, M. and Uoi, M., Macromolecules, 21, 2464 (1986).
Polystyrenes and Styrene Copolymers - An Overview NORBERT NIESSNER AND HERMANN GAUSEPOHL BASF AG, Ludwigshafen, Germany
1
INTRODUCTION
Polystyrene is one of the most widely used thermoplastic materials ranking behind polyolefins and PVC. Owing to their special property profile, styrene polymers are placed between commodity and speciality polymers. Since its commercial introduction in the 1930s until the present day, polystyrene has been subjected to numerous improvements. The main development directions were aimed at copolymerization of styrene with polar comonomers such as acrylonitrile, (meth)acrylates or maleic anhydride, at impact modification with different rubbers or styrene-butadiene block copolymers and at blending with other polymers such as polyphenylene ether (PPE) or polyolefins. Polystyrene ('general-purpose polystyrene', GPPS) can be understood as a linear polyethylene chain with laterally attached phenyl rings, being responsible for the enhanced glass transition temperature and high refractive index. Stiffness, brilliance, gloss and hardness are the main characteristics of this material. Consequently, applications such as audio/video cassette packs, beakers, transparent food packagings, shower cabinets, lamp covers, etc., are dominant. To overcome the brittleness of GPPS, the material was modified by incorporation of polybutadiene. Impact-modified polystyrene (IPS) was invented by Ostromislensky [1] and has been commercialized since the 1950s. IPS consists of a polystyrene matrix with embedded cellular rubber particles. By rubber
Modern Styrenic Polymers: Polystyrene and Styrenic Copolymers, Edited by J. Scheirs and D. B. Priddy © 2003 John Wiley & Sons Ltd
26
N. NIESSNER AND H. GAUSEPOHL
toughening, however, transparency is lost. The use of styrene—butadiene block copolymers instead of polybutadiene results in translucent impact polystyrene with a core/shell particle morphology. These particles are small and thus reduce scattering of visible light. Enhanced property demands in the packaging sector and also in the electric/ electronic and automotive sectors require improved product properties. Homogeneously miscible blends with, e.g., polyphenylene ether (PPE) combine the excellent processability of the amorphous polystyrene with the thermal stabilty of its blend partners. Styrenic copolymers are materials capable of thermoplastic processing which, in addition to styrene (S), also contain at least one other monomer in the main polymer chain. Styrene—acrylonitrile (SAN) copolymers are the most important representative and basic building blocks of the entire class of products. By adding rubbers to SAN either ABS (acrylonitrile-butadiene-styrene) or ASA (acrylate—styrene—acrylonitrile) polymers are obtained depending on the type of rubber component employed. These two classes of products yield blends composed of ASA and polycarbonate (ASA + PC) or ABS and polyamide (ABS + PA). MABS polymers (methyl methacrylate—acrylonitrile—butadiene—styrene) together with blends composed of polyphenylene ether and impact-resistant polystyrene (PPE/PS-I) also form part of the styrenic copolymer product range. Figure 2.1 provides an overview of the different classes of products and trade names. A characteristic property is their amorphous nature, i.e. high dimensional stability and largely constant mechanical properties to just below the glass transition temperature, Tg.
I
Transparent base polymer | I
I
ps | I
*-v
Addition of rubber 1
• • ' Impact-modified polymer |m «l I • I
Blend PPE/PS-I
ABS
Figure 2.1 Overview of the different classes of styrene polymers
27
POLYSTYRENES AND STYRENE COPOLYMERS
2
POLYMERIZATION
Styrene is one of the few monomers able to be polymerized under free radical, anionic, cationic and metal catalysed conditions. This is due to low polarity of the styrene molecule and to the resonance stabilization of the growing polystyryl species in the transition state. According to Mayo, [2,3] it is assumed that styrene forms a Diels—Alder adduct which isomerizes to phenyltetralin (a) or transfers a hydrogen atom to a further styrene molecule forming two radicals in a solvent cage. These radicals are stabilized by disproportionation (b, c) or recombination (d) to cyclic dimers and trimers. Diffusion of the radicals from the cage and subsequent polymerization to polystyrene is in fact only a side reaction in this scheme (Figure 2.2). Recently, strong evidence in support of this mechanism was obtained by Buzanowski et al. [4]. In addition to the thermal initiation, the use of peroxides or azo components is a common and well established method to start the chain reaction. Peroxides increase the rate of the polymerization process and improve the grafting efficiency in the case of IPS. More recently, multifunctional peroxides have also been used in order to obtain products with special molecular weight distributions. Up to relatively high conversions, the rate of polymerization can be satisfactorily represented as a first-order reaction [5] according to the following equation: -d[M]/df = rp = *p[P'][M] where rp= polymerization rate, kp= rate constant, [P*] = stationary radical concentration, and [M] = monomer concentration. This equation is valid under the assumption that the number of polymerizing monomer units is much larger than the monomer consumption which is needed for the initiation. A set of kp
+s
PS
Figure 2.2 Reactions during styrene oligomer formation in the chain initiation phase
28
N. NIESSNER AND H. GAUSEPOHL
values is listed in Ref. 6. A state-of-the-art method is pulsed laser initiated polymerization [7,8]. The polymerization reaction is terminated by disproportionation and recombination, with the ratio depending mainly on the temperature. Chain termination is represented by
where kt is the termination coefficient, values also being listed in Ref. 6. The mode of termination determines the dispersion index D: D = Pw/Pn = 2 for disproportionation and
for recombination, where Pw, and Pn are the weight- and number-average degree of polymerization, respectively. In addition to termination, transfer reactions to all other components of the polymerization occur. These transfer reactions do not affect the polymerization rate. However, they diminish the average molecular weight. The equation for transfer reactions is Rtr = Ktr[P*][TH]
where kp = transfer rate constant, and [TH] = concentration of transfer agent. The main disadvantage of radical polymerization reactions is their low selectivity, i.e. neither the molecular weight, the molecular weight distribution nor the molecular structure can be precisely controlled. In the literature, many experiments have been described to overcome this drawback by applying so called iniferters (initiator—transfer agent—termination). These substances act as initiators, transfer agents and termination agents. Typical examples for controllers are nitroxyl compounds such as 2,2,6,6-tetramethyl-l-piperidinyl oxide (TEMPO) and 2,2, 5,5-tetramethyl-l-pyrrolidinyl oxide (PROXYL) [9,10], thiocarbamate [11] and tetraphenylethane derivatives [12]. All these substances react with the growing macromolecular radicals by forming temporarily dormant species, minimizing termination by recombination or disproportionation [13,14]. The introduction of a dynamic equilibrium between dormant and active species leads to a low stationary concentration of free radicals and a relatively high concentration of reactive polymer chains. This results in a low overall polymerization rate. By this method, ordered copolymers from styrene and butadiene were synthesized with only limited success [15] owing to the formation of homopolymers.
POLYSTYRENES AND STYRENE COPOLYMERS
3
29
PROCESSES
Polystyrene was first commercially produced by BASF in 1931. The inventors of this mass polymerization process had to resolve two conflicting problems: heat removal from high-viscosity melts and development of an appropriate workup scheme. Heat removal was accomplished by heat-exchange tubes in the polymer melt. Polystyrene was obtained after exit of the melt from a hightemperature polymerization zone, using an extrusion screw. Today's processes are characterized by a fully continuous polymerization with heat removal by evaporation of styrene and solvent. The main advantage - consistent temperature — results in a high product quality with a narrow molecular weight distribution (Mw/Mn = 2.2–2.4) and high transparency. Impact-modified polystyrene is mainly produced by mass polymerization, either in tower cascades or tank/tower cascades. In the latter case, particle size and morphology can be defined by variation of the viscosity ratio between the continuous and the discontinuous phases, the stirrer velocity, the molecular weight of the polybutadiene rubber and the amount of rubber. Typical particles sizes are 2–20 u,m, this being the optimum for effectively dissipating impact energy. Styrene copolymers such as SAN and ABS are basically produced according to GPPS and IPS technology. However, effective impact modification in ABS is accomplished with smaller particles. Particles around or smaller than 1 fxm can also be produced by emulsion technology and thus emulsion polymerized ABS dominates by far in today's global ABS markets. In order to guarantee optimum impact modification, the rubber must be adapted to the base polymer (matrix). This is done by grafting with monomers which are miscible with the matrix. Thus, for example, the polybutadiene rubber for ABS is grafted with a mixture of styrene and acrylonitrile. The graft rubbers produced in this way consist of a flexible rubber core surrounded by a graft shell which provides linkages to the matrix in question (Figure 2.3). This process is performed in emulsion. The particles of rubber are finely dispersed in the rigid phase, i.e. the impact-modified styrenic copolymers thus have a multiphase structure. Figure 2.4 gives a brief overview of mass and emulsion polymerization processes.
4 STRUCTURE AND MORPHOLOGIES Commercial polystyrene manufacturing techniques are based either on a suspension process if the material is to be foamed or on a bulk polymerization process for GPPS and IPS. ABS-type polymers can also be produced via emulsion polymerization. Figure 2.5 shows the differences in emulsion and mass polymerization processes and the resulting morphology. Typically,
30
N. NIESSNER AND H. GAUSEPOHL
Elastomer core (rubber) Graft envelope Matrix Modification of properties due to • the chemical nature and quantity of the rubber • the structure morphology of the rubber • the particle size and particle size distribution
Figure 2.3 particles
Typical impact modification of styrene copolymer via emulsion core/shell
Mass ABS
Emulsion ABS Butadiene
H,O
S/AN Initiator
S/AN
Degassing 1r
H,O
Precipitation /Drying
Extruder
h
(residual
ABS
Features
Features Graft shell can be produced in a controlled manner independently from matrix polymer
•
High rubber content possible
•
High rubber efficiency
Monodisperse particles with defined particle size
•
Special HI PS-like morphology
Figure 2.4
Only few process steps
• Light color
Differences between mass and emulsion polymerization processes
POLYSTYRENES AND STYRENE COPOLYMERS Emulsion Polymerization Glossy surface
Figure 2.5
31
Mass Polymerization Lower Gloss Light and constant color
Differences between mass and emulsion polymerization: morphology
particles with diameters in the range of the wavelength of visible light can be produced by emulsion polymerization. As a consequence, glossy products with high-quality surface appearance are based on this technique. Larger particles often with inclusions of matrix polymer - result from mass polymerization. These morphologies typically cause a less glossy surface appearance, and are characterized by a high rubber efficiency, i.e. impact strength per unit amount of rubber. Rubber particles dissipate impact energy only if they can effectively initiate and terminate 'crazes'. By this 'crazing', energy is transformed into deformation of rubber particles, eventually accompanied by the formation of voids in the rubber particle itself. Deformation, however, initiates crazes, that can effectively be stopped by other rubber particles (Figure 2.6). A typical electromicrograph shows the formation of these micro-cracks after absorption of impact energy, and further magnification reveals the high inner surface caused by the formation of fibrils perpendicular of the craze direction (Figure 2.7). The predominant fracture mechanism in polymers with low polarity (such as IPS or ABS with low acrylonitrile content) is the craze mechanism (with an optimum particle size of approximately 2–6 |xm). With increasing polarity, the dissipation of energy by formation of shear bands ('shear yielding') becomes more pronounced. This second mechanism is facilitated by very small particles in the sub–100 nm range. As a rule of thumb, it is accepted that - starting with IPS an increase in the acrylonitrile content in rubber-modified styrene polymers causes a shift towards lower optimal particle size. For this reason, mass polymerized ABS typically has significantly smaller cellular particles than mass polymerized IPS. Nevertheless, anionic polymerization will be the pacemaking tool for the synthesis of innovative materials which shift the properties of the polystyrene
32
Figure 2.6
N. NIESSNER AND H. GAUSEPOHL
Formation of crazes after mechanical impact
Figure 2.7 TEM picture of crazes in ABS
family into the domain of thermoplastic elastomers and engineering resins [16]. The reason for this is that only anionic polymerization permits the control of the molecular architecture of the molecules, allowing the preparation of tailormade polymeric materials. Figure 2.8 gives an overview on the variety of the structural possibilities.
33
POLYSTYRENES AND STYRENE COPOLYMERS molecular composition 2-block
to
block and tapered block transition
multiblock D-M-D-i-0 D-1-D-l-D-i statistical to alternating monomer insertion
molecular structure linear
branched
star
comb
functionality monofunctionality
bifunctionality
molecular weight distribution mono-modal mono-modal narrow distribution broad distribution Figure 2.8
multimodal
Control of the molecular design by anionic polymerization
The molecular weight can be controlled by the ratio of the initator to the monomer, the molecular weight distribution by the type of polymerization (discontinuous or continuous) and the modality by single or multiple initiation. Sequential addition of different monomers leads to block copolymers with sharp or tapered transitions. In the presence of Lewis bases, statistical or alternating copolymers can be obtained. The molecular structure can be selected between linear, branched, star- and comb-like or dendritic molecules. Finally, functionalities can be introduced for changing the polarity of the resulting product or for extending the reaction profile. Styrene—butadiene block copolymers such as Kraton™, K-Resin™, Styrolux® and Styroflex® are examples of the versatility of this method and have long been manufactured via anionic polymerization on a commercial scale.
34
N. NIESSNER AND H. GAUSEPOHL
Most recently, block copolymers of the type ABC with completely new and previously unknown morphologies were synthesized by Stadler et al. [16]. By molecular self-assembly he obtained products with supramolecular structures and new property profiles, e.g. the block copolymer with 24% polystyrene, 7% polybutadiene and 69% poly(methyl methacrylate) (PMMA) exhibits higher toughness and higher tensile strength than PMMA. TEM pictures reveal polystyrene cylinders surrounded by a helical string of polybutadiene and embedded in a PMMA matrix (Figure 2.9a). This type of block copolymer but with a higher proportion of polybutadiene is very suitable also as elastomers because these products exhibit high tensile strength compared with classical ABA systems (Figure 2.9b). In order to upgrade polystyrene towards an engineering plastic material such as ABS, the long-term service temperature needs to be raised above 100 CC. Consequently, the polymer chain must be stiffened by radical copolymerization of styrene with a variety of ring-substituted styrenes. However, in these cases the influence on the glass transition temperature of the polymer is not very strong, but by using anionic polymerization technology a new polymer class with 1,1-diphenylethylene as comonomer could be introduced. It exhibits an overall advanced performance compared with GPPS without sacrificing the typical polystyrene property profile. The incorporation of additional phenyl rings into the polystyrene backbone results in a stiffened polymer chain. Conventional, butyllithium-initiated polymerization faces a major drawback. Owing to the living nature of anionic polymerization, all chains grow at the same time. As a consequence, turnover rates are exceedingly fast at high monomer concentrations, and removal of the heat of polymerization is difficult. Hence anionic polymerization has been restricted to dilute solutions and low temperatures. Under these conditions, polystyrene cannot be produced economically. However, with certain Lewis acids such as dibutylmagnesium this problem could be solved [18,19].
5
PROPERTIES
Commercial polystyrene is an amorphous material with a molecular weight Mw = 100000–400000. Low specific gravity, transparency and brilliance, absence of colour, low shrinkage and ease of fabrication are its most characteristic features. At temperatures sufficiently below the glass transition temperature (Tg) and at low deformations the material obeys Hooke's law of elasticity under external stress. The tensile strength and flexural strength are strongly dependent on the molecular weight of the polymer. Below a certain molecular weight (Mw « 120000) the material fails completely, whilst at higher values the tensile strength and the elongation at break increase until an upper limit is reached. At
POLYSTYRENES AND STYRENE COPOLYMERS
Figure 2.9
35
Morphology of special ABC block copolymer. Courtesy of BASF
higher impact and deformation rates, polystyrene tries to withstand the external stress by forming crazes, thus dissipating the impact energy over a broader area of the specimen. Crazes are voids which are crossed by orientated fibrilles.
36
N. NIESSNER AND H. GAUSEPOHL
They are the precursors of cracks and mainly determine the tensile strength of the material. The reason for the formation of crazes lies in the insufficient segment mobility of the molecular chain. Rubber-modified polystyrene exhibits a much higher toughness than crystal polystyrene because small dispersed rubber particles enhance the stress concentration and the larger ones stop their growth, thus preventing crazes from developing into cracks. The prerequisite for this mechanism is a well defined adhesion of the rubber particles to the matrix, which in turn depends on the grafting efficiency of the rubber [20,21]. Table 2.1 shows some important properties of polystyrene. Above its Tg, polystyrene is a viscoelastic melt. It is called viscoelastic because the polymeric material displays both a viscous and an elastic response to shear stress, depending on the rate and the temperature of the test. Two main factors influence the viscous and the elastic behaviour of the product, namely its molecular weight and the molecular weight distribution. For mondisperse polystyrenes the zero-state viscosity is proportional to M3.4. The ratio Mw/Mn has only a minor influence on the viscosity, but it greatly influences the steadystate shear compliance, i.e. small amounts of high molecular weight polystyrene blended into a medium molecular weight polystyrene enhance the elasticity and melt strength. Details concerning the rheology of polystyrene are given elsewhere [22,23]. The development of SAN was triggered by the idea of building a polar comonomer into polystyrene to improve its resistance to chemicals and to stress cracking. The relatively polar acrylonitrile presented itself as a suitable comonomer in this case. Styrene—acrylonitrile copolymers are further characterized by high rigidity and thermal shock resistance. Two parameters substantially determine the properties of SAN: molecular weight and the proportion of acrylonitrile. On account of its combination of properties, SAN is principally employed in a wide range of household articles. Other examples include transparent parts for kitchen appliances and coffee machines. In the field of household goods, its high stability in dishwasher conditions is the principal argument in favor of SAN. Other areas of application are packaging for cosmetics, toothbrushes and lamp covers. Relatively large industrial parts such as large industrial batteries (Figure 2.10) are a remarkable field of application, since the housings weigh up to 14kg. Fitted with lead electrodes and filled with sulfuric acid they have a service life of more than 10 years, which is a highly impressive testimony to SAN's durability and resistance to chemicals.
Table 2.1
Properties of atactic polystyrene
Property
Unit
Test method
Value (GPPS)a
Value (IPS)b
Density Refraction index Coefficient of expansion:
g/cm3 —
ISO 1183 ISO 489
1.050 1.590
1.050 —
cm3 /cm3 K cm3 /cm3 K
— — ISO 75-2
2.805 x 10–4 5.650 x 10–4 100 86
— — 73-85
ISO 75-2
98
81-96
W/mK — — Om MPa MPa %
ISO 306 DIN 52 612 IEC 60 250 IEC 60 250 IEC 6093 ISO 527-2 ISO 527-2 ISO 527-2
101 0.17 2.5 / 2.5 0.9 x 10–4/0.7 x 10–4 >1014 3300 42-59 1.5-3.0
91-101 0.17 2.5/2.5 1.0 x 10 - 4 /4x 10-4 >1014 1400–2800 21-43 1.4–2.0
kJ/m2 kJ/m2
ISO 179/leU ISO 179/leU
10-28
70-NBC 50–160
Tg
Glass transission temperature (Tg) Heat deflection temperature under 1.8MPa load(HDT/A) Heat deflection temperature under 0.45 MPa load (HDT/B) Vicat softening point A50 Thermal conductivity Dielectric constant at 100 Hz/1 MHz Dissipation factor at 100Hz/l MHz Volume resistivity Tensile modulus of elasticity Tensile stress at yield Elongation at yield Charpy impact strength: At +23 °C At-30°C GPPS: general-purpose polystyrene. IPS: impact polystyrene. NB: no break.
°c °c °c O/"~l
2
-4
38
N. NIESSNER AND H. GAUSEPOHL
Figure 2.10
6
Industry batteries made from SAN
PROPERTIES, RANGE AND APPLICATIONS OF MABS PRODUCTS
While amorphous, single-phase materials such as SAN or PMMA are transparent, rubber-modified, multiphase materials are opaque as a general rule. The reason for the loss of transparency is the scattering of light by the particles of rubber dispersed in the amorphous matrix. However, as with every rule there
POLYSTYRENES AND STYRENE COPOLYMERS
39
are also exceptions here. In the case of MABS the following concept is used: the rubber for MABS is built up in such a way that it has exactly the same refractive index as the matrix. In this way the rubber is rendered 'invisible', the light is not scattered and the material remains transparent (Figure 2.11). MABS is being increasingly used also for transparent parts for computer housings and household appliances (Figure 2.12), offering a high degree of freedom to accomplish even the most sophisticated designs. Graft rubbers based on polybutadiene (ABS) or acrylate esters (ASA) are generally used for the impact modification of styrene copolymers. Depending on the rubber component used, special features emerge in the final properties (Figure 2.13). Polybutadiene yields very high levels of toughness even at low temperatures but it contains numerous double bonds which can be relatively easily attacked by UV radiation and oxygen. On exposure to outdoor conditions or under the action of heat this attack results in yellowing and embrittlement of ABS even after a short time. Acrylate rubbers which are employed in ASA contain no double bonds. For that reason, ASA is substantially more resistant to weathering than ABS. Owing to the polar acrylate component, ASA is also more resistant to stress cracking than ABS. ABS in turn has significant advantages in low-temperature impact resistance on account of the very low glass transition temperature of the polybutadiene rubber.
Figure 2.11 Transparent impact-modified styrene copolymers by isorefractive phases
40
N. NIESSNER AND H. GAUSEPOHL
Figure 2.12 Transparent plastics (including styrene copolymers) for state-of-the-art household equipment
Figure 2.13
Differences between ASA and ABS polymers
POLYSTYRENES AND STYRENE COPOLYMERS
41
Figure 2.14 Electomicrographs of a cross-section perpendicular to the surface of ABS and ASA after 500 h of UV exposure (Xenotest, ISO 4892-2A) and subsequent treatment with soap
On account of its acrylate rubber content, ASA is superior to ABS with regard to its weathering resistance and thermostability. In outdoor applications ASA yellows to a significantly lower degree and also retains its impact resistance over a substantially longer period of time. The tendency to yellowing in both ABS and ASA can be reduced further by the addition of UV stabilizers. However, even stabilized ABS does not reach the level of unstabilized ASA (Figure 2.14).
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
Ostromislensky, I., US Patent 1613673 (1927). Mayo, F. R., J. Am. Chem. Soc., 15 (1953) 6133. Hiatt, P. D. and Bartelt, J., J. Am. Chem. Soc., 81 (1959) 1149. Buzanowski, W. C, Graham, J. D., Priddy, D. B. and Shero, E., Polymer, 33 (1992) 3055. Husain, A. and Hamielec, A. E., J. Appl. Polym. Sci, 22 (1978) 1207. Berger, K. C. and Meyerhoff, G. in Polymer Handbook, Vol. 3, Wiley, New York (1989). Olaj, O. F., Schnoll-Bitaj, I. and Hinkelmann, F., Makromol. Chem., 188 (1987) 1689. Deady, M., Mau, A. W. H., Moad, G. and Spurling, T. H., Makromol. Chem., 194 (1993) 1691. Georges, M. K., Veregin, R. P. N., Kazmaier, P. M. and Hamer, G. K., Macromolecules, 26(1993)2987. Brinkmann-Rengel, S. and Niessner, N., ACS Symp. Ser., 768 (2000) 394. Otsu, T., Yamashita, K. and Tsuda, K., Macromolecules, 19 (1987) 287.
42
N. NIESSNER AND H. GAUSEPOHL
12. 13. 14. 15.
Bledzki, A. and Braun, D., Polym. Bull., 16 (1986) 19. Webster, O. E., Science, 251 (1991) 887. Gretza, M. K., Madereu, D. and Matyjaszewski, K., Macromolecules, 27 (1994) 638. Georges, M. K., Veregin, R. P. N., Kazmaier, P. M. and Hamer, G. K., Polym. Prepr., 35 (1994) 582. Knoll, K. and Niessner, N., ACS Symp. Ser., 696 (1998) 112. Karppe, U., Stadler, R. and Voigt-Martin, I., Macromolecules, 28 (1995) 4558. Desbois, P., Fontanille, M., Deffieux, A., Warzelhan, V., Latsch, C. and Schade, C. H., Macromol. Chem. Phys., 200 (1999) 621. Ebara, K., Tanji, S. and Sawamoto, M., Patent WO/97 33923 (1997). Bucknall, C. B., Toughned Plastics, Applied Science Publishers, London (1977). Kramer, E. J. and Krauch, H. H., Crazing in Polymers, Springer, Berlin (1983). Ferry, J.D., Viscoelastic Properties of Polymers, Wiley, New York, 3rd edn, 1980. Carreau, P. J., Rheology of Polymeric Systems, Hanser, Munich (1997).
16. 17. 18. 19. 20. 21. 22. 23.
Preparation of Styrenic Polymers
This page intentionally left blank
BERNARD J. MEISTER AND CLARK J. CUMMINGS The Dow Chemical Company, Midland, Ml, USA
1
INTRODUCTION
Styrene is one of the easiest monomers to polymerize and of course this led to the early discovery and commercialization of polystyrene. Solid polystyrene was probably first prepared in 1845 [1] by heating the monomer in air. At first the reaction was considered to be an oxidation until Staudinger [2] first proposed the long-chain structure of polymer molecules. This proposal and a host of ensuing studies to understand the chemistry of polymerization led to commercial interest in producing this new type of material. The Dow Chemical Company had a major effort in development of a commercial process for polystyrene in the 1930s leading to the can process documented by Boyer [3]. In the same time frame, I. G. Farben in Germany developed the well known continuous tower process [4]. When the details of the I. G. Farben process became known to the US chemical companies after World War II, effort was placed on developing continuous processes due to the much lower labor costs and improved uniformity of the product. The reactors for these continuous processes took many forms, from the initial towers and tube tanks to agitated versions to stirred tank reactors with ebullient cooling, pipelines filled with static mixers and recirculated tube banks. Over the same time period, batch stirred tank reactors utilizing solution, suspension and emulsion polymerization were also developed. Because most of these processes did not go to complete conversion, devolatilization processes were developed that continue to evolve because of taste, odor and health related concerns with residual monomers and solvents. For the practitioner, the choice among all Modern Styrenic Polymers: Polystyrene and Styrenic Copolymers. Edited by J. Scheirs and D. B. Priddy >( 2003 John Wiley & Sons Ltd
46
B. J. MEISTER AND C. J. CUMMINGS
these possibilities depend somewhat on the technical base of the individual company (knowledge, patents and installed capital) and on the competing desires of product property range, product quality and uniformity, low manufacturing cost including maintenance and any copolymers or composites such as rubber-modified polymers that are included in the product mix. There are a number of prior works that tackle this same general subject [5–10]. The most important of these is the detailed review by Simon and Chappelear [5]-
2 TECHNICAL CONSTRAINTS THAT INFLUENCE REACTOR SELECTION 2.1
TEMPERATURE CONTROL
The first constraint that the engineer faces when the task is to develop a commercial process, is that a large amount of heat is generated when one converts styrene to polystyrene. The heat of polymerization is approximately 300 BTU/lb (700 J/g) at 100 °C and decreases with increasing temperature. This leads to a temperature increase of approximately 350°C for pure styrene if it is polymerized to completion and no heat is removed from the process. This alone does not rule out an adiabatic process; however, polystyrene degrades rapidly at temperatures above 250 °C and the molecular weight of the polystyrene produced decreases rapidly with temperature (see Figure 3.1). Therefore, no practical products could be produced with an adiabatic process without at least 50 % diluent to limit the temperature rise to 175 °C. The chemist in the laboratory does not have to deal with the heat generation issue. Ten grams of styrene are sealed in an ampoule and placed in a bath at the reaction temperature and the heat transfer is sufficient to maintain the temperature of the reactants under most conditions. The Dow can process [3] mentioned in the Introduction was a direct scaleup of the chemists ampoule to 10 gallon cans. There was a considerable rise in temperature in the middle of the 10 gallon can leading to high molecular weight on the outside and low molecular weight on the inside that was then ground up and blended. This lack of temperature control severely limits the product mix. A natural evolution from the batch bulk process was to the batch suspension process where water as the continuous phase supplies a large heat sink that allows control of the temperature of the polymerizing mass, yielding control of the molecular weight and the molecular weight distribution of the polystyrene. The alternative path of evolution was to the continuous solution process, first demonstrated with the tower process by I. G. Farben and implemented by Dow and others as either towers or tanks filled with heat transfer tubes. These
COMMERCIAL PROCESSES FOR THE MANUFACTURE OF POLYSTYRENE 47 50
20 10
0.50
0.20 0.15 2X10 6 1X10 6 5X10 5 2X10 5 1Xl05 Number-average molecular weight
Figure 3.1 Relationship between initial rate (wt%/h) of styrene polymerization and molecular weight at various temperatures. Reproduced with permission from Encyclopedia of Polymer Science and Engineering, Mark (Ed.), John Wiley & Sons, NY. Copyright John Wiley & Sons
towers or tanks were filled with as many heat transfer tubes as possible but they were severely limited in the rate of polymerization that they could control. The tube tanks that were used by Dow were limited to conversion rates of less than 5 %/h and therefore about 15 h of residence time. In order to produce 4000 kg/h of product at 75% conversion, 80000kg of reactor inventory were required. This large amount of inventory leads to slow product changeovers and large amounts of off-specification material. In addition, high solids material would collect on the tubes, leading to decreasing heat transfer with time so that periodic cleanouts were required. Because of the rate limitations of the tower and tube-tank processes that were primarily heat transfer constraints, further developments in the continuous solution process for crystal polystyrene (GP) were aimed at improving heat transfer. One obvious solution was to incorporate agitation of some type in the reactor. Although at Dow the incorporation of agitation in the reactors came about with the development of rubber-modified polystyrene [11], and this aspect will be discussed in a later section, agitation also significantly raises the heat transfer
48
B. J. MEISTER AND C. J. CUMMINGS
coefficient and allows a threefold increase in polymerization rate. This is extremely important in the economics of the process because the in-process inventory and the offgrade during transitions can be reduced by a factor of three. This also means less time in the finishing lines for any degradation to take place. The second major approach to improving heat transfer in the continuous solution process is boiling heat transfer. The first report of the utilization of an ebullient stirred tank reactor to produce polystyrene was a patent assigned to Union Carbide in 1950 [12]. The implications of a well mixed reactor on the product will be discussed in the section on mixing. An alternative approach to the utilization of boiling heat transfer is embodied in a patent assigned to Monsanto [13]. In this approach, the reactor is horizontal with reactor internals that segregate the reactor into a series of chambers that have a common vapor space. As the boiling point of a polystyrene solution increases with the concentration of polystyrene, the temperature in the chambers increases with conversion. This allows a temperature profile that is not possible in the single well mixed reactor. The use of boiling heat transfer raises the maximum conversion rates that can be controlled significantly beyond that of the agitated towers filled with heat transfer tubes. The main limitation that occurs is the removal of the vapor bubbles from the polymer solution. As the viscosity of the polystyrene solution increases rapidly with conversion, this becomes most limiting when the viscosity of the polystyrene solution exceeds 1000P (dPa/s). Below this viscosity, conversion rates of 40%/h can be controlled, but above this viscosity, the polymer mass foams up into the condenser and temperature control is lost, so that the maximum conversion rate decreases rapidly at high polymer concentrations.
2.2
CHEMISTRY-RELATED CONSTRAINTS
Thus far, all of the discussion has been related to the production of polystyrene that is initiated thermally. The main reactions that one must be concerned with that determine the rate of reaction and the molecular weight of the polymer chains that are produced are initiation, propagation, termination and chain transfer. The thermal initiation reaction has a high activation energy with the number of chains initiated per unit time increasing by a factor of seven on raising the reaction temperature from 120 to 140CC. The rate of reaction increases roughly by a factor of four so that the average chain molecular weight decreases by a factor of 1.75 over this same temperature range. The engineering consequence of this is that low molecular weight polystyrene can be produced at a conversion rate much greater than the high molecular weight polystyrene. If the same reactor is to be used to produce a molding resin of molecular weight 150000 and a high-grade extrusion resin of molecular weight 350000, then the
COMMERCIAL PROCESSES FOR THE MANUFACTURE OF POLYSTYRENE
49
molding resin could be produced at 140 °C at a rate of 40%/h and the extrusion resin could be produced at 110 °C at a rate of only 5%/h. Fortunately, this is about the full range of commercial products. However, it leaves the design engineer specifying the size of equipment with a difficult choice. This is what was meant when we titled this section 'chemistry-related constraints' as the reactor design is heavily influenced by inherent chemistry and the product mix. One possible solution to the product mix problem is to design the production unit to operate at the product rates of the highest molecular weight product and then use a chain transfer agent such as a thiol or high levels of a solvent such as ethylbenzene to produce the low molecular weight product. At this point we also need to mention the molecular weight distribution of the product. The lifetime of a growing polystyrene chain is fairly short. This can be calculated from T = DP/kp(M] where kp is the propagation rate constant (1420L/mol s at 140°C), DP is the number-average degree of polymerization (1200 at 140°C) and [M] is the monomer concentration in mol/L(7.68 mol/L at 140°C) . Therefore, at 140°C the lifetime of the growing chain is only 0.1 1 s. If a reactor in batch is operated at a time-temperature profile or a series of continuous reactors are operated at a temperature sequence, some chains will be produced at each temperature and a distribution of molecular weights will be obtained. The normal method of representing this distribution is to use the ratio of the weight-average molecular weight, Mw, to the number-average molecular weight, Mn, where
and
where ni is the number of chains of molecular weight Mi. Even at a fixed temperature, a distribution of molecular weights is produced because of the random nature of the chain termination process. If the chains all terminate by chain transfer, the ratio Mw/Mn is approximately 2.0, and if they all terminate by coupling, the ratio Mw/Mn is approximately 1.50. Under conditions where thermal polystyrene is produced, the ratio is approximately 1.90 because most chains are terminated by chain transfer to monomer, solvent and other agents. The use of the particular product may dictate an optimum molecular weight distribution. If the product is used to make an extruded foam, a relatively narrow distribution may be desirable to retain orientation and strength in the
50
B. J. MEISTER AND C. J. CUMMINGS
cell walls. However, a molding product might require a broad distribution because of the shear sensitivity and easier processing that results. The main point here being that the ability to adjust the molecular weight distribution is a desirable if not essential feature of a commercial reactor to produce polystyrene. This generally means the capability to react at sequentially higher temperatures or to add fast-acting chain transfer agents during the process is an important feature. One method commonly utilized commercially to surmount the chemistry constraints of thermally initiated polystyrene is the use of free radical initiators of either the peroxide or azo type. Initiators that have 1 h half-lives slightly below the thermal temperatures are most effective so that initiators with 1 h half-lives in the range 100–140°C are most often used. The utilization of chemical initiators under conditions where the thermal initiation reaction is suppressed generally leads to a boost in the rate at which a given molecular weight polystyrene can be produced. The main reason for this is that the Diels– Alder reactive dimer formed in the thermal initiation step is a chain transfer agent and plays a significant role in controlling the molecular weight. When the thermal initiation reaction is suppressed by lowering temperature or monomer concentration, but the radical concentration is maintained with an initiator, the resulting polystyrene has a higher molecular weight. During the 1980s, bifunctional initiators became widely used among the commercial producers of polystyrene. The reason for this is a larger boost in the rate-molecular weight relationship because a difunctional initiator with two peroxide groups in the same molecule can lead to two chains of average length DP and one chain of average length 2DP. Figure 3.2, taken from the review by Priddy [14], shows an illustration of the significance of this effect for a particular case. Each individual initiator will have somewhat different curves based on the half-lives and efficiency of generating free radicals of the specific monofunctional or bifunctional peroxide. Certainly, one can think of possible methods to extend this approach such as polymeric or cyclic peroxides that yield only diradicals or multifunctional peroxides that yield branched polystyrene. We do not know if any of these concepts are utilized commercially. There are other ways of chemically shifting the rate-molecular weight relationship. One of these is to add a small amount of divinylbenzene to the styrene, which leads to the incorporation of branch points in the polymer chain. This works well in a well mixed batch reactor in the laboratory but often leads to gels and crosslinked polymer in large-scale continuous reactors. A second approach investigated by Priddy [14] is the use of an acid to suppress thermal initiation. This results in a shift in the ratio of propagation to initiation that results in increased molecular weight. This also has not made it to the commercial realm. The final method of shifting free radical kinetics is emulsion polymerization. Unlike suspension polymerization where the reaction kinetics are identical with those of mass polymerization, in emulsion polymerization the droplets are small
COMMERCIAL PROCESSES FOR THE MANUFACTURE OF POLYSTYRENE
51
350
300
Spontaneous Monofunctional ' initiator
250
Difunctional initiator
200 10
20 30 40 50 Polymerization rate, %/h
60
Figure 3.2 Polymerization rate advantage of difunctional initiator. Reproduced with permission from Encyclopedia of Chemical Technology, Kirk (Ed.), John Wiley & Sons, NY (1997). Copyright John Wiley & Sons
enough so that only one polymer chain is growing within the droplet at a time. This essentially eliminates the termination mechanisms that involve two active radicals and therefore allows the production of higher molecular weight polystyrene at a given temperature. Back in the 1960s this was a competitive way to produce high molecular weight crystal polystyrene, but it is now used only for copolymers intended for latex applications and for small particle ABS. The second major way to alter kinetics is to utilize ionic polymerization instead of free radical polymerization. Styrene polymerization can proceed through a positively charged species (cationic polymerization) or a negatively charged species (anionic polymerization). These polymerizations are very sensitive to impurities so that extensive pretreatment of the monomer is required. Even with pretreatment, there is much chain transfer taking place during cationic polymerization so that the molecular weight is low and molecular weight distributions are similar to free radical polymerizations. Anionic polymerization, however, can be used to produce high molecular weight narrow distribution polystyrene. If all the chains are initiated at the same time and the temperature is kept low to minimize chain transfer, molecular weight distributions very close to monodisperse can be produced. The commercial uses of these polymers seem to be limited to instrument calibrations and laboratory studies of the effects of molecular weight on rheology and physical properties. However, anionic polymerization as a potential commercial method for producing polystyrene has been extensively studied by Dow and others. The potential for high polymerization rates, complete conversion of
52
B. J. MEISTER AND C. J. CUMMINGS
styrene, low oligomers, and the possibility of tailoring molecular weight distributions over a wider range was the driver behind this work. Much of this effort has been focused on backmixed reactors where the rate of addition of monomer could be used to control the rate of reaction [15]. Although anionic polymerization has been shown to be a serious contender for the production of crystal polystyrene, no commercial production facilities have been built. All of the above and later sections will deal only with atactic polystyrene. Styrene monomer can also be polymerized to stereoregular structure through the use of coordination catalysts. Until 1986, only isotactic (phenyl rings cis to each other) had been produced. Isotactic polystyrene crystallizes slowly, has a melting point of 230 °C and has found no commercial uses. Since 1986, considerable industrial research and development has been focused on syndiotactic polystyrene (phenyl rings trans to each other). This material crystallizes more rapidly and has a higher melting point of 270 °C. However, because it crystallizes during polymerization, processing is more complicated than for atactic polystyrene, and the specifics are best reviewed in a discussion of the manufacture of crystalline polymers.
2.3
CONSTRAINTS DUE TO REACTOR MIXING
An important characteristic of a polymerization reactor is the degree of mixing. Mixing is important to have control of the temperature distribution and the composition distribution in the reactor. In most cases, we can separate into two components of mixing, axial mixing in the direction of flow and radial mixing across the reactor perpendicular to flow. In a batch reactor, one might change this separation to mixing parallel to and perpendicular to the agitator shaft. The degree of axial mixing can vary from effectively zero in a perfect plug flow reactor to infinite in a perfectly backmixed reactor. In the perfect plug flow reactor with perfect radial mixing, every fluid element that enters the reactor at time t leaves the reactor at time (t + 0), where 9 is the residence time defined by the reactor volume divided by the flow rate. This could be accomplished by connecting a large number of stirred tanks in series or possibly using a pipeline filled with effective static mixers. However, if an empty pipe is used, because of the viscosities involved, laminar flow will take place, the velocity profile will be parabolic and material near the wall will have a much longer residence time than material in the center and will polymerize to much higher conversion. This was learned in a dramatic fashion in the early 1960s when a 1 inch diameter pilot pipeline was scaled up to commercial scale. The commercial unit operated for less than a day before it progressively filled with solid polymer with monomer squirting down the middle. In the 1 inch diameter pipe, radial diffusion was sufficient to accomplish elimination of the radial concentration gradient. In the large-scale pipeline, this was not the case. Once
COMMERCIAL PROCESSES FOR THE MANUFACTURE OF POLYSTYRENE
53
a wall layer has built up, it continues to grow because monomer diffuses into the layer faster than polymer diffuses out. This monomer polymerizes and adds to the layer. If any branching reactions can take place, the effect is amplified, because the molecular weight of existing polymer in the layer increases with time. This general phenomenon of buildup of solid polymer in static areas of the reactor is common to all continuous solution reactors. A major component of the design is to eliminate regions where this may occur. This is one of the drivers for incorporating agitators in the tower reactors which will be discussed in the next section and has also been a driver for removing heat transfer tubes from the reactor which add wall surface. The opposite of the large diameter pipeline with little axial or radial mixing is the perfect backmixed reactor with instantaneous mixing and uniformity. For polystyrene reactors with several hours of residence time, complete mixing in l–2min is usually adequate to satisfy a practical definition of perfectly mixed. The probability of exit of any fluid element from this type of reactor is independent of when it entered. The residence time distribution is exponential and the molecular weight distribution in the case of no termination is Mw/Mn = 2.0, which will spread out to 2.3 when chain transfer controls. If product requirements necessitate a narrower residence time distribution, one can utilize several of these reactors in series. This becomes necessary to control the grafting distribution in rubber modified polystyrene. The choice of agitator for the stirred tank reactor depends primarily on the viscosity range of the polymer solution in the reactor, ranging from turbines on the low end to helical agitators that function more like extruders on the high end. The details of the agitators are specific to the individual companies involved. These reactors have little problem with wall buildup, particularly when the walls are not used for heat transfer. Cooling is primarily from the cold feed and by operating the reactor at the pressure at which the polymer solution boils. Because the termination reaction is strongly diffusion controlled, these reactors are susceptible to a two-phase solution, particularly when operated at low levels of solvent. Also, as the percentage conversion is increased, the viscosity increases and mixing becomes less and less perfect. As this occurs, temperature gradients within the reactor become more significant. Increased temperature at the same pressure means increased conversion and product uniformity may suffer. As with most polymer process work, detailed modeling, pilot plant range finding, and plant-scale trial and error lead to optimized conditions. Figure 3.3 summarizes mixing for the various types of continuous solution reactors. Eliminating the wall buildup prevalent in nonradial mixed reactors drives the reactor designer from left to right. The choice from top to bottom depends on the product requirements and the degree of temperature and composition dispersion needed to accomplish them, and the capital cost of multiple control zones in series.
54
B. J. MEISTER AND C. J. CUMMINGS Instantaneous perfect mixing
Radial mixing
.With internal mixing Higher agitation
.
Recirculation loop
D
C
Perfect CSTR
C
Recirculation
_
c Agitated tower
Unagitated tower Empty pipe
Axially segregated agitated reactor
CSTRs in series With internals Pipe with static mixers
' No mixing Figure 3.3 Solution process polystyrene reactors as a function of axial and radial mixing. Reproduced with permission from Encyclopedia of Polymer Science and Engineering, Mark (Ed.), John Wiley & Sons, NY. Copyright John Wiley & Sons
2.4
CONSTRAINTS RELATED TO THE RUBBER MODIFICATION OF POLYSTYRENE
Rubber is incorporated with polystyrene in commercial high-impact polystyrene (HIPS). The rubber ends up as domains in the size range 0.5-5.0 um with the range 1.0–2.0 um being the most effective at toughening the polystyrene. It was learned early that compounding rubber into finished polystyrene was relatively ineffective. There must be some connecting links between the rubber phase and the polystyrene phase. This is usually accomplished by grafting some of the polystyrene chains to the rubber during the polymerization. The current process for
COMMERCIAL PROCESSES FOR THE MANUFACTURE OF POLYSTYRENE
55
rubber-modified polystyrene evolved from early work at Dow [16] that demonstrated that if rubber is dissolved in the feed and the polymerization was carried out in tower-type reactors with sufficient agitation, the final polystyrene product contained small rubber domains that contributed significant toughness to the product. The initial process development work is detailed in a paper by Amos [11]. In the continuous solution process for rubber-modified polystyrene, the rubber is first dissolved in the styrene feed. This rubber solution is then fed to the reactor. A complex sequence of morphology development takes place in the polymer solution as the styrene polymerizes. The phase diagram in Figure 3.4 illustrates that as soon as point A a small amount of polystyrene is formed in a rubberized feed, illustrated here as 8 % rubber, and two phases are formed, one phase containing polystyrene and styrene and the other containing rubber and styrene. The tie lines in Figure 3.4 such as line B-C show that the rubber phase is more dilute than the polystyrene phase. The polystyrene phase exists as small domains in the continuous rubber phase. The polymerization proceeds along the line A-E with increasing amounts of the polystyrene phase suspended as small droplets in the rubber phase. When point F is reached, the volume of the polystyrene phase is equal to the volume of the rubber phase. With sufficient agitation at this point, phase inversion starts to take place and polystyrene Styrene 100 wt%
40 wt%
40 wt%
Polystyrene
Equal phase volume line
Rubber
Figure 3.4 Ternary phase diagram for the system styrene-polystyrene-polybutadiene rubber. Reproduced with permission from Encyclopedia of Polymer Science and Engineering, Mark (Ed.), John Wiley & Sons, NY. Copyright John Wiley & Sons
56
B. J. MEISTER AND C. J. CUMMINGS
gradually becomes the continuous phase. Soon after phase inversion is complete, the rubber domains become fixed in size and these particles maintain their boundaries as polymerization continues in both phases. A typical set of rubber particles formed in this sequence of events, that has been polymerized to relatively high conversion, is shown in Figure 3.5. This material has particles 2-3 um in diameter with large internal occlusions of polystyrene. These occlusions tend to grow larger the more polymerization takes place after the boundaries are set. The size of the rubber particles and the size of the occlusions within the particle are key parameters determining the final properties of HIPS, in addition to the amount of rubber and the crosslink density of the rubber phase. The variables that determine the ultimate morphology include the reactor agitation, the amount of backmixing, and the variables that affect the grafting and crosslinking reactions between the two phases. These include the type and molecular weight of the rubber, any peroxide that might be used as an initiator or grafting agent and the molecular weight of the polystyrene being produced. Echte [17] has demonstrated how the morphologies developed during polymerization relate to the structures of block copolymers in solution. When a polystyrene chain is grafted to a rubber chain, the polystyrene part is compatible with the polystyrene phase and the rubber part is compatible with the rubber phase. This molecule migrates to the interface between the phases. If the
Figure 3.5 Typical HIPS particles
COMMERCIAL PROCESSES FOR THE MANUFACTURE OF POLYSTYRENE
57
polystyrene part is much shorter than the rubber part, the molecule can conform better to a small polystyrene occlusion in the rubber phase. Similarly, if the rubber part is much smaller than the polystyrene part, the molecule can conform better to a small rubber domain in the polystyrene phase. This would be the exterior shell of the rubber particle. Turley and Keskkula [18] showed clearly how increasing agitation decreases the particle size, the amount of occluded polystyrene within the particle and the rubber phase volume fraction when graft and graft molecular weight are held constant. How does all this physical chemistry affect the reactor design? It becomes very important because many reactor types used for crystal polystyrene are unsuitable for rubbermodified polystyrene. If one attempted to use a rubberized feed in the early unagitated tower or tube tank reactors, a very disappointing result would be obtained as phase inversion would not take place. If one feeds a rubber solution to one of the more modern single-stage backmixed boiling reactors operating at high conversion used for GP polystyrene, another relatively undesirable result is obtained. Because the reactor is operating well past the phase inversion point, the rubberized feed is immediately broken up and the monomer diffuses out, leaving large dense particles. Some of the particles stay in the reactor a long time and become overgrafted and others leave the reactor rapidly with no graft, leading to a rubber-modified polystyrene with poor properties. Using chemistry that promotes high rapid grafting can surmount most of these problems [19]. However, if the typical highly occluded uniform particles are desired and backmixed reactors are to be utilized, then four reactors in series are usually required. The first reactor operates with rubber continuous to graft the rubber, the second reactor operates just above phase inversion to size the particles and two finishing reactors are used to build the occlusions. The alternative to this approach is the original Dow process. [16,20,21] Three agitated tower reactors are operated in series with phase inversion and particle sizing occurring in the first reactor, and occlusion building occurring in the second and third reactors which are agitated at a slower rate.
2.5
REACTOR REQUIREMENTS FOR PRODUCING
COPOLYMERS
The production of copolymers leads to some additional constraints to reactor design beyond what is required for homopolymer. The most important of these is composition drift. The reactivity ratios of a monomer mixture define the composition of a copolymer that is instantaneously produced from a given monomer mixture. This is true in a plug flow reactor or a backmixed reactor. However, in the plug flow reactor, the copolymer composition drifts from that produced from the initial monomer composition to that produced by the monomer composition at the end of the polymerization. In contrast, in the backmixed reactor, all copolymer produced is of the same composition, which
58
B. J. MEISTER AND C. J. CUMMINGS
is determined by the ratio of unreacted monomers in the reactor, not the feed composition. The largest volume copolymer of styrene is styrene-co-acrylonitrile (SAN). The reactivity ratios for this system with styrene defined as monomer 1 are approximately r\ = 0.40 and r2 = 0.04. This set of values makes this polymer subject to a high degree of composition drift if the shift is not compensated by reactor design. Composition drift is undesirable in SAN copolymers because changes of only a few weight percent acrylonitrile can make two copolymer chains of high molecular weight incompatible and lead to haze in the final product. This is particularly disadvantageous if the copolymer is being sold as a clear molding resin. This is of less concern if the SAN is being blended to produce rubbermodified polymer ABS. Fortunately, in the styrene-acrylonitrile system, there is an azeotropic composition where the equilibrium monomer composition and polymer composition have the same ratio. This is illustrated in Figure 3.6, which shows the azeotropic point to be 76 wt% styrene and 24 wt% acrylonitrile. Polymers produced in linear flow reactors or axially segregated reactors usually operate close to the azeotropic composition. If not, they must have many addition ports to add the monomer that is being depleted in order to prevent a large composition drift. Reactors specifically designed for copolymers are usually of the backmixed type. If ebullient heat transfer is utilized, much care must be taken with the handling of the vapor stream, which will be rich in acrylonitrile owing to its low boiling point. Once this is condensed, it must be returned to the reactor and mixed in rapidly, so as not to produce the high-percentage acrylonitrile polymer chains that would be incompatible with the main polymer being produced.
'
I
I
1 . 1
0 0.2 0.4 0.6 0.8 1.0 Mole fraction styrene in feed monomers
Figure 3.6 Relationship between feed composition and copolymer composition of syrene-acrylonitrile. Point A indicates the azeotropic composition. Reproduced with permission from Encyclopedia of Polymer Science and Engineering, Mark (Ed.), John Wiley & Sons, NY. Copyright John Wiley & Sons
COMMERCIAL PROCESSES FOR THE MANUFACTURE OF POLYSTYRENE 59 3 3.1
POLYSTYRENE DEVOLATILIZATION DEVOLATILIZATION
CONCEPTS
Because the reaction processes described previously do not take the reaction to completion, a separate unit operation is required to remove monomer(s) and solvent from the polymer product. This is typically completed by heating the polymer solution and flashing off the unwanted monomer and solvent. There are several concerns such as equilibrium levels, polymer degradation, and mass transfer that must be considered. When completing a flash devolatilization, it is usually desirable to heat the polymer as much as practical to ensure the best separation. However, not unlike other polymers, polystyrene will degrade under typical devolatilization temperatures, 210-250 °C. In addition to loss of polymer molecular weight, styrene is formed when the polymer and/or oligomers degrade. When designing a devolatilization unit operation, it is important to minimize residence time at these elevated temperatures. Modern commercial devolatilization designs have focused on maximizing heat transfer and minimizing residence time. Devolatilization performance is usually measured against the equilibrium amount of volatile in the final polymer. The equilibrium level for the devolatilization conditions used can be calculated using a simplified Flory-Huggins equation for monomer activity in the polymer melt [6]. By equating the partial pressure of the monomer solution to the flash tank partial pressure, the following results: P = (j)P0e(l+'/J where P is the flash tank pressure, 4> is the volume fraction of monomer in the devolatilized product, p° is the vapor pressure of pure styrene at the flash temperature and X is the Flory-Huggins interaction parameter. For a polystyrene-styrene system, X — 0.34. Therefore,
^
3.8/>c
and under operating conditions (230 °C): W
P
where W is the weight fraction of styrene in the devolatilized polystyrene. In a commercial polystyrene manufacturing plant, the final residual levels can deviate significantly from the equilibrium levels. This deviation can be attributed to
60
B. J. MEISTER AND C. J. CUMMINGS
two factors. The first is the generation of monomer from the degradation of polymer and oligomers. Even if moderate temperatures are used, if there is significant residence time in the system after the flash devolatilization, there will be monomer formation. In a polymer solution containing 300 ppm styrene dimer held for 1 h at 230 °C, an additional 95 ppm of styrene will be generated [6]. The second factor causing a deviation between equilibrium and actual volatile concentrations is mass transfer. In the flash devolatilization of polystyrene, the generation of thin polymer films is critical. However, even with the high surface area created by bubble formation and growth, mass transfer significantly impacts the end results. Measuring the mass transfer rate in a dynamic foam is extremely difficult because the interfacial area is both unknown and constantly changing. However, in comparing commercial data to equilibrium calculations, Meister and Platt have developed correlations for the mass transfer effect [22]. In 1989, Meister and Platt published data comparing equilibrium volatile levels in polystyrene with actual commercial performance [22]. Figure 3.7 is a plot of styrene concentration versus devolatilizer vacuum. The two products shown differ in their molecular weight: product 1 has Mw = 304000 and product 2 has Mw = 196000. It is important to note that the offset between the data and the equilibrium levels is independent of both viscosity and devolatilizer pressure. Figures 3.8 and 3.9 are plots of styrene dimer and trimer levels as a function of devolatilizer pressure. Because these oligomers are of much higher molecular weight than styrene, their diffusion rates are much lower and so the departure from equilibrium is more significant. 3.2
DEVOLATILIZATION
EQUIPMENT
There are three basic types of devolatilization equipment that have been used for the commercial manufacture of polystyrene: wiped film evaporators, devolatilizing extruders and flash evaporators. In wiped film evaporators, the polymer solution is fed into a vessel under vacuum. The solution is moved into thin films along the vessel walls by a set of rotating blades. These blades continue to move the polymer through the vessel while continually renewing the surface area. The tank walls are heated to supply the required energy for devolatilization. These units are typically mounted vertically with the polymer solution fed at the top. At the bottom is a melt pool where a gear pump transfers the melt to the next unit operation, typically pelletization. Advantages of the wiped film evaporator include: • Because the devolatilization occurs in thin films on the heated vessel wall, it is possible to input heat during devolatilization. This can be important because for many heat-sensitive polymers, the temperature required for an adiabatic flash would cause excessive polymer degradation.
