Polyolefin Blends

POLYOLEFIN BLENDS POLYOLEFIN BLENDS Edited by Domasius Nwabunma 3M Company Thein Kyu University of Akron WILEY-INTE...

Author: Nwabunma D. | Kyu Th.

141 downloads 1982 Views 21MB Size Report

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

Report copyright / DMCA form

DOWNLOAD PDF

POLYOLEFIN BLENDS

POLYOLEFIN BLENDS Edited by Domasius Nwabunma 3M Company

Thein Kyu University of Akron

WILEY-INTERSCIENCE A JOHN WILEY & SONS, INC., PUBLICATION

Cover credit: (Third image on top right side) Reprinted from European Polymer Journal, vol. 40, Smit, G. Radonjic and D. Hlavata. Phase morphology of iPP/aPS/SEP blends, page 1439, 2004. With permission from Elsevier. Copyright ß 2008 by John Wiley & Sons, Inc. All rights reserved Published by John Wiley & Sons, Inc., Hoboken, New Jersey Published simultaneously in Canada 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 as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permission. Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com. Wiley Bicentennial Logo: Richard J. Pacifico Library of Congress Cataloging-in-Publication Data: Nwabunma, Domasius. Polyolefin blends / edited by Domasius Nwabunma, Thein Kyu. p. cm. Includes index. ISBN 978-0-471-79058-7 (cloth) 1. Polyolefins. I. Kyu, Thein, 1948- II. Title. TP1180.P67N928 2007 668.40 234–dc22 2007021318 Printed in the United States of America 10 9 8 7 6 5 4 3 2 1

Contents

Preface

xv

Contributors

Part I

Introduction

1. Overview of Polyolefin Blends 1.1 Introduction 1.2 Olefinic Monomers 1.3 Polyolefin Homopolymers, Copolymers, and Terpolymers 1.4 Polyolefin Blends 1.5 Trends in Polyolefin Blends Nomenclature References

2. Miscibility and Characteristics of Polyolefin Blends 2.1 Introduction 2.1.1 Polyolefins 2.1.2 Blends 2.2 Polymer Blend Miscibility 2.3 Interfaces in Liquid and Polymer Mixtures 2.4 Polyolefin–Polyolefin Blends 2.4.1 Blends between Polyethylenes 2.4.2 Blends between Isotactic Polypropylene and Ethylene Propylene Copolymers 2.4.3 Blends between iPP and High Comonomer Concentration Polyethylene Copolymers 2.4.4 Blends between iPP and PB1 2.5 Binary Immiscible Blends 2.5.1 Polyolefin–Polystyrene Blends 2.5.2 Polyolefin–Polyamide Blends 2.6 Ternary Blends of Polyolefins with Other Polymers and Compatibilizing Agents 2.6.1 Surfactants and Compatibilizing Agents 2.6.2 Polyolefin–Polystyrene Blends with Compatibilizing Agents

xvii

1 3 3 4 5 7 13 16 18 27 27 27 29 30 33 36 36 38 39 40 42 43 43 44 44 45

v

vi

Contents 2.6.3 Polyolefin–Polyamide Blends with Compatibilizing Agents 2.7 Conclusions Nomenclature References

Part II

Polyolefin/Polyolefin Blends

3. Miscibility, Morphology, and Properties of Polyethylene Blends 3.1 Introduction 3.2 Structure and Properties of Polyethylenes 3.3 Applications of Polyethylene Blends 3.4 Molar Mass and Branching Distributions 3.5 Crystallization, Melting, and Branching of Polyethylenes 3.6 Miscibility and Crystallization 3.7 Theoretical Prediction of Miscibility 3.8 Rheology of Melted Polyethylene Blends 3.9 Mechanical Properties of Polyethylene Blends 3.10 Additives 3.11 Conclusions Nomenclature References

4. Miscibility and Crystallization Behavior in Binary Polyethylene Blends 4.1 4.2

Introduction Miscibility 4.2.1 Linear and Short Branched Polyethylene Blends 4.2.2 Blends of Linear and Long Branched Polyethylenes 4.2.3 Blends of Short and Long Branched Polyethylenes 4.3 Crystallization Behaviors 4.3.1 Blends of Linear and Short Branched Polyethylenes 4.3.2 Blends of Linear and Long Branched Polyethylenes 4.3.3 Blends of Short and Long Branched Polyethylenes 4.4 Conclusions Nomenclature References

5. Microscopically Viewed Structural Characteristics of Polyethylene Blends Between Deuterated and Hydrogenated Species: Cocrystallization and Phase Separation 5.1 5.2 5.3

Introduction Cocrystallization and Phase Separation of PE Blends Aggregation Structure of Chains in Lamella

48 50 50 51

57 59 59 64 65 66 68 72 74 76 77 80 80 81 82

84 84 86 86 88 89 90 90 92 93 93 94 94

97 97 98 101

Contents 5.4 Crystallization Behavior of D/H Blend Samples 5.4.1 Crystallization in the Cooling Process from the Melt 5.4.2 Isothermal Crystallization Process 5.4.3 Blending Effect on Crystallization Rate 5.5 Mixing Behavior of D and H Components 5.6 Conclusions Acknowledgments Nomenclature References

6. Thermal and Structural Characterization of Binary and Ternary Blends Based on Isotactic Polypropylene, Isotactic Poly (1-Butene) and Hydrogenated Oligo (Cyclopentadiene) 6.1 Introduction 6.2 Binary Blends 6.2.1 Blend Preparation 6.2.2 Glass Transition Temperature 6.2.3 Morphology and Spherulite Growth Rate 6.2.4 Isothermal Bulk Crystallization Kinetics 6.2.5 Temperature Dependence of the Spherulite Growth Rate and the Overall Kinetic Rate Constant 6.2.6 Melting Behavior 6.2.7 Polymorphism and Phase Transformation of Poly (1-Butene)/ Hydrogenated Oligo (Cyclopentadiene) 6.2.8 Supermolecular Structure of Isotactic Polypropylene/ Hydrogenated Oligo (Cyclopentadiene) Blends 6.3 Ternary Blends 6.3.1 Blends Preparation 6.3.2 Morphology and Spherulite Growth Rate 6.3.3 Glass Transition Temperature 6.3.4 Nonisothermal Crystallization and Melting Behavior 6.3.5 Isothermal Bulk Crystallization Kinetics of Isotactic Polypropylene Component 6.3.6 Melting Behavior of the Isotactic Polypropylene Component 6.3.7 Supermolecular Structure 6.4 Conclusions Nomenclature References

7. Morphological Phase Diagrams of Blends of Polypropylene Isomers with Poly(Ethylene–Octene) Copolymer 7.1 Introduction 7.2 Blends of sPP/POE 7.2.1 Thermal Characterization and Morphological Phase Diagrams: Undulated Lamella, Sheaf, and Spherulite 7.2.2 Growth of Single Crystals: Length, Width, and Periodicity

vii 105 106 108 113 114 117 118 118 119

121 121 123 123 123 124 125 126 131 133 136 141 141 141 143 143 145 146 147 153 154 155

157 157 159 161 165

viii

Contents 7.2.3 Phase Field Modeling for a Single-Component System: Sectorization and Ripple Formation in sPP 7.3 Blends of iPP/POE 7.3.1 Morphology Development in Relation to Phase Diagrams 7.3.2 Sectorization in iPP/POE Blends 7.3.3 Crystal Growth Dynamics in Binary Blends of iPP and aPP 7.4 Blends of ePP/POE 7.4.1 Characterization of Neat Elastomeric Polypropylene 7.4.2 Melting Transitions and Morphology Phase Diagrams of ePP/POE Blends 7.5 Conclusions Nomenclature References

8. Structure, Morphology, and Mechanical Properties of Polyolefin-Based Elastomers 8.1 Introduction 8.2 Thermoplastic Polyolefin Elastomers 8.2.1 Reactor Blends of PP, PE, and EPR: Impact Copolymer PP 8.2.2 Postreactor Blends of PP–EPR and ICP–EPR 8.3 Thermoplastic Vulcanized Elastomers 8.3.1 Dynamic Vulcanization and Morphology 8.3.2 Origin of Rubber Elasticity 8.3.3 Several Factors that Influence Mechanical Properties 8.4 Polyolefin Copolymers, Blends, and Composites 8.4.1 Polyolefin Copolymers 8.4.2 Blends and Composites 8.5 Conclusion Acknowledgment Nomenclature References

9. Morphology and Mechanical Properties in Blends of Polypropylene and Polyolefin-Based Copolymers 9.1 Introduction 9.2 Morphology and Dynamic Mechanical Properties 9.2.1 Blends with Polyethylene or Poly(butene-1) 9.2.2 Blends with Ethylene–a-Olefin Copolymers 9.2.3 Ethylene–Isotactic Propylene Copolymers 9.3 Tensile and Rheo-Optical Properties 9.3.1 Principles for Rheo-Optical Characterization 9.3.2 Blends with Ethylene–a-Olefin Copolymers 9.3.3 Blends with Novel Ethylene–Isotactic Propylene Copolymers 9.4 Solidification Process and Final Morphology 9.4.1 Morphology Formation During Crystallization 9.4.2 Structure and Properties of Injection-Molded Products

172 177 177 181 182 187 188 188 193 195 196

198 198 199 199 201 206 206 208 211 214 214 219 221 222 222 222

224 224 225 225 226 236 241 241 242 247 250 250 257

Contents 9.5 Conclusions Nomenclature References

10. Functionalization of Olefinic Polymer and Copolymer Blends in the Melt 10.1 10.2

Introduction Scope of Review 10.2.1 Free-Radical Grafting of Unsaturated Monomers to PO Chains 10.2.2 The Use of Monomers and Initiators 10.3 Functionalization of PP/PE Blends 10.3.1 Effect of Reacting Blend Formulation on Grafting Efficiency and Rheological and High Elastic Properties of Melt of Functionalized PP/PE Blends 10.3.2 Structure and Mechanical Properties of Functionalized PP/PE Blends 10.4 Functionalization of PP/EPR Blends 10.5 Functionalization Features of Blends: PE/EPR, PP(PE)/EOC, and PP(PE)/Styrene Polymer 10.6 Use of Functionalized Polyolefin Blends 10.7 Conclusion Nomenclature References

11. Deformation Behavior of b-Crystalline Phase Polypropylene and Its Rubber-Modified Blends 11.1 11.2

Introduction Deformation Characteristics 11.2.1 Static Tensile Behavior 11.2.2 Strain-Induced b ! a Phase Transition 11.2.3 Impact Behavior 11.3 Fracture Toughness 11.3.1 General Aspects 11.3.2 Mode I LEFM Approach 11.3.3 Impact Fracture Toughness 11.3.4 Essential Work of Fracture 11.4 Conclusions Nomenclature References

12. Multiphase Polypropylene Copolymer Blends 12.1

Introduction 12.1.1 Commercial Production 12.1.2 Morphology of Commercial Impact PP Copolymers

ix 264 265 265

269 269 270 270 275 284

284 291 295 297 299 300 301 302

305 305 309 310 314 323 330 330 331 339 341 347 347 348

351 351 352 352

x

Contents 12.2 12.3 12.4

Dispersive Mixing during Processing Molecular Structure of Impact PP Copolymers Coarsening in Multiphase PP Copolymer Systems 12.4.1 Background 12.4.2 Coarsening of High Impact Polypropylene 12.4.3 Coarsening of Model Blends 12.4.4 Interfacial Effects in Polypropylene Copolymer Systems 12.5 Conclusions Nomenclature References

13. Heterogeneous Materials Based on Polypropylene 13.1 13.2 13.3

Introduction The Interphase: Definition Magnitude Orders in the Interphase 13.3.1 The Dispersed Phase 13.3.2 The Matrix 13.3.3 The Interphase: Designing the Interface 13.4 Interfacial Modification of Heterogeneous Materials Based on Polypropylene 13.4.1 Composites: When the Dispersed Phase is Rigid 13.4.2 Blends: When the Dispersed Phase is Flexible 13.4.3 The Role of the Interfacial Modifiers from the Matrix Side 13.5 Interfacial Modifiers Based on Polypropylene 13.5.1 The Kinetic Approach: Basic Aspects 13.5.2 Chemical Modification of Polypropylenes by Grafting of Polar Monomers 13.6 Conclusions Acknowledgment Nomenclature References

14. Polypropylene/Ethylene–Propylene–Diene Terpolymer Blends 14.1 14.2

14.3

Introduction PP/EPDM Blends 14.2.1 Toughness and Crystallization Behaviors of PP/EPDM Blends 14.2.2 Compatibilization of PP/EPDM Blends 14.2.3 Ternary Blends and Composites from PP/EPDM Blends 14.2.4 Application of Radiation Dynamically Vulcanized PP/EPDM Blends (or Thermoplastic Vulcanizates (TPVs)) 14.3.1 Effect of Cross-linking on the Properties of PP/EPDM TPVs

357 360 360 360 365 368 370 373 376 377

379 379 380 380 382 383 383 385 387 387 388 397 397 398 407 408 408 408

411 411 412 412 414 416 417 419 420

Contents 14.3.2 Microstructure of PP/EPDM TPV 14.3.3 PP/EPDM/Ionomer TPVs 14.3.4 Mechanical and Rheological Properties 14.4 Applications of PP/EPDM Blends 14.5 Conclusions Acknowledgments Nomenclature References

15. Ethylene–Propylene–Diene Rubber/Natural Rubber Blends 15.1 15.2

Introduction Miscibility, Compatibility, and Thermodynamics of Polymer Blending 15.3 Blend Preparation 15.4 Covulcanization 15.5 Filler Distribution in NR/EPDM Blends 15.6 Morphology of NR/EPDM Blends 15.7 Compatibilization of NR/EPDM Blends 15.8 Mechanical and Viscoelastic Properties 15.8.1 Mechanical Properties 15.8.2 Dynamic Mechanical Properties 15.9 Rheological Properties 15.10 Thermal Properties 15.11 Electrical Properties 15.12 Aging Properties 15.12.1 Thermal Aging 15.12.2 Ozone Resistance 15.13 Transport Properties 15.14 Applications 15.15 Conclusions Nomenclature References

16. Phase Field Approach to Thermodynamics and Dynamics of Phase Separation and Crystallization of Polypropylene Isomers and Ethylene–Propylene–Diene Terpolymer Blends 16.1 16.2

16.3

Introduction Experimental Phase Diagrams 16.2.1 Cloud Point Phase Diagram of iPP/EPDM Blends 16.2.2 Cloud Point Phase Diagram of sPP/EPDM Blends Thermodynamic Free Energy Description of Crystalline Polymer Blends 16.3.1 Flory–Huggins Free Energy of Amorphous–Amorphous Blends

xi 423 425 428 436 437 438 438 439 441 441 442 443 444 447 448 450 452 452 456 458 460 461 462 462 463 465 466 469 469 470

473 473 475 475 476 478 478

xii

Contents 16.3.2

Extension of the FH Theory to Crystal–Amorphous Blends 16.3.3 Prediction of Phase Diagram Topologies 16.3.4 Comparison with Experimental Phase Diagrams of PP/EPDM Blends 16.4 Phase Field Modeling on Polymer Phase Transitions 16.4.1 Theory on Phase Separation Dynamics and Morphology Evolution 16.4.2 Dynamics of Crystal Growth in a Phase Separating System: iPP/EPDM Blends 16.4.3 Dynamics of Crystal Growth in a Phase Separating System: sPP/EPDM Blends 16.5 Conclusions Nomenclature References

Part III

479 481 484 486 486 488 491 496 496 497

Polyolefin/Nonpolyolefin Blends

499

17. Compatibilization and Crystallization of Blends of Polyolefins with a Semiflexible Liquid Crystalline Polymer

501

17.1

Blends of Polyolefins (High Density Polyethylene and Isotactic Polypropylene) with a Semiflexible Liquid Crystalline Polymer 17.1.1 Introduction 17.1.2 Blends of High Density Polyethylene (HDPE) with LCP 17.1.3 Blends of Isotactic Polypropylene with LCP 17.2 Crystallization Behavior of Blends of Polyolefins with a Semiflexible Liquid Crystalline Polymer 17.2.1 Isothermal and Nonisothermal Crystallization of Blends of Linear Low Density Polyethylene with a Semiflexible Liquid Crystalline Polymer 17.2.2 Crystallization Behavior of PE-g-LCP Copolymers 17.2.3 Effect of PP-g-LCP Compatibilizer on the Morphology and Crystallization of PP/LCP Blends 17.2.4 Isothermal Crystallization Kinetics of Compatibilized Blends of Polyolefins with a Semiflexible LCP 17.2.5 Crystallization and Morphology of Fibers Prepared from Compatibilized Blends of Polyethylene with a Liquid Crystalline Polymer 17.3 Conclusions Acknowledgment Nomenclature References

501 501 502 509 513

513 517 519 519

522 523 523 523 524

Contents

18. Functionalized Polyolefins and Aliphatic Polyamide Blends: Interphase Interactions, Rheology, and High Elastic Properties of Melts 18.1 18.2 18.3 18.4 18.5 18.6

Introduction Compounding and Interphase Phenomena in PA/PO Blends Rheological and High Elastic Properties of PA/PO Melts Morphology and Impact Strength of PA/g-PO Blends that Give Melts of High Viscosity and Strength New Applications of PA/g-PO Blends Conclusions

Nomenclature References 19. Plastic Deformation and Damage Mechanisms of Ternary PP/PA6/POE Polymer Blends 19.1 19.2

Introduction Materials Presentation and Experimental Methods 19.2.1 Materials Presentation 19.2.2 Morphological Study under SEM and TEM 19.2.3 Video-controlled Tensile System 19.3 Microstructure 19.4 General Mechanical Properties 19.4.1 Dynamic Mechanical Thermal Analysis 19.4.2 Toughness by Impact Loading and Yield Stress by Tension 19.5 Plastic Deformation under Uniaxial Tension 19.5.1 Definition of Volume Strain 19.5.2 True Axial Stress–Strain Relation 19.5.3 Volume Strain 19.5.4 Under Cyclic Tension 19.6 Mechanisms of Plastic Deformation and Damage 19.6.1 Damage Mechanisms in Polymer Blends 19.6.2 Influence of Damage Type 19.6.3 Damage-induced Shear Banding 19.6.4 Microscopic Observation under SEM 19.6.5 Microscopic Observation under TEM 19.6.6 Discussion 19.7 Conclusions Nomenclature References

20. Reactive Compatibilization of Binary and Ternary Blends Based on PE, PP, and PS 20.1 20.2

Introduction Friedel–Crafts Alkylation Reaction

xiii

527 527 528 534 543 550 550

551 551

556 556 558 558 559 559 561 566 566 567 569 569 571 573 575 579 579 580 581 582 589 589 596 597 597

600 600 601

xiv

Contents 20.3

Binary Blends 20.3.1 PE/PS Blends 20.3.2 PP/PS Blends 20.4 Ternary Blends: PE/PP/PS 20.5 Conclusions Nomenclature References

21. Polyolefin/Epoxy Resin Blends 21.1 21.2 21.3 21.4 21.5 21.6 21.7 21.8 21.9 21.10

Introduction Blend Preparation Miscibility Studies and Phase Diagrams Cure Kinetics Crystallization Behavior Morphology Dynamic Mechanical Properties Mechanical Properties Conductivity Studies Conclusion

Nomenclature References Index

603 603 611 615 620 620 621

623 623 624 625 629 631 637 644 647 649 656

658 659 663

Preface

Polyolefins are the most widely used commodity thermoplastics. They are of immense interest to polymer community because of their simple chemical structures and fascinating hierarchical structural organizations possible. To date, the field of polyolefins remains one of the most vibrant areas in polymer research. Polyolefin blends are a subset of polymer blends that emerged as a result of the need to meet application requirements not satisfied by synthesized neat polyolefins. In comparison to other subsets of polymer blends, polyolefin blends have distinct advantages of lower density, lower cost, processing ease, and good combination of chemical, physical, and mechanical properties. In the last several years, research and usage of polyolefin blends have increased due to new application opportunities (e.g., in medical and packaging) and the development of novel polyolefins. Although a sizable number of books on polyolefins and general polymer blends are available, only a few chapters address polyolefin blends. Currently, there is no single book that focuses exclusively on the fundamental aspects and applications of polyolefin blends. This is the primary source of motivation behind this book. The second motivation stems from the fact that new research trends in polyolefin blends such as in situ reactor blending and compatibilization/functionalization in the melt have emerged that need to be covered in a book format. This book is structured as follows: Chapter 1 serves as a guide to polyolefin blends introducing this important class of materials, why they are important, typical systems studied, issues of fundamental and applied interest, and current trends. The contributed chapters are divided into two main categories: polyolefin/polyolefin blends (Chapters 2–16) and polyolefin/nonpolyolefin blends (Chapters 17–21). Issues covered in these chapters include miscibility, phase behavior, functionalization, compatibilization, microstructure, crystallization, hierarchical morphology, and physical and mechanical properties. Most of the chapters are in the form of review articles. Some original articles are included to capture the latest development in polyolefin blends research. This book is intended to serve as a valuable reference for academic and industrial professionals performing research and development in the specific area of polyolefin blends or in the general area of polymer blends. Some review chapters include introductory materials to attract newcomers including senior undergraduate and graduate students and to serve as a reference book for professionals from other disciplines. Some knowledge of polymer chemistry, physics, and engineering, although not strictly essential, would be helpful to better appreciate the technical information of some chapters. Since this book is the first of its kind devoted solely to polyolefin blends, it is hoped that it will be sought after by a broader technical audience. xv

xvi

Preface

The chapters in this book were contributed by highly reputed professionals from academia, industry, and government laboratories spanning several countries from various continents. All manuscripts were peer reviewed in accordance with the guidelines utilized elsewhere by top-rated polymer journals. The editors would like to thank all contributors for believing in the realization of this book and taking painstaking tasks of going through the processes of manuscript preparation, submission, review, revision, and seeking supporting documents. Finally, sincere thanks are extended to all reviewers for their invaluable help, which undoubtedly improved the quality of this book. DOMASIUS NWABUNMA THEIN KYU

Contributors

Susana Areso Capdepo´n, Department of Physics and Engineering of Polymers, Polymer Engineering Group, Institute of Science and Technology of Polymers, CSIC CL Juan de la Cierva 3, 28006 Madrid, Spain. [email protected]

Shu-Lin Bai, Centre for Advanced Composite Materials (CACM), Department of Mechanics and Engineering Science, School of Engineering, Peking University, 100871 Beijing, China. [email protected] Silvia E. Barbosa, Planta Piloto de Ingenierı´a Quı´mica, PLAPIQUI (UNS-CONICET), Camino La Carrindanga km. 7 (8000), Bahı´a Blanca, Argentina. [email protected]

Maurizio Canetti, C.N.R. Istituto per lo Studio delle Macromolecole, Via E. Bassini 15, I-20133 Milano, Italy. [email protected] Numa J. Capiati, Planta Piloto de Ingenierı´a Quı´mica, PLAPIQUI (UNS-CONICET), Camino La Carrindanga km. 7 (8000), Bahı´a Blanca, Argentina. [email protected]

Mo´nica F. Dı´az, Planta Piloto de Ingenierı´a Quı´mica, PLAPIQUI (UNS-CONICET), Camino La Carrindanga km. 7 (8000), Bahı´a Blanca, Argentina. [email protected]

Bejoy Francis, Department of Chemistry and Biochemistry, Laurentian University, 935 Ramsey Lake Road, Sudbury, Ontario, P3E 2C6, Canada. [email protected]

Soney C. George, Department of Basic Science, Amal Jyothi College of Engineering, Koovapally, Kottayam 686518, Kerala, India. [email protected]; [email protected] Christian G’Sell, Laboratoire de Physique des Mate´riaux, Ecole des Mines de Nancy, Parc de Saurupt, 54042 Nancy Cedex, France. [email protected]

Chang-Sik Ha, Department of Polymer Science and Engineering, Pusan National University, Busan 609-735, Korea. [email protected]

Jean-Marie Hiver, Laboratoire de Physique des Mate´riaux, Ecole des Mines de Nancy, Parc de Saurupt, 54042 Nancy Cedex, France. [email protected]

Benjamin S. Hsiao, Department of Chemistry, Stony Brook University, Stony Brook, NY 11794, USA. [email protected]

Boleslaw Jurkowski, Division of Plastic and Rubber Processing, Institute of Material Technology, Poznan University of Technology, Piotrowo 3, 60-950 Poznan, Poland. [email protected]

Wirunya Keawwattana, Department of Chemistry, Faculty of Science, Kasetsart University, Bangkok 10903, Thailand. [email protected]

xvii

xviii

Contributors

Gue-Hyun Kim, Division of Applied Engineering, Dongseo University, Busan 617-716, Korea. [email protected] Il Kim, Department of Polymer Science and Engineering, Pusan National University, Busan 609-735, Korea. [email protected]

Yuri M. Krivoguz, Laboratory of Chemical Technology of Polymeric Composite Materials, V.A. Belyi Metal-Polymer Research Institute of National Academy of Sciences of Belarus, 32a Kiroc Street, 246050 Gomel, Belarus. [email protected] Thein Kyu, Department of Polymer Engineering, University of Akron, Akron, OH 44325, USA. [email protected]

Jesu´s Marı´a Garcı´a Martı´nez, Department of Physics and Engineering of Polymers, Polymer Engineering Group, Institute of Science and Technology of Polymers, CSIC CL Juan de la Cierva 3, 28006 Madrid, Spain. [email protected]

Rushikesh A. Matkar, Department of Polymer Engineering, University of Akron, Akron, OH 44325, USA. [email protected]

Liliya Minkova, Institute of Polymers, Bulgarian Academy of Sciences, Acad. G. Bonchev str. Bl.103A, 1113 Sofia, Bulgaria. [email protected]

Francis M. Mirabella, Lyondell Chemical Co., Equistar Technology Center, Cincinnati, OH 45249, USA. [email protected] Koh-Hei Nitta, Department of Chemical Engineering, Graduate School of Material Sciences, Kanazawa University, 920-1192, Japan. [email protected]

Domasius Nwabunma, 3M Company, Safety, Security, and Protection Business Services Laboratory, St. Paul, MN 55144, USA. [email protected] Emilia Pe´rez Collar, Department of Physics and Engineering of Polymers, Polymer Engineering Group, Institute of Science and Technology of Polymers, CSIC, CL Juan de la Cierva 3, 28006 Madrid, Spain. [email protected]

Stepan S. Pesetskii, Laboratory of Chemical Technology of Polymeric Composite Materials, V.A. Belyi Metal-Polymer Research Institute of National Academy of Sciences of Belarus, 32a Kirov Street, 246050 Gomel, Belarus. [email protected]

Subhendu Ray Chowdhury Department of Materials Science and Engineering, Pennsylvania State University, University Park, PA 16802, USA. [email protected]

Moonhor Ree, Department of Chemistry, Polymer Research Institute, Pohang Accelerator Laboratory, National Research Lab for Polymer Synthesis and Physics, and Center for Integrated Molecular Systems, Pohang University of Science & Technology (Postech), Pohang 790-784, Republic of Korea. [email protected]

Robert A. Shanks, School of Applied Sciences, RMIT University, GPO Box 2476V, Melbourne, Vic 3001, Australia. [email protected]

Jesu´s Taranco Gonza´lez, Department of Physics and Engineering of Polymers, Polymer Engineering Group, Institute of Science and Technology of Polymers, CSIC, CL Juan de la Cierva, 3, 28006 Madrid, Spain. [email protected]

Contributors

xix

Kohji Tashiro, Department of Future Industry-Oriented Basic Science and Materials, Toyota Technological Institute, Tempaku, Nagoya 468-8511, Japan. [email protected] Sabu Thomas, School of Chemical Sciences, Mahatma Gandhi University, Priyadarshini Hills, Kottayam 686560, Kerala, India. [email protected]

Sie C. Tjong, Department of Physics and Materials Science, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong. [email protected] Shigeyuki Toki, Department of Chemistry, Stony Brook University, Stony Brook, NY 11794, USA. [email protected] Gong-Tao Wang, School of Aerospace, Mechanical & Mechatronic Engineering, The University of Sydney, Sydney, NSW 2006, Australia. [email protected] Min Wang, Centre for Advanced Composite Materials (CACM), Department of Mechanics and Engineering Science, School of Engineering, Peking University, 100871 Beijing, China. [email protected]

James L. White, Department of Polymer Engineering, The University of Akron, Akron, OH, 44325, USA. [email protected] Masayuki Yamaguchi, School of Materials Science, Japan Advanced Institute of Science and Technology, Nomi, Japan. [email protected] Jinhai Yang, Department of Polymer Engineering, The University of Akron, Akron, OH, 44325, USA. [email protected]

Part I

Introduction

Chapter

1

Overview of Polyolefin Blends Domasius Nwabunma1

1.1 INTRODUCTION Polyolefins are synthetic polymers of olefinic monomers. They are the largest polymer family by volume of production and consumption. Several million metric tons of polyolefins are produced and consumed worldwide each year, and as such they are regarded as commodity polymers. Polyolefins have enjoyed great success due to many application opportunities, relatively low cost, and wide range of properties. Polyolefins are recyclable and significant improvement in properties is available via blending and composite technologies. Polyolefins may be classified based on their monomeric unit and chain structures as ethylene-based polyolefins (contain mostly ethylene units), propylene-based polyolefins (contain mostly propylene units), higher polyolefins (contain mostly higher olefin units), and polyolefin elastomers (1). Ethylene-based polyolefins are normally produced either under low pressure conditions using transition metal catalysts resulting in predominantly linear chain structure or under high pressure conditions using oxygen or peroxide initiators resulting in predominantly branched chain structures of various densities and crystallinity levels. Propylene-based polyolefins are normally produced with transition metal catalysts resulting in linear chain structures with stereospecific arrangement of the propylene units or special stereoblock structures from a single-site catalyst. Higher polyolefins are normally produced using transition metal catalysts resulting in linear and stereospecific chain structures. Polyolefin elastomers based mainly on a combination of ethylene and propylene may be produced using metal or single-site catalysts with or without the inclusion of dienes (for cross-linking) and are mostly amorphous with high molecular weights and heterogeneous in phase structures. One

1 3M Company, Safety, Security, and Protection Business Services Laboratory, St. Paul, MN 55144, USA

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

3

4

Polyolefin Blends

may conclude that a given polyolefin may be a homopolymer, copolymer, or terpolymer depending on the number of monomers used in making the polyolefin, crystalline or amorphous depending on their chain conformation, configuration, and processing conditions. Today, polyolefins and polyolefin-based materials are used in many applications. These applications include transportation (automotive, aerospace), packaging, medical, consumer products (toys, appliances, etc.), electronics, cable and wire coating, thermal and acoustic insulation, and building and construction. Polyolefins can be extruded as filaments (fibers), films (cast and blown), and pipes/profiles. They can be molded into parts of various shapes. They can be foamed with physical and chemical foaming/blowing or/and can be coated onto other materials.

1.2 OLEFINIC MONOMERS The alkenes having one or more unsaturated double bonds in their structures are the monomers used to synthesize polyolefins. They have the general formula Cn H2n ; n 2. Table 1.1 shows the first 10 members of the olefinic monomers with one double bond, which are often called a-olefins. The monomers in Table 1.1 form a homologous series of hydrocarbon compounds. Thus, apart from having the same general formula, all compounds in the series have the same functional groups. Each member of the group differs from the next in the series by the CH2 group equivalent to 14 relative molecular mass units. All members of the series have similar chemical properties. The physical properties of the compounds in the series show a progressive change with increasing relative molecular mass. The first three members of the alkenes homologous series are gases at room temperature. Those containing between 5 and 15 carbon atoms are colorless liquids and the higher compounds are waxy solids at room temperature. These a-olefinic monomers may be obtained as products of the cracking of the gas– oil and naphtha fractions of petroleum distillations. They can also be obtained from synthetic organic chemistry methods.

Table 1.1

Alkene Monomers with One Double Used in the Synthesis of Polyolefins.

No. of carbon atoms (n) 2 3 4 5 6 7 8 9 10

Formula (Cn H2n ; n 2)

Name (other name)

C2H4 C3H6 C4H8 C5H10 C6H12 C7H14 C8H16 C9H18 C10H20

Ethene (ethylene) Propene (propylene) Butene-1 (butylene) Pentene-1 Hexene-1 Heptene-1 Octene-1 Nonene-1 Decene-1

Chapter 1

Overview of Polyolefin Blends

5

1.3 POLYOLEFIN HOMOPOLYMERS, COPOLYMERS, AND TERPOLYMERS Polyolefin homopolymers, copolymers, and terpolymers are foundation materials for polyolefin blends. They may be obtained via radical or ionic chain growth polymerization of alkenes using conventional free radicals (e.g., from peroxides) and organometallic complexing (Ziegler–Natta and metallocenes) catalyst systems. Polyolefin polymerization technologies and novel catalyst systems have enabled the rapid development of polyolefins with a wide range of molecular chain structures, morphologies, properties, and particle size and shape. Polyolefin homopolymers include polyethylene (PE), polypropylene (PP), polybutene-1 (PB), polymethylpentene-1 (PMP), and higher polyolefins. Table 1.2 shows the structures of commercial polyolefin homopolymers. Of these, PE and PP are the largest by amount produced yearly by the global polyolefin companies (1). PE comes in various forms differing in chain structures, crystallinity, and density levels. These are high density polyethylene (HDPE), low density polyethylene (LDPE), linear low density polyethylene (LLDPE), ultralow density polyethylene (ULDPE), and ultrahigh molecular weight polyethylene (UHMWPE). PP and higher polyolefins come in three stereospecific forms of varying densities: isotactic, syndiotactic, and atactic forms. Polyolefin copolymers involve two olefinic monomers. The process of copolymerization is normally used to control the properties of the polyolefins. Some of the consequences of copolymerization are reduced crystallinity, melting point, modulus, strength, hardness, and low temperature impact. Polyolefin copolymers are either random or block copolymers of same or different monomers and may be a single phase or heterophasic depending on the amount of comonomer, the polymerization catalyst, and the process. For polyolefin copolymer of same monomers, this can be achieved by having different segments of the copolymer with different tacticities. One can have polyolefin block copolymers of same block or of varying block lengths. One can also have polyolefin copolymers consisting of both block and random segments together in the same macromolecule. Polyolefin copolymers are usually not homogeneous in composition but are actually mixtures of copolymers of varying compositions. It is also possible with polyolefins to have block copolymers with only one monomer. These are called stereoblock copolymers and can be achieved by having sections of the polyolefin copolymer possess different tacticities. Polyolefin copolymers started with LLDPE and ethylene–propylene rubber (EPR). Today, there are polyolefin copolymers of ethylene with butene-1, hexene1, octene, cyclopentene, and norbornene and copolymers of propylene with butene-1, pentene-1, and octene-1 in addition to ethylene. There are copolymers of butene-1 with pentene-1, 3-methylbutene-1, 4-methylpentene-1, and octene in addition to its copolymers with ethylene and propylene. There are copolymers of 4-methylpentene1 with pentene-1 and hexene-1 in addition to its copolymers with butene-1 and propylene. The function of the comonomers is to reduce crystallinity, as compared to the homopolymers, resulting in copolymers that are highly elastomeric with very low

6

Polyolefin Blends

Table 1.2

Structures of Commercial Polyolefin Homopolymers.

Name (other name) Polyethylene (polyethene, polymethylene)

Chemical structure (repeat unit) CH

CH 2 n

2

Polypropylene (polypropene)

CH

2

Polybutylene (polybutene-1)

CH

n

C H2 CH3

Polyisobutylene (polyisobutene-1)

Polybutadiene

CH

CH

CH

2

CH

CH

2

CH

2 n

n

C H2

Poly-4-methylpentene-1

CH3 C H3

CH3 CH 3

Polyisoprene CH 2

C

CH

CH

2 n

glass transition temperatures, high impact strength, low modulus, low density, and are often optically transparent. The most widely used multiphase polyolefin copolymer is polypropylene impact copolymer. These copolymers are typically composed of isotactic polypropylene (iPP) and EPR. Impact polypropylene copolymers are produced by various processes, but are generally characterized by the synthesis of iPP in the first reactor and EPR in the second reactor. Therefore, these systems are typically reactor blends. Postreactor blending can be done, but the starting material is most often the reactor blend polypropylene copolymer. Polyolefin copolymers are often used for film applications or as impact modifiers. Polyolefin terpolymers contain three olefinic monomers. A well-known example is ethylene propylene diene monomer (EPDM). The diene (double bond) monomer is

Chapter 1

Overview of Polyolefin Blends

7

usually ethylidene norbornene or 1,4-hexadiene. EPDM was introduced because of the difficulty in cross-linking saturated polyolefin homopolymers and copolymers. There are also functionalized polyolefins. These are usually copolymer or terpolymer containing functional groups like epoxide, anhydride, hydroxyl, acrylate, and carboxylic acid. These functional groups are either grafted onto the polyolefin after polymerization or added directly in situ during polymerization reactions involving olefins and functional groups bearing polar monomers such as vinyl acetate, methyl acrylate, butyl acrylate, glycidyl methacrylate, and acrylic acid. Functionalized polyolefins are useful compatibilizers and impact modifiers in blends and composites containing polyolefins and nonpolyolefins. In this sense, functionalized polyolefins may be considered as additives rather than matrix materials in formulation of polyolefin blends. Commercial polyolefins often contain additives such as colorants, flame retardants, antioxidants, light stabilizers, nucleating agents, antistatic agents, lubricants (microcrystalline waxes, hydrocarbon waxes, stearic acid, and metal stearates), and so on. These additives aid the processing and fabrication of products from polyolefins. Detailed treatments about specific polyolefins, polymerization systems/ mechanism/processes, structures, properties, processing, and applications may be found in References 2–9.

1.4 POLYOLEFIN BLENDS Polymer blends (mixtures of structurally different polymers (10–19)) are of interest because synthesized polymers have not satisfied increasing application demands. Polyolefin blends are a subset of polymer blends and may be classified into two groups. The first group contains polyolefins only, which are formulated to broaden the range of structures, properties, and applications offered by polyolefins. The second group contains polyolefins and nonpolyolefins, which are formulated to mitigate some of the property drawbacks of the polyolefin or the nonpolyolefin. For a blend to be classified as a polyolefin blend, it is presumed that the polyolefin component is of significant composition in the blend. In terms of miscibility, polyolefin blends may also be classified as miscible and immiscible blends (10, 11). Polyolefin blending requires knowledge of the miscibility and crystallinity of the blend, in addition to the contributions of the components of the blend. Miscibility depends on molecular structure, blend composition, and mixing temperature. To characterize miscibility, a phase diagram is needed. Nonolefinic thermoplastic polymers that in principle may be blended with polyolefins include polyamides (nylons) such as polyamide 6, polyamide 66, polyphenylene sulfide (PPS), polyphenylene ether (PPE), and polyphenylene oxide (PPO); polyesters such as polyethylene terephthalate (PET), polybutylene terephthalate (PBT), polyethylene naphthalate (PEN), polytrimethylene terephthalate (PTT), polycarbonates, polyethers, and polyurethanes; vinyl polymers such as polystyrene (PS), polyvinyl chloride (PVC), polymethylmethacrylate (PMMA), and ethylene

8

Polyolefin Blends

vinyl acetate copolymer (EVA); block and graft copolymers (styrene–acrylonitrile copolymer, styrene–butadiene copolymer, styrene–ethylene–butadiene–styrene terpolymer, etc.); and liquid crystalline polymers (LCPs). Thermosetting resins that may be blended with polyolefins include, but are not limited to, the following: epoxies, unsaturated polyesters, phenol formaldehyde, melamine formaldehyde, urea formaldehyde, silicones, and so on. The properties that polyolefins normally contribute in blends with other polymers include high melt strength and elasticity, toughness, low viscosity for processability, low polarity, dielectric constant, and loss, and chemical resistance and moisture absorption resistance. Nonpolyolefins contribute to high modulus, heat resistance, and oxygen or solvent barrier properties. For example, barrier properties of polyolefins can be improved by blending with polymers such as ethylene vinyl alcohol and polyvinylidene chloride. Blends of polyolefins with nylons and polycarbonate (PC) allow balanced control of permeability and water retention. Polystyrene (PS) is an interesting candidate for blending with polyolefins for mechanical reasons as well as for paintability and printability. There are two classes of polyolefin blends: elastomeric polyolefin blends also called polyolefin elastomers (POE) and nonelastomeric polyolefin blends. Elastomeric polyolefin blends are a subclass of thermoplastic elastomers (TPEs). In general, TPEs are rubbery materials that are processable as thermoplastics but exhibit properties similar to those of vulcanized rubbers at usage temperatures (19). In TPEs, the rubbery components may constitute the major phase. However, TPEs include many other base resins, which are not polyolefins, such as polyurethanes, copolyamides, copolyesters, styrenics, and so on. TPEs are now the third largest synthetic elastomer in total volume produced worldwide after styrene– butadiene rubber (SBR) and butadiene rubber (BR). Two important types of elastomeric polyolefin blends are reactor-made iPP/ EPR blends and postreactor blend iPP/EPDM. The latter is called thermoplastic vulcanizates (TPVs), produced by dynamic vulcanization of blends containing a thermoplastic and an elastomer. To make iPP/EPDM TPV, the two polymers PP and EPDM are mixed with curatives, such as peroxides, phenolic resins, or sulfur with accelerators, and dynamically cured in an extruder resulting in a blend consisting of micrometer-sized elastomer particles dispersed in the PP matrix (20–24). Paraffinic oils are added in the melt mixing process for viscosity control and cost. In iPP/ EPDM TPV, the crystalline iPP resin is normally the minor phase. Recently, polyolefin plastomers have been added to the class of elastomeric polyolefin blends. Polyolefin plastomers are ultralow molecular weight linear low density polyethylenes (ULMW-LLDPE). Nonelastomeric polyolefin blends are blends of polyolefins with mostly nonpolyolefin (other thermoplastic) matrices as mentioned earlier. Polyolefin blends of commercial importance are normally made via two methods: blending in the melt either during polymerization or mechanically after the polymerization process. The first method, called in-reactor blending, involves the blending of different polyolefins (homopolymers, random, and block copolymers) in a polymerization reactor. This is enabled by the presence of multiple catalyst species

Chapter 1

Overview of Polyolefin Blends

9

in the polymerization recipe. A good example is in-reactor-made EPR/iPP blend, which is normally prepared by adding ethylene monomer to propylene monomer toward the end of propylene polymerization process. The function of EPR is to improve iPP flexibility; hence, EPR/iPP blend is often called toughened or impact PP and finds wide applications in the consumer industry. Another good example of in-reactor-made polyolefin blend is linear low density polyethylene (LLDPE), which often contains several ethylene/a-olefin copolymers that differ in ethylene contents. LLDPE and reactor-made EPR/iPP are often called thermoplastic olefins (TPOs). The second method, called postreactor blending, involves mechanical blending of a premade polyolefin with other polyolefins or nonpolyolefins in compounding extruders. A practical example of polyolefin blend made using this method is a blend of isotactic PP with cured EPDM, as described earlier. Other examples of polyolefin blends made by postreactor mechanical blending include polyolefin/polyamides (nylons), polyolefin/polyesters, polyolefin/polystyrene (PS), polyolefin/polyvinyl chloride (PVC) blends, and so on. EPR/iPP blend can also be made by postreactor mechanical blending as well as by in-reactor blending. Postreactor blending via single and twin screw compounding is still the preferred method of polyolefin blending because it is quick, easy, economical, and efficient. Table 1.3 shows a summary of specific polyolefin blends that have been studied in the literature extracted from references (25–310). These blends involve the following polyolefins: PP, PE (LLDPE, LDPE, HDPE, and UHMWPE), EPR, EPDM, and PB. Some of the blends listed in Table 1.3 are of commercial importance (6). There are many publications (journal articles and patents) on polyolefin blends in the literature. Tables 1.4 contains a summary of the journal articles (25–310) and patents published each year during the 6-year period 2000–2005. The choice of date range is arbitrary. The number of journal articles for each year was obtained from a search of electronic version of English-based polymer and polymer-related journals using the keywords polyolefin and blends. Within polyolefin keyword, the subkeywords used in the search were polyethylene (PE, LLDPE, LDPE, HDPE, UHMWPE, PE, etc.), polypropylene (PP, iPP, sPP, aPP, etc.), polybutene-1, poly-4-methylpentene-1, ethylene–diene monomer, ethylene–propylene– diene terpolymer, ethylene propylene rubber, thermoplastic olefins, natural rubber (NR), polybutadiene, polyisobutylene (PIB), polyisoprene, and polyolefin elastomer. For the polyolefin blends patent search, polymer indexing codes and manual codes were used to search for the patents in Derwent World Patent Index based on the above keywords listed in the search strategy. Table 1.4 shows an increasing trend in the number of publications. It should be noted that while generating Table 1.4, some publications may have been missed in the reference search period (2000–2005). There are several issues of interest in polyolefin blends research. They may be categorized into formulation design and processing; miscibility, structural, molecular, and property characterization; end-use properties and performance; and

10

PE/silicon rubber (76) PE/starch (85, 277, 289, 295) PE/hydrolyzed collagen (98) PE/natural rubber/PP (238, 254, 302) PE/poly(3-hydroxybutyrate) (251) PE/nitrile rubber (100, 183)

PE/polydianilinephosphazene (91)

PE/PS (47–48, 67, 79, 123, 145, 198, 219, 223) PE/PP/PA66 (71)

PE/PA6 (25, 27, 71, 88, 92, 96, 130, 140, 144, 149, 168, 204, 215, 224, 226, 272, 273) PE/PP (33, 34, 36, 47, 62, 69, 83, 93, 107, 116, 182, 196, 200, 203, 236, 237, 239, 243, 248, 255, 263, 297) PE/PP/cycloolefin copolymers (307) PE/ethylene-co-propylene-co-butene-1 (99) PP/PPE (216, 217) PP/EPDM (50, 66, 113, 121, 136, 199, 202, 207, 224, 267) PP/EPR (35, 63, 77, 118, 124, 152, 163, 195, 231, 233, 271) PP/epoxy (62, 120) PP/PA6/PS (205) PP/cyclopolyolefin (162) PP/natural rubber (64, 241, 247, 287) PP/recycle rubber(241, 247) PP/natural rubber/recycle rubber powder (249)

PE/PP/PA6 (71)

LLDPE/LDPE/wax (80)

Polyolefin blends PE/PP/EPR (35, 114) PE/PC (29, 173, 213) PE/poly(silsesquioxanes) (156) PE/ethylene acrylic elastomer (281)

Polyolefin Blends Studied in the Literature.

HDPE/UHMWPE (81) LDPE/HDPE (84, 153, 160, 221, 259, 284) LLDPE/HDPE (285) LLDPE/LDPE (72, 104, 108, 119, 128, 129, 181, 211, 214, 221, 242, 258, 261, 292, 309)

Table 1.3

PP/EPDM/epoxidized natural rubber (246) PP/epoxidized natural rubber (301) Butadiene/styrene–butadiene rubber (293) EPDM/PA66 EPDM/SBR (68) EPDM/NBR (234)

PP/EPDM/natural rubber (244–246)

PP/PC (25) iPP/sPP (40); iPP/aPP (103)

PP/PB (26, 109)

PP/PA6 (28, 54, 58, 71, 111, 170, 178, 262, 283, 286, 288, 305) PP/PA6/ethylene-co-octene (28, 31, 43, 46)

PP/PET (137) PP/PE/EPDM (49) PP/EVOH (135, 298) PP/PA66 (28)

11

PE/PBT (82, 300) PE/PS/PMMA (209) PE/wax (256)

PE/PA12 (142) PE/PTT (37) PE/PVC (52, 201) PE/starch/PCL (90) PE/ethylene-co-octene (46, 278) PE/ethylene-co-butene-1 (89, 171) PE/PA6,6 (127, 169, 273) PE/perfluoropolyether (126) PE/EVA (45, 51, 56, 78, 97, 161, 270) PE/PET (166, 167, 208, 213, 220, 274)

PP/ethylene-co-methyl acrylate (164) PP/P12 (304) PP/TPU (228) PP/SBS (74) PP/SAN (184) PP/EVA (177, 189, 191, 193, 294) PP/SEBS (74, 112, 218) PP/SEP (105, 218) PP/SEPS (105, 218) PP/PS (38, 39, 117, 155, 159, 227, 212, 229, 299, 310) PP/ethylene-co-octene (59, 87) PP/ethylene-co-butene-1 (59) PP/LCP (225, 232) Natural rubber/polystyrene (290) Natural rubber/reclaimed rubber (252) Epoxidized natural rubber/natural rubber (253)

EPDM/SAN (53, 308) EPDM/PA6 (101, 279) EPDM/polyaniline (122) EPDM/TPU (95) EPDM/PP/PA6 (262) EPDM/PA12,10 (262) EPDM/PP6,10 (266) EPDM/natural rubber (44, 65, 240, 250, 269) EPR/polydimethylsiloxane (70) EPR/PA6 (57, 60)

12

Polyolefin Blends

Table 1.4 Summary of Number of Electronic articles on Polyolefin Blends Published Between 2000 and 2005 in English Language-based Polymer and Polymer-related Journals in Comparison to the Number of Patents. Journal articles (25–310) Year Number of Articles

2000 36

2001 31

2002 42

2003 42

2004 68

2005 66

Total 285

2004 322

2005 365

Total 2349

Patents Year Number of Patents

2000 517

2001 444

2002 406

2003 295

modeling and simulation. Regarding formulation design and processing, some of the issues of interest include i. Use of block, random, and graft copolymers in compatibilization. ii. Effects of molecular structure, weight, and additives. iii. Phase behavior, miscibility, and compatibility issues. iv. Batch and continuous mixing/compounding, extrusion, and molding into films, fibers, and other articles. v. Physical blending versus reactive blending. Regarding miscibility, structural, molecular, and property characterization, some of the issues of interest include i. Morphological characterization using techniques such as scanning electron microscopy (SEM), transmission electron microscopy (TEM), atomic force microscopy (AFM), and polarized light microscopy (PLM). ii. Structural characterization using radiation scattering and diffraction techniques such as X-ray scattering (XRS), X-ray diffraction (XRD), and electron diffraction (ED). iii. Structural characterization using spectroscopic techniques such as nuclear magnetic resonance (NMR), Fourier transform infrared (FTIR), and temperature rising elution fractionation (TREF). iv. Melt rheology. v. Phase separation dynamics. vi. Thermal transitions and thermal stability. vii. Isothermal and nonisothermal crystallization behavior under quiescent and nonquiescent conditions. viii. Crystallinity and crystal structure determination by various methods such as density, differential scanning calorimetry (DSC), XRD, and ED. ix. Tacticity, copolymer sequence distribution, and comonomer composition.

Chapter 1 Overview of Polyolefin Blends

13

Regarding end-use properties and performance, some of the issues of interest include i. Mechanical (static and dynamic) behavior under tensile, shear, or compressive, and impact mode. ii. Plastic deformation. iii. Thermomechanical and thermal stability behavior. iv. Adhesion, interfacial, and interphase behavior. v. Fire resistance/flammability behavior. vi. Barrier and transport properties. vii. Optical properties (gloss, haze, and transparency). viii. Surface/tribological properties. ix. Electrical/dielectric properties. x. Aging effects (time–temperature-dependent behavior). Modeling and simulation involves theoretical analysis of processing, characterization, and performance behavior using phenomenological, atomistic, molecular dynamics, and Monte Carlo methods, among others, and comparisons with experimental results. Polyolefin blends are generally immiscible due to hydrophobic or nonpolar nature of polyolefins. The immiscibility leads to phase separation, which is responsible for the poor mechanical properties of these blends. Immiscibility of polyolefin/ nonpolyolefin blends can be mitigated by adding a proper compatibilizer. The function of a compatibilizer is to reduce interface tension (strengthen the interface between the phases) and thus improve mechanical properties of the stabilized blend (reduce the size and morphology of the phase-separated phases). The compatibilizer strengthens the interface by broadening it from a sharp change in composition and properties to a broader gradual transition interface. Thus, the chief concern in compatibilization of polyolefin/nonpolyolefin blends is phase morphology. One way to achieve compatibilization involves physical processes such as shear mixing and thermal history, which modify domain size and shape. The second way is the use of physical additives to increase attraction between molecules and phases. The third method is reactive processing, which is used to change the chemical structure of one or more of the components in the blend and thus increase their attraction to each other. Table 1.5 contains a list of compatibilizers used in the formulation of polyolefin blends. As can be seen from Table 1.5, most of the compatibilizers used in the formulation of polyolefin blends contain compounds such as maleic anhydride, acrylic and methacrylic acid, glycidyl methacrylate, and diblock and triblock copolymers involving styrene, ethylene, and butadiene.

1.5 TRENDS IN POLYOLEFIN BLENDS The first trend is the growing use of in-reactor blending technique to produce new elastomeric or toughened polyolefins for demanding applications traditionally reserved

14

Polyolefin Blends

Table 1.5

Compatibilizers Used in the Formulation of Polyolefin Blends.

Compatibilizer

Polyolefin blend

PP-g-maleic anhydride Ethylene-co-acrylic acid Ethylene-co-glycidyl methacrylate Styrene-b-ethylene-co-propylene-g-maleic anhydride PE-g-maleic anhydride Styrene-co-ethylene-co-butadiene-costyrene-g-maleic anhydride Sodium-neutralized ethyleneco-methacrylic acid PE-g-glycidyl methacrylate Styrene-co-ethylene-co-butadiene-costyrene-g-glycidyl methacrylate Styrene-co-ethylene-co-butadiene-co-styrene Styrene-co-ethylene-co-butadiene-costyrene-g-maleic anhydride PE-g-maleic anhydride EPR-g-maleic anhydride PE-g-maleic anhydride PP-g-maleic anhydride Styrene-co-ethylene-co-butadiene-co-styrene Styrene-co-butadiene Styrene-co-ethylene-co-butadiene-co-styrene Styrene-co-butadiene-co-styrene Liquid polybutadiene þ diakyl peroxides Chlorinated PE Ethylene-co-acrylic acid Bismaleimide-g-PE PE-g-maleic anhydride PE-g-epoxidized natural rubber PE-g-diethyl maleate Sodium neutralized ethylene-co-acrylic acid PE-g-glycidyl methacrylate PE-g-maleic anhydride Ethylene-co-glycidyl methacrylate Styrene-co-ethylene-co-butadiene-co-styrene Styrene-co-ethylene-co-butadiene-co-styrene PP-g-3-isopropenyl-a,a-dimethylbenzene isocyanate PP-g-succinic anhydride and PP-g-fluorescein PP-g-maleic anhydride Ethylene-co-butyl acrylate-g-maleic anhydride Styrene-co-ethylene-co-butadiene-co-styreneg-glycidyl methacrylate

PE/PA6 PE/PA6 PE/PA6 PE/PA6

(27, 88) (27, 92, 140, 204, 215, 224) (27, 144) (130)

PE/PA6 (130) PE/PA6 (130, 168) PE/PA6 (92) PE/PA6 (144) PE/PA6 (168) PE/PP/PA6,6 (71) PE/PP/PA6,6 (71) PE/PA12 (142) PE/PP (47) PE/PP (71, 239) PE/PP (71, 239) PE/PS (47) PE/PS (48, 115) PE/PS (79) PE/PS (145) PE/PS (223) PE/PVC (52) PE/hydrolyzed collagen (98) PE/hydrolyzed collagen (98) PE/acrylonitrile-co-butadiene (100) PE/scrap rubber powder (165) PE/PET (166) PE/LCP (188) PE/PET (220) PE/ethylene-co-acrylic acid (281) PE/plasticized tapioca starch (289) PP/PA6,6 (28) PP/PA6 (28) PP/PA6 (54) PP/PA6 (111) PP/PA6 (58, 170, 178, 262, 283, 286, 289) PP/PA6 (170) PP/PA6 (178)

Chapter 1 Overview of Polyolefin Blends Table 1.5 (Continued) Compatibilizer

Polyolefin blend

PP-g-bismaleimide PP-g-acrylic acid EPDM-g-maleic anhydride Ethylene-co-octene-g-maleic anhydride PP-g-maleic anhydride Styrene/AlCl3-catalyzed friedel–crafts alkylation PP-g-PS Styrene-co-ethylene-co-butadiene-co-styrene Styrene-co-ethylene-co-butadiene-costyrene þ ionomer resin Ionomer Zn2þ Styrene-co-ethylene-co-propylene-g-glycidyl methacrylate Styrene-co-ethylene-co-butadiene-costyrene-g-glycidyl methacrylate PP-g-monomethyl itaconate PP-g-dimethyl itaconate PP-g-maleic anhydride PP-g-acrylic acid PP-g-monomethyl itaconate PP-g-primary amine PP-g-secondary amine N,N-m-Phenylene bismaleimide N,N-m-Phenylene bismaleimide Polyoctenamer Polystyrene-modified natural rubber PP-g-maleic anhydride Hydroxylated EVA Ethylene-co-butyl acrylate-co-glycidyl methacrylate EPDM-g-maleic anhydride EPDM-g-styrene-co-acrylonitrile Ethylene-co-octene-g-maleic anhydride Mercapto-modified EVA EPDM-g-maleic anhydride EPDM-g-maleic anhydride EPDM-g-vinyltriethoxysilane EPDM-g-vinyloxyaminosilane EPR-g-maleic anhydride EPR-g-MAH þ epoxy-modified lignosulfonate Styrene-co-isoprene

PP/PA6 (262) PP/PA6 (262) PP/PA6 (279) PP/PA6 (286) PP/epoxy (62, 120) PP/PS (21, 117) PP/PS (155, 299) PP/PS (299) PP/PS/PA6 (205) PP/EVOH (135) PP/PET (137) PP/PET (137) PP/EPDM (139) PP/EPDM (139) PP/EPDM/PA6 (262) PP/EPDM/PA6 (262) PP/EPR (163) PP/TPU (228) PP/TPU (228) PP/natural rubber/PE (238) PP/EPDM/natural rubber (244–245) PP/PE/natural rubber (254) PP/natural rubber (288) PP/EVA (294) PP/EVA (294) PTT/PE (37) EPDM/natural rubber (44, 240, 250) EPDM/SAN (53) EPDM/TPU (91) EPDM/NBR (234) EPDM/PA6,10 (266) EPDM/PA12,10 (266) EPDM/PE (275) EPDM/PE (296) EPR/PA6 (57, 60) EPR/lignin (172) Natural rubber/PS (290)

15

16

Polyolefin Blends

for more expensive engineering thermoplastics. There is also a growing interest in the use of elastomer-rich thermoplastic olefin blends that contain 60–75% rubber and 40– 25% polypropylene for automotive interior applications because they provide the required amount of lower and upper service temperature performance. There is also growing use of beta nucleator functional additive in polyolefin blends, for example, PP/ EPR blend. A beta nucleating agent leads to blend with improved impact and ductility. The second growing trend is the impact modification of polyolefin blends using styrenic block copolymers, which are known to be clear, strong, have low glass transition, compatible with PP, form interpenetrating polymer networks, and very efficient in contrast to maleic anhydride-grafted polyolefins. The third trend is the growing interest in functionalization of polyolefin blends in their melt by means of reactive extrusion. Particular attention has been paid to blended systems PP/PE, PP/EPR, PE/ethylene–octene copolymer (EOC), PP/EOC, PE/PS, and PP/PS functionalized in melt by reactive extrusion. The major field for application of functionalized polyolefin blends is compatibilization of blends of condensation polymers, where they can be used in place of homopolyolefins. The fourth trend is spurred by environmental sustainability concerns and the need for increased recyclability and reuse of polyolefin blends. In this regard, there is increasing replacement of PVC by polyolefin–polyolefin blends. There is also an increase in recyclability of EPDM rubber vulcanizates since EPDM is the fastest growing elastomer among synthetic rubber and the most used of nontire rubbers. Also, cryogenically ground rubber tires are being used as fillers for polyolefin blends such as LLDPE/HDPE. The fifth trend is the use of beta nucleator as opposed to alpha nucleator in PP containing blends such as PP/EPR. The use of beta nucleating agent results in blend with improved impact and ductility The sixth trend is the growing interest in foams, nonwoven, and elastic materials based on polyolefin blends The seventh trend is the increasing use of novel processing methods. For example, there is growing use of supercritical fluids (e.g., supercritical carbon dioxide and nitrogen gases) to foam polyolefin blends for density reduction. There is use of ultrasound to, for example, devulcanize cross-linked rubber. There is use of solid-state shear mechanical processing to break the polyolefin blend material into submicron particles to make environment friendly (water-based) polyolefin dispersions. There is use of electrospinning technique to make polyolefin fibers and in particular nanofibers. In conclusion, the importance of polyolefin blends has increased dramatically in the last couple of decades and this is sure to continue as new polyolefins are made and as new applications are sought for these materials.