COMMERCIAL PROCESSES FOR THE MANUFACTURE OF POLYSTYRENE
1000
61
DATA-PRODUCT 2DATA-PRODUCT 1
750
500
EQUILIBRIUMPRODUCT 2
250
A O D O
10
PRODUCT PRODUCT PRODUCT PRODUCT
1 1 2 2
STREAM STREAM STREAM STREAM
15 20 25 30 VACUUM (TORR)
A B A B
35
Figure 3.7 Residual styrene as a function of devolatilizer vaccum. Reprinted with permission from B. J. Meister and A. E. Platt, Ind. Eng. Chem. Res., 28, 1662 (1989). Copyright 1989 American Chemical Society
• Wiped film evaporators are more effective at removal of components with lower diffusion constants such as styrene dimer and trimer. Although there are clearly some specific advantages with the wiped film evaporators, they have not been widely applied for commercial polystyrene production. Reasons for this are most likely the high equipment and maintenance costs associated with these types of units. The second type of equipment used for volatile removal from polystyrene is the devolatilizing extruder. In these devices, an extruder is equipped with one or more pressure let-down sequences where vacuum is applied. In these devices, polymer surfaces are constantly being renewed, giving excellent mass transfer. Another advantage with the devolatilizing extruder is the ability to add and mix additives after devolatilization. This is especially useful if the additive has a
62
B. J. MEISTER AND C. J. CUMMINGS
500 DATA-PRODUCT 2
400
02 UJ
^ 300 Q
DATA-PRODUCT 1 1
200
100 -
A O d O 10
I 15
PRODUCT PRODUCT PRODUCT PRODUCT I I 20 25
1 1 2 2
STREAM STREAM STREAM STREAM I I 30 35
A B A B
VACUUM (TORR)
Figure 3.8 Residual styrene dimer as a function of devolatilizer vacuum. Reprinted with permission from B. J. Meister and A. E. Platt, Ind. Eng. Chem. Res., 28, 1662 (1989). Copyright 1989 American Chemical Society
vapor pressure that would cause it to be removed under vacuum conditions. While these advantages are important, extruders are probably the most expensive of the three equipment types discussed here. The final type of equipment to be discussed is what has been termed the flash evaporator. While there are many variations of this type of equipment, they all have three main components. The polymer solution is first heated via some type of heat exchanger (Figure 3.10). It is then forwarded to a separation vessel, or flash tank, where the vapors formed are disengaged from the polymer melt. The polymer then collects at the bottom of this vessel until it is forwarded via a gear pump to the next unit operation, typically pelletization. The types of heat exchangers used in these processes can vary widely. In probably the simplest form, a standard shell and tube type exchanger can be
COMMERCIAL PROCESSES FOR THE MANUFACTURE OF POLYSTYRENE 63
5000 DATA-PRODUCT 2
4000
EQUILIBRIUM-PRODUCT 2
w I
3000 DATA-PRODUCT 1 2000
EQUILIBRIUM-PRODUCT 1
1000
A O d O
05
10
PRODUCT 1 PRODUCT 1 PRODUCT 2 PRODUCT 2
STREAM A STREAM B STREAM A STREAM B
15 20 25 30 VACUUM (TORR)
35
Figure 3.9 Residual styrene trimer as a function of devolatilizer vacuum. Reprinted with permission from B. J. Meister and A. E. Platt, Ind. Eng. Chem. Res., 28, 1662 (1989). Copyright 1989 American Chemical Society
used to heat the polymer solution. A pressure control valve may be installed after the heat exchanger to eliminate flashing before entering the flash tank. In this type of operation, the flash is completely adiabatic and care must be taken so that the polymer is not cooled to a point where the viscosity is too great. Conversely, if the temperature is excessive, the polymer will degrade. This type of process has been used very broadly within the polystyrene industry. Over the years, there have been many improvements made to the basic process. One basic improvement is to install a die plate at the top of the flash vessel to create strands of foaming polymer. The falling strand devolatilizer, as it has been called, is an effort to increase the polymer surface area and come closer to reaching equilibrium. In addition to having falling strands, other obstructions have been placed in the flash tank to increase the surface area [23–25],
64
B. J. MEISTER AND C. J. CUMMINGS Pressure Control Valve
Flash tank
Heat exchanger
Polymer melt pool Polymer solution from reactors
TLUJ
Polymer gear pump
Polymer forwarded to pelletization unit Figure 3.10
Basic flash devolatilization process
As mentioned previously, the process shown in Figure 3.10 involves an adiabatic flash. Another improvement to this process is to input heat while the flash is occurring. While this can be done by simply removing the control valve shown in the diagram, reducing this to practice is more difficult. When the back-pressure valve is eliminated, vaporization will begin in the heat exchanger and the two-phase mixture will be carried forward into the flash tank. Ensuring adequate heat transfer into the solution and proper vapor-melt disengagement are two important engineering aspects that must be addressed. Another variation of this process is to mount the shell and tube heat exchanger directly on the flash tank so it will discharge as falling strands directly into the vessel [26,27]. There have been other variations and improvements to this basic process. A common improvement is to insert static mixing elements within the heat exchanger tubes. This can increase heat transfer significantly. Also, by inserting these mixing elements into the polymer flow path, the pressure drop across the tubes is increased and the point of flashing and subsequent polymer-vapor disengagement can be better controlled. Another variation of the flash devolatilization process is the use of plate-type heat exchangers [2831]. These exchangers can be constructed such that they can be inserted directly into a flash vessel. The flowpath inside such an exchanger is typically a long, thin slot. Advantages of these exchangers include high surface area and the ability to design the flow channel to optimize where flashing occurs.
COMMERCIAL PROCESSES FOR THE MANUFACTURE OF POLYSTYRENE 3.3
STEAM
65
STRIPPING
Because of equilibrium and mass transfer limits, the conventional flash devolatilization processes cannot reach monomer/solvent levels significantly below 200 ppm. If levels below 200 ppm are required, an alternative process using a stripping agent is required [32,33]. Although water has been used widely because of its low cost, alternative stripping aids can also be used [34]. Alcohols and light hydrocarbons are potential alternatives. Supercritical carbon dioxide has also been explored [35]. A basic version of this steam stripping process is shown in Figure 3.11. The starting point for this process is a polystyrene melt that has already been nearly completely devolatilized. The stripping agent is injected and mixed into the system and the polymer-stripping agent solution is flashed in another flash vessel. The final product is then pumped from the flash vessel to a finishing unit operation. The purpose of the water injection is twofold: first, the presence of water during the flash significantly reduces the partial pressure of monomer and/or solvent, and second, the flashing water creates more surface area to aid mass transfer. Although steam stripping is used commercially, it is not without drawbacks. The primary drawback is the additional capital and operational costs to install such a system. In addition, there are other process issues that must also be addressed carefully. Water injection Mixer Heat Exchanger
First Devolatilizer Polymer solution from reactors
Second Devolatilizer
Polymer forwarded to pelletization unit Figure 3.11
Basic polystyrene steam stripping process
66
4
B. J. MEISTER AND C. J. CUMMINGS
CURRENT POLYSTYRENE POLYMERIZATION PROCESSES
As one examines the evolution of polystyrene production processes over the last 60 years, one finds that economics, quality of the products and the range of products that can be produced are the main drivers for this evolution. Processes that do not make money for their owners eventually get shut down and the business evolves. Because this is a high-capital business and growth is relatively modest, evolution is slow. However, if one compares the polystyrene business today with that in the 1960s, one finds tremendous changes in the size and type of process used, the companies that are competitive producers, and the global locations of the production plants. Table 3.1 summarizes some of the advantages and disadvantages of the various process choices already described. However, all of the styrenic plants being built today, with the exception of emulsion-based ABS, are continuous free radical polymerizations. If one were to construct a crystal polystyrene (GP) plant today, one would likely choose a backmixed reactor (CSTR) as the principal reactor. This reactor would operate in the 50–70 % solids range and would discharge by means of a gear pump to a devolatilizer system. In this case there would probably be two devolatilizers, the first raising the solids to the range 80–85 % and the second producing the final product. This final devolatilizer would probably combine a heater, a distributor, and a flash tank operated below l0 mmHg absolute pressure and a temperature above 230 °C. The heater might well serve as the distributor with tubes or channels open to the flash tank so vaporization begins in the heater and the flash is not completely adiabatic. From Figure 3.7, the final polymer is probably below 500 ppm residuals under these conditions. The vapor may be cooled in a desuperheater and then passed through a final condenser and recycled back to the feed. As ethylbenzene, cumene and other hydrocarbons that do not polymerize are present in the styrene, they build up in the recycle, causing this to be a solution process whether one wanted it or not. Some of these hydrocarbons leave in the product, but when operating below 10 mmHg absolute pressure in the devolatilizer, there generally is excess recycle generated in the process. As shown in Figure 3.12, the product recovered in the devolatilizer is pumped to a die and cut with a cutter into pellets. The residence time from the devolatilizer to the die is minimized to control the regeneration of styrene at elevated temperature. Plasticizer and other additives are often added at this point to avoid vaporizing them in the devolatilizer. Either a mechanical or a static mixer might be used to ensure the additives are uniformly mixed prior to pelletization. A world-scale plant today may produce as much as 10000 kg/h of polystyrene. Figure 3.13 illustrates what one might build today for a high-impact polystyrene (HIPS) plant today. The individual companies involved have patented a number of specialized reactor configurations. These are illustrated in Figure 27 in a review by Echte [36]. One common configuration is shown here in Figure 3.13 with two backmixed reactors (CSTR) followed by two linear flow
Table 3.1
Polymerization methods for manufacturing polystyrene.
Reactor type
Polymerizing system
Advantages
Continuous solution Free radical (backmixed reactor)
Styrene monomer Recycled solvent W or W/O initiator
Good Temperature Control Limited in final conversion Limited in product range Good for copolymers Good clarity and color Pumping difficulties Uniform product
Continuous solution Free radical (linear flow reactor)
Styrene monomer Recycled solvent W or W/O initiator
Good range of products Good for rubber extension Good clarity and color
Large number of control zones High capital Pumping difficulties Low-cost process for HIPS
Excellent heat control High conversion No devolatilization Good range of products
Need prereactor for HIPS Poorer clarity Poor uniformity Round beads are hazard
High operating costs Better for lowvolume products
Sensitivity to impurities Initiator cost Color of product Cannot produce HIPS
Not proven for high-volume GP
Batch or continuous Styrene monomer Suspension free radical Water carrier Stabilizing agent Several initiators Continuous solution Anionic
Pure styrene monomer Polymerize to completion Much recycled solvent Low residual monomer Anionic initiators High polymerization rate Good for spec, copolymer
Disadvantages
Economics High capital Low-cost process for high-volume GP
68
B. J. MEISTER AND C. J. CUMMINGS Recycle
Desuperheater
Excess Recycle
Feed
Additives
Product Figure 3.12 reactor)
General-purpose polystyrene plant (CSTR=continuous stirred tank
reactors (LFR), all in series, which follows the original patent by BASF [37]. The rubber, which usually arrives at the plant in bales, is ground up in a grinder and added to a large tank with the styrene feed. The rubber is dissolved in this agitated tank over a period of about 12hs. Two tanks are often used in parallel so that one tank feeds the first reactor while the other is dissolving more rubber. The first backmixed reactor is operated with rubber continuously and is used to graft the rubber prior to feeding the second backmixed reactor which operates above phase inversion and is used primarily to size the rubber particles and set the morphology for the two finishing reactors. The solids leaving the first reactor might be 12%, the second reactor might well operate at 30% solids and this conversion is carried to about 80% solids in the two linear flow finishing reactors. The finishing reactors are usually not recirculated or backmixed in order to generate a high level of rubber extension or high toughness for the amount of rubber used. Once the HIPS leaves the reactor system, the finishing and recycle systems may well be identical with those used for generalpurpose polystyrene illustrated in Figure 3.12.
COMMERCIAL PROCESSES FOR THE MANUFACTURE OF POLYSTYRENE Excess Recycle
69
Desuperheater
Product Figure 3.13 High-impact polystyrene plant (CSTR = continuous stirred tank veactor; LFR = linear flow reactor)
The system utilized in Figure 3.13 for HIPS can also be used to produce a solution polymerization ABS. This type of ABS is used in non-glossy applications. The glossy ABS is usually produced in an emulsion process in which emulsified polybutadiene latex is grafted and agglomerated and blended with a continuous phase of SAN. This blended material is then dried and pelletized. This process is not cost competitive with the continuous solution polymeriza tion, but it produces a product with a superior balance of properties that commands a premium price.
5
PROCESS SIMULATION AND CONTROL
Probably no part of the polystyrene production plant has changed as much over the last 30 years as the methods of process control. The early polystyrene processes required little process control because they were operated at reaction rates that were inherently stable. For polystyrene, a rule of thumb is that the reaction rate doubles with every increase in temperature of 10°C. If the reaction is conducted at rates that evolve heat at a rate that requires a temperature
70
B. J. MEISTER AND C. J. CUMMINGS
difference between the polymer and the heat transfer fluid of less than 10°C, then increases in polymer temperature lead to larger increases in heat transfer than increases in the heat evolved and the reactor is stable. However, the development of faster responding process control systems and the evolution to computer control have made possible the move into operational regimes that are not inherently stable, but are controlled by systems that have a faster response to an upset than the process. The other two parts of the major change in process control are first the implementation of gel permeation chromatography as a process control technique for molecular weight and molecular weight distribution. Second is the evolution of kinetic models for styrene polymerization for use as off-line and on-line computerized simulations of the reaction. Fundamental polymerization kinetics are the link between the operational parameters of temperature, monomer concentration and residence time and the resulting rates of reaction and molecular weight produced at any point in the reactor. Off-line simulations for the polymerization of polystyrene and the copolymerization of styrene and acrylonitrile based on fundamental kinetics were developed by Meister [38] at Dow Chemical in the early 1970s. The recognition that both the termination and propagation reactions were diffusion controlled and that the magnitude of diffusion control was primarily dependent on the polymer concentration was the key to obtaining accurate simulations that could be used to guide the operating conditions of the production plants. Models that are similar to the Dow models have been published in the literature and are widely used today. In particular, Hamielec and a series of students [39–41] developed models that are similar in format and predictions, although very different in some of the details. In the late 1970s, the Dow polystyrene production facilities made a major conversion from thermally initiated polystyrene to polymerization initiated by a bifunctional initiator. This brought about a new series of products that could be produced at higher rates and had improved grafting and a lower level of oligomers. The use of the simulation programs was a major factor in making possible the conversion to the new operating conditions and the development of new rules of thumb for adjusting operating conditions to maintain product properties. A method for the computation of the molecular weight of polystyrene formed when using a bifunctional initiator was a key to the success of the model. This was true even though some of the mechanisms involved only became understood later. Models for polystyrene polymerization using bifunctional initiators are now available in the literature [42,43]. Improving the predictions of the diffusion-controlled termination and propagation reactions is a subject of continuing interest [44]. This is particularly true because of the increasing use of backmixed reactors for the production of GP polystyrene and the increasing use of bifunctional initiators in these reactors at high conversions where the diffusion control is amplified. This search for polymerization conditions where high molecular weight polymers can be produced
COMMERCIAL PROCESSES FOR THE MANUFACTURE OF POLYSTYRENE
71
at high production rates often leads to situations where multiple steady states occur [45]. This can lead to oscillatory behavior [46] and it can also lead to the existence of two phases of differing concentration and temperature within the same reactor. Computer control of the operating parameters to maintain setpoints, the use of off-line kinetic models to determine operating conditions, and the use of online models to make predictions of the intermediate parameters and the final product properties as a function of the operating history have all become commonplace in industry over the last 30 years. This certainly has had a strong influence on the improvements in product quality and uniformity and the large increases in production rate that have occurred over this period of time.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.
Blyth, J., and Hofmann, A. W., Ann. Chem., 53, 289 (1845). Staudinger, H., Ber. Dtsch. Chem. Ges., 53, 1073 (1920). Boyer, R. F., J. Macromol. Sci. Chem., A15, 1411 (1981). Debell, J. M., et l., German Plastics Practice, ebell and Richardson, Springfield, MA (1946). Simon, R. H. M., and Chappelear, D. C., Technology of Styrente Polymerization Reactors and Processes, ACS Symposium Series 104, American Chemical Society, Washington, DC (1979). Meister, B. J., and Malanga, M. T., in Styrene Polymers, Encyclopedia of Polymer Science and Engineering, Vol. 16, Wiley, New York, pp. 21–61 (1989). Albright, L. F., Processes for Major Addition Type Plastics and Their Monomers, McGraw-Hill, New York (1974). Boundy, R. H., and Boyer, R. F., Styrene, Its Polymers, Copolymers, and Derivatives, ACS Monograph No. 115, Reinhold, New York (1952). Svec, P., et al., Styrene-based Plastics and Their Modification, Ellis Horwood, New York (1989). Albalak, R. J., Polymer Devolatilization, Marcel Dekker, New York (1996). Amos, J. L., Polym. Eng. Sci., 14, 1 (1974). Allen, L, et al, US Patent 2496653 (to Union Carbide) (1950). Latinen, G. A., US Patent 3794471 (to Monsanto) (1971). Priddy, D. B., in Styrene Plastics, Kirk Othmer Encyclopedia of Chemical Technology, 4th edn, Vol. 22, pp. 1015-1073, Wiley, New York (1997). Priddy, D. B., Pirc, M., and Meister, B. J., Polym. React. Eng., 1, 343 (1993). Amos, J. L., et al., US Patent 2694692 (to the Dow Chemical) (1954). Echte, A., Angew. Makromol. Chem., 58, 175 (1977). Turley, S. G., and Keskkula, H., Polymer, 21, 466 (1960). Meister, B. J., et al, US Patent 4 876 372 (to Dow Chemical) (1989). Platzer, N. Ind. Eng. Chem., 62(1), 6 (1970). Mott C. L., and Kozakiewicz, B. A., US Patent 4221883 (to Dow Chemical) (1981). Meister, B. J., and Platt, A. E., Ind. Eng. Chem. Res., 28, 1659 (1989). Aboul-Nasr, O. T., US Patent 4934433 (to Polysar Financial Services) (1990).
72
B. J. MEISTER AND C. J. CUMMINGS
24. 25. 26. 27. 28. 29. 30. 31. 32. 33.
Aboul-Nasr, O. T., US Patent 5069750 (to Polysar Financial Services) (1991). Desroches, D., and Krupinski, S., US Patent 5 874 525 (to Nova Chemicals) (1999). Gordon, R. E., and McNeill, G. A., US Patent 3 853 672 (to Monsanto) (1973). Hagberg, C. G., US Patent 3966538 (to Monsanto) (1976). Fink, P., et al., US Patent 4153 501 (to BASF) (1979). Aneja, V. P., and Skillbeck, J. P., US Patent 4 808 262 (to General Electric) (1989). Mattiussi, A., et al., US Patent 5084134 (to Montedipe) (1992). Cummings, C. J., and Meister, B. J., US Patent 5453 158 (to Dow Chemical) (1995). Szabo, T. T., US Patent 3 773 740 (to Union Carbide) (1971). Darribere, C., et al., Presented at the 6th International Workshop on Polymer Reaction Engineering (1998). Skilbeck, J. P., US Patent 5350813 (to Novacor Chemicals) (1994). Sacide, A., and Duda, J. L., AIChE J., 44, 582 (1998). Echte, A., in Rubber Toughened Plastics, ed. Riew, C. K. Advances in Chemistry Series 222, American Chemical Society, Washington, DC (1989). Bronstert, K., et al., US Patent 3658946 (to BASF) (1972). Meister, B. J., Dow Internal Documents (1974, 1975). Hui, A. W., and Hamielec, A. E., /. Appl. Polym. Sci., 16, 749 (1972). Friis, N., and Hamielec, A. E., /. Appl. Polym. Sci., 19, 97 (1975). Marten, F. L., and Hamielec, A. E., J. Appl. Polym. Sci., 27, 489 (1982). Villalobos, M. A., et al., J. Appl. Polym Sci., 42, 629 (1991). Dhib, R. et al., Polym. React. Eng., 8, 209 (2000). Vivaldo-Lima, E., et al., Polym. React. Eng. 2, 17 (1994). Henderson, L. S., Chem. Eng. Prog., 83, 42 (1987). Villa, C. M., et al., Polym. React. Eng., 7, 151 (1999).
34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46.
DUANE B. PRIDDY Dow Polystyrene R&D, Midland, Ml, USA
1
INTRODUCTION
Production of polystyrene (PS) in North America began in 1938 by The Dow Chemical Company. The first Dow process involved simply immersing metal cans (full of styrene monomer) in a heated oil bath until the monomer conversion reached 99+ %. Then the cans were opened and the PS was crushed to form granules. The PS prepared and isolated in this manner contained high levels of unreacted styrene monomer which ended up in the final fabricated articles. For the next several years, R&D efforts were focused upon the development of improved manufacturing processes and devolatilization processes to remove the unreacted monomer. The batch 'can process' finally gave way to continuous bulk polymerization technology (Figure 4.1) and suspension polymerization processes. Suspension polymerization had the advantage of yielding polymer granules direct from the polymerization reactor and utilized organic peroxide initiators which led to faster polymerization rates and very high monomer conversion (>99.5 %). The continuous bulk process, on the other hand, only allowed conversion of monomer to polymer of ~80 %, resulting in the need to remove the unreacted styrene monomer. The process developed by Dow to devolatilize the polymer syrup was to pass it through a heat exchanger to take the temperature up to 230-260 °C, and then forwarding it into a vacuum chamber at < 10 mmHg. The unreacted styrene monomer volatilizes from the polymer melt and is then condensed and recycled back into the polymerization process. However, this devolatilization process does not remove all of the monomer. Depending upon the temperature, vacuum, surface area of the polymer melt, and residence time in the Modern Styrenic Polymers: Polystyrene and Styrenic Copolymers. Edited by J. Scheirs and D. B. Priddy 2003 John Wiley & Sons Ltd
74
D. B. PRIDDY
vacuum tank, the level of monomer left in the PS can vary significantly. There is a theoretical minimum that can be achieved based upon the equilibrium partitioning of the monomer vapor with the molten polymer. This minimum varies with temperature and pressure inside the vacuum chamber as shown in Figure 4.2. There has been continuous improvement in PS purification technology over the past 60 years as efforts continued to focus on getting the residual monomer levels lower and lower. These efforts continue even though it is generally Styrene EB Initiator
75%PS
240°C 230°C. The problem with peroxides is that they are thermally unstable and do not survive the polymerization to reach the devolatilizer. Blakemore [23] solved this problem by using cyclic peroxides which have very high decomposition temperatures. Their thermal stability is due to the peroxide bond reforming once broken, because the two oxy radicals cannot escape each other so they recouple. If styrene happens to be in the vicinity of the cyclic peroxide while it is a dioxyradical, the diradical adds across the styrene double bond (Scheme 4.1).
Scheme 4.1 styrene
Hypothesized chemistry of how cyclic peroxides lower residual
Rather than adding peroxides that are very thermally stable, ICI researchers added peroxide generating enzymes to PS [25]. These enzymes (oxidases) constantly convert absorbed air to peroxides, which decompose forming radicals that allegedly scavenge the styrene. A class of scavengers developed by Dow researchers is benzocyclobutene (BCB) [26]. In the 1980s, Dow researchers began developing BCB derivatives for the electronics industry. BCB has the same molecular formula as styrene (C8H8), is inert in styrene polymerization, and becomes very reactive toward styrene when heated to >200°C. The chemistry of BCB is shown in Scheme 4.2. Above 200 °C, the strained ring of BCB opens to form o-xylylene, which Diels-Alder couples with the double bond of styrene to form phenyltetralin. Any residual BCB that does not react with styrene continues to react with itself and eventually is converted to C8H8 oligomers.
80
D. B. PRIDDY >200°C Ph
R
Scheme 4.2 Reaction of BCB with styrene to form a dimer. BCB also reacts with itself to form oligomers
2.4
ABSORBERS
It is well known that small molecules can be removed from larger molecules using inorganic materials such as molecular sieves. However, this approach contaminates the polymer with inorganic particulates. The only literature teaching this approach are Russian patents claiming the addition of silica gel [27] and montmorillonite clay [28] to absorb styrene from PS. The advances in nanocomposite technology in recent years may allow further development of this approach. 2.5 2.5.1
HIGH MONOMER CONVERSION
POLYMERIZATION
Free Radical
Typically, continuous bulk free radical polymerization processes produce partial polymer syrups at about 65-80 % solids. The unreacted styrene that remains is removed by evaporation. If the solids content of the polymer could be taken higher, the level of residual styrene in the polymer would be lower, especially when using a one-stage devolatilizer. In suspension polymerization, devolatilization is not even required because styrene is polymerized to >99.9 % monomer conversion. If it were possible to polymerize styrene to very high conversion in bulk polymerization processes, one should be able to achieve significantly lower residual styrene monomer. Owing to viscosity constraints, bulk polymerization reactors cannot operate at >80 % solids. Kelley patented a high-conversion bulk styrene polymerization process resulting in the formation of low residual PS (LRPS), which solves the viscosity problem and allows >99.9% solids to be achieved by finishing the polymerization off in an extruder [29]. Extruders are not very effective heat exchangers yet are designed for handling high-viscosity materials. Thus Kelley carried out the polymerization of pure styrene monomer without the use of a solvent in a conventional polymerizer to normal solids levels and then fed the partial polymer into an extruder where he finished off the polymerization. He used a mixture of initiators having different half-lifes so that radicals were continuously generated. More recently van der Goot and Janssen
APPROACHES TO LOW RESIDUAL POLYSTYRENE
81
[30] also looked at free radical polymerization of styrene directly in an extruder using 1 wt% peroxide initiator. They were not successful in making PS having a molecular weight > 100 000. They got around this problem by attaching a prepolymerizer to the front end of the extruder to polymerize to 25 % solids to make high-MW PS. The low-MW PS then made in the extruder ended up giving them a bimodal PS. The conversion they that achieved in the extruder was 98-99 %. By pulling a vacuum on the last zone of the extruder one could possibly get close to equilibrium styrene level. Injection of steam into the last zone might yield truly LRPS. However, the work published thus far on high conversion bulk polymerization does not appear to be aimed at achieving LRPS.
2.5.2
Anionic
Anionic polymerization chemistry is much better suited for polymerizing styrene to high conversion because of slower rates of termination of growing polymer chains. Therefore, during the polymerization, the concentration of active growing chains is much higher and the MW generally increases with monomer conversion whereas with free radical polymerization, the MW generally decreases rapidly at very high monomer conversions resulting in broadening of the overall polydispersity. Over the past 15 years, there have been significant advances aimed at developing anionic polymerization technology for industrial production of PS. Dow Chemical researchers [31] focused their efforts on solution polymerization in continuous stirred tank reactor (CSTR) processes with heat removal by ebullient cooling whereas BASF and Asahi researchers have focused on continuous plug flow reactors (CPFR) [32]. Since anionic polymerization in CSTR reactors operates at 99.9+ % monomer conversion at steady state and boiling is required to achieve ebullient cooling, a solvent is needed to lower the viscosity of the polymerization mass. The CPFR process, on the other hand, only reaches high monomer conversion in the final stage of the polymerization process where viscosity can be managed by increasing the temperature. The key problem that researchers faced with the CPFR process is the rapid rate of anionic polymerization. The heat removal capacity of CPFR is not sufficient to control the polymerization, resulting in runaway kinetics. The advance that BASF and Asahi researchers discovered is that the propagation rate can be slowed by the addition of certain electron-deficient organometallic reagents (e.g. dibutylmagnesium) [33,34]. Depending on the Mg:Li stoichiometric ratio, the polymerization rate can be adjusted to whatever is necessary to match the heat removal capability of the polymerization reactor being used (Figure 4.5). Anionic polymerization not only allows the production of low residual PS (i.e. typically PS produced using continuous anionic polymerization contains R R 4- R -> P
Diels-Alder forward Diels-Alder reverse Initiation Transfer to dimer Transfer to monomer Propagation Termination tft
Rate without acid /?i = K\M2 R^ =K-\DH R, = k,M^ Rm = kmMR Rm = kmMR Rp = kpMR = A:t/?2
Rate with acid /?, = KXM2 /?_, = R'- = Ri R'm = R^ = 0 l/2 /? R' = 01/2/?p /?[ = ^/?,
In order to estimate , the QSSA is again applied to DH, with and without acid: DH =
Ri-R>-Rm
where
DH' _ /?i - (j>Rt - 4>3/2Rm ( r\ TJ
JLJn
n
l\\
n
\
\
R
n
— Y\j — -f^m
If the rate of chain transfer to DH is significant, then is calculated by the model using successive substitution. Starting with = 1 on the right-hand side, the calculation converges sufficiently with three iterations. If Rm is negligible (for example, when considering chain transfer to monomer only), then can evaluated directly. This expression for (/> contains two unknown rate constants, k\ and ks. The Arrhenius coefficients for these rate constants were determined using the 140 °C conversion data from Figure 7.6. The parameters were estimated for both the case of chain transfer to monomer, and again for chain transfer to DH. Given that the two chain transfer models differ only in their predictions of A/w and that the fit was against conversion data, the optimum Arrhenius constants for both cases were the same: L = 6.91 x 102e758/;r
These model equations were implemented in a single computer program that does both standard autopolymerization and acid-mediated polymerization:
INCREASING PRODUCTION RATES OF HIGH MW POLYSTYRENE
143
by setting initial acid concentration to zero, (f> —* 1 and all the rates are calculated according to the Hui and Hamielec model.
5.3
MODEL RESULTS
During the parameter estimation, it was noted that the model is relatively insensitive to changes in k\, and therefore it should not be taken as a reliable prediction of the Diels-Alder reaction rate constant. Effectively, the rate R\ is very large compared with the rates of initiation and chain transfer, and therefore the DH ratio calculation could be simplified without significantly changing the model results: ' - l ~DH ~ l+ksS DH
There are two other implications: (1) the relative magnitude of the rates validates Hui and Hameilec's simplifying assumption that lead to a thirdorder rate equation for thermal initiation; (2) the uncertainty in the prediction of k\ precludes any argument for or against the possibility that acid catalyzes the Diels-Alder reaction as well as the aromatization of DH. Turning to model results, Figure 7.11 shows the predicted monomer conversion after 1 h (after 0.5 h for the 160 °C case). The effect of acid on conversion
0.4 - •
.2
0.3 ••
200
T = 120°C Figure 7.11
400 600 CSA (ppm)
T = 140°C
800
-T = 160°C
Effect of acid level on conversion and Mw at 140 °C
1000 • Data
144
B. MATTHEWS AND D. B. PRIDDY
appears to be well predicted, lending some credence to the mechanism and structure of the model. It should be noted, however, that what few data were available were used to make the fit. More data are required to validate the conversion predictions thoroughly. Molecular weight predictions are compared with data from Buzanowski et al. [14] in Figures 7.12 and 7.13. In both figures, plot (a) shows results assuming chain transfer to monomer only (Model A), while Model B assumes transfer to DH only and is represented in plot (b). Neither chain transfer model is adequate. With the exception of the 160°C run, Model A severely under-predicts Afw. Under-prediction can have severe consequences in the plant: run conditions predicted to be safe by the model would actually generate much higher Afw and therefore high viscosity and possibly gels leading to shut-down. The Mw predictions in Figure 7.12 show the correct overall trend (A/w increases with increasing acid concentration), but the magnitude is off by a factor of 2-3. Several possible explanations for the poor Mw predictions can be considered. Since Afw is the problem, a review of the moments equations contained in the base model is in order to see if they are compatible with the way in which $ is implemented. All of the moment concentration rate equations are based on either the rate of propagation or ratios of termination and chain transfer to the propagation rate. These ratios are very straightforward to correct using the Mayo dimer factor, . The zeroth- and first-order moment rates appear correct, as they are the concentration of terminated chains and the concentration of monomer that has been polymerized, respectively. Mn data are not available to validate the ratio of first to zeroth moment. Calculation of Mw requires the second-order moment, but re-deriving it as a check is beyond the scope of this work. Another source of model error could be the Cm equation shown earlier. A linear correction factor is applied to compensate for the gel effect on propagation rate. Since the datapoints available are at relatively low conversion where the Comftrbom whfc Bnwomki's Figore 3 u§iDg dnia transfer to noMaer model
0
(a)
|
200
T = UCfC
Figure 7.12
400 600 CSA (ppn)
800
i wit* BuMomfci't FigBt 3 r to M«yo 41
1000
T=I40°C ——T=160°C • Dau|
0
(b) |
T = I20°C
T = 140'C ——T = 160T
Effect of acid level on Mw at various temperatures
»Dau|
145
INCREASING PRODUCTION RATES OF HIGH MW POLYSTYRENE
gel effect is low, this correction might be interfering with the acid effects. The Cm calculation for chain transfer to dimer contains a significant correction for dimer concentration which could be duplicating the correction applied by the (j> factor. Assuming that the equations are reasonably accurate, the chain transfer mechanism could be called into question next. The model as written requires chain transfer to monomer or dimer, but not both. Since the two cases fall on either side of the experimental results, a modification to allow both to occur simultaneously might lead to a better fit. Although neither of the chain transfer models lead to a suitable tool for production planning, they both exhibited the correct trends for conversion and Mw relative to the addition of acid (Figure 7.13). Since the model modifications are based solely on altering the amount of DH available for initiation and chain transfer, the model results are consistent with the Mayo mechanism.
6
CONCLUSIONS
Significant improvements in the rate of manufacturing high molecular weight polystyrene can be achieved by adding small amounts of strong acid to the process. The acid should be soluble in styrene and must not initiate cationic polymerization. Other concerns such as corrosion, acid recovery, and allowable oligomer levels in the product must be considered before implementing the acid process. The acid works by catalytically aromatizing the Diels-Alder intermediate produced by the Mayo auto-initiation mechanism. A program written to simulate the Diels-Alder and aromatization mechanisms was successful in fitting the experimental conversion data. Molecular weight was underpredicted then over-predicted by changing the chain transfer assumptions. The Comparison with Buzanowski's Figure 5 using chain transfer to monomer model
300000 (a)
400000
500000
600000 Mw
700000
Comparison with Buzanowski's Figure 5 using chain transfer to Mayo dimer model
800000
| — w/o Acid — Sim with Acid OData with Acid > Data w.'o Acid|
Figure 7.13
300000
400000
500000
600000 Mw
700000
800000
(fc>) |—- w/o Acid — Sim with Acid» Data with Acid »Data w/o Acid|
Comparison of Rate-Mw curves with and without acid
146
B. MATTHEWS AND D. B. PRIDDY
general trend for Mw was correct in both cases, lending additional support to the choice of the Mayo mechanism with acid-catalyzed deactivation of the reactive dimer intermediate.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.
Cao, G.; Zhu, Z.; Le, H.; Zhang, M.; Yuan, W., /. Polym. Eng., 1999, 19, 135. Benbachir, M.; Benjelloun, D., Polymer, 2001, 42, 7727. Priddy, D. B., Adv. Polym. Sci., 1994, 111, 67. Drumright, R. E.; Ellington, E.; Kastl, P. E.; Priddy, D. B., Macromolecules, 1993, 26, 2253. Drumright, R. E.; Kastl, P. E.; Priddy, D. B., Macromolecules, 1993, 26, 2246. Toplikar, E. G.; Herman, M. S.; Buyle Padias, A.; Hall, H. K., Jr; Priddy, D. B., Polym. Bull. (Berlin), 1997, 39, 37. Dais, V. A.; Drumright, R. E.; Ellington, E.; Kastl, P. E.; Priddy, D. B., Macromolecules, 1993, 26, 2259. Drumright, R.; Terbruggen, R.; Priddy, D.; Koster, R., US Patent 5618900 (to Dow Chemical), 1997. Morioka, I.; Yamada, K., Japanese Patent 05 178914 (to Sekisui Plastics), 1993. Fuku, M.; Okada, Y.; Aoshima, K., Japanese Patent 62 197 407 (to Nippon Oils and Fats), 1987. Pike, W.; Priddy, D.; Vollenberg, P., US Patent 5 663 252 (to Dow Chemical), 1997. Sanchez, J.; Yormick, J.; Wicher, J.; Malone, K., World Patent 9807684 (to Elf Atochem North America), 1998. Cummings, C.; Hathaway, P. US Patent 5455321 (to Dow Chemical), 1995. Buzanowski, W. C.; Graham, J. D.; Priddy, D. B.; Shero, E., Polymer, 1992, 33, 3055. Flory, P. J., J. Am. Chem. Soc., 1937, 59, 241. Mayo, F. R., J. Am. Chem. Soc., 1968, 90, 1289. Pryor, W. A.; Coco, J. H.; Daly, W. H.; Houk, K. N., J. Am. Chem. Soc., 1974, 96, 5591. Pryor, W. A.; Graham, W. D.; Green, J. G., J. Org. Chem., 1978, 43, 526. Graham, W. D.; Green, J. G.; Pryor, W. A., J. Org. Chem., 1979, 44, 907. Pike, W. C.; Priddy, D. B.; Roe, J. M.; Rego, J. M., World Patent 0068281 (to Dow Chemical), 2000. Priddy, D. B.; Dais, V. A., World Patent 9618663 (to Dow Chemical), 1996. Roe, J. M.; Rego, J. M.; Priddy, D. B., US Patent 6084044, 2000. Pike, W. C.; Priddy, D. B., World Patent 9900432 (to Dow Chemical), 1999. Wesselmann, M. A., US Patent 4 585 825 (to Dow Chemical), 1986. Paquet, A.; Priddy, D.; Vo, C. V.; Pike, W.; Hahnfeld, J., US Patent 5650 106 (to Dow Chemical), 1997. Matthews, B.; Pike, W.; Rego, J.; Kuch, P.; Priddy, D., J. Appl. Polym. Sci., in press. Hui, A. W.; Hamielec, A. E., J. Appl. Polym. Sci., 1972, 16, 749.
8
Preparation of Styrene Block Copolymers Using Nitroxide Mediated Polymerization DUANE B. PRIDDY Dow Polystyrene R&D, Midland, Ml, USA
1
INTRODUCTION
Styrene is a very versatile monomer. It can be polymerized by most types of polymerization mechanisms, e.g. free radical (FR), Ziegler-Natta (ZN), anionic, and cationic. Classical ZN polymerization of styrene yields isotactic polystyrene. However, if methylalumoxane (MAO) is added as a co-catalyst, syndiotactic polystyrene is formed. The resulting polymers formed using the various mechanisms of polymerization are summarized in Scheme 8.1. Styrene-containing block copolymers are commercially very important materials. Over a billion pounds of these resins are produced annually. They have found many uses, including reinforcement of plastics and asphalt, adhesives, and compatibilizers for polymer blends, and they are directly fabricated into articles. Most styrene-containing block copolymers are manufactured using anionic polymerization chemistry. However, anionic polymerization is one of the more costly polymerization chemistries because of the stringent requirements for monomer and solvent purity. It would be preferred, from an economic cost perspective, to have the capability to utilize free radical chemistry to make block polymers because it is the lowest cost mode of polymerization. The main reasons for the low cost of FR chemistry are that minimal monomer purification is required and it can be carried out in continuous bulk polymerization processes.
Modern Styrenic Polymers: Polystyrene and Styrenic Copolymers. Edited by J. Scheirs and D. B. Priddy O 2003 John Wiley & Sons Ltd
148
D. B. PRIDDY Block Polym.
Low MW (2) Low cost (forgiving) No initiator required
Modes of polymerization of styrene
Nitroxide mediated radical polymerization (NMRP) was pioneered by Rizzardo and Solomon in the mid-1980s [1]. Their work went unnoticed for almost a decade until Georges et al. reported the preparation of narrow polydispersity (PD) (
7
+
V Major
Major
Scheme 8.2 Reactions of I upon thermolysis
Minor
150
D. B. PRIDDY
There are several ways in which block copolymers can be made. The three main methods are (1) sequential addition of monomers, (2) the preparation of a functionalized polymer followed by the use of the functionalized polymer as a macroinitiator or chain-stopper for initiation or termination of polymerization of the second monomer, and (3) use of a multiple-headed initiator. The purity of the block copolymers produced in these processes is dependent upon the livingness (lack of side reactions that lead to termination) of the chemistry used to make them. If the integrity of the chain-ends is maintained throughout the polymerization because all possible termination mechanisms are absent or eliminated, then pure block copolymers can be produced. If, however, impurities get into the process or if there are side reactions that lead to chain termination, the resulting block copolymers are contaminated with some homopolymer. Depending upon the application, some contamination of homopolymer in the block copolymer may be acceptable. In an effort to understand the limitations of NMRP for making functionalized and block copolymers, Priddy et al. carried out NMRP polymerization of styrene using model alkoxyamine I having a high extinction coefficient phenylazo chromophore attached to it, either on the initiating phenethyl radical (II) or else on the terminating TEMPO radical (III) [5]. This allowed quantification of the amount of the functionalized chain-ends during the polymerization of styrene using GPC-UV/VIS analysis. The results showed that a much higher percentage of polymer chain-ends have an attached chromophore group when using II versus III as the initiator (Figure 8.1). This finding suggests that there are more competing side reactions leading to termination than competing reactions leading to initiation of new chains. These data clearly show that highly pure block copolymers cannot be prepared using NMRP and that the purity of the block copolymer is inversely proportional to the molecular weight of the polymer segment formed using NMRP. The poor chain-end purity achievable using TEMPO-based alkoxyamine NMRP initiators led researchers to develop new nitroxyl radicals that will mediate vinyl polymerization more effectively. Hawker et al. utilized combinatorial techniques to synthesize and screen many different nitroxyl radical structures [6]. Their work led to the development of nitroxyl IV having a /^-hydrogen. Whereas TEMPO only works well for styrenic monomers, IV is claimed to work well for acrylates and diene monomers. Also, the polymerization rates achievable using IV are much faster than when using TEMPO. The alkoxyamine unimolecular initiator V, made from IV, has been successfully used to make a variety of block copolymers [7-9]. Recently, Hawker et al. synthesized V having a chromophore attached to either the initiating (VI) or the terminating (VII) radical (similar to the work of Priddy et al. on chromophore-labeled TEMPO mentioned previously) [10]. The results of the study showed that mediation of polymerizations using V yields polymers having greater end-group purity than polymerizations mediated using alkoxyamines based on TEMPO (Figure 8.2).
151
STYRENE BLOCK COPOLYMERS USING NITROXIDE l00-i "§95-
n
oe
o~\— -——— _
90-
^
4>
O
85-
n €
75-
~~"— '—
D^
a.
80-
\ \ .
\ \ DX
"?
.1 w
'~—~
\
o
o = k.
"""
O
0
70-
\
.c
—©— Initiated with II -a— Initiated with III
65^
\ \ \q
605
1
1
1
i
1
1
10
15
20
25
30
35
Mn/1000
Figure8.1 Comparison of PSchains having initiating versus terminating fragment derived from TEMPO-based alkoxyamine initiators. Polymerizations were carried out at 120°C 100 -i 98-
94o
92 -
86-
—e— Initiated using VI -B— Initiated using VII
84 100
50
150
Mn/1000
Figure 8.2 Comparison of PS chains having initiating versus terminating fragment derived from V-based alkoxyamine initiators. Polymerizations were carried out at 120 °C
152
D. B. PRIDDY
O
rv
VII
Chromophore
Chromophore
Another /?-hydrogen-bearing nitroxide having improved performance in NMRP is N-ter/-butyl-./V-( 1 -diethylphosphono-2,2-dimethyl)propylnitroxyl (VIII) [11,12]. A class of improved nitroxides (compared with TEMPO) for NMRP without a ^-hydrogen are the imidazolidone nitroxides (IX) developed by CSIRO researchers [13]. Both VIII and IX are claimed to give improved performance and higher end-group fidelity in NMRP, especially for acylates, compared with TEMPO.
O
N
•O—N O
I OEt
VIII 4
OEt
R
R
IX
BLOCK COPOLYMERS VIA THE MACROINITIATOR APPROACH
The highest volume commercial block copolymers are the styrene—butadiene (S-B) block copolymers. S–B block copolymers are manufactured using anionic polymerization with sequential addition of monomer (SAM) techniques. Attempts to make S—Bpolymers using NMRP via SAM have been limited because NMRP does not generally work well for diene monomers. Therefore, Priddy et al.
STYRENE BLOCK COPOLYMERS USING NITROXIDE
153
utilized the macromonomer approach. Butadiene was polymerized using traditional ani6nic chemistry to produce polybutadienyllithium (PBD-Li) [14]. The PBD-Li was then terminated with an alkoxyamine functional epoxide (X) to produce an alkoxyamine-functional macromonomer (XI). Addition of XI to bulk styrene polymerization at 120°C led to the formation of S—B copolymer (XII) (Scheme 3). The S—Bblock copolymer produced (XII) was characterized by a variety of techniques including NMR, gel permeation chromatography (GPC), thin-layer chromatography (TLC), and transmission electron microscopy (TEM). The results of these analyses clearly showed that the block copolymer was fairly pure with very little homopolymer contamination. One potential commercial application of this technology is the preparation of S—Bblock copolymers in situ during the manufacture of high-impact polystyrene (HIPS) and also acrylonitrile—butadiene—styrene (ABS) copolymers [15]. A significant proportion of HIPS and ABS polymers are manufactured in continuous bulk polymerization processes where an S–B block copolymer is dissolved in the monomer being fed to the reactor. Dow Chemical researchers demonstrated that addition of TEMPO-functionalized PBD can be added to the monomer feed instead of S–B block rubber, resulting in the formation of HIPS and ABS resins having similar properties to resins produced using block rubbers. Under appropriate conditions, transparent HIPS and ABS resins are formed. The transparency is achieved because the rubber and polystyrene or SAN domains are too small to scatter light. The TEMs of an S—Bblock copolymer XII and a transparent HIPS made using in situ-formed S—Bblock copolymer is shown in Figure 8.3.
PBD
Scheme 8.3 Synthesis of S—B block copolymers using sequential anionic/NMRP polymerization techniques
154
D. B. PRIDDY
Figure 8.3 TEMs of XII and TIPS made using /ns/fu-formed S—B by addition of XI to a continuous bulk styrene polymerization
Another example of the macroinitiator approach to making block copolymers is shown in Scheme 8.4. Since methyl methacrylate (MMA) polymerization cannot effectively be initiated by TEMPO-based alkoxyamine initiators, a poly(methyl methacrylate) macroinitiator (XIII) was prepared using conventional free radical polymerization [16]. However, the azo initiator was functionalized with a TEMPO-based alkoxyamine. Since the main mechanism of termination during bulk MMA polymerization is by radical coupling, most of the MMA polymer chain-ends are functionalized with alkoxyamine groups.
OH
CN
CN
+
Cl
o
N=
MMA/normal free radical polym. s I
Scheme 8.4 Synthetic approach to block copolymers using sequential normal/living radical polymerization
STYRENE BLOCK COPOLYMERS USING NITROXIDE
155
Addition of the XIII as a macroinitiator to styrene polymerization resulted in the formation of S—MMA—S triblock copolymer. The same procedure was also used to make styrene—butyl acrylate block copolymers. The success of this chemistry versus a control was demonstrated by conducting a parallel experiment where PMMA was prepared under the same conditions as the preparation of XIII, except using azobisobutyronitrile (AIBN) as the initiator instead of the alkoxyamine functional azo initiator. The AIBNand alkoxyamine functional azo-initiated PMMA were dissolved in styrene and heated at 130°C. A film of the resulting block copolymer made using alkoxyamine-functionalized XIII was translucent and flexible whereas a film of the polymer formed by polymerizing styrene in the presence of unfunctionalized PMMA control was opaque and very brittle. A final example of the use of the macroinitiator approach for block copolymer synthesis is the use of a macro azo initiator to initiate NMRP of styrene (Scheme 8.5). This process was used to make styrene-bl-siloxanes [17].
Scheme 8.5 azo initiator
5
Preparation of styrene-bl-siloxane using NMRP initiated with a macro
PREPARATION OF BLOCK COPOLYMERS USING ALKOXYAMINES AS CHAIN-STOPPERS IN STEP-GROWTH POLYMERIZATION
The molecular weights of polymers made using step-growth polymerization are typically controlled by the addition of a chain-stopper to the process. The chainstopper to monomer ratio determines the final molecular weight of the polymer. If functionalized chain-stoppers are used, functionalized polymers are produced. If macro chain-stoppers are used, triblock copolymers are formed during the step-growth polymerization. The highest volume commercial step-growth polymer is polycarbonate. There has been considerable interest in preparing block copolymers containing
156
D. B. PRIDDY
polycarbonate blocks. These materials have potential utility as compatibilizers for blends of polycarbonate with other polymers. Therefore, Priddy et al. set out to prepare polycarbonate-containing block copolymers utilizing NMRP techniques [18]. The approach is shown in Scheme 8.6. Polystyrene was prepared using the difunctional alkoxyamine initiator XIV. Since XIV contains a carbonate linkage at its center, the resulting polystyrene has a carbonate linkage in the center of the chain. Hydrolysis of the carbonate linkage yields polystyrene having a phenolic group on one end. Phenols are the most common chain-stoppers for the manufacture of polycarbonate. Addition of the phenolterminated polystyrene as a macro chain-stopper to the polycarbonate (PC) process led to the formation of S-PC-S block copolymer.
6
PREPARATION OF BLOCK COPOLYMERS VIA SEQUENTIAL ADDITION OF MONOMERS (SAM)
Prior to the development of the 'universal nitroxide' IV, the SAM technique was mainly limited to the preparation of block copolymers containing only vinylaromatic monomers. This is because TEMPO-based NMRP does not work well for other monomers. Although several papers appeared claiming to have successfully prepared block copolymers with acrylates using TEMPO chemistry, it is doubtful that they were very pure.
XIV
Styrene
N-O-PS.
120°C
PS—O—N
O—PC—O
PS—O—N
Scheme 8.6 Preparation of S-PC-S triblock copolymer using phenoxy functional PS prepared using NMRP as a chain-stopper in the PC process
STYRENE BLOCK COPOLYMERS USING NITROXIDE
157
There has been considerable debate over the reason why TEMPO does not work well for acrylates. CNRS researchers utilized NMR and matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) techniques to look at the terminal end-groups on poly(n-butyl acrylate) produced using TEMPO mediation [19]. They found that most of the chains were terminated by to-unsaturation resulting from either elimination or O—H TEMP from the chain-end or disproportionation between TEMPO and the polyradical (Scheme 8,7). The discovery of improved nitroxides IV, VIII, and IX led to the demonstration of a number of block copolymers via SAM. Table 8.1 list examples of block copolymers made using SAM and the type of mediating nitroxide used. The reactivity ratios of comonomer pairs in free radical polymerization are not changed during NMRP [20]. It is interesting that monomers such as MM A that do not give narrow PD polymer during NMRP mediated using TEMPO yield narrow PD copolymers when styrene is added. Examples of monomers that have been statistically copolymerized with styrene using NMRP include acrylonitrile, N, N-dimethylacrylamide, acrylic acid, methyl methacrylate, 2-hydroxyethyl acrylate, and glycidyl acrylate [8]. The ability of free radical chemistry to yield an infinite number of statistically random copolymers with a variety of comonomer pairs, coupled with the living nature of NMRP, provides the opportunity to synthesize a multitude of unique block copolymer compositions, for example block copolymers containing random and gradient copolymer blocks. NMRP can be used to make block copolymers by first polymerizing monomer 1, isolating the nitroxide-terminal polymer and then using it
^. . . Disproportionation
ao/wx,—CH2—f
+
o4
O-nBu
HO-N
)
V
TEMPO-H
Scheme 8.7 Proposed mechanism of termination during NMRP of n-butyl methacrylate using TEMPO
158 Table 8.1
D. B. PRIDDY Examples of block copolymers prepared using the SAM technique
Monomer 1
Monomer 2
Styrene w-Butyl acrylate Styrene
n-Butyl acrylate Styrene n-Butyl acrylate
Styrene Styrene Styrene Styrene Styrene Styrene Styrene Styrene Styrene Styrene Styrene Styrene tert-Eutyl acrylate Isoprene 4-Chloromethylstyrene
Nitroxide mediator
TEMPO [22], IX [13] IV[8], VIII [12] 4-Oxo-TEMPO [22], IV [8], VIII [12] Methyl methacrylate TEMPO [23] Ethyl methacrylate TEMPO [23] TEMPO [23] Octyl methacrylate Vinyl acetate TEMPO [23] N,N-Dimethylacrylamide TEMPO [23] 2-(Dimethylamino)ethyl acrylate TEMPO [23] 4-tert-Butylstyrene Di-tert-Butyl nitroxide [24] 4-Methylstyrene TEMPO [25], IX [13] TEMPO [26] Butadiene TEMPO [26], IV [9] Isoprene TEMPO [27] 4-Acetoxystyrene TEMPO 4-Chloromethylstyrene IV [9] Isoprene IV [9] Styrene TEMPO [28] Styrene
as a macroinitiator to carry out a random or gradient NMRP copolymerization of a binary monomer mixture. If the macroinitiated copolymerization of the binary monomer mixture is stopped at low monomer conversion, the second block is a random copolymer. However, if the monomer conversion is carried to high conversion, the second block is a gradient copolymer. Typically, normal free radical copolymerizations are stopped at low monomer conversion to eliminate copolymer composition drift. The composition drift occurs because the ratio of unreacted monomers is changing with conversion and new chains are always being initiated. Therefore, chains formed at low monomer conversion have a different monomer ratio than chains formed at high monomer conversion. However, since NMRP virtually eliminates termination processes, the second copolymer block on all of the macroinitiated chains have the same composition. As the ratio of unreacted monomers changes with conversion, a copolymer composition gradient from initiated to terminal end takes place. For example, if NMRP of a mixture of 10mol% acrylonitrile and 90mol% Styrene
STYRENE BLOCK COPOLYMERS USING NITROXIDE
159
is macroinitiated with polystyryl-TEMPO, and the monomer conversion is only carried to 10%, an S-bl-S-stat-AN is formed. However, if the monomer conversion is carried to 90%, a S-bl-S-grad-AN is formed. If the reactivity ratios of the two monomers are very low (i.e. « 1) such that an alternating copolymer block is formed (e.g. styrene and maleic anhydride), a triblock polymer can be formed. For example, if NMRP of a mixture of 90mol% styrene and 10mol% maleic anhydride is macroinitiated using polystyrylTEMPO, an alternating block consisting of 1:1 styrene and maleic anhydride is formed. Once all of the maleic anhydride has been consumed, a pure polystyrene block is formed. The result is an S-bl-S-alt-MA-bl-S triblock copolymer.
7
PREPARATION OF BLOCK COPOLYMERS USING MULTIPLE-HEADED INITIATORS
There is a lot of similarity between this concept and the macroinitiator approach in that a macromonomer is first formed by either (1) forming an alkoxyamine end-functional polymer by initiating a polymerization using an alkoxyamine functional initiator, or (2) using a functionalized alkoxyamine to initiate an NMRP resulting in the formation of a polymer that is capable of initiating a polymerization of another monomer by a mechanism other than NMRP. Hawker et al. developed a dual-functional molecule capable of initiating both NMRP and anionic ring-opening polymerization and then used it to demonstrate both approaches just described by using it to make block copolymers building the NMRP block either first or last (Scheme 8.8) [21]. In another example, Yildirim et al. photochemically generated anthracene radical cations in the presence of TEMPO [29]. TEMPO immediately trapped the radical to form the TEMPO-anthracene cation, which was subsequently used to initiate cationic polymerization of cyclohexene oxide (CHOX). The resulting alkoxyamine-functional polycyclohexene oxide was used to macroinitiate styrene polymerization, resulting in the formation of S-bl-CHOX (Scheme 8.9). Puts and Sogah were the first to propose the concept of making block copolymers via NMRP using multiple-headed initiators (Scheme 8.10). The initiator they developed (XV) is capable initiating blocks via four different polymerization mechanisms [30].
D. B. PRIDDY
160
HO
n Caprolactone
Scheme 8.8
Use of double-headed initiator to make block copolymers
161
STYRENE BLOCK COPOLYMERS USING NITROXIDE
Scheme 8.9 NMRP
Double-headed initiator that initiates both cationic polymerization and
Anionic Ring-opening Polymerization
HO
Cationic Ring-opening Polymerization of Oxazolines
Scheme 8.10
Anionic Vinyl Polymerization
Multiple-headed initiator to make copolymers via NMRP
162
D. B. PRIDDY
REFERENCES 1. Solomon, D. H.; Rizzardo, E.; Cacioli, P., Eur. Pat. Appl. 135280, 1985. 2. Georges, M. K.; Veregin, R. P. N.; Hamer, G. K.; Kazmaier, P. M, Macromol. Symp., 1994, 88, 89. 3. Li, I.; Howell, B. A.; Matyjaszewski, K.; Shigemoto, T.; Smith, P. B.; Priddy, D. B., Macromolecules, 1995, 28, 6692. 4. Moffat, K. A.; Hamer, G. K.; Georges, M. K., Macromolecules, 1999, 32, 1004. 5. Zhu, Y.; Li, I. Q.; Howell, B. A.; Priddy, D. B., ACS Symp. Ser., 1998, 685, 214. 6. Hawker, C. J.; Benoil, D.; Rivera, F., Jr; Chaplinski, V.; Nilsen, A.; Braslau, R., Polym. Mater. Sci. Eng., 1999, 80, 90. 7. Benoit, D.; Harth, E.; Helms, B.; Rees, I.; Vestberg, R.; Rodlert, M.; Hawker, C. J., ACS Symp. Ser., 2000, 768, 123. 8. Benoit, D.; Chaplinski, V.; Braslau, R.; Hawker, C. J., J. Am. Chem. Soc., 1999,121, 3904. 9. Benoit, D.; Harth, E.; Fox, P.; Waymouth, R. M.; Hawker, C. J., Macromolecules, 2000, 33, 363. 10. Rodlert, M.; Harth, E.; Rees, I.; Hawker, C., J. Polym. Sci., Polym. Chem. Ed., 2000, 38, 4749. 11. Benoit, D.; Grimaldi, S.; Robin, S.; Finet, J.; Turdo, P.; Gnanou, Y., J. Am. Chem. Soc., 2000, 122, 5929. 12. Robin, S.; Gnanou, Y., ACS Symp. Ser., 2000, 768, 334. 13. Chong, Y. K.; Ercole, F.; Moad, G.; Rizzardo, E.; Thang, S. H.; Anderson, A. G., Macromolecules, 1999, 32, 6895. 14. Kobatake, S.; Harwood, H. J.; Quirk, R. P.; Priddy, D. B., Macromolecules, 1998, 31, 3735. 15. Priddy, D. B.; Li, I. Q., US Patent 5721 320, 1998. 16. Li, I. Q.; Howell, B. A.; Dineen, M. T.; Kastl, P. E.; Lyons, J. W.; Meunier, D. M.; Smith, P. B.; Priddy, D. B., Macromolecules, 1997, 30, 5195. 17. Yoshida, E.; Tanimoto, S., Macromolecules, 1997, 30, 4018. 18. Li, I. Q.; Knauss, D. M.; Gong, Y.; Howell, B. A.; Priddy, D. B., Polym. Prepr. (Am. Chem. Soc, Div. Polym. Chem.), 1999, 40, 383. 19. Burguiere, C.; Dourges, M.-A.; Charleux, B.; Vairon, J.-P., Macromolecules, 1999, 32, 3883. 20. Baumann, M.; Roland, A.-I; Schmidt-Naake, G.; Fischer, H., Macromol. Mater. Eng., 2000, 280, 1. 21. Hawker, C. J.; Hedrick, J. L.; Malmstroem, E. E.; Trollss, M.; Mecerreyes, D.; Moineau, G.; Dubois, P.; Jerome, R., Macromolecules, 1998, 31, 213. 22. Listigovers, N. A.; Georges, M. K.; Odell, P. G.; Keoshkerian, B., Macromolecules, 1996, 29, 8992. 23. Yousi, Z.; Jian, L.; Rongchuan, Z.; Jianliang, Y.; Lizong, D.; Lansun, Z., Macromolecules, 2000, 33, 4745. 24. Jousset, S.; Hammouch, S. O.; Catala, J. M., Macromolecules, 1997, 30, 6685. 25. Li, I. Q.; Howell, B. A.; Koster, R. A.; Priddy, D. B., Macromolecules, 1996, 29, 8554. 26. Georges, M. K.; Hamer, G. K.; Listigovers, N. A., Macromolecules, 1998, 31, 9087. 27. Listigovers, N. A.; Georges, M. K.; Honeyman, C. H., Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.), 1997, 38, 410. 28. Tsoukatos, T.; Pispas, S.; Hadjichristidis, N., Macromolecules, 2000, 33, 9504. 29. Yildirim, T. G.; Hepuzer, Y.; Hizal, G.; Yagci, Y., Polymer, 1999, 40, 3885. 30. Puts, R.; Sogah, D., Macromolecules, 1997, 30, 7050.
P A R T III
This page intentionally left blank
(EPS) ROLF-DIETER KLODT Dow Central Germany, Schkopau, Germany
BRAD GOUGEON The Dow Chemical Company, Midland, Ml, USA
1
INTRODUCTION
Foamable polystyrene beads are generally produced by two basic processes: 1. polymerization of styrene in suspension into spherical beads containing a blowing agent, and its finishing in a multi-step process (Figure 9.1); 2. incorporation of a blowing agent during the extrusion process of bulk polystyrene, with the polymer strands quenched in a water bath to avoid foaming and consequent strand cutting. The resulting raw material, in the form of beads or granules, is called expandable polystyrene (EPS). The usual procedure is for the EPS beads or granules to be prepared in one location and transported to other locations, where they are expanded and/or molded into their final forms. This process has inherent advantages: the costs of shipping the voluminous foam is minimized and intricate molding shapes can be molded directly without postprocessing. The expandable polystyrene particles, based on suspension polymerization, are converted into foam in three steps: pre-foaming, temporary storage and final Modern Styrenic Polymers: Polystyrene and Styrenic Copolymers. Edited by J. Scheirs and D. B. Priddy © 2003 John Wiley & Sons Ltd
166
R.-D. KLODT AND B. GOUGEON
solid-liquid separation drying •
Expandable PolyStyrene beads
Figure 9.1 Schematic Diagram of the production of EPS and particle foam
foaming (Figure 9.1). The main applications for EPS particle foam based on suspension polymerization are in thermal insulation and in the packaging sector. Extruded EPS is normally processed directly into loose-fill material for packaging applications. The world-wide consumption of EPS is currently around 2.35 million tonnes per year (based on 1999). An annual increase in production of around 4% is forecast until 2005.
2
EPS BASED ON SUSPENSION POLYMERIZATION
2.1
2.1.1
PRODUCTION OF EPS RAW MATERIAL
General Description of Suspension Polymerization
Spherical beads that can be expanded into foam under the influence of heat or steam are produced directly by suspension polymerization in the presence of blowing agent. The term suspension polymerization describes a process in which water-insoluble monomers are dispersed as liquid droplets with suspension stabilizer and vigorous stirring to produce polymer particles as a dispersed solid phase. Initiators used in suspension polymerization are oil-soluble. The polymerization takes place within the monomer droplets. The kinetic mechanism of the suspension process is considered to be a free radical, water-cooled 'microbulk' polymerization [1]. Suspension stabilizing agents are present in the suspension to obtain and stabilize a desired droplet distribution of the dispersed phase. The suspension stabilizer has to be soluble or wetted in/by water. The particle size can cover
167
PARTICLE FOAM BASED ON EXPANDABLE POLYSTYRENE
diameters between 10 |xm and several millimeters. Commercial EPS is focused on a particle size range of 0.1–2 mm, preferably 0.4-1.6 mm. Suspension polymerization has the following characteristics: relatively easy heat removal and temperature control, low dispersion viscosity over the whole conversion, relatively low level of impurities in the finished product and fairly low separation costs. The greatest advantage, however, is to achieve a final expandable product in particle form, with special requirements for the morphology, which is simple to isolate, something which cannot be achieved by the other polymerization technologies.
2.1.2
The Technological Steps for the Manufacture of EPS
The suspension polymerization for the production of EPS is carried out in a jacket batch reactor with a stirrer and usually two or more baffles. The volume of the vessels is normally between 20 and 100m3. At the start of the production process, the water phase and the monomer phase are placed in the vessel, and additives are added either before or during the polymerization. These are typically water- and monomer-soluble components dissolved or dispersed in separate vessels before the start of reaction (Figure 9.2). The monomer/water phase ratio usually lies between 40:60 and 60:40. The filled reactor is heated to reaction temperature. In general, the polymerization is carried out in more temperature steps, gradually increasing the temperature [2]. During the free-radical polymerization a blowing agent (pentane) is added under pressure. Suspending agents Styrene
Water
^ Screening
Additives
tion
\/
hopper sr
X*" ~f
Hopper
, YY Y
ating
1
1
1
Reactor
UD OJ
Figure 9.2
Technological steps for the manufacture of EPS beads
168
R.-D. KLODTAND B. GOUGEON
After conversion of the styrene monomer droplets to EPS beads, the reactor is cooled and the suspension is transferred, usually into a stirred mixing vessel. The further finishing process is normally continuous. The expandable beads are separated from water by centrifuges or rotating sieves. In different methods, EPS beads are then washed and partly pretreated. They are then dried using flash, fluid bed-, or/and silo dryers [3], sieved, screened in various bead size fractions and coated, depending on their size and application. The coating is normally applied in batch or continuous blenders using solid or fluid coating materials. The sizes of the beads and the coatings used affect the processing properties of the EPS foams produced from EPS beads. The different EPS types are separately packed, usually in octabins, IBCs or silo trucks for use by EPS foam manufacturers. 2.1.3
Polymerization and Impregnation Process
2.1.3.1
Polymerization
Usually the polymerization process is carried out in two or more stages. During the first stage, the final particles are formed, and in the second stage blowing agent is added and this penetrates into the beads. The duration of the second step is preferably determined by the residual monomer concentration that needs to be achieved (normally 2) also results in a more desirable molecular weight distribution with a flat, low Mw side and a sharply decreasing high Mw flank. The procedure of later addition of the chain transfer agent before the gel effect takes off is preferred. A low concentration of branched macromolecules should result in a higher melt flow rate during the pre-foaming step of the expandable beads [12,13]. In the literature, the procedure of earlier addition of blowing agent is described to reduce the viscosity of the droplets. Hamielec and co-workers [16] pointed to the fact that temperature increase and, simultaneously, earlier addition of
M-
CH3 ^ ^ CH ru C— 3 CH,3 \^
*-°TrnrO~».H—O -C02 Q II -,CH3 — C — CH3+CH3 « CH I
\^ CH 3 -c-0« CH3
- 60 m3 to achieve a standard deviation of particle size distributions of < 25 %. An independent claim is included for a hydrodynamic system, where the ratios of stirrer diameter to
PARTICLE FOAM BASED ON EXPANDABLE POLYSTYRENE
177
reactor diameter and of stirrer surface blade height to stirrer and reactor diameter have to have definite dimensions. The aim is to reach ratios as low as possible between surface velocity, shear rate and pumping rate. Through sudden or stepwise reductions of stirrer speed during the polymerization, it should be possible to obtain narrower PSDs using the pickering stabilizer system [66]. In addition, a number of co-modifiers have been developed as replacements for, or in addition to, the standard extender of ABS type, to improve the particle size distribution. Sodium or potassium persulfate, sodium bisulfite, partly combined with hydroxy ceto compounds, unsaturated carboxy acids [73–75], small quantities of water-soluble sodium polystyrene sulfate [76] and/or copolymers of acrylic acid and 2-ethylhexyl acrylate [10] have been suggested as extenders for narrower PSD. Marclay [77] discussed the possible mechanism of extender action by potassium persulfate in suspension polymerization. Further suggestions for narrower particle size distribution, improved suspension stability and more accurate particle size control are combinations of water-insoluble pickering stabilizer of TCP type with compounds such as ethylenediaminetetraacetate (EDTA) and its salts [78], EDTA, CaCO3 and hydroxyethylcellulose [79], and sodium ^-naphthalene sulfonate and sodium polyacrylate [80]. A narrower PSD should also be achieved if the inorganic pickering salt is precipitated immediately before its use [2,128]. Vilchis et al. [81] presented a new idea to achieve better control of the particle size distribution by the synthesis in situ of a water-soluble copolymer of acrylic acid–styrene as suspension stabilizer without additional inorganic phosphate. Publications describe increasing the particle formation by using a physical (population balance, Maxwell fluid, power law viscosity, compartment mixing) modeling approach [22,60,98,105].
2.1.5
Additives
During or at the end of the EPS production, a number of additives can be incorporated to improve process and application properties. Additives can include nucleation agents, flame retardants, fast-cool agents, anti-lump and anti-static agents, stabilizers, plasticizers, pigments, etc. Some of the more important additives and their functions are described in more detail below.
2.1.5.1
Nucleation Agents
Nucleation agents are substances that are able to initiate and control cell formation and growth. They are purposely incorporated into the polymeric structure. At their location, the blowing agent (pentane) preferably evaporates when the EPS beads are exposed to saturated steam. However, unwanted
R.-D. KLODT AND B. GOUGEON
178
nucleation effects can arise from incorporated water, suspension stabilizer or hexabromocyclodecane (HBCD) [82–85]. Nucleation agents must be able to control nucleation independent of other effects and must have the ability to mask the bubble initiation of other sources. Waxes such as paraffins, chloroparaffins, and Fischer–Tropsch waxes and also esters and amides of fatty acids have been described [86–89] for this function. Finely divided phase-incompatible polymers, such as low molecular weight polyethylene (PE), are normally used in industry for cell size control [90,92–95]. Above its melting point, low molecular weight PE will dissolve in styrene. With growing conversion phase, inversion occurs and the domain size of polyethylene is set after the cooling step. The PE domains operate as gathering sites for the blowing agents, which are be able to form bubbles when treatment with heated steam occurs. The molecular weight, the branching degree, the concentration and the degree of distribution of PE determine the weak point of critical size in the polymer matrix resulting in the maximum diffusion path length. In general, the unbranched PE types with a degree of crystallinity >80% are, as previously described, suitable because of their higher incompatibility with polystyrene [90] (Figure 9.9). Investigations by Klodt et al. [91] showed that branched PE waxes with degrees of branching up to 1.5 and crystallinity above 40% can also be successfully used for the production of uniform cell size distribution (Figure 9.10). If the compatibility between PE and PS is further increased, for instance by increasing the degree of branching of PE or by
>/•>
vD
OO
O
—
«N
(N
30g/l) require temperature control by air addition. For low bulk densities (special,
\
90 •
« £ 85 . c •* "3 "~ 80 •
s.i
r./ s /
75
> £ .£ b 70 .
o
U
y
60
^^
-— m
// ^
s^
s
2 e 200 •y, .g
i I
Jff*
Q. O 1 00
so °
O
,/
S
^/
.2 § 150
/
'_/
5Q
^
?*•* ^ [
/
50
n. 0
(a )
s \s
o
10
20
30 40 50 60 Fusion degree ( %)
70
8
(b)
20 25 30 Foam density (g/1)
Figure 9.16 Compressive strain at 10% compression as function of (a) the degree of fusion (foam density 17 g/l) and (b) foam density at optimal fusion
2.3.2.2
Flexural Strength
The bending test shows a steady course of the compressive force-bending deflection, even though the relationship is not linear. At low loads, the foamed PS material behaves evenly (linear increase in stress-strain) and up to halfrupture there are only small deviations from a linear course. The flexural strength of foamed PS materials increases with increasing density. The degree of heat-sealing affects the flexural strength. The flexural strength is lower for poorly heat-sealed foamed PS materials.
2.3.2.3
Tensile Strength, Transverse Tensile Strength
The tensile test shows a steady course of the tensile force–elongation curve, where a linear relationship exists only at low tensile forces. There are only small deviations from a linear course at approximately half-rupture. The tensile strength increases with increasing density. The degree of heat-sealing has an effect here also.
2.3.4.3
Acoustic Properties
Owing to their dense, closed surface, untreated foamed PS slabs are acoustically ineffective. Small absorption effects can be achieved by slotting, perforating or corrugating the surfaces. Foamed PS slabs are suitable for footfall-sound insulation only if the dynamic E modulus is lowered by elastification.
188
2.4
R.-D. KLODT AND B. GOUGEON
APPLICATIONS
EPS foam has, like XPS, a closed cellular structure (Figure 9.17), but the production of lower densities is possible. The usual density range is 10–80 g/l (0.6–5.0 lb/ft3). Densities of 10g/l, or lower, are possible after several prefoaming operations.
2.4.1
Construction Industry
EPS foam is used in many building projects for thermal insulation and also, more and more often, for soundproofing in new buildings and modernization or renovation work. EPS foam slabs are used for the insulation of walls, roofs, floors and ceilings. Particle sizes between 0.9 and 1.6 mm are preferably used for this application. For the thermal insulation of walls, there is a difference between outside wall insulation, inside wall insulation, and core insulation (European applications). For the outside wall insulation the EPS foam is put directly on to the stone bearing structure. A fabric reinforced plastering or a ventilated facing protects it from the weather. Using sandwich panels of EPS plasterboard, modern heat insulation standards can be achieved on the walls of older buildings. For core insulation, the insulation layer is between the bearing wall and the external weather resistant wall. Finally, sound insulation is becoming increasingly important. Special elasticated slabs combine sound and thermal insulation.
Figure 9.17 scales
Comparison of EPS particle foam and XPS foam structure on different
189
PARTICLE FOAM BASED ON EXPANDABLE POLYSTYRENE
Another system of insulation is the use of EPS molded foam parts (insulated concrete forms) for a combination of outer and inner wall insulation. A wall is built with these molded foam parts and then filled with concrete. For the thermal insulation of roofs, there is a difference between flat roof and steep roof insulation. The insulation of a nonventilated flat roof is done with EPS insulation elements, which could, also, be pre-laminated with a sealing strip. Steep roof insulation is used in attic conversions for residential purposes. There is a difference between insulation under, on and between the rafters. For the between-rafter insulation, different systems are used to make installation very easy. Slabs from the Styrotect [125] system, tongued and grooved, are cut to size, fitted together and laid between the rafters. Also special wedge systems or the use of elasticated slabs are possible. For floor and ceiling insulation, different thermal and sound protection requirements have to be met. The priority for the insulation material of flat ceilings is to provide the necessary impact sound insulation protection. Impact sound insulation slabs are produced by a special manufacturing process [126]. In two or three pre-foaming steps the EPS beads are pre-foamed to a density of 10 g/l or lower. After molding, the formed blocks receive a special follow-up treatment. Before the EPS foam blocks are cut into slabs, the blocks are pressed in special presses to more than half their size and then immediately the load is removed. Because of the destruction of the cell structure in one direction, special elastic impact sound insulation properties are achieved (low dynamic stiffness with acceptable compressibility of the slabs, Figure 9.18). For decorative ceiling layouts and for echo regulation, specially produced slabs (pressed in design) or molded foam parts are available.
«s' 17/15 OS' 22/20 AS' 27/25 • s' 32/30 AS' 33/30 os' 38/35 • s'43/40
3 4 Compressibility (mm)
Figure 9.18 Dynamic stiffness as a function of compressibility of impact sound insulation slabs depends on its dimensions
190
R.-D. KLODT AND B. GOUGEON
Foam molding machine-produced perimeter insulation slabs (density 30 g/l or 1.85 lb/m3) are used for the insulation of outside walls and ground with soil contact, pressure load and moisture requirements. The good properties of molded EPS perimeter slabs (normally a domain of XPS) are achieved by means of hydrophobic coated small particles with very good pre-foaming behavior. For drainage purposes, bitumen banded EPS foam slabs are used. These slabs consist of poorly fused large particles resulting in large channels, where the water can drain away. 2.4.2
Packaging Materials
Foaming in foam molding machines produces packaging materials of EPS foam. In general, small EPS particles (0.4–0.9 mm) are used because the foam molding parts produced can have very complicated shapes and structures. Shapes with thin walls and bridges are also possible. Packaging materials using EPS foam protect packaged goods in storage locations and during transport against mechanical and thermal damage. The advantages over other packing materials are their low weight, resistance to water, good shock absorption behavior, excellent heat/cold insulation and easy processing and handling. Application examples are packaging materials for machines, machines parts, glass, china, optical, electrical and electronic equipment, toys and Pharmaceuticals, and also for edge and surface protection. Food packaging is possible, although, in this case, a special type of EPS with food certification is required. 2.4.3
Other Applications
Because of its diverse properties, EPS foam can be used in many other applications. The most important are: • • • • • • • • 3
additions for the production of lightweight building materials; bath support; safety helmets; foam cups; insulation of cooling equipment and cold boxes; life jackets, life buoys, fenders; plant containers, flower tubs; technical applications, for example full form molding. EPS BASED ON EXTRUSION PROCESS
For many years, wood excelsior and paper products dominated the dunnage packaging market. Dunnage here being defined as a packaging material that
PARTICLE FOAM BASED ON EXPANDABLE POLYSTYRENE
191
takes up space and resists the movement of packaged items in a box. As early as the 1960s, it was determined that small-sized EPS particles had two properties that would be highly desirable as a dunnage packing material. The low bulk density (product density plus airspace around the pieces) of shaped polystyrene foam material lowered the total package weight and thus shipping costs. The frictional resistance to movement of the foam particles, when tightly packed, greatly reduced the likelihood that packaged items might shift within the box, reducing the chance of breakage during transit. The Dow Chemical Company first entered the loose-fill packaging market in 1962 with a material that resembled spaghetti strands. Eventually the shape evolved to the 'S' shape that characterized the product from the early 1970s to the present. Other competitors in the polystyrene foam loosefill market include Flo-Pak, manufactured by Free-Flow Packaging, InterPac, manufactured by Inter-Pac, WingPac and C-Pac, manufactured by Rapac, and Alta-Pak, manufactured by Storopack. In 1993, Dow sold the trademark rights to the 'S' shape and Pelaspan-Pac to Storopack. The information about foamed polystyrene loose-fill that follows is an overview specific to materials that are formed in hard resin strands, cut to length, boxed and shipped to and later expanded for customer use at a convertor (expander) location. 3.1
3.1.1
EXTRUSION
Raw Materials
The recipe for EPS loose-fill resin is similar to that of extruded polystyrene foam insulation: polystyrene with about a 200 000 weight-average molecular weight, hexabromocyclododecane (HBCD) as fire retardant, magnesium oxide as oxygen scavenger, calcium stearate as lubricant and n-pentane as blowing agent. Different blowing agents of the CFC and HCFC types have been used in the past but were discontinued as a result of the environmental and economic requirements placed on the product.
3.1.2
Extrusion System
The solids are metered into the extruder screw by means of an accurate rate feeder. The blowing agent enters the mixture downstream of the screw but before the initial mixer. The polymer blend is pushed through a series of heat exchangers and mixers, a gel screen,and through a multi-hole die. The die may have up to 96 holes or individual strands but smaller number was found to be optimal owing to consistency in strand temperature at the outer edges. Figure 9.19 shows the equipment layout through the die.
192
R.-D. KLODT AND B. GOUGEON Accurate rate solids feeder
Blowing agent
Extruder screw
Gel screen
In-line mixer
Cooler
Cooler
-*
Cooler
Figure 9.19
3.2
Schematic diagram of extrusion process
POST EXTRUSION
After the strands leave the die, they are run through a cold water bath to solidify the polystyrene. The strands are then run up through a helical rotary cutter that cleaves the strands to a length of approximately 6–9mm. The resultant 'beads' or 'pellets' are passed through an annealing unit and flash cooled to set nucleation sites and cell size. The beads next go through a filtration system to remove chips, fines and oversize pieces (overs/unders) resulting from the cutting operation. After a dewatering step, the beads are air conveyed to holding silos for an 8–12 h period. Prior to packaging, the beads are run through a six-pass crossflow air declassifier which removes over 98% of the remaining fines. The beads are then transferred to a loading hopper where calcium stearate (antidumping agent) and an anti-static agent are applied to the beads prior to packaging. In the USA, the beads are packaged in 1000 lb net weight 'Gaylords', which are large corrugated boxes set on a wood pallet. The boxes are lined with a heavy-duty polyethylene bag, which, after being filled, is sealed at the top to prevent escape of blowing agent from the container and the beads. Work done in the late 1980s [132] showed that there are many variables to achieving a perfect cut on a strand in this process. The following are the equipment and polymer conditions that must be considered:
PARTICLE FOAM BASED ON EXPANDABLE POLYSTYRENE
• • • • • • • • • • • • •
193
die land length; die internal land angle; strand temperature at the die; variations in strand temperatures from center die openings to perimeter die openings; number of die openings (optimum); distance of the cooling water bath from the die; composition and hardness of pull rollers; external 'texture' of pull roller; variations in temperature at the cutter; cutter blade (helix) angle; rake angle/relief angle in the cutter; cutter gap distance; ability to discharge pellets cleanly from the cutter.
Failure to optimize one or more of these parameters can result in cuts that produce broken pieces or appendages to the pellet, which, when expanded, break off and create dust and partial pieces. This situation is difficult to rectify at a convertor location or at a customer's facility and results in a continuous, undesirable, housekeeping problem. Figure 9.20 shows an equipment layout from the water bath through the packaging station. Water bath
Helical cutter
Annealer Cold water
Surface lubricant Anti-static agent t
Packaging
Figure 9.20
Schematic diagram of post-extrusion process
194
3.3
R.-D. KLODT AND B. GOUGEON
STEAM EXPANSION OF EPS LOOSE-FILL RESIN: THEORY AND PRACTICE
EPS loose-fill resin expands under much the same conditions as EPS molding bead resin. The major difference is the orientation and size of the expander. Loose-fill expanders are generally horizontal and vary in size from 3 ft (laboratory size) to in excess of 20 ft in length. Unlike vertical bead expanders, loose-fill expanders rotate the entire drum. Loose-fill expanders do not have internal mixing arms attached to a shaft but, instead, have baffles attached to the interior wall of the rotating drum. These baffles are useful in maintaining the proper retention time inside the drum. They can be either fixed or adjustable. In addition to drum length and baffles, four other parameters influence the residence time in the expander: feed rate, drum pitch or angle from horizontal, drum rotation speed, and steam volume and pressure. Even under the best machine and operating conditions, unexpanded pieces of loose-fill entering the expander can be expected to exit at different times rather than moving through the expander at a constant rate. The output of the pieces follows a bell-shaped curve with the peak output occurring at about 4 min with a commercial size expander [133]. Polystyrene loose-fill normally undergoes three to five expansions with four being the norm. This will take the loose-fill from a bulk density of about 38–40 lb/ft 3 (pcf) (600–640 kg/m3) as an unexpanded pellet to about 0.20– 0.25 pcf (3.2–4 kg/m3) after four expansions or passes, as they are commonly called. Between expansions, a recovery or aging period of 16–24 h is required to refill the cells with air. For the first expansion pass, the unexpanded polystyrene loose-fill pellets are metered into one end of the horizontal expander at a rate of about 1000 lb (454 kg) per hour. This time can vary from 50 to 70 min and is dependent on the equipment and conditions listed above and on the bulk density that is achieved. Normally, the first-pass density aim is in the 0.80– 0.90 pcf (12.8–14.4 kg/m3) range with little (< 5%) shrink after cooling. Shrink or shrinkback is caused by the cooling of gases inside the cells of the foam. The cell walls are stretched during expansion by the heat and expansion of gases (blowing agent) inside the cells. When the cell walls are over-expanded and partially collapse during the cooling stage, this is shrink. A certain amount of shrink is acceptable as long as no cell walls are ruptured and the volume of gases filling the cells between expansions (aging period) is equal to the volume of gases that was in the cells at the peak of expansion. At the outfall of the expander, the first-pass loose-fill is air conveyed through ducts to a large ventilated holding bag or hopper. After the first aging is completed, the once expanded material is air conveyed back to an expander equipped with a large throat or opening to feed the expanded particles into the expander. Conditions are set so that the material exiting the expander will have a bulk density of about 0.38–0.42 pcf (6.0–6.7 kg/m3) with no more than 15% shrink after
195
PARTICLE FOAM BASED ON EXPANDABLE POLYSTYRENE
1 min. Third-pass fresh bulk densities should range from 0.26 to 0.30 pcf (4.2–4.8 kg/m3) with no more than 20% shrink. Fourth-pass fresh bulk densities should be in the 0.22–0.25 pcf (3.5–4.0 kg/m 3 ) range with no more than 25% shrink. A typical loose-full expansion setup is shown in Figure 9.21. Conditions for expansion and aging will vary significantly between sea level and higher altitudes such as found in Colorado Springs, Colorado (altitude 6035 ft/1839m). Because of the difference in boiling points, more steam and longer expansion times are required to achieve densities equivalent to those easily attainable at sea level to 1000 ft in altitude. First pass aging bag
Second pass aging bags
Return line to expander
I
^ ~4 *t
* i
Fill line to aging bags
/
Expander
Fill line to aging bags
Gaylord of unexpanded resin (first pass)
\
.*-•
Fourth pass assx\
i / a aging g bags \f
\
Figure 9.21
^
To pack stati
Typical loose-fill expansion setup
Third pass aging bags
196
R.-D. KLODT AND B. GOUGEON
Aging conditions will affect the rate at which the cells recover. Temperatures below 60 °F will slow the process so that more time is required or the convertor may have to accept higher densities on subsequent passes. The size of the aging bags may also affect the recovery rate. Larger bags, filled to the top, will restrain the expanded pieces as they recover and attain their former size. This will also inhibit air exchange for pieces at the center of the bag. A good firstpass bag (for 1000 lb of unexpanded beads) should have a volume of at least 1500 ft3. Other things that will affect ultimate density and appearance of the loose-fill are related to the air conveying (airveying) system used to move the loose-fill around the expansion system. Recently expanded loose-fill pieces need to be moved with a certain amount of care as physical abuse will crush cell walls and compress the loose-fill piece to an extent which it will not recover in the aging process. To this end, the airveying system must be engineered and constructed to minimize 90° angles, especially closely downstream from the blowers in the system where the air velocity is the greatest. The airveying system must also be constructed to minimize sharp edges, protrusions into the ducts, and other details that will create fines, broken pieces, and dust that are normally unacceptable to end-use customers. Loose-fill is either bagged or shipped in bulk truck to the customer. Bags vary in size from 10 to 20 ft 3 , depending on the customer need and the particular geographic market. Bulk systems are popular with companies that use large quantities of loose-fill and have many packaging stations. The trucks used to transport the loose-fill are specially equipped with flexible duct that connects to the duct and bagging system at the customer. This way, 2500–3000 ft3 of loosefill can be shipped to the customer at short notice and unloaded with a minimal amount of manpower involved.
3.3.1
Photodegradable Loose-Fill
In the late 1980s, environmental concerns led to the development of photodegradable resins being incorporated in the polymer mix. Methyl isopropenyl ketone (MIPK) at a level of 5 % was found to be effective in making the loosefill pieces photodegrade to polymer 'dust' with 10–12 months of exposure to sunlight [134].
3.3.2
Physical Properties
Polystyrene loose-fill packaging material used in US Federal Government applications must meet US Government Specification PPP-C-1683 (12/5/ 1988), 'Cushioning Material, Expanded Polystyrene Loose-Fill Bulk (for Packaging Application)'. Compliance with eight properties is the basis for this
PARTICLE FOAM BASED ON EXPANDABLE POLYSTYRENE
197
specification: flowability, vibrational settling (at two different loadings), electrostatic adhesion, dynamic cushioning, loaded bulk density, compressive creep after 7 days (three different loadings), compressive set after 24 h (same three loadings) and flammability.
3.3.3 End Uses Expanded polystyrene loose-fill is used in many markets today. The most common are dunnage (used to fill up space in a shipping box), cushioning (where the resilient qualities of the polystyrene foam are used), and texturizer, where it is ground into smaller particles and used in making textured ceilings in commercial and residential buildings.
REFERENCES 1. Munzer, M., Trommsdorff, E., Polymerizations in suspension, High Polym. 29 (1977) (Polymer Processes) 106. 2. Bilgic, T., Karali, M., Savasci, O. T., Effect of the particle size of the solid protective agent tricalcium phosphate and in-situ formation on the particle size of suspension polystyrene. Angew. Makromole. Chem. 213 (1993) 33. 3. JP 09059416 A2 (Achilles Corporation), 1995. 4. Riederle, K., Die Oligomerenbildung bei der thermischen Polymerisation von Styrol bis zu hohen Umsatzen des Monomeren, Dissertation, TU Munchen (1981). 5. Mayo, F. R., The dimerization of styrene, J. Am. Chem. Soc., 90 (1968) 1289. 6. Stein, D. J., Mosthaf, H., Oligomer formation in the thermal polymerization of styrene, Angew. Makromol. Chem. 2 (1968) 39. 7. Hohwiller, H., Neopor - a new EPS generation, in Particle Foam 2000, VDI Verlag, Dusseldorf (2000), 375-383. 8. Trommsdorf, E., Kohle, H., Lagally, P., Polymerization of methyl methacrylates, Makromol. Chem. 1 (1948) 169. 9. Mitta, I., Horie, K. J., Diffusion-controlled reactions in polymer systems, Makromol. Sci. Rev. Macromol. Chem. Phys. C27 (1987) 91. 10. Ethapol 1000 – Operating Manuals - Suspending Agent for Styrenic Polymerization Processes, CIRS, Padova, 1996. 11. AU 8661 177 (The Dow Chemical Company), 1985. 12. US 4652 609 (Atlantic Richfield Company), 1985. 13. US 4661 564 (Atlantic Richfield Company), 1985. 14. US 2857 342 (Monsanto Corporation), 1958. 15. US 3 072 581 (Monsanto Corporation), 1963. 16. Villalobos, M. A., Hamielec, A. E., Wood, P. E., Bulk and suspension polymerization of styrene in presence of n-pentane: an evaluation of monofunctional and bifunctional initiation, J. Appl. Polym. Sci. 50 (1993) 327. 17. Moritz, H. U., Influence of bifunctional initiators, DECHEMA Monogr. 131 (1995) 259. 18. EP 0 574 665 B1 (Huls AG), 1992.
198
R.-D. KLODT AND B. GOUGEON
19. Malta, A., More efficient initiators for the production of styrene (co)polymers, Organic Peroxides Symposium, Akzo Nobel, Conference Proceedings 8 (1998) 1. 20. Technical Documentation. Better Performance in Suspension Polystyrene, Elf Atochem, Paris, 1996, 1-22. 21. EP 0 488 040 (BASF), 1990. 22. Chen, Z., Pauer, W., Moritz, H.-U., Pruss, J., Warnecke, H. J., Modeling of the suspension polymerization process using a particle population balance, Chem. Eng. Technol. 22 (1999) 699. 23. EP 0 575 871 (BASF), 1992. 24. Konno, M., Arai, K., Seito, S., The effect of stabilizer on coalescence of dispersed drops in suspension polymerization of styrene, J. Chem. Eng. Jpn., 15 (1982) 131. 25. Ingram, A. R., Wright H. A., Composition of foam structures from expandable polystyrene beads, Mod. Plast. 41 (1963) 152, 156, 200, 203. 26. Villalobos, M. A., Diffusion of blowing agent into PS beads, Report PPR-01 to Plasti Fab Ltd, Calgary (1991). 27. Klodt, R. D., Wolfahrt, Ch., Blase, H., Hamann, B., Study on interaction of n-pentane with polystyrene in the polymerizing system, Conference Proceedings, Fachtagung Polymerwerkstoffe Merseburg (1994). 28. Busse, M., Hahn, O., Production of complex EPS moldings with defined properties, Plastverarbeiter 44 (1993) 46. 29. Ingram, A. R., J. Cell. Plast. 1 (1965) 69. 30. Matsumoto, S., Takeshita, K., Koga, J., Takashima, Y. J., A production process for uniform-size polymer particles, Chem. Eng. Jpn. 22 (1989) 691. 31. Polacco, G., Semino D., Rizzo C., Feasibility of methyl methacrylate polymerization for bone cement by suspension polymerization in a gel phase, J. Mater. Sci.: Mater. Med. 5 (1994) 587. 32. Hatate, Y., Uemura,Y., Ijichi, K., Kato, Y., Hano, T., Baba,Y. Kawano, Y., Preparation of GPC packed polymer beads by SPG membrane emulsifier, J. Chem. Eng. Jpn. 28 (1995) 656. 33. JP 96-340879 (Hitachi Chemical Co.), 1998. 34. DE 19 650 301 (The Dow Chemical Company), 1966. 35. JP 279602 (Hitachi Chemical Company), 1992. 36. Ahmed, S. M., Effects of agitation, and the nature of protective colloid on particle size during suspension polymerization, J. Dispers. Sci. Technol. 5 (1984) 421. 37. DE 75-2548524 (BASF), 1977. 38. EP 0 425 992 (BASF), 1989. 39. DE 3331 570 (Huls AG), 1983. 40. US 4 042 541 (The Dow Chemical Company), 1976. 41. Goodall, A. R., Greenhill-Hopper, M. J., Characterization of partially hydrolyzed poly(vinyl acetates) for use as stabilizers in suspension polymerization, Macromol. Chem., Macromol. Symp. 35–36 (1990) 499. 42. Fabini, M., Bobala, S., Rusina, M., Macho, V., Preparation of poly(vinyl alcohol) as the dispersant for suspension vinyl chloride polymerizations, Polymer 35 (1994) 2201. 43. Mendizabal, E., Castellanos-Ortega, J. R., Puig, J. E., A method for selecting a poly(vinyl) alcohol as stabilizer in suspension polymerization, Colloids Surf. 63 (1992) 209. 44. EP 0 732 343 A2 (Mitsubishi Chemical BASF Company), 1996. 45. Gotze, Th., Sonntag, H., Forces between quartz surfaces bearing adsorbed macromolecules in good solvents, Colloids Surf. 31 (1988) 181. 46. Hesselink, F.Th., On the theory of stabilization of dispersions with adsorbed polymer, J. Polym. Sci. Polym. Symp. 61 (1977) 439.
PARTICLE FOAM BASED ON EXPANDABLE POLYSTYRENE
199
47. EP 96-201906 (Royal Dutch Shell, Ltd), 1998. 48. DE 3 331 570 (Huls AG), 1983. 49. Klodt, R.-D., Wustneck, R., Ruhle, R., Thummler, W., Investigations on the interfacial rheological behavior of methyhydroxypropylcellulose, Abh. Akad. Wiss., N1 (1986) 257. 50. FR 76–32736 (BASF), 1978. 51. GB 97-21603 (Dyno Industries ASA), 1997. 52. DE 4 029 298 (Huls AG), 1991. 53. Witt, M., Dissertation, Technische Universitat Munchen (1989). 54. Wolters, D., Meyer-Zaika, W., Bandermann, F., Suspension of polymerization of styrene with pickering emulsifiers, Macromol. Mater. Eng. 286 (2001) 94. 55. Deslandes, Y., Morphology of hydroxyapatite as suspension stabilizer in the polymerization of poly(styrene-co-butadiene), J. Appl. Polym. Sci. 34 (1987) 2249. BE 0700533 (The Dow Chemical Co.), 1966. US 4303 784 (Atlantic Richfield Company), 1980. EP 0 575 872 (BASF), 1992. CS 70–5854 (Czech), 1978. Vivaldo-Lima, E., Wood, P. E., Hamielec, A. E., An updated review on suspension polymerization. Ind. Eng. Chem. Res. 36 (1997) 939. 61. Apostilidou, C., Stamatousdis, M., On particle distribution in suspension polymerization of styrene, Collect. Czech. Chem. Commun. 55 (1990) 2244. 62. Stamatousdis, M., Apostolidou, C., Characteristics of particle sizes produced by suspension polymerization of styrene, Part. Part. Syst. Charact., 9 (1992) 151. 63. Tanaka, M., Oshima, E., Dispersing behavior of droplets in suspension polymerization of styrene in a loop reactor, Can. J. Chem. Eng. 66 (1988) 29. 64. Leng, D. E., Guarderer, G. J. Drop dispersion in suspension polymerization, Chem. Eng. Commun. 14 (1982) 177. 65. Schroder, R., Piotrowski, B., On particle formation during suspension polymerization of styrene, Ger. Chem. Eng. 5 (1982) 139. 66. Masato, T., Hideyo, T., Isao, K., Natsakaze, S., Kazuhiko, H., Effect of stepwise and continuous reduction in impeller speed on particle size distributions in suspension polymerization of styrene, J. Chem. Eng. (Jpn.) 21 (1995) 118. 67. Yang, B., Kamidate, Y., Takahashi, K., Takeishi, M., Unsteady stirring method used in suspension polymerization of styrene, J. Appl. Polym. Sci., 78 (2000) 1431. 68. Tanaka, M., Izumi, T., Application of stirred tank reactor equipped with draft tube to suspension polymerization of styrene, J. Chem. Eng. Jpn. 18 (1985) 345. 69. Kuzmanic, N., Mitrovic-Kessler, E., Skansi, D., The influence of mixing on the styrene polymerization product, Chem. Biochem. Q. 6 (1992) 1. 70. DE 19 530 765 Al (BASF), 1997. 71. Tanaka, M., Hosogai, K. Suspension polymerization of styrene with circular loop reactor, J. Appl. Polym. Sci. 39 (1990) 955. 72. DE 19 816 461 C1 (Buna Sow Leuna Olefinverbund GmbH), 1998. 73. US 3 631 014 (Sinclair-Koppers), 1971. 74. JP 08301905 A2 (Sekisui Plastics), 1995. 75. US 3 755 282 (Sinclair-Koppers), 1973. 76. US 4 500 692 (Atlantic Richfield Company), 1985. 77. Maclay, W. N., The mechanism of extender action by potassium persulfate in suspension polymerization of styrene, J. Appl. Polym. Sci. 15, (1971) 867. 78. JP 95–118775 (Sekisui Plastics), 1996. 79. DE 3 728 044 A1 (Huls AG), 1987. 80. DE 19 650 301 (The Dow Chemical Company), 1966.
200
R.-D. KLODT AND B. GOUGEON
81. Vilchis, L., Rios, L., Guyot A., Guillot, J., Villalobos, M. A., In-situ formed copolymer of acrylic acid–styrene as stabilizer during suspension polymerization of styrene, DECHEMA Monogr. 134 (1998) 249. 82. DE 0142193 (Buna Schkopau), 1979. 83. Hinselmann, K., Physical laws in the pre-foaming of expandable polystyrene, Kunststoffe 61 (1971) 152. 84. BE 0 834 383 (BASF), 1974. 85. FR 1 493 947 (Monsanto Corporation), 1965. 86. EP 0000 120 (BASF), 1977. 87. BE 0815 185 (MonsantoCorporation), 1973. 88. BE 0819 346 (Cosden Technology Co.), 1973. 89. EP 0 396 046 b1 (Kanegafuchi Kagaku Kogyo), 1990. 90. Technical Release, Polywax Polyethylene, Petrolite Speciality Polymers Group, Sugar Land, Tx. 91. Klodt, R.-D., Nieter, E., Kuhnberger, W., Koller, F., Bunge, F., Nukleierungswirkung niedermolekularer Ethylenhomo-und-copolymerer im PS Partikelschaum, Polymerwerkstoffe 2000, Halle, Tagungsband (2000) 456. 92. Becker, G. W., Braun, D., In Kunststoffhandbuch, ed. Gausepohl, G. H., Gellert R., Polystyrol (4), Hanser, Munich, 1996, 579. 93. DE 2 448 476 (BASF), 1974. 94. US 3 503 908 (Koppers Corp.), 1969. 95. EP 0 050 968 (American Hoechst Corporation), 1981. 96. DE 3 843 536 Al (BASF), 1988. 97. Troitzsch, J., Brandverhalten von Kunststoffen, Hanser, Munich, 1982. 98. Vivaldo-Lima, E., Wood, P. E. Hamielec, A. E., Penlides, A., Calculation of the particle size distribution in suspension polymerization using a compartmentmixing model, Can. J. Chem. Eng., 76 495. 99. Troitzsch, J., Kunsstoffe 77 (1987) 1078. 100. Encyclopedia of Polymer Science and Engineering, vol. 16, 2nd edn, Wiley, New York. 101. EP 0 374 812 (BASF), 1988. 102. Hardy, M. L., Rouge, B., presented at 6th European Meeting on Fire Retardancy of Polymeric Materials, Lille, 1997. 103. Eichhorn, J., Synergism of free radical initiators with self-extinguishing additives in vinyl aromatic polymers, J. Appl. Polym. Sci. 8 (1964) 2497. 104. Fenimore, C. P., Inhibition of polystyrene ignition by tris(2,3-dibromopropyl) phosphate and dicumyl peroxide, Combust. Flame 12 (1968) 155. 105. Kalfas, G., Yuigen, H., Ray, H. W., Modeling and experimental stydies aqueus suspension polymerization processes, 2. Experiments in batch reactors, Ind. Eng. Chem. Res. 32 (1993) 1831. 106. Hamann, B., Klodt, R.-D., Gellert, R., Pelzers, T., Langzeit-Bewahrung von PSHartschaum der Baustoffklassen Bl bzw. B2 nach DIN 4102, Bauphvsik 21 (1999) 29. 107. Regulatory Status and Environmental Properties of Brominated Flame Retardants Undergoing Risk Assesment in the EU. 108. Product Bulletin - Fatty Amines, Akzo (1990). 109. Gachter, G., Muller, H., Taschenbuch der Kunsstoff-Additive, 3.Ausg., Hanser (1989). 110. US 3 389 097 (Koppers Corp.), 1964. 111. Informationsschriften Aerosil, Fallungskieselsauren undo Silikate, Degussa (1991). 112. DE 4 123 252 (BASF), 1991.
PARTICLE FOAM BASED ON EXPANDABLE POLYSTYRENE 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135.
201
DE 19816469 Cl (Buna Sow Leuna Olefinverbund GmbH), 1998. JP 72–53828 (Badische Petrochemical Co.), 1972. DE 2942865 (Hills AG), 1979. US 4286069 (Hoechst AG), 1980. US 4 278 731 (Atlantic Richfield Corporation), 1980. DE 2 226 168 (BASF), 1972. Ingram, A. R., Cobbs, R. R., Couchot, L. C., in Resinography of Cellular Plastics, ASTM STP 414, American Society for Testing and Materials, Philadelphia (1967) 53–67. US 3 139272 (Bloom), 1964. US 3023 175 (Koppers Corporation), 1962. Lindhof, W., Properties and uses of recycled, expanded poystyrene, Recycl. Recov. Plast. NA (1996) 631. Kunststofftechnik, Particle Foam 2000, VDI Verlag, Dusseldorf (2000). Guidlines for Transport and Storage of EPS Raw Beads, APME (1998) 1–9. RYGOL Dammstoffe Dammen leicht gemacht, Part: roof application (1999). Bollmann, W., Verarbeitungshinweise zur Herstellung von Trittschalldammplatten aus EPS, Kunststoffe 79 (1989) 77–2. Olayo, R., Garcia, E., Garcia-Corichi, B., Sanchez-Vazquez, L., Alvarez, J., Poly (vinyl alcokol) as a stabilizer in the suspension polymerization of styrene: the effect of the molecular weight, J. Appl. Polym. Sci. 67 (1998) 71. DE 4220 225 Al (BASF), 1993. Harre, K. H. Untersuchungen zum Einfluss hydrodynamischer Grossen auf die Suspensionspolymerisation von Styrol, Dissertation, Dortmund (1983). EP 781 638 A2 (Buna Sow Leuna Olefinverbund GmbH), 1997. DE 19816460 Cl (Buna Sow Leuna Olefinverbund GmbH), 1998. Russell, P. M., Hitchcock, M. K., et al. (Dow Chemical), unpublished work (1989). Tusim, M. H., Russell, P. M. (Dow Chemical), unpublished work (1992). Hitchcock, M. K. (Dow Chemical), unpublished work (1990). Hesselink, F. Th., Vrij A., Overbeck, J. Th. G., J. Phys. Chem. 75 (1971) 2094.
This page intentionally left blank
10
Rigid Polystyrene Foams and Alternative Blowing Agents KYUNG WON SUM AND ANDREW N. PAQUET The Dow Chemical Company, Midland, Ml, USA
1
INTRODUCTION AND GENERAL DESCRIPTION
The beginning of cellular polystyrene may be dated back to the time of the discovery of a material called styrene. Polystyrene is a clear, brittle, thermoplastic aromatic resin:
where n = 1900–2900 (general-purpose resin). It is one of the most versatile thermoplastic resins available for the production of low-cost plastic foams. In 1831, it was discovered that the vapors given off when a gummy material from the balsam tree (called storax) was heated contained the chemical substance styrene. Storax has also been found in substances from embalmed Egyptian mummies some 3000 years old. As early as 1831, scientists knew that liquid styrene was one of those unusual substances that could undergo a certain chemical change to become a hard solid. As scientists gained chemical knowledge about the natural styrene, they eventually learned how to produce synthetic styrene. By 1929, scientists at The Dow Chemical Company were able to devise a method to produce synthetic styrene from benzene and ethylene [1]. This Modern Styrenic Polymers: Polystyrene and Styrenic Copolymers. Edited by J. Scheirs and D. B. Priddy r 2003 John Wiley & Sons Ltd
204
K. W. SUH AND A. N. PAQUET
allowed The Dow Chemical Company to start the development of polystyrene as a material for molding and extrusion. The original concept of cellular polystyrene may be due to the Swedish inventors C. G. Munters and J. G. Tandberg, who filed a patent on 'Foamed Polystyrene' on 21 August 1931, and subsequently obtained US Patent 2023204 on 3 December 1935. In 1941, The Dow Chemical Company started research to develop a commercial process for the production of cellular polystyrene. This batch process, known as the tower process, consisted of blending polystyrene and a low-boiling compound, such as butylene or methyl chloride, in a large tower, with subsequent expansion into large foam logs, which were then cut into the desired boards or other shapes [2]. The Styrofoam (trademark of The Dow Chemical Company) was extruded as 30 cm diameter logs, which were cut into 90 cm lengths. This foam was used by the US Coast Guard and Navy as a flotation device for military equipment, and a floating buoy for the antisubmarine net during World War II. After the War, The Dow Chemical Company found new uses for Styrofoam; the material began to be used for decorative and novelty items, and also in those applications that took advantage of its insulating capability or its buoyancy. During these early stages of product development, advances were also made in process research at Dow. The first continuous extrusion process for producing a polystyrene foam was developed in the late 1940s through the early 1950s, and became the basis for the current extrusion process for the manufacture of polystyrene foams [3]. Other styrenic polymer foams were developed in the mid-1950s through the early 1960s. Examples are molded expanded polystyrene foam (MEPS), extruded polystyrene foam sheet, and expanded polystyrene loose-fill packaging material. Styrenic polymer foams have been commercially accepted in a wide variety of applications since the 1940s [1,4]. The total usage of polystyrene foams in the United States rose from about 4.10 x 105 metric tons in 1982 to an estimated 5.35 x 105 metric tons in 1987. It is expected to grow at a rate of 3–4% for the next several years [5]. For example, a recent Fredonia report on foamed plastics estimates that the 2008 volume will be 10.77 x 105 metric tons [6].