NOMENCLATURE AFM BR

Atomic force microscopy Butyl rubber

Chapter 1 Overview of Polyolefin Blends

DSC ED EOC EPDM EPR EVA EVOH FTIR HDPE LCP LDPE LLDPE NBR NMR NR PA PA12 PA12,10 PA6 PA6,6 PA6,10 PB PBT PC PCL PE PEN PET PIB PLM PMMA PMP PP iPP sPP aPP POE PPE PPO PPS PS PTT PVC SAN SB

Differential scanning calorimetry Electron diffraction Ethylene octene copolymer Ethylene propylene diene monomer Ethylene propylene rubber Ethylene-co-vinyl acetate Ethylene-co-vinyl alcohol Fourier transform infrared High density polyethylene Liquid crystal polymer Low density polyethylene Linear low density polyethylene Nitrile butadiene rubber Nuclear magnetic resonance Natural rubber Polyamide Polyamide 12 Polyamide 12,10 Polyamide 6 Polyamide 6,6 Polyamide 6,10 Polybutene-1 Polybutylene terephthalate Polycarbonate Polycaprolactone Polyethylene Polyethylene naphthalate Polyethylene terephthalate Polyisobutylene Polarized light microscopy Polymethylmethacrylate Polymethylpentene-1 Polypropylene Isotactic polypropylene Syndiotactic polypropylene Atactic polypropylene Polyolefin elastomer Poly(2,6-dimethyl-1,4-phenylene ether) Polyphenylene oxide Polyphenylene sulfide Polystyrene Polytrimethylene terephthalate Polyvinyl chloride Styrene-co-acrylonitrile Styrene-co-butadiene

17

18

Polyolefin Blends

SBR SBS SEM SEBS SEP SEPS TEM TPE TPO TPU TPV TREF UHMWPE ULDPE ULMW-LLDPE XRD XRS

Styrene butadiene rubber Styrene-co-butadiene-co-styrene Scanning electron microcopy Styrene-co-ethylene-co-butadiene-co-styrene Styrene-co-ethylene-co-propylene Styrene-co-ethylene-co-propylene-co-styrene Transmission electron microscopy Thermoplastic elastomer Thermoplastic olefin Thermoplastic polyurethane Thermoplastic vulcanizate Temperature rising elution fractionation Ultrahigh molecular weight polyethylene Ultralow density polyethylene Low molecular weight linear low density polyethylenes X-ray diffraction X-ray scattering

REFERENCES 1. M. Gahleitner, Prog. Polym. Sci., 26, 895 (2001). 2. J. L. White, Polyolefins, Hanser, Munich, 2005. 3. N. Pasquini (ed.), Polypropylene Handbook, Hanser, Munich, 2005. 4. K. Harutun (ed.), Handbook of Polypropylene and Polypropylene Composites, Marcel Dekker, New York, 2003. 5. C. Vasile (ed.), Handbook of Polyolefins, Marcel Dekker, New York, 2000. 6. A. J. Peacock (ed.), Handbook of Polyethylene: Structures, Properties, and Applications, Marcel Dekker, New York, 2000. 7. O. Olagoke, Handbook of Thermoplastics, Marcel Dekker, New York, 1997. 8. E. Moore (ed.), Polypropylene Handbook, Hanser, Munich, 1996. 9. J. Karger-Kocsis (ed.), Polypropylene: Structure, Blends, and Composites, Chapman and Hall, 1995. 10. L. A. Utracki (ed.), Polymer Blends Handbook, Vols. 1 and 2, Springer, New York, 2003. 11. C. Vasile and A. K. Kulshreshtha (eds.), Handbook of Polymer Blends and Composites, Rapra Technology Limited, 2003. 12. L. A. Utracki (ed.), Polymer Blends Handbook, Vols. 1 and 2, Kluwer Academic Publishers, Dordrecht, 2002. 13. D. R. Paul and C. B. Bucknall (eds.), Polymer Blends: Formulations and Performance, Vols. 1 and 2, Wiley, New York, 2000. 14. G. O. Shonaike and G. P Simon (eds.), Polymer Blends and Alloys, Marcel Dekker, New York, 1999. 15. T. Araki, Q. Tran-Cong, and M. Shibayama (eds.), Structure and Properties of Multi-Phase Polymeric Materials, Marcel Dekker, New York, 1998. 16. I. S. Miles and S. Rostani (eds.), Multi-Component Polymer Systems, Longman Scientific, Essex, 1992. 17. L. A. Utracki (ed.), Polymer Blends and Alloys: Thermodynamics and Rheology, Hanser, Munich, 1989.

Chapter 1 Overview of Polyolefin Blends

19

18. D. R. Paul and S. Newman (eds.), Polymer Blends, Academic Press, New York, 1978. 19. J. Manson and L. H. Sperling (eds.), Polymer Blends and Composites, Plenum Press, New York, 1976. 20. G. Holden and N. R. Legge (eds.), Thermoplastic Elastomers, Hanser, New York, 1996. 21. W. K. Fisher, US Patent 3,835,201 (1974). 22. S. Daniel and R. S. Porter, Polymer, 19, 448 (1978). 23. A. Y. Coran and R. P Patel, Rubber Chem. Technol., 53, 141 (1980). 24. S. Abdou-Sabet and R. P. Patel, Rubber Chem. Technol., 64, 769 (1991). 25. N. Chapleau, B. D. Favis, and P.J. Carreau, Polymer, 41, 6695 (2000). 26. Y. T. Shieh, M. S. Lee, and S. A. Chen, Polymer, 42, 4439 (2001). 27. V. Chiono, S. Filippi, H. Yordanov, L. Minkova, and P. Magagnini, Polymer, 44, 2423 (2003). 28. S.-L. Bai, G.-T. Wang, J.-M. Hiver, and C. G’Sell, Polymer, 45, 3063 (2004). 29. B. Lin and U. Sundararaj, Polymer, 45, 7605 (2004). 30. K. Shimizu, H. Wang, Z. Wang, G. Matsuba, H. Kim, and C. C. Han, Polymer, 45, 7061 (2004). 31. C. G’Sell, S.-L. Bai, and J.-M. Hiver, Polymer, 45, 5785 (2004). 32. J. I. Uriguen, L. Bremer, V. Mathot, and G. Groeninckx, Polymer, 45, 5961 (2004). 33. B. Na, K. Wang, Q. Zhang, R. Du, and Q. Fu, Polymer, 46, 3190 (2005). 34. D. Dikovsky, G. Marom, C. A. Avila-Orta, R. H. Somani, and B. S. Hsiao, Polymer, 46, 3096 (2005). 35. K.-H. Nitta, Y.-W. Shin, H. Hashiguchi, S. Tanimoto, and M. Terano, Polymer, 46, 965 (2005). 36. B. Na, Q. Zhang, K. Wang, L. Li, and Q. Fu, Polymer, 46, 819 (2005). 37. S. H. Jafari, A. Yavari, A. Asadinezhad, H. A. Khonakdar, and F. Bohme, Polymer, 46, 5082 (2005). 38. Q. Xing, M. Zhu, Y. Wang, Y. Chen, Y. Zhang, J. Pionteck, and H. J. Adler, Polymer, 46, 5406 (2005). 39. M. F. Diaz, S. E. Barbosa, and N. J. Capiati, Polymer, 46, 6096 (2005). 40. X. Zhang, Y. Zhao, Z. Wang, C. Zheng, X. Dong, Z. Su, P. Sun, D. Wang, C. C. Han, and D. Xu, Polymer, 46, 5956 (2005). 41. A. V. Machado and M. van Duin, Polymer, 46, 6575 (2005). 42. W. G. F. Sengers, M. Wu¨bbenhorst, S. J. Picken, and A. D. Gotsis, Polymer, 46, 6391 (2005). 43. S.-L. Bai, C. G’Sell, J.-M. Hiver, and C. Mathieu, Polymer, 46, 6437 (2005). 44. S. H. El-Sabbagh, J. Appl. Polym. Sci., 90, 1 (2003). 45. G. Takidis, D. N. Bikiaris, G. Z. Papageorgiou, D. S. Achilias, and I. Sideridou, J. Appl. Polym. Sci., 90, 841 (2003). 46. N. Phochalam, A. Tabtiang, and R. A. Venables, J. Appl. Polym. Sci., 90, 1655 (2003). 47. T. Kallel, V. Massardier-Nageotte, Mo. Jaziri, J.-F. Ge´rard, and B. Elleuch, J. Appl. Polym. Sci., 90, 2475 (2003). 48. I. Fortelny´, J. Mikesˆova´, J. Hromadkova´, V. Hasˆova´, and Z. Hora´k, J. Appl. Polym. Sci., 90, 2303 (2003). 49. T. Nedkov and F. Lednicky´, J. Appl. Polym. Sci., 90, 3087 (2003). 50. Y. Chen, Y. Cao, and H. Li, J. Appl. Polym. Sci., 90, 3519 (2003). 51. T. Wu, Y. Li, D.-L. Zhang, S.-Q. Liao, and H.-M. Tan, J. Appl. Polym. Sci., 91, 905 (2004). 52. Z. Fang, G. Ma, C. Liu, and C. Xu, J. Appl. Polym. Sci., 91, 763 (2004). 53. X. Qu, S. Shang, G. Liu, S. Zhang, Y. Zhang, and L. Zhang, J. Appl. Polym. Sci., 91, 1685 (2004). 54. G.-H. Hu, H. Cartier, L.-F. Feng, and B.-G. Li, J. Appl. Polym. Sci., 91, 1498 (2004). 55. M. Canetti, A. De Chirico, and G. Audisio, J. Appl. Polym. Sci., 91, 1435 (2004). 56. H. A. Khonakdar, J. Morshedian, H. Eslami, and F. Shokrollahi, J. Appl. Polym. Sci., 91, 1389 (2004).

20

Polyolefin Blends

57. S. C. George, K. N. Ninan, G. Geuskens, and S. Thomas, J. Appl. Polym. Sci., 91, 3756 (2004). 58. D. Shi, J. Yin, Z. Ke, Y. Gao, and R. Ky Li, J. Appl. Polym. Sci., 91, 3742 (2004). 59. J. Lee, C. W. Macosko, and F. S. Bates, J. Appl. Polym. Sci., 91, 3642 (2004). 60. Z. Oommen, S. R. Zachariah, S. Thomas, G. Groeninckx, P. Moldenaers, and J. Mewis, J. Appl. Polym. Sci., 92, 252 (2004). 61. B. C. Poon, S. P. Chum, A. Hiltner, and E. Baer, J. Appl. Polym. Sci., 92, 109 (2004). 62. X. Jiang. H. Huang, Y. Zhang, and Y. Zhang, J. Appl. Polym. Sci., 92, 1437 (2004). 63. M. Pires, R. S. Mauler, and S. A. Liberman, J. Appl. Polym. Sci., 92, 2155 (2004). 64. S. Varghese, R. Alex, and B. Kuriakose, J. Appl. Polym. Sci., 92, 2063 (2004). 65. S. Kiatkamjornwong and K. Pairpisit, J. Appl. Polym. Sci., 92, 3401 (2004). 66. L. Tang, B. Qu, and X. Shen, J. Appl. Polym. Sci., 92, 3371 (2004). 67. G. Chen, S. Guo, and Y. Li, J. Appl. Polym. Sci., 92, 3153 (2004). 68. T. Muraleedharan Nair, M. G. Kumaran, and G. Unnikrishnan, J. Appl. Polym. Sci., 93, 2606 (2004). 69. G. Liu, Y. Chen, and H. Li, J. Appl. Polym. Sci., 94, 977 (2004). 70. M. Carlberg, D. Colombini, and F. H. J. Maurer, J. Appl. Polym. Sci., 94, 2240 (2004). 71. R. Krache, D. Benachour, and P. Po¨tschke, J. Appl. Polym. Sci., 94, 1976 (2004). 72. H. Wu, S. Guo, G. Chen, and H. Shi, J. Appl. Polym. Sci., 94, 1917 (2004). 73. A. Colbeaux, F. Fenouillot, J.-F. Ge´rard, M. Taha, and H. Wautier, J. Appl. Polym. Sci., 95, 320 (2005). 74. F. O. M. S. Abreu, M. M. C. Forte, and S. A. Liberman, J. Appl. Polym. Sci., 95, 263 (2005). 75. F. Chen, R. A. Shanks, and G. Amarasinghe, J. Appl. Polym. Sci., 95, 1549 (2005). 76. A. Jalali-Arani, A. A. Katbab, and H. Nazockdast, J. Appl. Polym. Sci., 96, 155 (2005). 77. B. Patham, P. Papworth, K. Jayaraman, C. Shu, and M. D. Wolkowicz, J. Appl. Polym. Sci., 96, 423 (2005). 78. T. Wu, Y. Li, Q. Wu, G. Wu, and H.-M. Tan, J. Appl. Polym. Sci., 96, 261 (2005). 79. G. J. Nam, K. Y. Kim, and J. W. Lee, J. Appl. Polym. Sci., 96, 905 (2005). 80. S. Luyt and M. J. Hato, J. Appl. Polym. Sci., 96, 1748 (2005). 81. K. L. K. Lim, Z. A. Mohd. Ishak, U. S. Ishiaku, A. M. Y. Fuad, A. H. Yusof, T. Czigany, B. Pukanszky, and D. S. Ogunniyi, J. Appl. Polym. Sci., 97, 413 (2005). 82. J. Sook. Hong, J. L. Kim, K. H. Ahn, and S. J. Lee, J. Appl. Polym. Sci., 97, 1702 (2005). 83. S. Bertin and J.-J Robin, Eur. Polym. J., 38, 2255 (2002). 84. D. Gheysari, A. Behjat, and M. Haji-Saeid, Eur. Polym. J., 37, 295 (2001). 85. B. Raj, V. Annadurai, R. Somashekar, M. Raj, and S. Siddaramaiah, Eur. Polym. J., 37, 943 (2001). 86. J. Kim, J. H. Kim, T. K. Shin, H. J. Choi, and M. S. Jhon, Eur. Polym. J., 37, 2131 (2001). 87. X. Zhang, F. Xie, Z. Pen, Y. Zhang, Y. Zhang, and W. Zhou, Eur. Polym. J., 38, 1 (2002). 88. B. Jurkowski, Y. A. Olkhov, K. Kelar, and O. M. Olkhova, Eur. Polym. J., 38, 1229 (2002). 89. S. Glowinkowski, M. Makrocka-Rydzyk, S. Wanke, and S. Jurga, Eur. Polym. J., 38, 961 (2002). 90. P. Matzinos, V. Tserki, C. Gianikouris, E. Pavlidou, and C. Panayiotou, Eur. Polym. J., 38, 1713 (2002). 91. D.-H. Chen, L. Hong, X.-W. Nie, X.-L. Wang, and X.-Z. Tang, Eur. Polym. J., 39, 871 (2003). 92. H. Yordanov and L. Minkova, Eur. Polym. J., 39, 951 (2003). 93. S. Jose, A. S. Aprem, B. Francis, M. C. Chandy, P. Werner, V. Alstaedt, and S. Thomas, Eur. Polym. J., 40, 2105 (2004). 94. A. Krumme, A. Lehtinen, and A. Viikna, Eur. Polym. J., 40, 371 (2004). 95. X. Wang and X. Luo, Eur. Polym. J., 40, 2391 (2004). 96. A. Lahor, M. Nithitanakul, and B. P. Grady, Eur. Polym. J., 40, 2409 (2004).

Chapter 1 Overview of Polyolefin Blends

21

97. K. A. Moly, H. J. Radusch, R. Androsh, S. S. Bhagawan, and S. Thomas, Eur. Polym. J., 41, 1410 (2005). 98. M. C. Dasca˘lu, C. Vasile, C. Silvestre, and M. Pascu, Eur. Polym. J., 41, 1391 (2005). 99. A. C. Quental and M. I. Felisberti, Eur. Polym. J., 41, 894 (2005). 100. J. George, K. Ramamurthy, K. T. Varughese, and S. Thomas, J. Polym. Sci. B Polym. Phys., 38, 1104 (2000). 101. S. C. George, G. Groeninckx, K. N. Ninan, and S. Thomas, J. Polym. Sci. B Polym. Phys., 38, 2136 (2000). 102. M.-H. Kim, J. D. Londono, and A. Habenschuss, J. Polym. Sci. B Polym. Phys., 38, 2480 (2000). 103. Z.-G. Wang, R. A. Phillips, and B. S. Hsiao, J. Polym. Sci. B Polym. Phys., 38, 2580 (2000). 104. K. Lee, H. Kwag, B. Kim, G.-J. Kim, K.-H. Kim, and S. Choe, J. Polym. Sci. B Polym. Phys., 39, 218 (2000). 105. G. Radonjicˇ and Ivan sˇmit, J. Polym. Sci. B Polym. Phys., 39, 566 (2000). 106. H. Bashir, A. Linares, and J. L. Acosta, J. Polym. Sci. B Polym. Phys., 39, 1017 (2001). 107. C. Zhang, T. Mori, T. Mizutani, M. Ishioka, and Y. Cheng, J. Polym. Sci. B Polym. Phys., 39, 1741 (2001). 108. J. Lu and H.-J. Sue, J. Polym. Sci. B Polym. Phys., 40, 507 (2002). 109. Y.-T. Shieh, M.-S. Lee, and S.-A. Chen, J. Polym. Sci. B Polym. Phys., 40, 638 (2002). 110. J. Chatterjee and R. G. Alamo, J. Polym. Sci. B Polym. Phys., 40, 878 (2002). 111. C. Marco, E. P. Collar, S. Areso, and J. Ma Garcı´a-Martı´nez, J. Polym. Sci. B Polym. Phys., 40, 1307 (2002). 112. S. C. Tjong, S.-A. Xu, and Y.-W. Mai, J. Polym. Sci. B Polym. Phys., 40, 1881 (2002). 113. Y. Wang, Q. Fu, Q. Li, G. Zhang, K. Shen, and Y.-Z. Wang, J. Polym. Sci. B Polym. Phys., 40, 2086 (2002). 114. M. Moffitt, Y. Rharbi, J.-D. Tong, J. P. S. Farhina, H. Li, M. A. Winnik, and H. Zahalka, J. Polym. Sci. B Polym. Phys., 41, 637 (2002). 115. I. Fortelny´, D. Hlavata´, J. Mikesˇova´, D. Micha´lkova´, L. Potrokova´, and I. sˇloufova´, J. Polym. Sci. B Polym. Phys., 41, 609 (2002). 116. J. Finlay, M. J. Hill, P. J. Barham, K. Byrne, and A. Woogara, J. Polym. Sci. B Polym. Phys., 41, 1384 (2003). 117. M. F. Dı´az, S. E. Barbosa, and N. J. Capiati, J. Polym. Sci. B Polym. Phys., 42, 452 (2003). 118. A. Opdahl, R. A. Phillips, and G. A. Somorjai, J. Polym. Sci. B Polym. Phys., 42, 421 (2003). 119. K. Lee, S. E. Shim, B. H. Lee, S. U. Hong, and S. Choe, J. Polym. Sci. B Polym. Phys., 42, 1114 (2003). 120. X. Jiang, Y. Zhang, and Y. Zhang, J. Polym. Sci. B Polym. Phys., 42, 1181 (2003). 121. W. Jiang, D. Yu, L. An, and B. Jiang, J. Polym. Sci. B Polym. Phys., 42, 1433 (2003). 122. S. C. Domenech, J. H. Bortoluzzi, V. Soldi, C. V. Franco, W. Gronski, and H.-J. Radusch, J. Polym. Sci. B Polym. Phys., 42, 1767 (2003). 123. I. Banik, P. J. Carreau, and H. P. Schreiber, J. Polym. Sci. B Polym. Phys., 42, 2545 (2004). 124. C.-I Huang, C.-P. Chang, K. Shimizu, and C. C. Han, J. Polym. Sci. B Polym. Phys., 42, 2995 (2004). 125. C. Silvestre, S. Cimmino, and J. S. Lin, J. Polym. Sci. B Polym. Phys., 42, 3368 (2004). 126. E. Puukilainen and T. A. Pakkanen, J. Polym. Sci. B Polym. Phys., 43, 2252 (2005). 127. Z. Chen, T. Li, X. Liu, and R. Lu¨, J. Polym. Sci. B Polym. Phys., 43, 2514 (2005). 128. H. Wu, S. Guo, and Z. Li, J. Polym. Sci. B Polym. Phys., 43, 3030 (2005). 129. R. Pe´rez, E. Rojo, M. Ferna´ndez, V. Leal, P. Lafuente, and A. Santamarı´a, Polymer, 46, 8045 (2005). 130. S. Filippi, L. Minkova, N. Dintcheva, P. Narducci, and P. Magagnini, Polymer, 46, 8054 (2005).

22

Polyolefin Blends

131. C.-T. Lo, S. Seifert, P. Thiyagarajan, and B. Narasimhan, Macromol. Rapid Commun., 26, 533 (2005). 132. H. Hu, C. Chong, A. He, C. Zhang, G. Fan, J.-Y. Dong, and C. Han, Macromol. Rapid Commun., 26, 973 (2005). 133. X. Zhang, Z. Wang, M. Muthukumar, and C. C. Han, Macromol. Rapid Commun., 26, 1285 (2005). 134. J. Luettmer-Strathmann, J. Chem. Phys., 123, 014910–12 (2005). 135. M. Montoya, M. J. Abad, L. Barral Losada, and C. Bernal, J. Appl. Polym. Sci., 98, 1271 (2005). 136. S. Datta, K. Naskar, J. Jelenic, and J. W. M. Noordermeer, J. Appl. Polym. Sci., 98, 1393 (2005). 137. M. Pracella, D. Chionna, A. Pawlak, and A. Galeski, J. Appl. Polym. Sci., 98, 2201 (2005). 138. D. Xu, J. Wang, Z. Liu, Y. Ke, and Y. Hu, Macromol. Chem. Phys., 202, 1817 (2001). 139. M. A. Lo´pez-Manchado, J. M. Kenny, R. Quijada, and M. Yazdani-Pedram, Macromol. Chem. Phys., 202, 1909 (2001). 140. S. Filippi, V. Chiono, G. Polacco, M. Paci, L. I. Minkova, and P. Magagnini, Macromol. Chem. Phys., 203, 1512 (2002). ´ . Prieto, J. M. Peren˜a, R. Benavente, M. L. Cerrada, and E. Pe´rez, Macromol. Chem. Phys., 203, 141. O 1844 (2002). 142. H. Sato, S. Sasao, K. Matsukawa, Y. Kita, H. Yamaguchi, H. W. Siesler, and Y. Ozaki, Macromol. Chem. Phys., 204, 1351 (2003). 143. A. J. Mu˝ller, M. L. Arnal, A. L. Spinelli, E. Can˜izales, C. C. Puig, H. Wang, and C. C. Han, Macromol. Chem. Phys., 204, 1497 (2003). 144. Q. Wei, D. Chionna, and M. Pracella, Macromol. Chem. Phys., 206, 777 (2005). 145. J. Li, C.-M. Chan, B. Gao, and J. Wu, Macromolecules, 33, 1022 (2000). 146. M. L. D. Lorenzo, S. Cimmino, and C. Silvestre, Macromolecules, 33, 3828 (2000). 147. I. G. Economou, Macromolecules, 33, 4954 (2000). 148. Y. A. Akpalu, A. Karim, S. K. Satija, and N. P. Balsara, Macromolecules, 34, 1720 (2001). 149. H.-J. Kim, K.-J. Lee, Y. Seo, S. Kwak, and S.-K. Koh, Macromolecules, 34, 2546 (2001). 150. H. Wang, K. Shimizu, E. K. Hobbie, Z.-G. Wang, J. C. Meredith, A. Karim, E. J. Amis, B. S. Hsiao, E. T. Hsieh, and C. Han, Macromolecules, 35, 1072 (2002). 151. S. Costeux, P. Wood-Adams, and D. Beigzadeh, Macromolecules, 35, 2514 (2002). 152. A. Opdahl, R. A. Phillips, and G. A. Somorjai, Macromolecules, 35, 4387 (2002). 153. I. A. Hussein, Macromolecules, 36, 2024 (2003). 154. Y. Ma, J. P. S. Farinha, M. A. Winnik, P. V. Yaneff, and R. A. Ryntz, Macromolecules, 37, 6544 (2004). 155. L. Caporaso, N. Iudici, and L. Oliva, Macromolecules, 38, 4894 (2005). 156. F. M. Capaldi, G. C. Rutledge, and M. C. Boyce, Macromolecules, 38, 6700 (2005). 157. J. E. Wolak and J. L. White, Macromolecules, 38, 10466 (2005). 158. L. Minkova, M. Velcheva, and P. Magagnini, Macromol. Mater. Eng., 280-281, 7 (2000). 159. Z. Funke, C. Schwinger, R. Adhikari, and J. Kressler, Macromol. Mater. Eng., 286, 744 (2001). 160. G. Radonjicˇ and N. Gubeljak, Macromol. Mater. Eng., 287, 122 (2002). 161. B. Borisova and J. Kressler, Macromol. Mater. Eng., 288, 509 (2003). 162. J. Kolarˇ´ık, A. Pegoretti, L. Fambri, and A. Penati, Macromol. Mater. Eng., 288, 629 (2003). 163. M. Yazdani-Pedram, R. Quijada, and M. A. Lo´pez-Manchado, Macromol. Mater. Eng., 288, 875 (2003). 164. A. Genovese and R. A. Shanks, Macromol. Mater. Eng., 289, 20 (2004). 165. B. Guo, Y. Cao, D. Jia, and Q. Qiu, Macromol. Mater. Eng., 289, 360 (2004). 166. M.-B. Coltelli, S. Savi, I. D. Maggiore, V. Liuzzo, M. Aglietto, and F. Ciardelli, Macromol. Mater. Eng., 289, 400 (2004).

Chapter 1 Overview of Polyolefin Blends

23

167. Z.-M. Li, W. Yang, R. Huang, X.-P. Fang, and M.-B. Yang, Macromol. Mater. Eng., 289, 426 (2004). 168. S. Filippi, H. Yordanov, L. Minkova, G. Polacco, and M. Talarico, Macromol. Mater. Eng., 289, 512 (2004). 169. Z.-B. Chen, T.-S. Li, Y.-L. Yang, Y. Zhang, and S.-Q. Lai, Macromol. Mater. Eng., 289, 662 (2004). 170. U. Hippi, M. Korhonen, S. Paavola, and J. Seppa¨la¨, Macromol. Mater. Eng., 289, 714 (2004). 171. C. Li, Q. Kong, J. Zhao, D. Zhao, Q. Fan, and Y. Xia, Macromol. Mater. Eng., 289, 833 (2004). 172. G. Cazacu, M. Mihaies, M. C. Pascu, L. Profire, A. L. Kowarskik, and C. Vasile, Macromol. Mater. Eng., 289, 880 (2004). 173. Z.-M. Li, C.-G. Huang, W. Yang, M.-B. Yang, and R. Huang, Macromol. Mater. Eng., 289, 1004 (2004). 174. H. Kwag, D. Rana, K. Cho, J. Rhee, T. Woo, B. H. Lee, and S. Choe, Polym. Eng. Sci., 40, 1672 (2000). 175. R. Zacur, G. Goizueta, and N. Capiati, Polym. Eng. Sci., 40, 1921 (2000). 176. G. Flodberg, A. Hellman, M. S. Hedenqvist, E. R. Sadiku, and U. W. Gedde, Polym. Eng. Sci., 40, 1969 (2000). 177. E. Ramı´rez-Vargas, D. Navarro-Rodrı´guez, F. J. Medellin-Rodrı´guez, B. M. Huerta-Martı´nez, and J. S. Lin, Polym. Eng. Sci., 40, 2241 (2000). 178. J. D. Tucker, S. Lee, and R. L. Einsporn, Polym. Eng. Sci., 40, 2577 (2000). 179. H. Tang, B. Foran, and D. C. Martin, Polym. Eng. Sci., 41, 440 (2001). 180. M. A. Huneault, F. Mighri, G. H. Ko, and F. Watanabe, Polym. Eng. Sci., 41, 672 (2001). 181. I. A. Hussein and M. C. Williams, Polym. Eng. Sci., 41, 696 (2001). 182. K. Wang and C. Zhou, Polym. Eng. Sci., 41, 2249 (2001). 183. D. K. Setua, C. Soman, A. K. Bhowmick, and G. N. Mathur, Polym. Eng. Sci., 42, 10 (2002). 184. J. Kolarˇ´ık, L. Fambri, A Pegoretti, A. Penati, and P. Goberti, Polym. Eng. Sci., 42, 161 (2002). 185. A. Garcia-Rejon, A. Meddad, E. Turcott, and M. Carmel, Polym. Eng. Sci., 42, 346 (2002). 186. D. Roy, G. P. Simon, and M. Forsyth, Polym. Eng. Sci., 42, 781 (2002). 187. Y. C. Liang and A. I. Isayev, Polym. Eng. Sci., 42, 994 (2002). 188. Y. Son and R. A. Weiss, Polym. Eng. Sci., 42, 1322 (2002). 189. E. Ramı´rez-Vargas, F. J. Medelln-Rodrı´guez, D. Navarro-Rodrı´guez, C. A. Avila-Orta, S. G. Solı´sRosales, and J. S. Lin, Polym. Eng. Sci., 42, 1350 (2002). 190. J.-T. Yeh, and S.-S. Chang, Polym. Eng. Sci., 42, 1558 (2002). 191. F. Marguerat, P. J. Carreau, and A. Michel, Polym. Eng. Sci., 42, 1941 (2002). 192. M. L. Arnal, E. Can´izales, and A. J. Mu¨ller, Polym. Eng. Sci., 42, 2048 (2002). 193. C. Joubert, P. Cassagnau, A. Michel, and L. Choplin, Polym. Eng. Sci., 42, 2222 (2002). 194. I. Sendijarevic, A. J. McHugh, J. A. Orlicki, and J. S. Moore, Polym. Eng. Sci., 42, 2393 (2002). 195. C. Grein, C. J. G. Plummer, Y. Germain, H.-H. Kausch, and P. Be´guelin, Polym. Eng. Sci., 43, 223 (2003). 196. N. Kukaleva, G. P. Simon, and E. Kosior, Polym. Eng. Sci., 43, 431 (2003). 197. G. Flodberg, M. S. Hedenqvist, and U. W. Gedde, Polym. Eng. Sci., 43, 1044 (2003). 198. H. Padilla-Lopez, M. O. Va´zquez, R. Gonza´lez-Nu´n˜ez, and D. Rodrigue, Polym. Eng. Sci., 43, 1646 (2003). 199. W. Wang, Q. Wu, and B. Qu, Polym. Eng. Sci., 43, 1798 (2003). 200. G. Liu, M. Xiang, and H. Li, Polym. Eng. Sci., 44, 197 (2004). 201. N. Sombatsompop, K. Sungsanit, and C. Thongpin, Polym. Eng. Sci., 44, 487 (2004). 202. Y. Chen and H. Li, Polym. Eng. Sci., 44, 1509 (2004). 203. P. Rachtanapun, S. E. M. Selke, and L. M. Matuana, Polym. Eng. Sci., 44, 1551 (2004). 204. L. Canfora, S. Filippi, and F. P. La Mantia, Polym. Eng. Sci., 44, 1732 (2004).

24

Polyolefin Blends

205. M. A. Debolt and R. E. Robertson, Polym. Eng. Sci., 44, 1800 (2004). 206. M. H. Ha and B. Y. Kim, Polym. Eng. Sci., 44, 1858 (2004). 207. W. Feng and A. I. Isayev, Polym. Eng. Sci., 44, 2019 (2004). 208. Z.-M. Li, B.-H. Xie, R. Huang, X.-P. Fang, and M.-B. Yang, Polym. Eng. Sci., 44, 2165 (2004). 209. K. P. Tchomakov, B. D. Favis, M. A. Huneault, F. Champagne, and F. Tofan, Polym. Eng. Sci., 44, 749 (2004). 210. J.-T. Yeh, S.-S. Huang, and H.-Y. Chen, Polym. Eng. Sci., 45, 25 (2005). 211. Y. Fang, P. J. Carreau, P. G. Lafleur, and S. Ymmel, Polym. Eng. Sci., 45, 343 (2005). 212. V. Thirtha, R. Lehman, and T. Nosker, Polym. Eng. Sci., 45, 1187 (2005). 213. H.-S. Xu, Z.-M. Li, S. Y. Yang, J.-L. Pan, W. Yang, and M.-B. Yang, Polym. Eng. Sci., 45, 1231 (2005). 214. Y. Fang, P. J. Carreau, and P. G. Lafleur, Polym. Eng. Sci., 45, 1254 (2005). 215. F. P. La Mantia, L. Canfora, and N. Tzankova Dintcheva, Polym. Eng. Sci., 45, 1297 (2005). 216. S. S. Morye, Polym. Eng. Sci., 45, 1369 (2005). 217. S. S. Morye, Polym. Eng. Sci., 45, 1377 (2005). 218. Y. Matsuda, M. Hara, T. Mano, K. Okamoto, and M. Ishikawa, Polym. Eng. Sci., 45, 1630 (2005). 219. C. Lu, X. Yu, and S. Guo, Polym. Eng. Sci., 45, 1666 (2005). 220. F. Pazzagli and M. Pracella, Macromol. Symp., 149, 225 (2000). 221. J. Murı´n, J. Uhrin, and I. Choda´k, Macromol. Symp., 170, 115 (2001). 222. B. L. Lopez, C. Gartner, and M. Hess, Macromol. Symp., 174, 277 (2001). 223. D. Hlavata´, Z. Krulisˇ, Z. Hora´k, F. Lednicky´, and J. Hroma´dkova´, Macromol. Symp., 176, 93 (2001). 224. M. R. Arroyo and M. L. Bell, Macromol. Symp., 170, 181 (2001). 225. S. Bualek-Limcharoen, S. Saengsuwan, T. Amornsakchai, and B. Wanno, Macromol. Symp., 170, 189 (2001). 226. F. P. la Mantia, R. Scaffaro, A. Valenza, A. Marchetti, and S. Filippi, Macromol. Symp., 198, 173 (2003). 227. J. Pointeck, P. Po¨tschke, N. Proske, H. Zhao, H. Malz, D. Beyerlein, U. Schulze, and B. Voit, Macromol. Symp., 198, 209 (2003). 228. Q.-W. Lu, C. W. Macosko, and J. Horrion, Macromol. Symp., 198, 221 (2003). 229. Y. Roiter, V. Samaryk, S. Varvarenko, N. Nosova, I. Tarnavchyk, J. Pointeck, P. Po¨tschke, and S. Voronov, Macromol. Symp., 210, 209 (2004). 230. T. Trongsatikul, D. Aht-Ong, and W. Chinsirikul, Macromol. Symp., 216, 265 (2004). 231. H. H. Kausch, Macromol. Symp., 225, 165 (2005). 232. G. Guerrica-Echevarrı´a, J. I. Eguiazaba´l, and J. Naza´bal, J. Polym. Sci. B. Polym. Phys., 38, 1090 (2000). 233. F. Mighri, M. A. Huneault, A. Ajji, G. H. Ko, and F. Watanabe, J. Appl. Polym. Sci., 82, 2113 (2001). 234. M. G. Oliveira and B. G. Soares, J. Appl. Polym. Sci., 91, 1404 (2004). 235. J. Luettmer-Strathmann, J. Chem. Phys., 123, 014910 (2005). 236. D. Heine, D. T. Wu, J. G. Curro, and G. S. Grest, J. Chem. Phys., 118, 914 (2003). 237. R. Strapasson, S. C. Amico, M. F. R. Pereira, and T. H. D. Sydenstricker, Polym. Test., 24, 468 (2005). 238. A. Hassan, M. U. Wahit, and C. Y. Chee, Polym. Test., 22, 281 (2003). 239. C. Li, Y. Zhang, and Y. Zhang, Polym. Test., 22, 191 (2003). 240. S. H. El-Sabbagh, Polym. Test., 22, 93 (2003). 241. H. Ismail and Suryadiansyah, Polym. Test., 21, 389 (2002). 242. J.-Z. Liang, Polym. Test., 21, 69 (2002). 243. C. M. Tai, R. K. Y. Li, and C. N. Ng, Polym. Test., 19, 143 (2000). 244. Halimatuddahliana, H. Ismail, and H. Md. Akil, Polym. Plast. Technol. Eng., 44, 1429 (2005).

Chapter 1 Overview of Polyolefin Blends

25

245. Halimatuddahliana, H. Ismail, and H. Md. Akil, Polym. Plast. Technol. Eng., 44, 1217 (2005). 246. Halimatuddahliana and H. Ismail, Polym. Plast. Technol. Eng., 43, 357 (2004). 247. H. Ismail and Suryadiansyah, Polym. Plast. Technol. Eng., 43, 319 (2004). 248. A. P. Gupta, U. K. Saroop, and M. Verma, Polym. Plast. Technol. Eng., 42, 357 (2003). 249. H. Ismail and Suryadiansyah, Polym. Plast. Technol. Eng., 41, 833 (2002). 250. S. H. Botros, Polym. Plast. Technol. Eng., 41, 341 (2002). 251. A. A. Ol’khov, A. L. Iordanskii, G. E. Zaikov, and L. S. Shibryaeva, Polym. Plast. Technol. Eng., 39, 783 (2000). 252. T. D. Sreeja and S. K. N. Kutty, Polym. Plast. Technol. Eng., 39, 501 (2000). 253. B. T. Poh and G. H. Khok, Polym. Plast. Technol. Eng., 39, 151 (2000). 254. A. Hassan, M. U. Wahit, and C. Y. Chee, Polym. Plast. Eng., 44, 1245 (2005). 255. N. M. Abdel Monem, Z. I. Ali, H. M. Said, H. A. Youssef, and H. H. Saleh, Polym. Plast. Technol. Eng., 44, 1025 (2005). 256. I. Krupa and A. S. Luyt, Polym. Degrad. Stabil., 70, 111 (2000). 257. G. Luo, T. Suto, S. Yasu, and K. Kato, Polym. Degrad. Stabil., 70, 97 (2000). 258. I. A. Hussein, K. Ho, S. K. Goyal, E. Karbashewski, and M. C. Williams, Polym. Degrad. Stabil., 68, 381 (2000). 259. J. C. M. Suarez, E. B. Mano, and R. A. Pereira, Polym. Degrad. Stabil., 69, 217 (2000). 260. E. Ramı´rez-Vargas, D. Navarro-Rodrı´guez, A. I. Blanqueto-Menchaca, B. M. Huerta-Martı´nez, and M. Palacios-Mezta, Polym. Degrad. Stabil., 86, 301 (2004). 261. A. A. Basfar, K. M. Idriss Ali, and S. M. Mofti, Polym. Degrad. Stabil., 82, 229 (2003). 262. R. N. Darie, M. Brebu, C. Vasile, and M. Kozlowski, Polym. Degrad. Stabil., 80, 551 (2003). 263. W. Camacho and S. Karlsson, Polym. Degrad. Stabil., 78, 385 (2002). 264. R. A. Assink, M. Celina, K. T. Gillen, R. L. Clough, and T. M. Alam, Polym. Degrad. Stabil., 73, 355 (2001). 265. M. P. Anachkov, S. K. Rakovski, and R. V. Stefanova, Polym. Degrad. Stabil., 67, 355 (2000). 266. I. Vieira, V. L. S. Severgnini, D. J. Mazera, M. S. Soldi, E. A. Pinheiro, A. T. N. Pires, and V. Soldi, Polym. Degrad. Stabil., 74, 151 (2001). 267. T. Zaharescu, Polym. Degrad. Stabil., 73, 113 (2001). 268. R. Xie and B. Qu, Polym. Degrad. Stabil., 71, 375 (2001). 269. T. Zaharescu, V. Meltzer, and R. Vı˘lcu, Polym. Degrad. Stabil., 70, 341 (2000). 270. H. A. Khonakdar, S. H. Jafari, A. Yavari, A. Asadinezhad, and U. Wagenknecht, Polym. Bull., 54, 75 (2005). 271. T. Zaharescu and P. Budrugeac, Polym. Bull., 49, 297 (2002). 272. R. Gonza´lez-Nun˜ez, H. Padilla, D. De Kee, and B. D. Favis, Polym. Bull., 46, 323 (2001). 273. C. Albano, J. Trujillo, A. Caballero, and O. Brito, Polym. Bull., 45, 531 (2001). 274. C. G. In˜iguez, E. Michel, V. M. Gonza´lez-Romero, and R. Gonza´lez-Nun˜ez, Polym. Bull., 45, 295 (2000). 275. M. Alagar, S. M. Abdul Majeed, and S. Nagendiran, Polym. Adv. Technol., 16, 582 (2005). 276. J. K. Mishra and C. K. Das, Polym. Compos., 24, 83 (2003). 277. A. G. Pedroso and D. S. Rosa, Polym. Adv. Technol., 16, 310 (2005). 278. R. Adhikari, R. Godehardt, W. Lebek, S. Frangov, G. H. Michler, H.-J. Radusch, and F. J. Balta´ Calleja, Polym. Adv. Technol., 16, 156 (2005). 279. M. S. Kim and B. K. Kim, Polym. Adv. Technol., 15, 419 (2004). 280. D. Tomova and H.-J. Radusch, Polym. Adv. Technol., 14, 19 (2003). 281. J. K. Mishra, S. Roychowdhury, and C. K. Das, Polym. Adv. Technol., 13, 112 (2002).

26

Polyolefin Blends

282. R. R. N. Sailaja, Polym. Int., 54, 1589 (2005). 283. N. Abacha and S. Fellahi, Polym. Int., 54, 909 (2005). 284. Y. Lin, W. Du, D. Tu, W. Zhong, and Q. Du, Polym. Int., 54, 465 (2005). 285. I. A. Hussein, Polym. Int., 53, 1327 (2004). 286. N. Zeng, S. L. Bai, C. G’Sell, J.-M. Hiver, and Y. W. Mai, Polym. Int., 51, 1439 (2002). 287. A. S. Hashim and S. K. Ong, Polym. Int., 51, 611 (2002). 288. A. Tedesco, P. F. Krey, R. V. Barbosa, and R. S. Mauler, Polym. Int., 51, 105 (2001). 289. R. R. N. Sailaja, A. P. Reddy, and M. Chanda, Polym. Int., 50, 1352 (2001). 290. S. Chattopadhyay and S. Sivaram, Polym. Int., 50, 67 (2001). 291. J. K. Mishra, S. Raychowdhury, and C. K. Das, Polym. Int., 49, 1615 (2000). 292. P. Micic and S. N. Bhattacharya, Polym. Int., 49, 1580 (2000). 293. A. J. Marzocca, S. Cerveny, and J. M. Me´ndez, Polym. Int., 49, 216 (2000). 294. A. Maciel, V. Salas, and O. Manero, Adv. Polym. Technol., 24, 241 (2005). 295. F. J. Rodriguez-Gonzalez, N. Virgilio, R. A. Ramsay, and B. D. Favis, Adv. Polym. Technol., 22, 297 (2003). 296. M. Suresh Chandra Kumar and M. Alagar, Adv. Polym. Technol., 21, 201 (2002). 297. K. Wang, C. Zhou, H. Zhang, and D. Zhao, Adv. Polym. Technol., 21, 164 (2002). 298. J. H. Yeo, C. H. Lee, C.-S. Park, K.-J. Lee, J.-D. Nam, and S. W. Kim, Adv. Polym. Technol., 20, 191 (2001). 299. A. A. Adewole, A. Denicola, C. G. Gogos, and L. Mascia, Adv. Polym. Technol., 19, 180 (2000). 300. D. N. Saheb and J. P. Jog, Adv. Polym. Technol., 19, 41 (2000). 301. H. Huang, L. Yang, and X. Ni, J. Macromol. Sci. B Phys., 44, 137 (2005). 302. F. Herna´ndez-Sa´nchez, A. Manzur, and R. Olayo, J. Macromol. Sci. B Phys., 43, 1183 (2004). 303. A. K. Akinlabi, F. E. Okieimen, and A. I. Aigbodion, Polym. Adv. Technol., 16, 318 (2005). 304. J. Varga, A. Breining, and G. W. Ehrenstein, Int. Polym. Process., 15, 53 (2000). 305. C. Colletti, S. Piccarolo, and A. Valenza, Int. Polym. Process., 15, 46 (2000). 306. Y. Yu and J. L. White, Int. Polym. Process., 18, 388 (2003). 307. S. Pimbert, Int. Polym. Process., 19, 27 (2004). 308. Z. Hrnjak-Murgic, L. Kratofil, Z. Jelcic, J. Jelencic, and Z. Janovic, Int. Polym. Process., 19, 139 (2004). 309. H. Uehara, K. Sakauchi, T. Kanai, and T. Yamada, Int. Polym. Process., 19, 163 (2004). 310. H. Chen, U. Sundararaj, K. Handakumar, and M. D. Wetzel, Int. Polym. Process., 19, 342 (2004).

Chapter

2

Miscibility and Characteristics of Polyolefin Blends James L. White1 and Jinhai Yang1

2.1 INTRODUCTION It is our purpose here to summarize the characteristics of blends of polyolefins. We consider both blends of polyolefins with other polyolefins and with nonpolyolefins including polyamides and polystyrenes. In the case of interpolyolefin blends, our primary concern is with miscibility. In the case of blends of polyolefins with nonpolyolefins, the blends are all immiscible and our concern is with phase morphology. We also consider three component blend systems where the third component is a surfactant or compatibilizing agent, which collects at the interface.

2.1.1 Polyolefins Olefins are a group of unsaturated hydrocarbons of the general formula CnH2n. The polymers of olefins are known as polyolefins. Commercial polyolefins mainly include homopolymers of the olefin monomers: ethylene, propylene, butene-1, isobutene, and 4-methylpentene-1. These homopolymers have structure units as shown in Fig. 2.1, where the asterisk indicates asymmetric carbon atoms. Thus, polypropylene (PP), poly (butene-1) (PB1), and poly(4-methylpentene-1) (P4MP1) have different tactic forms. The most important commercial polyolefins are polyethylene, polyisobutene, and the isotactic forms, that is, iPP, iPB1, and iP4MP1. Polyisobutene was first polymerized by the IG Farbenindustries (BASF) in the late 1920s. Polyethylene was first polymerized by ICI in the late 1930s in a branched form (1). Linear polyethylene

1

Department of Polymer Engineering, The University of Akron, Akron, OH 44325, USA

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

27

28

Polyolefin Blends

CH2 CH2

PE

n

*

CH2 CH n CH3

PP

CH3 CH2 C n CH3

PIB

*

CH2 CH n CH2 CH3 PB1

*

CH2 CH n CH2 CH CH3 CH3 P4MP1

Figure 2.1 Structure units of polyolefins.

was developed by Ziegler and coworkers (2) and by Phillips Petroleum (3) in the 1950s. The other isotactic polyolefins were polymerized by Natta et al. (4,5) using modified Ziegler catalysts in the 1950s. It is a very interesting question why only these five types of polyolefins are commercialized, considering that there are many more types of olefin monomers. This is largely associated with their crystalline melting temperatures. Polyethylene, polypropylene, and polybutene-1 have reasonable melting temperatures, in particular, 135, 165, and 120 C, respectively. However, the melting temperatures decrease greatly with increase of the pendant group length in the isotactic polymers, for example, polypropylene (165 C), polybutene-1 (120 C), polypentene-1 (70 C), polyhexylene-1 (55 C), polyheptylene-1 (40 C), and polyoctylene-1 (38 C) (6,7). Introducing methyl groups onto the side group raises the melting temperatures, for example, poly(3-methylbutene-1) (300 C), poly(4-methylpentene-1) (235 C) (7). Polyisobutylene, which crystallizes with difficulty has become a significant synthetic rubber. In addition to the homopolymers, there are various polyolefin copolymers. Linear low density polyethylene (LLDPE) is probably the largest. This involves ethylene copolymerized variously with butene-1, hexene-1, octene-1, and 4methylpentene-1, and so on. Here the second comonomer content is around 5 mol%. In the 1990s, Exxon (8) and Dow Chemical (9,10) began to make polyolefin copolymers using new generation of Ziegler–Natta catalysts called metallocene catalysts. Polymers were also made with high comonomer contents and the density was as low as 0.86 g cc1 (compared to 0.83 g cc1 for amorphous polyolefins). It should be noted that these are random copolymers. Recently, Dow Chemical (11) made polyolefin block copolymers using ‘‘chain shuttling polymerization.’’ These block copolymers display melting temperature around 40 C higher than those of random copolymers of equivalent density. Ethylene–propylene copolymers, which usually contain roughly 50–75 mol% of ethylene, are also important polyolefin copolymers. These comonomers are elastomers and are defined as ethylene– propylene copolymer (EPM) or ethylene–propylene rubber. Terpolymers are commercially more important than EPM and are produced by polymerizing ethylene and propylene together with a diolefin, which are called ethylene–propylene diene terpolymer (EPDM). Since the late 1990s, another group of polyolefins, cyclopolyolefins, were developed (12,13). They are homo- or copolymers formed by the polymerization

Chapter 2 Miscibility and Characteristics of Polyolefin Blends

29

Figure 2.2 Norbornene form (I).

of norbornene derivatives of form (I) shown in Fig. 2.2. The bridged ring structure of norbornene does not readily fit into crystalline lattices, which together with comonomers make the resins completely amorphous with high transparency and high glass transition temperature, for example, 200 C with 60 mol% of norbornene (14). They are used in competition with polymers such as polycarbonate and polymethylmethacrylate for applications such as optical discs and lenses.

2.1.2 Blends Though more and more polymers have been synthesized since 1930s, however, they could not satisfy the fast growing millions of applications especially in electric, architecture, and automobile after 1960s. Thus, both blends and composites are compounded greatly to compensate the gap. So far, compounding is the most fast, economic, and efficient material development method for new applications. At present, about 36% of the synthetic resins are used in blends and about 39% in compounds with particle and fibers (15). Polymer blends date to the nineteenth century. In 1846, Parkes patented the first polymer blends of cis-1,4 polyisoprene (natural rubber) and trans-1,4 polyisoprene (gutta percha) (16). The modern era of polymer blending is thought to begin with the development of high impact polystyrene (HIPS) (17) and poly(acrylonitrile–butadiene–styrene) terpolymer (ABS) (18–21) in the 1950s and polyphenylene ether (PPE)/ polystyrene (PS) blends by General Electric (22) in the 1960s. Polymer blends can be divided into two groups: miscible and immiscible blends. Miscible blends are homogeneous and stable. Their properties tend to be intermediate. However, they are relatively few. Most polymer blends are immiscible. Their properties are strongly affected by their phase morphologies, which are decided by their viscosity, interfacial tension, and processing methods. In this review we will describe polyolefin blends. Many of these blends involve polar polymers with polyolefins.

30

Polyolefin Blends

These blends are immiscible and their interfaces are unstable. Special interfacial treatments are required to make them suitable as materials of commerce. A second group of blends are those in which the components are all polyolefins. These blends will be miscible or nearly so. Also in this paper the general questions of blend miscibility and interfacial characteristics will be treated.

2.2 POLYMER BLEND MISCIBILITY The question of polymer blend miscibility derives out of the nineteenth century development of thermodynamics and studies in the same periodic of binary mixtures of low molecular weight liquids. From a thermodynamic viewpoint, Gibbs (23) formulated the stability of multiphase systems in terms of the quantity G defined by (in modern notation (24)) G ¼ U þ PV TS ¼ H TS

ð2:1Þ

where U is the internal energy, P is the pressure, V is the volume, T is the absolute temperature, and S is the entropy. According to Clausius (25), equilibrium was associated with the maximization of Suniv , the entropy of the universe. Gibbs showed that this was equivalent to G for a system at constant temperature and pressure to reach a minimum. Thus, the formation of a solution requires DG < 0

ð2:2Þ

The quantity, DG, known today as the Gibbs free energy may be expressed DG ¼ DH TDS

ð2:3Þ

There is an extensive literature on modeling DH, the enthalpy or heat of solution, and DS, the entropy of solution. The simplest situation involves isotropic molecules with no energetic interactions. We may consider an ensemble of systems of equal numbers of component molecules and of energy (microcanonical ensemble). The entropy of mixing involves Boltzmann’s formulation of entropy as (24,26) S ¼ kb ln V

ð2:4Þ

where kb is the Boltzmann constant and V is the number of configuration. Considering N1 the number of isotropic molecules of molecule ‘‘1’’ and N2 the number of isotropic molecules of molecule ‘‘2’’ randomly distributed on a lattice, this leads to V¼

ðN1 þ N2 Þ! N1 !N2 !

ð2:5Þ

On the basis of Stirling’s approximation, as long as N is large the log of factorial, N!, is given by ln N! ¼ N ln N N

ð2:6Þ

Chapter 2 Miscibility and Characteristics of Polyolefin Blends

31

The entropy of mixing is then DSm ¼ kb ðN1 þ N2 Þ½x1 ln x1 þ x2 ln x2

ð2:7Þ

where x1 ¼

N1 N2 ; x2 ¼ N1 þ N2 N1 þ N2

ð2:8Þ

The second problem, which needed to be resolved was the heat of solution, DH. For solutions with very similar molecules such as benzene and toluene, or hexane and octane, the heat of mixing is near zero. For various pairs of molecules, for example, acetone and chloroform, significant heat is evolved (DH < 0) and the solution process is said to be exothermic. For other pairs of molecules, heat is absorbed (DH > 0) and the solution process is called endothermic. As early as 1906, van Laar (27) modeled the heat of mixing of a binary solutions using van der Waals equation of state. The views of van Laar were subsequently expanded by van Laar and Lorenz (28), Hildebrand (29,30), and Scatchard (31). The latter authors considered a mixtures of two fluids each with isotropic molecules with interaction energies, c11 and c22 . The interaction energy between the two fluid molecules is c12 . Scatchard (31) obtained the form DUm ¼ Um U1 x1 U2 x2 ¼ Vx1 x2 ðc11 þ c22 2c12 Þ where V is the mixture volume. He further presumed pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ c12 ¼ c11 c22

ð2:9Þ

ð2:10Þ

which leads to 1=2

1=2

DUm ¼ Vx1 x2 ðc11 c22 Þ2 Hildebrand (32) defined ‘‘solubility parameters’’ of liquids as DUi 1=2 1=2 ; i ¼ 1; 2 di ¼ cii ¼ Vi

ð2:11Þ

ð2:12Þ

Thus, Equation 2.11 is equivalent to DUm ¼ Vx1 x2 ðd1 d2 Þ2

ð2:13Þ

Hildebrand (32) also determined experimental values for various liquids. Typical values are listed in Table 2.1 taken from Hildebrand and Scott (33). A major problem with Equation 2.13 is that it leads to the heat of mixing being positive, that is, always endothermic, absorbing heat. Many solutions are exothermic, that is, heat is evolved when the solution is formed. There have been various efforts to model exothermic heats of mixing. These are generally dated to Dolezalek (34) who considered dissolution to represent processes similar to chemical reactions. The modeling of the formation of solutions that we have described assumes that there is no change in volume in mixing. Theories of liquids that consider cell models

32

Polyolefin Blends Table 2.1 Solubility Parameters of Organic Liquids at 25 C. Solubility parameter, (J cm3)1/2

Substance n-Hexane n-Octane n-Hexadecane Cyclohexane Benzene Ethyl benzene Naphthalene Fluorinated hexane (C6F14) Fluorinated octane (C8F18) Fluorinated cyclohexane (C6F12) Fluorinated benzene (C6F6) Dimethyl ether Carbon disulfide Chloroform

7.30 7.55 8.0 8.20 9.15 8.80 9.9 5.6 5.7 6.0 8.1 8.8 10.0 9.3

Adapted from Reference 33.

volume changes begin with Lennard-Jones and Devonshire (35) in 1937. This approach was extended to solutions by Prigogine and Garikian (36) and LonguetHiggins (37) in the 1950s. These developments up to the mid 1950s are described by Prigogine in his book ‘‘The Molecular Theory of Solutions’’ (38). Volume change in mixing can affect the change in DG and induce phase separations which are not predicted from Equations 2.7 and 2.13. Early investigations of the thermodynamics of polymer solutions indicated that solutions exhibited large deviations from ideality. Calorimetric measurement showed that heats of solution were small indicating that these deviations were due to the entropy of mixing. Subsequently, Meyer (39) proposed that the low entropy of mixing was associated with the reduced number of configurations available to polymer chains compared with low molecular weight molecules. More quantitative formulations were subsequently developed by Huggins (40) and notably Flory (41–43). The entropy of mixing DSm was shown to have the form DSm ¼ kb ðN1 þ N2 Þ½x1 ln f1 þ x2 ln f2

ð2:14Þ

where f1 and f2 are the volume fractions of solvent and polymer. The heat of mixing DHm in polymer solutions that has been expressed by Flory (43) is DHm ¼ xkb TN1 f2

ð2:15Þ

where x characterizes the interaction energy per solution molecule divided by kb T. According to Flory x may be positive or negative. The free energy of solution, DGm , may be expressed DGm ¼ kb TðN1 ln f1 þ N2 ln f2 þ xN1 f2 Þ

ð2:16Þ

Chapter 2 Miscibility and Characteristics of Polyolefin Blends

33

This formulation was subsequently extended to polymer blends by Scott (44) in 1949 as follows: V f1 f2 ð2:17Þ ln f1 þ ln f2 þ xf1 f2 DGm ¼ kb T M2 V s M1 where M1 and M2 are the numbers of segments of the two polymers, V is the total volume of mixture, and Vs is the volume of segments, which usually are regarded as the volume of repeating units of the polymers. The formulation of Scott (44) does not present the range of phenomena occurring in polymer blends. Various binary blends exhibit lower critical solution temperatures (LCST) where phase separations occur at lower temperature. Other blends exhibit upper critical solution temperatures (UCST) where miscible blends exhibit phase separations at higher temperatures (45). It was shown by McMaster (46) that volume changes occurred in mixing.

2.3 INTERFACES IN LIQUID AND POLYMER MIXTURES The concept of surface tensions between air and liquids seems to date back to Thomas Young (47). Measurements of surface tensions of liquids (48) and its relationship to waves on liquids (49) and the breakup of capillary jets and viscous filaments (50,51) occur throughout the nineteenth century. In thermodynamics point of view, the surface tension is the increase of Gibbs free energy as a result of creating a unit surface area. The surfaces of liquids can be easily understood as an elastic film that inherently tends to shrink to decrease its surface. A summary of surface tensions of various low molecular weight liquids is contained in Table 2.2. It can be seen that the highest values are for molten metals. For organic liquids, the highest values are for polar compounds. Aromatic and aliphatic compounds are similar and the lowest values are for fluorinated compounds. It was believed that the intermolecular interactions play similar role in determining cohesive energy density and surface tension. The surface tension was suggested to be estimated by the cohesive energy density based on the following equation (53) Vð298KÞ 4 2=3 ð2:18Þ gðTÞ ¼ 0:75 ecoh VðTÞ where g is the surface tension, ecoh is the cohesive energy density, V is the molar volume, and T is the absolute temperature. However, Equation 2.18 cannot be used to evaluate the polymers containing polar units such as amide, urea, and hydroxyl groups, which form strong hydrogen bonds because the effect of hydrogen bonding on surface tension is generally smaller than its effect on cohesive energy density (54). For high molecular weight materials, the surface tensions of these materials has been found to vary with their molecular weight roughly as (55) g ¼ g1

Ke M 2=3

ð2:19Þ

34

Polyolefin Blends Table 2.2 Surface Tensions of Low Molecular Weight Compounds. Surface tension, dyn cm1

Material Iron Gold Silver Tin Lead Mercury Water (H2O) Ethylene glycol Epsilon-caprolactam Phthalic acid Lactic acid Styrene Acrylic acid Methyl methacrylate Benzene Cyclohexane Hexafluorobenzene n-Octane n-hexane Octafluoro-2-butene Decafluorobutane

Temperature, C

1962 1083 882 540 442 318 74.5 50.7 47.1 42.8 39.5 32.6 29.0 26.1 28.8 25.2 22.6 21.5 18.5 12.7 11.8

1200 1200 1200 455 455 455 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20

Adapted from Reference 52.

where g 1 is the surface tension at infinite molecular weight, Ke is a constant, and M is the molecular weight. The surface tension of polymers seems to vary with temperature approximately as (56) g ¼ g 0 ð1 T=Tc Þ11=9

ð2:20Þ

where g 0 is the surface tension at T ¼ 0K and Tc is the critical temperature whose value is around 1000 K for most polymers. The temperature coefficient of surface tension is given by differentiation of Equation 2.20 to be dg=dT ¼ ð11=9Þðg 0 =Tc Þð1 T=Tc Þ2=9

ð2:21Þ

Thus, the coefficient is practically a constant at ordinary temperature, that is, T Tc ¼ 1000 K. The surface tensions of some polymers are listed in Table 2.3. It came to be realized that interfaces between low molecular weight liquids in immiscible liquid mixtures possess an interfacial tension. Experimental measurements of interfacial tension between low molecular weight liquids were first made in the nineteenth century (59,60). Various experimental techniques have been developed to measure interfacial tensions. A popular method has been the pendant drop measurement, which was originally developed by Andreas et al. (61)

Chapter 2 Miscibility and Characteristics of Polyolefin Blends

35

Table 2.3 Surface Tension (g) of Polymers at Different Temperature in dyn cm1 .

Polymer Linear polyethylene Atactic polypropylene Isotactic polypropylene Isotactic polypropylene

Molecular weight g mol1 Mw ¼ 67; 000 — — Mw ¼ 267; 000

g g (20 C) (140 C)

—

25.5

Polystyrene Polystyrene Poly(ethylene terephthalate)

Mn ¼ 44; 000 Mn ¼ 9290 Mn ¼ 25; 000

40.7 44.6

28.8 22.7 23.1 19.7 (210 C) 19.2 (180 C) 32.1 31.5 36.7

Poly(ethylene oxide) Poly(methyl methacrylate) Poly(e-caprolactam)

M ¼ 6000 Mn ¼ 3000 —

42.9 41.1

33.8 32.0

Poly(hexamethylene adipamide) Bisphenol-A-polycarbonate Poly(dimethyl siloxane) Polytetrafluoroethylene

Mn ¼ 19; 000

46.5

— Mn ¼ 75; 000 M ¼ 1088

42.9 20.9 21.5

30.2 (270 C) 35.7 14.3 13.7

Polybutene-1

35.7 29.4 30.1 30.0

g (180 C)

References

26.5 20.4 20.8 19.1 (220 C) 17.5 (220 C) 29.2 28.8 34.2

(55, 56) (55, 56) (55, 56) (57)

30.7 28.9 36.1 (265 C) 28.3 (300 C) 33.3 12.3 11.1

(55, 56) (55, 56) (55, 56)

(57) (55, 56) (58) (55, 56)

(55, 56) (55, 56) (55, 56) (55, 56)

Adapted from References 55–58.

for measuring surface tension. Another method involves the breakup of molten polymer threads. Interfacial tensions of some polymer pairs are listed in Table 2.4. Measurements of interfacial tensions of polymer melts were reviewed by Wu (55), Koberstein (65), and Demarquette (66). The measurements usually need long equilibrium time because of the high viscosities of polymer melts. The measurements can be divided into two groups: static methods in which interfacial tension is calculated based on the equilibrium profile of the drops and dynamic methods that study the evolution of fiber or drop profiles with time. Static methods include pendant drop method, sessile drop method, and rotating drop method. Dynamic methods include breaking thread method, imbedded fiber method, and deformed drop retraction method. Interfacial tension plays a very important effect in controlling the phase morphology of immiscible polymer blends. In mixing of polymer blends, the minor phase is deformed into long fibrils or thin films by shear or elongation flows, and then is broken into small particles. At the same time, the dispersed particles can collide and merge together, which is called coalescence process. The final morphology is decided by the balance between breakdown and

36

Polyolefin Blends

Table 2.4

Interfacial Tensions of Molten Polymer Blends in dyn cm1 .

Polymer A

Polymer B

Polyethylene Polyethylene Polyethylene

Polystyrene Polysulfone Poly(phenylene sulfone) Poly(ethylene terephthalate) Polycarbonate Polyamide 6 Polyethylene Polystyrene Polycarbonate Polyamide 6

Polyethylene Polyethylene Polyethylene Polypropylene Polypropylene Polypropylene Polypropylene

Interfacial tension, dyn cm1

Temperature, C

References

4.0 7.0 7.9

290 290 290

(62) (62) (62)

9.4

290

(62)

12.5 12.8 1.63 4.5 8 12.2

290 290 290 290 290 290

(62) (62) (63) (64) (64) (64)

Adapted from References 62–64.

coalescence. Large interfacial tension resists this process in the following three ways. First, larger interfacial tension indicates lower interfacial adhesion, which prevents passing shear stress from one phase to another phase. Second, the interface with large tension is like an elastic film as we mentioned earlier. Inherently, it tends to shrink the deformed phases into spherical shape. Thus, it resists both the shear or elongation deformation of the dispersed phase. Third, large interfacial tension increases the tendency to coalescence. The low interaction between the two phases leads the matrix molecules easy to be excluded when two dispersed particles collide together. Also, large interfacial tension accelerates the merging process of the collided particles.

2.4 POLYOLEFIN–POLYOLEFIN BLENDS There is a history of investigations of blends of polyolefins. Many of these blends were not produced on purpose but were the results of incompletely understood polymerization processes. Examples of these are the various studies of elastomeric polypropylenes, which are mixtures of polypropylenes of varying tacticity levels (67–71). These materials often have interesting technological properties. It is of more concern to consider the results of specially prepared blends of known characterized polymers.

2.4.1 Blends between Polyethylenes Hundreds of polyethylenes are commercially available, and their blends have been used to provide desired range of properties. The blends have high level of mis-

Chapter 2 Miscibility and Characteristics of Polyolefin Blends

37

cibility. As mentioned earlier, LLDPE is a copolymer of ethylene with other olefins, for example, butene, hexene, and octene. Thus, it includes many short branches. LDPE is synthesized by free radical polymerization at high pressure. It includes both short and long chains. Thus, except molecular weight, both branch type and branch content are the two most important molecular structures of polyethylenes. Although the blends of polyethylenes are nearly miscible, their miscibility has been studied extensively. Three methods have mainly been used to study the miscibility. Hill et al. (72–75) mainly used differential scanning calorimetry(DSC) combined with transmission electron microscope (TEM). The melts of the blends are quenched into acetone at freezing point. If DSC shows only one melting peak and there is only one group of crystal morphology based on TEM, the blend is thought miscible in the melt. If DSC shows double melting peak and there are two groups of crystal structures, the blend is thought to be separated in the melt. Hussein et al. (76– 79) used rheology method. They measured both the dynamic viscosity and viscosity at zero shear rate of the blends. If the two parameters of the blends show log-additive linear relationship based on blend content, the blends are considered miscible. If they deviate from the linearity, the blends are thought immiscible. Alamo et al. (80–82) used small-angle neutron scattering (SANS). One component of the blends is deuterated. It was concluded that deuterating can increase the interaction parameter. In some cases, it is enough to lead to phase separation. Tashiro et al. (83–86) argued that they found phase-segregated structures between polyethylene and deuterated polyethylene. These studies suggest the following conclusions regarding the miscibility of polyethylene blends. (1) The branch content is the most important factor to control the miscibility. The miscibility decreases with increase of branch content. (2) The branch type also affects the miscibility. The miscibility decreases with increase of branch length. For example, LLDPE with 4.4 mol% of butene is miscible with high density polyethylene (HDPE) at all compositions (82). However, LLDPE with 4 mol% of hexene is immiscible with HDPE when the HDPE fraction is less than 60 wt% (87), and LLDPE with 2.1 mol% of octene is claimed immiscible with HDPE when HDPE fraction is less than 50 wt% (73). It is hard to deduce the effects of long-branch chains in LDPE. (3) Decreasing molecular weight increases the miscibility. For example, in a study of Hill et al. (72), HDPE with molecular weight 105 g mol1 shows phase separation with the LDPE with 5.2 mol% of branches when the HDPE fraction is lower than 50 wt%. However, HDPE with molecular weight 2553 g mol1 is miscible with LDPE at all HDPE composition. (4) Most of the blends possessed an UCST. Thus, the blends become more miscible with the increase of melt temperatures. (5) Usually, the blends become more miscible with the increase of linear polyethylene content. (6) In summary, the miscibility between polyethylenes is dependent on molecular structures (molecular weight, branch content, and type), temperature, and composition. Thus, phase diagrams based on the composition and the temperature are necessary to characterize the miscibility of specific polyethylene blends.

38

Polyolefin Blends

2.4.2 Blends between Isotactic Polypropylene and Ethylene Propylene Copolymers There have been extensive applications of isotactic polypropylene (iPP)/EPM blends. These were used to produce rubber toughened polypropylene blends and subsequently polyolefin thermoplastic elastomers (88,89). Most commercial EPMs contain more than 50 mol% of ethylene, and these are elastomers. The solubility parameter of EPM should be intermediate to those of polyethylene and polypropylene dependent on ethylene content. Thus, it is often used to compatibilize PE/PP blends (90,91). IPP–EPM blends were generally considered to be immiscible (92–95). Chen et al. (96) have reported miscibility and a LCST in blends of isotactic polypropylene and an ethylene-propylene terpolymer (EPDM). Seki et al. (97) reported that iPP was miscible with EPM with both 19 and 47 mol% of ethylene based on both transmission electron microscopy (TEM) and SANS. However, the molecular weight of the iPP was below 10; 000 g mol1. More recently both Nitta et al. (98) and Kamdar et al. (99) reported that the miscibility between iPP and EPM, whose molecular weight is higher than 100; 000 g mol1 , is strongly dependent on ethylene content in EPM. The miscibility increases with decrease of ethylene content. IPP was miscible with EPM with ethylene content lower than around 17 mol% based on both dynamic mechanical analysis (DMA) and TEM. The EPM domain sizes decrease with decrease of ethylene content as shown in Fig. 2.3.

Figure 2.3 TEM micrographs for the iPP/ethylene–propylene copolymer (80/20) blend films. The propylene contents by mol% in the copolymers are (a) 89.3; (b) 84.3; (c) 76.5; (d) 52.5. (From Reference 98 with permission from Elsevier.)

Chapter 2 Miscibility and Characteristics of Polyolefin Blends

39

2.4.3 Blends between iPP and High Comonomer Concentration Polyethylene Copolymers Since about 1990, ethylene copolymers with substantially high level of olefin comonomers (e.g., butene, hexene, and octene) have become available (100–104). Compared with most commercial EPM, they are available in a pellet form that facilitates continuous processing with twin screw extruders. These copolymers are becoming widely used to substitute for EPM to toughen iPP (105–107). The miscibility between these copolymers and iPP is also strongly dependent on the comonomer content. In a 1996 paper, Yamaguchi et al. (108) of Tosoh studied the miscibility of isotactic polypropylene with ethylene–hexene-1 (EHR) and ethylene–butene-1 (EBR) copolymers. It was found that iPP was miscible with ethylene–hexene-1 copolymers with more than 50 mol% of hexene based on both TEM and DMA (as shown in Figs. 2.4 and 2.5). However, though DMA indicated miscibility between iPP and ethylene– butene-1 copolymers with 56 and 62 mol% of butene-1 (as shown in Fig. 2.4), TEM suggested that there were tiny particles in the blends (as shown in Fig. 2.5). Thomann et al. (109) showed that ethylene–butene-1 copolymers are miscible with iPP when the butene-1 content was higher than 78 mol%. Yamaguchi et al. subsequently investigated the mechanical properties of blends of polypropylene with ethylene–hexene copolymers (110,111). Interestingly

Figure 2.4 Variation of the mechanical loss modulus (E00 ) with temperature for the iPP (O), the copolymers: EHR57 (with 57 mol% of hexene-1); EBR56 (with 56 mol% of butene-1); EBR62 (with 62 mol% of butene-1) (D), their blends (.). (From Reference 108 with permission from John Wiley & Sons, Inc.)

40

Polyolefin Blends

Figure 2.5 TEM micrographs of thin sections of series of blends with 50wt% of iPP. EHR57 has 57 mol% of hexene-1; EBR56 has 56 mol% of butene-1; EBR62 has 62 mol% of butene-1. (From Reference 108 with permission from John Wiley & Sons, Inc.)

Yamaguchi and Miyata (112) found a lack of miscibility of these copolymers with syndiotactic polypropylene (as shown in Fig. 2.6).

2.4.4 Blends between iPP and PB1 Based on the research of Thomann et al. (109), iPP is miscible with ethylene–butene copolymer with 78 mol% of butene. It is an interesting question whether iPP is miscible with PB1. For iPP/PB1 (30/70 by weight) blends made by melt mixing, Cham et al. (113) reported that liquid–liquid demixing happened at 180, 220, and 250 C. The demixing

Chapter 2 Miscibility and Characteristics of Polyolefin Blends

41

Figure 2.6 Transmission electron micrographs for (a) iPP/EHR and (b) sPP/EHR. Hexene-1 content is 57.1 mol%. The samples were stained with RuO4. (From Reference 112 with permission from American Chemical Society.)

process was observed by optical microscope, which indicates that the refractive index difference between iPP and PB1 is large enough for microscopy observation. Bartczak et al. (114) also reported that PB1 particles were found to disperse in the iPP spherulites when the blends were isothermally crystallized at 125 C. These suggest that the blends are immiscible at least up to 250 C. Cham et al. (113) found some evidence of partial miscibility. The demixed blends were isothermally crystallized at 145 C. It was found that the crystals were nucleated in the iPP domains, then, they gradually extended to all the domains. Surprisingly they continued to grow into the continuous PB1 phase with low density. When the crystals growing in PB1 phase reach some iPP domains, they can nucleate the crystal grow in those domain. Since at 145 C PB1 cannot crystallize, the crystals growing in PB1 phase must be PP crystals. This indicates that in the continuous PB1 phase there are iPP molecules, indicating that they are partially miscible. The authors also showed that the iPP crystals growing in PB1 phase have lower melting temperatures than those growing in iPP domains. It was reported that iPP/PB1 blends showed only one tand peak in a DMA test (113,115,116). This was usually taken as a proof for the miscibility between the amorphous iPP and PB1. However, a phase separation was reported at 250 C by Cham et al. (113). They suggested an explanation in their paper. First, the difference of Tg of iPP and PB1 is not large, around 25 C. Second, the separated phases are not pure iPP or PB1 phases, but iPP-rich and PB1-rich phases. Thus, the difference of the Tg of the two phases should be less than 25 C. The small Tg difference may make the two tand peaks merge together. Here, we also want to mention another point based on the study of Hsu and Geil (116), where blends with iPP content lower than 30 wt% show only one peak, but, those with higher iPP content show two peaks (as shown in

42

Polyolefin Blends

Figure 2.7 DMA curves of iPP/PB1 blends. (From Reference 116 with permission from John Wiley & Sons, Inc.)

Fig. 2.7). First, the tand peak of PB1 is much larger than that of iPP. Thus, the iPP peak in the blends with low iPP content is easy to hide, but the PB1 peak in the blends with low PB1 content cannot. It was reported by Hsu et al. (116) and Lee et al. (117) that iPP can nucleate the crystallization of PB1 to increase its crystallization temperature, and PB1, especially in the blends with high content of PB1 more than 70 wt%, can obviously retard the crystallization of iPP. On the basis of the publications available so far, we can make the following conclusions: (1) iPP and PB1 are partially miscible and phase separation occurs at reported temperatures up to 250 C. Melt mixed blends should have a two-phase morphology. (2) The blends made by precipitation from dilute solution show homogeneous mixing. But they are in a metastable state and tend to demix at high temperature.

2.5 BINARY IMMISCIBLE BLENDS There is a large literature involving binary blends of polyolefins, especially polyethylene and isotactic polypropylene, with other polymers. Among the polymers included in these studies are polystyrenes and polyamides. Polyolefins contribute to increase toughness, processability, chemical resistance, and moisture absorption resistance to these polymers. The other polymers contribute high modulus, heat resistance, and oxygen or solvent barrier properties to polyolefins. However, these blends are all immiscible. Useful polymer pairs generally require a compatibilizing agent as described in Section 2.6.

Chapter 2 Miscibility and Characteristics of Polyolefin Blends

43

Figure 2.8 Scanning electron microscopy micrographs of PS/PP (70/30) blend prepared at 190 C.

2.5.1 Polyolefin–Polystyrene Blends Polyolefin–polystyrene blends have long been studied. They are immiscible showing two-phase morphologies. The blends showed poor mechanical properties, especially elongation at break and impact strength, much lower than those predicted based on an additive rule (118). Their fracture surfaces were observed by electron microscope (119,120). As shown in Fig. 2.8, the dispersed phase is easy to be pulled from the matrix and leaves very smooth surface, indicating low interfacial adhesion. Because of the high interfacial tension, the morphology of the blends is not stable. Coalescence readily occurs in the molten state. As suggested by Macosko et al. (121), in melt mixing of immiscible polymer blends, the disperse phase is first stretched into threads and then breaks into droplets, which can coalesce together into larger droplets. The balance of these processes determines the final dispersed particle sizes. With increase of disperse phase fraction (usually more than 5 wt%), the coalescence speed increases and the dispersed phase sizes increase (121–123).

2.5.2 Polyolefin–Polyamide Blends Polyolefin–polyamide melt blends are striking in not only the lack of miscibility, but also the large interfacial tensions between the two melt phases. Investigations of these phenomena in our laboratories (118,124–126) have made numerous studies of these polymer blend systems and found that their phase morphology are quite unstable and trend to coalesce especially under quiescent or low deformation rate conditions. Similar to polyolefin–polystyrene blends, they also show weak interfacial adhesion (118,124,127) (as shown in Fig. 2.9). The mechanical properties of the

44

Polyolefin Blends

Figure 2.9 Scanning electron microscopy micrographs of PA12/PP (70/30) blend prepared at 190 C.

blends are also much lower than those predicted based on the additive rule (125). The disperse phase particles increase in size with increase of concentration (126,128).

2.6 TERNARY BLENDS OF POLYOLEFINS WITH OTHER POLYMERS AND COMPATIBILIZING AGENTS There is a long history of the use of surfactants with a mixture of low molecular weight molecules. These were used as soaps in the preparation of emulsion and other applications. This science and technology were highly developed in the 1920s. The use of high molecular weight surfactants dates back to the 1950s and 1960s and was only studied scientifically in the 1970s.

2.6.1 Surfactants and Compatibilizing Agents The term surfactant has the same meaning with the term surface active agent. Actually it is a blend of ‘‘surface active agent.’’ It has been extensively accepted to substitute the term surface active agent since the early 1960s. Surfactants are usually organic amphipathic compounds with both hydrophobic and hydrophilic groups. They can lower the surface tension of liquids to allow them to spread easier. Their earlier name ‘‘surface active agent’’ was given because of this characteristic. Also they can lower the interfacial tension between two liquids. This characteristic leads them to be applied extensively to make emulsions. Soap is the earliest surfactant. The earliest soap was made by boiling fats together with ashes of plant, which dated as early as 2800 B.C. in ancient Babylon. In 1823, Michel Eugene Chevreul, a French chemist, worked out the structure of fats

Chapter 2 Miscibility and Characteristics of Polyolefin Blends

45

as compounds of organic acids and glycerol. He showed that alkalis converted fats into soaps. He isolated stearic acid and various other organic fatty acids. He sought to commercially develop stearic acid. This was the beginning of the scientific studies of modern soaps and surfactants. From then on more and more surfactants were developed and were used in more applications, for example, detergents, fabric softeners, emulsifiers, paints, and so on. Though surfactants were mainly used to stabilize the immiscible small molecular mixtures, they were also reported in more recent to be used to make immiscible polymer blends more compatible (129–132). They could suppress the coalescence of dispersed particles to decrease their sizes, but could not increase the interfacial adhesion to increase the mechanical properties. The term ‘‘compatibilizing’’ was defined by Gaylord in 1966 (133) as ‘‘rendering a mixture of two or more polymeric materials permanently miscible so as to form a homogeneous composition that has useful plastic properties and which does not separate into its component parts.’’ Actually, in his definition the term ‘‘miscible’’ was not appropriate and another term ‘‘uniform’’ would be better. A few years later, he defined another term ‘‘compatibilization’’ as ‘‘absence of separation or stratification of the components of the polymeric alloy during the expected useful lifetime of the product’’ (134). Different from surfactants, substance considered as compatibilizing agents are today considered not only to decrease the interfacial tension, but also to enhance mechanical behavior presumably by increasing the interfacial adhesion. They must be high molecular weight compounds or form high molecular weight compounds in the blends. They stay on the interface with the part of their molecular chain in one component of the blends and the other part of chain in the other component. Though it was usually thought that the most efficient compatibilizing agents are block copolymers of the two polymers in the blends, however, there are only few block copolymers commercially available. Thus, the most applicable way is to use some commercially available polymers whose surface tensions are between those of the two blend components. For blends between polyolefins and some polar polymers, for example, polyamide, polycarbonate, polyester, reactive group grafted polyolefins are usually used, for example, maleic anhydride grafted polyethylene for polyamide/PE blends. In mixing, the maleic anhydride can react with the amine groups on polyamide to produce polyamide grafted polyolefins, which have good compatibilizing effect. Table 2.5 indicates that compatibilizing agents can decrease the interfacial tensions of some polymer melt pairs greatly.

2.6.2 Polyolefin–Polystyrene Blends with Compatibilizing Agents 2.6.2.1 Polyethylene–Polystyrene Blends with Compatibilizing Agents In case of the first studies using polymer compatibilizing agents, Anderson et al. of Dow Chemical in 1960 (135) irradiated polyethylene with g rays and

46

Polyolefin Blends

Table 2.5 Interfacial Tensions between Polymer Melt Pairs Including Compatibilizing Agents. MAH–PP: Maleic Anhydride Grafted Polypropylene; SEBS: Hydrogenated Triblock Copolymer of Styrene and Butadiene; MAH-g-SEBS: Maleic Anhydride Grafted SEBS; PEMA–Zn: Poly(ethylene-co-methacrylic Acid) Ionomer Neutralized by Zinc. Binary system PE/PA 6

PE/PS

Compatibilizing agent None MAH-g-PP MAH-g-SEBS PEMA-Zn None SEBS

Temperature, C

Interfacial tension, dyn cm

230 230 230 230 180 180

12.5 2 1.5 1.8 5.8 1.1

Adapted from Reference 118 by Chen & White.

subsequently treated it with styrene to make styrene grafted polyethylene, which increased the tensile strength of PS/LDPE blends. Locke and Paul in 1973 (136) also reported that styrene-grafted polyethylene could improve the tensile strength, modulus, and elongation of LDPE/PS blends and greatly decreased the dispersed particle sizes. Kishimoto et al. in 1970 (137) masticated HDPE/ PS blends with the presence of a free radical initiator. They suggested that the formed block graft copolymers improved the properties of the blends. Carrick in 1970 (138) prepared polyethylene grafted polystyrene copolymers based on Friedel–Crafts (F-C) alkylation reaction of polystyrene. The aluminum chloride (AlCl3) was put into the solution of PE/PS blends in cyclohexane to initiate the reaction. After a period the solution was precipitated in methyl–ethyl ketone (MEK) to extract polystyrene. In the following years, Heikens et al. (139–142) used polyethylene-graft-polystyrene prepared as suggested by Carrick to compatibilize LDPE/PS blends and found that they showed good effect to increase interfacial adhesion and impact strength. In 1980, Sjoerdsma et al. (143) reported that tapered butadiene–styrene block copolymer (PS-(PS-PB)random-PB) could increase the interfacial adhesion of LDPE/ PS blends. In the followed years, hydrogenated butadiene–styrene copolymers (SEBS or SEB) was used as compatibilizing agents for HDPE/PS blends by Lindsey et al. (144), Fayt et al. (145), and for LDPE/PS blends by Fayt et al. (146–149). Fayt et al. (147,148) also reported that tapered hydrogenated butadiene–styrene copolymers (PS-(PS-PB)random-PB) showed better compatibilizing effect than pure hydrogenated butadiene–styrene diblock copolymer in decreasing dispersed phase sizes and increasing the elongation and tensile strength of LDPE/PS blends. Today these butadiene–styrene copolymers, for example, SB, SBS, SEB, SEBS, and so on, have been the major group of compatibilizers for PE/PS blends (150–157), presumably because of their large commercial availability. With 5–10 wt% of these compatibilizers in PE/PS blends, the dispersed phase sizes are decreased, and the elongation at break and impact strength are increased.

Chapter 2 Miscibility and Characteristics of Polyolefin Blends

47

Gao et al. (158) showed that LDPE-g-PS prepared based on Friedel-Crafts alkylation reaction could increase the impact strength, elongation at break and tensile strength of LDPE/PS blends. Wang et al. (159) reported that adding dicumyl peroxide (DCP) into HDPE/PS/SBS blends can obviously increase their impact and tensile strength. The mixing order affected the blend properties very much. However, DCP did not show such effect with HDPE/PS blends. 2.6.2.2 Polypropylene–Polystyrene Blends with Compatibilizing Agents The most commonly used compatibilizers for PP/PS blends are also di- or triblock copolymers of styrene and butadiene (SB and SBS) and their hydrogenated products (SEB and SEBS) (160–164). They form dispersed phases in both pure PP and PS. In PP/PS blends, they locate at the interface to connect both PP and PS phase together. Thus, the interfacial tension is decreased and the dispersed phase sizes are greatly decreased. Polystyrene-b-poly (ethylene-co-propylene) block copolymers were also used to compatibilize PP/PS blends (165,166). They showed similar effect to block copolymers of styrene and butadiene. Fig. 2.10 indicates that they locate in the blend interface. Graft and block copolymers of propylene and styrene have been developed to compatabilize PP/PS blends. Del Giudice et al. (167) and Xu and Lin (168) have synthesized PP-b-PS. Kim et al. (169) and Li et al. (170) first polymerized propylene together with some functional monomers, then polymerized styrene from these monomers units to form polystyrene branches. Diaz et al. (171,172) grafted PP chains onto PS chains based on F-C alkylation reaction when mixing PP/PS blends in the presence of AlCl3 catalyst and styrene. All these copolymers help form very

Figure 2.10 Transmission electron micrograph of iPP/aPS 70/30 blends compatibilized with poly (styrene-b-ethylene-co-propylene) (SEP) and stained with RuO4: 10 wt% of SEP. (From Reference 165 with permission from Elsevier.)

48

Polyolefin Blends

strong interfacial adhesion to decrease the dispersed phase size and increase the tensile strength.

2.6.3 Polyolefin–Polyamide Blends with Compatibilizing Agents To increase the affinity between polyolefins and polyamides, some polar groups, such as maleic acid or its anhydride, can be grafted onto polyolefin chains. In mixing the polar groups react with polyamides to form polyamide-grafted polyolefins, which are the real compatibilizing agents. The studies about polar group grafted polyolefins were stimulated by trial to increase the dyeability of polyolefin fibers or wettability of polyolefin films, which was earlier than their application to compatibilize polyolefin– polyamide blends. In the 1960s, Nowak and Jones (173–175) of Dow Chemical grafted reactive reagents such as acrylic acid onto melt polyethylene in a screw extruder. In the 1960–1970s, unsaturated acid group grafted polyolefins were seen to be used as compatibilizing agents for polyolefin/polyamide blends first in patents (176–181), then in publications by Ide and Hasegawa (182), Braun and Eisenlohr (183). Since the late 1980s, compatibilizing agent effects on polyolefin/polyamide blends were extensively studied. 2.6.3.1 Polyethylene–Polyamide Blends with Compatibilizing Agents The most important group of compatibilizers for polyethylene/polyamide blends is the free radical initiated graft copolymers of polyethylene, SEBS, EVA, EPDM, and polypropylene with unsaturated acid or its anhydride. They include graft copolymers of polyethylene with maleic acid or its anhydride, glycidyl methacrylate, and ricinoloxazoline maleinate; graft copolymers of SEBS with glycidyl methacrylate, maleic anhydride, and ricinoloxazoline maleinate; graft copolymers of EVA with mercaptoacetic acid and maleic anhydride; graft copolymer of EPDM with maleic anhydride; graft copolymer of polypropylene with maleic anhydride. The acid or anhydride groups of the grafted copolymers can react with the -NH2 groups in polyamide to form polyolefin-graft-polyamide, the real compatibilizers. These graft copolymers have been called precursor compatibilizers. An extensive study comparing various graft copolymers was given by Chen and White (118). Because the free radical initiated graft reaction can also lead to the cross-linking of polyethylene, copolymers of ethylene and with acrylic acid (184,185), glycidyl methacrylate (184,186), methacrylic acid and 10-undecenoic acid (187–189) were synthesized to compatibilize polyethylene/polyamide blends. The poly (ethyleneco-methacrylic acid) ionomers neutralized by sodium (184) and zinc (45,118,190– 192) has also used as compatibilizers. High energy irradiation, used to modify the surface of fibers or films at beginning, was also used to compatibilize the polyethylene/polyamide blends (193–196).

Chapter 2 Miscibility and Characteristics of Polyolefin Blends

49

2.6.3.2 Polypropylene–Polyamide Blends with Compatibilizing Agents The most commonly used compatibilizers for the blends are graft copolymers of polypropylene with maleic anhydride (PP-g-MAH) (197–203), glycidyl methacrylate (PP-g-GMA) (204,205), and acrylic acid (PP-g-AA) (198,206). They can increase the tensile strength and impact strength of the blends. Favis et al. (207– 210) used an ionomer, a copolymer of ethylene and a mixture of methacrylic acid, zinc methacrylate, and isobutylacrylate to compatibilize PP/PA blends and found it effective to decrease the dispersed particles and improve mechanical properties. To further increase the toughness of polyamides and polyamide-polypropylene blends, elastomers grafted with maleic anhydride were used, for example, SEBS-g-MAH (203,211–214), poly(ethylene-co-octene) grafted maleic anhydride (POE-g-MAH) (202,215), EPR-g-MAH (214) and EVA-g-MAH (216). For example, Fig. 2.11 shows the toughening effect of EPR-g-MAH and SEBS-g-MAH on PP/polyamide 6 blends. For PP-rich blends, the notched impact strength of the toughened blends is usually less than 20 kJ m2 , however, for polyamide-rich blends around 100 kJ m2 notched impact strength was obtained. For polypropylene/polyamide/grafted rubber ternary blends, TEM was proved to be a good tool to study their phase morphology. Without grafting, EPR (217) or SEBS (218) formed dispersed particles together with the minor component of the blends separately; however, after grafting with maleic anhydride they could encapsulate the dispersed minor component of the blends (214,217–221).

Figure 2.11 Izod impact strength of polyamide 6/polypropylene blends modified with 20% EPR-gMA or SEBS-g-MA. The composition shown on the abscissa is the percentage of polypropylene in the blend on a rubber-free basis. The dashed lines correspond to the properties of unmodified binary blends. (From Reference 214 with permission from Elsevier.)

50

Polyolefin Blends

2.7 CONCLUSIONS The miscibility between polyolefins is dependent on molecular structure, blend composition, and mixing temperature. Phase map is needed to characterize their miscibility. The binary blends between polyolefins and polar polymers are immiscible and have large interfacial tension. They show weak interfacial adhesion, poor mechanical properties, and trend to coalesce in the melt state. Compatibilizing agents are needed in these binary blends, which can decrease interfacial tension, prevent phase coalescence, and improve properties largely. Compatibilizing agents can be grafted or biblock copolymers of the two blend components, or can be prepared by functionalizing one blend component, which can graft onto the other component to form grafted copolymers through the functional groups in mixing. To be effective, these compatibilizing agents must stay at phase interface. Both their molecular structures and melt mixing procedure can affect their diffusion behavior to the phase interface.

NOMENCLATURE G U DUm P V Vs T Tc S DS DSm Suniv H DHm kb V N N1 N2 x1 x2 c11 c22 c12 d d1

Gibbs free energy Internal energy Internal energy change in mixing Pressure Volume Volume of segments of polymers Absolute temperature Critical temperature Entropy Entropy change Entropy change of mixing Entropy of the universe Enthalpy Enthalpy change in mixing Boltzmann constant Number of configuration Number of molecule Number of molecule 1 Number of molecule 2 Mole fraction of molecule 1 Mole fraction of molecule 2 Interaction energy between molecule 1 Interaction energy between molecule 2 Interaction energy between molecules 1 and 2 Solubility parameter Solubility parameter of molecule 1

Chapter 2 Miscibility and Characteristics of Polyolefin Blends

d2 f1 f2 x M1 M2 g g0 g1 ecoh M Ke

51

Solubility parameter of molecule 2 Volume fraction of molecule 1 Volume fraction of molecule 2 Interaction parameter Number of segments of molecule 1 Number of segments of molecule 2 Surface tension Surface tension at zero absolute temperature Surface tension of molecules with infinite molecular weight Cohesive energy density Molecular weight Molecular weight–surface tension constant

REFERENCES 1. E. W. Fawcett, R. O. Gibson, M. W. Perrin, J. G. Patton, and E. G. Williams, British Patent 471,590 (1937). 2. K. Ziegler, E. Holzkamp, H. Breil, and H. Martin, Angew. Chem., 67, 541 (1955). 3. J. P. Hogan and R. L. Banks, U.S. Patent 2,825,721 (1958). 4. G. Natta, P. Pino, P. Corradini, F. Danusso, E. Mantica, G. Mazzanti, and G. Moraglio, J. Am. Chem. Soc., 77, 1708 (1955). 5. G. Natta, J. Polym. Sci., 16, 143 (1955); Makromol. Chem., 16, 213 (1955). 6. G. Natta, P. Pino, G. Mazzanti, P. Corradini, and U. Giannini, Rend., Accad. Naz. Lin., 19, 397 (1955). 7. F. P. Reding, J. Polym. Sci., 21, 547 (1956). 8. A. Montagna and J. C. Floyd, MetCon’93, Houston, May 26–28, 1993. 9. J. C. Stevens, F. J. Timmers, D. R. Wilson, G. F. Schmidt, P. N. Nickias, R. K. Rosen, G. W. Knight, and S. Y. Lai, Eur. Patent 416,815-A2 (1991). 10. S. Y. Lai, D. R. Wilson, G. W. Knight, J. C. Stevens, and S. Chum, U.S. Patent 5,272,236 (1993). 11. D. J. Arriola, E. M. Carnahan, P. D. Hustad, R. L. Kuhlman, and T. T. Wenzel, Science, 312, 714 (2006). 12. W. Kaminsky and M. Arndt, Adv. Polym. Sci., 127, 143 (1997). 13. Y. Nishi, M. Ohshima, T. Kohara, T. Matsui, and T. Natsuume, U.S. Patent 5,599,882 (1997). 14. V. Dragutan and R. Streck, Cataytic Polymerization of Cycloolefins, Elsevier Science Ltd., New York, 2000. 15. L. A. Utracki, Introduction to polymer blends, in: Polymer Blends Handbook, Vol. 1, L. A. Utracki (ed.), Kluwer Academic Publishers, Boston, 2002. 16. A. Parkes, English Patent, 11,147 (1846). 17. W. H. Smyers and D. W. Young, U.S. Patent 2,610,962 (1952). 18. L. E. Daly, U.S. Patent 2,439,202 (1948). 19. J. D. Rourk, CA Patent 529, 621 (1956). 20. G. H. Fremon and W. N. Stoops, DE Patent 1,040,249 (1958). 21. T. G. Heggs and G. P. Lee, GB Patent 817,141 (1959). 22. NL Patent 6,517,240 (1966). 23. J. W. Gibbs, Trans. Conn. Acad. Sci., 3, 228 (1876). 24. K. Denbigh, The Principles of Chemical Equilibrium: With Applications in Chemistry and Chemical Engineering, 2nd edition, Cambridge University Press, Cambridge, 1966.

52

Polyolefin Blends

25. R. Clausius, Abhandlungen u¨ber die mechanische wa¨rmetheorie, Vol. 2, Vieweg, Braunschweig, 1867, p. 44. 26. T. L. Hill, An Introduction to Statistical Thermodynamics, Addison-Wesley Publishing Co., Reading, 1960. 27. J. J. Van Laar, Sechs Vortrage uber das Thermodynamische Potential, Vieweg, Braunschweig, 1906. 28. J. J. Van Laar and R. Lorenz, Z. Anorg. Allgem. Chem., 146, 42 (1925). 29. J. H. Hildebrand, J. Am. Chem. Soc., 38, 1452 (1916). 30. J. H. Hildebrand, Proc. Nat. Acad. Sci., U.S.A. 13, 267 (1927). 31. G. Scatchard, Chem. Rev., 8, 321 (1931). 32. J. H. Hildebrand, The Solubility of Non-Electrolytes, Reinhold, New York, 1936. 33. J. H. Hildebrand and R. L. Scott, The Solubility of Nonelectrolytes, 3rd edition, Reinhold, New York, 1950. 34. F. Dolezalek, Z. Phys. Chem., 64, 727 (1908). 35. J. E. Lennard-Jones and A. F. Devonshire, Proc. Roy. Soc., A163, 53 (1937). 36. I. Prigogine and G. Garikian, Physica, 16, 239 (1950). 37. H. C. Longuet-Higgins, Proc. R. Soc., A205, 247 (1951). 38. I. Prigogine, The Molecular Theory of Solution, North-Holland Publishing Co., Amsterdam, 1957. 39. K. H. Meyer, Z. Physik. Chem., B44, 383 (1939). 40. M. L. Huggins, J. Phys. Chem., 46, 151 (1942). 41. P. J. Flory, J. Chem. Phys., 10, 51 (1942). 42. P. J. Flory, J. Chem. Phys., 12, 425, 1944. 43. P. J. Flory, Principles of Polymer Chemistry, Cornell University Press, New York, 1953. 44. RJ. Scott, J. Chem. Phys., 17, 268 (1949). 45. R. Koningsveld, W. H. Stockmayer, and E. Nies, Polymer Phase Disgrams, Oxford University Press, Oxford, 2001. 46. L. P. McMaster, Macromolecules, 6, 760 (1973). 47. T. Young, Philos. Trans. R. Soc., 95, 65 (1805). 48. N. K. Adam, The Physics and Chemistry of Surfaces, Oxford University Press, London, 1941. 49. W. Thomson, Philos. Mag., 42, 368 (1871). 50. J. W. Strutt, Proc. Lond. Math. Soc., 10, 4 (1878). 51. J. W. Strutt, Proc. R. Soc., 29, 71 (1879). 52. C. L. Yaws, X. Y. Lin, L. Bu, and S. Nijhawan, Surface tension, in: Chemical Properties Handbook, C. L. Yaws (ed.), McGraw Hill, New York, 1999. 53. J. Bicerano, Prediction of Polymer Properties, 2nd edition, Marcel Dekker, New York, 2002. 54. D. W. van Krevelen, Properties of Polymers, 3rd edition, Elsevier, Amsterdam, 1990. 55. S. H. Wu, Polymer Interface and Adhesion, Marcel Dekker, New York, 1982. 56. S. H. Wu, Surface and interfacial tensions of polymers, oligomers, plasticizers, and organic pigments, in Polymer Handbook, J. Brandrup, E. H. Immergut, and E. A. Grulke (eds.), Wiley, New York, 1999. 57. T. J. Menke, Z. Funke, R. D. Maier, and J. Kressler, Macromolecules, 33, 6120 (2000). 58. G. W. Bender and G. L. Gaines, Macromolecules, 3, 128 (1970). 59. J. C. Maxwell, Surface tension, in: Encyclopedia Britannica, Samuel L. Hall, New York, 1878. 60. G. Quincke, Ann. Phys., 137, 1 (1870). 61. J. M. Andreas, E. A. Hauser, and W. B. Tucker, J. Phys. Chem., 42, 1001 (1938). 62. P. J. Yoon and J. L. White, J. Appl. Polym. Sci., 51, 1515 (1994). 63. G. Macaubas and P. H. Pierin, Polym. Eng. Sci., 43, 670 (2003). 64. G. Palmer and N. R. Demarquette, Polymer, 46, 8169 (2005).

Chapter 2 Miscibility and Characteristics of Polyolefin Blends

53

65. J. T. Koberstein, Interfacial properties, in: Encyclopedia of Polymer Science and Engineering, J. I. Kroschwitz (ed.), Wiley Interscience, New York, 1987. 66. N. R. Demarquette, Int. Mater. Rev., 48, 247 (2003). 67. E. D. Carlson, M. T. Krejchi, C. D. Shah, T. Terakawa, R. M. Waymouth, and G. G. Fuller, Macromolecules, 31, 5343 (1998). 68. R. Chen, M. R. Xie, Q. Wu, and S. G. Lin, J. Polym. Sci. A Polym. Chem., 38, 411 (2000). 69. Y. Hu, M. T. Krejchi, C. D. Shah, C. L. Myers, and R. M. Waymouth, Macromolecules, 31, 6908 (1998). 70. E. D. Carlson, G. G. Fuller, and R. M. Waymouth, Macromolecules, 32, 8094 (1999). 71. W. Wiyatno, J. A. Pople, A. P. Gast, R. M. Waymouth, and G. G. Fuller, Macromolecules, 35, 8488 (2002). 72. M. J. Hill, P. J. Barham, and A. Keller, Polymer, 33, 2530 (1992). 73. M. J. Hill, P. J. Barham, and J. V. Ruiten, Polymer, 34, 2975 (1993). 74. M. J. Hill, R. L. Morgan, and P. J. Barham, Polymer, 38, 3003 (1997). 75. R. L. Morgan, M. J. Hill, and P. J. Barham, Polymer, 40, 337 (1999). 76. T. Hameed and I. A. Hussein, Polymer, 43, 6911 (2002). 77. I. A. Hussein, Macromolecules, 36, 2024 (2003). 78. T. Hameed and V Hussein, Macromol. Mater. Eng., 289, 198 (2004). 79. I. A. Hussein and M. C. Williams, Polym. Eng. Sci., 44, 660 (2004). 80. G. D. Wignall, J. D. Londono, J. S. Lin, R. G. Alamo, M. J. Galante, and L. Mandelkern, Macromolecules, 28, 3156 (1995). 81. R. G. Alamo, W. W. Graessley, R. Krishnamoorti, D. J. Lohse, J. D. Londono, L. Mandelkern, F. C. Stehling, and G. D. Wignall, Macromolecules, 30, 561 (1997). 82. G. D. Wignall, R. G. Alamo, J. D. Londono, L. Mandelkern, M. H. Kim, J. S. Lin, and G. M. Brown, Macromolecules, 33, 551 (2000). 83. K. Tashiro, R. S. Stein, and S. L. Hsu, Macromolecules, 25, 1801 (1992). 84. K. Tashiro, M. M. Satkowski, R. S. Stein, Y. Li, B. Chu, and S. L. Hsu, Macromolecules, 25, 1809 (1992). 85. K. Tashiro, M. Izuchi, M. Kobayashi, and R. S. Stein, Macromolecules, 27, 1221 (1994). 86. K. Tashiro, M. Izuchi, M. Kobayashi, and R. S. Stein, Macromolecules, 27, 1234 (1994). 87. B. S. Tanem and A. Stori, Polymer, 42, 4309 (2001). 88. A. Y. Coran and R. Patel, U.S. Patent 4,297,453 (1981). 89. A. Y. Coran and R. Patel, U.S. Patent 4,310,638 (1982). 90. E. Nolley, J. W. Barlow, and D. R. Paul, Polym. Eng. Sci., 20, 364 (1980). 91. D. W. Bartlett, J. W. Barlow, and D. R. Paul, J. Appl. Polym. Sci., 27, 2351 (1982). 92. N. K. Kalfoglou, Angewandte Makromolekulare Chemie, 129, 103 (1985). 93. D. J. Lohse, Polym. Eng. Sci., 26, 1500 (1986). 94. L. Dorazio, C. Mancarella, E. Martuscelli, and G. Sticotti, J. Mater. Sci., 26, 4033 (1991). 95. E. Martuscelli, C. Silvestre, and G. Abate, Polymer, 23, 229 (1982). 96. C. Y. Chen, W. M. Z. W. Yunus, H. W. Chiu, and T. Kyu, Polymer, 38, 4433 (1997). 97. M. Seki, H. Nakano., S. Yamauchi, J. Suzuki, and Y. Matsushita, Macromolecules, 32, 3227 (1999). 98. K. H. Nitta, Y. W. Shin, H. Hashiguchi, S. Tanimoto, and M. Terano, Polymer, 46, 965 (2005). 99. A. R. Kamdar, Y. S. Hu, P. Ansems, S. P. Chum, A. Hiltner, and E. Baer, Macromolecules, 39, 1496 (2006). 100. P. S. Chum, W. J. Kruper, and M. J. Guest, Adv. Mater., 12, 1759 (2000). 101. V. B. F. Mathot and M. J. F. Pijpers, J. Appl. Polym. Sci., 39, 979 (1990).

54

Polyolefin Blends

102. F. Defoor, G. Groeninckx, H. Reynaers, P. Schouterden, and B. Van der Heijden, Macromolecules, 26, 2575 (1993). 103. K. Sehanobish, R. M. Patel, B. A. Croft, S. P. Chum, and C. I. Kao, J. Appl. Polym. Sci., 51, 887 (1994). 104. S. Bensason, J. Minick, A. Moet, S. Chum, A. Hiltner, and E. Baer, J. Polym. Sci. B, Polym. Phys., 34, 1301 (1996). 105. A. L. N. Da Silva, M. I. B. Tavares, D. P. Politano, F. M. B. Coutinho, and M. C. G. Rocha, J. Appl. Polym. Sci., 66, 2005 (1997). 106. F. Stricker, R. D. Maier, Y. Thomann, and R. Mulhaupt, Kunststoffe-Plast Europe, 88, 527 (1998). 107. T. McNally, P. McShane, G.M Nally, W. R. Murphy, M. Cook, and A. Miller, Polymer, 43, 3785 (2002). 108. M. Yamaguchi, H. Miyata, and K. H. Nitta, J. Appl. Polym. Sci., 62, 87 (1996). 109. Y. Thomann, J. Suhm, R. Thomann, G. Bar, R. D. Maier, and R. Mu¨lhaupt, Macromolecules, 31, 5441 (1998). 110. M. Yamaguchi, K. Suzuki, and H. Miyata, J. Polym. Sci. B, Polym. Phys., 37, 701 (1999). 111. M. Yamaguchi and K. Nitta, Polym. Eng. Sci., 39, 833 (1999). 112. M. Yamaguchi and H. Miyata, Macromolecules, 32, 5911 (1999). 113. P. M. Cham, T. H. Lee, and H. Marand, Macromolecules, 27, 4263 (1994). 114. Z. Bartczak, A. Galeski, and M. Pracella, J. Appl. Polym. Sci., 54, 1513 (1994). 115. A. Siegmann, J. Appl. Polym. Sci., 27, 1053 (1982). 116. C. C. Hsu and P. H. Geil, Polym. Eng. Sci., 27, 1542 (1987). 117. M. S. Lee and S. A. Chen, J. Polym. Sci. C, Polym. Lett., 25, 37 (1987). 118. C. C. Chen and J. L. White, Polym. Eng. Sci., 33, 923 (1993). 119. K. Min, J. L. White, and J. F. Fellers, J. Appl. Polym. Sci., 29, 2117 (1984). 120. I. Banik, P. J. Carreau, and H. P. Schreiber, J. Polym. Sci. B, Polym. Phys., 42, 2545 (2004). 121. C. W. Macosko, P. Gue´gan, A. K. Khandpur, A. Nakayama, P. Marechal, and T. Inoue, Macromolecules, 29, 5590 (1996). 122. U. Sundararaj and C. W. Macosko, Macromolecules, 28, 2647 (1995). 123. J. M. Willis, B. D. Favis, and J. Lunt, Polym. Eng. Sci., 30, 1073 (1990). 124. B. R. Liang, J. L. White, J. E. Spruiell, and B. C. Goswami, J. Appl. Polym. Sci., 28, 2011 (1983). 125. K. Min, J. L. White, and J. F. Fellers, Polym. Eng. Sci., 24, 1327 (1984). 126. C. C. Chen, E. Fontan, K. Min, and J. L. White, Polym. Eng. Sci., 28, 69 (1988). 127. A. Valenza, L. Gallo, G. Spadaro, E. Calderaro, and D. Acierno, Polym. Eng. Sci., 33, 1336 (1993). 128. Y. G. Cho and M. R. Kamal, Polym. Eng. Sci., 42, 2005 (2002). 129. M. Baer and E. H. Hankey, U.S. Patent 3,085,082 (1963). 130. E. H. Hankey and A. Kianpour, U.S. Patent 3,145,187 (1964). 131. M. Baer and E. H. Hankey, U.S. Patent 3,399,155 (1968). 132. D. Q. Khang, L. Starke, and Z. Funke, J. Macromol. Sci. Pure Appl. Chem., A33, 1973 (1996). 133. N. G. Gaylord, U.S. Patent 3,485,777 (1969). 134. N. G. Gaylord, in: Copolymers, Polyblends and Composites, ACS Advances in Chemistry Series, Vol. 142, ACS, Washington, 1975, p.76. 135. L. C. Anderson, D. A. Roper, and J. K. Rieke, J. Polym. Sci., 43, 423 (1960). 136. C. E. Locke and D. R. Paul, J. Appl. Polym. Sci., 17, 2597 (1973); 17, 2791 (1973). 137. A. Kishimoto, S. Hirata, and H. Ueno, DE Patent 1,950,479 (1970). 138. W. L. Carrick, J. Polym. Sci., 8, 215 (1970). 139. D. Heikens and W. Barensten, Polymer, 18, 69 (1977). 140. W. M. Barentsen and D. Heikens, Polymer, 14, 579 (1973). 141. W. M. Barentsen, D. Heikens, and P. Piet, Polymer, 15, 119 (1974).

Chapter 2 Miscibility and Characteristics of Polyolefin Blends

55

142. D. Heikens, N. Hoen, W. Barentsen, P. Piet, and H. Ladan, J. Polym. Sci. Polym. Symposia, 62, 309 (1978). 143. S. D. Sjoerdsma, J. Dalmolen, A. C. A. M. Bleijenberg, and D. Heikens, Polymer, 21, 1469 (1980). 144. C. R. Lindsey, D. R. Paul, and J. W. Barlow, J. Appl. Polym. Sci., 26, 1 (1981). 145. R. Fayt, R. Jerome, and P. Teyssie, J. Polym. Sci. B, Polym. Phys., 19, 1269 (1981). 146. R. Fayt, R. Jerome, and P. Teyssie, J. Polym. Sci. C, Polym. Lett., 19, 79 (1981). 147. R. Fayt, R. Jerome, and P. Teyssie, J. Polym. Sci. B, Polym. Phys., 20, 2209 (1982). 148. R. Fayt, R. Jerome, and P. Teyssie, Makromol. Chem., 187, 837 (1986). 149. R. Fayt, R. Jerome, and P. Teyssie, J. Polym. Sci. C Polym. Lett., 24, 25 (1986). 150. I. Fortelny, J. Mikesova, J. Hromadkova, V. Hasova, and Z. Horak, J. Appl. Polym. Sci., 90, 2303 (2003). 151. M. Tasdemir and H. Yildirim, J. Appl. Polym. Sci., 83, 2967 (2002). 152. C. Harrats, R. Fayt, and R. Jerome, Polymer, 43, 863 (2002). 153. J. S. Wu, B. H. Guo, C. M. Chan, J. X. Li, and H. S. Tang, Polymer, 42, 8857 (2001). 154. I. Luzinov, C. Pagnoulle, and R. Je´roˆme, Polymer, 41, 3381 (2000). 155. J. X. Li, C. M. Chan, B. H. Gao, and J. S. Wu, Macromolecules, 33, 1022 (2000). 156. H. F. Guo, S. Packirisamy, R. S. Mani, C. L. Aronson, N. V. Gvozdic, and D. J. Meier, Polymer, 39, 2495 (1997). 157. T. Li, V. A. Topolkaraev, A. Hiltner, E. Baer, X. Z. Ji, and R. P. Quirk, J. Polym. Sci. B, Polym. Phys., 33, 667 (1995). 158. Y. Gao, H. L. Huang, Z. H. Yao, D. Shi, Z. Ke, and J. H. Yin, J. Polym. Sci. B, Polym. Phys., 41, 1837 (2003). 159. Z. Wang, C. M. Chan, S. H. Zhu, and J. R. Shen, Polymer, 39, 6801 (1998). 160. A. Halimatudahliana, H. Ismail, and M. Nasir, Polym. Test., 21, 163 (2002). 161. D. Hlavata, Z. Horak, F. Lednicky, J. Hromadkova, A. Pleska, and Y. V. Zanevskii, J. Polym. Sci. B Polym. Phys., 39, 931 (2001). 162. P. H. P. Macaubas and N. R. Demarquette, Polymer, 42, 2543 (2001). 163. D. Hlavata, Z. Horak, J. Hromadkova, F. Lednicky, and A. Pleska, J. Polym. Sci. B, Polym. Phys., 37, 1647 (1999). 164. G. Radonjic, V. Musil, and I. Smit, J. Appl. Polym. Sci., 69, 2625 (1998). 165. I. Smit, G. Radonjic, and D. Hlavata, Eur. Polym. J., 40, 1433 (2004). 166. G. Radonjic and V. Musil, Die Angewandte Makromolekulare Chemie, 251, 141 (1997). 167. L. Del Giudice, R. E. Cohen, G. Attalla, and F. Bertinotti, J. Appl. Polym. Sci., 30, 4305 (1985). 168. G. X. Xu and S. G. Lin, Polymer, 37, 421 (1996). 169. J. Kim, J. Kwak, and D. Kim, Polym. Adv. Technol., 14, 58 (2003). 170. Z. Li, Y. C. Ke, and Y. L. Hu, J. Appl. Polym. Sci., 93, 314 (2004). 171. M. F. Diaz, S. E. Barbosa, and N. J. Capiati, Polymer, 46, 6096 (2005). 172. M. F. Diaz, S. E. Barbosa, and N. J. Capiati, J. Polym. Sci. B, Polym. Phys., 42, 452 (2004). 173. R. M. Nowak and G. D. Jones, U.S. Patent 3,177,269 (1965). 174. G. D. Jones and R. M. Nowak, U.S. Patent 3,177,270 (1965). 175. R. M. Nowak, U.S. Patent 3,270,090 (1966). 176. R. G. Armstrong, U.S. Patent 3,373,222 (1967). 177. H. Terada, T. Kono, M. Kawajima, and A.Hasegawa, Japan Patent 46,038,022 (1971). 178. T. Ozeki, T. Kono, and M. Kawajima, Japan Patent 46,038,023 (1971). 179. W. Nielinger, R. Dhein, K. Schneider, and P. Tacke, DE Patent 2,426,671 (1975). 180. J. H. Davis, GB Patent 1,403,797 (1975).

56

Polyolefin Blends

181. H. W. Starkweather, DE Patent 2,551,023 (1976). 182. F. Ide and A. Hasegawa, J. Appl. Polym. Sci., 18, 963 (1974). 183. D. Braun and U. Eisenlohr, Angewandte Makromolekulare Chemie, 58–59, 227 (1977). 184. A. Lahor, M. Nithitanakul, and B. P. Grady, Eur. Polym. J., 40, 2409 (2004). 185. A. Valenza, L. Gallo, G. Spadaro, and E. Calderaro, Polym. Eng. Sci., 33, 1336 (1993). 186. S. Filippi, V. Chiono, G. Polacco, M. Paci, L. Minkova, and P. Magagnini, Macromol. Chem. Phys., 203, 1512 (2002). 187. A. Valenza, G. Spadaro, E. Calderaro, and D. Acierno, Polym. Eng. Sci., 33, 845 (1993). 188. Y. J. Yoo, C. H. Park, S. G. Lee, K. Y. Choi, D. S. Kim, and J. H. Lee, Macromol. Chem. Phys., 206, 878 (2005). 189. M. Mehrabzadeh and M. Kamal, Polym. Eng. Sci., 44, 1152 (2004). 190. B. D. Favis, Polymer, 35, 1552 (1994). 191. P. Leewajanakul, R. Pattanaolarn, J. W. Ellis, M. Nithitanakul, and B. P. Grady, J. Appl. Polym. Sci., 89, 620 (2003). 192. G. Fairley and R. E. Prudhomme, Polym. Eng. Sci., 27, 1495 (1987). 193. A. Valenza, L. Gallo, G. Spadaro, and E. Calderaro, Polym. Eng. Sci., 33, 1336 (1993). 194. A. Valenza, G. Spadaro, E. Calderaro, and D. Acierno, Polym. Eng. Sci., 33, 845 (1993). 195. G. Spadaro, D. Acierno, C. Dispenza, E. Calderaro, and A. Valenza, Radiat. Phys. Chem., 48, 207 (1996). 196. S. S. Wu, B. O. Chi, H. Yan, and J. Shen, J. Appl. Polym. Sci., 99, 2029 (2006). 197. S. J. Park, B. K. Kim, and H. M. Jeong, Eur. Polym. J., 26, 131 (1990). 198. A. Valenza and D. Aciermo, Eur. Polym. J., 30, 1121 (1994). 199. S. Jose, B. Francis, S. Thomas, and J. Karger-Kocsis, Polymer, 47, 3874 (2006). 200. S. Jose, S. V. Nair, S. Thomas, and J. Karger-Kocsis, J. Appl. Polym. Sci., 99, 2640 (2006). 201. D. X. Li, D. M. Jia, and P. Zhou, J. Appl. Polym. Sci., 93, 420 (2004). 202. N. Zeng, S. L. Bai, C. G’Sell, J. M. Hiver, and Y. M. Mai, Polym. Int., 51, 1439 (2002). 203. J. D. Tucher, S. Lee, and R. L. Einsporn, Polym. Eng. Sci., 40, 2577 (2000). 204. X. M. Zhang, Z. H. Yin, T. H. Na, and J. H. Yin, Polymer, 38, 5905 (1997). 205. X. M. Zhang, G. Li, J. S. Li, and J. H. Yin, Angew. Makromol. Chem., 248, 189 (1997). 206. X. M. Zhang and J. H. Yin, Polym. Eng. Sci., 37, 197 (1997). 207. P. Van Gheluwe, B. D. Favis, and J. P. Chalifoux, J. Mater. Sci., 23, 3910 (1988). 208. P. Van Gheluwe and B. D. Favis, Polym. Mater. Sci. Eng., 58, 966 (1988). 209. J. M. Willis, V. Caldas, and B. D. Favis, J. Mater. Sci., 26, 4742 (1991). 210. J. M. Willis, B. D. Favis, and C. Lavallee, J. Mater. Sci., 28, 1749 (1993). 211. M. J. Modic and L. A. Pottick, Plast. Eng., 47, 37 (1991). 212. J. Roesch and R. Muelhaupt, Makromol. Chem. Rapid Commun., 14, 503 (1993). 213. M. Tasdemir, J. Appl. Polym. Sci., 89, 3485 (2003). 214. A. Gonza´lez-Montiel, H. Keskkula, and D. R. Paul, Polymer, 36, 4587 (1995). 215. S. L. Bai, G. T. Wang, J. M. Hiver, and C. G’Sell, Polymer, 45, 3063 (2004). 216. H. Z. Liu, T. X. Xie, L. L. Hou, Y. C. Ou, and G. S. Yang, J. Appl. Polym. Sci., 99, 3300 (2006). 217. J. Ro¨sch, Polym. Eng. Sci., 35, 1917 (1995). 218. A. N. Wilkinson, L. Laugel, M. L. Clemens, V. M. Harding, and M. Marin, Polymer, 40, 4971 (1999). 219. B. Lu and T. C. Chung, Macromolecules, 32, 2525 (1999). 220. G. M. Kim, G. H. Michler, J. Ro¨sch, and R. Mu¨thaupt, Acta Polym., 49, 88 (1998). 221. J. Ro¨sch and R. Mu¨lhaupt, Polym. Bull., 32, 697 (1994).

Part II

Polyolefin/Polyolefin Blends

Chapter

3

Miscibility, Morphology, and Properties of Polyethylene Blends Robert A. Shanks1

3.1 INTRODUCTION Polyethylenes (PE) are a group of polymers having the same chemical structure, but differing in molecular architecture. They have subambient glass transition temperatures so the amorphous phase is rubbery. Several additional subambient relaxation temperatures have been detected depending on the branching characteristics. Molecular flexibility contributes yield, drawing, and toughness. Strength and moderate thermal resistance are derived from crystallinity. Weak intermolecular forces means that interactions and crystallinity result from symmetry and close packing. Branching restricts symmetry and close packing. Branch points are excluded from the crystals. Polyethylenes have many applications and through their price and performance they rank as commodity polymers. They are used in films, sheets, fibers, pipes, and many types of moldings. Polyethylene products are prepared by extrusion, injection molding, thermoforming, and rotational casting. A particular polyethylene must be chosen for the molding process, product properties, and longer term product performance. Even with the large range of polyethylenes available, it is difficult to choose a polyethylene with the best properties. Blending is usually the solution to property optimization. More polyethylenes are manufactured as blends than they are as the pure polymers. Most polyethylene blends are with other polyethylenes, though some blends are formed with ethylene copolymers with polar monomers such as vinyl acetate (VAc), methyl acrylate (MA), butyl acrylate (BA), glycidyl methacrylate

1

School of Applied Sciences, RMIT University, GPO Box 2476V, Melbourne, Vic 3001, Australia

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

59

60

Polyolefin Blends

(GMA), acrylic acid (AA), or grafted with maleic anhydride (MAn). Hydrocarbon waxes, sometimes highly branched, are other compounds that may be blended with polyethylenes. Polyethylene in its ideal form is a linear polymer (LPE) that should correctly be called poly(methylene), Table 3.1. LPE can reach high crystallinity (Xc > 0:90) due to its compact symmetrical molecules. It has low strength dispersive molecular forces. The melting temperature (Tm ) of LPE is variable but usually about 135 C. The equilibrium melting temperature of LPE has been measured many times and is about 145 C. LPE is insoluble at ambient temperatures, but it can be dissolved in hydrocarbon, alkylaromatics, and halogenated hydrocarbons at temperatures approaching its melting temperature. Two crystalline morphologies have been observed; the chain folded form is most common and forms spontaneously from

Table 3.1

Properties of Linear Polyethylene.

Property

Value

Comment

Monomer

CH2 CH2

Polymerization

Ziegler–Natta, supported metal oxides such as Philips, Unipol; and metallocene 120 to 60 C 137 C

Symmetrical and compact for close packing, dispersive forces; no other intermolecular interactions, polarity, or bulky steric groups Stereospecific chain formation required; weak interaction/complexation with any initiators/catalysts; metal coordination complexes required

Glass transition (Tg ) Melting temperature (Tm )

Crystallinity

0.60–0.95

Molar mass

Wide range, 103–107 g mol1

Density

0.90–0.96 g cm3

Stability

No chemical groups to absorb ultraviolet light and no reactive tertiary hydrogens

An elastomer or viscous liquid at ambient folded chain crystals; strength properties derived from crystallinity; low thermal resistance. Extended chain crystals can melt at about 145 C Properties dependent on crystallinity, with amorphous phase providing flow, flexibility, and toughness. Kinetically stable folded chain crystals; thermodynamically stable extended chain crystals can be formed by gel drawing Molar mass must be high since intermolecular forces are weak; all values are available from waxes to ultrahigh molar mass. Dependant on crystallinity; often HDPE contains some butene comonomer to avoid brittleness The ideal polyethylene structure should be very stable; however, some unsaturated end groups, impurity oxygen containing groups, and catalyst residues accelerate degradation

Chapter 3 Miscibility, Morphology, and Properties of Polyethylene Blends

61

Figure 3.1 Density, crystallinity, and polyethylene type scale.

the melt, being kinetically favored. The chain extended form is thermodynamically favored, but can be formed by only flow elongation of molecules, typically by gel spinning. Significant characteristics of LPE are molar mass (M) and molar mass distribution. Molar mass and morphology are the only variables available to provide required properties with LPE. LPE is usually called high density polyethylene (HDPE) because of the high density arising from high crystallinity. Commercial HDPE often contains a small proportion of comonomer, typically 1-butene, to decrease its brittleness. The comonomer will lower the density, crystallinity, and melting temperature; Tm may be in the range 130–135 C. A schematic showing polyethylene type, density, melting temperature, and spherulitic crystal ranges is shown in Fig. 3.1. Branching of polyethylenes provides the second dimension, after molar mass, with which to control properties, Tables 3.2 and 3.3. Branching of polyethylenes is a complex topic; in this review it will be treated starting with the ideal copolymer structures formed with the new metallocene catalysts. Branched polyethylenes (BPE) provide increased toughness though decreased modulus and strength compared with LPE. Branches are obtained by copolymerization with 1-alkenes, such as Table 3.2 Properties of Typical Branched Polyethylenes. Properties

VLDPE-B VLDPE-O

Comonomer Butene Catalyst type M MFI (dg min1) 27.0 0.901 Density (g cm3) Mw 58,000 Mw/Mn 2.65 Comonomer 6.3 Content (%mol) Tm ( C) 92.7 Tc( C) 76.6

VLDPE-O(ZN)

LDPE-highM LDPE-lowM

Octene S 1.0 0.908 96,700 2.86 2.4

Octene ZN 1.0 0.912 120,000 3.80 4.2

P 7.0 0.919 474,000 23.3

P 22.0 0.918 89,000 4.4

105.4 90.3

123.0 100.5

106.5 87.8

103.9 87.1

Data were obtained from data sheets published by the manufacturer. ZN ¼ Ziegler–Natta catalyst, S ¼ Constrained geometry single-site catalyst, M ¼ Metallocene, and P ¼ Peroxide. Crystallisation (Tc) and melting (Tm) temperature were obtained using DSC at scanning rates of 10 C min1.