2
NOMENCLATURE
A cellular plastic is defined as a plastic, the apparent density of which is decreased substantially by the presence of numerous cells disposed throughout its mass [7]. In this chapter, the terms cellular polymer, foamed plastic, expanded plastic, and plastic foam are used interchangeably to denote all two-phase gas-solid systems in which the solid is continuous and composed of a synthetic polymer or rubber. The gas phase in a cellular polymer is usually distributed in voids or pockets called cells. If these cells are interconnected, the material is termed open-cell. If
RIGID POLYSTYRENE FOAMS AND ALTERNATIVE BLOWING AGENTS
205
these cells are discrete and the gas phase of each is independent of that of the other cells, the material is termed closed-cell. The nomenclature of cellular polymers is not standardized; classifications have been made according to the properties of the base polymer [8], the methods of manufacture, the cellular structure, or some combination of these. The most comprehensive classification of cellular plastics [9] has not been adopted and is not consistent with some of the current names for the more important commercial products. According to an ASTM test [10], foamed plastics are classified as rigid or flexible. A flexible foam is one that does not rupture when a 20 x 2.5 x 2.5cm piece is wrapped around a 2.5 cm mandrel at a uniform rate of 1 lap per 5 s, at 15–25 °C. Rigid foams rupture under this test. This classification is used here. In the case of cellular rubber, the ASTM uses several classifications based on the method of manufacture [11,12]. Cellular rubber is a general term covering all cellular materials that have an elastomer as the polymer phase. Sponge rubber and expanded rubber are cellular rubbers produced by expanding bulk rubber stocks, and are open-cell and closed- cell, respectively. Latex foam rubber, which is also a cellular rubber, is produced by frothing a rubber latex or liquid rubber, gelling the frothed latex, and then vulcanizing it in the expanded state. The term structural foam has not been exactly defined, but is used here to refer to rigid foams produced with densities greater than 320 kg/m3.
3 THEORY OF THE EXPANSION PROCESS Foamed plastics can be prepared by various methods. The most widely used, called the dispersion process, involves the dispersion of a gaseous phase throughout a fluid polymer phase, and the preservation of the resultant state. Other methods of producing cellular plastics include leaching out solid or liquid materials dispersed in the plastic, sintering small dispersed particles, and dispersing small cellular particles in the plastic. The latter processes are relatively straightforward techniques of lesser commercial importance. The expansion process has been the subject of extensive investigation because it is the foundation of foamed plastics [13–21]. In general, the expansion process consists of three steps: creation of small discontinuities or cells in a fluid or plastic phase, growth of these cells to a desired volume, and stabilization of the resultant cellular structure by physical or chemical means. 3.1
BUBBLE INITIATION
The development of bubbles within a liquid or polymer solution is generally called nucleation, although the term actually refers only to those bubbles that
206
K. W. SUH AND A. N. PAQUET
separate from the supersaturated liquid or polymer solution in the presence of an initiating site such as a surface irregularity, or at the interface of discrete phases. Gas for the bubbles has several sources: (1) dissolved gases that are normally present in the liquid or polymer solution and are forced into supersaturation by increased pressure; (2) low-boiling liquids that are incorporated into the system as blowing agents and are forced into the gas phase by increased temperature or decreased pressure; (3) gases produced as blowing agents, such as by the water–isocyanate reaction used for CO2 production in polyurethane foams; and (4) chemical blowing agents that decompose thermally to form a gas. Bubble nucleation is affected by a number of conditions. Physically, the effects of temperature, pressure, and in some cases humidity are fairly obvious. Other important parameters are surface smoothness of the substrate, surface characteristics of filler particles, presence and concentration of certain surfactants or nucleators, size and amount of second-phase liquid droplets, and the rate of gas generation. In many cases, bubbles of gas and other contaminants are already present in the liquid or polymer solution, and these serve as sites into which the gas may diffuse. The number and size of these gas bubbles may be another important factor in bubble development. 3.2
BUBBLE GROWTH
The initial bubble is ideally a sphere that grows as a result of the interaction of the differential pressure, A/>, between the inside and outside of the cell and the interfacial surface tension, y. The radius, r, of the bubble at equilibrium is related to these factors as follows: A/> = 27/r
(1)
Furthermore, the rate of growth of a cell depends on the viscoelastic nature of the polymer phase, the blowing agent pressure, the external pressure on the foam, the cell size, and the permeation rate of blowing agent through the polymer phase. Bubbles are enlarged by diffusion of gases coming out of solution into existing bubbles, coalescence of smaller bubbles, and thermal effects on internal bubble pressure according to the gas laws. As bubble size increases with decreasing density, the spherical shape of the bubbles is distorted into polyhedra with planar faces of uniform thickness. Low-density foams of cellular polymers tend to favor cell development into pentagonal dodecahedral cells, which have 12 five-sided membranes. Although this is not the configuration with the lowest surface energy, the angular symmetry is apparently more critical.
RIGID POLYSTYRENE FOAMS AND ALTERNATIVE BLOWING AGENTS 3.3
207
BUBBLE STABILIZATION
As the cell walls are squeezed into polyhedra, a wall-thinning effect takes place, and liquid is drained from cell-wall faces into the lines of cell intersections to form ribs or struts, which are typically triangular in cross-section. This cell wall membrane thinning can continue to the point where the cell walls collapse and the cells open. This becomes a very important characteristic of most plastic foams, and affects properties such as thermal conductivity, moisture absorption, breathability, and load bearing. If bubbles are sparsely distributed, as in high-density foams, they will occur as spherical cells because that is the stable shape (with the lowest surface energy). If the foam is less dense, cell-wall stability is achieved by careful control of the factors that influence membrane thinning [16]. Capillary action and gravity may cause drainage of material from the membrane into the ribs. Capillary drainage is proportional to the square of the distance between rib junctions. Increasing viscosity of the fluid reduces the drainage effect. Viscosity increase may be caused by chemical reactions that increase molecular weight through polymerization or cross-linking, or by temperature reduction of high molecular weight thermoplastics. Depression of surface tension in local areas of the cell membrane promotes rupture. Ultimate stabilization occurs as a result of the physical effect of cooling below the second-order transition point (which prevents polymer flow). As final solidification is approached, the previously formed bubbles may be distorted by the system flow or by gravity, thereby producing anisotropy in the cellular structure. This effect must be taken into account by obtaining samples oriented in specific directions to the process flow when the physical properties of plastic foams are to be evaluated.
4 4.1
PROPERTIES AND THEIR RELATION TO STRUCTURE TEST METHODS
Several countries have their own test methods for cellular plastics, and the International Organization for Standards (ISO) Technical Committee on Plastics TC-61 has been developing international standards. Information can be obtained from the American Standards Institute. The most complete set of test procedures has been developed by the ASTM and is published in a new edition every year.
4.2
PROPERTIES OF COMMERCIAL PRODUCTS
It is evident that polystyrene foams have a broad range of physical properties (Table 10.1) [22,23]; the manufacturer should be consulted for the properties of
Table 10.1 Physical properties of commercial extruded polystyrene foams: ASTM C578 - Standard for Rigid, Cellular Polystyrene Thermal Insulation Property
Method
Type X
Type IV
Type VI
Type VII
Type V
Density (min. pcf) (kg/m3) Max. use temperature, (°C) Thermal resistance of I in (25.4mm)
ASTM C303 or ASTM D1822
1.30 (21)
1.60 (26)
1.80 (29)
2.20 (35)
3.00 (48)
— ASTM C177 or ASTM C5 18
74
74
74
80
80
5.60 (0.99) 5.40 (0.95) 5.00 (0.88) 4.65 (0.82) 15.0 (104)
5.60 (0.99) 5.40 (0.95) 5.00 (0.88) 4.65 (0.82) 25.0 (173)
5.60 (0.99) 5.40 (0.95) 5.00 (0.88) 4.65 (0.82) 40.0 (276)
5.60 (0.99) 5.40 (0.95) 5.00 (0.88) 4.65 (0.82) 60.0 (414)
5.60 (0.99) 5.40 (0.95) 5.00 (0.88) 4.65 (0.82) 100.0 (690)
40.0 (276)
50.0 (345)
60.0 (414)
75.0 (517)
100.0 (690)
1.1 (63)
1.1 (63)
1.1 (63)
1.1 (63)
1.1 (63)
0.3 2.0 24.0
0.3 2.0 24.0
0.3 2.0 24.0
0.3 2.0 24.0
0.3 2.0 24.0
(min h ft2 F/Btu)(m 2 C/W): @25°F(-3.9°C) @40°F(4.4°C) @75°F (23.9°C) @110°F (43.3°C) Compressive strength (min. psi) (kPa) Flexural strength (min. psi) (kPa) Water vapor permeance of 1 .0 in. thick (25.4mm), max. perm. (ng/Paam) Water absorption (max. % by volume) Dimensional stability (max. % change) Oxygen Index (min. volume %)
ASTM D 1621 or ASTM C165 ASTM C203, Method 1 Proc. A ASTM E9-8 ASTM C272 ASTM D2128 ASTM D2843
RIGID POLYSTYRENE FOAMS AND ALTERNATIVE BLOWING AGENTS
209
a particular product. The properties of styrenic structural foams are shown in Table 10.2. These values depend upon several structural variables and should be used only as general guidelines [24]. The properties of a foamed plastic depend on the properties of the base polymer and the geometry of the foam structure, often referred to as structural variables. The polymer phase description must include the additives present. The condition or state of the polymer phase, including its orientation, crystallinity, previous thermal history, and chemical composition, determine the properties of that phase. Polymer state and cell geometry are intimately related because they are determined by forces exerted during the expansion and stabilization of the foam. Density has the most important influence on mechanical properties of a foamed plastic of a given composition. Its effect has been extensively studied.
4.3
CELLS
A complete knowledge of the cell structure of a particular polymer would require the size, shape, and location of each cell. Because this is impractical, approximations are employed. Cell size has been characterized by measurements of cell diameter [25] and of average cell volume [26,27]. Mechanical, optical, and thermal foam properties depend on cell size. Cell shape is governed predominantly by final foam density and the external forces exerted on the cellular structure before its stabilization in the expanded state. In a foam prepared without such external forces, the cells tend to be spherical or ellipsoidal at gas volumes less than 70–80% of the total volume, and the shape of packed regular dodecahedra at greater gas volumes. These shapes are consistent with surface chemistry [26,28,29]. In the presence of external forces, the cells may be elongated or flattened. Cell orientation can influence many properties [22,23], An important characteristic of the cell structure is the extent of communication with other cells. This is Table 10.2
Typical physical properties of commercial structural foams
Property Glass-reinforced Density (kg/m 3 ) Tensile strength (kPa) Compressive strength (kPa) at 10% compression Flexural strength (kPa) Flexural modulus (GPa) Max. use temperature (°C)
ASTM test
ABS
D1623 D1621
No 800 18600 6900 25500 0.86 82
D790 D790
NORYL
High-impact polystyrene
Yes 850 48000 34500
No 800 22700
No 700 12400
20% 840 34500
82700 5.2
41400 1.7 96
31000 1.4
58600 5.2
210
K. W. SUH AND A. N. PAQUET
expressed as fraction of open cells. When many cells are interconnected, the foam has a large fraction of open cells and is termed an open-cell foam. In contrast, numerous noninterconnecting cells result in a large fraction of closed cells and a closed-cell foam. The nature of the opening between cells determines how readily gases and liquids can pass from one cell to another. Because of variation of the flow from cell to cell, a single measurement of the fraction of open cells does not fully characterize this structural variable, especially in a dynamic situation. 4.4
GAS COMPOSITION
In closed-cell foams the gas phase in the cells can contain blowing agent (socalled captive blowing agent), air, or other gases generated during foaming. The thermal and electrical conductivity can be profoundly influenced by the cell-gas composition. In open-cell foams the presence of air exerts only a minor influence on the static properties but does affect the dynamic properties such as cushioning. Table 10.3 lists properties of common blowing agents used to make polystyrene foams. 4.5
RIGID CELLULAR POLYMERS
Compressive strength and modulus are readily determined and have been widely used to characterize rigid plastic foam. Rigid cellular polymers generally do not exhibit a definite yield point, but rather an increased deviation from Hooke's law as the compressive load is increased. The compressive strength is usually reported at some definite deflection (5 or 10%); the compressive modulus is extrapolated to 0 % deflection unless stated otherwise. Structural variables that affect the compressive strength and modulus of a rigid plastic foam are, in order of decreasing importance, plastic phase composition, density, cell structure, and plastic state. The effect of gas composition is minor, with a slight effect of gas pressure in some cases. The strong effect of density and polymer composition on compressive strength and modulus are illustrated in Tables 10.1 and 10.2 [8]. The cell shape or geometry also influences compressive properties [8,22,23,27,28]. In fact, the foam cell structure is controlled in some cases to optimize certain physical properties of rigid cellular polymers. In general, compressive strength and modulus of low-density foams may be expressed as Strength or modulus = A(7* where A and a constants and Q represents the foam density. This relationship is illustrated in Figure 10.1.
RIGID POLYSTYRENE FOAMS AND ALTERNATIVE BLOWING AGENTS Table 10.3
211
Physical properties of common blowing agents"
Agent
MW
Solubility
Permeability
BP
HV
TC
Crit. T.
MeCl CFC-12 EtCl HCFC-142b CO2 Ethane Butane Pentane Ethanol HCFC-22 HCFC-124 HFC-152a HFC-143a HFC-134a HFC-125 Water Nitrogen Oxygen Argon
50.5 120.9 64.5 100.5 44.0 30.1 58.1 72.2 46.1 86.5 136.5 66.1 84.0 102.0 120.0 18.0 28.0 32.0 39.9
4.6 1.7 20.0 6.7 0.4 0.3 1.9 13.0 24.0 1.6 ~ 1.0 1.8 ~ 1.5 ~1 ~ 1.0
980.0 0.17 84.0 0.1 1430.0 31.0 4.0 3.5 390.0 19.0 0.1 8.4 —»• 0)| is proportional to (M w /M e ) 35±01 for polystyrene and for many entangled flexible chain resins [35]. It is expected to apply similarly to SAN copolymers.
3.3
TIME-TEMPERATURE
SUPERPOSITION
The similarities of the isothermal curves permit the construction of a master curve (thick dashed curve in Figure 13.4) for any reference temperature TO by the well-known time-temperature superposition principle [30,36]. The superposed master curve expands the experimentally accessible w window for these examples by about three decades relative to that measured at any single temperature. This wider window is critical to understanding fabrication performance. Superposition shift factors a are identified numerically. They are designed to transform magnitude |n| and angular frequency w measured at temperature r to a reduced magnitude |n|/aj and a reduced angular frequency war at temperature TQ. Analysis of the T dependence of a allows the shifting of the data to any melt temperature T. This temperature dependence is best described by the WLF relation [30,36] for temperatures (< 1.27 t g ^460K w 190°C) just above the SAN copolymer glass-transition temperature Tg, or by the Arrhenius relation for higher temperatures [37]. Nevertheless, strategies are available to fit styrenic data to the WLF relationship to temperatures as high as 290 °C [38].
289
FEATURES IMPACTING SAN COPOLYMER PERFORMANCE 3.4
CROSS MODEL
Isothermal flow curves are often summarized by simple empirical models to understand fabrication performance. The Cross model [39,40], given by Equation (1), is well-suited for fitting SAN copolymer data as seen in Figure 13.4. (1)
1 + (COT)l-«
The three model parameters (n0, i and n) are often selected with a nonlinear least-squares algorithm which minimizes the squared difference between the measured and modeled ln |n*| for all w at temperature T. Application of a Levenberg-Marquardt algorithm [41,42] to the SAN copolymer data in Figure 13.4 yields fit parameters summarized in Table 13.4. Error bars are reported to two standard deviations. The parameter n 0 is the limiting viscosity at low w. The reciprocal () of relaxation time i marks the midpoint co for the transition from a power-law exponent of 0 at low wco to n — 1 at high co. Interpretation of these low strain amplitude parameters in nonlinear fabrication shear flows is enabled by the Cox-Merz rule [43].
3.5
NONLINEAR SHEAR FLOWS
Cox and Merz illustrated empirically the equivalence of the co dependence of rj* | to the shear rate y dependence of the steady-state shear viscosity n(y). This rule has been tested frequently, and found to be reliable for flexible-chain resins Table 13.4 Temperature dependence of Cross model parameters for a SAN resin (WAN = 0.51, Mw = 55kg/mol, Mw/Mn = 1.70) Temperature
r(°Q
160.97 ±0.11 171. 18 ±0.14 181.48 ±0.14 191.48 ±0.14 20 1.53 ±0.09 21 1.58 ±0.08 22 1.60 ±0.08 231. 70 ±0.07 241. 80 ±0.08 251. 98 ±0.09 262. 18 ±0.09
Zero-shear viscosity, 2021000 ± 94000 11 25000 ±72000 483000 ± 27000 219400 ± 7800 108000 ± 3000 58800 ± 1700 35100± 800 22100 ± 630 14130 ± 280 8800 ± 120 4860 ± 86
T(S)
Relaxation time, (Pa s)
Power-law exponent, n
14.5 ± 1.5 8.9 ± 1.5 4.59 ± 0.79 2.25 ± 0.28 1.14 ±0.12 0.659 ± 0.075 0.456 ± 0.042 0.336 ± 0.039 0.243 ± 0.021 0.1 665 ±0.0096 0.1076 ±0.0084
0.203 0.246 0.281 0.305 0.323 0.343 0.371 0.401 0.422 0.436 0.459
± 0.009 ± 0.015 ± 0.018 ± 0.015 ± 0.014 ± 0.016 ±0.012 ± 0.016 ±0.011 ± 0.008 ±0.011
290
R. P. DION AND R. L SAMMLER
[22]. In this viewpoint, parameter n0 is the limiting steady-state viscosity for creeping shear flows, i~l marks the shear rate for the transition between Newtonian flow [f/(y) = r\Q] and power-law flow [//(y) oc y""1], and (n — 1) is the limiting shear-thinning rate din >/(y)/dln y at high y. Of course, the powerlaw slope (n — 1) must vary from 0 to —1 to be physically meaningful, and correspondingly n must vary from 1 to 0. Fabricators will use //0 as a metric of melt strength for creeping flows. For example, selection of resins with higher values of rj0 should minimize sag in thermoformed parts, and allow a more uniform part wall thickness. Similarly, T and n are metrics of the onset and rate of shear thinning. They are used to identify low-f/(y) conditions (easy flow) in an extruder, which are often critical to fast part fabrication rates. Direct measurement of the y and T dependences of r/(y) for SAN copolymers can be made with a capillary rheometer [44,45]. A capillary rheometer is designed to probe the high shear rate (l-10 4 s -1 ) portion of the flow curve. Shear rates as high as 107 s-1 have been reported for SAN copolymers [44,45]. Capillary experiments also permit the study of thermomechanical degradation of polymer melts which may accompany Theological and fabrication experiments at y > 10 4 s -1 . Relative to dynamic experiments, much more material (20-200 g) and effort are required when isothermal y-sweep data are properly corrected by the prescriptions of Rabinowitsch and co-workers [46,47] and of Bagley [48,49]. Both features often preclude the use of capillary experiments for rapid screening of new materials available in research-scale quantities (5-20 g).
3.6
RELAXATION
SPECTRA
An isothermal shear flow curve may be transformed into a relaxation spectrum to fingerprint the time scales of molecular motion in the polymeric melt [50–53]. The spectrum represents the most fundamental aspects of melt rheology, and can be incorporated into a constitutive model to predict melt rheology in any flow [54–57]. The predictions, of course, are best for strains and strain rates within the experimental windows, but often remain very good just outside (< 1 decade) these windows. Fortunately, there are excellent approaches to improve the predictions. For example, large-strain data, measured during startup of uniaxial extensional flow, may be used to identify damping factors that improve the prediction of several types of nonlinear flows [54–57]. Alternatively, tandem use of isothermal creep recovery and dynamic experiments may be used to lower the lowest measurable shear rate from about 10-2 to 1 0 - 5 s ' if the molten resin has sufficient thermally stability to withstand the long creep/ creep recovery times [58]. Discrete spectra with less than five relaxation times are commonly used to coarse grain the Theological data [51–53]. The coarse graining is often necessary
FEATURES IMPACTING SAN COPOLYMER PERFORMANCE
291
to simulate fabrication performance with computation-intensive numerical algorithms [57] (or POLYFLOW™ from Fluent). In contrast, continuous spectra are useful for a better understanding of the long-chain structure. For example, a bimodal relaxation spectrum indicates the presence of two distinct polymer-chain length scales. The length scales may be due to two types of linear chains with much different lengths (molecular weights), or to identical comb-like chains with the backbone lengths much longer than the branch lengths.
3.7
EXTENSIONAL FLOW
Reliable measurement of the rheological properties of polymer melts in purely extensional flow and at constant strain rate has always been more challenging than for shear flow. Two complementary designs are in use for startup of uniaxial extensional flow [59–61]. The Meissner design [60,61] is more common today owing to its commercial availability. It remarkably stretches a strip of material to very large strains (Hencky strain E = 7; linear strains of e7 w 1100) in a small oven. The trajectories of small glass beads sprinkled on top of the strip are analyzed to verify that the stretching is spatially uniform and to determine the true strain. Trajectory analysis is especially important for SAN copolymers at temperatures below about 150 °C when sample slippage between the metal belts may become appreciable. No neutral-density oil bath is needed in this latest design to minimize sag of the sample strip. More details on sample preparation, data acquisition and analysis, and estimation of errors on Meissner design may be found elsewhere [62]. The maximum strain rate (e < 1 s - 1 ) for either extensional rheometer is often very slow compared with those of fabrication. Fortunately, time-temperature superposition approaches work well for SAN copolymers, and permit the elevation of the reduced strain rates saj to those comparable to fabrication. Typical extensional rheology data for a SAN copolymer (WAN = 0.264, Mw = 78kg/mol,Mw/Mn = 2.8) are illustrated in Figure 13.5 after time-temperature superposition to a reference temperature of 170°C [63]. The tensile stress growth coefficient n+E, 0 was measured at discrete times t during the startup of uniaxial extensional flow. Data points are marked with individual symbols (o) and terminate at the tensile break point at longest time t. Isothermal data points are connected by solid curves. Data were collected at selected £ between 0.0167 and 0.0840s-1 and at temperatures between 130 and 180 °C. Also illustrated in Figure 13.5 (dashed line) is a shear flow curve from a dynamic experiment displayed in a special format (3|n| versus a)-1) as suggested by Trouton [64]. The superposition of the low-strain rate data from two types (shear and extensional flow) of rheometers is an important validation of the reliability of both data sets.
292
R. P. DION AND R. L. SAMMLER
AS-2, \7Q°C
13 Cu
TV- 6
3n'(t)
3tl. eaT 0.0818 0.973 2.10 7.00 7.89 20.8 40.0 i
•o-' -3
-2
-1
0
1 log(taT -i/s)
Figure 13.5 Dependences of the reduced tensile stress growth coefficient ^|(e,0/aj at 170°C on reduced time teT and reduced strain rate ear for a SAN resin (WAN = 0.264, Mw = 78 kg/mol, Mw/Mn = 2.8) during the startup of uniaxial extensional flow. Also illustrated (dashed curve) are dynamic shear viscosity data displayed as 3|^*(a>, 770°C)| versus u-1 as suggested by Trouton [64]. Reproduced from L. Li, T. Masuda and M. Takahashi, J.Rheol., 34(1), 103(1990), with permission of the American Institute of Physics
Strain hardening, as indicated by */£(£, 0 > 3|f*l» is clearly evident in this SAN copolymer for the higher reduced strain rates (ea-r > 1 s -1 ). This is a very valuable feature in a material since it enhances its ability to deliver uniformly thick walls during extension. However, use of this isothermal feature may be underutilized in fabrication since large strains (e = tsaj > 2; e2 % 7.4) and high strain rates (e > 8) are needed. A more often used feature for hardening copolymer melts (e.g. thermoforming) arises from cooling since the melt viscosity has such a strong dependence on temperature. Here, the flows are often sufficiently slow that the tensile stress growth coefficient nE Ce, t) at time t may be estimated simply with Trouton's rule [3|>/*(a> = f-1)|]. A few rheometers are available for measurement of equi-biaxial and planar extensional properties polymer melts [62,65,66]. The additional experimental challenges associated with these more complicated flows often preclude their use. In practice, these melt Theological properties are often first estimated from decomposing a shear flow curve into a relaxation spectrum and predicting the properties with a constitutive model appropriate for the extensional flow [54–57]. Predictions may be improved at higher strains with damping factors estimated from either a simple shear or uniaxial extensional flow. The limiting tensile strain or stress at the melt break point are not well predicted by this simple approach.
FEATURES IMPACTING SAN COPOLYMER PERFORMANCE
3.8
293
BREAK POINTS
Wall thickness uniformity is often compromised when fabricating near the material break point. Thinner walls are expected in the most stretched regions, but only if the melt is not allowed to recoil after cessation of flow. Often fabrication conditions are selected to be well away from the break point to minimize these issues. Key break point metrics for the startup of data illustratrated in Figure 13.5 are the time /b and tensile stress coefficient ^(fi, ^) at break. Both quantities may be multiplied by the strain rate e to estimate the Hencky break strain (eb = t^e) and the break stress (0b = ^E(£, 0)- Similar metrics can be defined for other startup extensional flows. The break stress crb of entangled high polymer melts is thought to be dominated by the intrinsic strength of the polymer backbone, and offset by chain disentanglements when re < 1 and the type and level of stress concentrators (inhomogeneities, impurities, unmelts, dust, etc.) present in the melt. A zerothorder estimate of the intrinsic strength (1 MPa) of styrenic and olefinic resins has been reported by Laun and Schuch [67] at high strain rates e with a Rheotens setup [68,69]. This intrinsic stress can be used to improve predictions of extensional flow with constitutive models when break point data are not available. One simply truncates all predictions of flow when the tensile stress exceeds this limit. Nevertheless, it remains far more preferable to have break point data since stress concentrators may lower ab significantly and adversely affect fabrication performance. For example, the break stress a^ and strain eb are very low (0.01 MPa and 0.82, respectively) for the low £#T data reported in Figure 13.5.
3.9
BRITTLE BREAKS
The slopes of the tensile stress growth coefficient curves near the break point are also important. The steep slopes observed for the higher sa-r are typical of brittle breaks. The flat slopes observed for the low e«x are typical of ductile breaks. Strong elastic recoils are expected after cessation of flow for brittle breaks. The recoil may be characterized in uniaxial extensional flow in the Meissner design [55,60–63]. Rapid cooling and solidification of the stretched melt may minimize the melt recoil but may introduce warpage in the part as residual stress grows during the cooling to the part service temperature. The residual stress may later be responsible for part shrinkage if the part temperature approaches the glass transition temperature.
3.10
FLOW BIREFRINGENCE
The optical anisotropy of molten flexible-chain polymers is often very small in a quiescent (nonflowing) state owing to their nearly spherical chain configuration
294
R. P. DION AND R. L. SAMMLER
and random orientation. Flow will induce an optical anisotropy in the melt by deforming the entangled chains from their equilibrium configurations and aligning the deformed chains relative to the flow streamlines. This anisotropy can be sensed by measurement of the flow-induced birefringence A« [70,71]. Birefringence setups can be designed to characterize molten materials undergoing isothermal homogeneous flow. The ranges of strains and strain rates also often coincide with those of rheometers, and consequently may be limited relative to those used in fabrication. Similarly, time-temperature superposition approaches may be used to expand the rate window. State-of-the-art setups suitable for rapid screening of new materials with research-scale quantities (5–20 g) are available for shear flow [72] and startup of uniaxial extensional flow [73,74]. Birefringence experiments provide complementary information to rheological experiments. Both probe the response of the molten chains to the flow, and can be used to fingerprint their wide range of relaxation times. The birefringence experiment senses the alignment rather than the mechanical impedance of chains undergoing flow. It is widely used as a relative metric of stress fields in non-homogeneous flows at low strains [75-77] since it can be measured at any point in the fluid. The birefringence is transformed into stress via the stress-optic rule. This rule relates the refractive index tensor n to the extra stress tensor T at low strains by the stress-optical coefficient C. Extensions of the rule at higher strains remain an active research area [70,71].
4
MULTIPHASE SYSTEMS
Useful multiphase SAN copolymer mixtures can be made by the addition of an impact modifier to SAN in a polymer blending operation. Acrylonitrilebutadiene-styrene (ABS), which is an impact modified SAN copolymer, can be produced by melt mixing SAN copolymer grafted rubber with blending grade SAN copolymers. The properties of these immiscible high polymer blends are generally not a simple average of the properties of the components. They depend critically on the morphology of the immiscible domains that develop during blending and fabrication. An impact modifier with the incorrect graft phase may disperse poorly in the SAN continuous phase, and exhibit poor properties [78–81]. Extensive reviews of the rich field of polymer blends may be found elsewhere [26,82–84]. Since most high molecular weight polymers do not mix on a molecular scale, it is desirable to determine compositional ranges of miscibility. The solubility of SAN copolymers with other polymers has been measured by a variety of techniques. The tendency of the materials to mix or phase separate is determined by the enthalpy of interactions between 'mer' units in the polymers and by the molecular weight of the polymers. It was determined experimentally that
FEATURES IMPACTING SAN COPOLYMER PERFORMANCE
295
SAN copolymer compositions in solution will phase separate when they differ in composition by around 4 mol% [85]. The thermodynamics of this system has been described in detail, in terms of interaction parameters [86,87], and calculated phase diagrams agree with published experimental data. Blend morphology commonly depends on the weight fraction and viscoelastic properties of each component, the interfacial tension between components, the shape and sizes of the discontinuous phase, and the fabrication conditions and setup. Most rheological experiments applied to homogeneous melts can also be similarly applied to these immiscible blends [55,63,88,89]. The viscoelastic properties arising from these studies should be labeled with a subscript 'apparent' since the equations used to translate rheometer transducer responses to properties incorrectly assume that the material is homogeneous. Nevertheless, these apparent properties are often found to be excellent metrics of fabrication performance. Flow of the blend at melt temperatures generally stretches the discontinuous phase from its initial shape. Interfacial tension between the immiscible components will oppose this process and attempt to drive the system to a lowenergy spherical-morphology state. Studies of these phenomena in a rheometer permit the estimation of the time required for both processes and the interfacial tension [90–93]. Alternatively, one can estimate the time required for the latter process, and the interfacial tension, from the evolving shape of the discontinuous phase using either a fiber break-up [94,95] or fiber-retraction [33,96] experiment. Interfacial tension depends on the molecular weight of each component [96,97] and on temperature, so it is preferable to measure interfacial tension for the materials of interest at their fabrication temperature. The compatibility of SAN copolymers with an assortment of other polymers has been measured by a variety of techniques. Differential scanning calorimetry is used to determine the glass transition temperatures of copolymers. The values increase slightly with increasing acrylonitrile content and range from around 100 to 115 °C [98,99]. If the glass transition of a second polymer differs from that of the SAN copolymer by 20 °C or more, measurements of the composite transitions in blends can easily be used to measure compatibility. Blends with one glass transition are miscible, blends with the two original glass transitions are immiscible, and blends in which the two glass transitions have moved closer together are considered compatible. Compositional windows of miscibility of SAN copolymers with a-methylstyrene-co-acrylonitrile, styrene-co-acrylonitrile-cofumaronitrile and styrene-co-maleic anhydride polymers have been reported [98-100]. The commercial interest in these materials is due to the higher end use temperatures of impact modified styrene co- or terpolymers relative to SAN copolymers. The impact modifiers are elastomer cores, grafted with SAN copolymer shells, so the miscibility of the continuous phase polymer with the graft phase of the impact modifier influences the physical property balance of the blend [80,81]. Complex mixtures of SAN copolymers can be separated and
296
R. P. DION AND R. L. SAMMLER
identified using combinations of solvent precipitation, gel permeation chromatography and liquid chromatography [78]. Solid-state NMR has been used to examine compatibility in SAN copolymer blends with styrene-maleic anhydride copolymers [101]. Spin diffusion experiments indicate that the two polymers mix on a molecular scale, but the data suggest that there is no specific interaction between the nitrile group and the carbonyl groups in the maleic anhydride. This technique can provide some powerful chemical data that cannot be obtained by other methods. Unfortunately, the method requires the preparation of 13C-enriched polymers. 5
SOLID-PHASE BEHAVIOR
The micromechanical deformation behavior of SAN copolymers and rubberreinforced SAN copolymers have been examined in both compression [102] and in tension [103,104]. Both modes are important, as the geometry of the part in a given application and the nature of the deformation can create either stress state. However, the tensile mode is often viewed as more critical since these materials are more brittle in tension. The tensile properties also depend on temperature as illustrated in Figure 13.6 for a typical SAN copolymer [27]. This resin transforms from a brittle to ductile material under a tensile load between 40 and 60CC.
2
4 6 8 Elongation (%)
Figure 13.6 Temperature dependence of the tensile properties of a SAN resin. Reprinted with permission from Throne, J.TechnologyofThermoformina, Manser Publishers, Munich, Copyright 1996
FEATURES IMPACTING SAN COPOLYMER PERFORMANCE
297
Failure of unoriented SAN copolymers is dominated by crazing behavior. The total energy absorbed by the polymer during failure has been increased by optimizing both failure modes. Yielding can be enhanced by orienting the polymer [105], and crazing can be optimized by rubber modification. Semi-empirical approaches are available to predict selected solid-state properties of thermoplastics from molecular structure. Example predictions for about 30 types of thermoplastics, including SAN copolymers, may be found elsewhere [34]. The rheological and thermal properties of high molecular weight SAN copolymers can be modified by the addition of styrene plus acrylonitrile oligomers [106]. These low molecular weight species form during the copolymerization process [107], and are effective plasticizers. The changes in macroscopic properties are a mixture of desirable and detrimental effects. The melt viscosity is decreased, the toughness is increased, and the heat resistance is decreased. Several other properties of copolymers that are important in specific applications have also been measured. The surface properties of polymers determine the nature of adhesives that will stick to a substrate, and the nature of solvents that will wet the surface. The surface energy of some styrene and acrylonitrile have been measured, and the surface is rich in polystyrene when the acrylonitrile content of the copolymer is below 50% [108]. The properties of SAN and rubber-modified SAN copolymers have also been evaluated for oriented films. Biaxial orientation can increase the glass transition temperature by up to 15 °C and decrease the water vapor transmission rate by 30% [105]. SAN copolymers may be blended with polycarbonate (PC) to reduce the birefringence of magneto-optical storage disks to acceptable levels [109]. Methods to measure the birefringence in three different disk orientations are summarized elsewhere [109]. This approach takes advantage of the opposite sign of the birefringence of each blend component to annihilate the birefringence. PC-SAN blends perform better than PC-S blends in that the weight fraction WAN of acrylonitrile may be selected to compatibilize the blend. Blend compatabilization is desirable as it reduces the level of light scattering from the immiscible domains. It also reduces the driving force for additional phase separation and haze if the disk is exposed to heat. Other useful metrics of the optical properties of SAN copolymers are color [110], haze [111], gloss [112], and refractive index [113].
6
CONCLUSION
SAN copolymers are complex mixtures possessing heterogeneity in both chemical composition and molecular weight distribution. Consequently, analytical characterization of these materials is complex, but absolutely critical for
R. P. DION AND R. L. SAMMLER
298
Table 13.5 Summary of characterization techniques and associated structure-property relationships Scale
Characterization technique
Molecular
Ultraviolet spectroscopy Nuclear magnetic resonance Thermogravimetric analysis-mass spectrometry Macromolecular Gel permeation chromatography Liquid chromatography Differential scanning calorimetry Interfacial tension Melt flow rate Macroscopic Impact strength Tensile strength
Polymer attribute
Related property
Molecular conjugation Sequence distribution Pyrolysis products
Color
Molecular weight distribution Compositional distribution Thermal transitions
Surface energy Molecular mobility Energy absorption Resistance to flow in tension Resistance to flow Tensile modulus in tension Resistance to flow DTUL at high temperature Vicat softening point Resistance to flow at high temperature Flammability rating Flammability
Color, clarity Composition —» color, toughness, viscosity Viscosity, toughness Color, clarity Thermal resistance Toughness in blends Viscosity Toughness Toughness Stiffness Thermal resistance Thermal resistance Flammablility
understanding properties and developing commercial applications. Methodology exists to characterize SAN copolymers at the molecular, macromolecular and macroscopic scales (Table 13.5), and some of the structure-property relationships have been elucidated. This information can be used to modify physical properties and optimize relevant properties to best meet part fabrication and application needs.
REFERENCES 1. Priddy DB (1995) Thermal discoloration chemistry of styrene-co-acrylonitrile. Adv PolymSci 121:123–54. 2. Garcia-Rubio LH and Ro N (1985) Detailed copolymer characterization using ultraviolet spectroscopy. Can J Chem 63:253–63.
FEATURES IMPACTING SAN COPOLYMER PERFORMANCE
299
3. Schaefer J (1971) High-resolution pulsed carbon-13 nuclear magnetic resonance analysis of the monomer distribution in acrylonitrile-styrene copolymers. Macromolecules 4:107–10. 4. Borbely JD, Graham D, Hill DJT, Lang AP, Munro P and O'Donnell JH (1991) NMR studies of copolymerizations of acrylonitrile. Prog Pac Polym Sci, Proc Pac Polym Conf 1:237–48. 5. Zeng W and Shirota Y (1989) Studies on alternating radical copolymerization: analysis of microstructures of styrene-maleic anhydride, styrene-acrylonitrile, and styrene-methyl methacrylate copolymers by fluorescence spectroscopy. Macromolecules 22:4204–8. 6. Sargent M, Koenig JL and Maeker NL (1991) FT-IR analysis of the monomer sequence distribution of styrene-acrylonitrile copolymers. Appl Spectrosc 45:1726– 32. 7. Tsuge S, Kobayashi T, Sugimura Y, Nagaya T and Takeuchi T (1979) Pyrolysis gas chromatographic characterization of highly alternating copolymers containing styrene and tetracyanoquinodimethane, methyl acrylate, acrylonitrile, or methyl methacrylate units. Macromolecules 12:988–92. 8. Tsuge S, Sugimura Y, Kobayashi T and Nagaya T (1981) Microstructural characterization of various copolymers by pyrolysis-glass capillary gas chromatography. Polym Prepr (Am Chem Soc, Div Polym Chem) 22(l):284–5. 9. Vukovic R and Gnjatovic V (1970) Characterization of styrene-acrylonitrile copolymer by pyrolysis gas chromatography. J Polym Sci, Part A-l 8:139–46. 10. Del Rios JK (1988) Polymer characterization using the photodiode array detector. Am Lab 20:78-82. 11. Garcia-Rubio LH, MacGregor JF and Hamielec AE (1981) Size exclusion chromatography-characterization of copolymers. Polym Prepr (Am Chem Soc, Div Polym Chem) 22(l):292–3. 12. Allen DS, Birchmeier M, Pribish JR, Priddy DB and Smith PB (1993) Thermal styrene-co-acrylonitrile discoloration problem: the role of sequence distribution and oligomers. Macromolecules 26:6068-6075. 13. Gloeckner G and Van den Berg JHM (1987) Copolymer fractionation by gradient high-performance liquid chromatography. J Chromatogr. 384:135–44. 14. Gloeckner G, Wolf D and Engelhardt H (1991) Separation of copoly(styrene/ acrylonitrile) samples according to composition under reversed phase conditions. Chromatographia 32:107–12. 15. Gloeckner G, Wolf D and Engelhardt H (1994) Control of adsorption and solubility in gradient high performance liquid chromatography. Part 3. Sudden-transition gradient elution of styrene/acrylonitrile copolymers. Chromatographia 38:749-55. 16. Kuhn R (1983) Characterization of polymer blends, block copolymers, and graft copolymers by fractionation procedures using demixing solvents. Polym Sci Technol 20:45-58. 17. Lovric L and Res INA (1969) Fractionation of styrene-acrylonitrile copolymers. J Polym Sci, Part A-2 7:1357–66. 18. Teramachi S and Fukao T (1974) Cross fractionation of styrene-acrylonitrile copolymer. Polym J 6:532–6. 19. Cortes HJ, Jewett GL, Pfeiffer CD, Martin S and Smith C (1989) Multidimensional chromatography using on-line microcolumn liquid chromatography and pyrolysis gas chromatography for polymer characterization. Anal Chem 61:961–5. 20. Mori S (1996) Characterization of styrene-acrylonitrile copolymers by size exclusion chromatography/stepwise gradient elution-liquid precipitation chromatography. Int J Polym Anal Charact 2:185–92.
300
R. P. DION AND R. L SAMMLER
21. Ogawa T and Sakai M (1981) Column fractionation of acrylonitrile-styrene copolymers. J Polym Sci, Polym Phys Ed 19:1377–83. 22. Dealy JM and Wissbrun KF (1995) Melt Rheology and Its Role in Plastics Processing. Chapman and Hall, New York. 23. Agassant J-F, Avenas P, Sergent J-Ph and Carreau PJ (1991) Polymer Processing, Principles and Modeling. Hanser, New York. 24. Rosato DV and Rosato MG (2000) Injection Molding Handbook, 3rd edn. Kluwer, Dordrecht. 25. Rosato DV and Rosato DV (1988) Blow Molding Handbook. Hanser, New York. 26. Paul DR and Newman S (1978) Polymer Blends, Vols 1 and 2. Academic Press, New York. 27. Throne JL (1996) Technology of Thermoforming. Hanser, Munich. 28. Wilkinson AN and Ryan AJ (1998) Polymer Processing and Structure Development. Kluwer, Boston. 29. Bird RB, Armstrong RC and Hassager O (1977) Dynamics of Polymeric Liquids. Wiley, New York. 30. Ferry JD (1980) Viscoelastic Properties of Polymers, 3rd edn. Wiley, New York. 31. Middleman S (1968) The Flow of High Polymers. Interscience, New York. 32. Lodge AS (1964) Elastic Liquids. Academic Press, New York. 33. Sammler RL, Dion RP, Carriere CJ and Cohen A (1992) Compatibility of high polymers probed by interfacial tension. Rheol Acta 31:554–64. 34. Seitz JT (1993) The estimation of mechanical properties of polymers from molecular structure. J Appl Polym Sci 49:1331–51. 35. Berry GC and Fox TG (1968) The viscosity of polymers and their concentrated solutions. Adv Polym Sci 5:261–457. 36. Williams ML, Landel RF and Ferry JD (1955) The temperature dependence of relaxation mechanisms in amorphous polymers and other glass-forming liquids. J Am Chem Soc 77:3701–7. 37. Van Krevelen DW and Hoftyzer PJ (1976) Newtonian shear viscosity of polymer melts. Angew Makromol Chem 52:101–9. 38. Lomellini P (1992) Williams-Landel-Ferry versus Arrhenius behaviour: polystyrene melt viscoelasticity revised. Polymer 33:4983-89. 39. Cross MM (1965) Rheology of non-newtonican fluids: a new flow equation for pseudoplastic systems. J Colloid Sci 20:417–37. 40. Cross MM (1969) Polymer rheology: influence of molecular weight and polydispersity. J Appl Polym Sci 13:765–74. 41. Shrager R (1970) Nonlinear regression with linear constraints: an extension of the magnified diagonal method. J Assoc Comp Machinery 17:446–52. 42. Press WH, Flannery BP, Teukolsky SA and Vetterling WT (1988) Numerical Redpies in C, Chapt. 14.4. Cambridge University Press, Cambridge. 43. Cox WP and Merz EH (1958) Correlation of dynamic and steady flow viscosities. J Polym Sci 28:619–22. 44. Takahashi H (1990) Flow behavior of polymer melts under extremely high shear rates and flow analysis of injection molding. J Rheol (Jpri) 18:172–9. 45. Takahashi HIY, Yamamoto S and Kamigaito O (1990) Flow curves of mixtures of acrylonitrile-styrene copolymers with different molecular weights at extremely high shear rates. J Rheol (Jpn) 18:125–8. 46. Rabinowitsch BZ (1929) Uber die viskositat und elastizitat von solen. Z Phys Chem AbtA 145:1–26. 47. Eisenschitz R, Rabinowitsch B and Weissenberg K (1929) Mitt Dtsch Mat PrufAnst Sanderheft 9:91.
FEATURES IMPACTING SAN COPOLYMER PERFORMANCE
301
48. Bagley EB (1957) End corrections in the capillary flow of polyethylene. J Appl phys 28:624–7. 49. Bagley EB (1961) The separation of elastic and viscous effects in polymer flow. Trans Soc Rheol 5:355-68. 50. Honerkamp J and Weese J (1989) Determination of the relaxation spectrum by a regularization method. Macromolecules 22:4372–7. 51. Baumgartel M and Winter HH (1989) Determination of discrete relaxation and retardation time spectra from dynamic mechanical data. Rheol Acta 28:511–9. 52. Baumgartel M and Winter HH (1992) Interrelation between continuous and discrete relaxation time spectra. J Non-Newtonian Fluid Mech 44:15-36. 53. Winter HH (1997) Analysis of dynamic mechanical data: inversion into a relaxation time spectrum and consistency check. J Non-Newtonian Fluid Mech 68:225-39. 54. Larson RG (1988) Constitutive Equations for Polymer Melts and Solutions. Butterworths, Boston. 55. Solovyov SE, Virkler TL and Scott CE (1999) Rheology of acrylonitrile—butadienestyrene polymer melts and viscoelastic constitutive models. J Rheol 43:977-90. 56. Otsuki Y, Kajiwara T and Funatsu K (1997) Numerical simulations of annular extrudate swell of polymer melts. Polym Eng Sci 37:1171–81. 57. Baaijens FPT (1998) Mixed finite element methods for viscoelastic flow analysis: a review. J Non-Newtonian Fluid Mech 79:361-85. 58. Kraft M, Meissner J and Kaschta J (1999) Linear viscoelastic characterization of polymer melts with long relaxation times. Macromolecules 32:751-7. 59. Munstedt H (1979) New universal extensional rheometer for polymer melts. Measurements on a polystyrene sample. J Rheol 23:421–36. 60. Meissner J and Hosttler J (1994) A new elongational rheometer for polymer melts and other highly viscoelastic fluids. Rheol Acta 33:1–21. 61. Meissner J (1996) Elongation of polymer melts—experimental methods and recent results. Proc XIIth Int Congr Rheol 1:1–10. 62. Schweizer T (2000) The uniaxial elongational rheometer RME - six years of experience. Rheol Acta 39:428–43. 63. Li L, Masuda T and Takahashi M (1990) Elongational flow behavior of ABS polymer melts. J Rheol 34:103–16. 64. Trouton FT (1906) On the coefficient of viscous traction and its relation to that of viscosity. Proc R. Soc London, Ser. A 77:426–40. 65. Meissner J, Stephenson SE, Demarmels A and Portmann P (1982) Multiaxial elongational flows of polymer melts - classification and experimental realization. J Non-Newtonian Fluid Mech 11:221-237. 66. Meissner J (1985) Rheometry of polymer melts. Annu Rev Fluid Mech 17:45–64. 67. Laun HM and Schuch H (1989) Transient elongation viscosities and drawability of polymer melts. J Rheol 33:119–75. 68. Meissner J (1971) Dehnumgsverhalten von Polyathylen-schmelzen. Rheol Acta 10:230–42. 69. Wagner MH, Bernnat A and Schulze V (1998) The rheology of the rheotens test. J Rheol 42:917–28. 70. Janeschitz-Kriegl H (1983) Polymer Melt Rheology and Flow Birefringence. Springer, New York. 71. Fuller GG (1995) Optical Rheometry of Complex Fluids. Oxford University Press, New York. 72. Kumaraswamy G, Verma RK and Kornfield JA (1999) Novel flow apparatus for investigating shear-enhanced crystallization and structure development in semicrystalline polymers. Rev Sci Instrum 70:2097-104.
302
R. P. DION AND R. L. SAMMLER
73. Kotaka T, Kojima A and Okamoto M (1997) Elongational flow opto-rheometry for polymer melts - 1. Construction of an elongational flow opto-rheometer and some preliminary results. Rheol Acta 36:646–56. 74. Venerus DC, Zhu S-H and Ottinger HC (1999) Stress and birefringence measurements during the uniaxial elongation of polystyrene melts. J Rheol 43:795–813. 75. Quinzani LM, Armstrong RC and Brown RA (1994) Birefringence and laser-Doppler velocimetry (LDV) studies of a viscoelastic fluid through a planar contraction. J Non-Newtonian Fluid Mech 52:1–36. 76. Baaijens HPW, Peters GWM, Baaijens FPT and Mejer HEH (1995) Viscoelastic flow past a confined cylinder of a polyisobutylene solution. J Rheol 39:1243–77. 77. Liang RF and Mackley MR (2001) The gas-assisted extrusion of molten polyethylene. J Rheol 45:211–26. 78. Kuhn R, Mueller HG, Bayer G, Kraemer-Lucas H, Kaiser W, Orth P, Eichenauer H and Ott KH (1993) Characterization of bimodal bigraft ABS. Colloid Polym Sci 271:13–42. 79. Dion RP and Billovits GF (1996) Interfacial tension: a quantitative measure of styrenic blend compatibility. Polym Prepr (Am Chem Soc, Div Polvm Chem) 32:529-30. 80. Dion R and Warakomski J (1993) Heat resistant styrenic polymer blends. US Patent 5212240. 81. Henton D, Dion R and Lefevre N (1990) Process for preparing copolymers of alphamethylstyrene and acrylonitrile. US Patent 4972032. 82. Utracki LA (1989) Polymer Alloys and Blends. Hanser, New York. 83. Hobbs SY, Dekkers MEJ and Watkins VH (1988) Effect of interfacial forces on polymer blend morphologies. Polymer 29:1598–602. 84. Wu S (1990) Chain structure, phase morphology, and toughness relationships in polymers and blends. Polym Eng Sci 27:335–43. 85. Molau G (1965) Heterogeneous polymer systems. III. Phase separation in styreneacrylonitrile copolymers. J Polym Sci, Polym Lett Ed 3:1007–15. 86. Koningsveld R and Kleintjens LA (1985) Liquid-liquid phase separation in multicomponent polymer systems 22. Thermodynamics of statistical copolymers. Macromolecules 18:243-52. 87. Vukovic R, Bogdanic G, Karasz FE and MacKnight WJ (1999) Phase behavior and miscibility in binary blends containing polymers and copolymers of styrene, of 2,6dimethyl-l,4–phenylene oxide and of their derivatives. J Phys Chem Ref Data 28:851–68. 88. Takahashi M, Li L and Masuda T (1989) Nonlinear viscoelasticity of ABS polymers in the molten state. J Rheol 33:709–23. 89. Wetton RE, Corish PJ (1980) DMTA studies of polymer blends and compatibility. Polym Test 8:303–12. 90. Gramespacher H and Meissner J (1992) Interfacial tension between polymer melts measured by shear oscillations of their blends. J Rheol 36:1127–41. 91. Gramespacher H and Meissner J (1995) Reversal of recovery direction during creep recovery of polymer blends. J Rheol 39:151–60. 92. Gramespacher H and Meissner J (1997) Melt elongation and recovery of polymer blends, morphology, and influence of interfacial tension. J Rheol 41:27–44. 93. Levitt L and Macosko CW (1997) Extensional rheometry of polymer multilayers: a sensitive probe of interfaces. J Rheol 41:671–85. 94. Chappelear DC (1964) Interfacial tension between molten polymers. ACS Polym Prepr 5:363-72.
FEATURES IMPACTING SAN COPOLYMER PERFORMANCE
303
95. Elemans PHM, Janssen JMH and Meijer HEH (1990) The breaking thread method: the measurement of interfacial tension in polymer systems. J Rheol 34:1311–22. 96. Ellingson PC, Strand DA, Cohen A, Sammler RL and Carriere C (1994) Molecular weight dependence of polystyrene/poly(methyl methacrylate) interfacial tension probed by imbedded-fiber retraction. Macromolecules 27:1643–7. 97. Anastasiadis SH, Gancarz I and Koberstein JT (1988) Interfacial tension of immiscible polymer blends: temperature and molecular weight dependence. Macromolecules 21:2980-7. 98. Cowie, JMG, Elexpuru EM and McEwen IJ (1992) Observation of restricted miscibility in binary blends of poly(styrene-stat-acrylonitrile) and poly(alphamethylstyrene-stat-acrylonitrile). Polymer 33:1993–5. 99. Warakomski, J and Dion R (1992) The effect of chemical composition on the miscibility of styrene/acrylonitrile/fumaronitrile terpolymers with styrene/acrylonitrile copolymers. J Appl Polym Sci 46:1057–63. 100. Hall W, Kruse R, Mendelson R and Trementozzi Q (1983) New styrene—maleic anhydride terpolymer blends. ACS Symp Ser 229:49-64. 101. Heinen W, Wenzel CB, Rosenmoeller CH, Mulder FM, Boender GJ, Lugtenburg J, de Groot HJM, Van Duin M, Klumperman B (1998) Solid-state NMR study of miscibility and phase separation in blends and semi-interpenetrating networks of 13 C-labeled poly(styrene-co-acrylonitrile) and poly(styrene-co-maleic anhydride). Macromolecules 31:7404–12. 102. Kourtesis G, Renwick GM, Fischer-Cripps AC and Swain MV (1997) Mechanical property characterization of a number of polymers using uniaxial compression and spherical tipped indentation tests. J Mater Sci 32:4493-500. 103. Sultan JN and McGarry FJ (1974) Tensile crazing and shear banding of styrene A. Temperature and rate effects. Polym Eng Sci 14:282–7. 104. Donald AM and Kramer EJ (1982) Plastic deformation mechanisms in poly(acrylonitrile-butadiene styrene) [ABS]. J Mater Sci 17:1765-72. 105. Jabarin SA (1991) Orientation and properties of acrylonitrile copolymers. Polvm Eng Sci 31:644–51. 106. Schellenberg J and Hamann B (1993) Influence of styrene-acrylonitrile oligomers on the properties of ABS graft copolymers. Eur Polym J 29:727–30. 107. Hasha DL, Priddy DB, Rudolf PR, Stark EJ, De Footer M and Van Damme F (1992) Oligomer formation and the mechanism of initiation in the spontaneous copolymerization of styrene and acrylonitrile. Macromolecules 25:3046–51. 108. Adao MHVC, Saramago BJV and Fernandes AC (1999) Estimation of the surface properties of styrene—acrylonitrile random copolymers from contact angle measurements. J Colloid Interface Sci 217:94—106. 109. Siebourg W, Schmid H, Rateike FM, Anders S and Lower H (1990) Birefringence An important property of plastic substrates for magneto-optical storage disks. Polym Eng Sci 30:1133–9. 110. ASTM D 4674–89 (1989) Accelerated testing for color stability of plastics exposed to indoor fluorescent lighting and window-filtered daylight. 111. ASTM D 1003-92 (1992) Haze and luminous transmittance of transparent plastics. 112. ASTM D 2457-90 (1990) Specular gloss of plastic films and solid plastics. 113. Nikolov ID and Ivano CD (2000) Optical plastic refractive measurements in the visible and the near-infrared regions. Appl Opt 39:2067-70.
This page intentionally left blank
14
Rubber Particle Formation in Mass ABS GILBERT BOUQUET The Dow Chemical Company, Midland, Ml, USA
1
MANUFACTURE OF ABS
The commercial production of acryloritrile—butadiene—styrene (ABS) formulations is accomplished by a number of different methods based on free radical polymerization. The two main methods are based on emulsion or solution polymerization techniques. The solution polymerization is mostly called mass or bulk polymerization because only a low amount of solvent is used. Most of the ABS (~85%) is made using the emulsion process. Both techniques have been used in combination (emulsion/mass). Other combinations are with suspension polymerization as final step (mass/suspension and emulsion/suspension) [1]. Although the emulsion process is commercially the most important, the mass process cannot be neglected because it has a number of advantages that will become clear from the more detailed description of both processes.
1.1
EMULSION PROCESS
Although more complicated than the mass process, emulsion polymerization is still widely used because of its greater flexibility [2,3]. The first step is the preparation of the rubber latex using emulsifiers. The crosslinking of the rubber occurs simultaneously during polymerization and is controlled by initiator level, chain transfer agent and process conditions. It is common to increase the particle size by agglomeration, thereby achieving a reduction in cycle time. Modern Stvrenic Polymers: Polystyrene and Styrenic Copolymers. Edited by J. Scheirs and D. B. Priddy i(.-i 2003 John Wiley & Sons Ltd
306
G. BOUQUET
The next step is the polymerization of styrene and acrylonitrile in the presence of the rubber latex. Part of the polymerized styrene—acrylontrile is grafted on to the rubber. This grafted rubber concentrate is then either mixed with additional emulsion-prepared styrene-co-acrylonitrile (SAN) copolymer and then coagulated or first isolated and then compounded with SAN.
1.2
MASS PROCESS
In this process uncrosslinked rubber is dissolved in a mixture of the monomers and solvent(s). This solution is pumped into the first reactor which is connected to a series of reactors. The polymerization is started by increasing the temperature, eventually in the presence of an initiator. Most of the rubber grafting and particle sizing happen early in the process. Chain transfer agent level, initiator (type/amount) and shear have a great influence in this stage. Crosslinking of the rubber particles occurs later in the process. The final step is the removal of residual monomer and solvent. The advantages of the mass process are the absence of contaminants (emulsifiers) and a lower waste generation (no white water). Possible disadvantages are related to heat removal, viscosity and limitations in rubber particle sizing.
2
PHASE SEPARATION
When the homogeneous mixture of rubber in styrene—acrylonitrile—solvent mixture is heated, eventually in the presence of an initiator, SAN copolymer is formed. At about 2 % conversion the critical point is reached, meaning that phase separation has occurred. The first observation of opacity does not necessarily indicate the critical point, but often it is very close to it. It was reported as early as 1947 that a solution of two polymers in the same solvent, in most cases, separates into two phases [4]. This incompatibility is a direct result of the extremely small entropy of solution for high molecular weight polymer pairs. At the critical point there are two phases. The first is a continuous phase containing the rubber swollen in the monomer-solvent mix. The second is the dispersed or discontinuous phase consisting of SAN swollen in the monomersolvent mix. The system forms a polymeric oil-in-oil emulsion [5,6]. The rate of separation into two layers is greater the lower the temperature, even if the viscosity is higher. Kuhn and co-workers found that phase separation in a system close to its critical point can be reversed by shear and also by an increase in temperature [7–10].
RUBBER PARTICLE FORMATION IN MASS ABS
3
307
PHASE INVERSION
As the polymerization proceeds, more SAN is formed. This results in an increase in the discontinuous phase volume. At a certain degree of conversion the phase volume of each phase will be identical. If agitation is applied, phase inversion will occur. Thermodynamically, phase inversion is an instantaneous process; however, often there is a considerable delay in reaching the new equilibrium condition. The main reason is the necessity for significant mass transfer between the phases, which is retarded by the high viscosity. Furthermore, the graft copolymer has a stabilizing effect at the interface. The SAN phase will gradually become the continuous phase with dispersed rubber droplets. Although agitation is required for phase inversion to occur, it is not the cause of it. 4
PHASE DIAGRAM
The polymerization sequence can be plotted using a phase diagram (Figure 14.1) [11]. This diagram is representative for nearly all polymer (rubber)polymer (glassy polymer)-common solvent systems. Point a is the composition of the feed. The critical point (phase separation) is reached at point b. Further polymerization generates more glassy polymer (SAN) and phase inversion occurs at point c (rubber continuous —> discontinuous). Point d is reached when all monomer is converted. Monomer of Glassy Polymer
Rubber
Glassy Polymer
Figure 14.1 Phase diagram for rubber—glassy polymer-monomer system [11] a, Initial composition; b, phase separation; c, phase inversion; d, complete conversion. Reproduced from W. A. Ludwico and S. L. Rosen, J.Appl.Polym.Sci., 19, 757 (1975) with permission of John Wiley & Sons, Inc.
308
5
G. BOUQUET
RUBBER PARTICLE SIZING
After phase inversion, the dispersed phase is the rubber phase. This is the first time discrete rubber particles are present in the reaction mixture. The particle size in the final product is an important parameter to optimize the physical properties. To be successful in the manufacturing of mass ABS, it is necessary to understand and control which parameters can be used to control the final rubber particle size. The following parameters will be discussed: • shear; • viscosity; • interfacial tension. 5.1
SHEAR
Spontaneous phase inversion (no shear) has been described by Keskkula [12]. This was demonstrated during the quiescent polymerization of styrene-polybutadiene mixtures containing less than 3 wt% polybutadiene. For industrially important systems (higher rubber content), a minimum amount of shear is required [13]. If no adequate agitation is applied, the system will solidify in the emulsion state before the inversion point. The final product will then consist of a continuous phase of a crosslinked polybutadiene network with dispersed SAN particles. Such a material will not have the typical properties of ABS. A level of agitation above the minimum shear can be used to reduce the rubber particle size. However, agitation is not always effective in reducing the rubber particle size (see section 5.2) and there is also an upper limit based on hardware limitations (torque). Another aspect that plays a role related to shear is the feed rate. Increasing the feed rate means that the residence time in the reactor(s) is shorter. A lower amount of shear is transferred to the prepolymer, giving in general larger rubber particles in the end product. 5.2
VISCOSITY
Two factors related to viscosity play a role in the rubber sizing process: • ratio (p) of disperse phase viscosity (//) and continuous phase viscosity (/*); • viscosity of the continuous phase. Rumscheidt and Mason [14] described particle deformation in a shear field as a function of viscosity ratio (p). There is a minimum and a maximum viscosity ratio where it becomes impossible to reduce the droplet size. The limits described by Karam and Bellinger [15] are 0.005 and 4. Breakup of droplets readily occurs when the viscosity ratio is of the order of 0.2:1. Intuitively, a ratio of 1 would be best because in this case there is a maximum transfer of energy between the
309
RUBBER PARTICLE FORMATION IN MASS ABS
continuous and dispersed phase. This assumption is probably only valid if the interfacial tension is nearly zero and there is no viscous or rigid film at the interface. It was also found that the higher the viscosity of the continuous phase, the greater the ease of particle breakup. Both effects are illustrated in Figure 14.2. The rubber (dispersed) phase viscosity is determined by the rubber level and by the solution viscosity of the rubber. Furthermore, the grafting and crosslinking will also influence the viscosity. The SAN (continuous) phase viscosity is controlled by the molecular weight of the copolymer. With standard rubbers, the viscosity ratio tends to be higher than 1. This is not the optimum and to reduce it the rubber viscosity can be decreased or the SAN viscosity can be increased. Decreasing the rubber viscosity can be achieved by using low solution viscosity rubbers. These rubbers typically are star branched rubbers or block copolymers (styrene—butadiene block). Branching or introduction of a polystyrene block is done to control the cold flow of these materials to avoid problems during storage. Table 14.1 contains a short overview of typical rubbers that are used in mass ABS production. The polybutadiene chain contains a lot of double bounds (allylic hydrogens) that will react with radicals during the copolymerization. Possible reactions are grafting (see Section 5.3) and crosslinking. Both reactions, but especially crosslinking have an important 140
POLYGLYCOL-VEHICLE DC-200 DROPLET
120
RADIUS OF DROPLET 0.08 cm //-0.80 Poise
100
SHEAR RATE AT BREAK UP, Gb, s-1
/x-2.4 Poise
VISCOSITY RATIO, //'/// Figure 14.2 Influence of viscosity ratio and continuous phase viscosity (/i) on drop breakup [15]. Reprinted with permission from H. J. Karam and J. C. Bellinger, Ind. Eng. Chem. Fundam., 7, 575 (1968). Copyright 1968 American Chemical Society
310
G. BOUQUET
Table 14.1 Rubbers used in mass ABS process Producer Type
Solution A/w viscosity (mPas) (g/mol) Mw/M/n Type
Bayer Buna CB HX528 Bayer Buna CB HX565 Firestone Stereon 730
160 45 25
260000 205000 140000
1.4 1.2 1.3
Composition
Linear Polybutadiene Branched Polybutadiene Block 30%Styrene
influence on the rubber molecular weight, and hence viscosity. Crosslinking and the consequences it has on the rubber sizing process will be discussed later. To increase the SAN viscosity, the molecular weight generally has to be increased. Tools are acrylonitrile content, polymerization temperature, initiator (type/concentration), solvent level and chain transfer agent (amount and timing of addition). A higher acrylonitrile content also results in a higher matrix viscosity. Increasing polymerization temperature gives lower molecular weight. Increasing the concentration of monofunctional initiators also results in lower molecular weight. Increasing the functionality of the initiator increases the matrix molecular weight. Often the solvent used has some chain transfer activity, hence an increase in solvent level will give lower molecular weight. Furthermore, the solvent and unreacted comonomers are recycled. During the manufacturing process, side reactions occur, generating products that build up in the recycle. These products often show chain transfer activity. 5.3
INTERFACIAL TENSION
During the sizing process, the rubber phase is becoming increasingly finely dispersed in the SAN matrix. During this process, the surface area is increased. This process requires less energy when the interfacial tension is low. A reduction of the interfacial tension can be achieved by: • utilization of block rubbers; • grafting. Utilization of block rubbers is being successfully used for the production of high-impact polystyrene (HIPS). The styrene—butadiene block rubbers act as emulsifier for the polystyrene matrix and the rubber particles. Echte has extensively described the possible morphologies (cellular, core-shell, rods, droplets, etc.) that can be obtained [16,17]. The same principle cannot be used in ABS because the required block rubbers (SAN—butadiene) are not commercially available. The required SAN-polybutadiene copolymers can be generated in situ during the copolymerization of SAN. This grafting process will be discussed in detail in the next section because grafting is considered a major tool to control rubber particle morphology and thus the physical properties of mass ABS.
RUBBER PARTICLE FORMATION IN MASS ABS
6
311
GRAFTING
The mechanisms involved in grafting and the associated kinetics have not been completely resolved. Many articles have been published related to this topic, each describing specific aspects of the process [18–25]. Polybutadiene contains two reactive groups: double bonds and allylic hydrogens. A radical can add to the double bond, thereby creating a radical on the rubber backbone. If an allylic hydrogen is abstracted, again a radical is created on the backbone but the original radical does not become incorporated as is the case if addition occurs. Once a radical is present on the rubber, the addition of monomer units can start. During this propagation step the graft is created nearly instantaneously according to free radical kinetics. 6.1
GRAFT ANALYSIS
One of the problems in the study of the grafting reaction is the separation and quantification of the different species present in the prepolymer. The three polymeric components present in the prepolymer are: • free SAN; • free rubber; • graft rubber. The analysis of the prepolymer at high conversion and end product is even more problematic because additional crosslinking has occurred. Hughes developed a simple test based on the 'emulsifying' power of an artificial prepolymer containing the copolymer to be tested [26]. A more elaborate technique technique based on extraction was developed by Llauro [27] and Riess [28] based on the reversible crosslinking of the rubber (free + graft). This reversible crosslinking was achieved by attaching -COONa groups to the rubber. The dipole—dipole interaction between the -COONa groups in nonpolar solvents gives gelation of the rubber phase. The free SAN remains soluble and can be removed. Adding methanol breaks up the gel. Huang and Sundberg have described a method based on gel permeation chromatography using a dual detector set-up [29]. However, this method can only be used when the rubber and matrix molecular weights are different. Furthermore, block rubbers cannot be used. Bouquet et al. have developed a method to separate and measure the different fractions, but the actual technique is not disclosed [30]. 6.2 6.2.1
EFFECT OF PROCESS
PARAMETERS
Nomenclature
Graft efficiency = ratio of graft SAN to total SAN
312
G. BOUQUET
Yield graft rubber = ratio of graft rubber to total rubber Graft density = average number of graft chains per grafted rubber molecule
6.2.2
Monomer
Locatelli and Riess showed that the molecular weight of the SAN (free and graft) increased as the monomer concentration was raised [31]. This is in agreement with classical free radical kinetics. Increasing the monomer concentration results in a lower graft efficiency, but the graft density is higher [32]. At higher monomer concentration it is possible that an increasing amount of free SAN is formed in the rubber phase, explaining why the graft efficiency is lowered.
6.2.3
Initiator
Some initiators are more effective in creating radicals on the rubber chain compared with others. Initiators yielding phenyl, benzoyloxy and tert-butoxy radicals are described as very efficient for grafting. A possible explanation is that this type of radical can stabilize the negative charge present in the transition state during hydrogen abstraction. According to Allen et al., benzoyloxy radicals preferentially add to the double bond [24]. The phenyl radicals formed from the benzoyloxy radicals by CO2 loss are more reactive towards allylic hydrogens. 2,2'-Azobis(isobutyronitrile) (AIBN), which generates tertiary carbon radicals, reacts only with 1,2-vinyl units according to Locatelli and Riess [24]. The higher reactivity of the 1,2-vinyl groups is caused by the labile tertiary allylic hydrogens. Hayes and Futamara [25] concluded that AIBN and benzoyl peroxide only generate grafting through copolymerization. None of the graft copolymer is generated by hydrogen abstraction. They consider the reaction of polybutadiene with styrene and acrylonitrile as a true terpolymerization.
6.2.4
Rubber
Riess and Locatelli studied the influence of the 1,2-vinyl content of the elastomer on the overall polymerization rate [33]. The rate is increased as the 1,2-vinyl content of the rubber is higher. The presence of 1,2-vinyl units automatically means that allylic tertiary hydrogens are present, which are easily abstractable by a radical. This gives an increase in the radical sites on the polybutadiene backbone, thereby indirectly boosting the initiator efficiency. Supporting the previous statement, Locatelli and Riess showed that the graft efficiency increased as the 1,2-vinyl content of the rubber is higher [32]. The polystyrene fragment in a block rubber is considered inert for the grafting reaction.
RUBBER PARTICLE FORMATION IN MASS ABS
313
The effect of rubber concentration on grafting was also studied [31]. Adding more rubber resulted in an increase in the graft SAN molecular weight. The explanation given is based on the Trommsdorf effect. A higher rubber concentration means that the rubbery phase becomes more viscous, thereby reducing the termination rate, hence higher molecular weights are obtained. When benzoyl peroxide was used, the graft efficiency increased as rubber concentration was higher. This was not the case when AIBN was used [32]. The higher the rubber concentration, the higher is the probability that a graft active site will be created. When AIBN is used, the grafting is minimal and seems to be limited to a plateau value.
6.3
MASTER CURVE
To understand the grafting process better and to be able to control product properties, it is important to know how fast the rubber is converted to grafted rubber. The yield of graft rubber is a measure of this. If the yield of graft rubber is zero, none of the rubber molecules are grafted. If the yield of graft rubber is 1, then all the rubber chains are grafted with at least one SAN graft. The yield of graft rubber has been determined for a number of initiators at different temperatures [30]. In both cases the observed trend was very similar. At the beginning of the copolymerization the yield increases very fast, reaching a plateau around a solids level of 25 %. A higher yield is obtained when the initiator concentration is increased and when the polymerization temperature is higher. The ideal case would be to combine this information, making it possible to determine the yield under any conditions (initiator type/concentration/temperature). This has been attempted based on the idea that mainly the primary radicals initiate the grafting reaction. The initiator decay determines the rate of primary radical formation. The yield at 25 % solids (plateau) is plotted versus the mean of the initiator decay rate during the timeframe to obtain 25 % solids. This is called the master curve. All the data points for one initiator fall on the same line, which is an indication that primary radical formation indeed plays a dominant role in the grafting process.
6.4
GRAFT MODEL [30]
A graft model has been proposed to explain why it is practically impossible to graft all the rubber during the manufacturing of ABS. With a high amount of initiator (l000ppm benzoyl peroxide), only 70% of the rubber molecules are grafted. For AIBN-initiated runs, the yield even drops to 0.3 (70% of the rubber is not grafted). This is unexpected because the rubber used (Bayer Buna CB HX528) contains 14500 active sites (double bonds/allylic hydrogen)
314
G. BOUQUET
per molecule and per rubber molecule 16 primary radicals are in theory available to initiate the grafting. Brydon et al. performed experiments in HIPS under high dilution and showed that a yield close to the theoretical maximum of 1 is obtainable [34]. A partial explanation for the shielding of the free rubber is offered by Rosen [35]. The reacting system places an upper limit on the amount of rubber that can be grafted, regardless of the chemical nature of the chemistry involved in the grafting process. The limit is caused by the inherent compatibility of the two polymeric phases. If the initiator decomposes in the SAN phase, the primary radical will never be able to initiate the graft reaction because there is no rubber in its proximity. Only initiator that decomposes in the rubber phase has the potential to start the graft reaction. This means that when the polymerization is done in a one-phase system, theoretically all the rubber can be grafted (high dilution). Based on this theory it can be concluded that most of the grafting takes place in the early stages of the polymerization (small SAN phase versus rubber phase). The free rubber is also protected by a chemical restriction. The only species that can create a radical on the rubber backbone is the primary radical. An initiator molecule, present in the rubber phase, decomposes in a solvent cage, and in the case of ABS, the cage contains solvent, styrene, acrylonitrile and eventually chain transfer agent. The primary radical has to escape from the cage to reach a polybutadiene fragment before it reacts with the surrounding monomers or chain transfer agent, which are very efficient radical scavengers. Combining the physical and chemical restrictions imposed upon the grafting process can explain why it is impossible to graft all the rubber. This model is summarized in Figure 14.3.
7
CROSSLINKING
During the grafting process, a radical site (addition/abstraction) is created on the rubber backbone. Propagation of styrene—acrylonitrile results in the SAN graft. The final step in the grafting sequence is the termination of the radical site. Figure 14.4 gives an overview of the different possibilities. Chain transfer and disproportionation result in a grafted rubber with 'normal' SAN molecular weight. Termination by combination with a growing SAN chain also gives a grafted rubber but in this case the molecular weight of the graft is higher (double). A final possibility is combination with another growing graft. The result in this case is a crosslink between two rubber chains. This type of termination increases the molecular weight dramatically and will be reflected in the viscosity of the rubber phase. The viscosity of the rubber phase has a large influence on the sizing process that takes place during the production of ABS. As the viscosity ratio (rubber/SAN) increases, sizing will become more difficult.
315
RUBBER PARTICLE FORMATION IN MASS ABS
Figure 14.3
Model for grafting process showing the physical and chemical restrictions
Termination by transfer or disproportionation
rubber backbone SAN graft
Termination by combination with SAN
Termination by combination with graft 'pre-crosslinking' Figure 14.4
Termination processes during grafting
The most common way to measure crosslinking is by swelling. For systems containing only rubber, the Flory—Rehner equation can be applied. For ABS that is not the case because grafted and occluded SAN will interfere with the swelling behavior. Nevertheless, this method is being used frequently for practical reasons. Karam and Tien have developed a quantitative analysis for the swelling behavior of a heterogeneous gel [36]. They described quantitatively the crosslink density when the proportion of occlusions is known. DMA and DSC techniques are typically insufficiently sensitive to detect small changes in crosslink density. Relaxation experiments based in 13C NMR have been performed but is was difficult to obtain absolute data using this technique [37]. The
316
G. BOUQUET
physical properties of ABS are greatly influenced by the degree of crosslinking. Optimization of crosslinking to control product properties has been described in the patent literature [38,39]. Peng has studied the kinetics of the crosslinking process for HIPS in the absence of initiator [40]. In the absence of styrene and oxygen, polybutadiene cannot be crosslinked by heating alone. For HIPS most of the crosslinking takes place at high conversion. In the presence of acrylonitrile (ABS), the onset of crosslinking starts earlier in the polymerization.
8
SIZING WINDOW
The sizing window is defined as the period during the manufacturing of mass ABS where the rubber particle size can be controlled. The sizing window starts at the point of phase inversion. When the required minimum amount of shear is applied, the rubber phase will change from continuous to discontinuous phase. At this point the rubber particles are very large and the particle size distribution is very broad. The rubber particles will become smaller if shear is applied at the right viscosity ratio. The presence of graft (emulsifier) will facilitate the sizing process. At a certain point in time the rubber phase viscosity will increase because the monomer concentration is reduced through copolymerization and crosslinking becomes important. If the viscosity becomes too high, the sizing will cease even if shear is applied. The period between phase inversion and an excessive viscosity ratio is called the sizing window. Only shear applied in this region will be efficiently used to reduce the particle size and make the particle size distribution narrower. If small particles are required it is important to apply high shear where the viscosity ratio is optimal (0.2-1) and interfacial tension is reduced by grafting. Related to particle sizing, Molau and Kesskula described the concept of type I and II occlusion [5]. The prepolymer is viscous and has a retarding effect on the phase inversion. In most cases multiple emulsions are formed after the phase inversion point. If the agitation is not extremely high these multiple emulsions survive the further copolymerization and give SAN occlusions in the rubber particles. These occlusions are called type I. Type II occlusions are formed when monomer dissolved in the rubber phase is copolymerized. Because SAN is not compatible with the rubber, separation occurs within the rubber particle, giving type II occlusions.
317
RUBBER PARTICLE FORMATION IN MASS ABS 9
RUBBER PARTICLE MORPHOLOGY
In HIPS a wide variety of rubber particle morphology is possible. Echte et al. summarized this in an excellent review (see Figure 14.5) [41]. The key to these different structures is the composition of the styrene—butadiene block rubber, which is an emulsifier for the polystyrene—polybutadiene system. Additional grafting can generate a shift from one structure to another. In the case of mass ABS, the variety of rubber particle morphology is less diverse. Typical examples of morphology are shown in Figure 14.6. If polybutadiene rubber is used (linear or star), cellular particles are obtained with SAN occlusions. In the case of styrene—butadiene block rubber (typically 30% styrene) also cellular particles are obtained but besides the SAN occlusions, polystyrene domains are clearly visible in the particles. To be able to make the other morphologies that are possible in HIPS, the interfacial tension has to be manipulated. Controlling the grafting reaction is a way to achieve this but the possibilities are limited with the tools (mainly initiator) that are currently available.
if
Droplets
•j
V *>%?*'• Ml
Rods
Capsules
.
-
. ft
Rod clusters
Figure 14.5 Possible rubber particle morphology in HIPS
Droplet clusters
318
Figure 14.6
G. BOUQUET
Rubber particle morphology in mass ABS as function of rubber type
REFERENCES 1. Adams ME, Buckley DJ, Colborn RE, England WP, Schissel DN (1993) Acrylonitrile-Butadiene-Styrene Polymers. Rapra Review Reports, Vol. 6, No. 10. 2. Calvert W, US Patent 3 238 275 (to Borg Warner). 3. Bovey FA, Kolthoff IM, Medalia IM, Meehan EJ (1955) Emulsion Polymerization, Interscience, New York. 4. Dobry A, Boyer-Kawenoki F (1947) J. Polym. Sci. 2, 90. 5. Molau GE, Keskkula H (1966) J. Polym. Sci. A-l 4, 1595. 6. White JL, Patel RD (1975) J. Appl. Polym. Sci. 19, 1775. 7. Silberberg A, Kuhn W (1952) Nature (London) 170, 450. 8. Silberberg A, Kuhn W (1954) J. Polym. Sci. 13, 21. 9. Burkhardt F, Majer H, Kuhn W (1960) Helv. Chim. Acta 43, 1192. 10. Kuhn W, Majer H, Burkhardt F (1960) Helv. Chim. Acta 43, 1208. 11. Ludwico WA, Rosen SL (1975) J. Appl. Polym. Sci. 19, 757. 12. Keskkula H (1979) Plasti. Rubb.: Mater. Appl. 16, 71. 13. Freeguard GF, Karmarkar MJ (1957) J. Appl. Polym. Sci. 15, 1657. 14. Rumscheidt FD, Mason SG (1961) J. Colloid Sci. 16, 238. 15. Karam HJ, Bellinger JC (1968) Ind. Eng. Chem. Fundam. 7, 576. 16. Echte A, (1977) Angew. Makromol. Chem. 58/59, 175. 17. Echte A, Haaf F, Hambrecht J (1981) Angew. Chem., Int. Ed. Engl. 20, 44.
RUBBER PARTICLE FORMATION IN MASS ABS 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.
319
Huang NJ, Sundberg DC (1995) J. Polym. Sci. A: Polym. Chem. 33, 2533. Huang NJ, Sundberg DC (1995) J. Polym. Sci. A: Polym. Chem. 33, 2551. Huang NJ, Sundberg DC (1995) J. Polym. Sci. A: Polym. Chem. 33, 2571. Huang NJ, Sundberg DC (1995) J. Polym. Sci. A: Polym. Chem. 33, 2587. Walling C, McElhill EA (1951) J. Am. Chem. Soc. 73, 2979. Locatelli JL, Riess G (1973) Angew. Makromol. Chem. 32, 161. Allen PW, Ayrey G, Moore CG (1959) J. Polym. Sci. 36, 55. Hayes RA, Futamura S (1981) J. Polym. Sci. Polym. Chem. Ed. 19, 985. Hughes LI, Brown GL (1963) J. Appl. Polym. Sci. 7, 59. Llauro MF (1970) Thesis, Ecole Superieure de Chimie, Mulhouse. Riess G, Locatelli JL (1975) Adv. Chem. Ser. 142, 186. Huang NJ, Sundberg DC (1994) Polymer 35, 5693. Bouquet G, Kentie WC, De Theije PJG, Van Damme F (1996) Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.) 37, 536. Locatelli JL, Riess G (1972) Angew. Makromol. Chem. 28, 161. Locatelli JL, Riess G (1973) Angew. Makromol. Chem. 32, 117. Riess G, Locatelli JL (1973) Angew. Makromol. Chem. 32, 101. Brydon A, Burnett GM, Cameron GC (1973) J. Polvm. Sci., Polym. Chem. Ed. 11, 3255. Rosen SL (1973) /. Appl. Polym. Sci. 17, 1805. Karam HJ, Tien L (1985) J. Appl. Polym. Sci. 30, 1969. Curran SA, Padwa AR (1987) Macromolecules 20, 625. Vanspeybroeck RS, Galobardes MR, Maes D, Jones MA, Ceraso JM (2001) US Patent 6/211/298 (to The Dow Chemical Company). Iwamoto M, Nakajima A, Takaku M, Morita H, Ando T, Shirafuji T, Uchida M (1996) US Patent 5/552/494 (to Mitsui Toatsu Chemicals). Peng FM (1990) J. Appl. Polym. Sci. 40, 1289. Echte A, Haaf F, Hambrecht J (1981) Angew. Chem., Int. Ed. Eng. 20, 344.
This page intentionally left blank
15
High Heat Resistant ABS Technology RONY VANSPEYBROECK, ROBERT P. DION, AND JOSEPH M. CERASO The Dow Chemical Company, Midland, Ml, USA
1
INTRODUCTION
Conventional acrylonitrile—butadiene—styrene terpolymer (ABS) is a thermoplastic polymer, consisting of a rigid styrene-co-acrylonitrile (SAN) continuous phase and a dispersed rubber phase [1,2]. Some SAN copolymer is chemically bonded (grafted) to the rubber, which is usually polybutadiene or a styrene—butadiene copolymer. The rubber reinforcement of SAN increases its ability to withstand high speed impact at the expense of most other properties [3,4]. Compared with rubber-modified high-impact polystyrene (HIPS), the modification of the polystyrene continuous phase by copolymerization with acrylonitrile results in resins that have higher tensile strength, higher toughness, better solvent resistance and improved heat resistance, while maintaining excellent processability [5,6]. The typical physical properties of ABS are affected by the rubber and the acrylonitrile content, the rubber particle size, rubber particle size distribution, rubber morphology, and the rigid phase molecular weight (both grafted and free). Table 15.1 shows typical values for the properties of selected SAN, HIPS and ABS. The heat resistance performance of a resin is the temperature at which a part will start to be dimensionally unstable and to distort under load. There are various laboratory tests to predict this performance. The most commonly used are the measurement of the Vicat softening point (ASTM D1525, DIN 53460, ISO 306) and the deflection temperature under load (DTUL, ASTM D648). Both measurements monitor the modulus change with temperature, and determine an endpoint when a macroscopic change can Modern Styrenic Polymers: Polystyrene and Styrenic Copolymers. Edited by J. Scheirs and D. B. Priddy C) 2003 John Wiley & Sons Ltd
322 Table 15.1
R. VANSPEYBROECK ETAL. Properties of SAN copolymers, ABS and HIPS
Vicat softening point ( °C) Tensile yield (MPa) Tensile modulus (MPa) Notched Charpy impact (kJ/m2) Melt flow rate (cm3/10/min)
SANb
ABSC
HIPSd
103 63
101 48
88 25
3400
2300
1770
2 25
24 16
18 4.0'
a
Vicat: ISO 306; 50 N/ 50°C/hr. Tensile properties: ISO 527-1,2. Charpy: ISO 179. MFR:ISO1133;220°C, 10kg. * TYRIL® 990 (The Dow Chemical Company). c MAGNUM® 9020 (The Dow Chemical Company). d STYRON® A-TECH 1110 (The Dow Chemical Company). e HIPS @ 200 °C, 5kg.
be detected in a part under load. The Vicat is the temperature at which a flatended needle penetrates a test sample to a depth of 1 mm under a particular load and uniform heating rate, while the DTUL is the temperature at which a test specimen on edge in an oil bath deflects 0.025 cm under a stress of 0.46 or 1.82 MPa (66 or 264 psi) at a heating rate of 2 °C/min. The DTUL test result is influenced by the sample thickness, the molding conditions, and whether or not the sample has been annealed. Since these tests are influenced by both fabrication conditions and part geometry, additional evaluations are required for final qualification in an application. Often, molded parts are placed in an oven at a specified heat condition and the parts are measured for dimensional changes. The heat resistance performance strongly correlates with the glass transition temperature (7g) and the stiffness (modulus) of the resin. In this two-phase system, the Tg of the rigid SAN continuous phase is modestly influenced by the acrylonitrile content and increases approximately 0.3–0.4 °C per weight percent of added acrylonitrile. Within the range of commercial ABS resins, the acrylonitrile content is typically 20–30 wt%. The molecular weight of the SAN continuous phase is very important for determining the toughness and flow balance of the resultant ABS, but has no significant influence on the heat resistance. However, low molecular weight SAN oligomers, residual monomers, or additives such as flow promoters that act as plasticizers all have a negative effect on the heat resistance. The thermal properties are also influenced by the rubber phase. A higher rubber content or a higher rubber phase volume results in a lower modulus. The heat performance of conventional ABS correlates in general with the glass transition temperature (Tg) of the rigid phase. Table 15.2 lists some typical rgs of amorphous polymers. Also listed are the crystalline melting points (r m ) for semi-crystalline polymers. Typically, the heat performance of a neat semicrystalline polymer under low load correlates with its Tm.
HIGH HEAT RESISTANT ABS TECHNOLOGY Table 15.2 tures
323
Typical polymer glass transition and semi-crystalline melting tempera-
Amorphous
~r g (°C)
Semi -crystalline
T*
PS HIPS SAN (30%AN) ABS HH ABS PC
100 100 108 105 110-120 151
Syndiotactic-PS Nylon 6,6 PBT Acetal
270 265 222 175
Tm f( O/*"~*\ C)
Although syndiotactic polystyrene (SPS) consists of pure polystyrene, the tacticity has a dramatic effect on chain dynamics and causes crystallization to occur. This will be examined in a separate chapter. High heat ABS resins described in this chapter refer to a family of multiphase polymers which are dimensionally stable at temperatures where conventional, general-purpose ABS deforms. The maximum end-use temperature at which fabricated parts can meet the functional requirements of the specific application determines whether high heat resistant ABS or conventional ABS will be used. Typical applications for high heat ABS are automotive interior components that can be exposed to temperatures of more than 90 °C. Conventional ABS warps unacceptably in that environment. The heat resistance of ABS can be improved by adding a high-Tg polymer, by chemically modifying the SAN copolymer, or by removing low molecular weight plasticizers. This review will not cover the multiphase blend approach, exemplified by extruded mixtures of polycarbonate and ABS. The focus will concentrate on the chemical modification of ABS by the use of a different, continuous-phase styrenic copolymer. This review describes technical solutions for high heat resistance which make use of the incorporation of new monomers to increase the Tg of the continuous phase through chain stiffening and/or modification of the cohesive energy density. The new monomers can be added directly through terpolymerization with styrene and/or acrylonitrile. Alternatively, they can be copolymerized with styrene and post-blended with conventional high rubber content ABS impact modifiers. Some of the process chemistry influence on residuals will be briefly discussed for tx-methylstyrene termonomer high heat ABS. An overview of the product technologies is also described. The processes are designated but their description will be reviewed in other chapters. The most commonly used commercial monomers that increase the Tg of ABS are maleimides (most recent commercial addition), maleic anhydrides (MA), and substituted styrenes. The majority of ABS produced in North America and Europe relies on a-methylstyrene (aMeS) or N-phenylmaleimide (PMI) as
324
R. VANSPEYBROECK ETAL
termonomers to raise the ToR of the continuous *phase. A small fraction of the ABS market contains MA.
2 SUBSTITUTED STYRENES Table 15.3 lists a series of substituted styrenes that can be used to produce more heat-resistant polymers [7]. The substituent results in a higher rotational energy of the C—C bonds in the polymer backbone and consequently a stiffer polymer chain and higher Tg. The most widely practiced technology uses a-methylstyrene. This is mainly due to its favorable price/performance balance. ot-Methylstyrene is produced by cumene oxidation and occurs as a byproduct in the manufacture of phenol and acetone (Figure 15.1). About 40% of the monomer is used to produce heat-resistant ABS. For styrenic polymers containing 20 wt% of AN, the glass transition temperature increases by about 0.4 °C for every 1 % of styrene that is replaced with otMeS. a-Methylstyrene can be incorporated into the continuous ABS phase using the known processes for conventional ABS, including emulsion [8-12], mass [13], suspension [14], and mass/suspension [15]. However, the incorporation of aMeS into SAN is complicated by two factors, which determine the limitations on the use of aMeS for producing heat-resistant ABS at the rapid rates desirable in a manufacturing environment. The reactivity ratios of aMeS and AN are r\ =0.07 and r2 = 0.15, respectively, which results in the polymer tending toward alternating compositions. The azeotropic composition is approximately 50 mol% (70 wt%) aMeS. As one attempts to produce copolymers with very high aMeS content, the polymerization rate decreases. As the ceiling temperature for polymerization of aMeS is 61 °C [16,17], high-temperature polymerization of copolymers containing high levels of aMeS proceeds slowly and favors the production of low molecular weight or even oligomeric species, which results in poor toughness and heat properties. Table 15.3
Glass transition temperatures of substituted styrene homopolymers
Position
Substituent
— 4 2 2,5 4 a
— tert- Butyl Methyl Dimethyl Phenyl Methyl
Tg
("C) 100 128 136 143 161 168
325
HIGH HEAT RESISTANT ABS TECHNOLOGY CH7
CH
Figure 15.1
H2SO4
Phenol-acetone process to produce a-methylstyrene
The consequences of the aMeS kinetics are that a complete mass ABS product is possible but is limited in the heat resistance attainable. At the applied hightemperature range (100–160°C), polymerization rates are slow and oligomer formation is high if the weight percent of acrylonitrile in the monomer feed is too low [14]. The high levels of oligomeric species plasticize the ABS and greatly reduce its heat resistance, decreasing the benefit of incorporating aMeS. The aMeS incorporated in commercial mass ABS resins is as high as 45 wt%. Most of the ABS producers around the world produce HHABS with aMeS technology. Owing to the composition drift and the need for economical rates of reaction, a continuous stirred tank reactor (CSTR) configuration is required. The reaction is run off azeotrope with high acrylonitrile feed around 50 wt% to enhance the rate of polymerization. An approximately 30:70 AN—aMeS copolymer rigid phase is formed by this process. Some of the HHABS is still manufactured by utilizing 100 % of the emulsion polymerization process. After the polymerization of the polybutadiene latex, aMeS is added along with styrene and acrylonitrile to form the grafted and nongrafted rigid phase. Most of the commercial grades consist of an emulsion SAN grafted polybutadiene containing impact modifier blended with a solution polymerized a-methylestyrene-co-acrylonitrile copolymer. In this approach, aMeS-AN copolymer with up to 30wt% AN is produced in a mass (solution) process. Impact modification is achieved by compounding in an emulsion SAN-grafted rubber concentrate. The aMeS—AN copolymer manufacturing rates are significantly lower than the rate of manufacture for SAN copolymer. aMeS—AN copolymers are not miscible with SAN copolymers [18], but they are sufficiently compatible to produce heat-resistant ABS blends with acceptable properties. Replacing part of the styrene by ot-MeS results in S—aMeS—AN terpolymers. The slower aMeS kinetics result in a rate reduction compared with generalpurpose emulsion ABS. A typical emulsion process for S—aMeS—AN terpolymers is described in a patent [12].
326
R. VANSPEYBROECK ET AL.
Separate manufacturing and blending of aMeS–AN copolymers, S–AN copolymers and grafted rubber concentrates is the oldest way for obtaining high heat ABS and it is practiced by a large number of manufacturers [9–12]. By blending the components in different ratios, families of resins can be designed which vary in heat resistance, melt flow, and toughness. Although SAN and aMeS–AN are generally not miscible, they are sufficiently compatible to produce heat resistant ABS with acceptable property balances. In addition to the traditional free radical polymerization techniques, anionic polymerization at high temperature (100 °C) has been described for S–aMeS copolymers containing up to 67 wt% aMeS [19]. The anionic synthesis of S–otMeS copolymers is not plagued by the low ceiling temperature of 61 °C for radical polymerization. The use of aMeS in polymers of PS and HIPS has not been commercially viable. Much of the higher heat-resistant ABS grades commercially available are produced using emulsion polymerization technology and have high gloss esthetics. When low gloss esthetics are required, such as in many of the automotive interior applications, mass polymerization technology has been utilized. Low gloss applications require larger rubber particles that are most easily produced by a mass polymerization process. Low gloss high heat ABS containing a-methylstyrene or N-phenylmaleimide is commercially produced by the mass (solution) process. When polymers such as ethylene glycidylmethacrylate or styrene-acrylonitrile-methacrylate are added to the ABS formulation, low gloss surface aesthetics are attainable. However, the processing window for fabricating parts with uniform low gloss surfaces is narrower than with mass ABS. Also, reducing the gloss of emulsion polymerized ABS with gloss-reducing polymeric additives tends to increase the melt viscosity of the resin.
3
IMIDES
Incorporation in a polymer of imide groups, that are five-membered planar rings which completely hinder the rotation of the imide residues around the backbone chain of the macromolecule, leads to (co)polymers with great structural stiffness and higher thermal stability. Poly(N-n-alkylmaleimide)s exhibit a Tg corresponding to the number of the carbon atoms in the n-alkyl groups [20], e.g. from 97 °C for poly(TV-n-octadecylmaleimide) (n = 18) to 185°C for poly(TV-n-butylmaleimide) (n = 4). A higher Tg can be obtained by incorporation of N-phenylmaleimide (PMI) [21] and with N-(alkyl-substituted phenyl)maleimides [22]. All the corresponding polymaleimides exhibit excellent thermal stability up to at least 370 °C. Poly(N-phenylmaleimide) undergoes decomposition without softening and melting, when heated to above 400 °C. Other substituted poly(N-phenyl maleimides) soften and melt between 400 and
HIGH HEAT RESISTANT ABS TECHNOLOGY
327
450 °C, accompanying their decomposition. Commercial supply of maleimide monomers started in the mid-1980s. Since that time, the use of imides has increased dramatically. A number of applications have been described, making use of N-alkylmaleimides, such as N-isopropylmaleimide (IPMI) [23] and Ncyclohexylmaleimide (CHMI) [24] for heat-resistant transparent polymer resins, useful, for example, for optical lenses, liquid crystal displays and disks [25]. The most important reason for this is that N-substituted maleimides having an aliphatic or alicyclic group are colorless and most suitable for yielding colorless transparent copolymers, whereas N-arylmaleimides are orange or yellow. However, the majority of process and product developments have been carried out with N-phenylmaleimide (PMI). PMI has also been the most widely used imide for ABS modification. Imide modification leads to some unique properties because it both elevates transition temperatures and thermal decomposition temperatures, allowing for easier fabrication than less stable styrene-containing terpolymers [26]. The deformation temperature of ABS has been shown to increase by 2-3 °C on adding 1 % of PMI. By adding 5–10%, the temperature can be elevated to over 125 °C [27]. Furthermore, facile copolymerization rates and resistance to hydrolysis reactions overcome many of the reaction engineering obstacles encountered by other high heat monomers when scaling to mass or emulsion processes. The synthesis of N-substituted maleimides (Figure 15.2) involves two steps. The first step is a quantitative reaction at ambient temperature between maleic anhydride and a primary amine, yielding a maleamic acid. The second step, requiring elevated temperatures, is ring closure and water elimination. A description of a typical process can be found in the patent literature [28]. Figure 15.3 illustrates that the free radical polymerization of PMI and styrene proceeds in an alternating manner [22,29,30]. Over a wide range of monomer ratio, the copolymerization (at low conversions) results in polymers with PMI content between 45 and 55 % and a Tg between 225 and 245 °C. The same type of alternation is observed in the copolymerization of styrene with maleic anhydride [31] and various N-alkyl-substituted maleimides [32–35]. It is 6
NH,
Figure 15.2
PMI synthesis
328
R. VANSPEYBROECK ET AL.
Figure 15.3 Copolymerization of PMI and styrene: diagram for radical copolymerization of PMI (M1) with styrene (M2) in benzene at 35 °C with AIBN as initiator [22]; f(1) = initial molar fraction of PMI in the monomer feed; F(1) = molar fraction of PMI in copolymers at conversion of 5 % (±0.5 %)
generally accepted that this type of polymerization implies the participation of a charge-transfer complex between styrene, being the electron donor, and maleic anhyride or the maleimide, being the electron acceptor [36–40]. The miscibility of PMI-containing polymers with SAN has been studied by different researchers and is of particular interest to understand the possibilities of blending approaches to increase the heat resistance of ABS. There is a surprisingly high degree of miscibility between styrene-co-Af-phenylmaleimide (SPMI) and SAN copolymers, that increases as the PMI volume fraction increases [41]. Various types of copolymerizations of PMI and styrene are described in the literature [42,43] and various grades of SPMI copolymers are commercially available and used to increase the heat resistance of ABS [44,45]. In addition to blending with SPMI copolymers, PMI can be incorporated into ABS using mass, emulsion [46–50] or suspension [42] free radical polymerization techniques. The high heat ABS resin can be completely produced by mass polymerization, or mass polymerized PMI-SAN can be blended with (emulsion polymerized) SAN-grafted rubber concentrates and/or conventional mass ABS. Sumitomo Naugatuck determined an empirical relation for the compatibility of SAN/SAN-PMI blends based on the polar monomers in each component [51]. Figure 15.4 shows that the miscibility window with SANs becomes wider with increasing PMI level in the terpolymer [52]. A complication with mass polymerization is that at the high temperatures, high concentrations of PMI lead to the generation of oligomeric species, so that
HIGH HEAT RESISTANT ABS TECHNOLOGY
329
% PMI in SAN-PMI Figure 15.4
Miscibility of SAN and SAN–PMI [52]
the boost in heat distortion per unit of PMI decreases. Another complication is that severe composition drift can occur since PMI and styrene tend to form an alternating copolymer with only small amounts of AN incorporated [53]. Owing to this composition drift, a continuous stirred tank reactor (CSTR) configuration is preferred to produce mass PMI–SAN. For PMI-modified complete mass ABS, composition drift can be minimized by adding PMI at different stages in the process. In this way, one can still benefit from the advantages of plug flow reactor technology, such as rubber particles sizing capability, and ease of producing low-gloss materials that are required in the primary automotive interior market. By adding a portion of the total PMI after rubber phase inversion, PMI can be divided between the rubber graft and the continuous phase, thus avoiding incompatibility [54,55]. Improved impact strength and fatigue resistance have been claimed when the total maleimide monomer content and the content of the continuous phase differ by 20 % of the maleimide monomer at a point in the process after phase inversion [56]. Blends containing mass PMI–SAN and emulsion graft rubber concentrates (a mix of mass and emulsion technologies referred to as hybrids) are preferred for high-gloss applications. For automotive applications, where low-gloss esthetics are often required, hybrids containing mass PMI–SAN and both emulsion and bulk-polymerized SAN-grafted rubber result in excellent balances of toughness, heat resistance, and tensile properties [57]. Another advantage for automotive applications of PMI-containing resins produced by the continuous mass-solution process is the interrelated low carbonemission, fogging and odor properties due to the absence of emulsifiers and processing aids [58]. ABS producers with existing emulsion or suspension capital may also produce high heat grades via imide modification. Mitsubishi Monsanto Chemical described methods of preparing high heat ABS by blending 'graft copolymer'
330
R. VANSPEYBROECK ET AL.
PMI-ABS, suspension PMI-SAN, SAN and ABS [59]. They claimed a preferred blend combination of these components which has excellent heat distortion temperatures, superior impact (when mass and emulsion grafted rubber are present), and good stability at processing temperatures. There are a few alternative approaches to imide copolymers that allow the resin producer to make imide-modified high heat ABS without incurring the cost of the synthesized imide monomer. One is by reacting styrene–maleic anhydrides with a primary amine, either during the polymerization reaction with styrene or in a separate step. Mitsubishi Monsanto has practiced imidization on a commercial scale and described a process which follows the formation of S-MA with addition of amine and AN [60]. They described the manufacture of maleimide copolymers by heating the SMA copolymers with aniline in an extruder [61]. The maleimidation of the anhydride function is not complete, as there is unreacted amine or maleic anhydride in the product. The polymer stability and physical properties depend on the mole percent of maleimidation. Another approach is free radical copolymerization of a maleimic acid with styrene. Quantitative ring closure to the imide form was described by Newman of Dow Chemical Company [62,63]. Heat-resistant moldings can be obtained by pellet or melt blending of conventional ABS with maleimide containing master batch thermoplastics or thermoplastics containing colorants, with very high glass transition temperatures. Denki Kagaku Kogyo described master batches of maleimide copolymers and SAN-grafted rubber that has a Ts >140°C [64,65]. Nippon Shokubai Kagaku Kogyo holds a recent patent claiming an extrusion process for SPMI (54:46) copolymer pellets with a Tg of 206 °C, which exhibits no discoloration or thermal deterioration [66]. The excellent thermal stability of PMI-containing high heat ABS resins ensures that the heat resistance of the produced parts is not significantly influenced by thermal exposure applied during processing. Prolonged exposure at high temperatures of aMeS-based resins can result in an increase in the amount of residuals (Figure 15.5), a decrease in polymer molecular weight and a decrease in Vicat softening point of up to 6°C (Figure 15.6).
4
MALEIC ANHYDRIDE
As mentioned in the previous section, maleic anhydride (MA) undergoes copolymerization with styrene, and the five-membered ring anhydride lends rigidity to the resulting polymer chain backbone, resulting in a higher Tg. It has gained commercial utility in several forms: (1) transparent and glass-filled styrene–maleic anhydride copolymers (SMA); (2) filled and unfilled rubber modified SMA (DYLARK11); and (3) blends of SMA with ABS (CADON*).
HIGH HEAT RESISTANT ABS TECHNOLOGY
granules
331
245 degC/3 min. 270 degC/5 min. 295 degC/10 min. Processing Conditions
PMI Mass ABS Figure 15.5
D alphaMeS ABS
Residual monomers of high heat ABS as a function of thermal exposure
granules
245 degC/3 min. 270 degC/5 min. 295 degC/10 min. Processing Conditions
• Figure 15.6
PMI ABS
D alphaMeS ABS
Vicat softening point of high heat ABS as a function of thermal exposure
The predominant commercial synthesis of MA is by vapor-phase oxidation of hydrocarbons, e.g. benzene, n-butane, or a C-4 hydrocarbon mixture, over a solid catalyst [67]. The oxidation of benzene over a supported vanadium oxide catalyst is the preferred procedure. In a typical process, the reactor gas containing low concentrations of MA is passed through a heat exchanger and
332
R. VANSPEYBROECK ETAL
cooled to about 60 °C where most of the crude product is collected as a liquid. The product is sold as briquettes, flakes, or molten monomer in tank cars or trucks. The monomer is a white solid which melts at 53 °C and boils at 202 °C. The MA residue on the polymer backbone is known to promote hydrophilicity and adhesion, improve dyeability, give functionality for crosslinking, promote compatibility with other polymers and fillers, in addition to improving the heat distortion [68]. Early work by Baer at Monsanto showed that a nonhomogeneous resin could be produced very easily in a mass polymerization [69]. Similarly to PMI, the MA monomer tends to form 1:1 alternating copolymers with styrene arising from charge-transfer complexes [68]. Moore and co-workers at Dow Chemical showed that a three stirred tube mass polymerization train could produce a homogeneous rubber-modified SMA copolymer if each reactor was fully recirculated [70]. They employed vigorous recirculation and separate MA feeds to produce resins with about ca 25 wt% MA in the rigid phase. The excellent impact strength attained was indirect evidence of a single Tg, homogeneous rigid phase. Table 15.4 illustrates selected commercialized high heat styrenic resins with MA as the high heat modifier. Rubber-modified SMA resins fulfill specific market needs such as automobile instrument panel substrate materials where low-temperature ductility is not required. However, high tensile and high impact strengths cannot be simultaneously attained. Efforts to terpolymerize styrene, maleic anhydride, and acrylonitrile have revealed crosslinking behavior at high temperatures. Monsanto researchers reported a systematic study where rubber-modified S–MA–AN resins were made via mass polymerization with different AN levels [71]. They defined the degree of crosslinking which occurred during compression molding by testing for solubility in refluxing methyl ethyl ketone (MEK). The data are summarized in Table 15.5. Monsanto claimed that a maximum of 11 wt% AN can be added before severe crosslinking occurs. Melt processing of materials with this tendency to crosslink is challenging. Table 15.4 Selected MA-containing products and properties (all data from manufacturer's published data sheets) Product
Technology
MFR (g/l0 min) Notched Izoda (J/m) Vicar6 ( = C)
DYLARK® 232 SMA-transparent 1.9C DYLARK 8 378 Impact-SMA 1.0r CADON® 152 SMA-ABS 1.0^ a b c d
Notched Izod impact, AST D256. Vicat softening point, ASTM D1525. Melt How rate (230°C/2.16kg), ASTM D1238. Melt How rate (230 °C/3.8 kg), ASTM D1238.
10.7 154.8 133.5
118 127 126
HIGH HEAT RESISTANT ABS TECHNOLOGY
333
Table 15.5 Solubility properties of SMA–AN impact modified terpolymers after compression molding S:MA:AN
Solubility in MEK after devolatilization
After compression molding
68:26:6 68:23:11 63:23:14 56:26:19
Dispersible Dispersible Not Dispersible Not Dispersible
Dispersible Dispersible Not Dispersible Not Dispersible
To design a resin with the property enhancements of AN without the crosslinking problem, it was found that SMA copolymers and terpolymers could be blended with ABS resins to form miscible blends with properties of HHABS. A fundamental look at the miscibility of SMA copolymers with SAN copolymers indicated that the optimum thermodynamic interaction occurs when the AN content of the SAN is nearly equal to the MA content of the SMA [72]. Kim et al. also found low impact strengths at all modifier levels when blending SMA with SAN-g-polybutadiene (GRC = grafted rubber concentrate) [73]. Blends of SMA with SAN and GRC (SAN + GRC = emulsion ABS) exhibited ductility behavior similar to HHABS. The impact strengths of the polymers were 2-5 ftIb/in, in a notched Izod test at ambient temperature. Dow and Monsanto, among others, have investigated the manufacture of SMA resins both with and without rubber modification. Moore at Dow Chemical Company described a method of producing SMA copolymers via a recirculated coil reactor [74]. In general, SMA copolymers, impact-modified SMA copolymers, and glassfilled SMA copolymers have good competitiveness and reasonable processibility in applications which require heat properties greater than general-purpose HIPS and ABS. They provide a low-cost solution where low-temperature ductility is not required. The more ductile SMA–ABS blends have had limited success owing to their poorer flow and tendency to crosslink and decompose at higher temperatures.