62

Polyolefin Blends

Table 3.3

Types of Polyethylenes and Their Properties.

Polyethylene

Structure

Comment

HDPE or linear polyethylene, poly(methylene) LDPE

Linear carbon chain

See Table 3.1

Random branching with either short (average C4) or long (many C) branches; dependant on autoclave or tubular reactor type

LLDPE

Short branches introduced by a comonomer, usually butene, hexene, or octane; with many others; comonomer amount and structure limited with ZN catalysts

mLLDPE

Metallocene-catalyzed PE with more uniform short branch distribution; increased monomer content and structure available Increased comonomer content giving lower density and lower Tm. ZN and metallocene type available Highest comonomer contents, lowest Tm ; crystallization can continue below ambient; can be dissolved at ambient

Many short branches limit crystallinity and lamella thickness (reduce Tm ); long branches provide nonNewtonian rheology and melt strength Short branch content is controlled by comonomer amount, distribution of comonomer not controlled. Slurry (or solution) and gas phase variations. Bimodal crystal distributions and possibly two liquid phases Structure and property combinations highly specialized; new applications possible with a wide range of molar mass and branching Soft PE that can be blended with stronger PE to improve toughness, heat sealing, and elasticity Elastomeric properties that require cross-linking if used alone. Often used in blends

VLDPE

ULDPE

1-butene, 1-hexene, and 1-octene. Polar polyethylenes are formed by copolymerization with VAc, MA, BA, GMA, or AA. Grafting of MAN can be performed in a solution reaction or in an extruder. The properties of BPE are determined by the proportion of branches, or more specifically the methylene sequence length (MSL) between branches, in addition to molar mass and its distribution. When the comonomers are unevenly distributed along polyethylene molecules, MSL will be a distribution or this can be expressed and a comonomer distribution. A two-dimensional contour chart must define the architecture of BPE, with molar mass distribution and comonomer distribution axes. The comonomer distribution can be intermolecular, such that some molecules have mainly linear structure while others have many branches. The comonomer distribution may be intramolecular, where the density of comonomer units changes along the length of a molecule, so that segments

Chapter 3 Miscibility, Morphology, and Properties of Polyethylene Blends

63

Figure 3.2 PE branch distributions, intermolecular branch distribution (top and middle), and intramolecular branch distribution (bottom).

of a molecule may be mainly linear, while other segments have many comonomers or branches (1). Inter- and intramolecular branch distributions are illustrated in Fig. 3.2. The morphology is independent of branch length, that is, comonomer structure, since the branches are excluded from crystals. BPE polymerized by Ziegler–Natta, Philips, Unipol, or similar multisite catalysts show increased dispersion of comonomers. The distributions are typically bimodal. Gas phase and solution polymerizations differ in distribution, with solution or slurry prepared BPE generally showing broader distributions. Bimodal distributions of branches result in bimodal distributions of lamella thickness and hence melting temperatures. Bimodal BPE behave like blends though there is poor control over the structure and composition of each component. The amount of braches that can be included is limited to low amounts compared with single-site-catalyzed BPE where highly branched elastomeric BPE can be produced. There is controversy as to whether the components are miscible in the liquid phase and if they become immiscible at what difference in branch composition and at what temperature. The phase diagram is proposed to be of an upper critical solution temperature type. The BPE described are prepared by copolymerization and only short branches are present. PBE with both short and long branches are prepared from radical polymerization in the traditional high pressure process. These PE are called low density polyethylene (LDPE). A molecular model of a conformation of a typical LDPE is shown in Fig. 3.3. LDPE with different proportions of short and long branches are formed using autoclave or tubular reactors. There are many short branches with average

Figure 3.3 Molecular model of typical LDPE with 1000 carbons, 3 long branches, and 30 short branches.

64

Polyolefin Blends

length of four carbons formed by intramolecular chain transfer by the growing chain. There are relatively few long branches, about 10% of the short branches, formed by intermolecular chain transfer by a growing chain to a terminated chain. The special properties of LDPE compared with other PBE are brought about by the long branches. Long branches contribute to melt rheology during processing by providing increased shear thinning and melt strength. The branch distribution in LDPE is random because of the mechanism of formation, and the overall distribution demonstrated by melting is narrow compared with most multisite-catalyzed BPE. Techniques are now available for polymerization with single-site catalysts that allow long branches to be included by copolymerization with alkene-terminated BPE formed during the same polymerization reaction. Alternatively, the long branches can be added to any PE composition by blending with LDPE.

3.2 STRUCTURE AND PROPERTIES OF POLYETHYLENES The important factors (in order of most to least) in controlling the properties of polyethylenes are as follows: Proportion of branches (i.e., the composition of the comonomer). Each comonomer unit in the polymer is unable to crystallize, so this breaks the ability of the polyethylene chain to take part in crystallization. The distribution of the comonomer units in the chain (are they distributed evenly or are they clustered within some molecules or within parts of some molecules). This is determined by the type of catalyst—metallocenes give relatively even distribution of branches. Ziegler–Natta catalysts give broad distribution. Bimodal distribution of branches, similar to the above two, except that the melting shows two endotherms. Usually a broad lower temperature melting and a sharp higher temperature melting, the latter for crystals of molecules with few branches. This is caused by uneven reactivity giving two phases in the reactor. Type of branches—1-butene comonomer gives ethyl branches, 1-hexene comonomer gives butyl branches, and 1-octene comonomer gives hexyl branches. The length of branches has a small influence, but the number of branches is more important since the branches do not take part in crystallization. Molar mass (inversely proportional to melt flow index, MFI) has a small influence. Generally in the chain folded crystal structure, the molar mass will not be important. It will mean that a particular molecule has more folds in the crystal. Melting is determined by the crystal thickness. Crystal thickness is determined by crystallization conditions and the presence of branches that sterically prevent crystallization of some segments. Often molecules with different numbers of branches will be immiscible in the liquid state (melt). So a polymer with a broad branch distribution may have two

Chapter 3 Miscibility, Morphology, and Properties of Polyethylene Blends

65

liquid phases and so crystallize with two crystal populations. This is another manifestation of the above three, within the one polymer. Blends of polyethylenes with much different branch contents will also show two immiscible liquid phases.

3.3 APPLICATIONS OF POLYETHYLENE BLENDS Rheological control is a primary reason for the preparation of PE blends. The PE chosen to provide optimum properties for a particular application may not have desired processing characteristics, Table 3.4. Ideally shear thinning is preferred so that high molar mass PE required for a product can be processed under moderate conditions. Even a small amount of BPE with significant shear thinning can impart shear-thinning properties when used in a blend. Branching increases the elasticity, elongation at break, and toughness of PE films, though modulus, yield stress, and break stress are decreased. Heat sealing of films is enhanced by BPE since branching reduces the melting temperature. PE films are usually produced by the blown film extrusion method. The alternative method is sheet extrusion followed by passing through chilled rollers. Blown film is more economic; biaxial orientation is introduced by the draw-off and blow ratios. Bubble stability is critical to the blown film process. Bubble stability is provided by the melt strength and rheological characteristics of the PE. Long

Table 3.4 Some Examples of Typical Polyethylene Blends and Their Applications. Polymer 1

Polymer 2

HDPE

LDPE

HDPE

LLDPE

LLDPE mLLDPE UHMWPE

HDPE

Application

Toughened HDPE with improved processing, especially melt strength; HDPE and LDPE are immiscible LLDPE Pipe grade HDPE required rheology for (or LDPE) extrusion; rigidity and impact resistance for pipes LDPE Typical film combination for increased melt strength, elasticity combined with LLDPE strength. These are the most widely used PE blends mLLDPE or Packaging films with improved melting and mVLDPE flow character for heat seal LDPE Introduction of long branches to assist processing into specialized metallocene PE HDPE Lower molar mass (HDPE) to moderate the properties of ultrahigh molar mass PE for processing and flexibility HDPE Bimodal molar mass distributions provide (‘‘reactor blends’’) unique properties that cannot be achieved by mixing the two PE grades

66

Polyolefin Blends

branches and high molar mass increase melt strength and so LDPE is preferred for blown film manufacture, but other properties may be required of the film after formation. The solution to optimization of production and application properties is to use blends of PE to contribute their unique characteristics. Such an important blend use for films is LLDPE–LDPE. Better-defined properties can be obtained using particular BPE formed by single-site catalysts. These BPE are more expensive, so a blend with lower cost PE will often provide the desired contribution to properties. Often combinations of properties are required and these are more applicable to a PE blend. A combination of high melt strength and shear-thinning rheology requires long branches, with high film strength from low branch content, to allow thin films to be made. Heat sealable films require a low temperature melting component, together with a strength-providing component. PEs are being increasingly used for cable coverings, either for large cables or for flexible appliance coverings. Many competing requirements are expected of these cable coverings, most specified by regulations or standards. The inherent properties of the PE are often supplemented by peroxide or grafted silane-induced cross-linking. Metallocene PE have suitable properties for cable covering, but blends are needed to satisfy all specifications. Plasticised poly(vinyl chloride) has been used so a PE covering will often need to possess similar properties to be considered as a substitute. Extruded pipes need to be rigid yet tough so a significant component is likely to be HDPE, yet toughness must be included and flexibility to enable coiling. Toughness of HDPE can be improved by inclusion of some comonomer, though a blend with a PE plastomer or elastomer will produce dispersed phase elastomer that is know to provide efficient toughness. Other extruded profiles for building components will require similar properties. PE moldings such as garbage bins need a combination of stiffness, strength, and toughness. In addition high melt flow is needed to completely fill the mold before solidification. HDPE is the base polymer, but other blended polymer such as a PE elastomer will be required to increase the toughness while retaining strength and dimensional stability, particularly when the bin is used to carry a load. PE is used for large rotationally molded tanks. The tanks must be strong, stiff, and tough as for the other molded products. Rotomolding is a special molding technique where flow of the molten polymer is required under a small centrifugal or gravitational force and the walls must be void free. The slow molding process usually does not produce stresses within the walls. Low molar mass could provide suitable melt flow, but the tank would have insufficient strength so the PE must be a blend to obtain the individual distinct properties. See Table 3.5 for a summary of processing methods, the PE and product requirements.

3.4 MOLAR MASS AND BRANCHING DISTRIBUTIONS Metallocene catalysts provide PE with control of molar mass and a narrower molar mass distribution than other catalysts. These PE will have precise control of

Chapter 3 Miscibility, Morphology, and Properties of Polyethylene Blends

67

Table 3.5 The Main Processing Methods for Polyethylenes and Their Requirements. Processing method

Polyethylene

Comment

Extrusion

All types

Cast-film extrusion

LLDPE, HDPE

Laminate-film extrusion

LLDPE, LDPE

Coextrusion

VLDPE–mPE

Blown-film extrusion

VLDPE–LDPE blends

Injection molding

HDPE

Vacuum forming

HDPE, VLDPE

Rotational casting

LDPE, LLDPE

Powder coating

LDPE, LLDPE

Extrusion is the main process for compounding with additives, fillers, pigments, blended polymers, and then forming pellets for subsequent processing. Further processing may form various profiles such as pipes, beading, wire, and cable coating Film is extruded through a slit-die then passed through chilled rollers to crystallize and smooth the surface; orientation by cold drawing PE can be extruded onto other substrates, paper, and paperboard mainly; though other polymer films and metal foils are used Two or more polymer films are extruded together for combination properties not available by blending; strength and heat-seal layers The tubular extrudate in expanded into a thinner film; melt strength to resist bursting is needed Molding of articles of many sizes, from rubbish bins to small fittings requires melt high flow and lubrication Partially melted sheet is rapidly shaped with high extensional flows and critical edge strengths Melting, flow, and particle coalescence are required; particularly flow under little more than gravity, though log times are involved Powder must adhere to hot objects, then flow under adsorption and surface tension forces

properties associated with molar mass, but this may not be effective in providing a broad range of properties. A broad molar mass distribution will combine good processing with sufficient strength, but with less precise control. Multisite catalysts give a broad molar mass distribution, but the molar mass can be controlled to give a range of grades for each polymer. Radical polymerization to manufacture LDPE gives the least control of molar mass and distributions or offer very broad due to the many long branches.

68

Polyolefin Blends

Bimodal molar mass distributions have become of increasing interest due to their combination of processability and mechanical properties. Combinations of toughness, tensile strength, and ultimate strain can be found in bimodal PE. Bimodal HDPE can retain modulus and yield strength, while providing some flexibility and toughness. The alternative of blending with a BPE often decreases modulus and tensile strength too much for applications such as extruded pipe. Resistance to creep is an important property retained in bimodal HDPE. The bimodal PE are usually not blends of differing molar mass, they are created in a polymerization reactor. Bimodal LLDPE are available where the high molar mass component contributes to tensile strength and creep resistance required in films, while the low molar mass facilitates processing and flexibility of the film. Branch distribution is the orthogonal distribution to molar mass distribution. Branching determines crystallinity and melting temperature range, so the mechanical properties are more dependent on branch distribution than molar mass distribution. As described above, catalyst technology has mainly revolutionized the copolymerization reaction. The single-site or metallocene catalysts produce a more uniform structure, in both molar mass and branch distribution. These catalysts are less selective of monomer, and the branch distribution becomes more statistical compared with multisite catalysts such as Zeigler–Natta. The distribution of short branches approaches that of LDPE, except that there are normally no long branches. Long branches can be introduced through single-site copolymerization. This is done by maximizing termination by disproportionation so that unsaturation is formed at chain ends. The end unsaturated chains then become macromonomers for further copolymerization with ethylene and the respective 1-alkene (2). The long branches are desired for their contribution to shear thinning and melt strength during processing. These rheological properties assist blown film production. Many LLDPE are compositionally heterogeneous, since they contain molecules with few branches and molecules that are highly branched. The more highly branched fraction can be separated by solvent extraction. Small-angle neutron scattering (SANS) has confirmed the prediction that structurally disperse LLDPE display liquid–liquid phase separation, even if the overall branch content is low. These LLDPE behave like polyethylene blends even though they are the result of a single polymerization. PE with high levels of 1-alkenes can be produced using single-site catalysts. The near statistical highly branched PEs are elastomeric. They have similar compositions to EP and EPDM, but they are formed from comonomers such as 1-butene, and their comonomer distribution is more uniform. They have processing characteristics similar to thermoplastics, as opposed to rubbers.

3.5 CRYSTALLIZATION, MELTING, AND BRANCHING OF POLYETHYLENES The melting temperature of PE varies over a broad range with the equilibrium melting temperature as the upper limit for LPE. The melting temperature varies with lamella

Chapter 3 Miscibility, Morphology, and Properties of Polyethylene Blends

69

thickness that in turn varies with crystallization conditions. Slow crystallization will give larger lamella thickness. Slow crystallization will occur at higher isothermal conditions or upon slow cooling. As the crystallization temperature (Tc ) approaches the melting temperature, the melting temperature approaches the equilibrium melting temperature (Tm ). Branching produces a new limiting temperature called the copolymer melting temperature (Tmc ). Tmc still depends on crystallisation conditions and represents the limiting case for Tc ¼ Tmc ; however, Tmc depends on the mole fraction of branches, or more specifically the noncrystallizable component. Branches suppress lamella morphology due to their causing variable crystallizable sequences (3). Branches are excluded from the lamella, thus the lamella thickness is limited by the length of ethylene segments, the methylene sequence length (MSL). Methyl and sometimes ethyl branches can be included in lamella if crystallization is fast; the resulting crystals are less well formed and the melting temperature is further reduced. Linear polyethylene (LPE or HDPE) has been blended with LLDPE with varying branch distributions. The critical branch content for phase separation was lower for mPE, where the branches were more evenly distributed, than for Ziegler– Natta LLDPE, where the branches are heterogeneously distributed. Where cocrystallization occurred, the unit cell of the crystals showed marked expansion (4). Crystallization conditions were shown to be important for the cocrystallization of blends of linear and branched polyethylenes, low isothermal temperatures promote cocrystallization as did quench cooling (5). Figure 3.4 shows the crystallization (on cooling at 10 C min1) followed by melting (on heating at 10 C min1) of ZN–octene–LLDPE. This slurry or solutionpolymerized LLDPE shows considerable molecular or branching heterogeneity according to the broad bimodal melting temperature range. There is a sharp higher

Figure 3.4 Crystallization and melting of ZN–octene–LLDPE after cooling and then heating at 10 C min1 using a Perkin-Elmer Pyris 1 DSC.

70

Polyolefin Blends

Figure 3.5 DSC melting curves for HDPE, LDPE, and their 1:1 blend.

melting double peak at 125 C, where the double peak is attributed to melting– recrystallization–melting during the DSC scan. Crystallization similarly shows a sharp higher temperature exotherm followed by a broad large exotherm (not both exotherms and endotherms are positive since the heat flow data have been converted to apparent specific heat by baseline corrections and calibration). DSC melting scans of blends of polyethylenes reveal changes in all aspects of crystallinity as shown by the melting peaks, though some peaks of the original components are retained in modified form. This indicates different crystal populations that may arise from liquid–liquid immiscibility or fractionation of a singlephase liquid during crystallization. Figure 3.5 shows melting of HDPE and LDPE and their blend. The HDPE in the blend melts similarly to the pure HDPE but at a slightly lower temperature. The melting of LDPE in the blend occurs at the same temperatures as the pure LDPE, but the main peak at 110 C is much reduced indicating that LDPE is miscible with the HDPE in the amorphous phase and that crystallinity of the blend is much reduced. Figure 3.6 shows the melting of a blend of LDPE–VLDPE and the melting of the pure components. The individual peak melting temperatures are retained in the blend, but the distribution of crystals melting under each peak has changed. This indicates miscibility of the components with similar branching distributions to provide a new branching distribution in the liquid phase, which may exist as a phase separated liquid phase. A small low temperature endotherm shifted from about 90 C in pure LDPE, to 80 C in the blend. The LDPE–ULDPE blend shown in Fig. 3.7 provides the greatest difference between the PE in these examples. ULDPE melts between 30 and 75 C because it is highly branched; such PE are called thermoplastic polyolefin elastomers (TPO). ULDPE and LDPE in the blend seem to mainly melt independently, though with overall decreased crystallinity. The main peak shift is the LDPE lower melting temperature peak shifting from 90 C to about 80 C. The melting endotherms of LDPE and

Chapter 3 Miscibility, Morphology, and Properties of Polyethylene Blends

71

Figure 3.6 DSC melting curves for LDPE, VLDPE, and their 1:1 blend.

ULDPE do not overlap, so it can be assumed that the branching distributions do not overlap making it unlikely that any components from each polymer would be dissolved in the other. Within a BPE with a distribution of branches, there will be a distribution of lamella thickness. This will result in a broad melting range. BPE with a bimodal distribution of branches will have a bimodal distribution of lamella thickness and a corresponding melting temperature range. When two polyethylenes are blended, assuming they are miscible, they will cocrystallize only where they have common MSL. Some molecular segments in each BPE will crystallize independently of

Figure 3.7 DSC melting curves for LDPE, ULDPE, and their 1:1 blend.

72

Polyolefin Blends

the other BPE in the blend. Note that it is the segments that enter lamella, not whole molecules. Where there are intramolecular branch distributions, then segments of the one molecular can join different lamella as appropriate to their MSL. Thermal fractionation techniques have been used to separate crystals with differing lamella thickness formed from molecular segments with differing methylene sequence lengths. Stepwise isothermal cooling (SC) has been used to allow each fraction to crystallize according to the temperature appropriate to each methylene sequence length, where longer sequences can crystallize at higher temperatures (6–8). Another technique that provides better resolution of the fractions is successive self-nucleation and annealing (SSA) (9). SSA has been used to fractionate ethylene–hexene copolymers to distinguish the contribution of catalyst on the branch distribution and structure (10).

3.6 MISCIBILITY AND CRYSTALLIZATION Polyethylenes have a broad distribution of melting temperatures resulting from varying lamella thickness. The lamella thickness is determined by the fold length, which is limited by MSL. At each crystallization temperatures, only those molecules with sufficient MSL can crystallize. If a blend is homogeneous, then crystallization will proceed, which is dependent on MSL and temperature, irrespective of the component from where the molecules originated. The overall distribution of lamella thickness will be a combination of the lamella thicknesses if each component were crystallized separately. Difference in the distribution will depend on the conditions of crystallization, such as rate of cooling. This situation will provide a broadened and probably bimodal melting endotherm. In this case phase separation does not occur, but fractionation according to MSL occurs for the combined blend components. If the PEs in a blend are immiscible at the temperature of crystallization, then each phase must crystallize independently and there will normally be a disperse phase and a continuous phase, except for nearly equal volume fractions where the phases could be cocontinuous. Though the phases crystallize independently, they will not crystallize at the same temperatures, the crystallization temperatures will depend on branch density. This situation will provide a bimodal melting endotherm and a double crystalline morphology. Note that the two crystalline phases will not have the composition of the original PE. The composition of the phases will be according to the tie-lines of the phase diagram (Fig. 3.8). As the blend is cooled and the mutual solubility changes, the composition of the two phases will change. Descriptively, the more linear PE will dissolve the more linear molecules from the more branched PE of the blend. Conversely, the more branched PE will dissolve the more branched molecules from the less branched PE. In this case phase separation occurs before crystallization and continues during crystallization. The phase equilibria correspond to a typical U-shaped upper critical solution temperature diagram. The competition between liquid–liquid phase separation and crystallization from homogeneous melt has been observed for

Chapter 3 Miscibility, Morphology, and Properties of Polyethylene Blends

73

Figure 3.8 Temperature–composition phase diagram for an immiscible polyethylene blend with differing branching.

blends of ethylene copolymers with hexene and butene, using time resolved SAXS and WAXS, complemented by optical and atomic force microscopies (11). Similar conclusions were made concerning blends of ethylene copolymers of butene, hexene, and octane (12–14). In both miscible and immiscible blends, the resulting crystalline morphology will be complex and there will be a broad melting temperature range. Properties will be significantly modified by the blend morphology and events such as impact; elongation or creep conditions will differ. Only a small amount of a modifying PE may be necessary to provide significant changes. If miscibility is required then the two polyethylenes must be similar, but different enough to change morphology in the blend. Blends of LLDPE with LDPE with only small differences in branch composition (3% and 5%, respectively) were found to be immiscible in the liquid (15). Blends of HDPE with long-chain branched polyethylenes (HBPE) prepared from metallocene catalysts have been studied by DSC and their crystal structures interpreted in terms of phase behavior. The HBPE contained long-chain branches and short branches formed form octane comonomer. HBPE with 7.5–12.0% octane exhibited phase separation, whereas HBPE with 2% octane were found to be miscible with HDPE over the whole composition range. Long branches were few and did not contribute to the immiscibility (16). Crystallization and melting of ternary blends containing mPE, LLDPE, and LDPE have been studied where the mPE and LLDPE varied in content and had the same melting temperature, Tm ¼ 122 C and a fixed content of LDPE, Tm ¼ 114 C. Crystallinity increased with mPE content, and it was proposed that mixed crystals may be formed since no separated melting peaks were observed for the component PE (17). Blends of LPE and several different poly(ethylene-cohexene)s were shown,

74

Polyolefin Blends

by DSC measurements of melting, to be immiscible even with only 1.8 mol % hexene (18). SSA has been used to determine the miscibility of linear and branched polyethylenes and the results showed that only those PE fractions that were similar in branch content and distribution of short branches were miscible in the melt. The SSA thermal fractionation helped to distinguish miscibility effects from cocrystallization effects (19). Phase diagrams have been proposed for blends of polyethylenes where the first component is linear, or less branched, and the second component is more branched. The method involves quench cooling each blend composition from the melt at various temperatures, so that there is insufficient time for liquid–liquid phase separation. Separated blends are considered to be separated at the particular temperature of the melt prior to quenching. TEM and DSC are used to characterize the quenched blends. The phase diagrams exhibit UCST behavior, depending on the difference in branching content of the component polyethylenes. Branch length has been found not to be important since it is the branch points that are excluded from the crystals (20,21).

3.7 THEORETICAL PREDICTION OF MISCIBILITY The miscibility of polyolefins in the melt is difficult to predict. Their miscibility or immiscibility is in contrast to the chemical similarity of their structural groups: methyl, methylene, and methyne. Polyolefin melts have similar structures, including density–temperature relationships and optical characteristics such as the refractive index, making most characterization techniques inappropriate for the detection of phase separation. The general rule of like dissolves like being the determinant of miscibility is inadequate for polyolefin miscibility. Some blends of polyethylenes are considered to phase-separate the above crystallization temperatures of the components according to an upper critical solution temperature spinodal phase diagram (Fig. 3.8). Small-angle X-ray scattering and hot-stage optical microscopy have not provided definitive evidence of phase separation. Crystallization of polyethylene from solution has been shown to follow selfnucleation to form gels where the network links are formed by the crystals. Crystallization occurs at a temperature-dependent critical gelation concentration. The lamella may join edge-to-edge when near coplanar (22). Crystallization from miscible polyethylene blends is expected to be consistent with solution gelation studies. Two processes are predicted to occur on solidification of polyolefin blends. Crystallization from a homogeneous melt of the blend where the composition of the crystalline phase will initially contain molecules with the longest methylene sequence length irrespective of which component from which the molecules originated. Alternatively liquid–liquid separation may occur first, followed by independent crystallization of the two liquid phases, and in this case the composition of the liquid phases will be determined by the temperature and associated tie-lines of the phase diagram for the blend. Analogous phase separation mechanisms have been observed for polyethylene–polypropylene blends (23–25).

Chapter 3 Miscibility, Morphology, and Properties of Polyethylene Blends

75

Studies, where one of the blended polyethylenes was deuterated, have revealed immiscibility of polyolefins with SANS, by exploiting the difference in coherent scattering lengths between proton and deuterium nuclei to provide suitable enhanced scattering contrast. SANS techniques have been applied to polyolefin blends to investigate the thermodynamic interactions relating to phase behavior and purecomponent pressure–volume-temperature (PVT) relationships. Qualitative predictions of the miscibility of model homopolymer blends of PE and deuterated PE have been achieved by simulation. Optical microscopy, differential scanning calorimetry, rheological measurements, and scanning electron microscopy have been applied. Molecular architecture and the effects of molar mass have been studied in relation to phase separation and PVT data (26). Thermodynamic interactions in polyolefin blends have been shown to originate from induced- dipole forces but they depend upon the component structures. SANS has been used to determine interactions of blends, including solubility data, light scattering to determine phase boundaries, and PVT measurements to establish cohesive energies. Isotope and microstructure interactions have been shown to contribute to the Flory–Huggins interaction parameter in mixtures of deuterated and protonated PE. Lack of a single dominant thermodynamic parameter determining phase behavior means polyolefin blends are influenced by small variations in molecular architecture. Binary phase behavior was correlated with statistical segment length asymmetry and its dependence on the Flory–Huggins interaction parameter. Additional contributions from temperature, composition, and molar mass have been established, so Flory–Huggins theory of polymer blends and their thermodynamics has been extended to include them. Miscibility loop behavior has been predicted and described using a temperature-dependent interaction parameter, leading to an UCST phase diagram. The predictions have been attributed to nonrandom packing effects where constituents have sufficient asymmetry in the polymerization indices (27). Molecular packing may have a dominant influence on polyolefin miscibility, so the accurate prediction of polyolefin melt structures would enable understanding of reasons for the difficulty of PE miscibility prediction to be obtained. In the melt, molecules of a homopolymer polyolefin, such as LPE, generally exist in random conformations, but some order is present because of weak intramolecular and intermolecular interactions, such as dispersive forces. Molecular conformation is determined by the configuration of branches, which influence the packing of the PE chains, determining the melt structure, density, and, hence, the miscibility of PE in the melt. The physical properties of the melt are thus determined by conformation. The branch structure, length, frequency, and distribution determine the most favorable conformations. The conformations allow or limit interactions to the extent that PE with differing branch structures may be immiscible. The thermodynamic stability of a single liquid phase has been theoretically predicted by computation of pure component and blend parameters dependant upon entropic considerations (28). Microphase separation has been ascribed to a considerable increase in unfavorable noncombinatorial entropy, using a Flory–Huggins equation to predict the free energy of mixing (29).

76

Polyolefin Blends

3.8 RHEOLOGY OF MELTED POLYETHYLENE BLENDS Linear polyethylene rheology is dependant on the critical molar mass for entanglements (940 g mol1 (30), which is low due to lack of chain stiffening groups or side chains. Chain entanglements and branch entanglements provide non-Newtonian rheology for higher molar mass PE. Entanglements of branches increase the melt strength of PE with long branches, especially LDPE, and also some newer mPE that are polymerized to contain long branches in addition to the short branches from a comonomer. Short branches do not interact or entangle sufficiently to provide melt strength. PE with differing branch content have been shown to form immiscible liquid–liquid mixtures. The rheology of immiscible PE blends deviate from additivity expected for a homogeneous liquid. For example, HDPE and LLDPE are miscible when the branch content is low (<4 branches/100Cs). However for >8 branches/ 100 Cs, phase separation takes place. The rheology of melted polyethylene blends is of importance in processing (31). Processing is often by extrusion where the main product is film, either from the blown film process or cast film. The molecular structure and hence meltingtemperature profile determines screw design. Usually PE blends melt over a broad temperature range, so a long compression zone is required. Orientation of blown film requires high melt strength that is enhanced by long branches while subsequent film strength is dependent on crystallinity requiring long methylene sequence lengths and so few branches. The optimum structure for processing and strength of PE films is best achieved by blending different PE. Injection molding requires high melt flow and hence low molar mass, while the mechanical properties will require high molar mass; the optimum properties may require a blend of PE with differing molar mass, though bimodal molar mass distribution. Foams are prepared by gas injection molding, and successful foams require high melt strength. Rotational casting, powder coatings, heat sealing all require melt flow under relatively low shear, so non-Newtonian characteristics may not have the opportunity to become apparent. Rheology can be modified by processing aids and lubricants and some of these are low molar mass PE, such as PE waxes and microcrystalline waxes. Rheology has been used to interpret the miscibility of polyethylene blends in lieu of direct measurement of phase separation. The sensitivity of various models has been shown suitable for distinguishing homogeneous melts from liquid–liquid phase separated melts (32,33). Such phase separated melts may be of advantage in extrusion processing of polyethylene blends; particularly in blown film formation where melt strength and rheological characteristics are critical to bubble stability, film uniformity, orientation, and clarity. Metallocene poly(ethylene-cobutene)s were blended with LDPE and the viscosities were increased at low frequencies though shear thinning was more evident. The relaxations were retarded in the presence of LDPE and the effects were more noticeable for the metallocene copolymers with higher butene content (34). Metallocene HDPE and metallocene LLDPE blends were confirmed by rheological analysis to be miscible, contrary to blends of the same metallocene HDPE with LDPE (35).

Chapter 3 Miscibility, Morphology, and Properties of Polyethylene Blends

77

3.9 MECHANICAL PROPERTIES OF POLYETHYLENE BLENDS PE blends are compatible; they can be miscible, cocrystallize, or crystallize with separate crystallite distributions. These morphologies effect modulus, yield and any double yield phenomena, strain hardening, elasticity and recovery, creep, creep recovery, stress relaxation, tear strength. These properties are all measured using static stress or strain. Typical tensile stress–strain curves for LLDPE, LDPE, and a blend of LDPE–EP rubber are shown in Fig. 3.9. LDPE shows a sharp yield, plastic flow with some strain hardening with increasing strain. The LDPE–EP blend is a thermoplastic vulcanizate, though it still yields, it extends with constant stress. LLDPE shows a sharp yield followed by plastic flow, until extensive strain hardening occurs over the strain range 4 to 13. Strain hardening is a problem for films where rapid stretching to high strain is required, such as pallet wrap, stretch wrap, and agricultural films. The film exhibits decreased elasticity during strain hardening, as crystallinity is oriented in the direction of flow and the crystalline morphology changes. Figure 3.10 shows atomic force microscopy of the Ziegler–Natta-catalyzed LLDPE containing 5% hexene, shown in the stress–strain curve (Fig. 3.9). The original crystals become elongated and orientated during straining with loss of film elasticity, increase in surface roughness, and loss of film transparency. Blending with LDPE or metallocene VLDPE will reduce the strain hardening. Polyethylenes undergo interlaminar deformation during tensile elongation. This can give rise to a double yield phenomenon. At the onset of the first tensile yield, chain slip and lamella rotation occur and this process is reversible. The second tensile yield is irreversible and coinsides with lamella fragmentation (36). These mechanically induced morphological changes and the observation of any double yield phenomena are dependent on several structural factors that can be controlled by

Figure 3.9 Tensile stress–strain curves for LLDPE, LDPE, and LDPE–EP blends.

78

Polyolefin Blends

Figure 3.10 Atomic force microscopy of ZN–hexene–LLDPE before strain (a) and after strains of 3 (b) and 6 (c) (10 mm 10 mm).

polyethylene blending. Homogeneous or heterogeneous slips have been identified as responsible for different macroscopic deformation regimes. Homogeneous slip, arising from dislocations within the crystals, has a higher thermal activation and predominates as temperature increases or strain rate decrease. Heterogeneous slip, operating through defective crystal block boundaries, proceeds with lower strain hardening (37). Dynamic mechanical properties exhibit side chain or branch motions; short main chain segment motions, main chain segmental motions, recrystallization, and melting. These transitions are observed as inflections in the storage modulus curve with temperature, peaks in either the loss modulus or damping factor (tan(d)) curves. Figure 3.11 shows the dynamic mechanical spectroscopy (DMS) of a ZN–VLDPE at 1 Hz in tensile mode. The glass transition temperature (maxima of the loss modulus

Chapter 3 Miscibility, Morphology, and Properties of Polyethylene Blends

79

Figure 3.11 Dynamic mechanical spectroscopy of ZN–VLDPE in tensile mode using frequency of 1 Hz.

peak, or the damping factor peak, as often used) is at about 40 C. The transitions of individual components of a blend are often detected independently. The activation energy for each of the transitions can be determined using multifrequency data and correlation with the Arrhenius equation. Williams–Landel–Ferry (WLF) analysis of the multifrequency data is used to construct time (frequency)–temperature master curves to obtain plateau or terminal viscoelastic properties. These properties reveal the contributions of differing branching levels or distributions within a blend. This is illustrated in Fig. 3.12 that shows a LDPE–EP rubber blend 1, 2, 5, 10, and 20 Hz in

Figure 3.12 Dynamic mechanical spectroscopy of LDPE–EP rubber blend in tensile mode using synthetic frequency multiplexing of 1, 2, 5, 10, and 20 Hz.

80

Polyolefin Blends

tensile mode measured using synthetic frequency multiplexing. The glass transition temperature (50 C) is shifted to higher temperatures with higher frequencies. A second, lower temperature transition is shown in the range 70 to 90 C. Metallocene poly(ethylene-cooctene)s showed changes in the b-transition region that were related to motions of segments located at interfacial regions (38).

3.10 ADDITIVES Many additives are included in PE and their blends. These include colorants, stabilisers, ultraviolet protection, flame retardants such as antimony oxide and decabromodiphenyl ether, or halogen free flame retardants such as magnesium and aluminum hydroxides. Prodegradents are an alternative to stabilizers, such as prooxidants that are currently being used in some packaging films to reduce environmental impact. Surface properties such as tack, antiblocking, corona discharge treatment for printing, and its subsequent reversion over time are modified by additives. Additives for processing include internal and external lubricants. Microcrystalline waxes, hydrocarbon waxes, stearic acid, and metal stearates such as calcium, magnesium, and zinc stearates are common lubricants. Often blends with lower melting temperature polyethylenes, particularly those with long branches are preferred for processing enhancement. Elastomeric or plastomeric polyethylenes often require cross-linking, so radical initiators such as cumyl peroxide are added, or alkoxyvinylsilanes grafted for subsequent water-induced cross-linking.

3.11 CONCLUSIONS The polyethylenes are a large and diverse family of chemically similar polymers; increased understanding is required to learn how to combine them to achieve blends with a custom range of properties. Single-site (metallocene)-catalyzed PE have brought a new life cycle to PE developments and applications. These new PE have initially found applications in blends with existing PE whose characteristics are well known but require processing and performance improvement. Most PE blends are with other PE to provide combinations of molar mass, short branching, and long branching. Other blends are with PE copolymers with polar or functionalized comonomers so that increased surface polarity and functionalization can be included. Production and cost pressures increase the performance requirements for PE. PE can be enhanced to replace other polymers as well as expanding into areas occupied by other materials. The similarity of the chemistry and lack of distinguishing features for many characterization techniques make investigation of the structure–property– performance of PE blends a difficult and often controversial task. Critical requirements for PE are rapid processing, thinner yet stronger films, ease of direct printing or adherence of labels, rapid heat sealing, high strain rate at high strain for stretch wrap, plastic memory for shrink wrap, cross-linkability to resist creep and stress relaxation, cross-linking to resist elevated temperatures when the

Chapter 3 Miscibility, Morphology, and Properties of Polyethylene Blends

81

low melting temperature of PE would otherwise be a constrain, increased elasticity covering the range from traditional elastomers to stiff thermoplastics. PE compositions need performance compatible with newer high speed processing equipment and high speed packaging and finishing machinery. The new PE and their blends are meeting these needs and opening new opportunities where they easily meet or exceed expectations. All of these properties and structural diversity are now being derived from the polymer with the simplest chemical structure.

NOMENCLATURE Tg Tm Tm Tmc Tc PE LPE mPE M HDPE BPE LLDPE VLDPE ULDPE LDPE TPO TPE VAc MA BA GMA AA MAn MSL SANS PVT UCST SAXS WAXS SSA SC HBPE DMS ZN

Glass transition temperature Melting temperature Equilibrium melting temperature Copolymer melting temperature Crystallization temperature Polyethylene Linear polyethylene Metallocene polyethylene Molar mass High density polyethylene Branched polyethylene Linear low density polyethylene Very low density polyethylene Ultralow density polyethylene Low density polyethylene Thermoplastic polyolefin Thermoplastic polyolefin elastomer Vinyl acetate Methyl acrylate Butyl acrylate Glycidyl methacrylate Acrylic acid Maleic anhydride Methylene sequence length Small-angle neutron scattering Pressure–volume–temperature Upper critical solution temperature Small-angle X-ray scattering Wide-angle X-ray scattering Successive self-nucleation and annealing Stepwise isothermal cooling Long-chain branched polyethylenes Dynamic mechanical sprectroscopy Ziegler–Natta catalyst

82 EP TPO

Polyolefin Blends

Ethylene–propylene copolymer elastomer Thermoplastic polyolefin elastomer

REFERENCES 1. S. P. Chum, C. I. Kao, and G. W. Knight, Structure, properties and preparation of polyolefins produced by single-site catalyst technology, in: Metallocene-Based Polyolefins: Preparation, Properties and Technology, Wiley Series in Polymer Science, Vol. 1, J. Schiers and W. Kaminsky (eds.), Wiley, New York, 2000, p. 261. 2. G. D. Wignall, R. G. Alamo, J. D. Londono, L. Mandelkern, and F. C. Stehling, Macromolecules, 29, 5332 (1996). 3. F. M. Mirabella, J. Polym Sci. B: Polym. Phys., 44, 2369 (2006). 4. Y. Zhao, S. Liu, and D. Yang, Macromol. Chem. Phys., 198, 1427 (1996). 5. M. J. Galante, L. Mandelkern, and R. G. Alamo, Polymer, 39, 5105 (1998). 6. G. Amarasingh, F. Chen, A. Genovese, and R.A. Shanks, J. Appl. Polym. Sci., 90, 681 (2003). 7. R. A. Shanks and G. Amarasinghe, Application of differential scanning calorimetry to analysis of polymer blends, in: Polymer Characterization Techniques and Their Application to Blends, G. Simon (ed.), Oxford University Press, New York, 2003, pp. 22–67. 8. P. Starck, P. Lehmus, and V. Seppala, Polym. Eng. Sci., 39, 1444 (1999). 9. A. Muller and M. L. Arnal, Prog. Polym. Sci., 30, 559 (2005). 10. M. Bialek, K. Czaja, and B. Sacher-Majewska, Thermochimica Acta, 429, 149 (2005). 11. Z. Wang, H. Wang, K. Shimizu, J. Y. Dong, B. S. Hsiao, and C. C. Han, Polymer, 46, 2675 (2005). 12. G. Matsuba, K. Shimizu, H. Wang, Z. Wang, and C. C. Han, Polymer, 44, 7459 (2003). 13. G. Matsuba, K. Shimizu, H. Wang, Z. Wang, and C. C. Han, Polymer, 45, 5137 (2004). 14. K. Shimizu, H. Wang, Z. Wang, G. Matsuba, H. Kim, and C. C. Han, Polymer, 45, 7061 (2004). 15. M. J. Hill and C. C. Puig, J. Appl. Polym. Sci., 65, 1921 (1997). 16. J. Schellenberg and B. Wagner, J. Thermal Anal., 52, 275 (1998). 17. M. Run, J. Gao, and Z. Li, Thermochim. Acta, 429, 171 (2005). 18. B. S. Tanem and A. Stori, Polymer, 42, 5389 (2001). 19. M. L. Arnal, J. J. Sanchez, and A. J. Muller, Polymer, 42, 6877 (2001). 20. M. J. Hill, P. J. Barham, A. Keller, and C. C.A. Rosney, Polymer, 32, 1384 (1991). 21. M. J. Hill and P. J. Barham, Polymer, 38, 5595 (1997). 22. J. H. Kim and R. E. Robertson, Polymer, 39, 5371 (1998). 23. M. Razavi-Nouri and J. N. Hay, Polym. Eng. Sci., 46, 889 (2006). 24. J. Li, R. A. Shanks, and Y. Long, J. Appl. Polym. Sci., 87, 1179 (2003). 25. J. Li, R. A. Shanks, and Y. Long, Chin. J. Polym. Sci., 20, 497 (2002). 26. H. M. Freischmidt, R. A. Shanks, G. Moad, and A. Uhlherr, J. Polym. Sci. B: Polym. Phys., 39, 1803 (2001). 27. H. W. Kammer, Polymer, 40, 5793 (1999). 28. G. H. Fredrickson and A. J. Liu, J. Polym Sci. B: Polym. Phys., 33, 1203 (1995). 29. H. W. Kammer, Polym. Netw. Blends, 5, 69 (1995). 30. H. G. Elias, An Introduction to Polymer Science, VCH Publishers Weinheim, New York, 1997, Chapter 7. 31. L. A. Utracki, Polymer Alloys and Blends: Thermodynamics and Rheology, Hanser Publishers, Munich, 1989, p. 201.

Chapter 3 Miscibility, Morphology, and Properties of Polyethylene Blends

83

32. Y. Fang, P. J. Carreau, and P. G. Lafleur, Polym. Eng. Sci., 45, 1254 (2005). 33. F. Chen, R. A. Shanks, and G. Amarasinghe, J. Appl. Polym. Sci., 95, 1549 (2005). 34. J. Xu, X. Xu, Q. Zheng, L. Feng, and W. Chen, Eur. Polym. J., 38, 365 (2002). 35. C. Liu, J. Wang, and J. He, Polymer, 43, 3811 (2002). 36. M. F. Butler and A. M. Donald, Polymer, 38, 5521 (1997). 37. V. Gaucher-Miri and R. Seguela, Macromolecules, 30, 1158 (1997). 38. A. G. Simanke, G. B. Galland, L. Freitas, J. A. H. da Jornada, R. Quijada, and R. S. Mauler, Polymer, 40, 5489 (1999).

Chapter

4

Miscibility and Crystallization Behavior in Binary Polyethylene Blends Moonhor Ree1

4.1 INTRODUCTION Polyethylene (PE) is currently the most widely used commercial polymer throughout the world, and the industrial market for this polymer is still growing due to the variety of applications that are based around its superior properties, such as high chemical and mechanical resistances, easy processability, and low specific gravity, as well as low manufacturing costs. So far, several types of PE materials have been developed, each with different levels of mass density, branching content, and molecular weight. These PE materials include linear high density PE (HDPE), linear ultrahigh molecular weight PE (UHMWPE), linear low density PE (LLDPE), and low density PE (LDPE) (Fig. 4.1) (1–5). These polymers are usually synthesized by the conventional polymerization of the ethylene monomer, or with one or more a-olefin comonomers (1-butene, 1-hexene, 1-octene, etc.) with the aid of Ziegler–Natta catalysts (1–5). Advances in the catalysts used for the polymerization of olefins have allowed manufacturers to expand their product line to accommodate a wide range of densities and branching structures. In particular, new high performance PEs are mainly produced via the use of metallocene catalysts (i.e., single-site catalysts), which have been shown to have superior properties compared with conventional PEs (6,7). The single-site nature of the catalysts allows for the formation of polymers

1

Department of Chemistry, Polymer Research Institute, Pohang Accelerator Laboratory, National Research Lab for Polymer Synthesis and Physics, and Center for Integrated Molecular Systems, Pohang University of Science & Technology (Postech), Pohang 790-784, Republic of Korea Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

84

Chapter 4 Miscibility and Crystallization Behavior in Binary Polyethylene Blends

85

Figure 4.1 Schematic molecular structures of LDPE, LLDPE, HDPE, and UHMWPE.

with a narrower molecular weight distribution and a more uniform distribution of comonomers, as compared with conventional PEs (6,7). HDPE gives rise to stiffer materials with good tensile properties but poor impact and tear resistance (1,5). Moreover, UHMWPE has excellent mechanical properties but very poor melt properties, causing severe processing difficulties (4,5). LLDPE, which is synthesized by copolymerization of ethylene with one or more a-olefin comonomers, has short chain branching and presents excellent mechanical properties, such as impact and tear resistance, as well as high tensile strength (1–5). However, due to a narrow molar mass distribution, LLDPE exhibits an elevation in the melt viscosity and a lowering of the melt strength, causing relatively poor processability (1–5). In general, metallocene-catalyzed LLDPE is synthesized with a much narrower molecular weight distribution, compared to that of Ziegler–Nattacatalyzed LLDPE, and exhibits very poor melt properties that cause processing difficulties (6–15). On the contrary, LDPE is a polymer composed of both long and short chain branches and is well known to have good melt properties and therefore efficient processability (16). In general, the processing behavior of PEs mainly depends on parameters such as the weight-average molecular weight and distribution, comonomer content, and the long-chain-branching content and distribution (17–22). In particular, the longchain-branching content is known to have a large impact on polymer processability; PE polymers with long-chain-branching demonstrate excellent processing properties (18,23).

86

Polyolefin Blends

These PE polymers offer a broad spectrum of structures, properties, and applications. However, the blending of different types of PEs (HDPE, LDPE, LLDPE, and UHMWPE) has attracted growing interest because of the potential for obtaining low cost materials with improved mechanical properties and better processabilities, as compared to those of the pure constituents (1–34). Nowadays, 70% of PEs in the market are blends (24). The processability and properties of PE blends are dependent on the melt miscibility. Moreover, the properties are also dependent on the morphological structure of the blend, which is basically a combination of the crystallization behavior and melt miscibility. Therefore, the miscibility and crystallization behavior of PE blends have been prevalent research topics over the last two decades.

4.2 MISCIBILITY 4.2.1 Linear and Short Branched Polyethylene Blends Ree (5) first studied quantitatively the miscibility of two HDPE/LLDPE blend systems using the small-angle neutron scattering (SANS) technique: (i) blends of a HDPE-d (237,000 weight-average molecular weight Mw , 2.06 polydispersity PDI, and 98% deuteration) and a LLDPE-B (114,000 Mw , 4.50 PDI, and 18 ethyl branches per 1000 backbone carbons); (ii) blends of a HDPE-d (109,000 Mw , 1.80 PDI, and 98% deuteration) and a LLDPE-O (96,000 Mw , 4.50 PDI, and 15 hexyl branches per 1000 backbone carbons). For these blends in the melt, the Flory–Huggins interaction parameter x was determined to be small, with positive values of the order 10 4, but did not exceed the critical x values, that is, the upper limit of the stability of a single miscible phase. Moreover, these blends revealed no SANS cloud point over the temperature range 140–300 C, indicating that the HDPE blend systems with LLDPE-B and LLDPE-O are both miscible in the melt. Nicholson et al. (35) extended the SANS analysis approach to accommodate blends of HDPEs with hydrogenated and deuterated polybutadienes (PB, a LLDPE-B and PB-d, a LLDPE-B-d) with different amounts of ethyl branches ranging from 18 to 106 per 1000 backbone carbons: HDPE (111,830 Mw and 1.23 PDI), HDPE-d (116,580 Mw and 1.27 PDI), LLDPE-B-d(1) (102,660 Mw , <1.05 PDI, and 18 ethyl branches), LLDPE-B-d(2) (22,620 Mw , <1.05 PDI, and 22 ethyl branches), LLDPE-B(1) (161,650 Mw , <1.05 PDI, and 39 ethyl branches), LLDPE-B(2) (101,230 Mw , <1.05 PDI, and 65 ethyl branches), LLDPE-B(3) (76,850 Mw , <1.05 PDI, and 88 ethyl branches), and LLDPE-B(4) (179,670 Mw , <1.05 PDI, and 106 ethyl branches). Alamo et al. (36) and Agamalian et al. (37) also carried out SANS measurements on the binary blends prepared from hydrogenated and deuterated HDPEs (94,000–201,000 Mw and 1.03–3.4 PDI) and hydrogenated and deuterated PBs (LLDPE-Bs: 21–106 branches, 78,000– 123,000 Mw , and <1.07 PDI). Both groups found that the melt miscibility of the HDPE/LLDPE-B blend is dependent on the extent of ethyl branches; in the blend, the LLDPE-B component containing more than 60–80 branches per 1000 backbone carbons induced phase separation (35–37). The miscible blends in the melt

Chapter 4 Miscibility and Crystallization Behavior in Binary Polyethylene Blends

87

were found to have x values on the order of 10 4 (35–37). The effect of the ethyl branches on the melt miscibility of the HDPE/LLDPE-B blends was further confirmed by molecular dynamics simulations (38–41). A similar branching effect on the miscibility was observed in the HDPE/ LLDPE-O blends, which were prepared from HDPE (49,400 Mw and 3.60 PDI) and LLDPE-O polymers (69,200–104,000 Mw and 1.8–3.40 PDI) with 2–87 hexyl branches per 1000 backbone carbons (42). It was observed that the critical branch number in the LLDPE-O component capable of causing immiscibility in the HDPE/ LLDPE-O blend was 50 branches per 1000 backbone carbons, as determined using inverse gas chromatography, rather than the SANS technique. The miscibility of HDPE/LLDPE blends was also studied using differential scanning calorimetry (DSC), dynamic mechanical thermal analysis (DMTA), rheological analysis, excimer fluorescence analysis, and transmission electron microscopy (TEM). Lee et al. (43) studied blends of HDPE (121,000 Mw and 16.8 PDI) with LLDPEs (LLDPE-B: 89,300 Mw and 3.8 PDI; LLDPE-O: 93,100 Mw and 3.6 PDI) having 15–16 branches per 1000 backbone carbons using DSC and DMTA, and found the blends to be miscible. Lee and Denn (44) conducted rheological measurements on the blends of HDPE (40,000 Mw and 3.1 PDI) with LLDPE-O polymers (68,300–134,000 Mw , 3.2–3.8 PDI, and 10–16 hexyl branches per 1000 backbone carbons), and found that the HDPE/LLDPE-O blends are miscible in the melt over the whole composition range. Zhao et al. (45) attempted to examine the miscibility of HDPE/LLDPE blends using excimer fluorescence analysis; the LLDPE component in these blends was labeled with a chromophore. They found that the LLDPE component is miscible with the blended HDPE. It is noteworthy that in such studies using chromophores, the effect of the labeled chromophore on the blend miscibility should be considered. Tabtiang et al. (46) used rheological analysis to study the observed immiscibility of blended HDPE (45,000 Mw ) with LLDPE-O having a very high content of branches, typically 125 hexyl branches per 1000 backbone carbons. Furthermore, Hill et al. (47, 48) reported on the use of DSC and TEM analyses to study the partial miscibility of quenched HDPE/LLDPE blends (quenched from the melt). The authors studied the miscibility of blends of HDPEs (50,000–100,000 Mw and 2.8– 3.5 PDI) with LLDPE-B polymers (10,000–160,000 Mw , 2.0–5.0 PDI, and 0.5–22 ethyl branches per 1000 backbone carbons) as well as LLDPE-O polymers (37,000– 130,000 Mw , 2.0–2.2 PDI, and 10.5–59 hexyl branches per 1000 backbone carbons). They found that the blends with LLDPE polymers having 10–40 branches per 1000 backbone carbons exhibited a closed loop phase diagram at low HDPE concentrations. The obtained phase diagrams are quite different from those obtained from the SANS results discussed above. In fact, in the DSC and TEM study, Hill et al. assumed that the phase structure of the blends in the melt is retained in the blend samples quenched from the melt, and if any crystallization is involved in such the quenched blend samples, the structure of the formed crystals is dependent on the phase structure. It was determined that the quenched blend samples do not have zero crystallinity. On the contrary, for the above SANS studies it is noted that the used LLDPE-B-d samples were prepared from deuterations of the polybutadienes, which

88

Polyolefin Blends

are different from LLDPE-B samples synthesized by the copolymerization of ethylene and 1-butene. As reviewed above, the miscibility of the HDPE/LLDPE blends seems to depend on the molecular weight, polydispersity, composition, and temperature, as well as on the branch content and length of the components. However, the HDPE and LLDPE samples used in the blend studies were limited to certain ranges of molecular weight, polydispersity, and branch content. Further, there was a distinct lack of information regarding the phase diagram. Therefore, more quantitative analysis is still needed in order to completely understand the miscibilities and phase diagrams of HDPE/ LLDPE blends.

4.2.2 Blends of Linear and Long Branched Polyethylenes Ree (5) first studied in detail the miscibility of HDPE/LDPE blends using the SANS technique: HDPE (198,500 Mw and 2.28 PDI), HDPE-d (237,000Mw , 2.06 PDI, and 98% deuteration) and LDPE (286,000 Mw , 15.98 PDI, and 27.6 branches per 1000 backbone carbons: 26 short and 1.6 long branches). For the blends in the melt, x was determined to have small, positive values on the order of 10 4, which did not exceed the critical x values, that is, the upper limit of the stability of the single miscible phase. Moreover, the blends did not show a SANS cloud point over the temperature range 140–300 C. These SANS results indicated that the HDPE/LDPE blends are miscible in the melt. Alamo et al. (49) extended the SANS analysis to blends of HDPEs (101,000–201,000 Mw and 2.9–3.6 PDI) with LDPEs (110,000–212,000 Mw , 5.1–17.2 PDI, and 12–16 short branches and 1.5–2.8 long branches per 1000 backbone carbons), and found that the HDPE/LDPE blends in the melt state form a homogeneous liquid mixture throughout the whole composition. In addition to the SANS technique, other analytical methods were employed to investigate the miscibility of HDPE/LDPE blends. Lee et al. (43) investigated the miscibility of HDPE/LDPE blends using DMTA: HDPE (121,000 Mw and 16.8 PDI) and LDPE (73,000–98,000 Mw and 8.7–9.2 PDI; 32–34 branches per 1000 backbone carbons). They found that HDPE/LDPE blends are miscible in the amorphous phase. Lee and Denn (44) conducted rheological measurements on the blends of HDPE (40,000 Mw and 3.1 PDI) and LDPE (388,000 Mw and 10 PDI) at 160 C; no detailed information on the branches in the LDPE was given. These results suggest that the HDPE/LDPE blends are partially miscible in the melt, depending on the composition. Barham et al. (50) studied the miscibility of HDPE/LDPE blends using DSC and TEM: HDPE of 98,000 Mw and LDPE of 208,100 Mw (16 short and 10 long branches per 1000 backbone carbons). They observed that the HDPE/LDPE blends are partially miscible in the melt depending on the composition. Hill et al. (51–54) conducted DSC, rheology, and TEM measurements on blends of HDPEs (2550–2,000,000 Mw and 1.4–12.0 PDI) and LDPE polymers (90,000–208,100 Mw and 8.0–8.5 PDI; 10–16 short and 1–16 long branches per 1000 backbone carbons), and constructed phase diagrams, indicating that the blends are partially miscible

Chapter 4 Miscibility and Crystallization Behavior in Binary Polyethylene Blends

89

depending on the composition, molecular weight, and branch content; phase separation took place in the LDPE-rich region and the size of the phase separation region in the phase diagram decreased with decreasing molecular weight of HDPE. As pointed out in the earlier section for HDPE/LLDPE blends, the validity of the techniques of Hill et al. is still controversial because the authors assumed that quenching a sample from the melt state to the solid state does not change the morphology of the blends in the melt. In other work in this field, Fan et al. (40) investigated the miscibility of HDPE/LDPE blends using molecular dynamics simulations and suggested that the critical branch content of LDPE required to induce immiscibility with HDPE is 25 branches per 1000 backbone carbons. As discussed above, there is still no consensus on the miscibility of the HDPE/ LDPE blends. Therefore, this subject needs more detailed investigation.

4.2.3 Blends of Short and Long Branched Polyethylenes Kyu et al. (2) and Ree et al. (3,5) studied blends of LLDPE-B (114,000 Mw , 4.50 PDI, and 18 ethyl branches per 1000 backbone carbons) and LDPE (286,000 Mw , 15.98 PDI, and 26 short and 1.6 long branches per 1000 backbone carbons) using in situ small-angle light scattering (SALS) and DSC and found that these blends are miscible across the whole composition range. Similar miscibility results were reported for LDPE blends with LLDPE-B and LLDPE-O polymers by other research groups. The DSC and DMTA analysis of Lee et al. (43) confirmed that the blends of LLDPEs (LLDPE-B: 89,300 Mw , 3.8 PDI, and 15–16 branches per 1000 backbone carbons; LLDPE-O: 93,100 Mw , 3.6 PDI, and 15–16 branches per 1000 backbone carbons) and LDPEs (73,000–98,000 Mw and 8.7–9.2 PDI; 32–34 branches per 1000 backbone carbons) are miscible at all compositions. Chen et al. (55, 56) obtained miscible blends from LDPEs (89,000–474,000 Mw and 4.4–23.3 PDI) with LLDPEB (58,000 Mw , 2.65 PDI, and 31.5 ethyl branches per 1000 backbone carbons) and LLDPE-O polymers (96,700–474,000 Mw , 2.86–3.80 PDI, and 12.0–64.5 hexyl branches per 1000 backbone carbons). In these blends, miscibility was achieved for all compositions, regardless of the branch lengths and contents considered. The steady-state rheological, thermal, and mechanical properties of the binary blends consisting of LLDPE-H (690,000 Mw , 1.86 PDI, and 18.6 butyl branches per 1000 backbone carbons) and LDPE (660,000 Mw , 9.43 PDI, and 9 short and 5 long branches per 1000 backbone carbons) were investigated by Yamaguchi et al. (29, 30). According to the rheological measurements, LLDPE is miscible with LDPE in the melt state. Silveira and Choi (57) also studied the miscibility of blends of LDPE with six different LLDPE polymers using inverse gas chromatography; the molecular characteristics of the used PE polymers were not given. These blends were found to be miscible regardless of the molecular characteristics of the LLDPEs used. Lee and Denn (44), however, reported somewhat different results on LLDPE/ LDPE blends. They investigated blends of LDPE (388,440 Mw , 10.0 PDI, and 32 branches per 1000 backbone carbons) with LLDPE polymers (68,300–134,000 Mw and 3.2–3.8 PDI; 10–16 short branches per 1000 backbone carbons) at 160 C using

90

Polyolefin Blends

rheological analysis and found that the blends are partially miscible in the melt depending on the composition. Muller et al. (58) studied the miscibility of LDPE (110,000 Mw ) and LLDPE-O (130,000 Mw ) blends by means of DSC. These blend samples too were found to be partially miscible, depending on the composition. Using rheological analysis, Hameed and Hussein (33) also investigated the miscibility of blends of LDPE (100,000 Mw , 4.14 PDI, and 7.8 branches per 1000 backbone carbons), LLDPE-B (110,000 Mw , 1.78 PDI, and 20.4 ethyl branches per 1000 backbone carbons), and two LLDPE-H (67,000 Mw , 1.85 PDI, and 21.1 butyl branches per 1000 backbone carbons; 108,000 Mw , 1.83 PDI, and 18.0 butyl branches per 1000 backbone carbons) and found that the blends were partially miscible. Interestingly, the blends of LDPE with LLDPE-H (67,000 Mw ) were almost miscible throughout the whole composition range. In contrast, the blends of LDPE with LLDPE-H (108,000 Mw ) were partially miscible. The miscibility of the blend was nonsymmetric with respect to composition; the addition of a small amount of LLDPE-H (108,000 Mw ) to LDPE was more likely to cause immiscibility than the addition of a small amount of LDPE to the LLDPE-H polymer. However, increasing the branch length of LLDPE from ethyl to butyl showed no effect on the miscibility with LDPE. Hill et al. (59) examined the miscibility of LLDPE-O (40,000 Mw , 4.2 PDI, and 15 hexyl branches per 1000 backbone carbons) and LDPE (112,000 Mw , 12.0 PDI, and 15 short and 10 long branches per 1000 backbone carbons) using DSC and TEM and found that the blend produces a complex phase diagram. The realization of the phase diagram indicates that the blend is partially miscible, depending on the composition and temperature window. However, the phase diagram was constructed from the DSC and TEM data measured for the quenched samples. Thus, such phase diagrams still need to be verified by direct measurements. Contrary to the reports discussed above, Liu et al. (60) reported rheological results indicating that the LLDPE-B (45,700 Mw , 2.1 PDI, and 13.5 ethyl branches per 1000 backbone carbons) and LDPE (102,000 Mw , 6.8 PDI, and 27.5 branches per 1000 backbone carbons) blend is an immiscible system. As reviewed above, there is still no consensus on the miscibility of LLDPE/ LDPE blends, or the effects of short and long branches on the miscibility. Furthermore, the lack of detailed SANS analysis on LLDPE/LDPE blends is surprising, although the miscibility specifications of these blends are highly demanded, particularly in relation to their various applications in the polymer industry.