5
MODIFIED NITRILES
Heat-resistant ABS resins can also be produced by polymerizing styrene with modified nitriles, such as fumaronitrile and maleonitrile. Fumaronitrile can be produced from acrylonitrile (Figure 15.7) in a two-step process involving the addition of hydrogen cyanide followed by oxydehydrogenation over metal oxide catalysts. The compatibility of styrene-co-acrylonitrile-co-fumaronitrile (SANF) terpolymers and SAN has been studied [75]. High-gloss, heat-resistant
334
R. VANSPEYBROECK ETAL. NC H7C
\
HCN
CN
Figure 15.7
maleonitrile
Synthesis of fumaronitrile
ABS technology has been developed whereby SANF terpolymers produced in a solution process were melt blended with ABS impact modifiers [76]. However, there are no major applications for the monomer and it is commercially available only in small quantities at high cost.
6
VARIOUS HIGH HEAT-RESISTANT ABS GRADES
In the positioning of ABS products against generic competition, the balance of several key properties for the cost is the key. ABS is sold primarily for its toughness, appearance, heat resistance, and flow. The flow relates to processability, particularly for injection molded applications. Essentially, the flow is balanced with the other properties (excluding appearance). In the case of injection molding one would want the highest toughness and highest flow for the best resin. Typically, increasing rubber increases the toughness, but the flow is decreased with the rubber phase being the highest viscosity. The reality is that toughness, flow, and heat are intertwined properties. One is usually maximized at the expense of others. Table 15.6 lists a series of ABS grades, from general-purpose molding grade to high heat grades in which the heat resistance has been increased over conventional ABS. The Vicat softening points have been increased by the incorporation of various levels and types of heat-boosting monomers, and range from 98 to 112oC. Heat-resistant resins can be divided into several classes of heat resistance: 100–103, 104–106, 107–110, and 111 + oC Vicat softening point. The data in Table 15.6 and in Figure 15.8 show that the toughness/melt flow balance decreases with increasing Vicat softening point. Within the 100–112CC range, mass ABS and mass/emulsion blended resins containing PMI exhibit an enhanced property balance when compared with emulsion ABS resins based on aMeS. An alternative route to obtain the equivalent heat/toughness/flow balance seen with PMI-containing ABS is to melt blend polycarbonate into the ABS resin. The blends contain less than 40 wt% PC and the ABS remains the continuous phase. Also included in Table 15.6 for comparison is a PC–ABS blend where PC is the continuous phase and the ABS is less than 40 wt% of the blend (PULSE8 2000EZ).
Table 15.6 Selected HHABS and PG–ABS resins (all data from manufacturer's published data sheets unless noted)a Heat range General-purpose
Grade
CYCOLAC® GPM5500 RONFOLIN® RT-51 RONFOLIN® GG-20 Medium heat MAGNUM 3325 MT CYCOLAC® BDT5510 CYCOLAC®X11 SINKRAL® C442 POLYLAC® PA-777B High heat RONFOLIN® HH-20 XZ-96502.00 PC modified NOVODUR® KU2-5300 RONFOLIN® HX-05 RONFOLIN® HX-03 CYCOLAC® G365 Highest heat CYCOLAC® Z48 RONFOLIN® HX-10 MAGNUM® 3416 SC MAGNUM® 358 HP MAGNUM® 357 HP Super high heat CYCOLAC® X17 PC–ABS(>60%PC) PULSE® 2000EZ
Supplier
Technology
General emulsion GE Plastics BASF Corporation aMS, emulsion aMS, emulsion BASF Corporation Low residuals, mass Dow Chemical Co. GE Plastics aMS, mass aMS, emulsion GE Plastics Low residuals, mass Enichem Chi Mei Industrial Co., Ltd aMS, emulsion aMS, emulsion BASF Corporation PMI, low residuals, mass Dow Chemical Co. Bayer Corporation Emulsion ABS/PC modified BASF Corporation aMS, emulsion aMS, emulsion BASF Corporation aMS, emulsion GE Plastics aMS, emulsion GE Plastics BASF Corporation aMS, emulsion PMI, low residuals, mass Dow Chemical Co. PMI, mass Dow Chemical Co. PMI, mass/emulsion hybrid Dow Chemical Co. aMS, emulsion GE Plastics PC/mass ABS blend Dow Chemical Co.
Charpy Vicat MFR (gVlOmin) (kj/m 2 ) (°C)
24.0 17.0 10.0 10 15.0 6 6.0 6.5 5.0 6.0 5.0 3.5 7.0 6.0 3.0 3.5 6.5 6.5 6.0 5.0 I5h
16 20 25 18 13 28 18 20 20 25 32 17 9 11 14 12 18 18 16 10 40
98 101 101 101 100 103 104 105 106 106 106 107 108 108 109 110 108 108 109 112 124
a
MFR: ISO 1133; 220 °C, 10kg. Charpy, notched @ 23 °C: ISO 179. Vicat: ISO 306; 50N/50°C/h The MFR for PULSE® 2000EZ is estimated for ISO 1133; 220°C, 10kg from the measured MFR of 7.00g3/10min value at 230°C/3.8kg. h
w
CO Ol
336
R. VANSPEYBROECK ETAL 46
PC/ABS
PC > 60 wt % 42
PC/ABS PC < 30 wt %
38
i| 34 3 30
§:
General Purpose ABS
5 26
! n o 18
O A
O
\
A
14 10
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
18.0 20.0
22.0 24.0
26.0
MeltFlow Rate (g/10') x 100 - 103 deg C o 104-106 deg C
• 106 degC ; PC modified ABS A 107 - 110 deg C
» 106degC;PMI-ABS
o 112degC
--•Linear (107- 110 deg C)
• PC/ABS General Purpose ABS
Linear (100- 103 deg C)
Figure 15.8 Scatter plot of the balance of impact strength and melt flow for ABS and PC-ABS resins with different heat resistance (data from Table 15.6)
There are two linear trend lines plotted in Figure 15.8 for the medium heat range (100 – 103 °C) and the highest heat range (107 – 110°C). The trends indicate that as the heat is increased from the medium heat range into the highest heat range, the toughness flow balance of the highest heat resins becomes significantly lower (dotted line) than the toughness flow balance of the medium heat range of resins (solid line). The second major point to note in Figure 15.8 is that within a heat series, the slope of the trend lines is negative. This clearly indicates that as flow is increased, it is at the expense of toughness. One can also note from Table 15.6 that several resins have increased heat due to low residuals and high AN contents (not shown). An illustrative way to visualize the flow/toughness/heat balance is to view the key resin properties in a bubble plot as shown in Figure 15.9. It can easily be
337
HIGH HEAT RESISTANT ABS TECHNOLOGY
OMFR
i
0 10 20 30
k. 01
1 : £1e» _:
I
PC/A
f I
i a*
I
.
.
>M
4>-o
i
i
-'0
!
i
i
'i
"o
z :
! i
O
0
6 J... n-
o '; °
O«MeS
\.
A High AN Low ReisduaL
i
30
20 15
:
i
•
35
25
j
^ ' &
&« : d " 1 : _o -
IPC/AlU —^^^
! o^— ^ j i
O
i /"A iGPABSl
|
;
\
i
i
10 5 n
95
100
105 110 115 VICAT Heat Distortion (degrees Celsius)
120
125
Figure 15.9 Bubble plot of melt flow rate (bubble) versus Vicat heat distortion and notched Charpy
seen that the high heat ABS decreases in flow and toughness as the Vicat heat resistance is increased. Also, as one changes from a high heat ABS to a PCABS, where the ABS becomes the dispersed or co-continuous phase, a dramatic change in the flow/toughness/heat property balance occurs. The improved property balance is also accompanied by an increase in cost. In Figure 15.9, the open circles with solid lines are for high heat ABS made with aMeS, the open circles with horizontal parallel lines represent high heat ABS made with PMI, and the gray solid circles represent high heat ABS made with low residuals and high AN content with no high heat monomer. The dotted black circle is general-purpose grade ABS and the two bold black circles represent PC–ABS blends. The diameter of the bubble represents the flow of the resin as indicated in Table 15.6. As a final note, the Vicat heat distortion, as noted earlier, is an indicator of heat resistance. The actual heat performance of the part is influenced by the molding process conditions and the flow rheology of the resin itself (molded-in stress can noticeably affect the heat performance of an actual part. Thus the heat ranking as noted on the plot can be different to that in the final part owing to actual differences in rheology, which will then impart different levels of molded-in stress.
338
R. VANSPEYBROECK ET AL.
REFERENCES 1. Adam M. E., Buckley D. J., Colborn R. E., England W. P., Schissel D. N., Rapra Rev. Rep., 6, No. 10 (1993). 2. Mark H. F., et al. (Eds), Encyclopedia of Polymer Science and Engineering, Vol. 16, Wiley, New York, pp. 1–246 3. Product Bulletin SR-606E, Borg Warner Chemicals Inc., Parkers burg, WV. 4. MAGNUM ABS Resins, Product Bulletin 301-899-686, The Dow Chemical Company, Midland, MI. 5. Product Bulletin 301-665-687, The Dow Chemical Company, Midland, MI. 6. Product Bulletin 6364C, Monsanto Co., St. Louis, MO. 7. Maul J., Meyer H. H., Makromol. Chem., Macromol Symp. 53, 23 (1992). 8. Henton D., Dion R. P., Lefevre N. A., US Patent 4972032 (to The Dow Chemical Company) (1990). 9. Minematsu H., Matsumoto T., Saeki T., Kishi A., US Patent 4294946 (to Sumitomo Naugatuck Co.) (1980). 10. Abe M., Iwama M., Tsuchikawa S., Morikawa T., US Patent 4306043 (to Japan Synthetic Rubber Co.) (1981). 11. Mathumoto S., Jagoshi F., US Patent 4526 928 (to Kanegafuchi Kagaku Kogyo Kabushiki) (1985). 12. Grabowski T., GB Patent 1253 226 (to Borg-Warner Coporation) (1971). 13. Schwier C. E., Wu W. C, US Patent 4874829 (to Monsanto Co.) (1989). 14. Rinehart M., US Patent 4169195 (to Borg-Warner Corporation) (1979). 15. Sakano Y., Miyaki N., Japanese Patent JP 02051515 (to Denki Kagaku Kohyo KK) (1990). 16. Rudin A., Chiang S., J. Polym. Sci., 12, 2335 (1974). 17. Driscoll K., Dickson J., J. Macromol. Sci. Chem., A2, 49 (1968). 18. Cowle J. M. G., Elexpuru E. M., McEwen I. J., Polymer, 33, 1993 (1992). 19. Priddy D. B., Traugott T. D., Seiss R. H., J. Appl. Polym. Sci., 41 383 (1990). 20. Mastsumoto A., Oki Y., Otsu T., Polym. J., 23, 201 (1991). 21. Aida H., Kimura M., Fukuoka A., Hirobe T., Kobunshi Kagaku, 28, 354 (1971). 22. Barrales-Rienda J. M., Gonzalez de la Campa J. I., Gonzales Ramos J., J. Macromol. Sci. Chem., All, 267 (1977). 23. Kita Y., Kishino K., Nakagawa K., J. Appl. Polym. Sci., 63, 1055 (1997). 24. Sato H., Jpn. Kokai, 87 156115 (1987); Sato H., Matsuo M., Jpn. Kokai, 89 62315 (1989). 25. Maekawa M., Adachi M., Yasuda Y., Japanese Patent JP 06116338 (to Toray Indsutries) (1994). 26. Robinson J. C., Ziegelmeyer T. A., European Patent Application 0415120A2 (to General Electric Company) (1991). 27. Daihachi Chemical Industry, Plas. Ind. News (Jpn.) 30 (11) 163 (1984). 28. Kita Y., Kentaro S., Maseo B., Atsushi O., US Patent 4623734 (to Nippon Shokubai KK) (1986). 29. Iwatsuki S., Kubo M., Wakita M., Matsui Y., Kanoh H., Macromolecules, 24, 5009 (1991). 30. Matsumoto A., Kubota T., Otsu T., Macromolecules, 23, 4508 (1990). 31. Young L. J., in Polymer Handbook, Brandrup J., Immergut E. H. (Eds) Wiley– Interscience, New York, 2nd edn, pp. 11/105–386 (1975). 32. Coleman L. E., Jr, Conrady J. A., J. Polym. Sci., 38, 241 (1959). 33. Van Paesschen G., Timmerman D., Makromol. Chem., 78, 112 (1964). 34. Yamaguchi H., Minoura Y., J. Polym. Sci., A-l, 8, 1467 (1970).
HIGH HEAT RESISTANT ABS TECHNOLOGY 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70.
339
Yamada M., Takase I., Kobunshi Kagaku, 23, 348 (1966). Furukawa J., /. Polym. ScL, Polym. Symp. Ed., 51, 105 (1975). Kokubo T., Iwatsuki S., Yamashita Y., Macromolecules, 1, 482 (1968). Shirota Y., Tomono T., Makromol. Chem., 141, 265 (1971). Seiner J. A., Litt M., Macromolecules, 4, 308 (1971). Abayasekara D. R., Ottenbrite R. M., ACS Polym. Prepr., 27, 462 (1986). Aoki, Y. Macrocomolecules, 21, 1277 (1988). He, J., Wang J., Li S., Gongneng Gaofenzi Uuebo, 12 (1), 19 (1999). Dedovets G. S., Kondratovich A. A., Ivanov V. S., Vestn. Leningr. Univ., Fiz. Khim. (4), 132 (1977). Yatagai H., Plast. Compounding, 45 (1993). Denki Kagaku Kogyo, Mod. Plast. Int., 26 (4), 93 (1996). Toyooka Y., Fujii S., Japanes Patent 03097703 (to Mitsubishi Rayon Co.) (1991). Maeda Y., Miyazaki H., Japanese Patent 03143910 (to Monsanto Chemicals) (1990). Byrdina N. A., et al., Plast. Massy, 10, 36 (1990). Jang B., Jung H., Park H., Korean Patent 9605 08 (to Cheil Industries Inc.) (1992). Maeda Y., Myazaki H. (Monsanto Chemicals, Japan), Japan Kokai Tokyo Koho. Masatsune K., Ogura S., Koiti K., US Patent 4877 833 (to Sumitomo Naugatuck Co.) (1989). Ogura S., Mastsune K., Katsuji U., 45th ANTEC Conference Proceedings, pp. 1365 (1987) Florjanczyk Z., Krawiec W., Makromol. Chem., 190, 2141 (1989). Iwamoto I., Ito N., Sugazaki K., Matsubara T., Ando T., US Patent 4808 661 (to Mitsui Toastsu Chemicals) (1989). Iwamoto I., Ito N., Sugazaki K., Matsubara T., Ando T., US Patent 4954517 (to Mitsui Toatsu Chemicals) (1990). Traugott, T. D., Workentine, S. L., US Patent 5412036 (to Dow Chemical Co.) (1995). Shields N., Van de Langkruis G., US Patent 5 270 387 (to Dow Chemical Co.) (1993). Reunis A., Hogg A., Naughton P. Schoppmann T.D. Nickel R., Markhardt A. SAE Technical Paper 1999-01-0853, International Congress and Exposition, Detroit, MI (March 4, 1999). Aoki Y., Hiroaki M., US Patent 4879343 (to Mitsubishi Monsanto Chemical Company) (1989). Ikuma S., US Patent 4381373 (to Mitsubishi Monsanto Chemical Company) (1983). Oshida T., Kajiwara T., Japanese Patent 02004806 (to Mitsubishi Monsanto Chemical Company) (1990). Newman T., US Patent 5015 712 (to The Dow Chemical Company) (1991). Newman T., US Patent 5077343 (to The Dow Chemical Company) (1991). Noguchi A., Shimura T., Miyashita S., Japanese Patent 10036614 (to Denki Kagku Kogyo) (1998). Shinmura, T., Konishi, K., US Patent 5532317 (to Denki Kagaku Kogyo) (1996). Kinoschit F., Yoshikawa K., Fujioka K., Japanese Patent 2000271928 (to Nippon Shokubai Kagaku Kogyo) (2000). Sharma R., Creswell P., Newson P., Am. Inst. Chem. Eng. Symp. Ser., 87 (1984). Trivedi B. C, Culbertson B. M., Maleic Anhydride, Plenum Press, New York (1982). Baer M., US Patent 2971939 (to Monsanto Chemical Company) (1961). Moore E., Lehrer C., Lyons C., McKeever L., US Patent 3919354 (to The Dow Chemical Company) (1975).
340 71. 72. 73. 74. 75. 76.
R. VANSPEYBROECK ET AL Lee Y, Trementozzi Q., US Patent 4262096 (to Monsanto Chemicals) (1981). Kim J., Barlow J., Paul D., J. Polym. Sci., Polym. Phys. Ed., 27, 223 (1989). Kim J., Kesskula H, Paul D., J. Appl. Polym. Sci., 40, 183 (1990). Moore E., Ind. Eng Chem. Prod. Res. Dev., 25, 315 (1986). Warakomski J. M., /. Appl. Polym. Sci., 46, 1057 (1992). Dion R., Warakomski J. M., US Patent 5 212240 (to The Dow Chemical Company) (1993).
16
Synthesis, Properties and Applications of AcrylonitrileStyrene-Acrylate Polymers G. E. McKEE, A. KISTENMACHER, H. GOERRISSEN, M. BREULMANN BASF AG, Ludwigshafen, Germany
1
INTRODUCTION
Acrylonitrile-styrene-acrylate (ASA) constitutes a versatile member of the group of styrenic copolymers used for housings, covers and other applications which require excellent surface appearance and environmental stability combined with high impact resistance and stiffness. It consists of a poly (styrene– acrylonitrile) matrix modified with small rubber particles. From its architecture, ASA is closely related to ABS; however, instead of polybutadiene rubber particles grafted with poly(styrene–acrylonitrile) (PS AN), poly(alkyl acrylate)-based graft rubber particles are used as the impact modifier (Figure 16.1). Generally, the poly(alkyl acrylate) core is slightly crosslinked to exhibit the needed elastomeric properties. Grafted chains, tethered to the acrylate rubber surface, serve as chemically bonded compatibilizers between the rubber particles and the PSAN matrix (Figure 16.2). These grafted chains usually consist of a PSAN, and optionally additional monomers such as methyl methacrylate are added. The density of the grafting points, the average chain length of the graft polymers, the graft comonomer composition and the relative amount of graft shell to poly(butyl acrylate) core play a significant role in the ASA mechanical and processing properties. In comparison with ABS, where the double bonds of polybutadiene are prone to oxidation and crosslinking due to oxygen, ultraviolet (UV) radiation Modern Styrenic Polymers: Polystyrene and Stvrenic Copolymers. Edited by J. Scheirs and D. B. Priddy f- 2003 John Wiley & Sons Ltd
342
Figure 16.1
G. E. McKEE
Architecture of acrylonitrile–styrene-acrylate
9 w
Styrene Acrylonitrile Rubber
Figure 16.2
Grafted PSAN chains tethered on a rubber surface
or heat, ASA is free of double bonds. Therefore, ASA, while having similar basic properties to ABS, has significant advantages in terms of UV stability and long-term heat resistance. Furthermore, the chemical resistance of ASA is significantly improved in comparison with ABS.
2
ASA MARKET
Only a few manufacturers of ABS also produce ASA. The reason for this is that in spite of the obvious similarities, there are also significant differences in the
ETAL..
SYNTHESIS, PROPERTIES AND APPLICATIONS OF ASA
343
production processes, and the know-how needed to produce a competitive ASA exists in only a few companies. Thus, within the different global regions there are only a few major players with significant ASA capacities, e.g. BASF, General Electric, Bayer, Hitachi and LG Chemicals. Since most of the manufacturers of ASA make use of multipurpose production plants that are used for both ABS and ASA production, the ASA output can be adapted to demand. It is estimated, that the annual demand for ASA is in the range 1-5 % of the annual ABS demand (ABS demand 2001: ca4.5 x 106 /t [1]).
3
PRODUCTION OF ASA
3.1
EARLY DEVELOPMENTS
The first attempts to produce an impact-modified PSAN using an acrylatebased rubber date back to the early 1960s when Herbig and Salyer [2] of Monsanto used a butyl acrylate–acrylonitrile rubber with a high acrylate content and Otto [3] of BASF used a rubber composed of an acrylate copolymerized with a crosslinking agent as the impact modifier. Further work at BASF by Siebel and Otto [4] polymerized butyl acrylate or ethylhexyl acrylate with butadiene and a vinyl alkyl ether in emulsion as the base rubber, which was then grafted with styrene and acrylonitile and used as an impact modifier for the brittle but stiff poly(styrene–acrylonitrile). This was followed by investigations of Willersinn et al. [5] of the same company in which the impact modifier was produced by polymerizing butyl acrylate and a crosslinking agent using free radicals in emulsion. The resulting polymer dispersion was grafted with styrene and acrylonitrile and after its isolation blended with PSAN. A similar patent by van der Werth and Lederer of BP Chemicals followed in 1968 [6]. This emulsion polymerization process is the basic process used today for the commercial production of ASA.
3.2
EMULSION POLYMERIZATION PROCESS
The production process for ASA is shown schematically in Figure 16.3 and is briefly discussed below.
3.2.1
Base Rubber Production
The base rubber is produced by polymerization of an acrylate in emulsion. A crosslinker can be employed, e.g. a diol diacrylate such as butanediol diacry late, divinylbenzene allyl(meth)acrylate, dioldiallylcarbonate, acrylic esters of
344
G. E. McKEE ETAL. base poly(butyl acrylate) rubber grafting with styrene and acrylonitrile polymerisation of the graft rubber in emulsion storage tanks
isolation of rubber extruder ASA
Figure 16.3
Preparation of ASA
tricyclodecenyl alcohol, diallyl maleate, diallyl fumarate and diallyl phthalate [5,7–11]. In most of the commercial products butyl acrylate is employed; however, in the majority of the ASA patent literature, other acrylates, especially ethylhexyl acrylate, are mentioned. The advantage of the latter over poly(butyl acrylate) is its lower glass transition temperature (Tg), -65°C, compared with poly(butyl acrylate) with a value of —45°C. This lower T% opens up the way for better lowtemperature properties for the ASA. The low-temperature properties of ASA are touched on briefly in Section 5. The mean weight particle size is in the usual range for the emulsion polymerization and is less than 1 u,m.
3.2.2
Grafting of Base Rubber
The base rubber is then grafted with styrene and acrylonitrile. The purpose of the PSAN graft shell is to anchor the rubber particles in the PSAN matrix and also to ensure their good dispersion in the PSAN matrix. The graft shell is usually not crosslinked; however, this is not always the case [9].
3.2.3
Isolation of the Grafted Rubber
The grafted rubber is usually isolated by either spray drying or by coagulation followed by drying. The blending of the coagulated rubber containing residual water with the PSAN melt in an extruder is also possible [12].
SYNTHESIS, PROPERTIES AND APPLICATIONS OF ASA
3.2.4
345
Preparation of the PSAN
The PSAN matrix can be prepared in suspension, solution or emulsion. The first two are the most commonly used methods. After the PSAN preparation, the unreacted monomers, water and if necessary solvent are removed.
3.2.5
Blending of PSAN and the Impact Modifier to give the End Product
Blending of the PSAN and the impact modifier is usually carried out in an extruder at a temperature range of 220-300 °C to give the end sales product. During this blending step, additives such as lubricants, pigments and antioxidants can also be added. The morphology of an ASA product with a mean weight particle size of approximately 0.1 jxm is shown in Figure 16.4. As can be seen, the rubber particles have agglomerated, which according to Ramsteiner [13] leads to better mechanics.
3.3
BULK POLYMERIZATION PROCESS
The preparation of ASA in bulk or bulk-suspension polymerization processes has been described by McKee et al. [14–18]. The system is similar to that used in the preparation of high-impact polystyrene (HIPS) and in bulk-produced ABS. Thereby the rubber was prepared using free radical polymerization, dissolved in the SAN monomers which were then polymerized using free radicals. Phase separation between the rubber and PSAN occurred, followed by phase inversion. In the preparation of HIPS and ABS prepared in bulk, the polybutadiene rubber is easily grafted at the pendant 1,2-C—C double bond. In the case of
Figure 16.4
TEM of an ASA containing an emulsion-produced impact modifier
346
G. E. McKEEE7V\L.
poly(butyl acrylate) the rubber is not easily grafted; however, grafting of the SAN on the poly (butyl acrylate) rubber can be achieved by: • Copolymerizing the acrylate monomer with a monomer containing two or more polymerizable double bonds. Thereby some of the double bonds should survive the polymerization stage of the butyl acrylate and serve as grafting points in the SAN polymerization [14]. • Copolymerizing the butyl acrylate with a monomer containing a peroxide group. The peroxide group is stable at the polymerization temperature of the butyl acrylate. After the rubber formation the temperature is increased to decompose the peroxide side group in poly(butyl acrylate), leading in the presence of SAN monomers to a grafting reaction [15]. Alternatively, if higher temperatures are not desired, the peroxide side group can be decomposed using a redox system [16]. • Copolymerizing the butyl acrylate with a SAN macromonomer [17]. An elegant variation of this process is the anionic polymerization of ethylhexyl acrylate or butyl acrylate using, e.g., styrene as the solvent [18,19]. With the correct choice of catalyst only acrylate double bonds are polymerized and not the styrene. Grafting points were introduced via Copolymerizing the ethylhexyl acrylate with a monomer containing two double bonds. The first double bond is chosen to be an acrylate and is thus copolymerized into the rubber backbone. The second double bond, which is not an acrylate, remains intact and can be used as a grafting point in the free radical polymerization step of the SAN monomers. The morphology (Figure 16.5) [20] is different to that obtained using the emulsion process (Figure 16.4). In the case of the emulsion-prepared product the particles are smaller than those produced in bulk and they contain less inclusions (salami structure) than the bulk-prepared product.
Figure 16.5
JEM of ASA produced in bulk
SYNTHESIS, PROPERTIES AND APPLICATIONS OF ASA
3.4
347
MICROSUSPENSION POLYMERIZATION PROCESS
A further method for producing ASA uses rubber particles prepared in a microsuspension process [21-24] (for morphology, see Figure 16.6). Thereby the monomer-water mixture containing a protective colloid is intensively sheared to produce monomer droplets which are then polymerized to give polymer particles of approximately the same size. These particles are then grafted with styrene and acrylonitrile and used as the impact modifier for the ASA. The particles may also be used in polymer formulations as delustering agents [22]. Owing to their large size they lead to an uneven surface of the moulded article (Figure 16.7) which leads to light scattering and thus matt surfaces. Brandstetter et al. [25] produced similar particles without the intensive shearing step by polymerizing acrylate monomer in water in the presence of a long-chain alcohol, an emulsifier and a water-insoluble initiator. The particles have a mean particle size of 0.2–6 u>m. After their preparation the particles can be grafted with SAN and then used as impact modifiers for ASA or as delustering agents for thermoplastics [26].
Figure 16.6 ASA containing a rubber produced using the microsuspension process
Figure 16.7 Scanning electron micrograph of ASA containing an ASA impact modifier prepared using the microsuspension process. These large particles give delustered surfaces
348
4 4.1
G. E. McKEE ET AL.
PROPERTIES OF ASA AGEING PROPERTIES
At first glance, ASA possesses a similar chemical structure to ABS, since both consist of a SAN matrix containing a graft rubber. However, while the core of the graft rubber of ABS consists of polybutadiene, that of ASA consists of poly(n-butyl acrylate) (Figure 16.8), and this accounts for important differences in the properties of the two plastics. In ABS, the double bonds of polybutadiene are prone to oxidation and crosslinking due to oxygen, UV radiation or heat [27,28] (Figure 16.9). The result is deterioration of the rubber, leading to loss of impact strength and discoloration. In contrast, the butyl acrylate rubber of ASA is free of C—C double bonds which gives ASA clear advantages in terms of weatherability [29,30] (Figures 16.10 and 16.11) and resistance against heat ageing (Figure 16.12). ABS rubber: Polybutadiene
n
CH2
= CH - CH = CH2
>
... - C H 2 - C H = CH-CH 2 -CH 2 - CH = C H - C H 2 - ... (1,4 Addition) or ... -CH 2 -CH-CH 2 -CH-... I I CH CH II II CH2 CH2
(1,2 Addition)
ASA rubber: Poly(n-butyl acrylate)
n
Figure 16.8
CH2 = CH I O = C - O - C4 H9
>
...-CH2-CH-... I O = C - O - C4 H9
Chemical structure of the rubber components of ABS and ASA
... - CH = CH - CH2 - CH2 - ...
02
+
... - CH - CH - CH2 - ... I
oo«
». ... - CH - CH = CH - ... —*• I OOH Figure 16.9
Alcohols, diols, ketones, esters or chain cleavage, crosslinking
Attack by oxygen and hydroperoxide formation in polybutadiene
349
SYNTHESIS, PROPERTIES AND APPLICATIONS OF ASA
Figure 16.10 Scanning electron micrographs of the surface of ABS (UV stabilized) and ASA after 500 h of xenon arc weathering according to ISO 4892-2A (specimens struck on exposed side)
0 Figure 16.11 Florida
1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 Hours of sunshine Change of multiaxial impact strength during outdoor weathering in
20
40 Time (weeks)
60
100
Figure 16.12 Multiaxial impact strength (ISO 6603–2) of ASA and ABS after heat ageing at 90 °C
G. E. McKEE ET AL.
350
A further advantage of ASA vs ABS is its higher resistance against environmental stress cracking, especially against alcohols and many cleaning agents [31] (Figure 16.13). ASA also exhibits advantages over other thermoplastic housing materials such as polycarbonate, PBT and polypropylene, as shown in Table 16.1 The low moulding shrinkage of ASA and of PC is advantageous for housings and covers, because warpage problems are almost nonexistent with these products. However, PC has only a limited resistance against environmental stress cracking, for example by alcohols and cleaning agents, and it yellows much more than ASA during outdoor exposure. Compared with polypropylene, a material widely used because of its low price, ASA has advantages in terms of stiffness, impact strength, heat distortion temperature and weatherability. As an example of the last aspect, the change in gloss during outdoor weathering of UV-stabilized, white-coloured ASA and polypropylene is shown in Figure 16.14. Further, owing to its high moulding shrinkage, polypropylene has a high tendency to warp during or after processing, and it is much more sensitive to scratches than ASA. Thus, because of its superior physical properties, and because it can be easily processed by injection moulding or extrusion, ASA has widespread use and can be found in a large variety of applications. 60 ASA
50 40 30 20
X
ABS
10
100
200
300
400
Ball oversize(fim)
Figure 16.13 Resistance against environmental stress cracking by 2-propanol (ball indentation test ISO 4600, 1 h exposure)
351
SYNTHESIS, PROPERTIES AND APPLICATIONS OF ASA Table 16.1 door use
Comparison of unreinforced thermoplastic housing materials for out-
Property
Units
ASA PC
Moulding shrinkage (typical) Tensile modulus Tensile strength Notched Charpy impact strength HDT/A Chemical resistance/ environmental stress cracking resistance
% MPa MPa kJ/m 2 °C
0.5 2300 48 15 97 High
0.5 2400 63 100 129 Low
PBT
PP homopolymer
1.5 2500 60 6 65 Very high
1.5 1500 35 3 60 Very high
Polypropylene UV-stabilised. white
9
12
15
18
20
24
Time (months) Figure 16.14 Change of gloss of UV-stabilized, white-coloured ASA and polypropylene during outdoor weathering in Germany
4.2
IMPACT BEHAVIOUR
Ramsteiner et al. [32] investigated the rubber toughening of PSAN. In ASAcontaining particles with a diameter of ca 0.5 |xm and at a graft rubber concentration of 50 wt %, the distance between the rubber particles is small enough for
352
G. E. McKEE ET AL.
energy dissipation via stretching and shearing to occur. With large particles of > 1 jxm and low rubber concentrations such as occur in bulk-prepared ABS, the shearing process at large inter-particle distances cannot occur; however, the large particles are able to dissipate the energy via crazes. If the rubber particles are smaller and in the range of 80 nm the impact strength is low; however, if they agglomerate during processing then a large increase in impact strength occurs [13,32]. To compensate for the low toughness in ASA when using small rubber particles, large particles prepared using the microsuspension polymerization process can be added to the products (700 nm to 100|xm) [22]. Further work showed that the use of particles with a diameter of 0.15–0.8 fim brings better toughness than a particle size of less than 0.1 (xm [33].
5
ADDITIONAL AREAS OF INVESTIGATION
Additional areas of investigation include: • To reduce the surface gloss of articles made from ASA, e.g. by copolymerizing additional monomers into the base rubber or the graft shell or by incorporating large particles into the ASA formulations [22,23,25,34–55]. • To develop flame retarded grades [56–58]. A lot of work has been carried out using ASA—polycarbonate blends. However, this field will not be covered here. • To reduce the glass transition temperature of the rubber to improve the lowtemperature properties of the ASA products. Especially preferred is ethylhexyl acrylate [59,60]. A further method is the incorporation of silicone rubber into the ASA particles [55,61–65].
6
ASA BLENDS
Blends of ASA with many of the common thermoplastic materials are state of the art. The properties of the finished products depend to a large extent on the polymer compatibility often induced by the use of reactive polymers. Only a few of these blends have reached a significant commercial status. The largest blend products in this area are ASA-PC blends followed by ASA—PBT and ASA-PC-PMMA blends. Combinations of ASA with high-Tg matrix polymers are also frequently found in commercial products. Although polycarbonate is an engineering thermoplastic material which provides high toughness, flexibility and thermal stability, it suffers from certain limitations due to poor chemical resistance and low flow characteristics in injection moulding. These shortcomings can be circumvented by blending PC
353
SYNTHESIS, PROPERTIES AND APPLICATIONS OF ASA
with ABS or ASA. Generally, the impact strength and heat deflection temperature increase with increasing PC content (Figure 16.15). ASA as a blend component provides improved weatherability, which is a prerequisite for outdoor applications. In blends of polycarbonate with an ASA (PSAN containing 25 wt% of grafted rubber) extruded at 270 °C [66], the notched Izod impact strength surprisingly showed a maximum at 20 wt% ASA. This striking result is probably due to the critical thickness of the test specimen employed. It is known that for polycarbonate above a critical thickness of the test specimen the notched impact strength deteriorates steeply. This unfavourable critical thickness effect can be eliminated by adding amounts of >5% of ASA to PC [67]. The improved flowability of ASA-PC blends was addressed by Yu [68], by suggesting the use of a three-stage impact modifier, consisting of an acrylate core that is covered by a first shell of crosslinked PSAN and a further shell comprised of linear PSAN. These blends, while retaining the impact properties of the material, exhibit an improved flowability by a factor of 3. Another option in improving the flowability of these blends is the use of a branched polycarbonate. Brandstetter et al. [69] found that a dianhydride branched polycarbonate combined with ASA can be injection moulded at temperatures 5–10 K lower than similar blends based on linear polycarbonates. Increased toughness and colourability over conventional ASA-PC blends can be attained by using a two-stage grafted n-butyl acrylate core [70]. In the first grafting step pure styrene is added followed by a styrene—acrylonitrile Vicat softening temperature,
Charpy notched impact strength, [kJ/m2] —*60 -,
30 70 wt-% PC in ASA/PC blend
100
Figure 16.15 Charpy notched impact strength and Vicat temperature versus PC content in ASA-PC blends
354
G. E. McKEE ETAL.
comonomer composition. Core-shell grafted ASA rubbers containing a styrene core are used in ASA-PC blends to yield products with high gloss and excellent colourability [71]. Ternary blends of ASA—PMMA—PC exhibit improved toughness and flexural strength in comparison with ASA—PMMA blends [72]. An optimum in tensile modulus was found for blends containing >50% ASA. To overcome the critical thickness impact deficiencies of polycarbonate blends, a two-phase crosslinked acrylate—crosslinked styrene—acrylonitrile modifier has been suggested [73]. The same two-phase acrylate—crosslinked styrene—acrylonitrile product is used for significantly increasing the notched impact strength of polycarbonate-poly(l,4-butylene terephthalate) blends containing 18– 30 wt% of PC [74]. Inferior processablility and multiaxial impact strength of conventional ASA—polycarbonate blends can be improved by a three-stage grafted acrylate—rubber of average particle size (d50 weight average) of 200– 700 nm, with a first grafting stage comprised of styrene, a second stage of styrene—acrylonitrile and a third stage of methyl methacrylate [75]. A further improvement of impact strength and colourability can be obtained by using an ASA component with a graft rubber composition having a bimodal particle size distribution [76]. A different approach to solve shortcomings in impact strength and colourability of an ASA-PC blend is the addition of a SAN-grafted EPDM rubber [77]. The low-temperature impact strength can be significantly raised by adding a methyl methacrylate grafted polydimethylsiloxane—poly (n-butyl aerylate) interpolymer [78]. Ternary blends of ASA-PC and an amorphous polyester were shown to have improved melt flow and thick section impact strength in comparison with ASA-PC blends [79]. To improve the notched impact strength and to prevent hydrolysis induced embrittlement of PBT, PBT—styrenic blends have been thoroughly investigated. In comparison with ASA-PC, blends of ASA and PBT generally have to be compatibilized in order to yield acceptable mechanical properties. In the case of glass fibre reinforced PBT—ASA blends, an added and important bonus is reduced warpage properties on injection moulding in comparison with reinforced PBT. One option for compatibilizing PBT with ASA is the use of reactive polymeric additives that are compatible with the PSAN matrix and bear groups (e.g. carboxylic, carboxylic anhydride, epoxide or isocyanate functionalities) for forming chemical bonds with the hydroxyl and carboxyl groups in PBT. A recent study reported on the hydrothermal ageing of ASA—PBT blends containing up to 30% ASA [80]. It was revealed that the moisture uptake at 60CC corresponds to the percentage of modifier rubber. Strikingly, the mechanical properties deteriorate with increasing ASA content, probably resulting from insufficient compatibilization between the two phases. An ASA—PBT with improved hydrolysis resistance and reduced warp was reported for a resin composition containing a difunctional epoxy compound such as bis(3,4-epoxycyclohexylmethyl) adipate [81]. To increase the heat distortion temperature of a PBT—ASA blend by 10–20°C, the addition of talc at a
SYNTHESIS, PROPERTIES AND APPLICATIONS OF ASA
355
concentration of 0.5% is recommended [82]. To overcome disadvantages in colorability, a blend of 40 wt% PMMA, 10 wt% PSAN and a graft butyl acrylate rubber blend with a bimodal particle size distribution was developed [83]. In blends of polyamide with styrenics, compatibilization of the phases is generally a prerequisite for acceptable mechanical properties. This can be accomplished by having reactive groups in the PSAN phase that react with the amino groups of the polyamide phase. Blending is used to improve the main weaknesses of the polyamide, namely moisture sensitivity, toughness and warpage on injection moulding. Another motivation for blending ASA is to improve the heat distortion temperature of the material. This can be achieved by increasing the Tg of the continuous PSAN phase, for example by substituting the PSAN matrix partially or completely with poly(a-methylstyrene-acrylonitrile) [84]. This leads to an increase in the Vicat temperature of up to approximately 110°C. On increasing the amount of the a-methylstyrene-acrylonitrile copolymer, the ease of processability decreases, so often a mixture of both copolymers is used. [85] Alternatively, the Tg of the PSAN phase can be increased by using a terpolymer of a-methylstyrene-acrylonitrile-acrylamide [86]. Another approach to increase the Tg of the continuous phase of the ASA products is to blend it with maleic anhydride- or N-phenylmaleimide-containing resins. The use of matrix tetrapolymers of styrene, acrylonitrile, maleic anhydride and Nphenylmaleimide is recommended to improve considerably the heat distortion temperature [87]. Similarly, a matrix consisting of N-phenylmaleimide—acrylonitrile-styrene used in combination with a core-shell ASA rubber with a polyorganosiloxane core was found to have good colourability and excellent lowtemperature impact strength [88]. It must be remembered, however, that it may be necessary to adjust the graft comonomer composition when adding polar high-Tg copolymers to the matrix. For a matrix containing 22–36 wt% of a styrene—maleic anhydride copolymer and 36–50 wt% of PSAN, the resulting material exhibits an optimized impact strength when the acrylonitrile content of the rubber graft shell is in the range 45–55 wt% [89]. Alternatively, a matrix blend consisting of PSAN, a-methylstyrene-acrylonitrile copolymer and poly-N-methylglutarimide can be used, yielding ASA with Tg up to 120°C [90].
7 7.1
APPLICATIONS OF ASA GENERAL
Owing to its favourable combination of properties, ASA has grown substantially from a small niche product to an important plastic material since its
356
G. E. McKEE ET AL.
introduction into the market some 30 years ago, and it is now being used for many interior and exterior applications. The largest area of application for ASA is the automotive sector [91,92]. Almost all automotive manufacturers use ASA for unpainted exterior parts such as mirror housings, radiator grills, cowl vent grills and pillar covers. ASA is also increasingly being used in the transportation industry [93] and for truck and motor scooter segments, for example for truck door steps, radiator grills, scooter fairings, and other, mostly unpainted, exterior parts. A long-standing use for ASA is for housings for various equipment, machines, etc. [94–99] A further important field of application for ASA is the household sector. Typical applications include: • housings for electric toothbrushes and interdental cleaners; • housings and covers for sewing machines and kitchen appliances; • garden equipment, such as lawnmower housings, parts for lawn and garden irrigation devices and housings for garden lamps [100]; • children's toys [101]. In the building sector, ASA has found various applications such as window frames and sanitary equipment [102–105]. The leisure and sports sector constitutes a further important area of application. Thus, ASA is used to manufacture boat hulls, roofs and sills for camper vans and other exterior plastic parts for recreational vehicles. Since ASA has been on the market for over three decades and has thus acquired a certain maturity, it would be easy to conclude that no more major innovations can be expected from such a product. However, several important new fields of application are currently being developed. Among these are solar and safety technology and automotive body panels. 7.2
SOLAR TECHNOLOGY
Alternative sources of energy such as solar power are gaining significance because of the threat of global warming induced by the use of fossil fuels. A number of new applications involving ASA in the use of solar energy, such as housings for solar collectors and carriers for solar cells, have been developed in recent years ASA is also being used for the housings of solar-powered street numbers, which are already being made by several manufacturers. They contain photovoltaic cells in order to deliver electricity to rechargable batteries. Further applications for ASA in the area of solar energy include, for example, sunlight sensors, solar-powered battery chargers and solar-charged flashlights.
SYNTHESIS, PROPERTIES AND APPLICATIONS OF ASA
7.3
357
SAFETY IN THE HOUSE AND IN THE OFFICE
Another relatively new but strongly growing area of application is that of security and safety components. More and more people are relying on electronic systems to protect themselves, their homes and their offices. Such systems are comprised of a variety of components such as sensors, movement detectors and transmitter modules. Since many of these parts are mounted on the exterior of buildings, a main requirement is long-term weather resistance. Examples of applications in which ASA is already in use are [106] • • • • •
sensors and transmitter modules for electronic alarms; movement detectors; safety lighting and alarms; housings for access control units: housings for fire alarms.
7.4
A very promising new field of application for ASA and ASA-PC blends is automotive body panels. Until now, these have been made almost exclusively from painted metal. Since painting is costly, automotive companies are attempting to save costs by developing technologies that dispense with this operation. As in the case of the 'smart' car, bulk coloured thermoplastics (e.g. PBT—PC blends) may be used to make body panels. However, such parts must still be sprayed with a top coat in order to achieve the desired scratch resistance and UV stability. A promising alternative for the production of unpainted body panels is the use of a co-extruded thermoplastic film as a decorative layer which is backmoulded with a thermoplastic material [107]. This technique has received the name 'paintless film moulding' (PFM®) (Figure 16.16). The thermoplastic films for the PFM® technology normally have a total thickness of approximately 1 mm (Figure 16.17) and have an outer layer of PMMA which provides high gloss, scratch resistance and excellent weatherabilty. The base layer of the film is made from coloured ASA, chosen for its good adhesion both to PMMA and the backmoulding material (e.g. ABS) and for its excellent weatherability and long-term heat resistance. If special colour effects, e.g. metallic effects, are required the film may contain an additional intermediate layer made from high-impact PMMA. In order to make parts by means of the PFM® technology, the co-extruded film is thermoformed and then backmoulded with a compatible thermoplastic. Alternatives for the backmoulding material include, for example, ABS (without or with glass fibre reinforcement) and glass fibre reinforced PBT—ASA blends.
358
G. E. McKEE ET AL.
Preform design: 1. PMMA — r2. Coloured PMMA 3. ASA, ASA/PC -
Layer for layer toward resistance 1. Top layer • gloss • scratch resistance • hardness • weatherability
I and 2= top layers
3.-Film base layer • Injection moulded backing material 4a. ABS (trim, mirror housing, etc.) 4b. e.g. ABS-GF, PBT/ASA-GF (large-area body components)
Figure 16.16
2. Color layer • colour • UV resistance
4. Injection moulded backing material • strength • rigidity 3. Film base layer • impact strength • thermal expansion • toughness • resistance to thermal • colour mechan. strength deformation
Process sequence for paintless film moulding
Process sequence Coextrusion Multiple layer film extrusion
Film preforming Thermoforming
Component production Film insertion into mold
Trimming
In-mold decoration
-oFigure 16.17
Structure of the films used for the PFM* technology
The PFM® films may also be backed with a glass fibre reinforced polyurethane foam. This technology is already being used in thermoplastic roof modules and gives rise to parts having low weight, high stiffness and excellent thermal insulation.
SYNTHESIS, PROPERTIES AND APPLICATIONS OF ASA
8
359
FUTURE PERSPECTIVES
Although ASA has been marketed for more than 30 years, it is still a product with a considerable potential for new applications owing to the well-balanced cost—property relationship. The strength of ASA lies in the unique combination of its good weatherability, toughness, surface properties and resistance to chemicals. To be successful as an ASA supplier in the future the following prerequisites seem to be mandatory: back integration for the main raw materials, innovative process and development, development of new applications coupled with a reliable and skilled technical customer support service.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
Lasche H., van Roessel R., Kunststoffe 91, 276 (2001). Herbig J., Salyer I. O.; Monsanto Co., US 3 118 855. Otto H.-W.; BASF Aktiengesellschaft, DE 1182811. Siebel H. P., Otto H.-W.; BASF Aktiengesellschaft, DE 1 238207. Willersinn H., Otto H.-W., Paul R., Schuster L.; BASF Aktiengesellschaft, DE 1260135. van der Werth A., Lederer F.; BP Chemiclas, DE 1 960409. Luetje H.; BASF Aktiengesellschaft, DE 2 311129. Hambrecht J., Schmitt B., Rebafka W., Stephan R., Schwaab J.; BASF Aktiengesellschaft, DE 3 134 103. Yu J. R., Gallager R. E.; Stauffer Chemical Company, US 3944631. Swoboda H., Lindenschmitt G., Bernhard C; BASF Aktiengesellschaft, DE 2826925. McKee G. E., Koch J., Fischer W., Rosenau B., Czauderna B.; BASF Aktiengesellschaft, DE 19508312. Laber W., Gottschalk A., Schwaab J., Jeckel G., Mosthaf H.; BASF Aktiengesellschaft, DE 2 037 784. Ramsteiner F., Kunststoffe, 67, 517 (1977). Mckee G. E., Gausepohl H., Moors R., Rosenau B., Heckmann W.; BASF Aktiengesellschaft, DE 4440676. McKee G. E., Moors R., Gauspohl H., Seibring J.; BASF Aktiengesellschaft, EP 792298. McKee G. E., Rosenau B.; BASF Aktiengesellschaft, DE 19623661. McKee G. E., Rosenau B.; BASF Aktiengesellschaft, DE 19614845. McKee G. E., Jimgling S., Warzelhan V, Gausepohl H.; BASF Aktiengesellschaft, DE 19651300. Jungling S., Mckee G. E., Warzelhan V., Gausepohl H., Fischer M.; BASF Aktiengesellschaft, DE 19651299. Unpublished work. Mckee G. E., Renz G., Jahns E., Kastenhuber W.; BASF Aktiengesellschaft, DE 19633626.
360
G. E. McKEE ET AL.
22. McKee G. E., Rosenau B., Goerrissen H., Jahns E.; BASF Aktiengesellschaft, DE 19702733. 23. McKee G. E., Jahns E., Fischer W., Guentherberg N., Rosenau B.; BASF Aktiengesellschaft, DE 4443 886. 24. Kanebuchi Kagaku Kogyo KK, JP 101 2074. 25. Brandstetter F., Hambrecht J., Hildenbrand P., Echte A.; BASF Aktiengesellschaft, DE 3114090. 26. Unpublished work. 27. Mark H. F., et al. (Eds), Encyclopedia of Polymer Science and Engineering, Wiley, New York, (1985), pp. 388–425. 28. Vollmert B., Grundriss der Makromolekularen Chemie, Vol. 2, Vollmert-Verlag, Karlsruhe, 1979, pp. 219-225. 29. Rosenau B. Kunststoffe 85, 805 (1995). 30. Zahn A., Kunststoffe 87, 314, (1997). 31. Gorrissen H., Kunststoffe Plast. Eur. 89, 31 (1999). 32. Ramsteiner F., McKee G. E., Heckmann W., Fischer W., Fischer M., Ada Polym. 48, 553 (1977). 33. Mittnacht H., Priebe E.; BASF Aktiengesellschaft, DOS 1 911 882. 34. Eichenauer H., Doering J., Ott K.-H., Bottenbruch L.; Bayer AG, EP 139271. 35. Eichenauer H., Zabrocki K., Doering J., Ott K.-H., Bottenbuch L.; Bayer AG, DE 3421 353. 36. Ostermayer B., McKee G. E.; BASF Aktiengesellschaft, DE 3620684. 37. Niessner N., Seitz F.; BASF Aktiengesellschaft, DE 4131 729. 38. Niessner, N., Fischer W., Guentherberg N., Ruppmich K., Seitz F; BASF Aktiengesellschaft, DE 4142910. 39. Fischer W., Guntherberg N., Niessner N., Ruppmich K., Seitz F.; BASF Aktiengesellschaft, DE 4216549. 40. Niessner N., Seitz F., Fischer W., Guentherberg N., Ruppmich K., Moors R., Weiss R.; BASF Aktiengesellschaft, DE 4221 293. 41. Wittmann D., Schoeps J., Beicher H., Piejko K.-E., Weirauch K.; Bayer AG, DE 4229642. 42. Fischer W., Deckers A., Guentherberg N., Niessner N.; BASF Aktiengesellschaft, DE 4 234 296. 43. Fischer W., Guntherberg N., Niessner N.; BASF Aktiengesellschaft, DE 4235976. 44. Fischer W., Guentherberg N., Niessner N.; BASF Aktiengesellschaft, DE 4237 640. 45. Fischer W., Guntherberg N.; BASF Aktiengesellschaft, DE 4242485. 46. Fischer W., Guentherberg N.; BASF Aktiengesellschaft, DE 4439969. 47. McKee G. E., Rosenau B., Wendel K.; BASF Aktiengesellschaft, EP 732377. 48. Mckee G. E., Koch J., Fischer W., Gottschalk A., Guentherberg N., Rosenau B.; BASF Aktiengesellschaft, DE 19509514. 49. Rosenau B., McKee G. E., Schweiger C; BASF Aktiengesellschaft, DE 19536892. 50. McKee G. E., Rosenau B., Heckmann W.; BASF Aktiengesellschaft, DE 19614844. 51. McKee G. E., Rosenau B.; BASF Aktiengesellschaft, DE 19614845. 52. McKee G. E., Roseenau B., Heckmann H.; BASF Aktiengesellschaft, DE 19614846. 53. Bennet J. H., Muelbach K., Kogowski G.; BASF Aktiengesellschaft, WO 9933914. 54. Goerrisen H., Morgenstern H., McKee G. E.; BASF Aktiengesellschaft, DE 19837854. 55. Niessner N., Fischer W.; BASF Aktiengesellschaft, DE 4342045. 56. Seitz F, McKee G. E., Buschl R.; BASF Aktiengesellschaft, DE 4000544.
SYNTHESIS, PROPERTIES AND APPLICATIONS OF ASA 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91.
361
Eckel T., Ooms P., Wittmann D., Buysch H., Bayer AG, DE 4231 774. Techno Polymer KK Aktiengesellschaft, JP 11116827. Seitz F., Ruppmich K.; BASF Aktiengesellschaft, DE 4005210. Fischer W., Koch J., McKee G. E.; BASF Aktiengesellschaft, EP 725091. Niessner N.; BASF Aktiengesellschaft, DE 4124251. Niessner N., Seitz F.; BASF Aktiengesellschaft, DE 4342048. Mitsubishi Rayon, JP 05 339 324. Mitsubishi Rayon, JP 08041 149. Craig D., Hu R.; General Electric Co., WO 200034346. Kato T., Izumi M., Hayashiba Y., Suenaga K., Othake H.; Mitsubishi Rayon Co., DE 2 037 419 (1970). Schirmer H., Peilstocker G., Schuster H.; Bayer AG, DE 2259564 (1972). Yu A. J.; Stauffer Chem. Co., US 4148 842 (1978). Brandstetter F., Hambrecht J., Muenstedt H.; BASF Aktiengesellschaft, DE 3149812(1981). Mitulla K., Swoboda J., Schmitt B., Wassmuth G.; BASF Aktiengesellschaft, EP 111260(1982). Chen C., Peng F. M.; Bayer AG, WO 200060007 (1999). Gilles H. F., Sasserath J. N.; General Electric Co., US 4579909 (1984). Peascoe W. J.; General Electric Co., EP 269861 (1986). McHale A. H., Peascoe W. J.; General Electric Co., EP 272425 (1986). Mitulla K., Hambrecht J., Echte A., Swoboda J., Siebel P., Schwaab J., Frank H.; BASF Aktiengesellschaft, EP 164513 (1984). Wassmuth G., Ruppmich K., Seiler E., Gausepohl H., Benker K.; BASF Aktiengeselschaft, EP 244856 (1986). Wassmuth G., Ruppmich K., Seiler E.; BASF Aktiengesellschaft, EP 244857 (1986). Fujiguchi T., Saito A., Itoi H.; General Electric Co., EP 663425 (1994). Udipi K.; Monsanto Co., EP 440007 (1989) Mohd Ishak Z. A., Ishiaku U.S., Karger-Kocsis J. J. Appl. Polym. Sci. 74, 2470 (1999). Talibuddin S. H., Sastri V.R., Mercx F., Cheret E., Gallucci R.; General Electric Co., WO 200046296 (1999). Cheret E., De Vries R., Mercx F., Kwiecinski V.; General Electric Co., WO 200042105 (1999). Niessner N., Ruppmich K., Mosbach N.; BASF Aktiengesellschaft, EP 603674 (1992). Osima, A., Casale, A., Orsatti, E. Dakli, I.; Montecatini Societa Generate, DE 1569194 (1964). Stein D., Haaf F., Priebe E.; BASF Aktiengesellschaft, DE 2140437 (1971). Jansen U, Ott K.-H., Suemmermann K., Frohberg E.; Bayer AG, EP 330038 (1988). Robison J. C., Ziegelmeyer T.A.; General Electric Co., EP 415 120 (1989). Mitsubishi Rayon Co., JP 08 199025 (1995). Lindner C., Braese H.-E.; Bayer AG, DE 4110975 (1991). Guentherberg N., Deckers, A., Niessner N.; BASF Aktiengesellschaft, DE 4 229 913 (1992). Lindenschmidt G., Ruppmich K., in Proceedings of the Conference 'ABS and Related Polymers in the Automotive Industry', Sueddeutsches Kunsststoff-Zentrum, Wurzburg, Sept. 28–29, 1993.
362
G. E. McKEE ET AL.
92. Naarmann H., McKee G. E., Pirker A., Sterzel H.-J., Brandstetter F., Bernstorff B.-S., Rosenau B., Endemann U., Straube B.; BASF Aktiengesellschaft, DE 19630099. 93. Naarmann H., McKee G. E., Pirker A., Sterzel H.-J., Brandstetter F., Bernstorff B.-S., Rosenau B., Endemann U., Straube B.; BASF Aktiengesellschaft, DE 19630063. 94. Naarmann H., McKee G. E., Pirker A., Sterzel H.-J., Brandstetter F., Bernstorff B.-S., Rosenau B., Endemann U., Straube B.; BASF Aktiengesellschaft, DE 19630142. 95. Naarmann H., McKee G. E., Pirker A., Sterzel H.-J., Brandstetter F., Bernstorff B.-S., Rosenau B., Endemann U., Straube B.; BASF Aktiengesellschaft, DE 19630143. 96. Naarmann H., McKee G. E., Pirker A., Sterzel H.-J., Brandstetter F., Bernstorff B.-S., Rosenau B., Endemann U., Straube B.; BASF Aktiengesellschaft, DE 19630144. 97. Naarmann H., McKee G. E., Pirker A., Sterzel H.-J., Brandstetter F., Bernstorff B.-S., Rosenau B., Endemann U., Straube B.; BASF Aktiengesellschaft, DE 19630117. 98. Naarmann H., McKee G. E., Pirker A., Sterzel H.-J., Brandstetter F., Bernstorff B.-S., Rosenau B., Endemann U., Straube B.; BASF Aktiengesellschaft, DE 19630120. 99. Naarmann H., McKee G. E., Pirker A., Sterzel H.-J., Brandstetter F., Bernstorff B.-S., Rosenau B., Endemann U., Straube B.; BASF Aktiengesellschaft, DE 19630103. 100. Naarmann H., McKee G. E., Pirker A., Sterzel H.-J., Brandstetter F., Bernstorff B.-S., Rosenau B., Endemann U., Straube B.; BASF Aktiengesellschaft, DE 19630061. 101. Naarmann H., McKee G. E., Pirker A., Sterzel H.-J., Brandstetter F., Bernstorff B.-S., Rosenau B., Endemann U., Straube B.; BASF Aktiengesellschaft, DE 19630135. 102. Naarmann H., McKee G. E., Pirker A., Sterzel H.-J., Brandstetter F., Bernstorff B.-S., Rosenau B., Endemann U., Straube B.; BASF Aktiengesellschaft, DE 19630098. 103. Naarmann H., McKee G. E., Pirker A., Sterzel H.-J., Brandstetter F., Bernstorff B.-S., Rosenau B., Endemann U., Straube B.; BASF Aktiengesellschaft, DE 19630095. 104. Naarmann H., McKee G. E., Pirker A., Sterzel H.-J., Brandstetter F., Bernstorff B.-S., Rosenau B., Endemann U., Straube B.; BASF Aktiengesellschaft, DE 19630118. 105. Naarmann H., McKee G. E., Pirker A., Sterzel H.-J., Brandstetter F., Bernstorff B.-S., Rosenau B., Endemann U., Straube B.; BASF Aktiengesellschaft, DE 19630097. 106. Treede H. J., Euro Security No. 6, 31 (1998). 107. A. Grefenstein, Metalloberflache 53 (Oct.) 2 (1999).
P A R T IV
Syndiotactic Polystyrene
This page intentionally left blank
17
NORIO TOMOTSU Polymer Research Laboratory, Idemitsu Petrochemical Co., Ltd, Chiba, Japan
MICHAEL MALANGA R&D Engineering Plastics, The Dow Chemical Company, Midland, Ml, USA
JUERGEN SCHELLENBERG R&D Engineering Plastics, Dow Central Germany Schkopau, Germany
1
INTRODUCTION
Polystyrene was commercialized by I. G. Farben in 1931 and it has long been used as a commodity plastic. Although polystyrene is endowed with excellent properties not found in other commodity plastics such as polyolefins, its amorphous nature (relatively low heat and solvent resistance) limits its use in some application areas. Karl Ziegler first discovered in 1953 that transition metal compounds could be activated by aluminum alkyls and used as organometallic catalysts to polymerize ethylene. Giulio Natta discovered stereoregular polymers such as isotactic polypropylene and isotactic polystyrene (IPS) were also prepared using this same family of new catalysts [1,2]. IPS is a semi-crystalline polymer with a melting point of ~ 240 °C. Some companies have tried to commercialize IPS with the idea that it should be a plastic with higher heat resistance. Unfortunately, the crystallization rate of IPS is too slow to be practical in injection molding. One of the most important recent achievements in the field Modern Styrenic Polymers: Polystyrene and Styrenic Copolymers. Edited by J. Scheirs and D. B. Priddy r 2003 John Wiley & Sons Ltd
366
T. NORIO ET AL.
of polymerization catalysts has been the introduction of methylaluminoxane (MAO) by Sinn and Kaminsky [3,4]. The controlled reaction between water and trimethylaluminum produces MAO. The polymerization activities of metal compounds with MAO are higher than those attainable with traditional alkylaluminum compounds. Additionally, this soluble catalyst system can be more easily used to control the stereoregularity of produced polymers by varying the ligand structure of the metallocene. Ishihara et al. first succeeded in the synthesis of syndiotactic polystyrene (SPS) in 1985 [5–7]. With a melting point of 270 °C and a crystallization rate much faster than that of IPS, SPS has been commercialized as a new Engineering Plastic. The detailed physical and mechanical properties are described in Chapter 18. The Dow Chemical Company and Idemitsu Petrochemical Co. Ltd started joint research work on SPS in 1988 and have together vastly improved the catalyst system, polymerization technology, manufacturing process and application areas for this new material. The first commercial plants for the production of SPS were built in Japan in 1996 and in Germany in 1999. Both companies supply SPS products to the plastics industry (tradenames: XAREC1 from Idemitsu Petrochemical Co. Ltd and QUESTRA* from Dow Chemical Co.). One of the most important issues for the commercialization was cost reduction for the production. Idemitsu Petrochemical Co. Ltd and Dow Chemical Co. succeeded in reducing the production cost by catalyst activity improvement, polymerization condition optimization and process development.
2 CATALYTIC SYSTEMS FOR SPS Since 1985, many different transition metal compounds have been examined for their ability to produce syndiotactic polystyrene in combination with counterions based on methylalumoxane, borane, borate and other chemicals. 2.1 2.1.1
TRANSITION METAL COMPLEXES Metal
Typical transition metal complexes investigated are summarized in Table 17.1 together with the polymerization conditions, the polymer properties, and the catalytic activities. Yang et al. examined rare earth coordination catalysts. The Nd(naph)3/ Al(iBu)3 catalyst system was found to produce syndiotactic-rich polystyrene [8]. They proposed that the catalytically active species might be an ionic complex because the addition of CCU increased the catalytic activity.
SYNTHESIS OF SYNDIOTACTIC POLYSTYRENE
367
Table 17.1
Polymerization of styrene using various transition metal compounds
Compound
Metal MAO Conversion (mmol) (mmol) (%)a
TiCl4 Ti(OMe)4 CpTiCl3 Ti(acac)2Cl2 ZrCl4 Zr(CH2Ph)4 VOC13 Nb(OEt)5 Ta(OEt)5 Cr(acac)3 MoO(acac) Fe(acac)3 Co(acac)3 Ni(acac)2
0.05 0.05 0.05 0.01 0.05
0.2 0.05 0.25 0.25 0.02 0.02 0.02 0.02 0.25
40 40 40 8 10 16 40 20 20 10 10 10 10 20
4.1 3.8 92.3
0.4 0.7 2.0 0.2 0.2 Trace
1.4 0.5 0.5 1.8 80.8
syndiotactic syndiotactic syndiotactic syndiotactic atactic syndiotactic atactic atactic atactic atactic atactic atactic atactic atacticc
Stereospecificity polymerization conditionsb
1 1 2 1 3 5 4 1 1 4 4 4 4 4
a
Conversion from styrene to polymer. Polymerization conditions:styrene:toluene (ml/ml): (1) 180:100; (2) 23:50; (3) 100:50; (4) 50:100; (5) 40:90. Polymerization temperature and time: (1)–(4) 50 °C, 2h; (5) 90 °C, 4h. c Iso-rich polymer. b
Recently, Wakatsuki and co-workers have shown that samarium compounds produce SPS with lower syndiotacticity than titanium compounds [9]. Group 4 transition metal complexes showed higher activity and higher syndiospecificity than the other metal complexes. The ansa-zirconocene compounds show lower activity and lower syndic-directing Stereospecificity than the corresponding ansa-titanocenes. Zambelli and co-workers also found that Zr compounds [e.g. Zr(CH2C6H5)4, Zr(C7H8)2, ansa-Cp2ZrCl2] catalyze the syndiospecific styrene polymerization [10–14]. Among Group 4 transition metals, titanium compounds show the highest performance.
2.1.2
--Ligand
The polymerization activities of bis-cyclopentadienyl titanium compounds are lower than those of bridged bis-cyclopentadienyl titanium compounds. Miyashita et al. reported the polymerization activities of several bridged bis-cyclopentadienyl titanium compounds [15]. They found that the catalytic activity of CH2(Cp)2TiCl2 is the highest among bis-cyclopentadienyl titanocene compounds. The data indicate that the polymerization activities and also syndiospecificity increase with a decreasing angle between the Cp centroid—Ti—Cp centroid in bis-cyclopentadienyl titanocene compounds.
368
T. NORIO ET AL.
Among the titanium complexes producing SPS, monocyclopentadienyltitanium compounds show the highest polymerization activities and highest syndio-directing stereospecificity as compared with non-cyclopentadienyl, biscyclopentadienyl and bridged bis-cyclopentadienyl titanium complexes. A comparison of the polymerization behavior of CpTiCl3 and [Me5C5]TiCl3 (Cp*TiCl3) under equivalent experimental conditions shows that the Cp* derivative gives a higher degree of syndiotacticity in the resulting polymer, much higher molecular weights and predominantly an increased polymerization activity. The structure of the cyclopentadienyl ligand of the metal complex has been extensively varied. Some of the investigated complexes with substituted monocyclopentadienyl ligands are shown in Table 17.2 and Figure 17.1. The data indicate that substituents on the cyclopentadienyl ligand which are electron releasing generally yield higher polymerization activities. This result suggests stabilization of the active site by electron-releasing substituents. The catalytic activities of bulky substituents such as tBu or SiMe3 are lower than is expected from the electron density of titanium. These bulky groups are believed to sterically hinder the coordination of styrene monomer to the metal catalyst. Recently, Aoyama et al. have shown that Cp with a side ring ligand such as 4,5,6,7-tetrahydroindenyltitanium trimethoxide led to good activity [16]. The strain of the side ring increases the electron donation of the Cp ligand and probably improves the catalyst activity of titanocene. 49Ti NMR (49TiCl4 as standard) shows an electron density on titanium of [l,2,3-Me3-5,6,7,8-tetrahydroindenyl]TiCl3 of –95.3ppm, which is higher than that of Cp*TiCl3 (—80.7 ppm). The polymerization results are shown in Figure 17.2. The catalytic activity of [1,2,3-Me3-5,6,7,8-tetrahydroindenyl]TiCl3 is higher than that of Cp*TiCl3. Ready et al. [17] observed that IndTiCl3 is a significantly better catalyst than CpTiCb. Takeuchi et al. [18] also examined similar compounds at almost the same time, but their performance is lower than that of CpTiCl3. The difference in these results might be due to the difference in polymerization conditions. Takeuchi's et al. results suggest that substituents on the five-membered ring generally increase the catalytic activity among indenyl ligands. Table 17.2
Relationship between catalyst structure and catalytic activity
Compound
Catalytic activity (kg/g Ti)
(EtMe4Cp)Ti(OMe)3 Cp*Ti(OMe)3 [(tBu)2Cp]Ti(OMe)3 (Me4Cp)Ti(OMe)3 [(Me3Si)2Cp]Ti(OMe)3 CpTi(OMe)3
270 265 15 150 30 10
1
H
NMR: (OMe) (ppm) 4.045 4.050 4.072 4.081 4.088 4.112
369
SYNTHESIS OF SYNDIOTACTIC POLYSTYRENE 300
250 S 200
•>
150
100
U
50 01— 4.04
4.06
4.08
4.10
4.12
Chemical shift of MeO/ppm Figure 17.1 Relationship between chemical shift of methoxy group and catalytic activity 80 70
^ 60 .1 50 £ 40 c o U 30 20 10
0 50
60
70
80
Temp.fC) Figure 17.2 Catalytic activity of (•) [1,2,3-Me3-5,6,7,8-tetrahydroindenyl]TiCl3 and (+)Cp*TiCI3
2.1.3
a-Ligand
The polymerization activities in the presence of Cp*Ti compounds containing different ancillary ligands, i.e. alkoxide and chloride ligands, with MAO are as follows in order of decreasing catalytic activity [19]: Cp*Ti(OiPr)3 and
370
T. NORIO ETAL.
*Ti(OMe)3 > Cp*Ti(OPh)3 > Cp*Ti(OC6H4CH3)3 > Cp*TiCl3 > Cp*Ti(Oi C3HF6)3. Kaminsky showed that the catalytic activity of CpTiF3 is better than that of CpTiCl3 [20]. The chloride ligand and the electron-withdrawing alkoxide, OiC3HF6, decrease the conversion. The catalyst activity of all these systems is increased by the addition of triisobutylaluminum (TIBA) to the MAO—Ti complex mixture. By the addition of TIBA, all a-ligands are found to be substituted and the activity of the final catalyst system is practically the same. The ESR spectra of mixed solutions of Cp*Ti compounds, MAO and TIBA supported this hypothesis (Figure 17.3) [21]. 2.1.4
Other Catalysts
As a catalyst with a noncyclopentadienyl ligand, Kakugo et al. [22] examined the performance of bridged bisphenolato titanium complexes [(OC6H4-4-CH3-6tC4H9)2ZTiX2]. The catalyst activity of [(OC6H4-4-CH3-6-tC4H9)2S]Ti(OiPr)2 is higher than that of [(OC6H4-4-CH3-6-tC4H9)2CH2]Ti(OiPr)2 [22]. Okuda and Masoud showed that the catalytic activity depends on the nature of the bridging group Z and increases in the order—CH2_ < —CH2CH2_ < —S— < —SO—[23]. A titanium complex with a pyrazolylborate ligand was studied by Campbell and Malanga [19]. Similarities between the cyclopentadienyl ligand and the hydridotris(pyrazolyl)borate ligand have been noted for transition metal complexes. Catalyst efficiencies are much lower than those of the analogous pentamethylcyclopentadienyl complexes. 2.2 2.2.1
CO-CATALYSTS MAO
MAO is a useful co-catalyst for titanium metal complexes in the syndiotactic polymerization of styrene. Its complete structure and role are as yet not Cp*Ti(OMe)3
3315
3340
3365 G
Figure 17.3
ESR spectrum of Cp*TiR3 with MAO and TIBA
371
SYNTHESIS OF SYNDIOTACTIC POLYSTYRENE
clarified. Several types of MAO with different catalytic activities were produced by different synthesis conditions and the performance evaluated. MAO is produced by the reaction between trimethylaluminum (TMA) and water. The reaction is controlled by the reaction temperature. Small amounts of TMA always exist in MAO. The concentration of TMA in MAO solution has been measured by a titration method, but the reproducibility of the measurement is low. An improved method was reported using 1 HNMR of an MAO solution in toluene in the presence of a small amount of dioxane (Figure 17.4) [21]. This method is easier and more reliable than the titration method. The relationship between catalytic activity and TMA content in MAO is shown in Figure 17.5. Residual TMA in the MAO decreases the catalytic activity. The reactivity of TMA is too strong to control the active site formation and decomposes the active sites for polymerization. The MAO solution was 'dried' in many attempts to remove small amounts of TMA. It was found that not all the TMA could be removed under any conditions. Some amount of TMA is thought to coordinate strongly with MAO and seems to effect changes in the reactivity of MAO.
-0.5
-1.0
Shift/ppm
Figure 17.4
1
H NMR spectrum of MAO in dioxane
-1.5
T. NORIO ET AL.
372
10
20
30
40
50
60
TMA contents/% Figure 17.5 Effect of TMA in MAO on the catalytic activity. Polymerization conditions: 70 °C, 1 h polymerization, Cp*TiOMe3/MAO
MAO is an oligomer produced by the reaction between TMA and water. The molecular weight of MAO has been measured by a cryoscopic method. Dioxane can be used as a solvent for this method. The relationship between catalytic activity and molecular weight of MAO is shown in Figure 17.6. The catalytic activity showed the maximum around a molecular weight of 400 g/mol. It seems probable that a rapid alkylation of the metallocene by MAO takes place first, and the active species arises from a methyl transfer reaction between the metallocene alkyls and MAO. The active species formed from these reactions are probably complexed by a bulky MAOcoordinating non-quenching anion.
2.2.2
Borate
Some types of borate compounds act as co-catalysts for the syndiospecific polymerization of styrene in these catalyst systems. The active borate compounds have a tetraphenylborate anion. The effect of anions on the catalytic activity is summarized in Table 17.3. Fluorine substituents at the of 3,4,5positions increase the catalytic activity and tetrapentafluorophenylborate showed the highest performance. The cations formed react with the titanium complex and form active sites. The reaction equations are as follows: Cp*TiR3 + [RjNH][(C6F5)4B] -+ [Cp*TiR2]+[(C6F5)4B]
RH
373
SYNTHESIS OF SYNDIOTACTIC POLYSTYRENE
100
300
400
600
500
Mw
Figure 17.6 Effect of molecular weight of MAO on catalytic activity. The molecular weight of MAO was measured by a cryoscopic method. Dioxane was used as solvent Table 17.3
Effects of fluorine in borate
Borate
Catalytic activity (kg/g Ti)
[Me2HNPh][(C6H5)4B] [Me2HNPh][(C6H4F)4B] [Me2HNPh][(2,4-F2C6H3)4B] [Me2HNPh][(l, 3,4-F3C6H2)4B] [Me2HNPh][(C6F5)4B] [Me2HNPh][(3-CF3C6F4)4B]
0 0 5 10 250 20
Cp*TiR3 + [RJN][(C6F5)4B] -> [Cp*TiR2]+[(C6F5)4Br + RjN + RR Cp*TiR3 + [R^C][(C6F5)4B] -» [Cp*TiR2]+[(C6F5)4B]- + R3CR The use of borate with a small amount of TIB A as co-catalyst for polymerization of styrene to SPS was examined by Campbell and Malanga [19], Tomotsu [21] and Kucht et al. [24]. TIB A was found to be a good scavenger of impurities in styrene and to increase the syndiotacticity of the resultant polymer. The effects of cations on the catalytic activity are summarized in Table 17.4. Ammonium borate compounds with a lower pK a , such as [2-CN-pyridine(N)Me][B(C6F5)4] (pKa = —0.3), showed higher activity than those with higher pK a . The by-products of the active site formation reaction are thought to coordinate strongly to the active site. On the other hand, the by-product of the reaction between titanium compounds and [Ph3C][B(C6F5)4] is Ph3CH and
T. NORIO ET AL.
374
does not coordinate to the active site. In this case, however, the [Ph3C][B(C6F5)4] reacts with TIBA and is decomposed. [Ph(PhOMe)2C][B(C6F5)4] and [(PhOMe)3C][B(C6F5)4] are more stable compounds and do not react with aluminum alkyls. Lower reactivity against aluminum alkyls results in an increase in the apparent catalytic activity (Figure 17.7). The decomposition of the borate by TIBA is observed by 1H NMR and the excess amount of borate increases the catalyst activity.
2.2.3
Supported and Heterogeneous Catalysts
It has been demonstrated that mixtures of highly isotactic and highly syndiotactic polystyrene are obtained when using titanium compounds such as TiCl3 or TiCl4 supported on Mg compounds in the presence of MAO [25] (Table 17.5). In this situation, the two types of polymer polymerize simultaneously from two Table 17.4
Effects of pyridinium cations in borate
Borate
pKa
Catalytic activity (kg/g Ti)
[PyMe][(C6F5)4B] [4-CN-PyMe] [(C6F5)4B] [3-CN-PyMe] [(C6F5)4B] [2-CN-PyMe] [(C6F5)4B]
5.2 1.9 1.0 -0.3
2.0 2.8 31.8 40.5
[(MeOPh)3C]+
>-^
[(MeOPh)2PhC]+
40
Figure 17.7
60 80 TIBA/Ti
100
120
Catalytic activity of Cp*TiOMe3 with carbenium borate
375
SYNTHESIS OF SYNDIOTACTIC POLYSTYRENE Table 17.5
Polymerization by supported catalysts
Catalyst
Ti (mmol)
TiCl3(AA)
1.0 0.2 1.0 0.2 2.0 0.02 0.02 40 5 0.2 2.0 2.0 0.2
TiCl3 (Solvay) Mg(OEt)2/EB/TiCl4 TiCl4 Ti(OEt)4
a
Al/Ti
100 1000
20 1000
50 500 1000 10 40 500 10 50 500
a
Conv. (%)
Stereospecificity
8.2 2.0 1.9 0.9 2.9 1.1 1.4 7.2 0.4 0.7 0.3 2.5 0.9
Mixture of IPS and SPS Mixture of IPS and SPS Mixture of IPS and SPS Mixture of IPS and SPS IPS (84%) + SPS(16%) IPS(12%) + SPS (88%) IPS(10%) + SPS (90%,) IPS Mixture of IPS and SPS SPS Atactic PS SPS SPS
Polymerization conditions: styrene 50 ml, toluene 100ml, 50 °C, 2h.
different stereospecific active sites. Toluene-soluble titanium catalysts with MAO produce the syndiotactic polystyrene and isotactic polystyrene was produced by the heterogeneous titanium catalyst on the Mg support. The amount of titanium soluble in toluene was increased with increasing addition of MAO. Soga and Nakatani [26] examined Ti(OC4H9)4 supported on SiO2 with MAO and a reacted mixture of Ti(OC4H9)4 and MAO supported on SiO2. The syndiotacticity of the polymers with both catalysts was almost 100%. They found that the catalytic activity was independent of Al/Ti molar ratio. The yields in this case were very low. They suggested that SiOTi(OC4H9)3 heterogeneous species are more stable against reduction than the active species in the soluble system.
3 3.1
COPOLYMERIZATION POLYMERIZATION OF SUBSTITUTED STYRENES
When various ring-substituted styrenes were polymerized using CpTiCl3 with an MAO catalyst system, the corresponding syndiotactic polystyrenes were obtained. 13CNMR spectra of the phenyl C-l carbon of the poly(ringsubstituted) styrenes poly(p-methylstyrene), poly(m-methylstyrene), poly (p-tert-butylstyrene), poly(p-chlorostyrene), poly(m-chlorostyrene) and poly(p-fluorostyrene) were examined [27]. The spectra of each atactic poly (ring-substituted) styrene show many main peaks corresponding to their various
376
T. NORIO ETAL
configurational sequences. These spectra are similar to that of atactic PS. The spectrum of each syndiotactic poly(ring-substituted) styrene shows a single sharp peak at a higher magnetic field corresponding to the rrrr pentad configuration. These spectra are similar to that of SPS. In poly(p-fluorostyrene), the peak at high magnetic field is split, owing to coupling to the 19F nucleus. Soga and co-workers examined the relation between the Hammett's a value of each substituent and the reactivities in copolymerization (Figure 17.8). It is observed that the monomer reactivity is enhanced by electron-releasing substituents in the aromatic ring. Even p-ter/-butylstyrene with a substituent of large steric hindrance shows a high reactivity. This indicates that there is a strong polar effect of the substituent on the rate of addition [28]. Similar effects were observed by Ishihara et al. in the homopolymerization of these monomers [27]. In the syndiospecific polymerization of styrene, the monomer addition has been demonstrated as a secondary addition mechanism. Additionally, the monomer reactivity is enhanced by electron-releasing substituents in the aromatic ring. The electronic effect of the substituent of the ring is transmitted more efficiently to the methine than to the methylene carbon. Styrene and alkylstyrenes form co-syndiotactic polymers. Almost all of the syndiotactic polymers produced have a high degree of crystallinity, and all the Tm values of syndiotactic polymers are higher than those of isotactic polymers with the corresponding substituent. These results imply that the syndiotactic polymers might be useful in industry. 0.4
p-MeSt 0.2
p-BrSt o -0.2 m-ClSt -0.4 -0.2
Figure 17.8 merization
-0.1
0.0
0.1 a
0.2
0.3
0.4
Hammett's a value for substituted styrene and reactivity ratio of copoly-
SYNTHESIS OF SYNDIOTACTIC POLYSTYRENE
3.2
377
COPOLYMERIZATION OF STYRENE AND ETHYLENE
Mani and Burns [29] examined the copolymerization of ethylene and styrene by using TiCl3 with MAO. The polymer obtained by this catalyst system is isotactic polystyrene. There is no ethylene-styrene bond in the polymer. Tazaki [30] reported that a copolymer, which was obtained from ethylene and styrene using monocyclopentadienyl- or tetraalkoxytitanium compounds with MAO, had a small number of ethylene units in the polymer backbone. Seppala and coworkers [31] also examined the copolymerization of styrene and ethylene using titanium compounds with bulky alkoxy ligands and also CpTiCl3 with MAO. The products of the polymerization were a mixture of polyethylene and syndiotactic polystyrene. This suggests that there are more than two kinds of active sites in this catalyst system active for ethylene and styrene independent of each other. Pellecchia et al. [32] examined Ti(benzyl)4 with borane for producing ethylene-styrene copolymer, and it was reported that all sequences showed an alternating copolymer. Inoue and Shiomura [33] examined zirconium complexes for the ethylenestyrene copolymerization. They found alternating copolymer and homopolymer. The stereoregularity of the polymers was atactic. Kakugo et al. [22] reported that a catalyst based on 2,2'-thiobis(4-methyl-6tert-butylphenoxy)titanium dichloride [(TBP) TiCl2 and MAO produced a mixture of syndiotactic polystyrene and the alternating ethylene-styrene copolymer. They stressed the role of sulfur as essential to obtain the alternating copolymer. They reported that the copolymer was not produced using a similar compound having a methylene bridge [2,2/-methylenebis(4-methyl-6-ferf-butylphenoxy)titanium dichloride] instead of a sulfur bridge in the catalyst ligand. Subsequently a patent disclosed ethylene-styrene copolymerization promoted by catalysts based on bridged amidomonocyclopentadienyltitanium complexes such as [dimethylsilyl(phenylamido)(Cp*)]titanium dichloride and MAO. The copolymer obtained was a random copolymer but it did not contain any regioregularity in the arrangement of the styrene-styrene sequences. Suzuki and co-workers copolymerized styrene and ethylene by ansa-zirconocene compounds [34]. The polymer produced was isotactic polystyrene with ethylene. They found that the phenyl group of the monomer coordinates to the active site and decreases the catalytic activity.
3.3
COPOLYMERIZATION OF STYRENE AND DIENES
Pellecchia et al. copolymerized isoprene and styrene [35] and examined the copolymerization rate. They found a value for the product of the reactivity ratios of r1r2 — 2.3. The difference in the catalytic activity of styrene and isoprene may be due to the difference in coordination strength.
378
4
T. NORIO ETAL
MECHANISMS OF POLYMERIZATION OF STYRENE
Zambelli et al. reported on the mechanism of styrene polymerization [36]. They showed that the main chain of the syndiotactic polymer has a statistically transtrans conformation. It was established then the double-bond opening mechanism in the syndiospecific polymerization of styrene involves a cis opening. The details in the control of the monomer coordination for this polymerization mechanism were examined by Newman and Malanga using detailed 13C NMR. It was shown through the analysis of tacticity error (rmrr) that the tacticity in the polymer is chain-end controlled and that the last monomer added directs the orientation and coordination of the incoming monomer unit prior to insertion [37].
4.1
ACTIVE SITE SPECIES
The polymerization activity of a titanium catalyst increases with an increasing molar ratio of MAO to Ti. The amount of cationic Ti(III) species measured by ESR also increases with increasing ratio of MAO to Ti. This suggests that MAO acts as a reducing agent for Ti(IV) to Ti(III). Cationic Ti(III) might be an active species in the synthesis of SPS. Newman and Malanga [38] synthesized Cp*Ti(OMe)2 via reduction of Cp*Ti(OMe)3 and found that the catalytic activity of this complex with smaller amounts of MAO is almost the same as that of Cp*Ti(OMe)3. Metal alkyl compounds are known to reduce titanium compounds. The effect of reductants on the catalytic activity was evaluated and the data are summarized in Table 17.6 [21]. Very strong reducing reagents such as A1(CH3)3 and A1(C3H5)3 have a negative effect on the catalytic activity. In this case the titanium compound is reduced to Ti(II) by these reagents. As discussed earlier, however, the catalytic activity is significantly increased by the addition of TIBA. The details of the effect of TIBA were evaluated and are shown in Figure 17.9 [21]. TIBA increases the catalytic activity and reduces the molecular weight of the polymer. TIBA reacts with the titanium compound and reduces its oxidation state. Moreover, TIBA reacts with metal alkyl bonds and therefore acts as a chain transfer agent during polymerization by insertion and a new polymer end is started. The polydispersity (M w /M n ) produced by the titanocene catalyst with MAO is always found to be about 2.0. These results show that there is only one active site and that chain transfer occurs during the polymerization. The numbers and the kind of active sites were clarified by the stopped flow method. Without a chain transfer reaction, the polymerization proceeds like a 'living polymerization' and is an easy to analyze reaction. /j-Methylstyrene (PMS) was used in this experiment to avoid plugging during polymerization due to crystallization. The
379
SYNTHESIS OF SYNDIOTACTIC POLYSTYRENE
Table 17.6 Polymerization activities of [Me5Cp]TiCl3 with MAO and alkylation reagents Reagent
Relative activity"
Mw
None A1(CH3)3 A1(C2H5)3 Al(nC4H9)3 A1(iC4H9)3 Al(nC8H17)3 A1(C2H5)7(OC2H5) Zn(C2H5)2
100 13 23 76 560 100 140 48
750000 64000 84 000 570000 580000 670000 870000 130000
a
Catalyst activity without alkylation reagent as 100
300
200
2 3 T1BA/MAO
Figure 17.9
Effect of TIBA on the catalytic activity
results are shown in Figure 17.10. The Mw/Mn of the polymer was about 1.0. This strongly indicates that only one kind of active site is responsible for the polymerization of the styrene. The number of polymer chains indicated about half of the added titanium compound polymerize styrene. By the addition of hydrogen to PMS, the number of chains increased without any broadening of the molecular weight distribution. Hydrogen did not act as a chain transfer reagent but the number of polymer chains showed that almost all Ti compounds turned to active sites. After the polymerization with hydrogen, ethylmethylbenzene was observed as a by-product. The amount of ethylmethylbenzene is almost equimolar to the hydrogen added. These results showed that almost all of the titanium compounds turned to active sites, but about half of the active sites become dormant by some type of coordination with
380
T. NORIO ETAL
monomer. The addition of hydrogen activates this dormant site during polymerization. The structure of the active site is proposed in Figure 17.11 (a). The structure of the dormant sites might be an irregular coordination of the monomer [Figure 17.11(5)] or a change of the direction of the monomer coordination [Figure 17.11(c)]. The dormant site might be reactivated by the hydrogenation of styrene which irregularly coordinated to the active site. The misinsertion of styrene and the stereoirregularity are difficult to observe. We suppose that the polymerization reaction was stopped after an irregular coordination of styrene. This also supports the chain-end controlled mechanism of stereospecificity.
o
1.0
o
0
0
With hydrogen
0.8 o
Mw/Mn=1.05~1.15
H 0.6 c
•6 0.4
Without hydrogen
0.2 0.0
Figure 17.10 MAO
0.2
0.4 0.6 time/sec
0.8
1.0
Stopped flow polymerization of p-methylstyrene by Cp*TiOMe3 with
(c)
Figure 17.11 Proposed structure of active site and dormant site for styrene polymerization (a) Active site; (b) dormant site by the coordination error of monomer; (c) dormant site by the polymer chain rotation
SYNTHESIS OF SYNDIOTACTIC POLYSTYRENE
381
The active site structure was determinied by XANES or EXAFS because all of the titanium changed to active site. Figure 17.12 shows the results of these measurements. A sharp peak was observed in XANES but it disappeared on the addition of TIBA, MAO or borate. TIBA, MAO or borate changes the ligand coordination structure of the titanium complex. The position of the edge shifted to the left shows that the tetravalent titanium was reduced to trivalent titanium. The electron density of titanium was reduced by TIBA and MAO or borate and the cyclopentadienyl ligand came closer to the titanium. The structures of active sites produced using either MAO or borate are therefore thought to be almost the same. EXAFS shows a new signal outside the cyclopentadienyl ligand. The structure of the active site was examined by molecular modeling calculations. The reaction between styrene and CpTi(CH3)[CH(C6H5)CH3]+, CpTi(CH3)[CH (C6H5)CH3] and CpTi[CH(C6H5)CH3]+ was examined. The coordination of the monomer to Ti(III) is more stable than to Ti(IV) because the coordination of the vinyl group is fortified by the back-donation from the d-orbital to the vinyl group of styrene. We calculated the Ti(III) cation as an active site. This structure was proposed by Zambelli et al. [36] and involves a Ti(III) cation additionally coordinated to the phenyl group existing at the end of the growing polymer chain in a multi-hapto manner and forming a stabilized structure. However, this structure seems to be too stable to insert styrene by our calculation. The activation energy of the reaction of styrene and CpTi(CH3) [CH(C6H5)CH3] was 9kcal/mol and that of styrene and CpTi[CH(C6H5) CH3]+ was 28 kcal/mol. Therefore, we concluded that the active site is a neutral trivalent titanium instead of a trivalent titanium cation with a coordinated phenyl group. We suppose that MAO or borate exists near the active site and neutralizes the active site to prevent the coordination of the phenyl group. OMe or R
+TIBA+MAO or DMAS
Figure 17.12
XANES and EXAFS results for active sites
382
4.2
T. NORIO ETAL
KINETIC ANALYSIS OF STYRENE POLYMERIZATION
We assumed that the polymerization proceeds by a normal coordination polymerization. The effect of the catalyst concentration on the polymerization was examined by polymerization at different ratios of catalyst to styrene (Figure 17.13). The reaction rate increased in proportion to the catalyst ratio. However, the decay of the polymerization reaction was too fast to explain it as a firstorder reaction. We hypothesized that this polymerization proceeds by a single-site catalyst under different morphological conditions and variable monomer concentrations, i.e. polymerization in the crystalline polymer and in the amorphous polymer state. The frequency factor of the polymerization in the crystalline polymer should be lower than that in the amorphous polymer. The effect of monomer concentration on the polymerization rate is shown in Figure 17.14 using toluene as a solvent. The reaction rate is proportional to the monomer concentration. From these results, the polymerization reaction can be described by the following equations:
d[M]_ ~d0~~~*
conv. = 1 -exp^ —^-—— [1 -exp( - &<j,0)] — , 100
60
120
180
20
[ 1 -exp( -
240
Time/min Figure 17.13
Polymerization of styrene with different concentrations of catalyst
383
SYNTHESIS OF SYNDIOTACTIC POLYSTYRENE Conv/%
T>
40 30 20 10 0
20
40
60
8(
Monomer concentration/%
Figure 17.14
Effect of monomer concentration on the catalytic activity
where [M] = monomer concentration; 9 polymerization time; iq] = concentration of active site at higher monomer concentration; [qi = concentration of active site at lower monomer concentration; chain propagation rate constant of active site 1 , including effects of monomer concentration around the active site; chain propagation rate constant of active site 2, including effects of monomer concentration around the active site; decay rate constant of active site 1; decay rate constant of active site 2; *P,[C1*]0 =0.502exp( MQo = 0.050 exp ( - 894/f)[Ti]; fcdJCHo = 699000 exp ( - 5300/0; kd2[C*2]0 = 49400 exp ( - 5300/OUi]; t = polymerization temperature; [Ti] = titanium concentration. The polymerization rate constants were measured by adjustment of the equations above to the experimental polymerization results. The activation energy of chain propagation and decay for active site 1 and 2 were the same because this catalyst system is single site. The results of the fitting of calculation and polymerization are shown in Figure 17.15. The chain transfer reaction was also examined by these equations and the comparison of the calculated predictions and experimental results is shown in Figure 17.16.
384
T. NORIO ETAL
60
180
120
240 Time/min
Figure 17.15 Effect of polymerization temperature on catalytic activity. Line, calculated by kinetic equation; symbols, polymerization results Mw/1000 2000 1500 1000 500
0 50
60
70
80
90
Figure 17.16 Effect of polymerization temperature on Mw. Line, calculated by kinetic equation; symbols, polymerization results Mn = M
ka1 [C*][TIBA] + ka2[C*][MAO] where Mn = number average molecular weight; rp = poljonerization rate; rt — chain transfer rate; kt = termination rate constant;
SYNTHESIS OF SYNDIOTACTIC POLYSTYRENE
385
ka1 = chain transfer reaction rate constant of TIBA; ka2 = chain transfer reaction rate constant of MAO; kt = 1.19x 1010exp (-8157/r); ka1 = 1.64 x 10 7 exp( - 6860/0; ka2 = 1.19 x 108 exp(- 8615/0MAO and TIBA act as chain transfer agents, whereas ß-hydride elimination was not the main reaction for chain transfer with this catalyst system. Murata et al. [39] showed that a living polymerization was observed in low- temperature polymerization using this catalyst system. This result is in agreement with this model.
4.3
EFFECTS OF HYDROGEN
Triisobutylaluminum (TIBA) is a good activation agent for styrene polymerization: Cp*TiR3 + TIBA + MAO -> Cp*TiR(iBu) -> Cp*TiR(H) + (CH3)2C=CH2 As in the previous section, hydrogen increases the concentration of the active sites. Moreover, we think that the polymerization is preceded by the formation of a titanium hydride complex. If this is correct, hydrogen also increases the overall catalytic activity. The effects of hydrogen pressure on catalytic activity and molecular weight of SPS are summarized in Table 17.7. Hydrogen increases the catalytic activity and the polydispersity. The change in the molecular weight distribution might be the effect of the lack of uniformity of the polymerization system or the result of the formation of new types of active sites. Table 17.7
Effects of hydrogen pressure on styrene polymerization"
Hydrogen pressure (MPa)
Relative activity6
Mw/Mn
None 0.010 0.049 0.098
100 160 210 220
2.1 2.5 4.5 12.0
a b
Polymerization conditions: styrene bulk polymerization, 70 °C, 1 h. Catalyst activity without hydrogen as 100.
386
5
T. NORIO ETAL
CONCLUSION
Titanium compounds with MAO or borate as co-catalysts effectively produce syndiotactic polystyrene from styrene monomer. The design of high-performance catalyst systems is now well demonstrated. The basic structure of the active site, the mechanism of coordination and insertion and the kinetics are also now well understood for this new polymerization.
REFERENCES 1. Natta, G., Pino, P., Corradini, P., Danusso, F., Mantica, E., J. Am. Chem. Soc., 77, 1708 (1955). 2. Natta, G., Pino, P., Mantica, E., Danusso, F., Mazzanti, G., Peraldo, M., Chim. Ind. (Milan), 38, 124 (1956). 3. Sinn, H., Kaminsky, W., Vollmer, H.-J., Woldt, R., Angew. Chem., 92, 396 (1980). 4. Sinn, H., Kaminsky, W., Adv. Organomet. Chem., 18, 99 (1980). 5. Ishihara, N., Kuramoto, M., Uoi, M., to Idemitsu Kosan Co. Ltd., JP 62187708 (1985). 6. Ishihara, N., Kuramoto, M., Uoi, M., to Idemitsu Kosan Co. Ltd., EP 210615 (1986). 7. Ishihara, N., Kuramoto, M., Uoi, M., Macromolecules, 21, 3356 (1988). 8. Yang, M., Cha, C., Shen, Z., Polym. J., 22, 919 (1990). 9. Zhang, Y., Hou, Z., Wakatsuki, Y., Macromolecules, 32, 939 (1999). 10. Zambelli, A., Oliva, L., Pellecchia, C., Macromolecules, 22, 2129 (1989). 11. Grassi, A., Pellecchia, C., Longo, P., Zambelli, A., Gazz. Chim. Ital., 117, 249 (1987). 12. Zambelli, A., Pellecchia, C., Oliva, L., Shimin, H., Chin. J.. Polym. Sci., 6, 365 (1988). 13. Pellecchia, C., Longo, P., Grassi, A., Ammendola, P., Zambelli, A., Makromol. Chem., Rapid Commun., 8, 277 (1987). 14. Zambelli, A., Longo, P., Pellecchia, C., Grassi, A., Macromolecules, 20, 2035 (1987). 15. Miyashita, A., Mabika, M., Suzuki, T., presented at International Symposium on Synthetic, Structural and Industrial Aspects of Stereospecific Polymerization Proceeding, Milan, 1994. 16. Aoyama et al., in International Symposium on Future Technology, in press. 17. Ready, T. E., Day, R. O., Chien, J. C. W., Rausch, M. D., Macromolecules, 26, 5822 (1993). 18. Takeuchi, M., Shozaki, H., and Tomotsu, N., to Idemitsu Kosan Co. Ltd., EP0707013 (1993) 19. Campbell, R. E., Malanga, M. T., Metcon '93, (1993) 315. 20. Kaminsky, W., Metallocenes '96, 211 (1996). 21. Tomotsu, N., Polym. Prepr. Jpn., 42, 919 (1993). 22. Kakugo, M., Miyatake, T., Mizunuma, K., Stud. Surf. Sci. Catal., 56, 517 (1990). 23. Okuda, J., Masoud, E., Macromol. Chem. Phys., 199, 543 (1998). 24. Kucht, H., Kucht, A., Chien, J. C. W., Rausch, M. D., Appl. Organomet. Chem., 8, 393 (1994). 25. Soga, K., Monoi, T., Macromolecules, 23, 1558 (1990).
SYNTHESIS OF SYNDIOTACTIC POLYSTYRENE
387
26. Soga, K., Nakatani, H., Macromolecules, 23, 957 (1990). 27. Tomotsu, N., Ishihara, N., Newman, T. H., Malanga, M. T. J., Mol. Catal. A: Chem., US, 167(1998). 28. Monoi, T., Nakatani, H., Soga, K., Polym. Prepr. Jpn., 38, 1726 (1989). 29. Mani, R., Burns, C. M., Macromolecules, 24, 5476 (1991). 30. Tazaki, T., JP 2059871, 3087301. 31. Aaltonen, P., Seppala, J., Matilainen, L., Leskela, M., Macromolecules, 27, 3136 (1994). 32. Pellecchia, C., Pappalardo, D., D'Arco, M., Zambelli, A., Macromolecules, 29, 1158 (1996). 33. Inoue, N., Shiomura, T., Polym. Prepr. Jpn., 42, 2292 (1993). 34. Aral, T., Ohtsu, S., Suzuki, S., Macromol. Rapid Commun., 19, 327 (1998). 35. Pellecchia, C., Proto, A., Zambelli, A., Macromolecules, 25, 4450 (1992). 36. Zambelli, A., Giongo, M., Natta, G., Macromol. Chem., 112, 183 (1968). 37. Newman, T. H., Malanga, M. T., 4th SPSJ International Polymer Conf., (1993)27. 38. Newman, T. H., Malanga, M. T., J. Macromol. Sci. Pure Appl. Chem. A, 34, 1921 (1997). 39. Murata, M., et al., Macromol. Chem. Phys., 202, 1799 (2001).
This page intentionally left blank
18
Characterization, Properties and Applications of Syndiotactic Polystyrene KOMEI YAMASAKI Plastics Technical Center, Idemitsu Petrochemical Co., Ltd., Chiba, Japan
NORIO TOMOTSU Polymer Research Laboratory, Idemitsu Petrochemical Co., Ltd., Chiba, Japan
MICHAEL MALANGA R&D Engineering Plastics, The Dow Chemical Company, Midland, Ml, USA
1
INTRODUCTION
Highly syndiotactic polystyrene (SPS) was synthesized using a homogeneous catalytic system using a titanium compound and methylaluminoxane or borate [1]. The detailed syndiospecific polymerization of styrene is described in the previous chapter. Dow Chemical and Idemitsu Petrochemical produce SPS, which is commercialized the trade names Questra® and XAREC®, respectively. SPS is very different from conventional amorphous polystyrene (atactic polystyrene, APS) in chemical and thermal properties. The high stereoregularity allows SPS to obtain a high level of crystallinity, resulting in a high melting temperature (270 °C), fast crystallization rate [2-5] and high solvent resistance. With these characteristics given by crystallization, SPS has a low specific gravity, excellent electrical properties, hydrolytic resistance and good moldability, which were inherited from APS (Figure 18.1). The excellent balances of mechanical, electrical, solvent resistance, heat resistance and dimensional stability properties combined with a relatively low price have led to the common use of this material in engineering plastics. Modern Styrenic Polymers: Polystyrene and Styrenic Copolymers. Edited by J. Scheirs and D. B. Priddy | 2003 John Wiley & Sons Ltd
390
K. YAMASAKI ETAL
Crystallization
New Engineering Plastics SPS Additional Characteristics
Commodity Plastics
• Heat Resistance • Chemical Resistance • Dimensional Stability
Conventional PS
Original Characteristics + Low Specific Gravity ^ Electrical Properties • Hydrolytic Stability • Moldability
Figure 18.1 Additional characteristics brought by crystallization of polystyrene
2 2.1
CHARACTERIZATION STRUCTURE
Syndiotactic polystyrene is produced by well stereo-controlled coordination polymerization by titanium compounds. The polymer has a high percentage of rrrr pentad structure. The 13C NMR chemical shift for the phenyl-1 carbon and backbone methylene carbon are approximately 145.3 and 44.9 ppm, respectively [1]. In general, these polymers are found to be >99% pure in syndiotactic structure as defined by NMR. The most stereoselective catalysts produce a >99.6% pure syndiotactic structure (Figure 18.2)
2.2
CRYSTAL FORM
SPS takes two different conformations in its crystal, TT and TTGG conformations, depending on the crystallization conditions, and exhibits a very complex polymorphic structure [6-14]. One is the TT conformation which appears when SPS crystallizes from the melt. Two different transcrystals are reported when SPS takes the TT conformation in the crystal, a-form [15-19] and ß-form. The identity period is 5.06 A (Figure 18.3) in both crystal forms. The ß-form is more thermally stable than the a-form. Figure 18.4 shows the detailed structure of the ß-form [6,10]. SPS takes the TTGG conformation when it crystallizes in the presence of organic solvents. SPS takes the TTGG conformation in -y-form and 8-form crystals. The 8-form is a complex of SPS and solvent [20–24]. Figure 18.5 shows the complex of SPS and toluene.
391
SYNDIOTACTIC POLYSTYRENE
[rrrrr] >99.6%
416
45
44
4':
Spinning side band PPM
46.0
45.5
45.0
44.5
44.0
43.5
43.0
Figure 18.2 13C NMR of syndiotactic polystyrene produced by highly stereocontrolled catalyst system. Apparatus: JEOL Lambda 500 (13C: 125.65MHz). Frequency: 25 000 Hz. pulse: Q.OJJLS (45° pulse). Repetition time: 4s. Scans: 10000
Figure 18.3 Conformation of a-form of SPS
392
K. YAMASAKI ETAL
a = 8.81 A
Figure 18.4
13/4
13/4
1/4
1/4
p-Form of SPS
On thermal treatment, the -form is transformed into the -/-form and then into the a-form. Tsutsui et al. examined in detail solvent removal from the 8-form and found that the solvent was removed without a change of the conformation of SPS [25].
3
PHYSICAL PROPERTIES
3.1
THERMAL PROPERTIES
By programmable differential scanning calorimetry (DSC), the glass transition temperature (7"g) of SPS is 100 °C. When amorphous SPS is exposed to heat, it crystallizes above the Tg and a crystallization exotherm maximum is observed at about 150°C, then a melting endotherm peak is observed at 270 °C. For a series of SPS samples with different degrees of crystallinity, the heat of fusion was determined as 53.2 J/g. Gvozdic and Meier tried stepwise annealing of SPS in a DSC instrument and found that the melting temperature increased from ca 270 to 287.3 °C [26]. By a modification of the annealing procedure, they attained a higher melting temperature of 291.5 CC [27].
393
SYNDIOTACTIC POLYSTYRENE b = 13.26 A
2b
a = 17.58 A
y= 121.2°
c = 7.71 A
Figure 18.5 8-Form of SPS: solvent-induced crystallized form with incorporation of toluene in the lattice
3.2
CRYSTALLIZATION BEHAVIOR OF SPS
The crystallization rate is dependent on the crystallization temperature, T, and molecular weight of the polymer. It is important to study the effect of T and molecular weight on the crystallization nature of SPS. Takebe et al. [28] studied the effect of temperature and molecular weight on the crystallization rate of SPS by DSC. When SPS was melted at 300 °C, then rapidly cooled to the crystallization temperature, T, the evolution of crystallization showed a sigmoidal curve with reference to crystallization time (Figure 18.6). The crystallization rate becomes slower as T approaches close to melting point. When the crystallization rate of SPS is analyzed based on Avrami theory, the Avrami index, n, is equal to 3, which suggests that crystallization of SPS proceeds via three-dimensional heterogeneous growth [28,29]. Figure 18.7 shows the time evolution of the crystallinity of SPS of various weight-average molecular weight, Mw. The crystallization rate becomes slower as Mw becomes larger at a fixed T.
394
K. YAMASAKI ETAL 100 80
; 60 ' 40 20
OIL
2
345678
0.1
2
345678
1
2
345
10
Time (tnin) Figure 18.6 Time evolution of degree of crystallinity of SPS isothermally crystallized at various temperatures. Mw = 400 000; Mw/Mn = 2.4; Xct = degree of crystallinity at time t; X£° = degree of crystallinity at infinite time
1.0 0.8
0.4
0.2
0.0
56789
2 1
3 4 5 6 7 8 9 10
2
3
4 5 6 7 89 100
Time (min) Figure 18.7 Time evolution of degree of crystallinity of SPS of various Mw. Isothermal crystallization temperature T = 245 °C. Numbers on curves represent Mw of SPS
From these results, the effect of T and Mw on the crystallization rate constant, k:1/3, can be analyzed according to the Lauritzen-Hoffman nucleation theory [30,31], and the results are shown in Figure 18.8.
SYNDIOTACTIC POLYSTYRENE
395
5 8
4
1.4
(a)
1.6
1.8
T ; 4 •• -;.;->; 4
k•
*...
r= 290°C
4W-: J.42 x 10 5 M V
k1n=2.47
/^
6.C13 x K)5 7.1 7 x K)5
O
L ;i
..;.;.;.;; 4
2.53 2.39
i i iiiiiii i i i i i iiil -i i i i-iiili i i i iiiiu 1 100 102 10
i i iiiJ 104
waTaM ( Figure 18.11 Superposed viscosity curve of SPS. Shear viscosities measured for several Mw and temperatures were superposed with reference to temperature (reference temperature 290 °C) and Mw (reference Mw 289000). aT = shift factor for temperature; aM — shift factor for Mw 105
104
102
10' 105
•I---1---VT-
4
5 6 7 89
106
Figure 18.12 Relationship between zero-shear viscosity n0 and Mw. »;0 is proportional to
399
SYNDIOTACTIC POLYSTYRENE
temperature (Figure 18.13), the activation energy to flow was calculated to be 21 kcal/mol for SPS. This value is almost equivalent to that of APS (25 kcal/ mol) [33] and IPS (20 kcal/mol) [34].