4.3 CRYSTALLIZATION BEHAVIORS 4.3.1 Blends of Linear and Short Branched Polyethylenes Datta and Birley (61) first reported on the crystallization behavior of HDPE/LLDPE blends using DSC and wide-angle X-ray diffraction (WAXD). In the DSC measurements, the blends revealed only one endothermic peak for both fast and slow cooling conditions, indicating that the blend components were cocrystallized. The X-ray

Chapter 4 Miscibility and Crystallization Behavior in Binary Polyethylene Blends

91

analysis confirmed that the unit cell dimensions of the crystallized HDPE/LLDPE blends are identical to those of the HDPE. Here, it is noted that details of the molecular characteristics of the LLDPE were not given. Similar cocrystallization was evident in the DSC measurements for blends of HDPEs (29,000 Mw and 4.08 PDI; 121,000 Mw and 16.8 PDI) with an LLDPE-O (93,100 Mw , 3.6 PDI, and 15 hexyl branches per 1000 backbone carbons) (43) and LLDPE-Bs (64,000 Mw , 4.57 PDI, and 12.5 ethyl branches per 1000 backbone carbons; 89,300 Mw , 3.8 PDI, and 16 ethyl branches per 1000 backbone carbons) (25, 43). The cocrystallization was also observed in HDPE blends with LLDPE-O having a high number of hexyl branches (e.g., 32) (62). Hu et al. (1) studied in detail HDPE/LLDPE blend systems using DSC, WAXD, small-angle X-ray scattering (SAXS), small-angle light scattering (SALS), and Raman longitudinal-acoustic-mode spectroscopy (LAM). In their study, the blend samples were prepared from HDPE (160,000 Mw and 7.1 PDI) and LLDPE-B (114,000 Mw , 4.50 PDI, and 18 ethyl branches per 1000 backbone carbons). Their DSC results are similar to those of Datta and Birley (61). The WAXD, SAXS, and LAM analyses of Hu et al. found that the branches are excluded from the polyethylene unit cell, resulting in thinner lamellae. This finding was confirmed by molecular dynamics simulations (63). Tashiro et al. (64, 65) investigated the crystallization behavior of blends of HDPEs (80,000–126,000 Mw and 5.3– 5.7 PDI) with LLDPE-B polymers (61,000–75,000 Mw , 2.0–3.1 PDI, and 17 and 41 ethyl branches per 1000 backbone carbons) using DSC and infrared spectroscopy. They found that even under slow cooling conditions, the LLDPE-B polymer containing 17 branches cocrystallizes with the HDPEs over the whole composition range. In contrast, the LLDPE-B polymer with 41 branches was found to partially cocrystallize with HDPEs, depending on the composition. Rana (66, 67) studied in detail the cocrystallization kinetics of a HDPE/LLDPE blend system using DSC and SALS; in that study, a HDPE and an LLDPE-O having 30 hexyl branches per 1000 backbone carbons were used, but their molecular weight and polydispersity were not documented. It was found that, on the basis of Avrami exponent, both the HDPE and LLDPE-O polymers form cocrystallites. The incorporation of LLDPE segments into the HDPE crystallites progressively dilutes the properties of HDPE in the blend. The variation in crystallization half-time (t1/2) with increasing LLDPE-O content shows three distinct stages: t1=2 increases slowly in the region of 0–40 wt% LLDPE-O (HDPE-rich blend), remains constant in the 30–70 wt% LLDPE-O containing region (middle region of blend composition), and increases sharply with further increases of the LLDPE-O content. These variations of t1=2 qualitatively account for the change in crystallinity percentage, the Avrami exponent, and crystallite-size distribution. These observations were further supported by SALS measurements. The effect of branch distribution of LLDPE on the crystallization of HDPE/ LLDPE blends has also been investigated. The DSC and scanning electron microscopy (SEM) studies of Lee et al. (68) found that a Ziegler–Natta-catalyzed LLDPEB having a relatively heterogeneous branch distribution (179,000 Mw , 4.2 PDI, and 16 ethyl branches per 1000 backbone carbons) cocrystallizes with HDPE (147,000 Mw and 3.7 PDI) in a wide range of blend compositions, and at the same time,

92

Polyolefin Blends

separate crystallization occurs to a small extent. A metallocene-catalyzed LLDPE-B polymer with a relatively homogeneous branch distribution (170,000 Mw , 2.8 PDI, and 17.5 ethyl branches per 1000 backbone carbons) was also found to cocrystallize with HDPE in a wide range of blend compositions. On the contrary, the metallocenecatalyzed LLDPE-B polymer undergoes separate crystallization to a relatively larger extent compared to that of the Ziegler–Natta-catalyzed LLDPE-B (68). From the results discussed above, it is concluded that the crystallization behavior of HDPE/LLDPE blends depends on the number, length, and distribution of branches in the LLDPE component. In general, it was found that increases in the number and degree of branch distribution in the LLDPE component reduce the tendency of cocrystallization in the LLDPE blend with HDPE.

4.3.2 Blends of Linear and Long Branched Polyethylenes Lee et al. (43) studied the crystallization behavior in blends of HDPE (121,000 Mw and 16.8 PDI) and LDPE (73,000–98,000 Mw and 8.7–9.2 PDI; 32–34 branches per 1000 backbone carbons) and reported that the HDPE and LDPE components cocrystallize in the blend. In contrast, in DSC and WAXD studies on HDPE/LDPE blends cooled both rapidly and slowly, Datta and Birley (61) found that both the rapidly and slowly cooled blend samples exhibited two distinct endothermic peaks in the DSC measurements, indicating that the HDPE and LDPE components in the blends crystallize separately. However, no data were given on the molecular characteristics of LDPE. Similar separate crystallization behavior was also reported for HDPE/LDPE blends by Ree (5) and Song et al. (69, 70). In these studies, HDPE (160,000 Mw and 7.1 PDI) and LDPE (286,000 Mw , 15.98 PDI, and 26 short and 1.6 long branches per 1000 backbone carbons) were employed. In the DSC measurements, the blends revealed two peaks characteristic of both exotherms and endotherms, and indicative of the formation of separate crystals. This separate crystallization phenomenon was also found at the crystallite and lamellar levels using WAXD, SAXS, and LAM. The SAXS spectra of the blends were interpreted with respect to the segregation phenomena of interfibrillar scales during crystallization of the blend components. The SALS studies were performed in order to consider the morphology and crystallization behavior of the blends at the super structural level. Light scattering results demonstrated that the blend samples were predominantly volume-filled as a result of the primary crystallization of the HDPE component, whereby the LDPE component crystallized as a secondary process within the spherulites of the high density component. In addition, the SALS intensities from the blend samples of various compositions were interpreted on the basis of the spherulite radius, its internal disorder, and the volume fraction crystallinity. The crystallization process of the LDPE component in the blends was investigated by means of DSC. In this study, it was evident that the degree of crystallinity changes linearly with logarithmic time during isothermal crystallization of the LDPE component.

Chapter 4 Miscibility and Crystallization Behavior in Binary Polyethylene Blends

93

From the results discussed above, it is concluded that the crystallization behavior of HDPE/LDPE blends depends on the number, length, and distribution of branches in the LDPE component. In general, most HDPE/LDPE blends reveal a strong tendency to undergo separate crystallization. Cocrystallization is rarely observed in HDPE/LDPE blends.

4.3.3 Blends of Short and Long Branched Polyethylenes Kyu et al. (2) and Ree et al. (3, 5) studied the crystallization behavior of blends of LLDPE-B (114,000 Mw , 4.50 PDI, and 18 ethyl branches per 1000 backbone carbons) and LDPE (286,000 Mw , 15.98 PDI, and 26 short and 1.6 long branches per 1000 backbone carbons) using DSC, WAXD, SAXS, SALS, and LAM. They found that the LLDPE-B and LDPE components crystallize separately. The SAXS and SALS results indicate that when the LLDPE/LDPE blends are cooled from the molten state, the LLDPE-B crystallizes first and forms volume-filling spherulites, which is closely followed by crystallization of the LDPE within the spherulites formed by LLDPE-B. Similar separate crystallization behavior was reported by Lee et al. (43) for blends of LLDPE-B (89,300 Mw , 3.8 PDI, and 15–16 branches per 1000 backbone carbons) with LDPEs (73,000–98,000 Mw and 8.7–9.2 PDI; 32–34 branches per 1000 backbone carbons). Separate crystallization was also reported for blends of LDPEs with LLDPEs having butyl and hexyl branches. Using DSC analysis, Drummond et al. (71) observed that separate crystallization of the individual blend components takes place in the blends of LLDPE-H (146,000 Mw , 3.1 PDI, and 19.9 butyl branches per 1000 backbone carbons) and LDPE (78,700 Mw , 31.0 PDI, and 20.1 branches per 1000 backbone carbons). The DSC studies of Lee et al. (43) revealed that for blends of LLDPE-O (93,100 Mw , 3.6 PDI, and 15–16 branches per 1000 backbone carbons) with LDPEs (73,000–98,000 Mw and 8.7–9.2 PDI; 32–34 branches per 1000 backbone carbons), the blend components also undergo separate crystallization. Hill et al. (69) reported similar separate crystallization behavior in blends of LLDPE-O (40,000 Mw , 4.2 PDI, and 15 hexyl branches per 1000 backbone carbons) and LDPE (112,000 Mw , 12.0 PDI, and 15 short and 10 long branches per 1000 backbone carbons). As reviewed above, LLDPE/LDPE blends tend to undergo separate crystallization of the individual components, depending on the number, length, and distribution of branches in the two blend components.

4.4 CONCLUSIONS The miscibility and crystallization behavior of the three binary PE blends, HDPE/ LLDPE, HDPE/LDPE, and LLDPE/LDPE, were reviewed. In general, differences in the number, length, and distribution of branches in the PE blend components are the major factors governing their miscibility and crystallization phenomenon. In particular, the content and distribution of branches significantly affect both the miscibility and crystallization. Moreover, the presence of a few long chain branches, as well as

94

Polyolefin Blends

the observed differences in the molecular weight and molecular weight distribution, all affect the miscibility and crystallization. It is noted, however, that studies on both the miscibility and crystallization behavior of PE blends are quite limited due to experimental difficulties, in particular those caused by the chemical repeat units of linear and branched PEs having the same molecular nature. Therefore, the development of more quantitative analytical techniques is still needed in order to fully understand the miscibility and crystallization behavior of PE blends.

NOMENCLATURE DSC DMTA HDPE HDPE-d LDPE LLDPE LLDPE-B LLDPE-B (hydrogenated PB) LLDPE-B-d LLDPE-H LLDPE-O PB PE PDI SANS TEM UHMWPE Mw x

Differential scanning calorimetry Dynamic mechanical thermal analysis High density linear polyethylene Deuterated high density linear polyethylene Low density branched polyethylene Linear low density polyethylene Linear low density polyethylene prepared by the copolymerization of ethylene and 1-butene Linear low density polyethylene prepared by the hydrogenation of polybutadiene Deuterated LLDPE-B Linear low density polyethylene prepared by the copolymerization of ethylene and 1-hexene Linear low density polyethylene prepared by the copolymerization of ethylene and 1-octene Polybutadiene Polyethylene Polydispersity index Small-angle neutron scattering Transmission electron microscopy Ultrahigh molecular weight polyethylene Weight-average molecular weight Flory–Huggins interaction parameter

REFERENCES 1. S. R. Hu, T. Kyu, and R. S. Stein, J. Polym. Sci. B: Polym. Phys., 25, 71 (1987). 2. T. Kyu, S. R. Hu, and R. S. Stein, J. Polym. Sci. B: Polym. Phys., 25, 89 (1987). 3. M. Ree, T. Kyu, and R. S. Stein, J. Polym. Sci. B: Polym. Phys., 25, 105 (1987). 4. P. Vadhar and T. Kyu, Polym. Eng. Sci., 27, 202 (1987). 5. M. Ree, Ph.D. Thesis, University of Massachusetts, Amherst, 1987. 6. G. M. Benedikt and B. L. Goodall (eds.), Metallocene-Catalyzed Polymers: Materials, Properties, Processing and Markets, William Andrew, New York, 1998.

Chapter 4 Miscibility and Crystallization Behavior in Binary Polyethylene Blends

95

7. F. Garbassi, L. Gila, and A. Proto, Polym. News, 19, 367 (1994). 8. K. M. Drummond, J. L. Hopewell, and R. A. Shanks, J. Appl. Polym. Sci., 78, 1009 (2000). 9. C. S. Speed, Plast. Eng., 39, 4 (1982). 10. B. Schlund and L. A. Utracki, J. Polym. Eng. Sci., 27, 359 (1987). 11. A. J. Muller, V. Balsamo, and C. M. Rosales, Polym. Networks Blends, 2, 215 (1992). 12. C. T. Lue, T. H. Kwalk, D. Li, C. R. Davey, and D. Y. Chiu, Annu. Technol. Conf. Soc. Plast. Eng., 62(2), 2140 (2004). 13. A. A. Montaga, Chemitech, 25, 44 (1995). 14. B. Schlund and L. A. Utracki, Polym. Eng. Sci., 27, 359 (1987). 15. A. Rudin, H. P. Schreiber, and D. Duchesne, Polym. Plast. Technol. Eng., 29, 199 (1990). 16. F. Chen, R. Shanks, and G. Amarasinghe, J. Appl. Polym. Sci., 81, 2227 (2001). 17. I. B. Kazatchkov, N. Bohnet, S. K. Goyal, and S.G. Hatzikiriakos, Polym. Eng. Sci., 39, 804 (1999). 18. L. Wild, R. Ranganath, and D.C. Knobeloch, Polym. Eng. Sci., 16, 811 (1976). 19. S. K. Goyal, J. Auger, E. Karbashewski, and R. Saetre, Annu. Technol. Conf. Proc., 56, 1881 (1998). 20. S. G. Hatzikiriakos, Polym. Eng. Sci., 40, 2279 (2000). 21. B. H. Bersted, J. D. Slee, and C. A. Richter, J. Appl. Polym. Sci., 26, 1001 (1996). 22. B. H. Bersted, J. Appl. Polym. Sci., 30, 3751 (1985). 23. S. K. Goyal, J. Auger, E. Karbashewski, and R. Saetre, Annu. Technol. Conf. Proc., 56, 1881 (1998). 24. L. Zhao and P. Choi, Materi. Manufact. Process., 21, 135 (2006). 25. B. Neway and U. W. Gedde, J. Appl. Polym. Sci., 94, 1730 (2004). 26. C. F. Iwu and O. M. Egbuhuzor, Annu. Technol. Conf. Soc. Plast. Eng., 62(1), 1163 (2004). 27. K. Cho, B. H. Lee, K.-M. Hwang, H. Lee, and S. Choe, Polym. Eng. Sci., 38, 1969 (1998). 28. J.-H. Oh, W. Dong, and Y. Gu, Reinf. Plast. Compos., 18, 662 (1999). 29. M. Yamaguchi and S. Abe, J. Appl. Polym. Sci., 74, 3153 (1999). 30. M. Yamaguchi and S. Abe, J. Appl. Polym. Sci., 74, 3160 (1999). 31. F. Chen, R. Shanks, and G. Amarasinghe, J. Appl. Polym. Sci., 81, 2227 (2001). 32. I. A. Hussein and M. C. Williams, Polym. Eng. Sci., 41, 696 (2001). 33. T. Hameed and I. A. Hussein, Polymer, 43, 6911 (2002). 34. B. Crist and M. J. Hill, J. Polym. Sci. B: Polym. Phys., 35, 2329 (1997). 35. J. C. Nicholson, T. M. Finerman, and B. Crist, Polymer, 31, 2287 (1990). 36. R. G. Alamo, W. W. Graessley, R. Krishnamoorti, D. J. Lohse, J. D. Londono, L. Mandelkern, F. C. Stehling, and G. D. Wignall, Macromolecules, 30, 561 (1997). 37. M. Agamalian, R.G. Alamo, M. H. Kim, J. D. Londono, L. Mandelkern, and G. D. Wignall, Macromolecules, 32, 3093 (1999). 38. Z. Fan, Ph.D. Thesis, University of Alberta, 2001. 39. P. Choi, Polymer, 41, 8741 (2000). 40. Z. Fan, M. C. Williams, and P. Choi, Polymer, 43, 1479 (2001). 41. G. H. Fredrickson, A. J. Liu, and F. S. Bates, Macromolecules, 27, 2503 (1994). 42. L. Zhao and P. Choi, J. Appl. Polym. Sci., 91, 1927 (2004). 43. H. Lee, K. Cho, T. Ahn, S. Choe, I.-J. Kim, and I. Park, J. Polym. Sci. B: Polym. Phys., 35, 1633 (1997). 44. H. Lee and M. M. Denn, Polym. Eng. Sci., 40, 1132 (2000). 45. H. Y. Zhao, Z. L. Lei, and B. T. Huang, Polymer J., 30, 149 (1998). 46. A. Tabtiang, B. Parchana, R. A. Venables, and T. Inoue, J. Polym. Sci. B: Polym. Phys., 39, 380 (2001). 47. M. J. Hill, R. L. Morgan, and P. J. Barham, Polymer, 38, 3003 (1997). 48. M. J. Hill, P. J. Barham, and J. van Ruiten, Polymer, 34, 2975 (1993).

96

Polyolefin Blends

49. R. G. Alamo, J. D. Londono, L. Mandelkern, F. C. Stehling, and G. D. Wignall, Macromolecules, 27, 411 (1994). 50. P. J. Barham, M. H. Hill, A. Keller, and C. C. A. Rosney, J. Mater. Sci. Lett., 7, 1271 (1988). 51. M. J. Hill, P. J. Barham, A. Keller, and C. C. A. Rosney, Polymer, 32, 1384 (1991). 52. M. J. Hill and P. J. Barham, Polymer, 33, 4099 (1992). 53. M. J. Hill, P. J. Barham, and A. Keller, Polymer, 33, 2530 (1992). 54. M. J. Hill, Polymer, 35, 1991 (1994). 55. F. Chen, R. Shanks, G. Amarasinghe, J. Appl. Polym. Sci., 81, 2227 (2001). 56. F. Chen, R. Shanks, G. Amarasinghe, J. Appl. Polym. Sci., 95, 1549 (2005). 57. M. D. L. V. Silveira and P. Choi, Annu. Technol. Conf. Soc. Plast. Eng., 58(2), 2454 (2000). 58. A. J. Muller, V. Balsamo, F. Da Silva, C. M. Rosales, and A. E. Saez, Polym. Eng. Sci., 34, 1455 (1994). 59. M. J. Hill and C.C. Puig, J. Appl. Polym. Sci., 65, 1921 (1997). 60. C. Liu, J. Wang, and J. He, Polymer, 43, 3811 (2002). 61. N. K. Datta and A. W. Birley, Plast. Rubber Proc. Appl., 2, 237 (1982). 62. S. J. Mahajan, B. L. Deopura, and Y. Wang, J. Appl. Polym. Sci., 60, 1517 (1996) 63. M. Doran and P. Choi, J. Chem. Phys., 115, 2827 (2001). 64. K. Tashiro, M. Izuchi, M. Kobayashi, and R. S. Stein, Macromolecules, 27, 1221 (1994). 65. K. Tashiro, R. S. Stein, and S. L. Hsu, Macromolecules, 25, 1801 (1992). 66. S. K. Rana, J. Appl. Polym. Sci., 61, 951 (1996). 67. S. K. Rana, J. Appl. Polym. Sci., 69, 2599 (1998). 68. S. Y. Lee, J. Y. Jho, and W. Huh, J. Ind. Eng. Chem., 4, 258 (1998). 69. H. H. Song, R. S. Stein, D.-Q. Wu, M. Ree, J. C. Philips, A. LeGrand, and B. Chu, Macronolecules, 21, 1180 (1988). 70. H. H. Song, D.-Q. Wu, B. Chu, M. Satkowski, M. Ree, R. S. Stein, and J. C. Philips, Macronolecules, 23, 2380 (1990). 71. K. M. Drummond, J. L. Hopewell, and R. A. Shanks, J. Appl. Polym. Sci., 78, 1009 (2000).

Chapter

5

Microscopically Viewed Structural Characteristics of Polyethylene Blends Between Deuterated and Hydrogenated Species: Cocrystallization and Phase Separation Kohji Tashiro1

5.1 INTRODUCTION Much attention has been paid in these decades to polymer blends from scientific and practical points of view. Polyethylene (PE) blends between the species with different degrees of branching contents have also been widely investigated (1–23). In these studies, it is very important to trace the aggregation state of each component separately and at a molecular level. However, it had been difficult to distinguish the behaviors of each component separately because these components have almost the same chemical structure of carbon and hydrogen atoms. One possible and useful idea is to use a deuterated PE sample as one component. The deuterated species can be separated from the hydrogenated species by utilizing the difference in the vibrational frequency (24–26) and the neutron scattering cross-sectional area (27–31). Therefore, the infrared and Raman spectroscopies

1 Department of Future Industry-Oriented Basic Science and Materials, Toyota Technological Institute, Tempaku, Nagoya 468-8511, Japan

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

97

98

Polyolefin Blends

and the neutron scattering method can be used effectively by utilizing these characteristic differences between the deuterated and hydrogenated species. In the utilization of these two isotopic components, we need to notice the fact that the deuterated species does not behave in perfectly the same manner as the hydrogenated species. In contrast to the case of low molecular weight compounds, the chemical potential of the fully deuterated PE species is appreciably different from that of the hydrogenated PE species because the aggregation of many CD2 groups results in the large isotope effect even if the difference in chemical potential between single CD2 and CH2 groups is negligibly small, that is, the polymeric effect (32). Therefore, the deuterated high density PE [DHDPE, –(CD2CD2)n–], for example, is predicted quite reasonably to behave in a different manner from the normal high density PE [HDPE, –(CH2CH2)n–], giving the detectable difference in the melting and crystallization temperatures. This situation is seriously related to the phenomena of phase separation and cocrystallization between DHDPE and HDPE species. In fact, the blend sample between DHDPE and HDPE was reported to show the phase segregation when the blend sample was slowly cooled from the melt (27–31). The phase segregation or the self-aggregation of D (or H) chains becomes serious in some problems. For example, the phase segregation problem of DHDPE/HDPE blend sample is serious in the research theme of chain-folding mechanism where the clarification of the trace of one isolated chain is important (30). We prepare the blend sample of H chains with very low concentration of D chains, where the D chains must not aggregate together but should be isolated from each other to avoid the optical interference between the D chains in the small-angle neutron scattering experiment. Different from the case of HDPE, the chemical potential of linear low density PE (LLDPE) can be modified by controlling the degree of side group branching and the miscibility between D and H chains may be enhanced. In fact, the blend of DHDPE and LLDPE was found to be almost perfectly miscible and can cocrystallize together in a common crystal lattice. In this chapter, we will describe the structure and crystallization behavior of PE blends between DHDPE and LLDPE with various degrees of ethyl branching on the basis of experimental data of infrared spectra, wide-angle (WAXD), and small-angle X-ray scatterings (SAXS), neutron scattering, thermal analysis, and so on so that the chain aggregation state can be figured out concretely at the molecular level (33–47).

5.2 COCRYSTALLIZATION AND PHASE SEPARATION OF PE BLENDS Table 5.1 shows the PE samples used in a series of our research. The blend samples were prepared between DHDPE and LLDPE with different degrees of ethyl branches. The PE components were mixed together in boiling p-xylene solutions at the concentration of about 2 wt% followed by rapid precipitation into methanol at room temperature. Figure 5.1a shows the DSC thermograms measured in the cooling process from the melt for a series of DHDPE/LLDPE(2) blend samples. The crystallization temperature is originally different by about 15 C between the pure D and H

Chapter 5 Microscopically Viewed Structural Characteristics of Polyethylene Blends

99

Table 5.1 Characterization of PE Samples. Mw DHDPE HDPE LLDPE(2) LLDPE(3)

80,000 126,000 75,000 61,000

Mn 14,000 24,000 37,000 20,000

Ethyl branching per 1000 carbon atoms 2–3 1 17 41

components. The blend sample shows a single exothermic peak. The peak temperature changes continuously in between the original two points depending on the D/H content. The infrared spectra, WAXD, and SAXS measured in the cooling process from the melt supported the DSC data (33–36). For example, Figure 5.1b shows the temperature dependence of infrared band intensity, in which the crystalline bands of

Figure 5.1 (a) DSC thermograms and (b) the temperature dependence of infrared absorbance estimated for the D and H infrared crystalline bands in the cooling process from the melt measured for a series of DHDPE/LLDPE(2) blend samples. The starting temperature of crystallization is slightly different between these two curves probably because of the difference in the cooling rate, the monitoring point of temperature, and so on. But essential behavior is the same.

100

Polyolefin Blends

the two components start to appear simultaneously near the temperature region corresponding to the standing-up point of the DSC curve shown in Fig. 5.1a. Besides, as will be mentioned in a later section, the splitting width of infrared bands, due to the correlation coupling between the neighboring chain stems, changes continuously depending on the D/H relative content. All these experimental data allow us to apply the idea of cocrystallization to the DHDPE/LLDPE(2) blend system even when the crystallization occurs at slow cooling rate. As already pointed out, the D and H chains in the DHDPE/HDPE blends crystallize separately when the cooling rate is slow, although the degree of phase segregation depends on the blend ratio. At this point, therefore, the discovery of perfectly cocrystallizable pair of two different PE components is quite important, which can be utilized for the structural study of a single D chain embedded in a matrix of H chains without any concern about the D/H segregation problem (46). Different from the DHDPE/LLDPE(2) case, the blend of DHDPE with LLDPE(3) with higher degree of ethyl branchings shows the phase separation phenomenon. The DSC thermogram consists of the two peaks corresponding to the meltings of individual components (33–36). In the crystallization from the melt, the infrared bands and the WAXD peaks characteristic of each component are observed to appear at the different temperatures corresponding to their crystallization. In this way, the cocrystallization and phase separation are dependent on the degree of branching of the LLDPE component. This difference in the crystallization behavior affects sensitively the morphology or the chain aggregation state. Table 5.2 shows the summary of the various structural parameters estimated for these two typical blend samples (37). For example, the spherulite size (as estimated from the small-angle light scattering data) and the long period between the neighboring lamellae in the spherulite (as estimated from the SAXS data) are different depending on the blend. As for both of the spherulite radius and the long period, the DHDPE/ LLDPE(2) blend shows the intermediate size between those of the original two components. But, for the DHDPE/LLDPE(3) blend, the spherulite radius is Table 5.2

Morphology of PE Samples.

Spherulite radius, mm ˚ Long period, A ˚ Unit cell, A a b ˚ Lattice size, A (110) (200) a

LLDPE(3)

DHDPE/ LLDPE(3)

8.6 187

2.4 195

8.7 277

7.48 4.96

7.46 4.96

7.63 5.03

7.45a 4.96a

184 134

225 176

117 100

218a 190a

DHDPE

LLDPE(2)

5.3 231

9.3 158

7.42 4.93 272 216

DHDPE/ LLDPE(2)

The apparent values deduced from the overlapped profiles between DHDPE and LLDPE(3) components.

Chapter 5 Microscopically Viewed Structural Characteristics of Polyethylene Blends (a) DHDPE/LLDPE(2) blend

101

(b) DHDPE/LLDPE(3) blend

D H

Figure 5.2 Stacked lamellar models of polyethylene blend samples.

remarkably larger than that of the original two components and the long period is also very long compared with the original ones. In the DHDPE/LLDPE(2) case, both the D and H chains coexist to form the common lamellae. In the DHDPE/LLDPE(3) case, on the contrary, the D and H species form their own lamellae individually, which are mixed together to form the stacked lamellar structure in longer period. In Fig. 5.2 are illustrated schematically the lamellar stacking modes in these samples.

5.3 AGGREGATION STRUCTURE OF CHAINS IN LAMELLA As pointed out above, the blend of DHDPE and LLDPE(2) components show almost perfect cocrystallization phenomenon. Combining the various experimental data, we can discuss the chain aggregation state in the crystal lattice for each component individually. Figure 5.3a and b shows the infrared spectra taken at room temperature for a series of DHDPE/LLDPE(2) blend samples. The infrared spectra of pure components, that is, DHDPE and LLDPE(2) show the doublet bands due to the correlation splitting. The vibrational modes, in which the two chains arrayed along the (110) direction in the orthorhombic unit cell vibrate with the phase difference of 0 and p, are optically active and give the bands at the different frequency positions. The band splitting width is dependent on the strength of interactions between the neighboring chain stems. If the D chain stems invade the (110) array of the H chain stems and disturb the interaction between the H chain stems (H H) as illustrated in Fig. 5.4, then the vibrational coupling between the neighboring H chain stems is cut and the band splitting does not occur anymore (24–26). In other words, depending on the spatial arrangement of D and H chain stems in the crystalline lamella, the band splitting width is changed. As shown in Fig. 5.3c and d such a remarkable change of band splitting width cannot be seen in the case of DHDPE/LLDPE(3) blend samples. The systematic behavior of infrared spectra observed for DHDPE/LLDPE(2) blend samples can be interpreted reasonably as follows. Using the simply coupled

102

Polyolefin Blends

Figure 5.3 Infrared spectra measured at room temperature for a series of (a and b) DHDPE/ LLDPE(2) blend samples and (c and d) DHDPE/LLDPE(3) blend samples. Right: the CD2 rocking bands. Left: the CH2 rocking bands.

oscillator model, we can derive the simple equation concerning the band splitting width (46, 48, 49). As shown in Fig. 5.4c, an array of H chain stems along the (110) direction is assumed to be cut by an invasion of D stems. The band splitting width Dn is given by the following equation. Dn ¼ Dn0 cos½p=ðN þ 1Þ

ð5:1Þ

where Dn0 is a splitting width of the infinitely repeated array of oscillators (H chain stems) and N is the effective number of stems in the (110) direction. If we assume that the statistically random arrangement of D and H chain stems and the distribution are

Chapter 5 Microscopically Viewed Structural Characteristics of Polyethylene Blends

103

Figure 5.4 Chain-packing modes of orthorhombic polyethylene crystal. (a) The crystal lattice consists of all the same species. (b) A single chain stem (CH2) is surrounded by CD2 stems. (c) An illustration of CH2 and CD2 stem array along the (110) direction. (From Reference 46 with permission from the Society of Polymer Science, Japan.)

not affected by the states of the neighboring chains, then the probability to have a sequence of D(H)pD is expressed by f ðpÞ ¼ ð1 XÞ2 X p1

ð5:2Þ

where X is the fraction of the H stems and the probabilities of pairs of D–H (H–D) and H–H sequences are given by 1 X and X, respectively. Then the averaged size hNi of the D(H)pD clusters is calculated as hNi ¼ Spf ðpÞ=Sf ðpÞ ¼ 1=ð1 XÞ

ð5:3Þ

where the summation is over p ¼ 1 to infinity. By substituting the hNi into Equation 5.1, we can get the band splitting width Dn as Dn ¼ Dn0 cos½pð1 XÞ=ð2 XÞ

ð5:4Þ

In a similar way, we can predict the fraction of singlet H band component for the sequence of DHD. In Equation 5.2, f ðp ¼ 1Þ is the probability of getting a single band and f ðp > 1Þ is the probability of getting doublet bands. Therefore, the singlet

104

Polyolefin Blends

Figure 5.5 Dependence of (a) the splitting width and (b) the single band component fraction estimated for the CH2 rocking bands on the H chain fraction in the DHDPE/LLDPE(2) blends and in the blend of n-C30H62/n-C30D62. (From Reference 46 with permission from the Society of Polymer Science, Japan.)

fraction or f ðp ¼ 1Þ=½ f ðp ¼ 1Þ þ f ðp > 1Þ is given as 1 X by assuming the equal molar extinction coefficients for single and doublet bands. As shown in Fig. 5.5, where the data on the D/H blends of n-alkane (n-C30H62 and n-C30D62) are also given, these equations are found to satisfy the observed infrared data quite well. The wide-angle neutron scattering profiles measured for a series of DHDPE/ LLDPE(2) blend samples can be reproduced also reasonably by assuming the random distribution of the D and H chain stems in the lamella. In Fig. 5.6 are compared the observed and calculated wide-angle diffraction profiles measured for DHDPE/LLDPE(2) 50/50 blend sample. The neutron scattering profiles were calculated using a software Cerius2 for the model in which 20 D and 20 H chains were randomly packed in the crystal cell and each chain consisted of 20 methylene units. As seen in Fig. 5.6, the relatively intense low angle scattering is characteristic of the D and H blend lattice. In the case of DHDPE/LLDPE (3) blend, this low angle scattering becomes intense when the sample is melted, but it is quite weak when the sample is solidified because of the segregation of two components.

Chapter 5 Microscopically Viewed Structural Characteristics of Polyethylene Blends

105

Figure 5.6 Wide-angle neutron scattering profiles measured at the various temperatures for (c) DHDPE/LLDPE(2) and (d) DHDPE/LLDPE(3) blend samples in comparison with the profiles of (a) the amorphous state and (b) the crystalline state calculated for the pure DHDPE and the blend of D and H chain components. In (c), the low angle broad scattering is detected at any temperature due to the homogeneous mixing of D and H species in the same crystallite state as well as in the melt state. In (d), the low angle scattering can be seen only in the molten state, while it becomes lower when the sample crystallizes into the separated phases of D and H chains. These observations are consistent with the simulated results of (a) and (b).

From all these data analyses, we can definitely say that the D and H chain stems are distributed statistically randomly in the crystalline lamellae of the D/H cocrystallized blend. This conclusion is quite important in relation with the chain-folding problem, a controversial research theme that had been discussed for a long time (30). The random distribution of the D and H chain stems naturally supports the idea that the D and H chains reenter randomly into and out of the crystalline lamellae as shown in Fig. 5.7. The regular adjacent reentry model is impossible to apply at all as for as the melt-crystallized sample is concerned.

5.4 CRYSTALLIZATION BEHAVIOR OF D/H BLEND SAMPLES In order to investigate the cocrystallization and phase segregation behaviors of the above-mentioned PE blend samples, we performed the experiments for both the isothermal and nonisothermal crystallizations. In the nonisothermal crystallization, the temperature is changed gradually and the WAXD, SAXS, or infrared spectra are collected as a function of temperature. In the isothermal crystallization, the

106

Polyolefin Blends

Figure 5.7 Illustration of the chain-folding models. (a) A regular adjacent reentry model, (b) a random reentry model, and (c) a cluster model consisting of the mixed structures of (a) and (b). (From Reference 46 with permission from the Society of Polymer Science, Japan).

temperature is changed rapidly from the molten state to the crystallization temperature and is kept constant for a long time, during which the WAXD, SAXS, or infrared spectra are measured as a function of time.

5.4.1 Crystallization in the Cooling Process from the Melt Figure 5.8 shows the infrared spectral change of DHDPE/LLDPE(2) blend of 75/25 wt % measured during the nonisothermal crystallization or in the cooling process from the melt (36). Generally, the infrared CH2 and CD2 bending bands show approximately the singlet and the doublet profiles, respectively. The band profiles were separated into the singlet and doublet components by a deconvolution method. As shown in Fig. 5.9a, when the crystallization started, the singlet component of the H bands appeared at first, just when the D bands showed both the doublet and singlet components at almost the same ratio. That is to say, the H chain stems in the crystal lattice are isolated from each other by being surrounded by D chain stems at the early stage of cocrystallization. As the temperature deceased furthermore, the H and D

Chapter 5 Microscopically Viewed Structural Characteristics of Polyethylene Blends

107

Figure 5.8 Temperature dependence of infrared spectra measured for DHDPE/LLDPE(2) 75/25 blend sample during the cooling process from the melt. In these profiles, the contribution of the amorphous band was subtracted. (From Reference 36 with permission from the American Chemical Society.)

chain stems gathered randomly around the initially created nuclei. In the case of DHDPE/LLDPE(3) blend, on the contrary, the situation was a little different. As mentioned above, the D and H species crystallize separately in the different lamellae. But, more strictly speaking, some parts of the H species are trapped in the D crystallites. As shown in Fig. 5.9b, at the early stage of crystallization, the singlet band of H chain stems was detected in the crystallization temperature region of the D species. When the temperature was lowered down to 90 C, where the main part of LLDPE(3) started to crystallize, both the doublet and singlet band components were observed to increase in intensity. That is to say, some portion of H chains was induced to cocrystallize with the surrounding D chains. Below 90 C, the H chain stems crystallized separately from the growing CD2 crystals, as illustrated in Fig. 5.9b. This observation may be due to more or less inhomogeneous distribution of side group branchings along the skeletal chain; the parts with lower degree of branching can be cocrystallized with the D chains likely as the case of DHDPE/LLDPE(2) blend, but the parts with higher branchings are separated from the D chains.

108

Polyolefin Blends

Figure 5.9 Schematic illustration of aggregation behavior of D (o) and H (.) chain stems in the crystallite during the crystallization process from the melt. (a) DHDPE/LLDPE(2) and (b) DHDPE/ LLDPE(3).

5.4.2 Isothermal Crystallization Process Before the description of crystallization behavior of DHDPE/LLDPE(2) blend system, it is valuable to review the structural evolution process of LLDPE(2) itself during the isothermal crystallization process from the melt (44–46). The timedependent measurement of infrared spectra was performed during the temperature jump from 160 to 105 C, for example, where the degree of supercooling DT, evaluated as a difference between the actually-set crystallization temperature and the starting point of crystallization, was about 4 C. The data were interpreted on the basis of the information about the infrared bands characteristic of such various types of molecular conformation as all-trans, trans-gauche, and so on. After the temperature jump, the infrared bands (1368 cm1, etc.) characteristic of the so-called conformationally disordered trans-form consisting of long trans-segment with invasion of some gauche bonds were found to generate as shown in Fig. 5.10, where the time ¼ 0 s was defined as the time at which the temperature reached just the predetermined crystallization temperature. After that, these infrared bands started to decrease in intensity and the infrared bands intrinsic to the orthorhombic crystal form or the long trans-zigzag bands started to increase. Therefore, it is considered that the conformationally disordered but not perfectly random chain stems are generated at first from the melt. After that, they are regularized to the longer

Chapter 5 Microscopically Viewed Structural Characteristics of Polyethylene Blends

109

Figure 5.10 Time dependence of the integrated intensity estimated for the infrared bands characteristic of the partially disordered trans-form (1368 cm1) and the regular trans-zigzag form (728 cm1) in the isothermal crystallization process from the melt. The time 0 s was defined as the time when the temperature reached just the predetermined crystallization point (Tc ). (From Reference 44 with permission from the Society of Polymer Science, Japan.)

trans-zigzag stems in the orthorhombic crystal. As shown in Fig. 5.11, we performed the time-resolved WAXD measurement during the isothermal crystallization by using a synchrotron radiation source. In the early crystallization process, the amorphous halo peak was observed and shifted to higher scattering angle with the passage of time (47). This indicates the closer packing of the random coils in the melt. This is almost detected at the same time when the disordered trans-zigzag bands were observed. It might be speculated that the partially regularized trans-zigzag segments are created in the amorphous coils and these segments gather gradually to form a

Figure 5.11 Time-dependent wide-angle X-ray scattering pattern taken for LLDPE(2) in the isothermal crystallization process.

110

Polyolefin Blends

Figure 5.12 Time dependence of small-angle X-ray scattering patterns measured for (a) LLDPE(2) and (b) DHDPE in the isothermal crystallization process from the melt. It should be noticed that the intensity exchange (at (b)) can be observed between the firstly appeared peak and the secondly appeared peak in the crystallization process.

closer packing of chains in the amorphous region. Although the X-ray diffraction peaks characteristic of the hexagonal lattice was not detected clearly before the appearance of the peaks of the orthorhombic cell, there should exist more or less ordered structure consisting of partially ordered trans-segments, which might be similar to the nematic liquid–crystalline state. This structure changes finally to the orthorhombic lattice. The time-resolved SAXS measurements were made in parallel (45). As shown in Fig. 5.12a, immediately after the temperature jump (DT ¼ 4 C), the invariant Q, which is approximately proportional to the total crystallinity, could be detected at almost the same time when the partially disordered trans-bands (Fig. 5.13a) appeared. After that the SAXS peak corresponding to the stacked ˚ long period was detected. The lamellar thickness lamellar structure with about 800 A also increased gradually. At around 150 s after the jump, another SAXS peak ˚ long period was detected as a shoulder and increased corresponding to the 400 A in intensity, while the original SAXS peak decreased. This SAXS profile change was detected more clearly in the isothermal crystallization experiment of DHDPE as shown in Fig. 5.12b. The similar phenomenon was observed for the isothermal crystallization of polyoxymethylene (POM) (50–52). The amorphous region sandwiched between the neighboring lamellae crystallizes to a new lamella, resulting in ˚ long period. This type of stacked lamellar the observation of the structure of 400 A structure formation was proposed for the LLDPE/LDPE blends (2–6, 11, 12). The intensity exchange between these two SAXS peaks in Fig. 5.12 is not perfect but occurs only partly, indicating that the secondary crystallization occurs only locally at

Chapter 5 Microscopically Viewed Structural Characteristics of Polyethylene Blends

111

Figure 5.13 Comparison of vibrational spectroscopic data with SAXS data collected for (a) LLDPE (2) and (b) DHDPE samples in the isothermal crystallization process from the melt. Q: invariant, hdi: lamellar thickness, Li and I(Li): the long period and the corresponding SAXS peak intensity of lamellar ˚ period and L2 for 400 A ˚ period. stacking structure: L1 for 800 A

the various positions of original lamellae, as reported for the isothermal crystallization of POM (50–52). In parallel to these studies, we evaluated the radius of gyration Rg for D and H chains in the DHDPE/LLDPE(2) blend samples by measuring the small-angle neutron scattering at various temperatures (46). As shown in Fig. 5.14, Rg takes essentially the same value in both the melt and the solid state. In other words, the spatial size of a chain is not significantly affected even when drastic environmental changes occur in the crystallization process from the melt to the stacked lamellar structure. By combining the data shown in Fig. 5.13 (for both LLDPE(2) and DHDPE) and the information on Rg , we can describe the structural change of a chain during the crystallization process as illustrated in Fig. 5.15. The D and H chains are mixed homogeneously in the molten state. Once the crystallization starts to occur, the conformational ordering starts to occur and the partially disordered trans-segments are generated in the random coil, the details of which will be described in a later section. These partially disordered trans-chain segments are gathered together to form the lamella with Rg remaining almost unchanged. The D and H chains are randomly mixed and aggregate together with the random folding mode. As a result, the D and H chain stems are randomly mixed in the crystalline lattice. The lamellae

112

Polyolefin Blends

Figure 5.14 Temperature dependence of the radius of gyration (Rg ) estimated for the D (solid symbol) and H (open symbol) chains of the DHDPE/LLDPE(2) blend samples with the different D/H ratios. (From Reference 46 with permission from the Society of Polymer Science, Japan.)

Figure 5.15 Model of structural evolution process in the isothermal crystallization of polyethylene chains. The random chain segments change to the locally regularized trans-form. The stacked lamellar ˚ . The structure changes further to the stacked lamellar structure is formed with the long period 800 A ˚ period by inserting the new lamella into the original lamellae. It is noted that the structure of 400 A radius of gyration is kept unchanged during this process. (From Reference 46 with permission from the Society of Polymer Science, Japan.)

Chapter 5 Microscopically Viewed Structural Characteristics of Polyethylene Blends

113

˚ long period. With the passage of time, the lamellar are stacked with about 800 A ˚ long period changes to the final state of 400 A ˚ long period stacking structure of 800 A through the secondary crystallization process in the amorphous region sandwiched between the originally formed lamellae. The Rg is kept almost the same even after an occurrence of the secondary crystallization as seen from the SANS data (Fig. 5.14). On the basis of the accumulated knowledge on the isothermal crystallization behavior of LLDPE (and DHDPE) sample, we investigated the crystallization of the DHDPE/LLDPE(2) blend sample through the time-resolved infrared and SAXS measurements. Similarly to LLDPE(2), the infrared bands characteristic of partially disordered trans-segments appeared at first, suggesting the formation of closer packing structure of chain in the amorphous region as discussed in the previous paragraphs. Then the infrared bands of long trans-segments started to appear to show the regularization as the orthorhombic lattice. It was noticed that the doublet bands of the DHDPE component and the singlet of LLDPE(2) component were observed to appear at first and simultaneously. After that, both the doublet and singlet bands increased in intensity for both the D and H species. This observation is consistent with that in the nonisothermal crystallization of DHDPE/LLDPE(3) blend as described in the previous section. The H chain stems are surrounded by the D stems in a small nucleus at the early stage of crystallization, then the D and H chain stems gather together randomly around these small nuclei to produce larger crystallites (Fig. 5.9).

5.4.3 Blending Effect on Crystallization Rate In the preceding section, the structural evolution process was described at the molecular level as well as the lamellar level. In the experiment about the structural evolution in the crystallization process, we noticed the difference in crystallization rate between various kinds of PE samples. Figure 5.16 shows the increase in the crystallization sensitive infrared band intensity estimated for DHDPE, HDPE,

Figure 5.16 Time dependence of infrared band intensity measured for various PE samples during the isothermal crystallization from the melt. The degree of supercooling was about 3 C. (From Reference 38 with permission from the American Chemical Society.)

114

Polyolefin Blends

Figure 5.17 Comparison of crystallization rate between the pure components and their blend sample in DHDPE/LLDPE(2) system in the isothermal crystallization process from the melt. (From Reference 38 with permission from the American Chemical Society.)

LLDPE(2), and LLDPE(3) during the isothermal crystallization at almost the same degree of supercooling. The increasing rate of the band intensity is higher for PE sample with lower degree of branching: HDPE > DHDPE > LLDPE(2) LDPE(3). Figure 5.17 compares the crystallization rate between the D and H components in the DHDPE/LLDPE(2) blend and DHDPE. The D and H components in the blend, which are named D (blend) and H (blend), respectively, in this figure, are found to crystallize at almost the same rate, and this crystallization rate is much higher than that of the pure DHDPE.

5.5 MIXING BEHAVIOR OF D AND H COMPONENTS As already mentioned, a pair of DHDPE and LLDPE(2) cocrystallizes, while that of DHDPE and LLDPE(3) separates when cooled from the melt. In both the cases, however, the SANS experiment indicates that the D and H species are mixed homogeneously in the molten state (46). In what mechanism do the segregation and cocrystallization occur from such a homogeneously mixed molten state? For example, it might be reasonable to speculate that the D and H chains in the melt experience the diffusional motion at different rates and the chains of similar species gather to crystallize into their own lamellae, resulting in the phase segregation. In the cocrystallization phenomenon, the D and H chains diffuse at almost the same rate in the melt and are stabilized into the state of coexistence in the same lamella. (Of course, these two species are needed to be compatible to each other to be stabilized in the common lattice.) Therefore, we need to investigate the interpenetration behavior of the D and H chains in the molten state. We investigated the diffusion of D and H chains in the melt to understand the aggregation process of these two kinds of PE chains to form the crystalline lamellae (42). In the actual experiment, as shown in Fig. 5.18, the H and D films were contacted at the edges and the interpenetration process of D and H chains was observed by measuring the change in the infrared band

Chapter 5 Microscopically Viewed Structural Characteristics of Polyethylene Blends

115

Figure 5.18 Schematic illustration of the sample used for the IR microscopic observation at the interfacial boundary of DHDPE and LLDPE(2). (a) The cover slit for infrared spectral measurement was set at the open box position. (b) The sample was shifted from the contact position to measure the infrared spectra at the position x.

intensity of the H and D chains at the interfacial part as functions of time and position. The infrared spectral measurement at the interfacial part was made using an infrared microscope. Figure 5.19 shows the time dependence of infrared band intensity of D and H components measured for a pair of DHDPE and LLDPE(2) films at the interfacial part of the H film side. The H band decreased in intensity and the D band increased

Figure 5.19 (Left) Infrared spectral change during the interdiffusion procedure at the interfacial parts of DHDPE/LLDPE(2) films. H-side: profile changes of D and H bands measured at the LLDPE(2) side. D-side: profile changes of D and H bands measured at the DHDPE side. (Right) Time dependence of integrated intensity estimated for the D and H components at the H side.

116

Polyolefin Blends

with time. From these curves, the diffusion constant G was evaluated as 5:0 109 cm2 s1 pon the basis of Fick’s equation of diffusion; cðx; tÞ ¼ ð1=2Þc0 erfc ½x=ð2 ðGtÞÞ where a system of two infinitely long plates connected at one edge is assumed to solve the equation of diffusion @cðx; tÞ=@t ¼ Gð@ 2 cðx; tÞ=@x2 Þ. The cðx; tÞ is the concentration at a position x and at a time t, c0 is the initial concentration, and erfc is a complementary error function: p Rt erfcðtÞ ¼ 1 ð2= pÞ 0 expðy2 Þdy. Figure 5.20 compares the infrared spectra

Figure 5.20 (a) Infrared spectra measured at the various positions for a pair of contacted DHDPE and LLDPE (2) films. (b) Infrared spectra measured for a pair of DHDPE and LLDPE (3) films. The samples were quenched after being kept at 160 C for 10 min. The measured positions are indicated by figures with micrometer unit (refer to Fig. 5.18).

Chapter 5 Microscopically Viewed Structural Characteristics of Polyethylene Blends

117

Figure 5.21 Schematic illustration of diffusion process of the D and H chain components passing through the interface during the melt. (a) DHDPE/LLDPE(2) and (b) DHDPE/LLDPE(3). When the sample was quenched from the melt, the chains were crystallized with the relative positions remaining the same.

measured for DHDPE/LLDPE(2) and DHDPE/LLDPE(3) systems at the various positions apart from the interface for the samples quenched from the molten state. In the case of DHDPE/LLDPE(2) system, the spectral profile is found to change from singlet to doublet as the measurement point is shifted from the interface to the inner part, indicating mutual mixing of the D and H chains in the interfacial part during melt. Figure 5.21 shows this situation clearly, where the spatial distribution of D and H chains at the interfacial part between the D and H films is illustrated schematically. On the contrary, when the DHDPE and LLDPE(3) films are contacted at the edge, the D and H chains migrate also in the melt, but the distribution of the D and H chain stems is not perfectly homogeneous as seen in Fig. 5.20: at the positions distant from the interfacial part, the H chains and D chains are segregated and crystallize into the separated domains, as known from the infrared spectral patterns.

5.6 CONCLUSIONS In the present review, the crystallization behavior has been described for the D and H chains in the blend samples of DHDPE and LLDPE with various degrees of side

118

Polyolefin Blends

branching. The LLDPE sample with relatively low ethyl branching content was found to show almost perfect cocrystallization phenomenon with the D species even when the sample was cooled quite slowly from the melt. On the contrary, the LLDPE with relatively higher ethyl branching and the HDPE without any branching were found to show phase segregation phenomenon when the blend sample with DHDPE component was cooled slowly from the melt. This finding is quite important in the study of chain-folding mode in the lamella. The D and H chain stems are statistically randomly distributed in the lamella as clarified by the quantitative interpretation of the infrared band profile. The thermal data and the wide-angle and small-angle neutron scattering data also support this concept. From all these data, we concluded that the PE chains show almost perfectly random reentry mode when crystallized from the melt, as far as a pair of DHDPE and LLDPE(2) is concerned. The study of cocrystallization is important also from the industrial point of view. For example, the two kinds of pipes, each made of a different kind of PE, are connected to make a water pipe. The strength of the connecting part of the two pipes is determined by the degree of cocrystallization of the two kinds of PE used. The concept deduced from the study of a pair of DHDPE and LLDPE(2) might not be applied as a universal concept to any other kind of PE pair. But the knowledge accumulated for the D/H blend samples described here should be basically important and useful for the structural study of any type of PE blend samples.

ACKNOWLEDGMENTS The author wishes to thank Dr. Richard S. Stein and Dr. Shaw Lin Hsu of the Department of Polymer Science and Engineering, University of Massachusetts at Amherst, USA; Dr. M. Satkowskii of Procter & Gamble Co., USA; Dr. Sono Sasaki of Japan Synchrotron Radiation Institute, SPring-8, Japan; and Mr. Masaaki Izuchi, Mr. Kouji Imanishi, and Mrs. Gose Naomi of the Graduate School of Science, Osaka University, for their contribution in these studies.

NOMENCLATURE erfc PE DHDPE HDPE LLDPE WAXD SAXS LLDPE(2) LLDPE(3)

Complementary error function Polyethylene Deuterated high density PE High density PE Linear low density PE Wide-angle X-ray diffraction Small-angle X-ray scattering Linear low density polyethylene with 17 ethyl branchings per 1000 carbon atoms Linear low density polyethylene with 43 ethyl branchings per 1000 carbon atoms

Chapter 5 Microscopically Viewed Structural Characteristics of Polyethylene Blends

Dn Dn0 N f ðpÞ X hNi DT Q Rg G c(x,t) c0

119

Band splitting width Band splitting width of infinitely repeated array of oscillators Effective number of chain stems in the 110 direction Probability to have a sequence of D(H)pD Fraction of the H stems Averaged size of the D(H)pD clusters Degree of supercooling Invariant Radius of gyration Diffusion constant Concentration at a position x and at a time t Initial concentration

REFERENCES 1. D. R. Norton and A. Keller, J. Mater. Sci. 19, 447 (1984). 2. S. Hu, T. Kyu, and R. S. Stein, J. Polym. Sci. B Polym. Phys. Ed., 25, 71 (1987). 3. T. Kyu, S. Hu, and R. S. Stein, J. Polym. Sci. B Polym. Phys. Ed., 25, 89 (1987). 4. M. Ree, T. Kyu, and R. S. Stein, J. Polym. Sci. B Polym. Phys. Ed., 25, 105 (1987). 5. P. Vadhar and T. Kyu, Polym. Eng. Sci., 27, 202 (1987). 6. H. L. Marand, R. S. Stein, and G. M. Stack, J. Polym. Sci. B Polym. Phys. Ed., 26, 1361 (1988). 7. J. M. Rego Lopez and U. W. Gedde, Polymer, 29, 1037 (1988). 8. J. M. Rego Lopez, M. T. Conde Brana, B. Terselius, and U. W. Gedde, Polymer, 29, 1045 (1988). 9. J. M. Rego Lopez and U. W. Gedde, Polymer, 30, 22 (1989). 10. S. Hosoda and Y. Gotoh, Polym. J., 21, 17 (1988). 11. H. H. Song, R. S. Stein, D. Q. Wu, M. Ree, J. C. Philips, A. Legrand, and B. Chu, Macromolecules, 21, 1180 (1988). 12. H. H. Song, D. Q. Wu, B. Chu, M. Satokowski, M. Ree, R. S. Stein, J. C. Philips, A. Legrand, and B. Chu, Macromolecules, 23, 2380 (1990). 13. C. Recknger, F. C. Larbi, and J. Raut, J. Macromol. Sci. Phys., B23, 511 (1984–1985). 14. U. Wendt, J. Mater. Sci. Lett., 7, 643 (1988). 15. P. J. Barham, M. J. Hill, A. Keller, and C. C. Rouney, J. Mater. Sci. Lett., 7, 1271 (1988). 16. J. Martinez-Salazar, M. SanchezCuesta, and J. Plans, Polymer, 32, 2984 (199l). 17. J. Plans, M. SanchezCuesta, and J. Martinez-Salazar, Polymer, 32, 2989 (1991). 18. M. J. Hill, P. J. Barham, and A. Keller, Polymer, 33, 2530 (1992). 19. A. Prasad, Polym. Eng. Sci., 38, 1716 (1998). 20. M. Yamaguchi and S. Abe, J. Appl. Polym. Sci., 74, 3153 (1999). 21. G. D. Wignall, R. G. Alamo, J. D. London, and L. Mandelkern, Macromolecules, 33, 551 (2000). 22. C. C. Puig, Polymer, 42, 6597 (2001). 23. L. Y. Zhao and P. Choi, Mater. Manuf. Process., 21, 135 (2006). 24. M. Tasumi and S. Krimm, J. Polym. Sci. B Polym. Phys. Ed., 6, 995 (1968). 25. M. J. Bank and S. Krimm, J. Polym. Sci. B Polym. Phys. Ed., 7, 1785 (1969). 26. T. Cheam and S. Krimm, J. Polym. Sci. B Polym. Phys. Ed., 19, 423 (1981). 27. F. C. Stehling, E. Ergos, and L. Mandelkern, Macromolecules, 4, 672 (1971). 28. J. Schelten, D. G. H. Ballard, G. D. Wignall, G. W. Longman, and W. Schmaz, Polymer, 751 (1976).

120

Polyolefin Blends

29. J. Schelten, G. D. Wignall, D. G. H. Ballard, and G. W. Longman, Polymer, 18, 1111 (1977). 30. Special issue on chain folding problem: Faraday Discuss. Chem. Soc., 68 (1979). 31. M. Stamm, E. W. Fischer, and M. Dettenmaier, Faraday Discuss. Chem. Soc., 68, 263 (1979); Pure Appl. Chem., 50, 1319 (1978). 32. G. Jancso, L. P. N. Rebelo, W. A. Vanhook, Chem. Soc. Rev., 257 (1994). 33. K. Tashiro, R. S. Stein, and S. L. Hsu, Macromolecules, 25, 1801 (1992). 34. K. Tashiro, M. M. Satkowski, R. S. Stein, Y. Li, B. Chu, and S. L. Hsu, Macromolecules, 25, 1809 (1992). 35. K. Tashiro, M. Izuchi, M. Kobayashi, and R. S. Stein, Macromolecules, 27, 1221 (1994). 36. K. Tashiro, M. Izuchi, M. Kobayashi, and R. S. Stein, Macromolecules, 27, 1228 (1994). 37. K. Tashiro, M. Izuchi, M. Kobayashi, and R. S. Stein, Macromolecules, 27, 1234 (1994). 38. K. Tashiro, M. Izuchi, F. Kaneuchi, C. Jin, M. Kobayashi, and R. S. Stein, Macromolecules, 27, 1240 (1994). 39. K. Tashiro, K. Imanishi, Y. Izumi, M. Kobyashi, K. Kobayashi, M. Satoh, and R. S. Stein, Macromolecules, 28, 8477 (1995). 40. K. Tashiro, K. Imanishi, M. Izuchi, M. Kobayashi, Y. Itoh, M. Imai, Y. Yamaguchi, M. Ohashi, and R. S. Stein, Macromolecules, 28, 8484 (1995). 41. K. Tashiro, Acta Polym., 46, 100 (1995). 42. K. Tashiro and N. Gose, Polymer, 42, 8987 (2001). 43. K. Tashiro, S. Sasaki, and M. Kobayashi, Macromolecules, 29, 7460 (1996). 44. K. Tashiro, S. Sasaki, N. Gose, and M. Kobayashi, Polym. J., 6, 485 (1998). 45. S. Sasaki, K. Tashiro, M. Kobayashi, Y. Izumi, and K. Kobayashi, Polymer, 40, 7125 (1999). 46. S. Sasaki, K. Tashiro, N. Gose, K. Imanishi, M. Izuchi, M. Kobayashi, M. Imai, M. Ohashi, Y. Yamaguchi, and K. Ohyama, Polym. J., 31, 677 (1999). 47. K. Tashiro and H. Hama, Macromol. Res., 12, 1 (2004). 48. R. G. Snyder, M. C. Goh, Svivdtsavoy, H. L. Strauss, and D. L. Dorset, J. Phys. Chem., 96, 10008 (1992). 49. R. Zbinden, Infrared Spectroscopy of High Polymers, Academic Press, New York, 1964. 50. H. Hama and K. Tashiro, Polymer, 44, 6973 (2003). 51. H. Hama and K. Tashiro, Polymer, 44, 3107 (2003). 52. H. Hama and K. Tashiro, Polymer, 44, 2159 (2003).

Chapter

6

Thermal and Structural Characterization of Binary and Ternary Blends Based on Isotactic Polypropylene, Isotactic Poly(1-Butene) and Hydrogenated Oligo (Cyclopentadiene) Maurizio Canetti1

6.1 INTRODUCTION In the last 25 years, research into polymer blends has increased significantly with the aim to obtain new high performance materials without synthesizing totally new polymers. Polymer blending has been utilized to improve the achievement of polymeric materials by merging the characteristics of each constituent, making it possible to fit new thermal and mechanical specifications at relatively cheap price. Three different series of binary blends were prepared by mixing hydrogenated oligo (cyclopentadiene) (HOCP), isotactic polypropylene (iPP), and poly(1butene) (PB-1), alternatively. The influence of the presence of HOCP, on the morphology, crystallization, melting behavior, and supermolecular structure of

1

C.N.R. Istituto per lo Studio delle Macromolecole, Via E. Bassini 15, I-20133 Milan, Italy

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

121

122

Polyolefin Blends

iPP (1–3) and PB-1 (4,5) was investigated. The influence of PB-1 on thermal properties of iPP was also analyzed (6). Ternary iPP/PB-1/HOCP blends were prepared at different composition to investigate the possibility of improving the miscibility of the two polyolefins by the addition of HOCP. The composition has been found to affect structure, miscibility, crystallization, and melting properties of the ternary blends and the crystal modification of PB-1 (7,8). In the present work, the thermal and structural behaviors of binary and ternary blends based on iPP, PB-1, and HOCP are summarized. Morphological characterization was carried out by using electronic and optical microscopic techniques (SEM and POM). Structural and supermolecular characterizations were performed employing small-angle X-ray scattering and wide-angle X-ray diffraction techniques (SAXS and WAXD). Isothermal crystallization kinetics, melting properties, and glass transition temperatures were investigated by means of differential scanning calorimetry (DSC) and dynamical mechanical thermal analysis (DMTA). Studies carried out on melt-extruded films prepared from iPP/HOCP blends showed that the addition of HOCP causes the formation of smectic phase of iPP at temperature where the iPP crystallizes in the monoclinic a-form (9). These phenomena are attributed to the effect of HOCP on the glass transition temperature, equilibrium melting temperature, crystallization temperature, and rate of crystallization of iPP. Properties and plastic deformation of the oriented films were also investigated (10). Oxygen transport through membranes of iPP/HOCP blends was studied as a function of the weight fraction of the two components. The results showed that increasing HOCP content lowers oxygen permeability and diffusivity through the films. The HOCP has an antiplasticizing effect and produces a limitation in penetrant mobility (11). Binary blends prepared by mixing iPP with a fully saturated alicyclic hydrocarbon resin (obtained by cationic oligomerization of indene, a-methyl styrene, and vinyl toluene, followed by hydrogenation) showed good miscibility in the melt state (12). The influence of HOCP on the morphology, the phase structure, and the thermal behavior of its blends with PB-1 revealed that the spherulite growth rate, the overall crystallization rate, and the melting properties are strongly dependent on crystallization conditions and blending composition (13–15). Besides, the results suggested that PB-1 and HOCP components are not completely miscible, and they can form two conjugated amorphous phases (one rich in PB-1 component and the other rich in HOCP component), at least in the range of composition up to 30% of HOCP. Several authors have analyzed the miscibility of iPP and PB-1, by means of different analytical approaches. Piloz et al. (16) found a single, compositiondependent, glass transition behavior for these blends, and concluded that they are compatible in the amorphous state. Sjegmann (17,18) reported that the composition dependence of tensile properties evidences a high degree of compatibility of iPP and PB-1 and observed a marked effect of the composition on the morphology of melt-crystallized samples. Conversely, the analysis of the crystallized blends indicates the presence of separated crystal phases of the two polymers, even if a mutual influence during the crystallization cannot be excluded.

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

123

A partial inclusion of PB-1 chains in iPP crystal can occur under fast quenching of the blends from the melt, causing the formation of solid solution (19). Several papers report about properties of ternary blends with iPP as the main component. The majority of the ternary blends possess a multiphase morphology, and adequate mechanical and physical properties are correlated to the dimension of the particles of the dispersed phase. The compatibility could be related to the modification of the interfacial properties of the blend. To improve the compatibility in the ternary blends based on iPP and polyamides, different polymers grafted with maleic anhydride were added to the blend as interfacial modifier (20–24). Generally, the addition of the compatibilizer led to finer dispersion of the particles of the minor component and an upgrading of the mechanical properties of the blend. Different ternary blends were prepared by mixing iPP with several polyolefins with the aim to modify the mechanical properties of the matrix (25–29). The effect of the composition on the morphology, the rheological properties, and the crystallization behavior was investigated. The nature of the components can act as a nucleant agent on iPP crystallization and can produce different effects on mechanical performance of the blends. Studies on the morphology and on the melt rheological, tensile, and impact properties were carried out on ternary blend of iPP with two of the following polymers: low and high density polyethylene, styrene-b-ethylene butylene-b-styrene triblock copolymer, polystyrene, and acrylonitrile–butadiene–styrene terpolymer (30–33). The results are interpreted for the effect of each individual component by comparing the ternary blends with the respective iPP-based binary blends as the reference systems.

6.2 BINARY BLENDS 6.2.1 Blend Preparation Binary blends of iPP (Moplen T 30 S, Mw 300,000, Montecatini) and HOCP (Escorez, Mw 630, Esso Chemical) and these of iPP (Rapra, Mw 307,000) and PB-1 (Petrotex, Mw 1,060,000) were prepared by mixing them in the weight ratios of 90/10, 70/30, and 50/50. The blends of PB-1 (PB 8340, Mw 700,000, Shell) and HOCP were prepared in the weight ratio of 95/5, 90/10, 80/20, and 70/30. Binary blends were prepared by mixing the components in a laboratory miniextruder. Pure iPP and PB-1 were processed under similar conditions.

6.2.2 Glass Transition Temperature The blends prepared by mixing iPP and PB-1 with HOCP were investigated by DSC to obtain information about their glass transition temperature, Tg . Before scans, the

124

Polyolefin Blends Table 6.1 Experimental and Theoretical, Calculated by Equation 1, Glass Transition Temperature ðTg Þ, for PB-1/HOCP and iPP/HOCP Blends. Tg , C Sample PB-1/HOCP 100/0 PB-1/HOCP 95/5 PB-1/HOCP 90/10 PB-1/HOCP 80/20 PB-1/HOCP 70/30 PB-1/HOCP 0/100 iPP/HOCP 100/0 iPP/HOCP 90/10 iPP/HOCP 80/20 iPP/HOCP 70/30 iPP/HOCP 0/100

Experimental 26 24 20 19 18 67 13 6 7 25 67

Calculated by Equation 1 — 23 19 12 4 — — 7 7 22 —

sample was melted and immediately immersed in liquid nitrogen to obtain a completely amorphous material. The amorphous blends showed a single Tg with numerical value dependent on composition. The dependence of Tg on the weight fraction of HOCP (referred to the overall amorphous content in the blends) is shown in Table 6.1. The Tg values of the blends were also calculated by using the theoretical relation of Fox (34): 1=TgðblendÞ ¼ ½WðPÞ =TgðPÞ þ ½WðHOCPÞ =TgðHOCPÞ

ð6:1Þ

where WðHOCPÞ and TgðHOCPÞ are the weight fraction and the glass transition temperature of the HOCP component, respectively, and WðPÞ , and TgðPÞ alternatively, are the weight fraction and the glass transition temperature of the PB-1 or iPP components. A good agreement was observed between the experimental values obtained for the iPP/HOCP blends and the ones calculated by Fox equation. The appearance of a single glass transition temperature should suggest that the blend presents a single homogeneous amorphous phase, that is, the two components are miscible in the amorphous phase. For the PB-1/HOCP blends, the experimental values of Tg always below the theoretical values. This behavior could be an indication of specific interactions between the two components in the amorphous phase and/or an indication of incomplete compatibility between the components (15).

6.2.3 Morphology and Spherulite Growth Rate Thin films of pure iPP and PB-1 homopolymers and iPP/HOCP, PB-l/HOCP blends, when observed under the optical polarizing microscope during the isothermal

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

125

crystallization process, show birefringent spherulitic structures. After complete crystallization, the samples appear completely filled with impinged spherulites for all the HOCP concentrations studied. The absence of HOCP domains from both the intraspherulitic regions and the interspherulitic contact zones suggests that the HOCP is incorporated in the interlamellar or interfibrillar zones of iPP or PB-l spherulites. The morphology of the iPP spherulites in the iPP/PB blends results are affected by the PB-1 content at concentrations of PB-1 30%, the spherulites show a less regular texture with a coarse fibrillar structure, indicating that the growth process of the iPP crystals depends on the melt composition. The microscopic examinations do not reveal the presence of PB-1 phase separation in the interspherulitic contact zones and suggest that the molten component can be incorporated during the crystallization of iPP in the amorphous regions of the spherulites. Investigations carried out by SEM on the etched surfaces of iPP /PB-1 samples, crystallized from the melt at isothermal crystallization temperature, Tc 125 C, and cooled to room temperature have demonstrated the presence of small PB-1 inclusions, quite homogeneously dispersed within the spherulites, for PB-1 concentrations larger than 30%. These morphological features depend on factors such as composition, crystallization rate, and diffusion rate of the components in the melt (35,36). Thus, at moderate PB-1 concentrations, the PB-1 component should segregate during the iPP crystallization and affect both the spherulite growth and the kinetics of bulk crystallization. Moreover, for samples crystallized from the melt in a lower temperature range, under nonisothermal conditions, different morphologies have been observed depending on the composition ratio and crystallization rates of both components (17). Bartczak et al. (37) found that iPP and PB-1 are miscible in the melt but the thermal treatment induces partial phase separation of components and the formation of iPP- and PB-1-rich phases. The complete phase separation needs high temperatures and a long time of melt annealing. For pure iPP and PB-1 homopolymers and their respective blends, the spherulite radius Rs increases linearly with time t for all Tc , investigated. For all samples, the isothermal radial growth rate G was calculated at different Tc as G ¼ dRs =dt. Generally, the G values decrease an increase in the Tc values and with increase in the amount of noncrystallizable component in the blend. As shown in Fig. 6.1, where the relative G values of the iPP-based blends are reported for Tc ¼ 125 C, the depression of the G values was more pronounced for the blends prepared with HOCP as the second component. Analogously, for a given Tc , the addition of HOCP to PB-l causes a depression of the G values, allowing the control of the isothermiticy of the PB-l crystallization at lower Tc values.

6.2.4 Isothermal Bulk Crystallization Kinetics The DSC crystallization isotherms of pure iPP, pure PB-l, and blends compared at the same Tc demonstrate that the overall crystallization rate constant progressively decreases with increase in the amount of the diluent component in the sample.

126

Polyolefin Blends

Figure 6.1 Relative G values, GðblendÞ =GðiPPÞ , of iPP/PB-1 (&) and iPP/HOCP (*) blends for the Tc ¼ 125 C.

The isotherms of crystallization were obtained by plotting the crystalline weight fraction at time tðXt Þ versus t. The half-time of crystallization, t0:5 (defined as the time taken for half of the crystallinity to develop), is obtained from the curves and plotted against blend composition, for some Tc , in Fig. 6.2. The reduction in the overall crystallization rate of iPP in the iPP/HOCP blends was more drastic to the one engendered by the addition of PB-1 as the second component. The addition of HOCP to PB-1 causes a striking reduction in the overall crystallization rate. The overall kinetic rate constant Kn was calculated by using the Avrami equation (38): Xt ¼ 1 expðKn tn Þ

ð6:2Þ

where n is a parameter depending on the type of nucleation and on the geometry of the growing crystals. The values of n and Kn were derived, for each Tc , from the slope and the intercept, respectively, of the straight lines obtained by plotting log10 ½ lnð1 Xt Þ versus log10 t (Fig. 6.3). The value of the Avrami exponent n was about 3 for pure iPP, pure PB-1, and all blends for all the investigated Tc . This indicates a threedimensional growth of crystalline units developed by heterogeneous nucleation (39).

6.2.5 Temperature Dependence of the Spherulite Growth Rate and the Overall Kinetic Rate Constant Assuming that in the examined temperature range, the iPP/HOCP, iPP/PB-1, and PB-1/HOCP blends constitute a polymer–diluent system, the experimental G and Kn values were analyzed according to the polymer–diluent theory (40–42).

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

127

Figure 6.2 Half-time of crystallization, t0:5 , versus composition, of (a) iPP/HOCP blends, Tc ¼ 125 C; (b) iPP/PB-1 blends, Tc ¼ 127 C; and (c) PB-1/HOCP blends, Tc ¼ 84 C.

128

Polyolefin Blends

Figure 6.3 Avrami plot (Eq. 6.2) for different Tc of PB-1/HOCP 70/30 blend.

According to this theory, the equation describing the growth rate G of spherulites of a crystallizable polymer in a one-phase melt containing a second polymer acting as a diluent assumes the form log G log f2 þ ½U =RðTc T1 Þ ½ð0:2Tm log f2 Þ=DT ¼ a1 ¼ log G0 ðDGc =TDTf Þ

ð6:3Þ

where G0 is the preexponential factor, which includes all the terms that are taken as effectively independent of temperature. The U =RðTc T1 Þ term contains the energetic contribution to the growth rate of diffusional processes of the amorphous and crystallizable material: U is the sum of the activation energies for the chain motions in the melt of crystallizable and noncrystallizable molecules and T1 ðT1 ¼ Tg – C), where C is a constant) is the temperature below which such motions cease (R is the universal gas constant). Tm is the equilibrium melting temperature and f2 is the volume fraction of the crystallizable polymer. The term f is a correction factor ð f ¼ 2Tc =Tm þ Tc, where Tm is the observed melting temperature); the term DGc contains the free energy required to form a nucleus of critical size. The slope of the straight lines obtained, for each mixture, by plotting a1 versus Tm =ðTc DTf Þ gives the DGc value. As an example, in Fig. 6.4 the plots relative to the PB-1/HOCP blends are reported. Analogously, the temperature dependence of the overall kinetic rate constant ðKn Þ can be expressed by the relation: 1=n log f2 þ ½U =RðTc T1 Þ ½ð0:2Tm log f2 Þ=DT ¼ a2 ¼ log An ðDGc =TDTf Þ

ð6:4Þ

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

129

Figure 6.4 a1 versus Tm =ðTc DTf Þ according to Equation 6.3 for PB-1/HOCP blends: 100/0 (), 95/5 (~), 90/10 (&), 80/20 (^), and 70/30 (*).

vPB;I0 where An is a constant equal to log Gð4=3ÞN 1=3 , where N is the nucleation density. As heterogeneous nuclei form instantaneously as soon as the temperature reaches values below Tm, the nucleation process can be considered independent of crystallization temperature. Straight lines were obtained by plotting a2 versus Tm =ðTc DTf ), where the slopes give the DGc values; in Fig. 6.5 the plots relative to the iPP and iPP/PB-1 blends are reported. As reported in Reference 3, the same values of C ¼ 30K and U ¼ 6:28 kJ moll were used in both Equations 6.3 and 6.4.

Figure 6.5 a2 versus Tm =ðTc DTf Þ according to Equation 6.4 for iPP/PB-1 blends: 100/0 (*), 90/10 (~), 70/30 (&), and 50/50 (^).

130

Polyolefin Blends Table 6.2 Folding Surface Free Energy, s e , for PB-1/HOCP, iPP/PB-1, and iPP/HOCP Blends. Sample PB-1/HOCP 100/0 PB-1/HOCP 95/5 PB-1/HOCP 90/10 PB-1/HOCP 80/20 PB-1/HOCP 70/30 iPP/HOCP 100/0 iPP/HOCP 90/10 iPP/HOCP 70/30 iPP/HOCP 50/50 iPP/PB-1 100/0 iPP/PB-1 90/10 iPP/PB-1 70/30 iPP/PB-1 50/50

s e from G, erg cm2 17 15 13 11 10 116 121 108 99 — — — —

s e from Kn , erg cm2 19 16 15 13 14 116 122 111 98 112 92 78 56

In our range of crystallization temperatures, according to Lauritzen (43), DGc can be expressed as follows: DGc ¼ ð4b0 ss e Tm Þ=DHk

ð6:5Þ

where s and s e are the lateral and folding surface free energy, respectively, Tm is the equilibrium melting temperature, DH is the enthalpy of fusion per unit volume, and k is the Boltzmann constant. The s e values were calculated for the PB-1/HOCP blends with b0 ¼ 7:49 108 cm and s ¼ 0:1b0 DH and, alternatively, from the DGc values obtained either by Equation 6.3 or by Equation 6.4 (44). In Table 6.2 it can be noted that the two series of s e values, so obtained, have a similar dependence on blend composition (except for the PB-l/HOCP 70/30 blend in the series of s e values obtained from Kn ). The s e values decrease with increase in the HOCP fraction in the blend. Such depression may be accounted for by assuming that amorphous HOCP is likely to segregate in interlamellar regions inducing an increase in the surface entropy of folding. As a consequence, PB-1 lamellar crystals with a less regular fold surface are obtained when the PB-l/HOCP blends are isotherma1ly crystallized from the melt in the examined range of Tc values. For the iPP-based blends, the s e values were calculated with b0 ¼ 5:25 108 cm (44). The s e values decrease with increase in the PB-1 fraction in the iPP/PB-1 blends. Such depression may be accounted for by assuming that amorphous PB-1 is likely to segregate in the interlamellar regions of iPP spherulites inducing an increase in the surface entropy of folding. As a consequence, iPP lamellar crystals with a less regular fold surface are obtained when the iPP/PB-1 blends are isothermally crystallized from melt.

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

131

A good agreement was observed between the s e values calculated form Kn as well as from G, for iPP/HOCP blends. The 70/30 and 50/50 blends show s e values that are meaningfully lower than that for pure iPP.

6.2.6 Melting Behavior The observed melting temperature ðTm Þ of pure iPP, pure PB-l, and blends linearly increases with the crystallization temperature. The experimental data can be fitted by the Hoffman equation (45): Tm ¼ ð1=gÞTc þ ð1 1=gÞTm

ð6:6Þ

where 1=y is the stability parameter, which depends on the crystal thickness and Tm is the equilibrium melting temperature. By increasing the fraction of the diluent component in the blend, a depression of the Tm value can be observed for every Tc explored. As reported in Table 6.3, for all blend systems examined, the lower the extrapolated Tm value, the higher the content of the second component in the blend. In Equation 6.6, 1=y assumes values between 0 (Tm ¼ Tm for all Tc ) and 1 (Tm ¼ Tc ). Therefore, the crystals are most stable at 1=y ¼ 0 and inherently unstable at 1=y ¼ 1 (46). For the PB-1/HOCP and the iPP/PB-1 blends, the 1=y values decrease slightly with increase in the second component content, indicating an increase in the crystal stability. However, the values of 1=y are very similar for the iPP/HOCP blends, that is, independent of composition.

Table 6.3 Equilibrium Melting Temperature, Tm , and Stability Parameter, 1=g, for PB-1/HOCP, iPP/PB-1, and iPP/HOCP Blends. PB-1/HOCP PB-1/HOCP 100/0 PB-1/HOCP 95/5 PB-1/HOCP 90/10 PB-1/HOCP 80/20 PB-1/HOCP 70/30 iPP/PB-1 100/0 iPP/PB-1 90/10 iPP/PB-1 70/30 iPP/PB-1 50/50 iPP/HOCP 100/0 iPP/HOCP 90/10 iPP/HOCP 70/30 iPP/HOCP 50/50

Tm ( C)

1=g

130 125 122 118 113 183 181 177 173 188 186 182 178

0.39 0.34 0.31 0.28 0.22 0.44 0.41 0.39 0.35 0.37 0.37 0.36 0.36

132

Polyolefin Blends

According to the Flory–Huggins theory, the equilibrium melting point depression can be related to the polymer–polymer interaction parameter, x12 , by (46,47): ½ðDHV1 =RV2 Þð1=Tmb 1=Tmp Þ þ ðln f2 =m2 Þ þ ð1=m2 1=m1 Þf1 ¼ b ¼ x12 f21

ð6:7Þ where subscripts 1 and 2 represent the noncrystallizable and the crystallizable polymer, respectively. DH is the perfect crystal heat of fusion of the crystallizable polymer, V is the molar volume of the polymer unit at the equilibrium melting and Tmb are the equilibrium melting temperature, R is the universal gas constant, Tmp temperatures of the pure crystallizable component and of the blend, respectively, f is the volume fraction of the components in the mixture, and m is the degree of polymerization. A plot of the left-hand side of Equation 6.7 b versus f1 should give a straight line passing through the origin if the interaction parameter is assumed to be composition independent. For the PB-1/HOCP blends, the fol1owing parameter values have been used in our calculations: DH ¼ 4:184 kJ mol1(48); V1 ¼ 69:87 cm3 mol1(48); V2 ¼ 72:58 cm3 mol1 (48); m1 ¼ 9:3, and m2 ¼ 8911. The Tm values used are those reported in Table 6.3. The straight line calculated with the aforementioned parameters is plotted in Fig. 6.6, it shows an intercept on the ordinate axis of about 1:3 102 and a slope of 0:196 which corresponds to the x12 of the mixture. Relatively to the iPP/PB-1 blends, the parameter values used in the calculations were DH ¼ 10:142 kJ mol1 (48); V1 ¼ 72:04 cm3 mol1 (48); V2 ¼ 53:76 cm3 mol1 (48); m1 ¼ 18895, and m2 ¼ 7296. The Tm values used are reported in Table 6.3. From the straight line plotted in Fig. 6.6, an intercept of 1:807 102 and a x12 value of 0:257 were calculated. Finally, the parameters used for the calculation in the case of the iPP/HOCP were DH ¼ 10:142 kJ mol1 (48); V1 ¼ 59:70 cm3 mol1 (48); V2 ¼ 53:76 cm3 mol1

Figure 6.6 b versus f21 according to Equation 6.7 of iPP/HOCP blends (*), iPP/PB-1 blends (), and PB-1/HOCP blends (&).

133

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

(48); m1 ¼ 9:3, and m2 ¼ 7142. The Tm values used are those reported in Table 6.3. An intercept of 5:0 103 and a x12 value of 0:077 were calculated from the straight line plotted in Fig. 6.6. The negative value of x12 parameter in the three-blend system could suggest that the two components can form a compatible mixture that is thermodynamical1y stable above the equilibrium melting temperature. The high value of the intercept can be due to composition dependence of x12 and/or to contribution to Tm of morphological and annealing effects (41) that are not taken into account in Equation 6.7. However, as suggested by Cimmino et al. (14) for PB-1/HOCP system, the negative value of x12 parameter does not completely mean that the blend is miscible, since morphological effects can contribute to the equilibrium melting point depression.

6.2.7 Polymorphism and Phase Transformation of Poly (1-Butene)/Hydrogenated Oligo (Cyclopentadiene) The PB-1 crystal1izes from the melt in the tetragonal crystal modification (form II), and then transforms into the hexagonal crystal modification (form I). This transformation, which is affected by time, temperature, atmospheric pressure, and mechanical deformation, changes the thermal, mechanical, and physical properties of the polymer (49–55). The thermal behavior of PB-l is strongly influenced by the presence of HOCP. The variations of the kinetics of PB-l crystal transformation from form II to form I were considered to be a consequence of the crystallization of PB-l in the presence of HOCP. The samples were crystallized from the melt at different temperatures, Tc (Table 6.4), chosen so as to provide the same undercooling according to the relation Tc ¼ Tm DT, in which Tm values are reported in Table 6.3 and DT ¼ 46 C was kept constant for all isothermal crystallizations. The crystallized samples were kept for different periods of time, at the following different temperatures, Ta : 4, 20, 40, and 69 C. The presence of HOCP considerably slows down the melt crystallization process of PB-l. Therefore, the adopted Tc values, lowered by increasing the HOCP fraction, provided similar rates of crystallization for pure PB-l and blends. Previous calculations from the spherulite growth rate and from the overall kinetic rate constant showed that the number of nuclei per unit volume was similar for samples crystallized at equal undercoolings. Had we used a constant value of Tc , there would have

Table 6.4 Crystallization Temperature Tc , Melting Temperature Tm , and Crystalline Weight Fraction v0 of PB-1/HOCP Blends. PB-1/HOCP

Tc , C

100/0 90/10 80/20 70/30

83.8 75.5 71.9 67.8

Tm , C (form II) 112.7 107.3 104.8 102.9

Tm , C (form I)

v0

127.7 125.4 124.0 122.8

0.441 0.428 0.439 0.430

134

Polyolefin Blends

been large differences in nucleation; over the range of temperatures considered, nucleation would decrease as HOCP content increased. During the conversion process, the melting temperatures ðTm Þ of the two crystalline forms were quite constant at all the aging temperatures. The mean values for Tm are reported in Table 6.4. The Tm values decreased as the HOCP fraction increased in the blends due to the different Tc values adopted and the interference by HOCP. The apparent enthalpies of fusion ðDHII Þ were calculated from the area under the endothermic peaks of form II, obtained by DSC. The crystalline weight fractions of form II, vII , referred to the amount of PB-l present in the blends, were calculated from the relation: vII ¼ DHII =DHPBII f

ð6:8Þ

1

in which DHPBII ¼ 75 J g is the heat of melting of the 100% crystalline form II of PB-1 (56) and f is the PB-1 weight fraction (w/w) in the blend. Analogously, the crystalline weight fractions of form I of PB-l, vI , were calculated for pure PB-l and blends from the relation: vI ¼ DHI =DHPBI f

ð6:9Þ

in which DHI is the apparent enthalpy of fusion of form I of the sample, DHPBI ¼ 126 J g1 is the heat of melting of 100% crystalline form I of PB-l (56). Pure PB-l and blends, scanned immediately after isothermal crystallization, showed only the melting peak typical of the crystalline form II. The vII values calculated at aging time zero, v0 , were the mean values obtained from several DSC scans run immediately after isotherma1 crystallization (Table 6.4). As reported by Hong and Spruiell (51), the conversion of form II to form I starts immediately after crystallization from the melt. For both pure PB-1 and blends, the transformation continues until complete disappearance of form II at all aging temperatures. The amount of form II at time t ðXtII Þ was calculated by XtII ¼ DHIIt =DHIIt0

ð6:10Þ

in which DHIIt is the enthalpy of fusion at time t and DHIIt0 is the enthalpy of fusion at time zero. It can be seen that the isothermal conversion process became increasingly slower when the form II of PB-l was crystallized from the melt in the presence of increasing amounts of HOCP. This trend was observed at all aging temperatures except 69 C. It was also noted in the slow phase transformation curves that complete conversion of residual small amount of form II into form I was exceedingly slow. This might have been due to continuous supplies of form II from the amorphous phase by secondary crystallization of the PB-l (51). The half-life, t1=2 , is the time necessary for the form II crystallinity index XtII to reach 0.5. It can be noted that at a constant aging temperature, the t1=2 values increase with increasing HOCP content (Table 6.5). After isothermal crystallization from the melt, new materials with different amorphous phases were obtained due to the presence of HOCP. This amorphous phase showed a single glass transition, which increased with increasing content of HOCP, for which the Tg (67 C) is much higher

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

135

Table 6.5 Half-Life t1=2 (min) of Transformation of Crystalline Form II to Form I, for the PB-1/HOCP Blends, at Different Temperature of Aging Ta : PB-1/HOCP 100/0 90/10 80/20 70/30

t12 ðTa ¼ 4 C) 51.56 520.20 903.26 2223.09

t12 ðTa ¼ 20 C)

t12 ðTa ¼ 40 C)

t12 ðTa ¼ 69 C)

30.71 119.20 225.76 275.94

27.17 40.06 70.19 101.11

795.99 360.04 568.13 951.22

than that for PB-1 (Table 6.1). The miscibility of PB-1 with HOCP in the melt and the presence of a single glass transition suggest that the HOCP molecules are located in the interlamellar regions of PB-l, where they form a homogeneous mixture with uncrystallized PB-l molecules. The molecular mobility of the new amorphous phase decreases as function of the HOCP content. In Fig. 6.7 the t1=2 values are plotted against the aging temperatures. An analogous trend of the rate of PB-l crystal transformation as a function of aging temperature has already been observed (51,52). It was reported that the molecular mobility in the first decreasing part of the curve should control the conversion process, while the process should be controlled by nucleation after the minimum of the curve. This interpretation was proposed by Boor and Mitchell (52) who applied to the solid–solid transition the same approach used for crystallization from the melt. In our case, the HOCP greatly influenced transformation in the temperature range in which molecular mobility should be the controlling factor. This influence increases with increasing HOCP content in the blend and with decreasing aging temperature. Increasing the HOCP content and decreasing the aging temperature bring the

Figure 6.7 Half-life t12 of the form II for PB-1/HOCP blends, as a function of the temperature of aging, Ta : 100/0 (&), 90/10 (&), 80/20 (*), and 70/30 ().

136

Polyolefin Blends

material to a temperature near the Tg and thus to a situation of greater molecular rigidity. The data do not clarify the influence of HOCP in the region of aging temperatures in which nucleation controls the process. With the aging temperature of 69 C, the region of reorganization of the PB-l crystals was reached. In this region, a postcrystallization annealing or a perfecting of the crystals may happen (57), especially for the PB-1/HOCP 80/20 and 70/30 blends, whose Tc are close to 69 C. However, the verified effect of HOCP as a diluent rather than a nucleant induced to prefer to investigate the aging temperature region in which the transformation rate is mostly influenced by molecular mobility. Some previous investigators (51–53,58–60) have applied the Avrami equation to the analysis of the phase transformation of form Il to form I. The degree of crystal transformation changes with time according to the Avrami equation (Eq. 6.6), where Xt , in this case, represents the phase transformation of form II to form I. The n values are not influenced by the temperature of aging or by the amount of HOCP in the blends. The Avrami index n near one, calculated for all the analyzed samples, suggests that the nuclei of crystal form I appear spontaneously almost immediately after the melted polymer crystallizes into crystal form II. Crystal form I nuclei grow only along one direction (51,60,61). During the conversion process, the total crystallinity of the sample, vII þ vI , gradually increased with aging. For all samples, we observed more form I than the amount of form II that had disappeared. Several investigators (51,52,57) have suggested different possible mechanisms for this secondary crystallization of PB-l. The amorphous phase might crystallize to form II, which subsequently is transformed into form I. Another mechanism would be direct transformation of the amorphous phase to form I. The secondary crystallization can also be described as perfecting the imperfect crystal forms II and I. The secondary crystallization might result from any of the above mechanisms, and at present we cannot assess the relative importance of any of them. In Fig. 6.8, the weight fraction of form I at time t ðvIt Þ is plotted against the disappearing fraction of form II ðv0 vIIt ). At all temperatures of aging, straight lines passing through the origin were obtained for both pure PB-l and blends until complete conversion of form Il, that is, when the abscissa value approached the initial crystallinity value of the sample ðv0 vIit v0 Þ. The slopes of the straight lines (a) quantitatively evidence the remarkable enhancement of form I due to secondary crystallization (Table 6.6). It was never possible from the a values to extrapolate a homogeneous trend as a function of aging temperature or of the HOCP fraction.

6.2.8 Supermolecular Structure of Isotactic Polypropylene/Hydrogenated Oligo (Cyclopentadiene) Blends The crystallization and thermal behavior of iPP/HOCP blends suggested the two components are miscible in the melt. Besides, during crystallization of iPP, the molecules of HOCP are mainly ejected in interlamellar regions where they form a homogeneous phase with uncrystallized iPP molecules or part of them. Next studies

137

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

Figure 6.8 Amount of crystalline form I ðvIt ) in PB-1/HOCP 70/30 blend, aged at 40 C, as a function of the disappearing fraction of the crystalline form II ðv0 vIIt Þ.

regard the influence of different annealing processes on the phase structure of isothermally crystallized samples of iPP/HOCP. It was found that the presence of HOCP molecules in the iPP interlamellar regions can determine a loss of thickening tendency of crystalline lamellae promoted by the annealing. The amount of this effect is dependent on composition and undercooling. Blends and plain iPP were melted at 200 C for 10 min, then isothermally crystallized. Tc was calculated according to the relation: Tc ¼ Tm DT

ð6:11Þ

where Tm is the equilibrium melting temperature and DT is the undercooling, which was chosen constant for every isothermal crystallization. Then all the isothermally crystallized samples were annealed for 24 h at the annealing temperature ðTan Þ, calculated according to the relation: Tan ¼ Tc þ 17 C

ð6:12Þ

The details of the crystallization and annealing for each sample are reported in Table 6.7. Table 6.6 Slope Value (a) for the vIt versus v0 vIIt Plot as an Index of Secondary Crystallization of the PB-1/HOCP Blends at Different Temperature of Aging Ta : PB-1/HOCP 100/0 90/10 80/20 70/30

aðTa ¼ 4 C) 1.214 1.411 1.408 1.331

aðTa ¼ 20 C) 1.239 1.354 1.318 1.362

aðTa ¼ 40 C) 1.366 1.329 1.325 1.362

aðTa ¼ 69 C) 1.242 1.293 1.302 1.399

138

Polyolefin Blends

Table 6.7 Temperature of Crystallization Tc and Temperature of Annealing Tan for Pure iPP and iPP/HOCP Blends. DT ¼ 65:4 C iPP/HOCP 100/0 90/10 70/30 50/50

Tc ( C)

DT ¼ 59:3 C Tan ( C)

122.9 120.6 116.8 112.7

139.9 137.6 133.8 129.7

Tc ( C)

Tan ( C)

129.0 126.7 122.9 120.8

146.0 143.0 139.9 137.8

The thermal behaviors of the crystallized and annealed samples were investigated by DSC. The apparent enthalpies of fusion, DH, of plain iPP and blends were calculated from the area of the endothermic peaks. The crystalline and amorphous weight fractions were calculated from the following relations: vcr ¼ DH=DHPP ;

vm ¼ 1 vcr

ð6:13Þ

1

where DHPP (207 J g ) is the heat of melting per gram of 100% crystalline iPP (48). The crystalline weight fractions referred to the iPP were calculated from 1 vcrPP ¼ vcr =XPP

ð6:14Þ

where XPP is the weight fraction of iPP in the blends. As shown by the data reported in Table 6.8, the crystallinity of the blends vcr decreases on increasing the HOCP content, while the crystallinity of the iPP Table 6.8 Amorphous and Crystalline Weight Fractions, vam and vcr , and Crystallinity of iPP Components, vcrPP , of Pure iPP and iPP/HOCP Blends Isothermally Crystallized and Annealed. DT ¼ 65:4 C

Annealed samples

iPP/HOCP

vam

vcr

vcrPP

vam

vcr

vcrPP

100/0 90/10 70/30 50/50

0.49 0.53 0.61 0.68

0.51 0.47 0.39 0.32

0.51 0.52 0.56 0.64

0.47 0.49 0.57 0.66

0.53 0.52 0.43 0.34

0.53 0.57 0.62 0.68

DT ¼ 59:3 C

Annealed samples

iPP/HOCP

vam

vcr

vcrPP

vam

vcr

vcrPP

100/0 90/10 70/30 50/50

0.48 0.50 0.58 0.65

0.52 0.50 0.42 0.35

0.52 0.55 0.60 0.70

0.45 0.48 0.56 0.62

0.55 0.52 0.44 0.38

0.55 0.58 0.62 0.76

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

139

Figure 6.9 Melting point versus HOCP content: isothermal crystallization at DT ¼ 65:4 C (&) and annealing (&); isothermal crystallization at DT ¼ 59:3 C () and annealing (*).

component, vcrPP , increases. The annealing produced a general enhancement of the crystallinity. Plots of the observed calorimetric melting temperature, Tm , against HOCP weight fraction for pure iPP and iPP/HOCP blends are shown in Fig. 6.9. The Tm decreases linearly with HOCP content according to the thesis of compatibility in the melt at molecular level between iPP and HOCP. It can be observed that, for both pure iPP and blends, annealing produces an increase in melting temperature. The lamellar thickness ðlc Þ, the thickness of amorphous interlamellar region ðla Þ, and the long period (L), defined as the distance between the centers of two adjacent lamellae, were calculated by using the small-angle X-ray scattering technique (62) and are reported in Table 6.9. From the data of Table 6.9, it is obvious that the long spacing of iPP is slightly larger than that of iPP /HOCP 90/10 blends. On the contrary, 70/30 and 50/50 blends are characterized by values of L significantly larger than that of plain iPP; moreover, the lamellar thickness is almost independent of blend composition. Such a result, in addition to the trend of the thickness of the amorphous interlamellar region, suggests that the uncrystallizable component is located in the interlamellar regions especially in the case of blends with high HOCP content. The constant trend observed for the lamellar thickness lc is accounted for by considering that the crystallization is performed at DT constant. The annealing produced a general increase in the long period and lamellar thickness for pure iPP and blends. Thus, with a certain accuracy, the trend of Tm against 1=lc can be described by the following relation (40): Tm ¼ Tm ð2s e Tm =DHPP Þ1=lc

ð6:15Þ

Plotting the Tm obtained by DSC versus the inverse of the lamellar thickness ð1=lc Þ, a straight line was observed. From the intercept, the equilibrium melting temperature

140

Polyolefin Blends

Table 6.9 Long Period L (nm), Lamellar Thickness lc (nm), and Amorphous Interlamellar Thickness la (nm) of Pure iPP and iPP/HOCP Blends Isothermally Crystallized and Annealed. DT ¼ 65:4 C iPP/HOCP 100/0 90/10 70/30 50/50

Annealed samples

L

lc

la

L

lc

20.5 18.3 45.0 48.3

9.4 8.6 10.4 10.5

11.1 9.7 34.6 37.8

23.9 21.0 48.3 50.5

12.1 9.5 10.7 13.8

DT ¼ 59:3 C iPP/HOCP 100/0 90/10 70/30 50/50

la 11.8 11.5 37.6 36.7

Annealed samples

L

lc

la

L

lc

22.7 21.2 46.0 48.3

9.0 7.7 11.1 11.1

13.7 13.5 34.9 37.2

27.0 24.7 49.3 50.5

12.6 10.5 13.3 14.5

la 14.4 14.2 36.0 36.0

Tm , was deduced. Figure 6.10 reports the plots relative to the blends and the extrapolated Tm values for all the investigated samples. The extrapolated Tm values are in quite good agreement with those obtained by DSC measurements (see Table 6.3) for pure iPP, 90/l0, and 70/30 iPP /HOCP blends. On the contrary, for the 50/50 iPP /HOCP blend, a lower Tm value was extrapolated.

Figure 6.10 Plots of melting temperatures, Tm , versus the inverse of lamellar thickness, lc , for iPP/HOCP blends: 90/10 (&); 70/30 (); 50/50 (*).

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

141

Table 6.10 Weight Ratio of the iPP/PB-1/HOCP Ternary Blends. iPP PB-1 HOCP

80 10 10

70 10 20

70 20 10

60 10 30

60 20 20

60 30 10

50 10 40

50 20 30

50 30 20

50 40 10

6.3 TERNARY BLENDS 6.3.1 Blends Preparation Ternary blends of iPP(Moplen T 30 S, Mw 300,000, Montecatini), PB-1 (PB 8340, Mw 700,000, Shell), and HOCP (Escorez, Mw 630, Esso Chemical) were prepared by melt processing in the 50 ml mixing room of a Brabender Plasticorder PLE 330 mixer at 60 rpm and 195 C for 15 min, dry nitrogen was continuously purged into the mixing chamber. Neat iPP and PB-1 processed under similar conditions were investigated as reference materials. The weight ratio of the blend components are reported in Table 6.10.

6.3.2 Morphology and Spherulite Growth Rate Thin films of iPP/PB-1/HOCH blends were observed under the optical polarizing microscope during a cooling run from 260 to 90 C at 2 C min1, no melt phase separation was observed for all the examined compositions. A birefringent spherulitic structure relative to iPP crystallization appeared at temperature lower than 140 C. For all investigated compositions, when iPP crystallization was complete, no separate domains were observed either in the intraspherulitic region or in the interspherulitic contact zones. Cimmino et al. (63) found that the industrial films produced by using binary blends of iPP and HOCP, when cooled from 260 C to lower temperatures, showed a modulated structure on optical microscopy, indicating the formation of two phases. Examinations by SEM of blends crystallized from the melt at room temperature revealed a homogeneous surface without domains of a separated phase for 70/20/10, 50/20/30 iPP/PB-1/HOCP blends, and for all the ternary blends containing 10% of PB (Fig. 6.11a). For other blend compositions investigated, SEM micrographs exhibited the presence of bubbles as a separate phase having a diameter mostly lower than 2 mm (Fig. 6.11b). On the contrary, when those last blends were cooled from the melt at 120 C, kept in isothermal condition for a sufficient time to crystallize the iPP component and ultimately quenched in liquid nitrogen to avoid the PB-1 crystallization, they showed a homogeneous surface without the presence of bubbles or other separated domains (Fig. 6.11c). This suggests that the formation of bubbles can be reasonably attributed to the phase separation that occurs during the crystallization of the PB-1 component. For all the examined compositions, the above results indicate compatibility between the three components before PB-1 crystallization. The demixing phenomena observed after PB-1 crystallization were dependent on blend composition. The absence of separated domains observed by

142

Polyolefin Blends

Figure 6.11 SEM micrographs of iPP/PB-1/HOCP blends: (a) 50/10/40, crystallized from the melt at room temperature; (b) 50/30/20, crystallized from the melt at room temperature; (c) 50/30/20, crystallized under isothermal condition up to complete crystallization of the iPP component and eventually quenched in liquid nitrogen. (From Reference 8 with permission from John Wiley & Son, Inc.)

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

143

optical and electronic microscopy techniques suggests that before PB-1 crystallization, the PB-1 and HOCP components are incorporated in the interlamellar or interfibrillar zones of iPP spherulites. The spherulite dimension, at constant Tc , increases with increasing concentration of noncrystallizable component. The spherulite radius R increases linearly with crystallization time for pure iPP and iPP/PB-1/HOCP blends for all Tc s investigated. For all samples, the isothermal radial growth rate, G ¼ dR=dt, calculated at different Tc , is reported in Table 6.11. With the increase in the Tc , the G values appear to decrease for all investigated compositions. The blends prepared with the same fraction of iPP show G values that decrease with increasing of HOCP fraction at constant Tc value.

6.3.3 Glass Transition Temperature The ternary blends when crystallized at room temperature from the melt showed two Tg values (Table 6.12) as observed for other ternary blends (64). These Tg values correspond to the occurrence of PP-rich and PB-1-rich phases where the whole HOCP component is dissolved. In fact, no change of moduli, E0 and E00 , was observed in the zone where the Tg of HOCP should have appeared. Figure 6.12 reports the DMTA scan of the 50/30/20 blend quenched in liquid nitrogen after iPP crystallization from the melt (a) and crystallized at room temperature from the melt (b). In the first case, a single Tg value is shown (16 C) to be rather close to the value calculated by Fox equation (13 C). The same behavior has been observed for all the other ternary blends. The experimental Tg values were compared with the Tg calculated by the Fox equation 34: 1=TgðblendÞ ¼ ðWPP =TgPP Þ þ ðWPB =TgPB Þ þ ðWHOCP =TgHOCP Þ

ð6:16Þ

where TgPP , TgPB, TgHOCP , WPP , WPB, WHOCP are, respectively, the glass transition temperatures and the weight fraction of the three components. The weight fractions Wi have been calculated on the basis of iPP crystallization extent obtained by WAXD. The detection of a single glass transition is in good agreement with the theoretical values obtained through Fox equation (Table 6.13) suggesting that iPP, PB-1, and HOCP are miscible in the amorphous phase of the blend. The slight difference between the experimental and calculated Tg values could be attributed to minor demixing effects due to an in sufficient quenching. In the light of this fact, the shoulder at the left side of the main peak in Fig. 6.12a, observed only for some PB-rich blends, can be justified.

6.3.4 Nonisothermal Crystallization and Melting Behavior The ternary blends cooled from the melt during DSC measurements showed two exothermic peaks due to the iPP and PB crystallization, respectively. A decrease in the Tc values for both polyolefins was observed with the increasing of the HOCP fraction in the ternary blends (Table 6.14). After crystallization, the ternary blends

144

118 120 122 123 124 125 126 127 128 129 130 131 132 133 134

Tc , C

0.0865

0.0743 0.0539

0.1163

0.0926 0.0844 0.0627

0.2483

80/10/10

0.1367

100/0/0

0.0599 0.0546

0.0971

0.1289

0.2436

70/20/10

0.0594

0.1555 0.1333 0.1177 0.0938

70/10/20

0.1485 0.1265 0.0989 0.0817 0.0638

60/30/10

0.2262 0.1739 0.1448 0.1271 0.0889

60/20/20

0.1316 0.1192 0.0957 0.0718

0.1968

60/10/30

0.2599 0.1432 0.1346 0.0983 0.0744

50/40/10

0.0768

0.1335

0.2826 0.2381 0.1972

50/30/20

0.0885

0.2345 0.1909 0.1671 0.1355

0.2678 0.1863 0.1341 0.1133 0.0866

50/20/30 50/10/40

Spherulite Isothermal Growth Rate, G (mm s1), at Different Crystallization Temperature, Tc , of the iPP/PB-1/HOCP. Ternary

0.1894

Table 6.11 Blends.

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

145

Table 6.12 Values of Glass Transition Temperature, Tg , of the iPP/PB-1/HOCP Ternary Blends Crystallized at Room Temperature from the Melt. iPP/PB-1/HOCPa

iPP/PB-1/HOCPb

Tg , Cc

80/10/10 70/20/10 70/10/20 60/30/10 60/20/20 60/10/30 50/40/10 50/30/20 50/20/30 50/10/40

65.6/19.9/14.5 58.2/27.2/14.6 56.5/14.2/29.3 49.5/36.7/13.8 47.0/25.5/27.5 48.0/13.0/39.0 43.0/44.4/12.6 41.4/33.5/25.1 40.3/23.0/36.7 39.9/11.7/48.4

17 13 16 7 11 14 4 2 6 8

Tg , Cd 20 17 29 11 25 37 10 25 41 45

a

Nominal composition.

b

Composition corrected on the basis of the iPP and PB-1 crystallization extent determined by WAXD.

c

PB-1-rich phase.

d

IPP-rich phase.

were heated to 200 C at 20 C min1. The DSC scans relative to the ternary blends prepared with 60% of iPP are reported in Fig. 6.13. A part from the melting peak of the iPP component, every thermogram exhibits two endothermic peaks centering at 87–91 and 108–110 C, respectively, and an exothermic peak centering at 91–100 C. The two endothermic peaks correspond to the melting of PB crystallized in I0 and II forms, respectively. According to Boor and Youngman (50), we define I0 as the crystalline form of PB that shows the X-ray profile of form I (see below), but it melts at an appreciably lower temperature than that (130–140 C) of the material spontaneously deriving from form II. The melting of form I0 is immediately followed by the recrystallization of PB in form II and its melting (including the crystalline form II possibly present in the blend at the origin). Generally, the melting point of the two forms decreased with the increase in the HOCP fraction, for blends prepared with a constant content of PB (Table 6.14). The tendency of PB to crystallize in form I preferably than in form II resulted to be strongly influenced by blend composition. The melting point of PP slightly decreased with the decrease in its fraction in the blend or with the increase in the HOCP content.

6.3.5 Isothermal Bulk Crystallization Kinetics of Isotactic Polypropylene Component The isotherms of crystallization of pure iPP and iPP in the ternary blends showed that the overall crystallization rate constant in most cases decreases with the decrease in the amount of iPP in the blend and, with a constant fraction of iPP, increases when the fraction of HOCP decreases (Table 6.15). For all the examined samples, values of Avrami exponent, n, close to 3 have been obtained, suggesting a three-dimensional growth of crystalline unit, developed by heterogeneous nucleation.

146

Polyolefin Blends

Figure 6.12 DMTA analysis of pure iPP, pure PB-1, and iPP/PB-1/HOCP 50/30/20 blends after iPP crystallization from the melt and quenching (a) and after iPP and PB-1 crystallization from the melt at room temperature (b). (From Reference 7 with permission from John Wiley & Son, Inc.)

6.3.6 Melting Behavior of the Isotactic Polypropylene Component The observed melting temperature of blended PP, Tm , appeared to linearly increase with its isothermal crystallization temperature. The experimental data could be fitted by the Hoffmann equation (see Eq. 6.6). The extrapolated Tm values were influenced

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

147

Table 6.13 Calculated and Experimental Glass Transition Temperatures, Tg , of the iPP/ PB-1/HOCP Ternary Blends Quenched After iPP Crystallization from the Melt. Tg , C iPP/PB-1/HOCPa

iPP/PB-1/HOCPb

100/0/0 0/100/0 0/0/100 80/10/10 70/20/10 70/10/20 60/30/10 60/20/20 60/10/30 50/40/10 50/30/20 50/20/30 50/10/40

100/0/0 0/100/0 0/0/100 65.3/17.3/17.4 57.4/28.3/14.3 56.2/14.6/29.2 47.3/39.5/13.2 46.2/26.9/26.9 47.8/13.0/39.2 40.0/48.0/12.0 39.0/36.5/24.5 39.5/24.2/36.3 40.0/12.0/48.0

Experimental

Calculated

8 15 82 15 13 22 9 19 27 8 16 26 40

a

Nominal composition.

b

Composition corrected on the basis of the iPP crystallization extent determined by WAXD.

14 9 22 6 17 29 3 13 24 36

by the blend composition (Table 6.16), while the value of 1/g were close to each other (1=g 0:34). The volume blend composition in Table 6.16 have been calculated at the various Tm on the basis of reported temperature of the molar volumes of the considered components [48]. Assuming the absence of annealing and morphological effects [41], the values of Tm have been related to the melting temperature of the pure ) by the following equation (47): component (Tm;0 þ Bðn2 =DH2 ÞTm;0 ð1 f2 Þ2 Tm ¼ Tm;0

ð6:17Þ

where DH2 =n2 is the heat of fusion per unit volume of the pure crystal, f2 is the volume fraction of PP in the blend, and B is the parameter describing the enthalpic interaction between PP and the diluent fraction consisting of PB and HOCP. In Fig. 6.14, the B values are reported versus the HOCP volume fraction on the total volume of the diluent phase. For all the blend composition considered, negative value of B were obtained that decreased with the increase in the HOCP fraction in the diluent phase, thus indicating that such an increase leads to a higher stability of the blend.

6.3.7 Supermolecular Structure In Fig. 6.15 the WAXD profiles of two ternary blends containing 60% of iPP are reported. The X-ray diffusion due to the amorphous phase increased with the increase in the HOCP fraction in the blend. The crystalline index of the blend, vb , decreased with the increase in the HOCP fraction (Table 6.17). The diffractograms of the blends

148

Tc PP Tc PB Tm PP Tm PB I Tm PB II

107 58 166 94 111

103 55 164 90 110

107 59 165 91 111

99 50 162 87 108

103 54 162 88 108

107 64 164 91 110

95 44 160 82 104

101 49 161 84 105

102 57 163 88 108

60/30/10 50/10/40 50/20/30 50/30/20

Crystallization and Melting Temperatures ( C) of iPP/PB-1/HOCP Ternary Blends.

iPP/PB-1/HOCP 80/10/10 70/10/20 70/20/10 60/10/30 60/20/20

Table 6.14

106 65 163 88 109

50/40/10

107 — 170 — —

— 57 — — 114

Pure PP Pure PB

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

149

Figure 6.13 DSC scans for the melting of ternary blends prepared with 60% of iPP. The samples were heated from 20 to 200 C with a constant rate of 20 C min1. (From Reference 8 with permission from John Wiley & Son, Inc.)

showed that iPP crystallizes in the a-form for the whole range of the compositions analyzed, while the crystalline forms of PB-1 were strongly influenced by blend composition. In Fig. 6.16, the WAXD profiles of blends containing different amounts of PB-1 (and a constant fraction of iPP) show different diffraction intensities relative to the crystalline forms I0 and II. The contribution to the total crystallinity of the blend, vb , of the three different crystalline phases (a-form of iPP, forms I0 and II of PB-1) possibly present in the ternary blends were calculated by the following equations: vPB;I0 ¼ ½IPB;I0 a=ðIPB;I0 a þ IPB;II b þ IPP;a cÞvb

ð6:18Þ

vPB;II ¼ ½IPB;II b=ðIPB;I0 a þ IPB;II b þ IPP;a cÞvb

ð6:19Þ

vPP;a ¼ vb ðvPB;I0 þ vPB;II Þ

ð6:20Þ

where vPB;I0 and vPB;II are the weight fraction of forms I0 and II of PB-1, respectively, and vPP;a is the weight fraction of the crystalline iPP. IPB;I0 and IPB;II are the

150

116 117 118 119 120 121 122 124 125 126 127 128 129 130

Tc , C

28.100

1.570

0.559

0.128

0.035

5.750 3.070

0.904

0.155

0.060

80/10/10

30.000

100/0/0

0.020

0.073

0.436

3.310 1.530

20.000

70/20/10

0.054 0.033 0.014

0.831 0.266

2.900

9.340

70/10/20

0.043

0.175

0.555

60/30/10

0.053 0.039

0.729 0.220

2.71

42.900

60/20/20

0.380 0.340 0.091 0.032 0.016

1.410

4.600

60/10/30

Table 6.15 Crystallization Rate Constant, 108 Kn (s3), of iPP/PB-1/HOCP Ternary Blends.

0.220 0.109 0.042 0.016

1.570 0.491

50/40/10

0.028

0.429 0.133

2.460

8.380

21.600

50/30/20

0.022

0.156

0.592 0.057 0.036 0.019

0.845 0.263 0.182

50/20/30 50/10/40

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

151

Table 6.16 Values of Equilibrium Melting Temperature, Tm , of iPP in iPP/PB-1/HOCP Ternary Blends. iPP/PB-1/HOCPa

iPP/PB-1/HOCPb

Tm , C

100/0/0 80/10/10 70/20/10 70/10/20 60/30/10 60/20/20 60/10/30 50/40/10 50/30/20 50/20/30 50/10/40

1/0/0 0.822/0.097/0.081 0.723/0.195/0.082 0.735/0.099/0.0166 0.624/0.294/0.082 0.634/0.199/0.167 0.644/0.101/0.255 0.523/0.394/0.083 0.531/0.301/0.168 0.540/0.203/0.257 0.549/0.103/0.348

187.9 187.0 186.3 185.2 185.7 183.7 183.0 185.5 182.9 180.1 179.0

a

Weight percent composition.

b

Volume fraction composition.

diffraction intensities of the peaks at 9.8 and at 11.8 2u relative to the PB-1 crystallized in forms I and II, respectively, and IPP;a is the diffraction intensity of the peak at 14.1 2u relative to the (110) crystallographic plane of iPP crystallized in a-form; a, b, and c are the proportionality factors between the peaks calculated from the diffraction data of pure PB-1 and pure iPP; the form I of pure PB-1 was obtained from the total spontaneous evolution at 20 C of the pure PB-1 crystallized in form II. The influence of blend composition on the PB-1 crystalline forms was evidenced by calculating the fI0 index as follows: fI0 ¼ vPB;I0 =ðvPB;I0 þ vPB;II Þ

ð6:21Þ

Figure 6.14 Dependence of the interaction parameter B on the composition of PB-1/HOCP system.

152

Polyolefin Blends

Figure 6.15 WAXD profiles of iPP/PB-1/HOCP blends: (a) 60/20/20; (b) 60/10/30. (From Reference 8 with permission from John Wiley & Son, Inc.)

The values of fI0 reported in Table 6.17 show that the crystallization in form I0 is the favorite for blends prepared with small fractions of PB-1; by increasing the PB-1 content in the blends, the fraction crystallized in form II increased. These results are in good agreement with the trend observed for the thermal data obtained by DSC. PB-1/HOCP binary blends containing up to 30% of HOCP showed that PB-1 Table 6.17 Total Blend Crystallinity, vb , iPP Crystalline Fraction, vPP;a , and Form I0 Fraction of the Crystallized PB, fI0 , of iPP/PB-1/HOCP Ternary Blends. iPP/PB-1/HOCP

vb

vPP;a

fI0

100/0/0 0/100/0 80/10/10 70/10/20 70/20/10 60/10/30 60/20/20 60/30/10 50/10/40 50/20/30 50/30/20 50/40/10

— — 0.56 0.48 0.50 0.42 0.50 0.51 0.37 0.41 0.47 0.48

0.66 0.43a 0.53 0.45 0.43 0.39 0.43 0.40 0.34 0.35 0.36 0.34

— 0.00 1.00 1.00 0.82 1.00 0.91 0.23 0.88 0.41 0.32 0.04

a

Crystallinity index of PB crystallized in form II.

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

153

Figure 6.16 WAXD profile of ternary blends containing 50% of iPP. (From Reference 8 with permission from John Wiley & Son, Inc.)

crystallizes from the melt completely in form II [4]. All the examined ternary blends showed the presence of PB in form I0 as a consequence of the different crystallization conditions of PB-1 occurring after iPP crystallization. The results reported in Table 6.17 show a quite constant value of vPP;a for blends prepared with the same content of iPP.

6.4 CONCLUSIONS The morphology, the crystallization kinetics, and the melting behavior of iPP/PB-1 blends have been examined as a function of the composition in a temperature range where only the crystallization of iPP can occur. It has been found that both the spherulite growth rate and the overall crystallization rate of iPP in the blends are depressed by the presence of the molten PB-1 component. Moreover, a significant decrease in the equilibrium melting temperature of iPP has been observed for PB-1 contents higher than 30 wt%. iPP and PB-1 can form a homogeneous mixture in the melt at the adopted temperature. The investigations have shown that the spherulitic growth rate, the overall crystallization rate, and the melting temperature of iPP are depressed by the presence of HOCP. Together with the detection of a single glass transition, the results suggest that the two polymers are miscible in the melt. In agreement with the above conclusion, the supermolecular characterization of the iPP/HOCP blends suggested that in the crystallized blends the HOCP molecules are rejected in interlamellar regions of iPP spherulites where they form a homogeneous mixture with uncrystallized iPP molecules.

154

Polyolefin Blends

The spherulitic growth rate, the overall crystallization rate, and the melting temperature of PB-l are depressed by the presence of HOCP. The verified HOCP interferences on the kinetics of PB-l crystal transformation from form II to form I indicate that in the crystallized mixtures the HOC molecules are rejected in interlamellar and/or interfibrillar regions of PB-1 spherulites where, dependening on the blend composition, they can form a homogeneous mixture with uncrystallized PB-l molecules or a conjugated amorphous phase, one rich in PB-1 component and the other rich in HOCP component. The existence of a contribution of morphological effects on the melting behavior of PB-l in the blends is evidenced by the variation of the stability parameter l/g with composition. The interaction parameter x12 is also subjected to morphological effects. The results show that HOCP interacts stronger with PB-1 than with iPP. In spite of this difference, the possibility of employing HOCP as a second compatible component to prepare ternary blends has been demonstrated. The miscibility in the melt of iPP, PB-1, and HOCP has been evidenced by the presence of a single glass transition and by the interference of PB-1 and HOCP on spherulite growth rate, overall crystallization rate, and melting equilibrium temperature of iPP. The addition of the HOCP component to iPP and PB-1 has been found to increase the stability of the blends. The demixing phenomena in the ternary blends occur during and/or after the crystallization of the PB-1component. The possibility of preparing ternary blends where the PB-1 component directly crystallizes in form I0 , could offer new opportunities about their application.

NOMENCLATURE x12 DH DMTA DSC 0 G HOCP iPP K Kn lc la L Mw n N PB-1 POM R Rs

Polymer–polymer interaction parameter Enthalpy of fusion Dynamical mechanical thermal analysis Differential scanning calorimetry Spherulite isothermal radial growth rate Hydrogenated oligo(cyclopentadiene) Isotactic polypropylene Boltzmann constant Overall kinetic rate constant Lamellar thickness Thickness of amorphous interlamellar region Long period Molecular weight Avrami parameter Nucleation density Isotactic poly(l-butene) Polarized optical microscopy Universal gas constant Spherulite radius

Chapter 6 Thermal and Structural Characterization of Binary and Ternary Blends

s se SAXS SEM t t0:5 Tg Tm Tm WAXD

155

Lateral surface free energy Folding surface free energy Small-angle X-ray scattering Scanning electron microscopy Time Half-time of crystallization Glass transition temperature Observed melting temperature Equilibrium melting temperature Wide-angle X-ray diffraction

REFERENCES 1. E. Martuscelli, C. Silvestre, M. Canetti, C. de Lalla, A. Bonfatti, and A. Seves, Makromol. Chem., 190, 2615 (1989). 2. E. Martuscelli, M. Canetti, and A. Seves, Polymer, 30, 304 (1989). 3. E. Martuscelli, M. Canetti, A. M. Bonfatti, and A. Seves, Polymer, 32, 641 (1991). 4. M. Canetti, M. Romano`, P. Sadocco, and A. Seves, Makromol. Chem., 191, 1589 (1990). 5. A. M. Bonfatti, M. Canetti, P. Sadocco, A. Seves, and E. Martuscelli, Polymer, 34, 990 (1993). 6. M. Canetti, A. M. Bonfatti, P. Sadocco, A. Seves, and M. Pracella, Polym. Netw. Blends, 3, 83 (1993). 7. P. L. Beltrame, A. Castelli, G. Munaretto, M. Canetti, and A. Seves, J. Appl. Polym. Sci., 65, 1659 (1997). 8. M. Canetti, A. Seves, L. Bergamasco, G. Munaretto, and P. L. Beltrame, J. Appl. Polym. Sci., 68, 1877 (1998). 9. S. Cimmino, P. Guardata, E. Martuscelli, and C. Silvestre, Polymer, 32, 3299 (1991). 10. Z. Bartczak and E. Martuscelli, Polymer, 38, 4139 (1997). 11. B. Marcandalli, G. Testa, A. Seves, and E. Martuscelli, Polymer, 32, 3376 (1991). 12. M. Canetti, A. M. Bonfatti, A. Siciliano, and A. Seves, Angew. Macromol. Chem., 197, 59 (1992). 13. S. Cimmino, M. L. Di Lorenzo, and C. Silvestre, Thermochim. Acta, 321, 99 (1998). 14. S. Cimmino, M. L. Di Lorenzo, E. Di Pace, and C. Silvestre, J. Appl. Polym. Sci., 67, 1369 (1998). 15. C. Silvestre, S. Cimmino, and M. L. Di Lorenzo, J. Appl. Polym. Sci., 71, 1677 (1999). 16. A. Piloz, J. Decroix, and J. F. May, Ang. Macromol. Chem, 54, 77 (1976). 17. A. Sjegmann, J. Appl. Polym. Sci., 24, 2333 (1979). 18. A. Sjegmann, J. Appl. Polym. Sci., 27, 1053 (1982). 19. R. M. Gohil and J. Peterman, J. Macromol. Sci., Phys., B18, 217 (1980). 20. P. L. Beltrame, A. Castelli, M. Di Pasquantonio, M. Canetti, and A. Seves, J. Appl. Polym. Sci., 60, 579 (1996). 21. S. N. Sathe, S. Devi, G. S. S. Rao, and K. V. Rao, J. Appl. Polym. Sci., 61, 97 (1996). 22. Y. Seo, B. Kim, and K. Kim, Polymer, 40, 4483 (1999). 23. A. N. Wilkinson, M. L. Clemens, and V. M. Harding, Polymer, 45, 5239 (2004). 24. R. Krache, D. Benachour, and P. Po¨tschke, J. Appl. Polym. Sci., 94, 1976 (2004). 25. M. M. Dumoulin, C. Farha, and L. A. Utracki, Polym. Eng. Sci., 24, 1319 (1984). 26. J. Kolaı´k, J. Velek, G. L. Agrawal, and I. Fortelny´, Polym. Compd., 7, 472 (1986). 27. M. A. Lo´pez Manchado, L. Torre, and J. M. Kenny, J. Appl. Polym. Sci., 81, 1063 (2001).