3.6
MECHANICAL PROPERTIES OF NEAT SPS
SPS has high a melting point almost comparable to those of other engineering plastics (Table 18.2). In comparison with other crystalline engineering thermoplastics, of particular interest is that SPS has the lowest specific gravity and dielectric constant. In a practical fabrication method, such as injection molding, the mold temperature affects the degree of crystallinity, Xc, of SPS and the morphology [35]. When neat SPS is molded at high mold temperatures, a well developed lamellar structure is observed, and macroscopically spherulites are observed (Figure 18.14), and Xc stays constant across the whole cross-section of the molded part (Figure 18.15). In contrast, when SPS is molded at low mold temperatures far below the Tg, crystallization of SPS is restricted at the surface owing to fast cooling, resulting in a lower Xc and a poorly developed crystalline lamellar structure at the surface. However, in the center of the molded parts, spherulites are formed because the slower cooling allows the crystallization of SPS. The Xc profile across the cross-section is very different from what was molded at high mold 0.30 0.20 0.10 0.00
-0.10 -0.20
-0.30 -0.40 -0.50 -0.60 .700
1.750
1.800
1/TxlO3 (K- 1 )
Figure 18.13 Shift factor, aj, plotted against inverse of temperature. The activation energy to flow was calculated to be 21 kcal/mol for SPS
K. YAMASAKI ETAL
400
Table 18.2
Comparison of mechanical properties of neat polymers Units SPS
Property Specific gravity Melting point Glass transition temperature Flexural strength Flexural modulus Notched Izod impact DTUL(1.82MPa) Vicat softening temperature Dielectric constant (23 °C, 1 MHz)
Figure 18.14 (b)50°C
kg/m
3
°C
°c MPa MPa
kJ/m2
°C
°c
GPPS PBT
PA6
PA66 PPS
1040
1040
1310
1140
1140
1340
270 100 75
— 100 65
224 30 80
224 45 100
260 70 110
285 92 95
3000
2900
2400
2600
2800
3800
2..5 5.4 4.4 2.2 4.4 2.0 138 80 64 89 60 96 104 254 270 250 215 215 3.4 3.5 3.2 3. 1 2.6 2.6
Morphology of injection molded SPS. Mold temperature: (a) 160 and
1UU.U
90.0
_
90.0 80.0 -
80.0 70.0 60.0
>• 70.0
3
| 60.0
O
50.0
X
40.0
U
40.0
30.0 -
^
30.0 20.0
20.0 10.0 n ft
1
500
(a)
1
1
1000 1500 Distance (|im)
1
1
O
50.0
10.0 An
\
r
° i
i
i
i
i 3000
(b)
Distance (um)
Figure 18.15 Profile of degree of crystallinity of injection molded SPS in the crosssection. Mold temperature: (a) 160 and (b) 50 0C
401
SYNDIOTACTIC POLYSTYRENE
temperatures; at the surface of the molded part, Xc is low and it gradually increases towards the center of the part. A sufficiently high mold temperature is, therefore required in order to attain a high degree of crystallinity of SPS. The degree of crystallinity is important as long as the heat resistance of crystalline polymer is concerned. Figure 18.16 shows the dependence of the dynamic elastic moduli (E' and E") on temperature. The dynamic storage modulus, E', exhibits a strong dependence on Xc; for high Xc, E' drops at the TB, then it shows a gradual decline near the melting point, whereas for low Xc, E' shows a steep drop above Tg. Thus the degree of crystallinity is very critical to the heat resistance of SPS.
SPS has heat resistance and chemical resistance in addition to the inherent characteristics of conventional polystyrene. Of significant interest, moreover, is that SPS is cost-competitive because it is synthesized from styrene monomer, a well established and widely available raw material. Although SPS has a brittle nature like APS and is not suitable for use alone for structural material, reinforcement with glass fiber or impact modification by elastomers improves the mechanical properties of SPS. 104
101, 1000-
50
100
150
200
250
300
Temperature (°C) Figure 18.16 Temperature dependence of elastic modulus of SPS with various degrees of crystallinity, Xc: +, 58%; O, 46%; •, 10%
402
4.1
K. YAMASAKI ETAL
MECHANICAL AND FLOW PROPERTIES
The mechanical properties of glass fiber-reinforced materials are compared in Table 18.3. Glass fiber-reinforced SPS (GFSPS) has mechanical properties competitive with those of other engineering thermoplastics. Highlights of GFSPS are low specific gravity and a high heat distortion temperature. Moreover, GFSPS has good flowability; almost comparable spiral flow length with liquid crystal polymer (LCP) which is a representative material in connector applications where thin wall flowability is required (Figure 18.17).
4.2
ELECTRICAL PROPERTIES
GFSPS offers a low dielectric constant and low dissipation factor compared with other thermoplastic engineering plastics (Figure 18.18). Also, the low dissipation factor of GFSPS remains almost constant over a wide frequency range up to 10 GHz (Figure 18.19). GFSPS possesses a higher tracking resistance, CTI Class 2 to Class 0 (Figure 18.20), which suggests further reliable electrical properties of GFSPS. These electrical properties of GFSPS, along with heat resistance and flowability, allow GFSPS to be competitive with existing plastics in electrical applications such as various connectors, antennae, and other electrical devices. Table 18.3 polymers
Comparison of mechanical properties of 30% glass fiber-reinforced
Property Specific gravity Water absorption Mold shrinkage (MD) Tensile strength Tensile elongation Flexural strength Flexural modulus Notched Izod impact DTUL(1.82MPa) DTUL (0.45 MPa) CLTE (MD) Dielectric constant (1 MHz) Dissipation factor (1 MHz) Breakdown voltage 40%GF.
Units
GF-SPS
3
kg/m 1250 0.05 % 0.35 % MPa 118 2.5 % MPa 185 MPa 9000 11 kJ/m2 C 251 C 269 °C 2.5 x!05/°C 2.9 20 wt%, in contrast with the miscibility found for aPS/PVME blends. In fact, DSC experiments show two Tg values corresponding to an sPS-rich phase (83:17 wt%) and a PVMErich phase (13:87 wt%). The lack of miscibility is also confirmed by the absence,
446
L. ABIS ETAL.
13
in C NMR solid-state experiments, of cross-polarization from sPS protons to PVME methoxy carbons [28]. Later, Mandal and Woo [29] demonstrated that this system is miscible, and exhibits a behavior equivalent to aPS/PVME blends. The previously found immiscibility is due to the relatively low value of the lower critical solution temperature, which in 50:50 wt% blends induces a phase separation already at temperatures of ca 120°C. In fact, OM, SEM and DSC, applied to blends (70:30 and 50:50 wt%) prepared by casting films from 1-2% solutions of chloronaphthalene at about 120°C, or by precipitation from the same solution with «-heptane, show a substantial homogeneity. However, OM measurements, performed at various temperatures on a series of samples, show a cloud point at ca 120 °C and above, indicating the onset of segregation. At higher temperature (samples briefly treated at 300 °C and then quenched), DSC shows two Tgs at -30 °C (attributed to PVME) and at 95 °C (attributed to sPS), shifted with respect to the pure compounds and corresponding to two partially miscible phases, one rich in PVME and the other rich in sPS. Under slow cooling the process appears to be reversible.
4.1.2.2
Phase structure and morphology
According to Cimmino et al. [25], OM measurements show that in the melt (340 °C) demixing of PVME occurs at all compositions, and during crystallization segregated PVME particles remain occluded in the interspherulitic region of sPS crystals. At the end of the crystallization process, an increase in the spherulite dimensions is observed, indicating that PVME addition reduces the nucleation ability of sPS. Measurements carried out at different crystallization temperatures (Figure 20.3) show that, at the same Tci, G increases in the PVME blend with respect to neat sPS and it is concentration independent. This behavior is accounted for by considering that a lower energy is needed to transport the molecules in the melt, the Tg of the blend being lower than in neat sPS. The independence of concentration seems, according to the authors, to arise from the phase separation which keeps the composition of the sPS-rich phase constant.
4.1.3 4.7.3.7
Other Blends sPS/TMPC
Investigations carried out by DSC [30] on a series of sPS/TMPC blends show a single Tg transition at all compositions, as expected for a miscible blend. The addition of TMPC generally retards the crystallization of sPS, and at contents
POLYMERIC BLENDS BASED ON SYNDIOTACTIC POLYSTYRENE
447
greater than 40 wt% the blends become amorphous. On samples crystallized isothermally at various temperatures, Tm values and implicitly lamellae thickness increase with increase in sPS content. The equilibrium values of Tm, calculated with the Hoffman-Weeks method, give an interaction parameter between polymers of —0.92 J/cm3.
4.7.3.2
sPS/S-DPE
Gausepohl et al. [31] investigated the behavior of blends between sPS and random styrene-l,l-diphenylethylene copolymers obtained by anionic synthesis. The blends were miscible for copolymer contents of 1,1-diphenylethylene lower than 15 wt% as indicated by the occurrence of a single Tg (114°C). Tm and crystallization rate were not influenced. 4.2
IMMISCIBLE BLENDS
As discussed in the first part, blends containing immiscible components such as polyolefins could improve the performances of the inherently brittle sPS. Until now the reported investigations have concerned simple binary blends containing a polyolefin and sometimes SEBS as a compatibilizer. In addition, sPS/ polyurethane and sPS/sulfonated sPS blends were also investigated. All these studies tried to correlate the microscopic features of the blends with their mechanical properties.
4.2.1
sPS/Polyolefins
Abis and co-workers [32,33] reported on a multi-technique characterization of immiscible blends between sPS and several polyolefins, prepared in a mixer at 285 °C and thermo-compressed at the same temperature. SEM performed on the cryo-fractured surface of the blends shows a gross phase separation, while the presence of voids due to the easy pull-off of particles from cryo-fractured surfaces points to a lack of adhesion between components (Figure 20.4a). The particles of polyolefins present in 80:20 wt% blends have dimensions decreasing in the following order: PP « LLDPE > HDPE > EPR > PIB. Coherently, as expected for immiscible blends, Tg values measured by DSC show very small variations with respect to the pure components while the mechanical properties degrade with respect to neat sPS. In particular, for minimum polyolefin contents 40 wt% the elongation values remain lower than 1 %, whereas the tensile strength and the energy to break decrease. Only for higher contents does HOPE become the matrix and a gradual recovery of mechanical parameters, in particular of the elongation at break, is observed. A similar study performed on aPS/HDPE/SEBS blends [33,36] showed an improvement in both the morphological and mechanical properties of the blend. This different behavior is tentatively explained in Figure 20.9 where phase transitions occurring when the blend is cooled from the melt state to room temperature are sketched. When sPS and aPS blends are at 285 °C, each end-block constituting SEBS (PS and EB) interacts and penetrates into the chemically similar blend component (sPS, aPS and HOPE, respectively), by forming a mixed interface with extended
POLYMERIC BLENDS BASED ON SYNDIOTACTIC
POLYSTYRENE
453
2CL,
O
(a)
-60 -160 -140 -120 -100 -8 7TQ
-40
-20
0
2.5-
2.0-
1.5
(b)
2 3 sPS/HDPE, wt.ratio
Figure 20.8 (a) Real part of the dynamic shear modulus G' plotted against temperature for sPS/HDPE/SEBS at different compositions (wt%): (1) 0:90:10; (2) 18:72:10; (3) 36:54:10; (4) 54:36:10; (5) 72:18:10; (6) 90:0:10. (b) Real part of the dynamic shear modulus G' measured at -140°C for sPS/HDPE (solid symbols) and sPS/ HDPE/SEBS (open symbols) plotted against sPS/HDPE weight ratio. Reprinted from Ref. 33 by permission of Wiley-VCH
interchain entanglements. As the aPS blend is cooled from the meltdown to about 100°C, SEBS polystyrene blocks and aPS vitrify together, thus maintaining the entanglements developed during the melt mixing. At almost the same temperature very fast crystallization of HDPE occurs which immobilizes the SEBS aliphatic blocks within the polyethylene semicrystalline structure. Both effects result in a strengthening of the interfacial adhesion between the blend components. In the sPS blends the same phenomena are likely to occur at 100°C. However, when 225 °C is reached from the melt, the polystyrene end-blocks of SEBS disentangle from the fraction of sPS chains which crystallize at the interface.
454
L. ABIS ETAL ATACTIC PS SEES interface
aPS bulk
SYNDIOTACTIC PS HOPE bulk
melt mixing I 285°C|
sPS bulk
SEBS interface
melt
3
HOPE bulk
1
sPS I crystallization
|225°C
liqf,
melt
melt crystalline
liquid
HDPE crystallization 100°c I
V V I N glassy
aPSl vitrification
amorphous (liquid)
IsPS vitrification
' J crystalline
crystalline
amorphous (liquid)
amorphous (glassy)
.
liq..
V
J crystalline + amorphous (liquid)
Figure 20.9 Liquid/melt to solid transitions of sPS/HDPE/SEBS (right) and of aSP/ HDPE/SEBS blends (left), occurring from 285 °C down to room temperature. Reprinted from Ref. 33 by permission of Wiley-VCH
This process seems to be extended enough to produce a significant weakening of the interface bonding on the sPS side. The load applied in the tensile test to aPS/HDPE/SEBS blends generates in the aPS phase a stress which is efficiently transferred to HDPE dispersed particles through the thick interface and delays the brittle failure of aPS phase to higher elongational values. In the case of sPS/ HDPE/SEBS blends, the poor interfacial adhesion inhibits the stress transfer to the HDPE phase and causes the formation and propagation of cracks throughout the sPS domains which lead to brittle failure of the sample. However, the kind of technique used to mold these blends seems to be important in determining their mechanical properties. In fact, mechanical tests carried on injection molded samples [37] show, with respect to compression molded samples, a significant enhancement of the energy to break for all samples (Tables 20.2 and 20.3). Moreover the addition of 10 wt% of SEBS to an 80:20 wt% sPS/HDPE blend involves in the injection molded samples an increase in both the energy at break and the Izod impact strength, whereas in the thermo-compressed samples no improvement is observed. Differences between compression and injection molded samples are widely acknowledged [38] and
455
POLYMERIC BLENDS BASED ON SYNDIOTACTIC POLYSTYRENE
Table 20.2 Tensile tests on compression molded sPS/HDPE and sPS/HDPE/SEBS blends: Young's modulus (E); strength (trb), elongation (ej,) and energy at break (Eb); elongation at yield (ey). Reprinted from Ref. 33 by permission of Wiley-VCH sPS (wt%)
HDPE
SEBS
(wt%)
(wt%)
E (GPa)
Ob
1 1 1 1 60 700
3.52 2.62 1.81 1.48 1.24 0.91
34 24 14 16
5 5 5 5 5 5
2.95 2.42 1.80 2.46 0.98 0.72
44.7 23.3 15.3 9.7 17.4 15.2
90 72 56 36 18 0
0 18 34 54 72 90
10 10 10 10 10 10
2.49 1.89 1.45 1.06 0.80 0.50
29.4 13.1 11.1 15.7 17.1 12.9
aPS 95 76
HDPE
SEBS
E 2.75 2.16
16.5 mol% styrene. Further aspects of the viscoelastic behavior of ESIs which have been reported to date include linear stress relaxation behavior of amorphous ESI [40] and the creep behavior of amorphous ESI in the glass transition region [41]. Chen et al. [42] 0.25
-100
Figure 26.4 Dynamic mechanical spectroscopy (DMS) plot of land versus temperature, measured at 0.1 1, 10, 100rad/s, for ES7
ETHYLENE-STYRENE COPOLYMERS
613
also reported the large strain-stress relaxation and strain recovery of ESIs at temperatures above Tg, and found that the observed behavior could be rationalized in terms of various network models.
3.2
MECHANICAL PROPERTIES
The tensile stress-strain behavior (23 °C, S.Vmin"1) for ESIs, as depicted in Figure 26.5, generally exhibit large strain at rupture, and have been found to show uniform deformation behavior [22]. Tensile stress-strain behavior for ESIs differing in styrene content and molecular weight has been reported. The deformation response of the semicrystalline materials is predominantly controlled by the level of crystallinity and the connectivity between the crystalline and amorphous phases. The large-scale deformation behavior and recovery behavior of semicrystalline ESI have been studied by Chang et al. [43,44] as a function of temperature, comonomer content and crystallinity and compared with the behavior of ethylene-octene (EO) copolymers. Chen et al, [45] provided an in-depth comparison of the morphological structure and properties of ESIs and EO copolymers and confirmed that aspects of deformation which depended on crystallinity, such as yielding and cold drawing, were determined primarily by comonomer content for both sets of copolymers. Clearly a major factor which is different between the ESI and EO copolymers is the respective location of the Tg and the influence of chain mobility on the mechanical response observed.
100
200
300
400 500 Strain, %
600
700
800
Figure 26.5 Engineering stress-strain curves measured at 23 °C. RT denotes room temperature
614
Y. W. CHEUNG AND M. J. GUEST
The materials in the mid-styrene region (40–65 % S copolymers) are characterized by low modulus, and they show some elastomeric characteristics. Information on the deformation and recovery behavior of selected ESIs was reported by Mudrich et al. [46], with an ES45 copolymer of particular note in terms of strain recovery after deformation. The effects of styrene become dominant in the high-styrene region where the modulus and yield stress are seen to increase. ESIs having Tg above ambient temperature show some characteristics of glassy materials. Additionally, high-styrene (>70% S) ESI exhibit interesting shape-reshape functionality [47]. Good low-temperature toughness of ESI has been evident from impact testing and low-temperature tensile testing. As discussed above, DMS provides some evidence that there are sub-Tg chain motions which may contribute to energy dissipation and toughness. In addition, Chen et al. [40] provided estimates of the entanglement molecular weights (Me) for ESI. The low A/e values found suggest a high entanglement density in these polymers, and this is considered to contribute to the ability of the polymers to shear yield at temperatures below T"g rather than undergo brittle fracture.
3.3
MELT RHEOLOGY AND PROCESSABILITY
The melt rheology and associated information from solid-state DMS, melt strength and pressure-volume-temperature (PVT) property of ESIs have been reported by Karjala et al. [48]. The polymers were shown to have (1) good thermal stability at processing temperatures, (2) viscosities which are a strong function of styrene content in addition to molecular weight and (3) excellent processability, particularly regarding shear thinning characteristics at high shear rates and good melt strength. Melt Theological master curves could be generated via time-temperature superposition. In addition to providing fundamental structure-property understanding, melt rheology has further found utility in the design of polymer blends based on ESI [49]. Viscosity-shear rate dependence as a function of styrene content and molecular weight was analyzed with a Cross model [50], with the results for the zero shear viscosity 0/0), critical shear stress (T*), relaxation time (A.) and powerlaw index («) shown in Table 26.1. An example of a fit of the data to this model is shown in Figure 26.6. As can be seen from Table 26.1, at equivalent melt index and with increasing styrene level, the zero shear viscosities decrease, relaxation times increase and the ESI becomes more shear sensitive. Table 26.1 also contains an average relaxation time estimated from the inverse of the frequency at which the storage and loss modulus cross. These values are lower than those based on the Cross model, but show the same general trend of relaxation time decreasing with increasing melt index and with increases in styrene level at a given melt index/molecular weight.
615
ETHYLENE-STYRENE COPOLYMERS
Table 26.1 Cross model parameters: ^0 is the zero shear viscosity, T* critical shear stress, A relaxation time and n power exponent, where ES20 represents 20 wt% styrene ESI and MI denotes the melt index Material
fo (P)
ES20-0.1MI ES20-0.5MI ES20-3MI ES20-11MI ES60-0.1MI ES60-0.5MI ES60-3M1 ES60-10MI
1.63 3.63 4.10 9.33 4.20 1.48 2.98 8.21
x x x x x x x x
T* (dyn/cm2)
106 105 104 103 105 105 104 103
1.31 9.94 2.67 6.83 2.71 1.73 1.88 1.62
x x x x x x x x
105 104 105 10s 106 106 106 106
/(s)
/ cross-over (s)
n
12.4579 3.6561 0.1537 0.0137 0.1550 0.0860 0.0159 0.0051
1.2716 0.0042 0.0006 ND 0.0882 0.0250 0.0040 0.0012
0.423 0.491 0.569 0.550 0.140 0.206 0.214 0.238
The melt strength of the ESIs was measured at several temperatures and was shown to improve dramatically with both increasing styrene content and changes in temperature. Figure 26.7 shows the effect of styrene level at a constant melt index of 0.1. At equivalent melt index, the melt strength increases from approximately 12 to 25 cN as the styrene level increases from ~ 20 to ~ 60wt%. By decreasing the temperature, the melt strength of ESI is also dramatically increased, in part relating to its relatively high activation energy. Activation energies of melt strength are similar to those of viscous activation energy, of the order of 15 kcal/mol.
Reference Temp. = 190°C
1000
ES20-0.1MI ES20-0.5MI ES20-3MI ES20-11MI ES60-0.1MI ES60-0.5MI
0.01
100 104 co aT Frequency (rad/s)
Figure 26.6 indices (Ml)
108
Melt rheology master curve for ES20 and ES60 at various melt
616
Y. W. CHEUNG AND M. J. GUEST T=190°C;MI=0.1
50 100 Velocity (mm/s)
Figure 26.7
150
Melt strength of ESI measured at 190 CC
As a general comment, melt processability of interpolymers is favorable towards most fabrication techniques. These characteristics have allowed the fabrication of ESI articles by a wide range of standard melt processing techniques, including injection molding, film fabrication, blow molding operations and melt extrusion. The potential for ESI to be utilized in calendering operations has been described by Karjala et al. [51]. Selected ESIs were found to demonstrate the requisite Theological properties and thermal stability to be successfully calendered, and this was supported by commercial-scale validation.
4
MATERIALS ENGINEERING ASPECTS
Although most polymers are in fact modified to a greater or lesser degree to optimize their utility in end-use applications, certain basic characteristics of these ethylene-styrene interpolymers make them particularly likely to be modified, or to be used as modifiers. These characteristics are as follows: 1. polymer chain microstructure, aromatic/olefin functionality and inherent compatibility with a wide range of other polymers, fillers and low molecular weight materials, including bitumens [52], plasticizers [53], tackifiers [54] and processing aids; 2. the location of the glass transition temperature (Tg), in the range -20 to +35 °C, and the associated major changes in, e.g., modulus for relatively small changes in copolymer styrene content or temperature; 3. excellent processability.
ETHYLENE-STYRENE COPOLYMERS
617
This section provides more details on selected aspects of blend systems using interpolymers as components, filler composites and terpolymers. A related aspect which warrants mention in relation to multicomponent systems is the interfacial nature and behavior of these interpolymers. Ronesi [55] presented a study of the interfacial adhesion between LDPE and ESI, analyzing the significant effects of ESI copolymer styrene content, layer thickness and test temperature.
4.1
INTERPOLYMER BLENDS
Primarily because of the olefinic and styrenic functionality, ESI generally exhibit good compatibility with a wide range of polymers. Studies on ESI blends with atactic polystyrene show that significant toughening of brittle polystyrene can be achieved with selected ESI, in large part due to good compatibility between these polymers [56]. The olefinic nature of ESI helps to provide compatibility with olefinic polymers and copolymers, including polyethylenes, ethylene-a-olefin copolymers and polypropylene homopolymers and copolymers [57]. As reported by Diehl et al. [58], interpolymers are also compatible with a broader range of polymers, including styrene block copolymers [59], poly(vinyl chloride) (PVC)-based polymers [60], poly(phenylene ethers) [61] and olefinic polymers such as ethylene-acrylic acid copolymer, ethylene-vinyl acetate copolymer and chlorinated polyethylene. Owing to their unique molecular structure, specific ESI have been demonstrated as effective blend compatibilizers for polystyrene-polyethylene blends [62,63]. The development of the miscibility/ compatibility behavior of ESI-ESI blends differing in styrene content will be highlighted below.
4.2
BLENDS OF ETHYLENE-STYRENE INTERPOLYMERS: MISCIBILITY CONSIDERATIONS
Based on the pioneering work of Molau [64], it is evident that phase separation can occur in blends of two or more copolymers produced from the same monomers when the composition difference between the blend components exceeds some critical value. The mean field theory for random copolymercopolymers blends has been applied to ES-ES blends differing in styrene content to determine the miscibility behavior of blends [65,66]. On the basis of the solubility parameter difference between PS and PE, it was predicted that the critical comonomer difference in styrene content at which phase separation occurs is about 10 wt% S for ESI with molecular weight around 105. DMS plots for ES73 and ES66 copolymers and their 1:1 blend are presented in Figure 26.8.
618
Y. W. CHEUNG AND M. J. GUEST
10
20
30
40
50
60
Figure 26.8 DMS of ES73-ES66 (50:50 weight ratio) blend showing Tg intermediate to the pure component 7gs
The blend, having a 7 wt% styrene difference between blend components, can be considered miscible, as indicated by the presence of a single Tg intermediate of the two pure copolymers. The transition width of the 7"g for the blend is almost identical with those of the pure components, which further supports singlephase behavior in the blend. In accordance with copolymer model prediction, a difference in styrene content of greater than about 10wt% (for a 105 MW) between two ES copolymers is sufficient to drive phase separation. This criterion is further supported by the DMS plot shown in Figure 26.9, in which two distinct Tgs are clearly evident for an ES73-ES58 blend (1:1 weight ratio and having a 15wt% styrene difference between blend components), indicating blend immiscibility. Figure 26.10 shows two distinct Tgs for a blend containing five components for which the styrene content difference between successive components lies close to or below the critical composition difference for phase separation. On the basis of the Tg behavior, it may be rationalized that the five components have phase separated into two phases in which the ES66 and ES73 form one miscible phase and ES58, ES52 and ES46 form another miscible phase. It is interesting that the peak amplitude of the blend is much smaller than that of the pure blend components. These findings clearly indicate that the glass transition process, including peak temperature, width and amplitude, can be controlled by combining ESI with different compositions and hence different degrees of compatibility. Such ESI-ESI blends can therefore be designed to be immiscible in nature, and offer the opportunity to engineer materials with broadened Tg loss processes and enhancements in relaxation behavior, mechanical properties and melt rheology and processability.
619
ETHYLENE-STYRENE COPOLYMERS
-20
-10
0
10
20
30
40
50
60
Figure 26.9 DMS of ES73-ES58 (50:50 weight ratio) blend showing two distinct Tgs indicative of phase separation
20%ES73/20%ES66/ 20%ES58/20%ES52/ 20%ES46
T, °C
Figure 26.10 DMS of five-component blend which phase separates into two distinct phases in which ES73 and ES66 form a miscible phase and ES58, ES52 and ES46 form another miscible phase Chen et al. [67,68] further extended the study of binary blends of ESI over the full range of copolymer styrene contents for both amorphous and semicrystalline blend components. The transition from miscible to immiscible blend behavior and the determination of upper critical solution temperature (UCST) for blends could be uniquely evaluated by atomic force microscopy (AFM) techniques via the small but significant modulus differences between the respective ESI used as blend components. The effects of molecular weight and molecular weight distribution on blend miscibility were also described.
620
Y. W. CHEUNG AND M. J. GUEST
Extending the above rationale regarding miscibility to blends of ESI with styrenic polymers, it is evident that ESI having less than 80 wt% styrene content will be immiscible with polystyrene, unless the molecular weights of the respective polymers are very low. Similarly, ESI with more than 10wt% styrene content will be immiscible with polyethylene.
4.3
FILLER COMPOSITES
It is well known that the addition of fillers to polymers can enhance the stiffness, dimensional stability, upper service temperature, tensile strength, damping characteristics and ignition resistance, in addition to lowering the cost of fabricated parts [69]. Additionally, many durable applications of ESI can be pursued by modifying the balance of properties through the addition of inorganic fillers. The phenyl functionality of ESI is postulated to contribute to compatibility with a wide range of fillers via possible polymer-filler interactions. ESI has been reported to have good filler acceptance with a broad range of inorganic fillers, including calcium carbonate, barium sulfate, alumina trihydrate (ATH) and magnesium hydroxide (MgOH2) [70,71]. The resulting composite materials generally exhibit very good mechanical properties, even for relatively high loadings of fillers. Many combinations of filler types and ESIs with different copolymer styrene contents show exceptional tensile elongations at rupture. Impact testing of ESI-filler systems has further shown that these composites have good toughness. The mechanical properties, and particularly failure performance, indicate that good interfacial bonding exists between ESI and fillers. Figure 26. 1 1 shows the stress-strain behavior of ES30 filled with various levels of ATH. It can be seen that the yield stress increases with increasing level of ATH while the ultimate elongation is in excess of several hundred percent even for materials with more than 50wt% ATH. The modulus of composite materials can be modeled by the generalized Kerner equation:
where M\ is the modulus of the polymer, M the modulus of the composite, k^ the Einstein coefficient, m the maximum volume fraction, 2 the volume fraction of filler and A = kt — 1 • It can been from Figure 26. 1 2 that the relative flex modulus of ES30 filled with ATH and MgOH2 can be reasonably described
621
ETHYLENE-STYRENE COPOLYMERS
Figure 26.11
Engineering stress-strain curves (23 °C) for ATH-filled ES30
20
15
10
10
20
30
40
50
VOLUME % FILLER
Figure 26.12 Relative flex modulus of ATH- and MgOH2-filled ES30. The solid line represents the Kerner model [Equation (1)] prediction
by the Kerner model up to about 30 vol.% filler. The deviation from the Kerner model observed at higher filler levels could be related to agglomeration of fillers and/or the onset of percolation. Interestingly, the modulus of MgOH2-filled ES30 is much higher than that of the ATH-filled material when the filler level exceeds 40 vol.%.
622
Y. W. CHEUNG AND M. J. GUEST
In addition to the modulus prediction, the ultimate properties including elongation and ultimate tensile strength, assuming good adhesion between filler and polymer, can be modeled by the following equations: (3)
(4)
where e is the elongation of the composite, £Q the elongation of the polymer, (f>2 the volume fraction of filler and a the ultimate tensile strength. The relative elongation and ultimate tensile strength for ATH/MgOH2-filled ES30 are depicted in Figures 26.13 and 26.14, respectively. Both the elongation and ultimate tensile strength of the composites generally show positive deviations from the model prediction. Interestingly, the ATH-filled materials generally show higher elongation than that of the MgOFb-filled materials. At 50 vol.% filler loading, this difference vanishes and the elongation for both materials is lower than that predicted by the model. In the case of the ultimate tensile strength, a positive deviation from the model prediction is always observed. This analysis provides additional support that good interfacial adhesion is found in most ESI-filler composites The filled ESI show rheological characteristics suggesting that the compositions can be easily fabricated into final parts. Filled ESI compositions are
0.8
0.6
0.4
0.2
10
20 30 40 VOLUME % FILLER
50
60
Figure 26.13 Relative elongation of ATH- and MgOH2-filled ES30. The dotted line represents the model [Equation (3)] prediction
623
ETHYLENE-STYRENE COPOLYMERS
0.8
0.6
"...
f
0.2
10
20
30
40
50
60
VOLUME % FILLER Figure 26.14 Relative tensile strength of ATM- and MgOH2-filled ES30. The dotted line represents the model [Equation (4)] prediction
expected to find broad utility in applications including, for example, wire and cable [72], injection molded articles, film and sheet structures, profile extrusions and flooring systems.
4.4
TERPOLYMERS
INSITE™ Technology also permits the copolymerization of ethylene and styrene with additional monomers [73], including dienes, higher a-olefins and norbornene. Basic structure-property relationships of ESP terpolymers have recently been published [74]. There are limited, mainly patent, references to the preparation and basic characterization of terpolymers based on ethylene, styrene and propylene (ESP). The use of propylene as a comonomer in combination with ethylene and styrene results in a much broader polymer design space for materials development. Depending on the comonomer composition of the terpolymers, these materials exhibit solid-state microstructures ranging from semicrystalline through to essentially amorphous materials. The terpolymers are most differentiated from ESI when there is measurable crystallinity in the solid-state microstructure, in part because the propylene comonomer introduces methyl groups on the polymer chain which modify both the crystalline and amorphous phases of the solid-state microstructure [75]. Figure 26.15 shows the variation in crystallinty of ESP as a function of total comonomer (styrene and propylene). The crystallinity-composition dependence for ESI is also plotted for comparison. The incorporation of monomer
624
Y. W. CHEUNG AND M. J. GUEST
units such as styrene and propylene into a linear ethylene chain is generally accepted as introducing defects which suppress and inhibit the crystallization of —(CH2)W—sequences. For ESI produced with the constrained geometry catalysts of current interest, no detectable crystallinity is typically found for those copolymers containing more than about 18mol% styrene comonomer, as depicted in Figure 26.15. Owing to steric hindrance, the phenyl group introduced into the chain microstructure by the styrene comonomer is excluded from the crystalline regions. For EP copolymers, the critical propylene content above which crystallization is inhibited occurs at about 30mol%. This difference has been attributed to the partial inclusion of the methyl group of propylene into the ethylene crystalline lattice. Similarly to the EP copolymers, ESP terpolymers with more than 20mol% comonomer still exhibit a measurable level of crystallinity. The glass transition temperature, as measured by the tan S maximum from DMS, for the ESP terpolymers is presented in Figure 26.16. It is interesting that the Tg of the terpolymers is not only dependent on the total comonomer content but also strongly on the propylene/styrene ratio. For a given molar percentage of comonomer, Tg generally decreases with increasing propylene/ styrene ratio, primarily because propylene has a much lower Tg than styrene. In addition to perturbing the crystalline phase, propylene also profoundly affects the amorphous phase, as evidenced by the depression in TB of the semicrystalline ESP. An approach to modeling the Tg behavior of the ESP terpolymers has been introduced, with predictions from the model and experimental data showing satisfactory agreement, despite the simplifying assumptions that were made. 15
(P/S)=0.7 £>
(P/S)=4.5
EP
10
U
ESI
10
12
\
14 16 18 Mol% Comonomer
20
22
Figure 26.15 Crystallinity of ESI, EP and ESP. (P/S) denotes the propylene to styrene molar ratio
ETHYLENE-STYRENE COPOLYMERS
625
30
'(P/S)=0.25 20 • 0.29 10
0.4
0
'0.5
Semi-crystalline
0.9
-10 ^
-20'
(P/S)=0.7 n -7 A* A u. / A
-30
Amorphous
^-
1.6
- •^ "•
• 4.0
*4.5
EP -40 10
20
30
40
50
60
Mol% Comonomer Figure 26.16 Glass transition temperature, measured from the tan300%) when propylene is used as a comonomer.
5
ATTRIBUTES AND APPLICATIONS
Ethylene-styrene interpolymers exhibit a novel balance of properties that are uniquely different from polyethylenes and polystyrenes. In contrast to other ethylene-a-olefin copolymers, ESI display a broad range of material response ranging from semicrystalline, through elastomeric to amorphous. The styrenic functionality and unique molecular architecture of ESI are postulated to be the basis of the versatile material attributes such as processability (shear thinning, melt strength and thermal stability), viscoelastic properties, low-temperature toughness and broad compatibility with other polymers, fillers and low molecular weight materials.
626
Y. W. CHEUNG AND M. J. GUEST 40
ESP-14,5 ES20
/
30
/ 20
/
/
/ .'
/
.'
.' ESP-7,11
10
EP15
0
2
3
4
5
6
Strain Figure 26.17 Engineering stress—strain curves, measured at 23 CC, for ESI, EP and ESP. Compositions are given in wt% comonomer (S and P)
The current range of potential markets and applications for ESI which have been identified now includes paintable injection molded toys [76], wire and cable, footwear, automotive, bitumen modification, packaging, injection and blow molded articles, adhesives, building and construction. Although intermaterial substitutions in existing applications based on thermoplastic elastomers such as styrene—butadiene block copolymers, flexible PVC and other ethylene-a-olefin copolymers, including ethylene-vinyl acetate, are potential application areas for ESI, the novel combination of material attributes suggests that new application areas will emerge. Of particular note to date are developments using ESI-based materials in foam applications. Novel foam structures offer attractive properties and characteristics including softness, esthetics and drape for a wide range of thermoplastic and crosslinked foam applications. Other product technologies of interest are as injection molded structural foams, as foamed layers in multilayer structures and as foamed blends of interpolymers with styrenic and olefinic polymers [77–79]. Interpolymers also have potential for co-extruded film and sheet applications.
6
SUMMARY
Ethylene-styrene interpolymers are a novel class of polymers exhibiting a unique combination of material attributes that are not found in polyethylenes,
ETHYLENE—STYRENE COPOLYMERS
627
polystyrenes or their blends. The effective production of these novel copolymers has been enabled by INSITE™ Technology. INDEXTM interpolymers were introduced by The Dow Chemical Company in December 1998, and this technology includes ethylene-styrene interpolymers. A product development plant to produce Interpolymers (Sarnia, Canada) had a successful start-up in September 1999. This plant, having a nameplate production capacity of around 22 500t (5 x 107lb) per annum, is currently allowing further product and process developments and application validation. These interpolymers based on ethylene and styrene are an integral part of an exciting new generic class of materials, offering unique opportunities for innovative developments in basic polymer chemistry, catalyst and process development, materials science and engineering and application technology.
7
ACKNOWLEDGMENTS
The authors particularly wish to thank Steve Chum, Scott Mudrich and Teresa Karjala for helpful comments and discussions. They further thank researchers at Case Western Reserve University, including Professor Eric Baer, Professor Anne Hiltner, Hong-Yu Chen and Andy Chang, for their contributions to the understanding of structure-property relationships and material classification. The authors also thank many others, especially Joe Huang and Ken Reichek, for their help with providing the data presented in this chapter. Finally, The Dow Chemical Company and the Interpolymer Business Management team are thanked for permission to publish this work.
REFERENCES 1. European Patent Application EP 416 815 Bl; The Dow Chemical Company; Stevens, J.C., Timmers, F.J., Wilson, D.R., Schmidt, G.F., Nickias, P.N., Rosen, R.K., Knight, G.W. and Lai, S. 2. US Patent 5 703 187; The Dow Chemical Company; Timmers, F.J. 3. Stevens, J.C., in Abstracts of the Metcon, Houston, Tx, p. 157 (1993). 4. Guest, M.J., Cheung, Y.W., Diehl. C.F. and Hoenig, S.M. 'Structure, properties and applications of ethylene/styrene interpolymers', in Metallocene-based Polyolefins: Preparation, Properties and Technology, Vol. 2, ed. J. Scheirs and W. Kaminski, Wiley, Chichestes, Chapt. 12, pp. 271–292, (1999). 5. US Patent 3117945; Union Carbide Corporation; Gorham, W.F. and Farnham, A.G. 6. Kobayashi, S. and Nishioka, A., J. Polym. Sci, Part A, 2, 3009 (1964). Trademark of The Dow Chemical Company.
628
Y. W. CHEUNG AND M. J. GUEST
7. Kawai, W. and Katsuta, S., J. Polym. Sci., Part A, 8, 2421 (1970). 8. Soga, K. and Lee, D., Polym. Bull., 20, 237 (1988). 9. US Patent 4748209; Nippon Petrochemicals Inc.; Orikasa, Y. Kojima, S., Inoue, T., Yamamoto, K., Sato, A. and Kawakami, S. 10. US Patent 5652315; Mitsui Toatsu Chemicals, Inc.; Inoue, N., Tetsunosuke, S., Masahiro, K. 11. Longo, P., Grassi, A., and Oliva, L., Makromol. Chem., 191, 2387 (1990). 12. Pellechia, C. and Olivia, L., Rubber Chem. Technol., 72, 553 (1999). 13. Sernetz, F.G., Mulhaupt, R., Amor, F., Eberle, T. and Okuda, J. Polym. Sci. Part A: Polym. Chem., 35, 1571 (1997). 14. Sernetz, F.G., Mulhaupt, R., Fokken, S. and Okuda, J., Macromolecules, 30, 1562 (1997). 15. Sernetz, F.G. and Mulhaupt, R., Macromol. Chem. Phys., 197, 1071 (1996). 16. Thomann, Y., Sernetz, F.G., Thomann, R., Kressler, J. and Mulhaupt, R., Macromol. Chem. Phys., 198, 739 (1997). 17. Sernetz, F.G. and Mulhaupt, R., J. Polym. Sci., 35, 2549 (1997). 18. Sernetz, F.G., Mulhaupt, R. and Waymouth, R.M., Polym. Bull., 38, 141 (1997). 19. Suzuki, T., Tsuji, Y., Watanabe, Y. and Takegami, Y., Macromolecules, 13, 849 (1980). 20. US Patent 5 043 408; Sumitomo Chemical Company; Kakugo, M., Miyatake, T. and Mizunuma, K. 21. Miyatake, T., Mizunuma, K. and Kakugo, M., Makromol. Chem., Macromol. Symp., 66, 203 (1993). 22. Izzo, L., Oliva, L., Proto, A. and Trofa, M., Macromol. Chem. Phys., 200, 1086 (1999). 23. Arai, T., Ohtsu, T. and Suzuki, S., Polym. Prepr. Am. Chem. Soc. Div. Polym. Chem., 349, 38 (1997). 24. Arai, T., Ohtsu, T. and Suzuki, S., Macromol. Rapid Commun., 19, 327 (1998). 25. US Patent 5883213; Denki Kagaku KK; Arai, T., Nakamura, A., Suzuki, S., Otsu, T. and Okamoto, A. 26. Pasztor, A.J., 'Styrene polymers', in Encyclopedia of Polymer Science, 2nd edn, J Wiley, Chichester, Vol. 16, pp. 114–117 (1989). 27. 'Special Feature I. Ethylene-Styrene Interpolymers: Technologies, Products and Markets'. The Metallocene and Single Site Catalyst Monitor', The Catalyst Group, Spring House, PA (December 2000). 28. Chung, T.C. and Lu, H.L., J.Polym. Sci. A: Polym. Chem., 35, 575 (1997). 29. Lu, H.L., Hong, S. and Chung, T.C., Macromolecules, 31, 2028 (1998). 30. Chung, T.C. and Lu, H.L., and Hong, S., Rubber Chem. Technol., 72, 283 (1999). 31. Hou, Z., Tezuka, H., Zhang, Y., Yamazaki, H. and Wakatsuki, Y., Macromolecules, 31, 8650(1998). 32. Cheung, Y.W. and Guest, M.J., in Proceedings of SPE ANTEC 54th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 1634 (1996). 33. Chen, H., Guest, M.J., Chum, P.S., Hiltner, A. and Baer, E., J. Appl. Polym. Sci., 70, 109 (1998). 34. Guest, M.J., Cheung, Y.W. and Martin, J.M., in Proceedings of ACS Division of Polymeric Materials: Science and Engineering, Fall Meeting, New Orleans, LA, August 22-26, Vol. 81, pp. 371-372 (1999). 35. McCrum, N.G., Read, B.E. and William, G., Anelastic and Dielectric Effects in Polymeric Solids, J Wiley, New York (1967). 36. Flocke, H.A., Kolloid Z., 180. 118 (1962). 37. Boyd, R.H., Polymer, 26, 323 (1985).
ETHYLENE-STYRENE COPOLYMERS
629
38. Khanna, Y., Turi, E.A., Taylor, T.J., Vickroy, V.V. and Abbott, R.F., MacromoleculeslZ, 1302(1985). 39. Lobbrecht, A., Friedrich, C. Sernetz, F.G. and Mulhaupt, R., J. Appl. Polym. Sci., 65 209 (1997). 40. Chen, H.Y., Stepanov, E.V., Chum, S.P., Hiltner, A. and Baer, E., Macromolecules, 330, 8870 (2000). 41. Chen, H.Y., Stepanov, E.V., Chum, S.P., Hiltner, A. and Baer, E., J. Polym. Sci. B, Polym. Phys., 37, 2373 (1999). 42. Chen, H.Y., Stepanov, E.V., Chum, S.P., Hiltner, A. and Baer, E., Macromolecules, 32, 7587 (1999). 43. Chang, A., Stepanov, E.V., Guest. M., Chum, S., Hiltner, A. and Baer, E., in Proceedings of SPE ANTEC 56th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 1803 (1998). 44. Chang, A., Stepanov, E.V., Chum, S.P., Hiltner, A. and Baer, E., in Proceedings of SPE ANTEC 57th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 2121 (1999). 45. Chen, H.Y., Chum, S.P., Hiltner, A. and Baer, E., J. Polym. Sci. B, Polym. Phys., 39, 1578 (2001). 46. Mudrich, S.F., Cheung, Y.W. and Guest, M.J., in Proceedings of SPE ANTEC 55th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 1783 (1997). 47. US Patent 6156842; The Dow Chemical Company; Hoenig, S.M., Turley, R.R., Cheung, Y.W., Guest, M.J., Diehl, C.F., Stewart, K.B. and Sneddon, J. 48. Karjala, T.P., Cheung, Y.W. and Guest, M.J., in Proceedings of SPE ANTEC 55th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 1086 (1997). 49. Karjala, T.P., Hoenig, S.M., Guest, M.J., Cheung, Y.W., Finlayson, M.P., Walther, B.W. and Montanye, J.R., in Proceedings of SPE ANTEC 58th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 1833 (2000). 50. Karjala, T.P., Cheung, Y.W. and Guest, M.J., in Proceedings of SPE ANTEC 57th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 2127 (1999). 51. Karjala, T.P., Walther, B.W., Hill, A.S. and Wevers, R., in Proceedings of SPE ANTEC 57th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 2139(1999). 52. US Patent 6 107374; The Dow Chemical Company; Stevens, J.C., Fimmers, F.J., Gatzke, A.I., Bredeweg, C.J., McKay, K.W., Gros, W.A. and Diehl, C.F. 53. US Patent 5739200; The Dow Chemical Company; Cheung, Y.W., Gathers, J.J., Guest, M.J. and Bethea, J.R. 54. European Patent Application EP 0923619 Al: The Dow Chemical Company; Parikh, D., Guest, M.J. and Speth, D.R. 55. Ronesi, V.M., Polym. Mater. Sci. Eng., 84, 440 (2001). 56. Park, C.P. and Clingerman, G.P., in Proceedings of SPE ANTEC 57th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 2134 (1999). 57. US Patent 6 184294 Bl; The Dow Chemical Company; Park, C.P., Thoen, J., Broos, R., Guest, M.J., Cheung, Y.W., Chaudhary, B.I., Gathers, J.J. and Hood, L.S. 58. Diehl, C.F., Guest, M.J., Chaudhary, B.I., Chueng, Y.W., Van Volkenburgh, W.R. and Walther, B.W., in Proceedings of SPE ANTEC 57th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 2149 (1999). 59. US Patent 5741 857; The Dow Chemical Company; Esnault, C.P. and Edmonson, M.S. 60. US Patent 6 136923; The Dow Chemical Company; Cheung, Y.W. and Guest, M.J. 61. US Patent 6201067; The Dow Chemical Company; Cheung, Y.W., Guest, M.J., Chum, P.S. and Kao, C.I.
630
Y. W. CHEUNG AND M. J. GUEST
62. US Patent 5460818; The Dow Chemical Company; Park, C.P., Clingerman, G.P., Timmers, F.J., Stevens, J.C. and Henton, D.E. 63. Park, C.P. and Clingerman, G.P., in Proceedings of SPE ANTEC 54th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 1887 (1996). 64. Molau, G.E., Polym. Lett., 3, 1007 (1965). 65. Cheung, Y.W. and Guest, M.J., J. Polym. Sci. B, Polym. Phys., 38, 2976 (2000). 66. Chen, H.Y., Cheung, Y.W., Chum, S.P., Hiltner, A. and Baer, E., in Proceedings of SPE ANTEC 58th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 1828 (2000). 67. Chen, H.Y., Cheung, Y.W., Hiltner, A. and Baer, E., Polymer, 42, 7819 (2001). 68. Chen, H.Y., Chum, P.S., Hiltner, A. and Baer, E., Macromolecules, 34,4033 (2001). 69. Nielsen, L.E. and Landel, R.F., Mechanical Properties of Polymers and Composites, Marcel Dekker, New York (1994). 70. Guest, M.J., Cheung, Y.W., Betso, S.R. and Karjala, T.P., in Proceedings of SPE ANTEC 58th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 2116(1999). 71. US Patent 5973049; The Dow Chemical Company; Bieser, J.O., Cheung, Y.W., Guest, M.J., Thoen, J.A. and Gathers, J.J. 72. Betso, S.R., Guest, M.J., Remenar, R.M., Field, A.W. and Keen, F.E., in Proceedings of SPE ANTEC 58th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 3012 (2000). 73. US Patent 5872201; The Dow Chemical Company; Cheung, Y.W., Guest, M.J., Timmers, F.J. and Hahn, S.F. 74. Guest, M.J., Cheung, Y.W., Karjala, T.P., Ruiz, J.M. and Kolthammer, B.W.S., in Proceedings of SPE ANTEC 59th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 3012 (2001). 75. Alamo, R.G. and Mandelkern, L., Thermochimi. Ada, 238, 155 (1994). 76. Chang, D. and Cheung, Y.W., in Proceedings of SPE ANTEC 59th Annual Conference, Society of Plastics Engineers, Brookfield, CT, p. 1704 (2001). 77. US Patent 6048909; The Dow Chemical Company; Chaudhary, B.I., Barry, R.P. and Cirihal, S.C. 78. US Patent 5993707; The Dow Chemical Company; Chaudhary, B.I., Hood, L.S., Barry, R.P. and Park, C.P. 79. Chaudhary, B.I., Barry, R.P. and Tusim, M.H., J. Cell. Plast., 36, 397 (2000).
Properties of Styrenic Polymers
This page intentionally left blank
27
Fracture Behavior of High Impact Polystyrene and Acrylonitrile-ButadieneStyrene T. VU-KHANH Universite de Sherbrooke, Sherbrooke, Quebec, Canada
1
INTRODUCTION
One of the most important weaknesses of polystyrene is its poor impact resistance. To improve this performance, much attention has been paid to adding an elastomeric, dispersed phase to the polymer matrix. The role of rubber toughening in the impact performance of polymers has been extensively investigated. Most plastics can be made tougher by the addition of a small amount of rubbery material, dispersed as second-phase particles on a microscopic scale. The improvements in toughness can be seen in several commercialized systems based on styrenic polymers that are brittle in their unmodified state. Thus, the glassy, brittle polystyrene (PS) is transformed into high-impact polystyrene (HIPS), used successfully in applications such as refrigerator linings, packaging, vacuum cleaners, fans or even shoe heels. Similarly, styrene-acrylonitrile (SAN) has been modified into the tough acrylonitrile-butadiene-styrene (ABS) and is utilized in applications such as telephone sets, luggage, computer casings and interior car fittings. These materials possess good impact properties and have been widely used as engineering materials. In rubber toughened plastics, the rubber particles constitute the dispersed phase in the polymer matrix. Their essential role is to act as a stress concentrator. When the toughened material is Modern Styrenic Polymers: Polystyrene and Styrenic Copolymers. Edited by J. Scheirs and D. B. Priddy f 2003 John Wiley & Sons Ltd
634
T. VU-KHANH
subjected to a uniaxial stress, the localized stress experienced by the matrix in the immediate vicinity of a rubber particle is magnified by the local stress concentration effect. The matrix will yield locally in response to this localized stress field, thus avoiding a global brittle catastrophic failure of the material. The concentration of stress that initiates the local yielding in the matrix is just the first step in a complex process; if the overall applied stress is increased further the matrix then continues to deform by shear yielding or crazing. Shear yielding involves the creation of bands of highly oriented, stretched material at 45c to the direction of the applied stress. Crazes are cracks, spanned by elongated fibrils of the material which can carry load and thus maintain the structural integrity of the material. Dilatometric studies are often used to distinguish shear yielding, which is essentially a constant-volume process, from the dilatational processes of crazing or voiding. These tests are carried out under tension since a compressive or pure shear state of stress does not allow crazing to occur. Measurements of volume changes in rubber-modified plastics, in tensile creep under constant load have been performed to elucidate the mechanisms of local, irreversible deformation in toughened polymers. HIPS has been shown to exhibit the greatest volume change because the PS matrix forms crazes in preference to shear bands. ABS, however, shows a transition from shear yielding to crazing or large-scale voiding when the deformation level becomes important. In addition to the multiplication of energy dissipation by local plastic deformation, the rubber particles can also deform, stretching to bridge an expanding crack opening. The fracture and cavitation of the rubber particles also relieve the triaxial stress at the crack tip. These mechanisms constitute the basic toughening process. When the temperature and/or loading rate change, the fracture mechanisms become more complex. At relatively low temperatures or high loading rates, HIPS and ABS can become brittle. Like most polymers, the fracture behavior of HIPS and ABS is also strongly time and temperature dependent. Depending on the loading rate and temperature, the fracture mode and performance can be very different. HIPS and ABS can break in a brittle, semi-ductile or ductile manner. Brittle fracture usually results in the shattering phenomenon of the part. In this case, the elastic energy stored in the part is much higher than the energy dissipated in the fracture process. The excess of energy is transformed into kinetic energy and transferred to the debris so that it can fly away with a very high velocity. In large structures, the kinetic energy can assist the crack to propagate without external loading. In terms of safety, this fracture behavior is a real concern. Conversely, ductile fracture generally occurs under stable crack propagation, with more plastic deformation. In this case, after initiation, the cracks can only propagate with additional supply of energy by external loads so that the character of the failure is less catastrophic. It is therefore essential to use a quantitative method to characterize the fracture performance of HIPS and ABS. In fact, the field of fracture of materials has received much attention and research effort after several catastrophic failures of major structures, usually
FRACTURE BEHAVIOR OF HIPS AND ABS
635
made of high-strength metallic materials. The investigations carried out during the two decades following the Second World War have clearly demonstrated that brittle, catastrophic fracture is generally initiated by defects that exist in most materials and structures. This has led to the development of fracture mechanics that has brought significant progress in the understanding of failure of materials. Using the theories in fracture mechanics (developed mainly for metals), quantitative methods have been proposed for the characterization of the impact performance of polymers [1–5]. Most of the work in this area was published in the 1970s. The proposed methods provide a new interpretation of the results of the common Charpy and Izod tests, but using samples containing a sharp initial crack instead in order to simulate the presence of a defect in the material. This chapter presents recent results on quantitative analyses of the fracture behavior of HIPS and ABS. The emphasis is put on the prediction of fracture performance over a wide range of deformation rates and temperatures. Different approaches currently used to characterize fracture performance are applied to HIPS and ABS and recommendations on the use of these methods are presented. The fundamental mechanisms controlling the dependence of fracture performance and yield on deformation rate and temperature are discussed.
2
QUANTITATIVE CHARACTERIZATION OF FRACTURE
It has been shown that fracture is a very complex process and the fracture performance depends on both the initiation and the propagation of a defect [6-10] in the material. Under impact, most polymers break in very distinct manners. Several types of fracture have been identified depending on the amount of plastic deformation at the crack tip and the stability of crack propagation. For each type, an appropriate analysis has been developed to determine the impact fracture energy of the material. These methods have also been verified in various plastics [11,12]. The different fracture behaviors in most polymers are illustrated in Figure 27.1, which shows a schematic drawing of the load-deflection diagram of Charpy tests on HIPS [13] under an impact velocity of 2m/s at various temperatures. With increasing temperature, the fracture mode changes from brittle to semiductile at about –50°C and then becomes ductile at temperatures higher than –30 °C.
2.1
BRITTLE FRACTURE
Brittle fracture occurs when the strain energy stored in the sample up to the point of fracture is much larger than the energy dissipated in the creation of the
636
T. VU-KHANH Load T