156

Polyolefin Blends

28. M. Moffitt, Y. Rharbi, J. Tong, J. P. S. Farhina, H. Li, M. A. Winnik, and H. Zahalka, J. Polym. Sci. B Polym. Phys., 41, 637 (2003). 29. N. Kukaleva, G. P. Simon, and E. Kosior, Polym. Eng. Sci., 43, 431 (2003). 30. A. K. Gupta and S. N. Purwar, J. Appl. Polym. Sci., 30, 1777 (1985). 31. A. K. Gupta and S. N. Purwar, J. Appl. Polym. Sci., 30, 1799 (1985). 32. A. K. Gupta, A. K. Jain, and S. N. Maiti, J. Appl. Polym. Sci., 38, 1699 (1989). 33. A. K. Gupta, A. K. Jain, B. K. Ratnam, and S. N. Maiti, J. Appl. Polym. Sci., 39, 515 (1990). 34. T. G. Fox, Bull. Am. Phys. Soc., 2, 123 (1956). 35. H. D. Keith and F. J. Padden, J. Appl. Phys., 35, 1270 (1975). 36. R. S. Stein, F. B. Khambatta, F. P. Wamer, T. Russel, A. Escala, and E. Balizer, J. Polym. Sci. Polym. Symp., 63, 313 (1987). 37. Z. Bartczak, A. Galeski, and M. Pracella, J. Appl. Polym. Sci., 54, 1513 (1994). 38. M. Avrami, J. Chem. Phys., 7, 1103 (1939). 39. L. Mandelkern, Crystallization of Polymers, McGraw Hill, New York, 1964. 40. J. D. Hoffman, Polymer, 24, 3 (1982). 41. J. Boon and J. M. Azcue, J. Polym. Sci. A-2, 6, 885 (1968). 42. S. Cimmino, E. Martuscelli, C. Silvestre, M. Canetti, C. de Lalla, and A. Seves, J. Polym. Sci. Polym. Phys. Ed., 27, 1781 (1989). 43. J. I. Lauritzen Jr., J. Appl. Phys., 44, 4353 (1973). 44. E. Martuscelli and G. B. Demma, in: Polymeric Blends: Processing, Morphology and Properties, E. Martuscel1i, R. Palumbo, and M. Kryszewski (eds.), Plenum Press, New York, 1980. 45. J. D. Hoffman and J. J. Weeks, J. Chem. Phys., 37, 171 (1962). 46. T. Nishi and T. Wang, Macromolecules, 8, 909 (1975). 47. P. J. Flory, Principles of Polymer Chemistry, Cornell University Press, Ithaca, 1953. 48. D. W. Van Krevelen, Properties of Polymers, Elsevier, Amsterdam, 1976. 49. J. Boor Jr. and J. C. Mitchell, J. Polym. Sci., 62, S70 (1962). 50. J. Boor Jr. and E. A. Youngman, J. Polym. Sci., Polym. Lett., 2, 903 (1964). 51. K. Hong and I. E. Spruiell, J. Appl. Polym. Sci., 30, 3163 (1985). 52. J. Boor Jr. and J. C. Mitchell, J. Polym. Sci. A, 1, 59 (1963). 53. T. Asada, I. Sasada, and S. Onogi, Polym. J., 3, 350 (1972). 54. A. Tanaka, N. Sugimoto, T. Asada, and S. Onogi, Polym. J., 5, 529 (1975). 55. C. Nakafuku and T. Miyaki, Polymer, 24, 141 (1983). 56. B. Wunderlich, Macromolecular Physics, Academic Press, New York, 1973. 57. R. Zanetti, P. Manaresi, and O. C. Buzzoni, Chim. Ind. (Milan), 43, 735 (1961). 58. J. Powers, J. D. Hoffman, J. J. Weeks, and F. A. Quinn Jr., J. Res. Natl. Bur. Stand. Sect. A, 69, 335 (1965). 59. T. Oda, M. Maeda, S. Hibi, and S. Watanabe, Kobunshi Ronbunshu, 31, 129 (1974). 60. P. H. Geil, Y. C. Yang, K. W. Chan, C. C. Hsu, and A. Agarwal, SPE ANTEC Conference Proceedings, Vol. XXIX, USA, 1983, p. 404. 61. T. Oda, M. Maeda, S. Hibi, and S. Watanabe, Kobunshi Ronbunshu (Engl. Ed.), 3 (2), 1249 (1975). 62. O. Glatter and O Kratky, Small Angle X-Ray Scattering, Academic Press, London, 1982. 63. S. Cimmino, E. Di Pace, F. E. Karasz, E. Martuscelli, and C. Silvestre, Polymer, 34, 972 (1993). 64. P. M. Remiro and J. Nazabal, J. Appl. Polym. Sci., 42, 1475 (1991).

Chapter

7

Morphological Phase Diagrams of Blends of Polypropylene Isomers with Poly(Ethylene–Octene) Copolymer Wirunya Keawwattana,1 Rushikesh A. Matkar,2 and Thein Kyu2

7.1 INTRODUCTION Polymer crystal morphology has attracted immense interest because of a wide variety of morphologies ranging from spherulites with intricate textures to single crystals (1–18). It is well documented that polymer single crystals are generally grown from very dilute solutions, whereas more complicated hierarchical crystalline morphologies such as axialite, hedrite, and spherulite emerge from concentrated solutions or the melt (1–6). Recently, it became apparent that various single crystals can be grown from the melt. The existence of supramolecular structures is not unique to polymers, but such organizations have been found in a large variety of small molecule systems such as inorganic substances, bioorganisms, and metals. In general, spherulite has been characterized as a rounded aggregate of radiating lamellar crystals with a fibrous appearance, which originates from a nucleus such as particle of contaminant, catalyst residue, or fluctuation in density created by chance. These structures often grow through stages—first, formation of lamellar plates or

1

Department of Chemistry, Faculty of Science, Kasetsart University, Bangkok 10903, Thailand.

2

Department of Polymer Engineering, University of Akron, Akron, OH 44325, USA

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

157

158

Polyolefin Blends

needles; second, lamellar bundles via stacking; third, lamellar aggregation leading to sheaflike structure; and finally, the spherulites with lamellar branching, the diameter of which may range from submicrons to several hundred microns (1–11). In the case of polymers, spherulites may be conveniently viewed under a polarized optical microscope, which consist of a large number of lamellae growing radially outward from a primary nucleus at the core. Often some lamellae may be twisted about their long axes, which result in the concentric bands or spiral structures (1, 3, 6, 10). The spherical shape arises usually due to side branching and splaying of microstructures, while such a structure in the initial to intermediate stages may not be spherical, but rather resembles a sheaf-like morphology (11–13). In the polymer spherulitic growth, the melt crystallized lamellae often stack into bundles from a single nucleation site and grow radially outward until they impinge on the neighboring lamellar bundles growing from other nucleating sites. The lamellar stacks are laterally constrained to some extent such that they form a ribbonlike structure. The lateral constraint instability is perceived to develop from impurities at the growing lamellar sides. Impurities include a number of things such as dirt, chain segments of improper tacticity, branched segments, end groups, and other amorphous components that cannot crystallize at the temperature of crystallization. However, some of these amorphous chains may crystallize at a lower temperature to facilitate secondary crystallization occurring in the interlamellar or interspherulitic regions, creating branch points that allow the spherulite to grow into a three-dimensional object. Although the main lamellae grow in some preferential crystallographic axis, for example, b-axis in polyolefin, the side branching in a spherulite is noncrystallographic unlike the dendrites in small molecule systems. In low molecular weight materials such as snowflakes or ice crystallites, branching predominantly occurs along low index crystallographic planes. In polymer spherulites with the lamellae radiating from the core, the lamellar orientation is random in a global sense, and thus there is little or no relationship between the crystallographic planes and the direction of branching. The release of latent heat gives a temperature distribution along the crystal–melt interface, resulting in a nonuniform boundary having density fluctuations, some of which eventually serve as likely sites for nucleation of lamellar branching. The crystal structure of PP greatly depends on crystallization conditions such as supercooling and crystallization time as well as chain stereoregularity, molecular weight, and its distribution (14–30). The microstructure of polypropylene has a profound effect on its morphology and physical properties. Syndiotactic (sPP) and isotactic polypropylene (iPP) are crystalline thermoplastic, whereas pure atactic polypropylene (aPP) is amorphous (1). However, metallocene catalyzed aPP contains a low level of crystallinity, arising from the isotactic blocks (24–26). These polymers are often known as elastomeric polypropylene (ePP), which has the properties of a thermoplastic elastomer. The properties of such a material were first postulated by Natta (31) in terms of a stereoblock structure. Although not as stable as their thermoset counterparts in regard to solvent and heat resistance, for example, softening point, thermoplastic elastomers are a

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

159

rapidly growing sector of the elastomer market due to their low production costs and ease of processing (23). The elastomeric polypropylene materials studied in this chapter are from a class of thermoplastic elastomers since they possess the physical properties of elastomers along with the processing characteristics of thermoplastics. These materials are characterized by a low degree of crystallinity (23–26), where the crystalline regions dispersed in the amorphous matrix essentially provide physical cross-links to the amorphous elastomeric segments of the chain (19, 20). The size and distribution of these crystalline regions in the amorphous matrix thus have important influences on the mechanical properties. Blends of two semicrystalline polymers (PP and POE) are complex due to possible involvement of the crystal–amorphous, amorphous–crystal, and crystal– crystal interactions in addition to the conventional amorphous–amorphous chain interaction of the binary amorphous blends. However, it offers interesting possibilities for probing the relation between complex phase behavior and structural phase transitions. The purpose of this chapter is to demonstrate the morphological development as a result of crystallization and to elucidate the effect of supercooling on the crystalline morphology in blends of POE and polypropylene isomers. Moreover, it is of interest to determine the emerged morphology in relation to melting temperature–composition phase diagram of the system. Various morphological structures at different points of the phase diagram were investigated under various isothermal supercooling conditions for each POE/PP isomeric blend. The development of sectorization, ripple formation, and amorphous concentration profiles have been simulated based on the phase field free energy approach for crystal solidification.

7.2 BLENDS OF sPP/POE In view of the low melting temperature of sPP and concomitant inferior mechanical performance relative to iPP, the morphology of sPP thus far has received limited attention. Another drawback is the development of poorly defined structures during crystallization, such as incomplete (or imperfect) spherulitic structures. Consequently, most studies have been directed to the morphological characterization of curved single crystals that develop predominantly during solution crystallization with or without sectorization (14–19). Although the sectorization is apparently similar to the truncated single crystal of solution-grown polyethylene (PE), the preferential growth occurs in the (020) direction of the sPP single crystals as opposed to the (200) direction of the PE crystals. The phenomenon of the sectorization is believed to be a consequence of different crystallization habits of the individual crystal planes, reflecting the difference in the fold energies of the chains in the (200) and (020) planes (17,18). In the past decade, the morphological characterization of sPP received renewed interest when Lovinger and coworkers demonstrated that sPP single crystals can be grown from the melt (27,28). At a shallow supercooling, lath-shaped sectorized

160

Polyolefin Blends

single crystals emerge with transverse undulations preferentially at the lateral region of the sectors. The origin of these modulated ridges was not clearly understood, but it was speculated to be due to thermal contraction during cooling. Another plausible account was due to surface stress relief caused by the uniaxial contraction of the sPP crystal (29). Interestingly, the periodic patterns, similar to those of sPP, were observed over 40 years ago in solution-grown truncated single crystals of polyethylene (1,13), which was attributed to the collapse of the nonplanar pyramid-shaped single crystal during solvent evaporation. Regardless of the differing opinions, these studies invariably suggest that mechanical deformation might play a role in the morphological development during melt crystallization that supposedly occurs under quiescent crystallization conditions. An immediate question is whether the same single-crystal structure would develop if crystallization were to take place in the presence of a polymer diluent that is miscible with sPP. It is the source of our motivation for (i) investigation of melting/crystallization behavior of sPP in the blends with poly(ethylene–octene) copolymer and (ii) establishment of the morphological phase diagram of sPP/POE blends. In the present study, syndiotactic polypropylene was supplied by FINA Chemical and Oil Company, having the reported weight-average and number-average molecular weights of sPP of 174,000 and 74;700 g mol1 , respectively, with a polydispersity (Mw/Mn) of 2.3 and 92.6% content of syndiotacticity. Poly(ethylene– octene) copolymer was provided by the Dow Chemical Company (Mw ¼ 41; 800 g mol1 , Mw/Mn ¼ 2.26, and octene content ¼ 10% by weight). Blends of sPP/POE were prepared by dissolving these polyolefins in xylene at about 100 C (at a polymer concentration of 2–3 wt%) and stirred thoroughly for about 1 day. Film specimens were prepared by solvent casting on optical microscope glass slides. To ensure complete removal of the solvent, the glass slides were immersed in distilled water (nonsolvent) for 1 h and then dried at the ambient temperature. Subsequently, the samples were further dried in a vacuum oven at room temperature for another 2 days. All samples were heated to 160 C for 10 min to provide the blend with a thermal history similar to the melt-mixed samples and to further remove any residual solvent. The thickness of the blend films used for microscopy was approximately 10 mm (31). The morphology of the blends was studied using a Nikon Optiphot 2-POL optical microscope (OM). The halogen light source was operated at 12 V and 100 W. The sample heating chamber (Mettler FP82 HT), connected to programmable temperature controller (Mettler Toledo FP90 central processor), was used to control the sample temperature. A real-time record of the morphology evolution as a function of temperature was made possible using a color video camera (Sony, HyperHAD, digital), interfaced to a computer. Atomic force microscopy (AFM) experiments were performed using a Digital Instruments NanoScope IIIA multimode scanning probe microscope (SPM) equipped with a heating stage. Crystallization kinetic experiments of various sPP/POE blends were performed using differential scanning calorimetry (DSC) under isothermal crystallization conditions. The samples were premelted at 180 C for 10 min and then rapidly quenched to desired crystallization temperatures.

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

161

7.2.1 Thermal Characterization and Morphological Phase Diagrams: Undulated Lamella, Sheaf, and Spherulite Before discussing the details of the morphology development in sPP/POE blend, it is worthwhile to establish the phase diagram as revealed by DSC and light scattering (LS). Figure 7.1a shows a liquid–solid phase diagram established for blends of two semicrystalline polymers, that is, sPP/POE. The melting points (Tm) were obtained by measuring the DSC melting peak at various heating rates and subsequently extrapolating to zero heating rate. Although this peak temperature at zero rate does not represent a true equilibrium (i.e., not at Tm ¼ Tc ), it may be sufficient for describing the relative trends of the Tm values as a function of the blend concentrations. Hence, the zero rate crystal melting temperatures of sPP and POE are displayed as a function of volume fraction of the blend in the phase diagram. These blends show little or no depression of melting transition of sPP. It should be noted that POE also shows no change in the crystal melting peak. A lack of depression in the melting temperature of sPP coupled with the fact that there is no identifiable liquid–liquid phase separation suggests that the POE melt acts like a theta solvent to sPP melt, showing virtually no interaction with the sPP crystals. The phase diagram displays three distinct regions: isotropic melt (I), coexistence of crystal–isotropic melt (C1 þ I), and crystal–crystal (C1 þ C2 ). The crystal melting temperatures as obtained by light scattering and DSC indicate the same trend as the crystal–isotropic line by DSC. Figure 7.1a is shown together with the micrographs to indicate the positions of temperatures to which the samples were quenched from the isotropic melt of 160 C in reference to the phase diagram. Figure 7.1b–f represents the optical micrographs displaying the crystalline morphologies for the 10/90 sPP/POE blend after quenching from 160 C to several indicated crystallization temperatures. Usually the spherulitic structure develops at a deeper supercooling, whereas faceted single crystals emerge predominantly at shallow supercooling. In deep quenching, the emerged melt–crystal boundary is rough, and even small amplitude of the irregular interface is amplified by the latent heat liberated (32, 33). By virtue of the nonuniform thermal transport, the lamellae undergo directional growth through extensive tip splitting and side branching from the main lamellae, which eventually evolve into the seaweed, alternatively known as dense branching morphology (33). At a moderate supercooling, the lamellar structure seemingly grows predominantly along the long lamellar axes relative to the transverse direction because of the differential growth rates corresponding to the crystallographic axes. As can be witnessed in Fig. 7.1b–d, these individual lamellae are anisotropic and tend to stack against each other at the nucleating stage. These lamellar stacks continue to grow predominantly along the long axes because the lateral growth is hindered by the presence of the neighboring lamellae. When these stacked lamellae reach a certain length, the ones at the most outer sides tend to curve toward the outer free space. With continued growth, the multilayer crystals splay out progressively while additional lamellar side branching takes place from the branches, and

162

Polyolefin Blends

Figure 7.1 (a) Liquid–solid phase boundaries as obtained by means of differential scanning calorimetry (DSC) and light scattering (LS) for the sPP/POE blend displaying isotropic (I), coexistence of isotropic– crystal (I þ C1 ) and crystal–crystal (C1 þ C2 ) regions (C1 and C2 correspond to crystals of sPP and POE, respectively). (b–f) Optical micrographs displaying the dependence of supercooling on crystalline morphologies of the sPP/POE (10/90) blend after quenching from 160 C to several indicated temperatures.

eventually form the sheaflike structure. These lamellae were locally anisotropic; however, the overall structure becomes isotropic in a global sense with the progression of the spherulitic growth. It was found that at high growth temperatures (e.g., 125 C) the aggregates of lamellar crystals emerge, while at lower temperatures (i.e.,

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

163

115 and 110 C) these lamellae branch out eventually and form spherulites. That is to say the initial nuclei grow into rectangular shaped lamellae, reflecting the crystallographic axes of sPP crystals. These lamellae tend to stack into lamellar bundles that transform to the sheaflike structure before emerging to dense lamellar branching morphology. It appears that a temperature quench to a lower temperature (107 C), which is below the pseudo-equilibrium melting temperature of POE, the coexistence of both crystals can be observed. At a lower temperature, for example, 100 C, the crystallization of POE competes with that of sPP. Of particular interest is that the dark and white regions, which are assigned to be the interference of undulated structure (ripples), can be observed in sPP lamellae crystals at a temperature range of 110–125 C, forming the banded-like texture. Similar experiments have been conducted for other blend compositions. It is reasonable to infer that the supercooling has strongly affected the crystalline morphologies of sPP in the blends as shown in the morphological phase boundary (Fig. 7.2). As mentioned earlier in the 10/90 sPP/POE blend, the crystalline structure of sPP obtained by means of OM shows the dark and white regions all over the rectangular laths of sPP, leading to the observations of banded-like texture in blends isothermally crystallized at 115 C (Fig. 7.1c). The development of single crystal and their aggregates, including spherulitic structures, can be seen qualitatively by OM. However, the spatial resolution of OM and scattering techniques strongly restricts their application to characterizing polymer morphology. Indeed, the size of basic 135

Temperature, °C

130

S+A

S+A

A

A

120 S S+A

S+A

A

P

P

P

115 S+A A

A

P

P

P

P

110 S+A P

P

P

P

P

P

125 S

105 0

20

40

60

80

100

wt% sPP

Figure 7.2 Morphological phase boundary for the sPP/POE blends at several isothermal crystallization temperatures and compositions (S, A, and P correspond to single crystal, single-crystal aggregate, and spherulite of sPP, respectively).

164

Polyolefin Blends

structural elements in semicrystalline polymers (crystalline lamellae and lamellar bundles) falls in most cases below the diffraction limits of these techniques. In order to dynamically follow polymer crystallization with a better resolution beyond OM, AFM was utilized to mimic the intricate surface morphology of sPP single crystals. The temporal evolution of the sPP single crystals and the internal undulation cannot be observed in situ under AFM. However, knowing when undulation develops is important because it may be related to nucleation and chain folding at the very onset of crystallization, or to crystal thickening during annealing, or during quenching from a high temperature. Figure 7.3a–c represents the AFM images of crystals grown at various temperatures for the 10/90 sPP/POE blend. It is clearly seen that the sPP crystals obtained at 115 C show the multilayer aggregates of rectangular laths originating from a central nucleus and fan out, leading to incipient branching. In addition, all the rectangular laths of sPP show the existence of undulations over the entire crystals resulting in the banded-like texture observed under OM. A feature of great importance relating to these ripples (undulations) is their periodicity, which will be discussed later. Unlike the higher crystallization temperatures, the structure crystallized at 105 C simply shows the splaying and bending of sPP lamellae without any banded or corrugated texture.

Figure 7.3 AFM images showing the dependence of supercooling on the morphologies of sPP crystals in blends of sPP/POE (10/90) (a) at 125 C and (b) crystallized isothermally at 120 C for 3 h and taken at 110 C after reheating. AFM images in (a), (c), and (d) were carried out at room temperature after the samples were isothermally crystallized at a present temperature and subsequently cooled to room temperatures.

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

165

A logical question follows: Is the observed multilayer aggregate of rectangular laths with transverse undulations unique to sPP? Note that AFM images are captured at room temperature after the samples were isothermally crystallized at a preset temperature for a certain period of time and subsequently cooled to room temperature. Therefore, there is a possibility that the lamellar aggregates thus formed are POE crystals rather than sPP. To resolve this issue, crystal melting experiments were carried out by means of AFM. The melting temperature of POE is approximately 110 C. Hence, if the lamellar aggregates do not melt at this temperature (110 C), these crystals should be assigned to sPP. Figure 7.3d shows the AFM height images of the 10/90 sPP/POE undertaken at room temperature after it was isothermally crystallized at 120 C for almost 3 h and subsequently quenched to room temperature. After increasing the temperature back to 110 C, it was found that the multilayer aggregates still persisted, while the other spherulitic crystals, presumably those of POE, disappeared (Fig. 7.3d). Therefore, it can be confirmed that single-crystal aggregates are truly sPP crystals with ripples running across the fastest growth direction.

7.2.2 Growth of Single Crystals: Length, Width, and Periodicity Figure 7.4 depicts the time evolution of single crystals in the blend of 1/99 sPP/POE grown at 125 C, which is close to the melting temperature of sPP. It took more than 10 h for a nucleus to develop. As soon as nuclei (particles larger than a critical size) form, these crystals begin to grow to a visible size under microscopic investigations. Similar experiments were carried out at various temperatures, for example 120 and 115 C. As shown in Fig. 7.4e and f, isolated sPP single crystals develop at high temperatures (120 and 125 C), whereas aggregation of these single crystals can be seen at a lower temperature (115 C). It is of interest to measure the crystal size as a function of time from which the slope of this relationship yields the growth rate of a true single crystal (Fig. 7.5). It is noticed that the crystal size increases in both growth directions (i.e., in length and width) with increasing crystallization temperature or decreasing supercooling. The lamellar crystals continue growing with time for some periods and then level off, indicating that the crystallization rate, that is, the slope, gradually declines, which may be attributed to the depletion of crystallizable polymer chains or the impingement by neighboring single crystals. It should be emphasized that the initial nucleus is isotropic, but the emerging single crystal becomes anisotropic with elapsed time by virtue of the difference in the growth habits of sPP in which the directional growth along the b-axis (length) is faster than that along the a-axis (width), reflecting the differential fold energies along the two growth fronts (15,16). As seen in Fig. 7.5, the average length scales along the lamellar length (b-axis) are approximately four times larger than those along the width (a-axis). A slope of unity at the initial stage is shown to guide eyes. At a deep supercooling of 115 C, the slope is closer to 1 in both length and width directions, suggesting the linear growth habit of the lamellar single crystal, which is consistent with the prediction by

166

Polyolefin Blends

Figure 7.4 (a–d) Temporal evolution of sPP/POE (1/99) isothermally crystallized at 125 C and (e) morphological patterns of sPP/POE (1/99) isothermally crystallized at 120 C and (f) at 115 C.

Lauritzen and Hoffman (34). However, the slope departs from the linear growth, suggesting that the secondary crystallization must have taken place with continued elapsed time. Shallower quenches reveal a similar trend; however, the timescale covered here is too short to substantiate any claim. As expected, the growth rate in the b-axis direction is faster than that in the a-axis direction. In the neat sPP or high sPP contents, the melt crystallized lamellae certainly stack into bundles from a single nucleation site and grow until they impinge on the neighboring lamellae. Consequently, the growth along the transverse direction may be hampered, thereby showing a larger differential growth rate in the two directions. The present observation is much simpler relative to the spherulitic growth that the lamellar crystals are free to grow without any perturbance from the neighboring lamellar crystals. Moreover, the observed growth rate in the length and width directions of the isolated sPP lamellae in the continuum of POE melt (i.e., 1 or 2 wt%) is well defined as opposed to the literature reported for the growth of the

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

167

Single-crystal size (length in µm)

40 120°C

115°C

30

125°C

20 1

10

100

1000 Time, min (a)

8 7

Single-crystal size (width in µm)

6

115°C

5

120°C

4 1 125°C

3

2

1 100

1000 Time, min (b)

Figure 7.5 Plot of single-crystal size versus time of sPP/POE (1/99) at different crystallization temperatures: (a) length and (b) width plotted in a log–log scale. The slope of unity was provided to guide the eyes to judge any departure from the linear growth.

spherulites, which is complicated by lamellar branching as well as aggregation of neighboring lamellae. Another intriguing observation is that there is a movement of single crystals grown at a temperature slightly below the melting temperature of sPP (125 C) as

168

Polyolefin Blends

shown in Fig. 7.4. There are two possibilities to explain this phenomenon. One explanation is that sPP single crystals are seemingly floating in the molten POE due to the buoyancy effect. The second scenario involves the melting–recrystallization process. Since the crystallization is a heat-liberation process, it might melt the crystals especially at very shallow supercooling (or high temperature) and then nucleate again at different spots. Nuclei first appear after a short period of time and subsequently disappear, and then new nuclei take their place while some other prior nuclei grow. However, this phenomenon does not occur at lower temperatures or at a larger supercooling. A similar observation was made in the blends of 2/98 and 5/95 sPP/POE after quenching from an isotropic melt (160 C) to 125 C. A rectangular-shaped crystal grows into a sizable single crystal and the tiny periodic striations appear running across the long axis of the single crystal (26). Although the crystallization was carried out from the melt, it is conceptually similar to the solution crystallization because POE melt effectively acts like a theta diluent to the sPP melt. The atomic force microscopic investigation was undertaken to identify these periodic ripples, which turned out to be the corrugated ridges (Fig. 7.6) similar to those reported by others (1,13), except that these ridges primarily formed in the thin sectors of the sPP single crystal. Unlike Lovinger (35) and Reneker’s (36) observations, these undulations are frequently observed in perpendicular to the long axis of the lamella within the thin sectors belonging to the b-axis direction. According to Lovinger et al. (35) and Reneker (36), these periodic undulations are due to the surface instability of the growing single crystals. It should be emphasized that the present optical micrograph was taken at the crystallization temperature; thus, these periodic ridges emerge in situ during the course of crystallization, and thus discounting earlier explanation of thermal contraction of the single crystal upon cooling (29). The contour map of the atomic force micrograph shows clearly that these periodic ridges are confined to thin sectors of the fastest growing front (37). This observation gives a hint that some types of mechanical deformation might have occurred in the aforementioned anisotropic crystallization, thereby influencing the morphology of single crystals. Physically, the polymer melt is highly constrained along the fastest growing b-axis direction. When the crystallization takes place following thermal quenching, there is a sudden increase in density of the emerging crystals, which in turn makes the volume of the constrained melt to shrink at the solid–liquid interface (i.e., crystallizing front). If the shrinkage were to take place preferentially along the constrained direction, one possible scenario for releasing the internal stress is through formation of periodic ripples as the propagating lamellar front solidifies. If the lateral shrinkage occurs, a mechanical torque would develop causing the emerging lamellae to twist although the crystallization is supposedly under quiescent conditions. A natural question is why the lamellar single crystal has to deform in a periodic manner. A possible account is due to the induced mechanical field during isothermal crystallization. This periodicity is consistent over the entire crystal and seems to be sensitive to crystallization temperature. The effect of supercooling on the periodicity of undulations obtained in 1 and 2 wt% sPP blend compositions is

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

169

Figure 7.6 Optical micrographs and AFM images displaying the crystalline morphologies of the 2/98 sPP/POE blend crystallized at various temperatures (AFM images were obtained at room temperature after the samples were isothermally crystallized at a present temperature and subsequently cooled to room temperature).

shown in Fig. 7.7. At a high temperature of crystallization (i.e., a lower supercooling), the periodicity is larger relative to that at a low temperature. The periodic ripples form predominantly in the sectors that belong to the long single-crystal axis. It is noticed that there are some differences between the experiments of Lovinger et al. (35) and the present study. That is to say, their periodic ripples are

170

Polyolefin Blends 3.0

Periodic distance, µm

2.5

1/99 2/98 sPP/POE

2.0

1.5

1.0

0.5

0.0 110

115

120

125

130

Crystallization temperature, °C

Figure 7.7 The dependence of crystallization temperature on periodic distance of ripples in the blends of 1/99 and 2/98 sPP/POE.

located in the thick transverse sectors of the sPP single crystals (35) as opposed to our finding in which the periodic ridges are confined to the thin sectors along the long lamellar axis. Another difference is the crystallization condition, in which crystallization was carried out in the neat sPP from the melt in Lovinger’s case as opposed to the crystallization in the ideal blends of sPP/POE (i.e., POE acts like a polymeric theta solvent to sPP) in the present study. It should be emphasized that single crystal of sPP can obviously grow, not only from the neat sPP but also from the blend of sPP/POE. The dependence of the composition on the evolution of the structure was also carried out by means of OM as shown in Fig. 7.8. Blends of two semicrystalline polymers, that is, sPP and POE, with low concentration of sPP (<30 wt%) exhibit rich and complex crystalline morphology encompassing single-crystal lamellae, sheaflike (or aggregates of single-crystal lamellae) to spherulites. Of particular importance is that the supercooling exerts significant influence on the length scale (size) of the structures, and also changes in morphology depending on the temperature and concentration of the blends. That is to say, faceted single crystals can be grown in a rectangular shape at high crystallization temperatures (i.e., low supercooling) from the melt. With decreasing crystallization temperature, both individual lamellar crystals and lamellar aggregates are observed. The lamellae are seen to fan out, leading to incipient branching, showing a spherulitic-shaped single crystal, a

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

171

Figure 7.8 Optical micrographs displaying the dependence of composition on the crystalline morphologies in blend of sPP/POE at 120 C.

texture called ‘‘premature’’ spherulite. At even lower temperature, for example, 105 C, branching occurs more frequently and the morphology exhibits a fully developed 2D spherulitic texture. The development of such internal textures inside the single crystal is intriguing and deserves to be scrutinized in more detail from a theoretical point of view. In the past few decades, a new and unique approach to study the spatiotemporal dynamics of phase transitions pertaining to order parameters has allowed researchers to gain insight into the growth dynamics and pattern forming aspects, evolving during such a process. These models that embodied a diffuse interface have been used to model epitaxial growth of snowflakes (37), crystallization in metal alloys (38–42) as well as in polymers (32, 33, 43–47). In the preceding sections, we have employed the aforementioned approach to elucidate the thermodynamic and dynamical model of

172

Polyolefin Blends

crystallization from the melt and also to investigate the origins of the ripple formation associated with stress relaxation and the observed sectorization due to differential growth rates of the crystallographic planes. We have also formulated theoretically the phase field model of crystallization in polyolefin blends in order to determine the concentration profiles around the emerging spherulites.

7.2.3 Phase Field Modeling for a Single-Component System: Sectorization and Ripple Formation in sPP The total free energy of crystal ordering in a neat crystalline polymer may be described to consist of local, nonlocal gradient, and elastic terms as ð F ¼ ðflocal þ fgrad þ felastic ÞdV ð7:1Þ where F is the free energy density of crystal ordering consists of a local term, a nonlocal gradient term, and an elastic deformation term, and dV indicates the integration over all volume. flocal is given in the form of Landau expansion of a nonconserved crystal order parameter, c, namely (32, 33, 41, 42) flocal ðcÞ ¼ W

c

ðc

cðc z0 Þðc zÞdc ¼ W

0

c

c2 c3 c4 ðz0 þ zÞ þ zz0 2 3 4

ð7:2Þ

This local free energy has an asymmetric double-well potential for crystal ordering with respect to c, in which c ¼ z0 1 represents the crystalline solid, whereas c ¼ 0 represents the amorphous melt (Fig. 7.9) (32,33). z0 is the solidification potential, which is equal to unity at equilibrium, but it usually has a smaller value for different supercooling reflecting the metastable nature of imperfect crystals. W c is a dimensionless constant representing the strength of the potential field pertaining to c, and z represents the peak position of the energy barrier. The nonlocal free energy term fgrad is customarily given as (44,45) fgrad ðcÞ ¼

ðkc rcÞ2 2

ð7:3Þ

where kc is the tensor representing the coefficient of the interface gradients of the c field, the dot product of kc and rc is vector, and thus its square is scalar. Note that the interface gradient free energy is nonzero only at the interface ð0 < c < 1Þ. Although the polymer chains are flexible in the melt, these molecules become significantly stiffer upon incorporating into the crystals. However, the crystal–melt interface behaves like liquid crystals or liquid membranes of colloidal systems having intermediate properties between the liquid and solid polymer single crystal. To appreciate such a picture, a schematic sketch displaying a chain-folded lamella with periodic undulations is shown in Fig. 7.10, in which the chain tilting was depicted through a tilt angle u between the chain stem and the normal to the plane of the lamella. Since the solid–liquid interface has an intermediate degree of order, it is necessary to take into consideration a higher order curvature elastic term to reflect the

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

173

0.05 T=Tm° T=T1 T=T2

T2
0.04

f (y)

0.03 0.02

ζ

0.01 0

ζ0 –0.01 –0.02 –0.2

0

0.2

0.4

y

0.6

0.8

1

Figure 7.9 Schematic plot of local free energy density and crystal order parameter c for various temperatures showing the metastable energy barrier for phase transition from the melt (c ¼ 0) to the crystalline state (c < 1) at a given crystallization temperature. The crystal order parameter less than unity implies the imperfect crystals containing some noncrystalline components.

crystal–melt interface. Thus, any out-of-plane chain tilting imposes a curvature elastic free energy penalty. We define the chain tilting angle u as the angle made by the polymer chain stems with the normal to the lamellar surface (Fig. 7.10). Then the curvature elastic free energy can be expressed as (46) fce ¼ 12 ½ku ðruÞ2 þ eðr2 uÞ2

ð7:4Þ

where ku and e are coefficients of the gradients of the u field, representing the secondorder and fourth-order curvature elastic terms, respectively. Physically, the first term in Equation 7.4 denotes the nonlocal free energy arising from the tension, whereas the second term refers to nonlocal free energy due to the curvature elasticity due to

Figure 7.10 Schematic sketch displaying a chain-folded lamella with periodic undulations, where chain tilting occurs, represented by a tilt angle u made by the chain stem with the normal to the plane of the lamella and the solid–liquid interface having an intermediate degree of order.

174

Polyolefin Blends

bending. The detailed derivation of this free energy may be found in a paper by Guenthner and Kyu (43). This chain tilting process was originally introduced to describe the development of banded textures in liquid crystalline polymers after cessation of shear, in which the emergence of banded textures in the liquid crystalline polymers has been explained satisfactorily. When the crystallization is taking place, there is a stress built up at the interface. One means of releasing the stress is through a relaxation process, which may occur in the form of local reorientation such as the chain tilting. Assuming that the deformation is small, the strain in the melt at the interface may be given as l=lr ¼ cos u, where lr is the maximum recoverable strain in the material and u is the chain tilt angle. The free energy density of elastic deformation may be written in the context of a neo-Hookean type potential as fel ¼ W u ½ðlr cos uÞ2 þ 1=ðlr cos uÞ2 2

ð7:5Þ

where W u is the elastic modulus. For small deformation strains we have cos u 1 u2 =2 and 1=cos u 1 þ u2 =2. If the chain tilt angle u is small, then the recoverable strain must also be small; that is, defining the variable w ¼ 4ðlr 1Þ, w must also be small. Neglecting higher powers of w we can write l2r ¼ ð1 þ w=4Þ2 1 þ w=2, and 1=l2r ð1 w=4Þ2 1 w=2. Substituting these approximations in Equation 7.5, one obtains fel ¼ W u

4 u 4ðlr 1Þu2 2

ð7:6Þ

Physically, Equation 7.6 represents the strain recovery potential associated with the deformation or volume p contraction ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ during crystallization. This free energy has two minima at w1=2 ¼ 2 ðlr 1Þ, representing the two stable orientations. The ordering in the orientational field takes place in conjunction with the propagation of the interface in the crystal order parameter field. Since the two order parameter fields do not occur independently during crystallization, these two processes must be coupled through a term composed of a linear and/or quadratic dependence of order parameters as follows: fcouple ¼ auðc c2 Þ

ð7:7Þ

a is the coupling strength, which is normally weak, that is, a << W c . This coupling term was chosen to be nonsymmetric in u so that the system can discriminate the chain tilting in one sector from that in the other sector in the crystal. The total free energy functional FðcÞ is then given as ð c2 c3 c4 ðkc rcÞ2 ðz0 þ zÞ þ þ FðcÞ ¼ W c zz0 2 3 4 2 4 1 2 u 2 2 u u 2 auðc c2 ÞdV: þ 4ðlr 1Þu k ðruÞ þ eðr uÞ þ W 2 2 ð7:8Þ

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

175

The total free energy of Equation 7.8 is further inserted in two nonconserved TDGL equations, namely @c=@t ¼ Gc dF=dc and @u=@t ¼ Gu dF=du, to give @c ¼ Gc ½W c cðc z0 Þðc zÞ ðkc Þ2 r2 c þ auð1 2cÞ @t @u ¼ Gu fW u u½4ðlr 1Þ u2 ku r2 u þ er4 u þ acð1 cÞg @t

ð7:9Þ ð7:10Þ

where Gc is the mobility representing the propagation of the interface that is inversely proportional to the viscosity or the frictional coefficient, whereas Gu is the rotational mobility associated with the orientation of the chain in the dissipative medium. Physically, Equation 7.9 signifies the spatiotemporal evolution of crystal order parameter, whereas Equation 7.10 arises due to the strain recovery deformation. It may be noted here that the coupling term acð1 cÞ in Equation 7.10 is nonzero only at the interface and therefore induced deformation must occur at the interface. On the contrary, the coupling term auð1 2cÞ in Equation 7.9 promotes crystallization in regions that simultaneously deform to accommodate the coupling effect. It is apparent that the auð1 2cÞ term changes its sign during crystallization; that is, it is positive in the melt ðc < 1=2Þ and becomes negative when it is in the crystalline state ðc > 1=2Þ. During crystallization at the interface where c > 1=2, the coupling terms in c and u have opposite signs. Therefore, the two propagating waves mutually interfere during solidification, resulting in the transformation from the solitary wave to the oscillatory wave, which in turn generates a rich variety of morphological patterns. If the coupling terms in the two equations have the same sign, the two waves propagate in the same direction without any interference. As a consequence, there will be no pattern formation. Hence, the opposite sign of the coupling terms in the two corresponding fields is the essential criterion in order to discern any pattern formation. It should be emphasized that a simple linear coupling in the two model A equations will not generate any patterns. This observation gives a hint that some types of mechanical deformation might have occurred in the aforementioned anisotropic crystallization, thereby influencing the morphology of single crystals. Physically, the polymer melt is highly constrained along the fastest growing b-axis direction. When the crystallization takes place following thermal quenching, there is a sudden increase in density of the emerging crystals, which in turn makes the volume of the constrained melt to shrink at the solid–liquid interface (i.e., crystallizing front) (Fig. 7.10). If the shrinkage takes place preferentially along the constrained direction, one possible scenario in releasing the internal stress is through formation of periodic ripples as the propagating lamellar front solidifies. If the lateral shrinkage occurs, a mechanical torque would develop causing the emerging lamella to twist although the crystallization is supposedly under quiescent conditions. We therefore attempted to couple the two nonlinear processes of crystallization and mechanical deformation in order to explain the unique morphology observed in the case of sPP single crystal. As expected, the solution of the coupled Equations 7.9 and 7.10 gives the oscillatory wave propagation leading to emergence of periodic

176

Polyolefin Blends

texture. The initial nucleus is isotropic (picture not shown), but the emerging single crystal becomes anisotropic with elapsed time by virtue of the difference in the growth habits of sPP in which the directional growth in the b-axis direction is faster than that along the a-axis, reflecting the differential fold energies along the two growth fronts. Concurrently, the periodic ripples form in the emerging single crystal, predominantly in the sectors that belong to the long single-crystal axis (Fig. 7.11). It is striking that the simulated periodic pattern is in close agreement with that observed experimentally by us for sPP (Fig. 7.6). It has to be noted that the parameters used in the simulation are nondimensionalized using the diffusion constant and characteristic length and are denoted by tilde symbols over the respective parameters as can be

Figure 7.11 Spatiotemporal growth of syndiotactic polypropylene single crystals, exhibiting ripple formation in the sectors belonging to the long axis in both fields: (a) the crystal order parameter and ~ u ¼ 0:4, ~ku ¼ 0, ~e ¼ 0:3, and a ¼ 0:1: (b) the tilt angle. The simulation was carried using parameters G

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

177

seen in the figure captions. Interested readers can refer to the paper by Mehta and Kyu for the complete details of the simulation (45).

7.3 BLENDS OF iPP/POE Metallocene-based iPP was provided by Exxon Chemical Company. The weightaverage and number-average molecular weights of iPP were reported to be 372,000 and 67; 100 g mol1 , respectively, with a polydispersity (Mw/Mn) of 5.54. Isotactic polypropylene usually forms spherulites when crystallized from the melt (7–9). In addition to the spherulitic morphology, it also displays a crosshatched structure at lower crystallization temperatures (24, 25). Lotz and coworkers reported parallel rows of single crystals grown from the melt of neat iPP displaying rough (or serrated) textures on the crystal surface (19). This type of texture has been attributed to the epitaxial crystallization of the g-form on the existing a-crystal modification (19). A different morphology such as a single crystal develops during crystallization from solution or from the melt by blending iPP with a theta solvent, for example, POE. In this section, we shall first discuss the establishment of melting temperature versus composition phase diagram of iPP in the presence of POE polymeric solvent.

7.3.1 Morphology Development in Relation to Phase Diagrams A liquid–solid phase diagram established for blends of iPP/POE by means of DSC and LS is presented at the top left of Fig. 7.12, displaying four distinct regions: isotropic (I), coexistence of crystal–isotropic (C1 þ I), coexistence of crystal– crystal–isotropic (C1 þ C2 þ I), and crystal–crystal (C1 þ C2 þ C3 ). C1 and C2 represent the a- and g-form crystals of iPP, respectively, while C3 is designated for POE crystals (31). The blend preparation is identical to the procedure conducted for sPP/POE blend specimens. These iPP/POE blends are found to be completely miscible in the melt state, showing little or no depression of the melting point with composition. Optical micrographs in Fig. 7.12a–e demonstrate the effect of supercooling (or crystallization temperature) on the crystalline structure in the 10/90 iPP/POE blend, following a temperature quench from isotropic melt (180 C) to various crystallization temperatures under the unpolarized conditions. At a lower crystallization temperature such as 120 C, a lamellar branching (or splay) pattern was observed. Of particular interest is the intricate textures displayed by the single crystals of iPP at smaller supercoolings (T 120 C) in which these crystals tend to grow in the form of two parallel rows, but these two rows merge at the tips reminiscent of a tweezer. It is therefore reasonable to infer that the isothermal crystallization temperature (supercooling) has resulted in the emergence of a variety of crystalline structures of iPP in the blends, as summarized in the morphology map in Fig. 7.13. Figure 7.14 shows the time sequence of the parallel crystals of the iPP/POE blend (10/90), obtained at an isothermal crystallization of 130 C following the

178

Polyolefin Blends

Figure 7.12 (a) Solid–liquid phase diagram for the iPP/POE blends displaying isotropic (I), coexistence of crystal–isotropic (C1 þ I), two crystals form–isotropic (C1 þ C2 þ I) regions, and crystal–crystal (C1 þ C2 þ C3 ) (C1 and C2 correspond to a- and g-forms of iPP and C3 corresponds to crystal of POE), obtained by means of LS and DSC. (b–f) Optical micrographs displaying the dependence of supercooling on crystalline structure in the blend of iPP/POE (10/90) isothermally crystallized at various indicated temperatures.

temperature quench from the isotropic melt (180 C). Although these seemingly twin crystals appear to originate from a common center, it is hard to distinguish whether the two rows of these needlelike crystals are two individual single crystals growing side by side or one single crystal emerging from a common center. In order to gain

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

179

180

Temperature, °C

160

140

120

S

S+A

A

A

S+C

L+C

L+C

L+C

S+C

L+C

L+C

L+C

P

L+C

L+C

P

P

P

L+C

P

L+C L+C L+C

100 0

20

40

60

80

100

wt% iPP Figure 7.13 Morphological phase diagram of iPP/POE blends at several isothermal crystallization temperatures and compositions (S, A, L, P, and C correspond to single crystal, single-crystal aggregates, sheaflike, spherulite, and crosshatched structure, respectively).

Figure 7.14 Structural evolution of 10/90 iPP/POE isothermally crystallized at 130 C showing both the sectorized curved single crystal and crosshatched structure of iPP.

180

Polyolefin Blends

Figure 7.15 AFM phase image of the 30/70 iPP/POE blend isothermally crystallized at 130 C with different magnification, showing the coexistence of lamellae and crosshatched structure.

insight into the above phenomenon, we performed the numerical simulations (45–47) to show the spatiotemporal emergence of the single-crystal patterns. It was found that two parallel rows of crystals indeed belong to the two parallel (010) sectors emerging from a single nucleus (45). It seems that the other (100) sectors are either depleted or too thin to be visible under the optical microscope. It may be concluded that the observed iPP single crystal is composed of the two parallel needlelike sectors growing from a common nucleus. In addition, these crystals appear to have a rough (or serrated) surface. Figure 7.15 exhibits the phase mode AFM images of the single crystals of iPP in the (30/70) iPP/POE blend isothermally crystallized at 130 C. These AFM micrographs show the two dominant sectors of the single crystals appearing nearly parallel to each other. At a higher magnification, the single crystals show periodic undulations (or roughness in the microscopic view) on their surfaces. In a closely related case of sPP, periodic ripples have been observed in single crystals of sPP; the details have been described earlier (45). The appearance of these periodic ripples has been ascribed to the buckling of the crystal occurring as a result of contraction that takes place in the melt, near the crystal–melt interface. However, it should be pointed out that the undulations observed in the iPP single crystals (47) are very different from those of sPP as they are at least an order of magnitude smaller in their periodicity as compared to the ripples found in the sPP single crystals (45). It is plausible that these undulations are the result of the g-phase

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

181

crystals of iPP, which have grown epitaxially on the a-phase single-crystal substrate of the larger rectangular or needlelike crystals. This finding is consistent with those by other researchers (13), who observed serrated lamella in iPP crystals that were grown under similar conditions. It is of particular importance that the surface of iPP crystals isothermally crystallized at various temperatures is not smooth, resulting in the black and white periodicity observed via optical microscopy, which is ascribed to the formation of undulations with interrupted edges. That is to say, undulations can be observed not only in sPP single-crystal lamellae, but also in iPP crystals, though in the different sector. Crystallization studies in blends of iPP/POE reveal that the crystallization process of iPP is affected by the addition of POE and vice versa. It has been demonstrated how the POE promotes the nucleation and crystal growth processes of iPP, this effect being more appreciable at low POE concentration (<10 wt% POE). Analysis of the crystallization kinetics of the iPP crystals isothermally crystallized at different temperatures in blends of iPP/POE is supported by the morphological observations (lamellae, dendritic, and eventually spherulitic textures) through optical microscopy.

7.3.2 Sectorization in iPP/POE Blends In order to gain further insight into the phenomenon described above, we performed the numerical simulations based on Equations 7.9 and 7.10 to illustrate the spatiotemporal emergence of the single-crystal patterns. Since the crystal–melt interface can be easily deformed like liquid membranes or liquid crystals, for example, bending or twisting, any deformation at the solid–liquid interface must be accompanied with a free energy penalty, which is represented by the curvature elasticity, as described in Equation 7.4. In the absence of the shrinkage deformation, the orientation of the polymer chains would be different in each growth front, resulting in the formation of sectorization inside the single crystals. In the present simulation, although the nucleation event could have been initiated via thermal noise, a single nucleus is triggered via perturbation at the center of the frame in order to avoid over crowding. Figure 7.16 illustrates the temporal evolution of the calculated single-crystal structures in both the crystal order parameter and chain tilt angle fields based on the ~ u ¼ 0:4, ~ku ¼ 0, ~e ¼ 0:25, and a ¼ 0:1. Although most parameters in the parameters G c field can be determined experimentally, in this particular case, the above model parameters were chosen arbitrarily to simply demonstrate the capability of the model for describing sectorization. Moreover, different values of the interface gradient coefficients along the (100) and (010) growth planes, that is, ~k100 ﬃ 3 ~k010 , have been utilized reflectingtheanisotropicgrowthintheexperimentalrectangularcrystals.Theindividual sectors are seen distinctly in the u field while appearing weakly in the c field, but are still discernible. It is also noticed that the boundaries of these sectors, which can bevisualized at all time steps during the crystal growth, are trajectories that demarcate the different chain orientations in the different sectors. These demarcating lines undoubtedly result from different chain orientations in the two different sectors.

182

Polyolefin Blends

Figure 7.16 Spatiotemporal growth of isotactic polypropylene single crystals, exhibiting sectorization ~ u ¼ 0:4, ~ku ¼ 0, ~e ¼ 0:3, and a ¼ 0:1: (a) the crystal order parameter as simulated using parameters G c and (b) the tilt angle u. (Reprinted with permission from Reference 45. Copyright (2004) by the American Physical Society.)

7.3.3 Crystal Growth Dynamics in Binary Blends of iPP and aPP From the previous sections, it has become apparent that crystallization from the melt has led to the formation of hierarchical crystalline morphologies including axialites, dendrites, and spherulites. Spherulites developed from a common center with a large number of lamellae growing radially outward. In the lamellar growth, the crystallizing chains are incorporated into the lamellae by folding back and forth, while the noncrystallizing components are rejected from the crystallizing fronts. These noncrystallizable materials are either rejected into the interlamellar regions or excluded

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

183

outside the growing spherulites. In this manner, spherulites incorporate within themselves considerable amounts of amorphous material. The amorphous regions may include in-chain structural defects such as stereoisomeric irregularities, chain ends, and branching disorders. Especially in the case of a blend of crystalline and amorphous polymers, crystallization occurs through solid–liquid phase transition in which the amorphous phase consists of the noncrystallizing polymer and the amorphous part of the crystallizing polymer. Entrapment of the amorphous materials inside a growing crystal causes large interlamellar separations and thus gives the spherulite an open appearance. In such systems, it is seen that greater the amorphous phase, more open the internal texture of the spherulites. We are intrigued by the experiment of Billingham et al. (48), who studied the rejection of noncrystalline materials from a crystallizing front in a blend of iPP and aPP. The uniqueness of their approach is that aPP was labeled with a fluorescent probe, and thus the rejection of labeled aPP from the emerging spherulitic morphology could be observed directly under a fluorescence microscope. Fig. 7.17 exhibits a ring of high concentration of the fluorescent aPP around the boundary of the growing spherulite of iPP, implying that the growing iPP spherulite rejects the noncrystallizing aPP. To better appreciate the rejection of the amorphous materials from the growing spherulite front, the temporal evolution of these concentration profiles has been simulated in the context of a phase field model based on the time-dependent Ginzburg–Landau (TDGL) model C equations (40, 49). The total free energy density of mixing of a crystal–amorphous polymer blend may be expressed as the weighted sum of the free energy density pertaining to crystal order parameter (7) of

Figure 7.17 Fluorescence micrograph of a sample of iPP containing 10% aPP. Sample was quenched after isothermal crystallization at 140 C and is viewed at room temperature. Scale bar: 200 mm. Dark line running through the micrograph is the densiometer trace along the diameter marked by the arrows showing the composition of aPP (46, 48). (Reprinted from Reference 46 with permission from Wiley.)

184

Polyolefin Blends

the crystalline constituent with its volume fraction (f) and the free energy of liquid– liquid mixing as described by the Flory–Huggin’s theory of mixing (50, 51). The free energy density of solidification was weighted by its volume fraction to ensure that the solidification potential vanishes in the limit of zero concentration of the crystalline constituent. We consider all possible interaction terms such as amorphous–amorphous and crystal–amorphous interactions for a binary crystalline system, namely (44, 49) f ð1 fÞ lnð1 fÞ þ fxaa þ xca c2 gfð1 fÞ ð7:11Þ f ðc; fÞ ¼ ff ðcÞ þ lnðfÞ þ r1 r2 where xaa is the Flory–Huggins interaction parameter representing the amorphous– amorphous interaction of the constituent chains in the isotropic melt. xca represents the crystal–amorphous interaction parameter. Note that the subscripts 1 and 2 denote constituent 1 (crystal) and constituent 2 (amorphous polymer), respectively. r1 and r2 correspond to the statistical segmental lengths of the respective components and n1 and n2 are the number of moles. In the case of polymer blend, crystallization can be competed by liquid–liquid phase separation. f is the volume fraction (i.e., composition) of crystallizing species defined as f ¼ n1 r1 =ðn1 r1 þ n2 r2 Þ (51). The overall free energy is given as a weighted addition of the local and nonlocal free energies in c and f fields, namely ð Fðf; cÞ ¼ ðff ðcÞ þ f ðfÞ þ fcoupling ðf; cÞ þ kf jrfj2 =2 þ kc jrcj2 =2ÞdV ð7:12Þ where the weighted addition of ff ðcÞ þ fcoupling ðf; cÞ ¼ xca c2 fð1 fÞ ensures that as f ! 1, the free energy tends to f ðcÞ and as f ! 0, the free energy approaches zero. kf and kc , the interface gradient energy coefficient, are positive. In the phase field model, we consider equations of motions of the nonconserved order parameter c and the conserved order parameter f, that is (48), @c ¼ Gc ½fW cðc z0 Þðc zÞ þ 2xca cfð1 fÞ kc r2 c ð7:13Þ @t 2 3 1 þ ln f 1 þ lnð1 fÞ þ xð1 2fÞ 6 7 r1 r2 6 7 @f 6 7 f ð7:14Þ ¼ r G r6 þW½ðc4 =4 ðz0 þ zÞc3 =3 þ zz0 c2 =2Þ 7 6 7 @t 4 5 kf þxca c2 ð1 2fÞ þ ðrfÞ2 kf r2 f 2 f where G is proportional to the diffusion coefficient of the polymer molecules D. It may be assumed that D is independent of f. Gc is the mobility representing the propagation of the interface that is inversely proportional to the viscosity or the frictional coefficient. It should be emphasized that the model variables such as z, kc , W, xca , and Gc can be related to physical parameters such as experimentally accessible material parameters and experimental conditions. These model parameters are shown explicitly to be supercooling dependent. Equations 7.15 and

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

185

7.16 were solved numerically in one-dimensional (1 200) and two-dimensional square lattices (1024 1024) using an explicit central difference scheme for spatial discretization and explicit forward difference scheme for time steps with periodic boundary condition (46,47). The crystal nucleation event is triggered with a single perturbation at the center of the grid to avoid overcrowding. With the progression of the crystallization, the nucleus emerges to spherulitic morphology (Fig. 7.18). The equation of motion under consideration has a solitary wave solution exhibiting the propagation of a simple domain wall. This observation is consistent with the emerging iPP spherulite that shows no particular texture like concentric rings or corrugated ridges. In the literature, it has been well documented that the melting point of iPP shows little or no change with the addition of aPP, and thus the FH interaction parameter (xFH) is estimated to be very small. The analysis of the melting point depression based on the Flory diluent theory was questioned recently; especially, the complete immiscibility assumption of the solvent in the solid crystal, that is, polymeric solvent, was completely rejected from the polymer crystals. In our recent paper (49), we pointed out the need for distinguishing the segmental amorphous–amorphous interaction and the crystal–amorphous chain interaction in explaining the miscibility of crystalline polymer blends. With this modification, we are able to account for various coexistence regions bound by the solidus and liquidus lines in the phase diagrams and also the discrepancy between the xaa parameter as determined from the melting point depression and small-angle neutron scattering experiments (49). Interested readers on the details are referred to our original paper. In the present calculation, the amorphous–amorphous interaction between the two polypropylene isomers is taken as nearly zero, for example, 1 106 . The corresponding xca parameter for the blend, representing the repulsive interaction between the iPP crystals and aPP amorphous chains, was taken as 1 105 in the simulation. The temporal evolution of the two order parameters (f and c) for a 70/30 iPP/ aPP blend is shown in Fig. 7.18. The dark color at the core represents the crystalline regions of the iPP. A distinct bright ring is evident in the concentration order parameter plot representing a large concentration of aPP surrounding the spherulitic boundary of iPP. A similar observation was made in other compositions. The cross section of the concentration profile of the 90/10 iPP/aPP blend was depicted in Fig. 7.19 as a function of isothermal crystallization time. As expected, the amorphous concentration shows concave curvature at the nucleation site of the iPP spherulite, suggesting the amorphous aPP (noncrystallizing materials) is deficient, but never depleted. As the spherulite grows, the spherulitic boundary propagates toward the surrounding undercooled melt; the rejected aPP amorphous materials are displaced in the radially outward direction of the interface propagation. As can be witnessed, the concentration profile of the aPP fluctuates locally between 7% and 15%, for example, 7% at the core and 15% at the rejected amorphous peak near the advancing spherulitic front and the average concentration of 10% in the surrounding melt of the 90/10 iPP/aPP blend. Both the amorphous concentration profile at the spherulitic growth front and the concave dimple at the core certainly

186

Polyolefin Blends

Figure 7.18 The calculated evolution of spherulitic morphology for iPP crystallizing in a blend of 70/30 iPP with aPP, at crystallization temperature of 125 C. Row (a) shows crystal order parameter field c and row (b) exhibits composition order parameter field f. A ring that formed at the spherulitic front represents the rejected amorphous material. The size of the simulation image represents 128 mm 128 mm.

resemble the experimental concentration profile of the labeled aPP of Billingham et al. (48). The possible entrapment of aPP in the core of iPP spherulite was first reported by Lohse and coworkers based on small-angle neutron scattering and thermal analysis experiments (52, 53), who found that aPP and iPP are not only miscible in the melt, but also the atactic PP chains can be trapped inside the spherulites of iPP on a scale comparable to that of the crystalline lamellae. Billingham and coworkers (48) emphasized the rejection of the amorphous concentration at the growing front, but

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

187

122 s

f

a-PP

0. 2 0. 1 0

0

50

100

150

200 (mm)

150

200 (mm)

150

200 (mm)

244 s

f

a-PP

0. 2 0. 1 0

0

50

100 367 s

f

a-PP

0. 2 0. 1 0

0

50

100

Figure 7.19 One-dimensional simulations of the evolution of the concentration profiles in a 90/10 iPP/aPP undergoing crystallization at 125 C. The concentration profiles are analogous to the experimental results obtain by Billingham et al. (48).

their results indicate that significant amount of amorphous materials were left within the core of the spherulite, which was left unexplained (48). That is to say, the dimple at the core is undoubtedly due to the nucleation event of iPP crystals and thus the aPP materials are deficient there; nevertheless, the aPP materials are not completely depleted. The present calculation also points to the same fact that there is significant amount of amorphous aPP left within the iPP spherulites.

7.4 BLENDS OF ePP/POE Elastomeric polypropylene was provided by Exxon Chemical Company. The weightaverage and number-average molecular weights of ePP were reported to be 32,000 and 6100 g mol1 , respectively, with a polydispersity (Mw/Mn) of 5.22.

188

Polyolefin Blends

7.4.1 Characterization of Neat Elastomeric Polypropylene In this section, we present the morphology development of neat elastomeric polypropylene as well as its blend with POE by means of optical microscopy and atomic force microscopy. ePP is essentially atactic polypropylene, which is amorphous in character, but contains some level of isotacticity (26). The incorporation of crystallizable iPP blocks provides the physical junctions to these amorphous ePP to possess elastomeric properties like a network in the solid state, but these crystal junctions can be removed upon melting, and thus affords melt processability like thermoplastics. Prior to studying the blends, it is desirable to understand the crystal morphology of neat ePP. The isothermally crystallized ePP specimen at 130 C shows the presence of the crosshatched morphological development as shown in Fig. 7.20. It is of interest to note that there is appearance of lamellae growing from the center of crosshatched structure toward the undercooled melt while seemingly splaying out with crystallization time. Similar experiments have been conducted for different isothermal crystallization temperatures. The low magnification AFM micrograph reveals the preferred initial growth direction as well as the branching and splaying of stacks of lamellae at acute angles (54). At intermediate stages of development of spherulitic growth at 120 C, the AFM images of ePP crystallized display the so-called hedrites in Fig. 7.21. Moreover, it was found that at both crystallization temperatures, a higher magnification micrograph shows edge-on lamellae and a dense crosshatching pattern, characteristics of epitaxial growth. These tapping mode (TM)-AFM images accord reasonably well to optical micrographs.

7.4.2 Melting Transitions and Morphology Phase Diagrams of ePP/POE Blends The investigations of neat ePP under isothermal crystallization by means of OM and AFM revealed that ePP shows two types of crystalline morphologies, that is, splayed

Figure 7.20 Optical micrographs and AFM images displaying the crystalline morphologies of the neat ePP crystallized at 130 C (AFM images were carried out at room temperature after the samples were isothermally crystallized at a present temperature and subsequently cooled to room temperature).

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

189

Figure 7.21 TM-AFM height image of neat ePP crystallized isothermally at 120 C showing lamellar edges splaying from a common center.

(or branched) lamellae and crosshatched pattern. It is therefore of interest to further investigate the influence of POE concentration on the morphological development of ePP crystals in ePP/POE blends. As shown in the top left of Fig. 7.22, the melting temperature versus composition phase diagram of the blends of ePP/POE shows virtually no depression of melting transition in both ePP and POE crystals. Moreover, the lack of identifiable liquid– liquid phase separation implies that POE acts like a theta solvent to ePP crystals. The effect of supercooling (crystallization temperature) on the morphological development in the blend of 70/30 ePP/POE is shown in Fig. 7.22a–e. A similar experiment has been conducted for other blend compositions. It was found that the lamellae often showed a curved growth habit. The crystallization temperature exerts strong influence on the content of the a- and g-modifications in which both crystal forms are present in the temperature range of 100–125 C. However, at higher temperatures (>130 C), which are slightly below the melting temperature of the g-phase, only the a-phase (lamellae) are observable. The microscopic observation of the sample over a wide range of crystallization temperatures shows that with decreasing crystallization temperature, the morphology changes from curved single crystals to sheaflike (seaweed), and eventually to spherulites. From the foregoing results, it may be inferred that the supercooling resulted in a change in the emergence of a variety of microstructures. To fully appreciate the dependence of microstructures on composition and supercooling, the morphological phase diagram has been established based on OM (Fig. 7.23). Interestingly, under OM observations, the dark and white regions, which might be the consequence of optical interference, can be discerned clearly on lamellae crystals of ePP grown in blend rich in ePP, for example, 70–90 wt%. It is thus of interest to confirm further if these regions can be attributed to the formation of ripples. In order to accomplish this, AFM pictures were taken. Figure 7.24a and b

190

Polyolefin Blends

Figure 7.22 (a) Liquid–solid phase diagram for the ePP/POE blends displaying isotropic (I), coexistence of isotropic–crystal (I þ C1 ), coexistence of isotropic–crystal (I þ C1 þ C2 ), and crystal–crystal (C1 þ C2 þ C3 ) regions (C1 and C2 correspond to crystals of a- and g-phase of ePP, respectively, while C3 corresponds to POE). (b–f) Optical micrographs obtained displaying the dependence of crystallization temperature on crystalline morphologies of the 70/30 ePP/POE blend isothermally crystallized at various temperatures.

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

Tma (DSC) Tmg (DSC)

160

Temperature, °C

191

140

S+A

S+C L+C

S+C

S

L+C

S+C

L+C L+C

L+C

A+C

L+C

L+C L+C

120 S+A S+A

P

100 0

20

40

60

80

100

wt% ePP Figure 7.23 Morphological phase diagram of ePP/POE blends at several isothermal crystallization temperatures and compositions (S, A, L, C, and P correspond to single crystal, single-crystal aggregates, sheaflike, crosshatched structure, and spherulites, respectively).

shows the TM-AFM pictures in both height and phase images with various magnifications in the blend of 70/30 ePP/POE isothermally crystallized at 130 and 110 C, respectively. According to TM-AFM images, it is evident that over a wide crystallization temperature range (110 and 130 C), elongated curved single crystals of ePP grown from the melt of thin films in blends of ePP/POE exhibit sectorization. Two sector lines are more likely to be found along the diagonal direction of the single crystals, dividing the crystal into four sectors. These four surface sectors, also called fold domains, are in a twin form. Two sectors can be assigned as the (100) sectors in the transverse direction and the (010) sectors in the longitudinal direction, that is, the long axis. The (100) sectors are thicker than the (010) sectors. According to the figure, the dark areas in the middle of the crystals represent the (010) thin sectors and the bright areas are the (100) thick sectors. This implies that at the same Tc, two different sectors with different thicknesses developed in the ePP single crystal. More importantly in these AFM images, for example, the 70/30 ePP/POE blend crystallized at 130 and 110 C, the interrupted edges (lamellar-like texture) might be

192

Polyolefin Blends

Figure 7.24 TM-AFM images of the ePP/POE (70/30) blend crystallized isothermally at 110 C ((a) height image; (b) phase image) showing the elongated single crystals with interrupted edges.

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

193

envisaged as undulations oriented more likely in the thick area of ePP single crystals corresponding to the dark and white regions observed via OM under isothermal crystallization. The lateral periodicity of these undulations is seemingly consistent, which is very similar to the morphological observations of curved single crystals of iPP grown from the melt in blends of iPP/POE discussed in the previous section. It is thus of interest to say that the formation of ripples (undulations) is not unique to syndiotactic polypropylene crystals. Even in the elastomeric polypropylene (ePP) of very low crystallinity (10%), the undulation formation can be observed. Figure 7.25a and b depicts the OM and AFM images of crystals grown at various indicated temperatures for the 30/70 ePP/POE blend, respectively. It should be noted, however, that the AFM images were taken at room temperature after the samples were crystallized at a preset Tc for a certain length of time and then cooled to room temperature, while the optical micrographs were taken at the experimental temperatures. Therefore, the regions of interest displayed via OM and AFM are not the same. Two types of crystals are discernable, which are curved single crystals and the crosshatched structure in ePP crystals grown at 125 and 120 C. It should be noted that there is appearance of sectorization on the ePP curved single crystals and the existence of the surface roughness of these curved structures is presumably thought to be either the formation of ripples (undulations) or the g outgrowth on the a-phase as discussed above. The present study has shown that low crystallinity ePP crystallizes from the melt into well-defined morphologies. This study presented definitive evidence that this class of materials, when crystallized isothermally from the melt, exhibits morphologies that are reminiscent of classical semicrystalline polymers. The presence of lamellae, crosshatching, and spherulites was revealed by high resolution tapping mode AFM and optical microscopy. The nonisothermally crystallized ePP specimens also display the hierarchical ordering as seen in the case of iPP.

7.5 CONCLUSIONS This chapter demonstrated the morphology development in relation to the phase diagrams of isomeric blends of PP/POE. The melt crystallization of sPP in the blends of sPP/POE revealed that the rectangular single crystal of sPP can be grown even in the blends, in which molten POE acts like a theta diluent to sPP at a high temperature above Tm of POE. At high crystallization temperatures, rectangular single crystals can be grown, while the lamellae branch out to form sheaflike and eventually become spherulitic texture with decreasing isothermal crystallization temperature or increasing supercooling gap. In addition, the transverse undulations on the emerging sPP single crystals were theoretically explicable in the context of the phase field theory of solidification. In the case of iPP/POE blends, various types of structures developed in a manner dependent on blend compositions and temperatures of crystallization. With decreasing crystallization temperature, the morphology changed from the curved single crystals to sheaflike (dendritic) and eventually to spherulites. Another interesting

194

Polyolefin Blends

Figure 7.25 (a) Optical micrographs and (b) AFM phase images displaying the crystalline morphologies of the 30/70 ePP/POE blend isothermally crystallized at various temperatures (AFM images are carried out at room temperature after the samples were crystallized at a preset Tc for a certain time and subsequently cooled to room temperature, while the optical micrographs are taken at certain temperatures).

result is that over a wide crystallization temperature range, elongated curved single crystals of iPP can be grown from the melt state of the blends of iPP/POE. The demarcation lines develop along both diagonal directions, thereby forming the (010) and the (100) sectors.

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

195

The crystal morphologies of the ePP exhibit a more complex trend relative to that of iPP in POE solvent. All polypropylene isomers (sPP, iPP, and ePP) in their blends with POE show undulation that may be attributed to the sudden change of density from the melt to solid at the crystallizing solid–melt interface front. It may be concluded that the supercooling has strongly affected not only the length scale, but also the emergence of a variety of hierarchical crystalline topologies in all PP/POE isomeric blends. Last but not the least, it may be inferred that POE seems to behave like a theta solvent to all polypropylene isomers investigated.

NOMENCLATURE c f u z e a kc ku Gc Gu Gf kf l; lr z0 xaa xca AFM C1 C2 dV F fce fcouple fel fgrad flocal I

Crystal order parameter Concentration order parameter, same as volume fraction Chain tilt angle Peak position of energy barrier Coefficient of the fourth-order gradients of the u field Coupling strength Tensor representing the coefficient of the interface gradients of the c field Coefficient of the second-order gradients of the u field Mobility representing the propagation of the crystal–melt interface Rotational mobility associated with the orientation of the chain in the dissipative medium Translational mobility equivalent to the diffusion constant Coefficient of the second-order gradients of the f field Recoverable strain, maximum recoverable strain Solidification potential, equal to unity at equilibrium Amorphous–amorphous interaction energy Crystal–amorphous interaction energy Atomic force microscopy Crystal phase of the PP component Crystal phase of the POE component Volume integral Free energy Curvature elastic free energy penalty Free energy density pertaining to the coupling of the two fields c and u Free energy density of elastic deformation Free energy density of a system undergoing phase transition related to the spatial variation of the order parameters Free energy density of a system undergoing phase transition formulated as a function of order parameters Isotropic melt of PP/POE blend

196

Polyolefin Blends

OM POE r 1, r 2 sPP, iPP, ePP Tc Tm Wc Wu

Optical microscopy Poly(ethylene–octene) copolymer Statistical segment lengths of components 1 and 2 Polypropylene with prefixes indicating the tacticity of the methylene groups on the polymer backbone Zero heating rate crystallization temperature obtained by measuring the DSC crystallization peaks at various cooling rates Zero heating rate melting point obtained by measuring the DSC melting peaks at various heating rates Coefficient determining the strength of the potential field pertaining to c Elastic modulus

REFERENCES 1. P. H. Geil, Polymer Single Crystals, Krieger, Huntington, NY, 1973. 2. F. Khoury and E. Passaglia, Treatise on Solid-State Chemistry, N. B. Hannay (ed.), Plenum, New York, 1976. 3. D. C. Bassett, Principles of Polymer Morphology, Cambridge University Press, New York, 1981. 4. G. Strobl, The Physics of Polymer, Springer, Berlin, 1996. 5. L. Mandelkern, Crystallization of Polymers, McGraw-Hill, New York, 1964. 6. B. Wunderlich, Macromolecular Physics, Vol. 2, Academic Press, New York, 1976. 7. J. J. Point, Bull. Acad. R. Med. Belg. 41, 982 (1955). 8. H. D. Keith and F. J. Padden Jr., J. Appl. Phys., 34, 2409 (1963). 9. H. D. Keith and F. J. Padden Jr., J. Appl. Phys., 35, 1270, (1964). 10. A. Keller, J. Polym. Sci., 36, 361 (1959). 11. R. Norton and A. Keller, Polymer, 26, 704 (1985). 12. F. P. Price, J. Polym. Sci,. 37, 71 (1959). 13. D. C. Bassett and A. Keller, Philos. Mag., 6, 345 (1961). 14. E. Martsucelli, N. Pracella, and L. Grispino, Polymer, 24, 693 (1983). 15. A. J. Lovinger and R. E. Cais, Macromolecules, 17, 1939 (1984) 16. B. Lotz, A. J. Lovinger, and R. E. Cais, Macromolecules, 21, 2375 (1988). 17. A. J. Lovinger, D. D. Davis, and B. Lotz, Macromolecules, 24, 552 (1991). 18. J. Varga, J. Mater. Sci., 27, 2557 (1992). 19. B. Lotz, S. Graff, and J. C. Wittmann, J. Polym. Sci. B Polym. Phys., 24, 2017 (1986). 20. W. Stocker, M. Schuhmacher, S. Graff, A. Thierry, J. C. Wittmann, and B. Lotz, Macromolecules, 31, 807 (1998). 21. G. Natta, J. Polym. Sci., 34, 531, (1959). 22. M. Yamaguchi, H. Miyata, and K. Nitta, J. Appl. Polym. Sci., 62, 87 (1996). 23. A. J. Kinloch and R. J. Young, Fracture Behavior of Polymer, Applied Science, London, 1989. 24. E. D. Carlson, M. T. Krejchi, C. D. Shah, T. Terekawa, R. M. Waymouth, and G. G. Fuller, Macromolecules, 31, 5343 (1998). 25. E. D. Carlson, G. G. Fuller, and R. M. Waymouth, Macromolecules, 32, 8094 (1999). 26. Y. Hu, E. D. Carlson, R. M. Waymouth, and G. G. Fuller, Macromolecules, 32, 3334 (1999).

Chapter 7 Morphological Phase Diagrams of Blends of Polypropylene Isomers

197

27. A. J. Lovinger, B. Lotz, D. D. Davis, and M. Schumacher, Macromolecules, 27, 6603 (1994). 28. B. Lotz, A. J. Lovinger, and R. E. Cais, Macromolecules, 21, 2374 (1988). 29. V. V. Tsukruk, D. H. Reneker, Phys. Rev. B, 51, 6089 (1995). 30. J. Brandrup, E. H. Immergut, E. A. Grulke, A. Abe, and D. R. Bloch, Polymer Handbook, Wiley, New York, 1999. 31. W. Keawwattana, Phase behavior, crystallization, and morphological development in blends of polypropylene (PP) isomers and poly(ethylene–octene) copolymer, Ph.D. Dissertation, The University of Akron, Akron, OH, 2002. 32. H. Xu, R. Matkar, and T. Kyu, Phys. Rev. E, 72, 11804 (2005). 33. H. Xu, W. Keawwattana, and T. Kyu, J. Chem. Phys., 123, 124908 (2005). 34. J. I. Lauritzen Jr. and J. D. Hoffman, J. Appl. Phys., 44, 4340 (1973). 35. A. J. Lovinger, B. Lotz, D. D. Davis, and F. J. Padden Jr., Macromolecules, 26, 3494 (1993). 36. V. V. Tsukruk and D. H. Reneker, Macromolecules 28, 1370 (1995). 37. R. Kobayashi, Physica D, 63, 410 (1993) 38. A. Wheeler, W. J. Boettinger, and G. B. McFadden, Phys. Rev. A, 45, 7424 (1992) 39. J. W. Cahn, J. E. Hillard, J. Chem. Phys. 28, 258 (1958). 40. J. D. Gunton, M. San Miguel, and P. S. Sahni, in: Phase Transitions and Critical Phenomena, C. Domb and J. L. Lebowitz (eds.), Academic Press, New York, 1983, Chapter 8. 41. S.-K. Chan, J. Chem. Phys., 67, 5755 (1977) 42. P. R. Harrowell and D. W. Oxtoby, J. Chem. Phys., 86, 2932 (1987) 43. A. L. Guenthner and T. Kyu, Macromolecules, 33, 4463–4471 (2000) 44. T. Kyu, R. Mehta, and H.-W. Chiu, Phys. Rev. E, 61, 4161 (2000) 45. R. Mehta, W. Keawwattana, A. L. Guenthner, and T. Kyu, Phys. Rev. E, 69, 061802/1-9 (2004). 46. R. Mehta and T. Kyu, J. Polym. Sci. B Polym. Phys., 42, 2892 (2004) 47. R. Mehta, W. Keawwattana, and T. Kyu, J. Chem. Phys., 120, 4024 (2004). 48. N.C. Billingham, P.D. Calvert., and A. Uzuner, Polymer 31, 258 (1990). 49. R. A. Matkar and T. Kyu, J. Phys. Chem. B., 124, 224902 (2006). 50. R. Konigsveld and W.H. Stockmayer, Polymer Phase Diagrams, Oxford University Press, Oxford, NY, 2001. 51. O. Olabisi, L. M. Robeson, and M. T. Shaw, Polymer–Polymer Miscibility, Academic Press, New York, 1979 52. D. J. Lohse, Polym. Eng. Sci. 26, 1500 (1986) 53. D. J. Lohse and G. E. Wissler, J. Mater. Sci. 26, 743 (1991). 54. A. Keller, in: Growth and Perfection of Crystals, R. H. Doremus, B. W. Roberts, and D. Turnbull (eds.), Wiley, New York, 1958.

Chapter

8

Structure, Morphology, and Mechanical Properties of Polyolefin-Based Elastomers Shigeyuki Toki1 and Benjamin S. Hsiao1

8.1 INTRODUCTION Polyolefin-based thermoplastic elastomers and corresponding polyolefin blends are third largest synthetic elastomers in total quantity produced worldwide, right behind styrene–butadiene rubber (SBR) and butadiene rubber (BR). There are three types of polyolefin-based thermoplastic elastomers: (1) blends of polypropylene (PP) and EPM copolymers based on ethylene and propylene monomers (EPM is sometimes termed as EPR), (2) blends of PP and EPDM terpolymers based on ethylene, propylene, and diene monomers, and (3) copolymers of polyolefins. The blends of EPM (EPR) copolymers and isotactic PP are also termed as thermoplastic polyolefin elastomer (TPO), which can be classified into two categories: (i) reactor-made copolymers, the so-called ‘‘impact copolymer PP (ICP), impact PP copolymer (IPC), or high impact polypropylene (hiPP)’’ and (ii) postreactor blends of isotactic PP and EPM (EPR) as well as postreactor blends of ICP and EPR. PP is the biggest products (over 20 million ton per year) in plastics and one quarter of PP products is ICP. Therefore, ICP has huge markets and has been growing at about 10% per year. The blends of EPDM terpolymers and isotactic PP with curing agents, such as peroxide, phenol resins, and sulfur, are termed as thermoplastic vulcanized elastomer (TPV) since the rubber domains are vulcanized. Polyolefin copolymers, such as random copolymer of propylene with ethylene, copolymers of other olefins, elastomeric PP, and elastomeric PE, are developed with recent advances of 1

Department of Chemistry, Stony Brook University, Stony Brook, NY 11794, USA

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

198

Chapter 8 Structure, Morphology, and Mechanical Properties

199

Figure 8.1 The classification of polyolefin-based elastomers.

single-site catalyst (i.e., metallocene catalyst) and other new catalysts in order to replace EPR. The above classification is illustrated in Fig. 8.1. Since all TPO materials are crystallizable, the final crystalline morphology and crystallinity ultimately affect the elastomeric properties. In this chapter, we aim to discuss the basic relationships among mechanical properties, morphology and structures of TPO, TPV, and copolymers of propylene and ethylene.

8.2 THERMOPLASTIC POLYOLEFIN ELASTOMER 8.2.1 Reactor Blends of PP, PE, and EPR: Impact Copolymer PP ICP is also termed as IPC or hiPP, which is manufactured through two-step polymerization. The first step involves the polymerization of propylene, and the second step involves the copolymerization of propylene and ethylene. In the second step, a mixture of PE and EPM (or EPR) is produced. As a result, ICP should not be considered as a copolymer of PP and polyethylene (PE), instead of a blend of PP, PE, and EPM (EPR). The typical morphology of ICP, composed of the PP matrix and dispersed PE–EPR domains (1,2), is shown in Fig. 8.2. In this transmission electron microscopy (TEM) image, the bright right side represents the PP matrix and the dark left side is the domain of PE and EPR. In the PE and EPR domain, the PE component forms a distinct lamellar structure and EPR surround PE. PP makes finer lamella structure than PE in the PE–EPR domain. The dark region is due to the staining by ruthenium tetraoxide (RuO4), which has been successfully demonstrated to examine the crystalline morphology of ultrathin sections of TPO. There is a limitation of the EPR content (around 20%) that can be produced by the second step of polymerization due to the high viscosity and strong adhesion of the

200

Polyolefin Blends

Figure 8.2 TEM image of ICP stained with RuO4: bright right side is the PP matrix; dark left side is a domain of EPR and PE (PE lamellae is visible). (Reproduced from Reference (1) with permission from the Society of Polymer Science, Japan.)

resulting EPR phase. Therefore, the additional content of EPR is often added to ICP in order to increase the toughness and decrease the modulus. For example, some car bumpers are produced in such a manner, composed of ICP, EPR, and talc mixtures. Since talc increases the modulus of the compound, the proper balance of ICP, EPR, and talc ratios is important to fine-tune the impact resistance and the modulus. The morphology of the final ICP products can be directly correlated with the mechanical properties. The morphology such as the size, shape, and distribution of PE–EPR domains depends on many material characteristics such as molecular weight (Mw ) and molecular weight distribution (MWD) of PP, PE, and EPR, the crystallization kinetics and crystallinity of PP and PE, and the strength of adhesion between the two phases. The processing variables such as mixing procedures and different shaping steps can also change the morphology of ICP. The temperature rising elution fractionation (TREF) (3–5) and temperature gradient extraction fractionation (TGEF) (6) techniques have been proven to be the powerful tools to analyze the fractions, Mw , MWD, and stereo regularities of both amorphous and crystalline components of PE or PP. The separated fractions can be examined further by incorporation of light scattering, gel permeation

Chapter 8 Structure, Morphology, and Mechanical Properties

201

Figure 8.3 SEM images of fractured surface ICP specimens after the impact test at 20 C (Reproduced from Reference (4) with permission from Elsevier Science.)

chromatography (GPC), differential scanning calorimeter (DSC), size exclusion chromatography (SEC), and Fourier transform infrared spectroscopy (FTIR) methods. These combined techniques have been demonstrated to effectively design the molecular structures of ICP with improved physical properties. The relationship between the impact properties and morphology in ICP has been studied extensively. For example, Tan et al. (4) and Cai et al. (5) examined the effect of morphology on the impact strength using ICP samples with similar ethylene content, molecular weight, and molecular weight distribution. A typical SEM micrograph of fractured surface ICP specimens after the impact test at 20 C is shown in Fig. 8.3. Poor interfacial adhesion between the disperse phase and the continuous matrix as well as the large dimension and nonuniform distribution of the disperse phase are the two main reasons for the low impact toughness. Zacur et al. (6) reported the distribution of the PE–EPR domains and the crystalline fraction in the domains by using the combined TREF and TEM techniques. The large dimension and nonuniform distribution of the dispersed phase in ICP observed by TEM is clearly seen in Fig. 8.4.

8.2.2 Postreactor Blends of PP–EPR and ICP–EPR Isotactic PP and ICP can be physically blended with EPR to adjust the modulus and the impact resistance. Recently, new types of polyolefin copolymers have been developed to replace EPR with higher compatibility with PP using metallocene catalysts. These polyolefin copolymers will be discussed in Section 8.4. Nitta et al. (7) compared the PP/EPR blends with different EPR in terms of Mw , MWD, and crystallinity. The different compatibility of EPR with PP creates different morphologies, which are shown in Fig. 8.5. The blends containing EPR of low molecular weight and high crystallinity usually exhibited homogeneous morphology, suggesting that both EPR and PP components are miscible. In contrast, the blends containing EPR of high molecular weight and low

202

Polyolefin Blends

Figure 8.4 The typical TEM image of RuO4-stained ICP. (Reproduced from Reference (6) with permission from Society of Plastics Engineers Inc.)

crystallinity exhibited a phase-separated morphology, where the size of the largest detectable EPR domain was in the order of 0.1 mm. The above results indicate that in the miscible blends, the EPR copolymers are partially incorporated in the amorphous regions of PP. On the contrary, in the immiscible blends, the EPR copolymers do

Figure 8.5 TEM images for blend films: PP/EPR¼80/20, EPR are (a) lowest Mw and high crystallinity, (b) low Mw and high crystallinity, (c) high Mw and low crystallinity, (d) highest Mw and lowest crystallinity. (Reproduced from Reference (7) with permission from Elsevier Science.)

Chapter 8 Structure, Morphology, and Mechanical Properties

203

Figure 8.6 TEM images of injection-molded PP alloys with 40 wt% EPR: (a) viewed along the edge direction; (b) viewed along the end direction. (Reproduced from Reference (10) with permission from Elsevier Science.)

not apparently affect the crystalline morphology of PP. The miscibility between PP and EPR thus strongly affects the resulting morphology and the corresponding mechanical properties. Nitta et al. (8) also studied the mechanical properties of PP with ethylene-a-olefin copolymers. They found that the miscible blends exhibited plastic deformation whereas the immiscible blends exhibited the brittle behavior. Seki et al. (9) investigated the miscibility between PP and EPR using small-angle neutron scattering (SANS) with samples containing either conventional EPR or new metallocene catalyst made EPR (the molecular weights of the chosen PP and EPR were very low). They concluded that the phase separation always took place due to the crystallization of PP. Wu et al. (10) demonstrated a successful way to decrease the coefficient of linear thermal expansion (CLTE) of the PP/EPR blends. The low CLTE is important for manufacturing of thin or large products (e.g., car bumpers) with precise dimensions. Although the coefficient of linear thermal expansion of rubber is large, it can be reduced by the addition of EPR in PP blends. The typical crystalline morphology of an injection-molded PP alloy with 40 wt% EPR is shown in Fig. 8.6 ((a) TEM image viewed along the edge direction and (b) TEM image viewed along the end direction). The variation of CLTE for PP/EPR alloys over the entire EPR concentration range is shown in Fig. 8.7. A ‘‘U-shaped’’ curve was observed. The value of CLTE

204

Polyolefin Blends

Figure 8.7 Rubber-concentration dependence of the CLTE in the flow direction (FD) for injectionmolded PP/EPR. (Reproduced from Reference (10) with permission from Elsevier Science.)

increased at EPR concentrations below 20 wt%, decreased at concentrations from 20 to 70 wt%, and then increased rapidly at concentrations above 70 wt%. The last stage of the CLTE increase was due to the phase inversion of the dispersed EPR domain in the matrix, which was verified by TEM. Thus, by the addition of EPR, the value of CLTE can be reduced from 13.5105 to 4.3105 C1. Wu et al. (10) also found that the cocontinuous microlayer morphology was mainly present near the skin portion of the injection-molded part using PP/EPR blends (Fig. 8.6). In the core portion, the polymer domains were less orientated. This confirmed the skin–core morphology of the injection-molded PP/PER blends, where the reduction of CLTE is expected because of the ‘‘skin–core–skin sandwich’’ structure. It has been argued that the thermal expansion coefficient of the polymer alloys along the flow direction (FD) and the thickness direction (TD) can be further reduced by decreasing the thickness of injection-molded specimens. The studies of impact fracture in PP/EPR blends have been carried out by Tam et al. (11), Hayashi et al. (12) and Yang et al. (13). In particular, Hayashi et al. (12) examined the microdeformation of EPR domains in the PP matrix after the impact test by using TEM. They reported that certain part of the sample showed whitening, where the corresponding TEM images are illustrated in Fig. 8.8. Figure 8.9 illustrates the TEM images of the PP/EPR blend with 20 wt% of EPR impacted by different energy (from 5 to 20 J). These images clearly exhibited the microfracture deformation through the formation of craze and voids, which was responsible for the brittle–ductile transition observed during mechanical deformation. The origin of the ductile fracture could be attributed to the relaxation of strain constraint by the microvoids in the craze. Nomura et al. (14) investigated the mechanical fracture behavior of PP/EPR blends containing additional ethylene-a-olefin copolymers (ECP). They found that the addition of a small content (8.2 wt%) of ECP did not change the rubber domain distributions in the matrix (corresponding TEM images are shown in Fig. 8.10).

Chapter 8 Structure, Morphology, and Mechanical Properties

205

Figure 8.8 TEM images of stress-induced whitening zone after the impact test. The EPR content was (a) 6 wt%, (b) 16 wt%, and (c) 20 wt%. To each specimen, the energy just below the fracture energy was applied. (Reproduced from Reference (12) with permission from John Wiley & Sons Inc.)

However, the structures in the rubber domains were significantly affected by the use of ECP with different crystallinity (corresponding TEM images are shown in Fig. 8.11). It was observed that ECP with high crystallinity formed distinct crystalline lamellar structure in the rubber domains, whereas ECP with low crystallinity did not. The average lamella thickness in the blend of PP/EPR/ECP with high crystallinity was 10–20 nm. The scratch resistance of the surface of these blends was measured since the surface toughness of the products, such as car bumpers, is an important performance factor. The surface resistance usually depends on the

Figure 8.9 TEM images showing the microdeformation process of the PP/EPR blend with 20 wt% of EPR. The applied impact energy level was varied by controlling the height of the weight: (a) 5 J, (b) 10 J, and (c) 20 J. (Reproduced from Reference (12) with permission from John Wiley & Sons Inc.)

206

Polyolefin Blends

Figure 8.10 TEM images of PP/EPR (left) and PP/EPR/ECP with 8.2.wt% of ECP (right). (Reproduced from Reference (14) with permission from John Wiley & Sons Inc.)

Figure 8.11 TEM images of PP/EPR/ECP with low crystallinity (left) and PP/EPR/ECP with high crystallinity (right). (Reproduced from Reference (14) with permission from John Wiley & Sons Inc.)

rigidity of the rubber domains. However, the use of ECP with high crystallinity did not increase the scratch resistance. Instead, the addition of ECP with low crystallinity was found to have better scratch resistance probably because the low crystalline ECP was homogeneously distributed in the rubber domain and reinforce the rubber domain evenly.

8.3 THERMOPLASTIC VULCANIZED ELASTOMERS 8.3.1 Dynamic Vulcanization and Morphology TPVs are prepared from blends of isotactic PP and EPDM with curatives, such as peroxides, phenolic resins, or sulfur, by using a process called dynamic vulcani-

Chapter 8 Structure, Morphology, and Mechanical Properties

207

Figure 8.12 TEM images of blends of PP/EPDM and corresponding TPV (PP/EPDM ¼ 20/80 wt%), all samples were stained with RuO4. (Reproduced from Reference (19) with permission from the Society of Rubber Industry, Japan.)

zation. The process of dynamic vulcanization involves the mixing of nonvulcanizable crystalline polyolefin with vulcanizable elastomer in the presence of curatives at temperatures above the melting point of the crystalline polyolefin. Before the onset of vulcanization, the crystalline polyolefin is dispersed in the rubber phase, whereas phase inversion can take place during molten state mixing and vulcanization. TPV usually exhibits higher elastic property and higher temperature endurability than TPO (15–18). As a result, TPV can be used in an engine compartment of automobile. The typical commercialized TPV is composed of PP, EPDM, and a high concentration of extender oil, which are used to reduce the modulus and improve the processability. The performance of TPV is also determined by the morphology of the PP/EPDM blend, Mw , and MWD of each components and the degree of vulcanization in the final products. Dynamic vulcanization is the key process of TPV to create suitable morphology and thus desired physical properties. Before dynamic vulcanization, the blend of PP and EPDM often shows that the major EPDM component is the matrix and the minor PP component is the dispersed domain, as shown in Fig. 8.12 (19). However, after dynamic vulcanization, the PP component becomes continuous phase and the EPDM component forms the dispersed domain. This can be explained as follows. It is known that PP and EPDM are immiscible materials and they exhibit a lower critical temperature (LCST) phase diagram (19). During mixing, especially at high shear rates, the LCST curve elevates with temperature and shear-induced mixing takes place. Thus, in the process of dynamic vulcanization, PP and EPDM can be considered as miscible materials under high shear rates. As a result, after the cross-linking reaction, the unmixed EPDM component forms the dispersed domain, while the matrix consists of mixed PP (dominant) and EPDM (minor) components connected by chemical cross-links (20). TPV can be processed under the same condition as typical thermoplastics such as PP. This unique characteristic distinguishes itself from the process of vulcanized

208

Polyolefin Blends

Figure 8.13 TEM images of TPV samples showing different morphologies (PP/EPDM¼20/80 wt%, stained with RuO4); the samples were prepared with the same recipe at different mixing conditions. (Reproduced from Reference (19) with permission from the society of rubber industry, Japan.)

rubber, which is cumbersome. However, the morphology of TPV can be significantly affected by the processing conditions such as mixing severity (19). Figure 8.13 shows different morphology of the same recipe at different mixing conditions, where strong mixing creates smaller EPDM domains.

8.3.2 Origin of Rubber Elasticity There have been many discussions on the mechanism of the elastic behavior of TPV based on morphological observation, mechanical property measurements, and finite element modeling (FEM). Yang et al. (20), Okamoto et al. (21), and Kikuchi et al. (22) proposed that TPV exhibited the elastic behavior similar to vulcanized rubber because the PP crystals are very fragmented in the matrix caused by the high shear mixing and subsequent dynamic vulcanization. As a result, the matrix becomes less ductile and more elastomeric than the neat PP matrix. As the PP dominant matrix becomes less plastic, it can facilitate the mechanism of strain recovery. Wide-angle X-ray diffraction (WAXD) studies confirmed that the PP crystallites in TPVare much smaller than those in neat PP. The occluded EPDM component in the PP dominant matrix apparently plays an important role as the noncrystallizable specie that hinders the growth of PP crystals. The small PP crystallites suffer less plastic deformation than large PP crystals, and they behave more like cross-linking points to give rise to the elastic properties of the PP dominant matrix. Wright et al. (23–24) discussed the elastic response of TPV using the microcellular modeling, in which three types of deformation such as elastic and plastic deformation of PP, elastic deformation of EPDM, and localized elastic and plastic rotation about PP junction points were considered. Their microcellular model suggests that EPDM domains are surrounded by the PP struts and the struts are connected with hinges. They were successful to elucidate the permanent deformation and stiffness and stress–strain relation of TPV. Boyce et al. (25–27) simulated the stress–strain relationship during compression of TPV using a one-dimensional constitutive model as well as the FEM technique.

Chapter 8 Structure, Morphology, and Mechanical Properties

209

Their results suggest that a pseudocontinuous rubber phase develops as a result of drawing of PP ligaments and shearing of rubber particles. The thickness of PP ligaments appears to control the initial stiffness and stress. Asami et al. (28) investigated the mechanical behavior of TPV using threedimensional FEM in combination with SEM and TEM characterization under uniaxial deformation. The rubber domains were dominantly deformed and elongated through localized yielding in the PP region between neighboring EPDM domains, which becomes perpendicular to the stretching direction. The PP region between the adjacent EPDM domains in the stretching direction remained undeformed, suggesting that the nonstretched PP region also plays a role in connecting the rubber domains. The deformation-induced morphological changes of TPV by uniaxially stretching were examined using SEM, the results of which are shown in Fig. 8.14. The offset in the deformation curve is a consequence of the permanent set imposed by the yielding of the PP dominant matrix, where the dispersed EPDM domains were highly elongated. According to rough estimation of the extension of EPDM domains, the localized strain in the EPDM domains was much greater than the overall applied strain. This means that the strain was concentrated on the rubber domains.

Figure 8.14 (a) The load–elongation curve of TPV measured at 23 C and the elongation speed of 200 mm min1. The SEM images of TPV samples drawn at (b) strain ¼ 1.7, (c) strain ¼ 5.5, (d) strain ¼ 7.4. (Reproduced from Reference (28) with permission from Elsevier Science.)

210

Polyolefin Blends

Figure 8.15 TEM images of TPV samples after drawn to a strain of 5.5. The drawing was performed at 23 C and the elongation speed was 200 mm min1. The EPDM rubber phases were stained with RuO4. The arrow denotes the stretching direction. (a) Before extension, (b) the equatorial PP region between neighboring elongated rubber domains, the microcraze-like fracture is perpendicular to the extension direction, (c) the polar PP region between adjacent rubber domains in the extension direction. (Reproduced from Reference (28) with permission from Elsevier Science.)

The morphological feature of drawn TPV samples was also examined with TEM by Asami et al. (28). The sample preparation involved thin sectioning of the central part of the drawn TPV sample (uniaxially extended to a strain of 5.5). The observed TEM images are shown in Fig. 8.15 (TEM pictures of undrawn TPV and drawn TPV samples). The crystalline lamellar organization of PP was clearly observed in the PP dominant matrix of the undrawn TPV. The TEM images of the drawn TPV sample indicated that the dispersed EPDM domains (dark portion in the image) are highly elongated along the extension direction, and the drawn process promotes a rather heterogeneous deformation in the matrix (Fig. 8.15). It should be noted here that the PP dominant matrix and the dispersed EPDM domains were completely bonded, and the interfacial separation between the matrix and the EPDM domains was not observed in the drawn TPV samples. The microcraze-like fracture was clearly observed in the equatorial PP region between neighboring elongated rubber domains and was aligned perpendicularly to the extension direction; moreover, the lamellar morphology of PP was not clear in this area. On the contrary, the ligament portion of PP crystalline in the matrix located between adjacent rubber domains remained undeformed in the stretching direction. A three-dimensional FEM analysis of a two-phase model corresponding to TPV was carried out and results are shown in Fig. 8.16. In this modeling, complete bonding between the matrix and the EPDM domains was assumed, based on TEM observations. The unloading state of the model is shown in Fig. 8.16a, in which the dark portion represents the EPDM rubber phase. The loading state at strain ¼ 1 is shown in Fig. 8.16b, in which the rubber regions were eliminated in order to visualize the deformed state of the matrix. The darker color represents the higher strain concentrated state. This computational result can explain the experimental data that the deformation state in the matrix is rather heterogeneous and the failure (or yielding) process takes place in the localized PP portion between neighboring rubber domains, resulting in microcraze-like features that aligned perpendicularly to the stretching direction. Moreover, these results suggest that the yielding process in the equatorial PP region between adjacent rubber domains takes place at relatively small strains.

Chapter 8 Structure, Morphology, and Mechanical Properties

211

Figure 8.16 Three-dimensional FEM analysis for uniaxial extension of TPV. The dark region denotes the dispersed cross-linked EPDM phase, and the bright region represents the dominant PP matrix. The arrow denotes the stretching direction. (a) FEM model for undrawn TPV, (b) FEM model for drawn TPV (strain ¼ 1). The darker color indicates the more strained state. (Reproduced from Reference (28) with permission from Elsevier Science.)

8.3.3 Several Factors that Influence Mechanical Properties The effect of the size of EPDM domains in the PP dominant matrix has been investigated since the initial study of TPV (15–19). It is well known that the size of the rubber domain affects the mechanical properties of both TPE and TPV. Wu (29) suggested that the critical rubber domain size and the critical distance between rubber domains are very important for toughness in rubber-filled nylon. This also appears to be the case of TPV. The typical stress–strain curve for TPV is shown in Fig. 8.17. In this curve, the symbol X represents the breaking points as a function of the EPDM domain size (dm ). The smaller EPDM domain size renders the higher tensile strength, which was first observed by Coran et al. (15). They as well as Ellul et al. (30) and Kojina (19) also showed the effect of cross-link densities of the EPDM dispersed phase on the tensile strength and tension set are shown in Fig. 8.18. The higher crosslink density in the EPDM phase was found to result in higher tensile strength and lower permanent set. To improve the compatibility between the PP dominant matrix and the EPDM dispersed phase, various compatibilizers have been prepared and incorporated into

212

Polyolefin Blends

Figure 8.17 The typical stress-strain relationship in TPV, where the symbol X represents the breaking points as a function of the EPDM domain size (dm). (Reproduced from Reference (15) with permission from the rubber division, American Chemical Society Inc.)

the blends. In addition, some reactive processing methods were introduced. The reactive processing is a simple way to react polymer with reactive polymer or monomer in the extruder. For example, various vinyl monomers can be grafted or block-copolymerized onto the backbone polymer such as PP, PE, EPR, and EPDM. One of such compatibilizers is maleic anhydride (MAH). Mixing PP with MAH can make PP-grafted maleic anhydride (PP-g-MAH). However, the molecular weight of PP-g-MAH is not sufficiently high to prepare blend of PP/EPDM with tougher properties. Kim et al. (30) reported polymeric compatibilizers that are based on

Figure 8.18 The effect of cross-link density of the dispersed EPDM phase on the tensile strength and tension set. (Reproduced from Reference (15) with permission from the rubber division, American Chemical Society Inc.)

Chapter 8 Structure, Morphology, and Mechanical Properties

213

Figure 8.19 A schematic representation of preparation for making polymeric compatibilizers based on nylon 66. (Reproduced from Reference (31) with permission from John Wiley & Sons Inc.)

the incorporation of PP-grafted maleic anhydride (PP-g-MAH) and EPDM-grafted maleic anhydride (EPDM-g-MAH) to nylon 66. The schematic diagram for preparing such compatibilizers is shown in Fig. 8.19. They found that the introduction of the compatibiliziler reduced the yield stress, but increased the elongation at break ratio, the results of which are shown in Fig. 8.20. Injection-molded TPV products usually exhibit three regions: skin, middle layer, and core. The distributions of EPDM domains are different in each region. Wang et al. (32) studied the change of morphology in these three regions and the effect on the impact strength. Typical SEM images of these three regions are shown in Fig. 8.21, in which the black holes represent the EPDM particles, which have been

Figure 8.20 Stress–strain curves of PP/EPDM/compatibilizer blends at 240 C (PP/EPDM was 65/35 by weight). (Reproduced from Reference (31) with permission from John Wiley & Sons Inc.)

214

Polyolefin Blends

Figure 8.21 SEM images of static sample for PP/EPDM (80/20) blends, showing three regions: (a) the skin, (b) the middle layer between the skin and the core, (c) the core. (Reproduced from Reference (32) with permission from Elsevier Science.)

dissolved by toluene. The dispersed rubber particles are highly oriented in the skin (Fig. 8.21a) due to the high shear rate from the nozzle during injection molding. This is because the elongated rubber particles were frozen quickly and could not recover back to the original shape. But in the middle layer and the core, since the deformed particles had sufficient time to recovery, the resulting particles became spherical as shown in Figs. 8.21b and c. The impact strength was measured in three fractured directions: along the shear flow direction, perpendicular to, and 45 angle with respect to the flow directions. They showed the fracture directions and the morphology of rubber domains could be strongly correlated with the impact strength. The curing agent can be used to control the cross-link density and the phase morphology of TPV. For example, Ellul et al. (30) reported that with the use of phenolic curative, the viscosity and the domain size of EPDM could be increased, and the swelling could be decreased. Such a curing agent can fabricate TPV with enhanced endurable temperature, suitable for use in goods such as boots, hoses, and air ducts in the engine compartment of automobile. Several other kinds of curing agents have also been developed especially for applications in an engine compartment.

8.4 POLYOLEFIN COPOLYMERS, BLENDS, AND COMPOSITES 8.4.1 Polyolefin Copolymers As discussed earlier, ethylene propylene rubber (EPR or EPM) has been blended with PP and PE to improve the impact strength and to render the materials softer. Recently, metallocene catalysts or postmetallocene catalysts provide new pathways to generate elastic copolymers that can replace EPR. These pathways possess cheaper manufacturing cost and generate new materials with better compatibility to PP or PE. Such new materials included ethylene–propylene random copolymers with dominant ethylene component (33–34) or propylene-dominant component (35–41), propylene– ethylene block copolymer (42), ethylene–octene copolymer (43), poly(propyleneco-ethylene) (44), ethylene–hexene copolymer (45), ethylene–butene copolymer (46), low isotactic PP (47), and stereoblock PP (48). These materials are generally compatible with PP or PE, thus can be used to tailor the toughness (or the softness) of

Chapter 8 Structure, Morphology, and Mechanical Properties

215

the blends. All of these materials exhibit the hard and soft segment morphologies, where the hard segments are composed of PP or PE crystals. The soft segments are due to the random insertion of propylene–ethylene to the backbone or the low stereoregularity in the backbone structure. Many advanced characterization techniques have been used to investigate the structure and property relationship of polyolefin copolymers. For example, positron annihilation lifetime spectroscopy (PALS) (38) has been carried out to analyze the free volume in the amorphous phase confined by the crystallinity of hard segments. TEM and SEM (39–41, 43) techniques were used to investigate the crystal nucleation and growth of these copolymers. The TREF was also used to analyze the compositional heterogeneity of the copolymer. TREF was proven to be a useful technique to correlate the relationships between the chain structure (i.e., comonomer composition and sequence distribution) and the crystalline morphology (44). Dynamic studies of the stress–strain relationship with simultaneous measurements of small-angle X-ray scattering (SAXS) and WAXD have been carried out to examine the structure change during tensile deformation in ethylene-based ethylene–propylene random copolymer (33–34) and propylene-based ethylene–propylene random copolymer (35–36). These studies showed that the original crystallites can be mechanically destroyed (i.e., mechanical melting) during initial deformation, and then new strain-induced crystallites are created to enhance the stress response (so called strain hardening). The results from the in situ X-ray study of propylene-based ethylene–propylene copolymer during tensile deformation (36) are described as follows, because the deduced molecular mechanism appears to be universal in all elastomeric polyolefin copolymers. The stress–strain curves and selected WAXD patterns during extension up to strain 5.0 and subsequent retraction back to strain 1.7 (i.e., at zero stress) in the first deformation cycle are shown in Fig. 8.22. At strain zero, an isotropic broad amorphous halo superimposed with a sharp crystal reflection ring ((040) from the a-form of PP crystal) was observed. The isotropic crystal reflection ring suggested the existence of randomly distributed crystals created in the quenching process during molding. It was seen that during extension, the isotropic (040) crystal ring was first transformed into two arcs and then two distinct spots on the equator. At strains 3.0, two additional equatorial reflection peaks ((110) and (130)) of the -form of PP crystal also appeared, representing the existence of oriented a-form crystals. The total crystal fraction, oriented crystal fraction (IOC), and unoriented crystal fraction (IUC) during extension and retraction have been analyzed. Results can be described as follows. At strain zero, the total crystal fraction was the same as the unoriented crystal fraction, which was about 14%, and both values decreased with strain. This indicates that a fraction of the original crystals was destroyed at the initial deformation stages. (This phenomenon was also reported in ethylene-based ethylene– propylene copolymer (34).) At strain 0.7, the total crystal fraction decreased about 4% (from 14% to 10%); at strains above 0.7, both total crystal fraction and oriented crystal fraction increased with strain, indicating the occurrence of straininduced crystallization, whereas the unoriented crystal fraction decreased continuously with strain. The increase of the total crystal fraction was slower than that of the oriented crystal fraction, suggesting that some unoriented crystals were reoriented by

216

Polyolefin Blends

Figure 8.22 Stress–strain curves and selected WAXD patterns during extension and retraction in the first cycles. Each image was taken at the average strain indicated by the arrow. (Reproduced from Reference (36) with permission from American Chemical Society.)

extension. It is interesting to note that (1) at strain 5.0, the total crystal fraction increased to 25%, that is, a 15% increase due to strain-induced crystallization, and (2) the unoriented crystal fraction did not return to 0% (i.e., completely vanish). In fact, the decrease in unoriented crystal fraction exhibited a two-stage process: at strains between 0 and 1.9, the unoriented crystal fraction decreased at a fast rate (from 14% to about 4%); at strains between 1.9 and 5.0, the unoriented crystal fraction decreased at a slow rate (from 4% to 2.5%). During retraction, the fraction of the unoriented crystal increased very slightly (2.5% to 3.5%), while both total crystal fraction and oriented crystal fraction decreased notably. For example, at strain 1.7, where the stress returns to zero, the total crystal fraction was about 17%, that is, an 8% decrease from strain 5.0. This indicates that a small fraction of strain-induced crystallites is not thermally stable under the unstrained state at room temperature, since they melt away upon relaxation. However, the majority of the strain-induced crystallites are thermally stable, and they form a network structure and can be considered as being permanent set. To investigate the change of lamellar structure in TPE during deformation, in situ SAXS experiments were also carried out. During the extension and retraction processes in the first cycle, the stress–strain relationship and selected SAXS patterns are shown in Fig. 8.23.

Chapter 8 Structure, Morphology, and Mechanical Properties

217

Figure 8.23 Stress–strain curves and selected SAXS patterns during extension and retraction in the first cycles. Each image was taken at the average strain indicated by the arrow. (Reproduced from Reference (36) with permission from American Chemical Society.)

Each image was taken at the average strain indicated by the arrow. During extension, the SAXS image changed from a circular pattern (at strain <0.7), to a (4 þ 2)-point pattern (four scattering peaks on the two off-axes and two scattering peaks on the meridian, at strain between 0.7 and 1.0), and then to a two-point pattern (two scattering peaks on the meridian, at strain >1.0), which are consistent with the WAXD results. Upon retraction, the image changed from a two-point pattern to a four-point pattern (four scattering peaks on two off-axes). The circular pattern represents the existence of randomly oriented lamellar structure, where the 2-, 4-, and (2 þ 4)-point patterns indicate the transformation and the reorientation of the lamellar structure during extension. Two major conclusions can be drawn from the SAXS results. (1) During deformation, two discrete populations of lamellar arrangements—the population with tilted lamellae and the population with the lamellar normal parallel to the stretching (machine) direction—can be produced, depending on the applied strain in the extension or retraction process. The population with the lamellar normal parallel to the stretching direction is the dominant population at high strains. It is most affected by the applied stress and thus is largely responsible for the mechanical response during deformation. (2) The most insightful structure parameter is the long period of

218

Polyolefin Blends

Figure 8.24 Change of long period L from the major lamellar population with strain during extension and retraction in the first cycle. (Reproduced from Reference (36) with permission from American Chemical Society.)

the dominant lamellar population, which can also be calculated directly from the position of the most intense scattering peak, as described in the general literature. The relationship between the long period L of the dominant lamellar population and the applied strain during extension and retraction in the first cycle is shown in Fig. 8.24. In this figure, the values of L below strain 1.0 were omitted because results were so fluctuated that the experimental uncertainty was too high to be meaningful. These fluctuations are due to the process of mechanical melting, which will be discussed later. It was found that the value of L decreased with strain during extension. This is a very interesting behavior because the distance in the material usually extends with increasing strain under normal stretching conditions. However, the observed feature is due to a different behavior, that is, strain-induced crystallization. On the basis of the WAXD results, it is clear that the crystal fraction is increased during extension. As the strain-induced crystalline lamellae most likely occur in the stretched and orientated amorphous regions between existing lamellae, the process would decrease the average long spacing between the lamellae during extension. This peculiar behavior has also been reported in ethylene-based ethylene–propylene copolymer (33). During retraction, L was found to further decrease with decreasing strain. This is due to the relaxation behavior of the permanent set crystalline lamellar network instead of the further crystallization process. In contrast, some portions of defective crystals were melted away during retraction, but which did not affect much on the average distance between the lamellae in the major lamellar population. The process of lamellar rearrangement has been confirmed by the AFM measurement (in tapping mode) on the TPE samples before and after deformation. The corresponding AFM images are shown in Fig. 8.25. The AFM image of the sample before deformation exhibited the existence of lamellar structure in the spherulitic arrangement without preferred orientation. However, the AFM image of the sample after deformation clearly showed an oriented lamellar structure having the lamellar normal parallel to the stretching direction.

Chapter 8 Structure, Morphology, and Mechanical Properties

219

Figure 8.25 AFM images of the TPE sample (a) before deformation and (b) after deformation. (Reproduced from Reference (36) with permission from American Chemical Society.)

On the basis of experimental results in this study, one can propose several schematic diagrams shown in Fig. 8.26 to illustrate the permanent set mechanism during the deformation of TPE in the first cycle. Figure 8.26a indicates the sample before deformation, having folded-chain lamellae without any preferred orientation. Figure 8.26b indicates that during extension below strain 1.0 some original crystallites are destroyed and new crystallites with thinner thickness are developed. These strain-induced thinner lamellae are probably fringe-micelle-like without the folded-chain conformation. The process of permanent set is probably completed at a strain about 1.0. Figure 8.26c indicates that at higher strains (>1.0) all lamellae (the residual of the original crystallites and the straininduced crystallites) are aligned with their normals parallel to the stretching direction. In addition, a large fraction of the amorphous chain segments are also oriented along the stretching direction, storing the entropic recovery force. Figure 8.26d indicates that upon complete retraction back to zero stress the permanent set crystalline network structure persists, where the network lamellar stacks change slightly with respect to the fully stretched lamellar structure. The relaxed lamellar stacks have a smaller average long spacing and preferred orientation having their normals tilted against the original stretching direction. These diagrams may represent the general mechanism of these elastomeric polyolefin copolymers containing crystalline hard segments and flexible soft segments during deformation.

8.4.2 Blends and Composites Practically, some of the above polyolefin copolymers have already been used to blend with PP in applications such as car bumpers and impact elastomeric goods. Although ethylene–propylene random copolymer has been the main component for such

220

Polyolefin Blends

Figure 8.26 Schematic diagrams to illustrate the permanent set mechanism during extension and retraction in the first cycle: (a) before deformation, (b) during extension below the permanent set strain (about 1.0), (c) during extension at strains larger than the permanent set strain, and (d) during retraction to zero stress (reproduced from Reference (36) with permission from American Chemical Society.)

applications, ethylene-1-octene copolymer is making gains in fabrication of elastomers for car bumpers. However, it is relatively rare to find systematic research results of these blends in the open literature. One of the limited examples was the recent work of Paul et al. (43), who showed the improvement on impact resistance by blending ethylene-1-octene copolymer with PP. The above polyolefin copolymers have also been used to prepare conventional composites and nanocomposites. However, similar to the case of polymer blends, not too many studies have been reported thus far. Recently, Kelarakis et al. (49) have mixed 10 wt% of surface-modified carbon nanofiber (MCNF) with propylene– ethylene random copolymer (propylene 84.3%). The MCNF acted as a nucleating agent for crystallization of the a-form of PP in the matrix. During deformation at room temperature, strain-induced crystallization took place, while the transformation from the g-phase to a-phase also occurred for both unfilled and 10 wt% MCNFfilled samples. The tensile strength of the filled material was consistently higher than that of pure copolymer. These results are illustrated in Fig. 8.27. However, when compared with pure copolymer, the highly stretched nanocomposite exhibited a higher amount of unoriented crystals, a lower degree of crystal orientation, and a higher amount of g-crystals. This behavior indicated that polymer crystals in the filled nanocomposite experienced a reduced load, suggesting an effective load transfer from the matrix to MCNF. At elevated temperatures, the presence of MCNF resulted in a thermally stable physically cross-linked network, which facilitated strain-induced crystallization and led to a remarkable improvement in the mechanical properties. For example, the toughness of the 10 wt% nanocomposite was found to increase by a factor of 150 times at 55 C. Although nanofillers

Chapter 8 Structure, Morphology, and Mechanical Properties

221

Figure 8.27 Stress–strain curves and selected WAXD patterns acquired during stretching of pure copolymer (unfilled circles) and 10 wt% nanocomposite (filled circles) at room temperature. (Reproduced from Reference (49) with permission from Elsevier Science.)

such as MCNF, carbon nanotubes, and nanoclays seem to be too expensive to use in modification of polyolefin today, extensive research activities on the nanocomposite products have already been explored. If the price of some nanofiller can fall in the future, more applications will be realized.

8.5 CONCLUSION The production of polyolefin-based elastomers has increased significantly in recent years, especially in applications such as car instrumentation and general goods, because of their good mechanical properties, easy processing conditions, and cheap cost. The particular advances are expected in the following areas. (1) The development of the metallocene catalyst has greatly enhanced the production of a wide range of rubbery components, suitable for fabrication of ICP. The postreactor blend of EPR or new polyolefin copolymers with ICP thus may be integrated into the production line during ICP polymerization; the filler addition step can also be incorporated in the same ICP polymerization line. (2) The integrated processing of filler and rubber additions can be used to produce reactor TPO (RTPO). The production of RTPO may become more cost effective than before, which would allow RTPO to eventually replace PVC (polyvinyl chloride) (this is because PVC is considered as an environmentally hazard material to produce). However, to achieve such a goal, TPO must be made with softer properties and the surface must be made resistant to scratch. (3) For TPV, higher temperature endurability will be required due to the more applications in the engine compartment of automobile. For this purpose, more

222

Polyolefin Blends

effective curing agents must be developed for vulcanization of EPDM while maintaining good adhesion between the EPDM domains and PP ligaments. (4) Conventional EPR may be replaced by new types of polyolefin copolymers using metallocene catalyst or postmetallocene catalyst in the future because of the cheaper cost and higher design freedom. (5) The new polyolefin copolymers can be reinforced by nanofillers such as clay, carbon nanofiber, and carbon nanotubes, but the high price of such fillers may restrict broad applications. In conclusion, we believe that the future of polyolefin-based elastomers is bright with many opportunities, while the study of the morphology, crystallization, and structure of different composition will become even more essential to the different applications.

ACKNOWLEDGMENT We wish to acknowledge the financial support of this work by the National Science Foundation (DMR-0405432).

NOMENCLATURE BR ECP EPM EPDM EPR ICP IPC hiPP MAH PE PP SBR TPO TPV

Butadiene rubber Ethylene-a-olefin copolymer Ethylene–propylene copolymer Ethylene–propylene diene terpolymer Ethylene–propylene rubber. Impact copolymer polypropylene Impact polypropylene copolymer High impact copolymer polypropylene Maleic anhydride Polyethylene Polypropylene Styrene–butadiene rubber Thermoplastic polyolefin elastomer Thermoplastic vulcanized elastomer

REFERENCES 1. H. Sano, T. Usami, and H. Nakagawa, Polymer, 27, 1497 (1986). 2. H. Sano and T. Komoto, Kobunshi (High Polymers, Japan), 54, 674 (2005). 3. F. M. Mirabella, Polymer, 34(8), 1729 (1993). 4. H. Tan, L. Li, Z. Chen, Y. Song, and Q. Zheng, Polymer, 46, 3522 (2005). 5. H. Cai, X. Luo, X. Chen, D. Ma, J. Wang, and H. Tan, J. Appl. Polym. Sci., 71, 103 (1999). 6. R. Zacur, G. Goizueta, and N. Capiati, Polym. Eng. Sci., 40,1921 (2000). 7. K. Nitta, Y. Shin, H. Hashiguchi, S. Tanimoto, and M. Terano, Polymer, 46 965 (2005). 8. K. Nitta, K. Okamoto and M. Yamaguchi, Polymer, 39, 53 (1998).

Chapter 8 Structure, Morphology, and Mechanical Properties

223

9. M. Seki, H. Nakano, S. Yamaguchi, J. Suzuki, and Y. Matsushita, Macromolecules, 32, 3227 (1999). 10. G. Wu, K. Nishida, K. Takagi, H. Sano, and H. Yui, Polymer, 45, 3085 (2004). 11. W. Y. Tam, T. Cheung, and R. K. Y. Li, Polym. Test., 15, 363 (1996). 12. K. Hayashi, T. Morioka, and S. Toki, J. Appl. Polym. Sci., 48, 411 (1993). 13. J. Yang, Y. Zhang, and Y. Zhang, Polymer, 44, 5047 (2003). 14. T. Nomura, T. Nishio, K. Iwanami, K. Yokomizo, K. Kitano, and S. Toki, J. Appl. Polym. Sci., 55, 1307 (1995). 15. A. Y. Coran, Rubber Chem. Technol., 53, 141 (1980). 16. A. Y. Coran, Rubber Chem. Technol., 68, 351(1995). 17. M. D. Ellul, Rubber Chem. Technol., 71, 244 (1998). 18. M. D. Ellul, Rubber Chem. Technol., 76, 202 (2003). 19. J. Kojina, J. Soc. Rubber Ind. Jpn., 76, 310(2003). 20. Y. Yang, T. Chiba, H. Saito, and T. Inoue, Polymer, 39, 3365 (1998). 21. M. Okamoto, K. Shiomi, and T. Inoue, Polymer, 35, 4619 (1994). 22. Y. Kikuchi, T. Fukui, T. Okada, and T. Inoue, Polym. Eng. Sci. 31,1029 (1991). 23. K. J. Wright, and A. J. Lesser, Polym. Eng. Sci. 43, 531 (2003). 24. K. J. Wrigh,t and A. J. Lesser, Rubber Chem. Technol. 74, 678 (2001). 25. M. C. Boyce, K. Kear, S. Socrate, and K. Shaw, J. Mech. Phys. Solids 49 1073 (2001). 26. M. C. Boyce, S. Socrate, K. Kear, O. Yeh, and K. Shaw, J. Mech. Phys. Solids 49 1323 (2001). 27. M. C. Boyce, O. Yeh, S. Socrate, K. Kear, and K. Shaw, J. Mech. Phys. Solids 49 1343 (2001). 28. T. Asami and K. Nitta, Polymer 45, 5301 (2004). 29. S. Wu, Polymer, 26, 1856 (1985). 30. M. D. Ellul, A. H. Tsou, and W. Hu, Polymer, 45, 3351 (2004). 31. B. C. Kim, S. S. Hwang, K. Y. Lim, and K. J. Yoon, J. Appl. Polym. Sci., 78, 1267 (2000). 32. Y. Wang, Q. Zhang, B. Na, R. Du, Q. Fu, and K. Shen, Polymer, 44, 4261 (2003). 33. L. Liu, B. S. Hsiao, B. Fu., S. Ran, B. Chu, A. H. Tsou, and P. K. Agarwal, Macromolecules, 36, 1920 (2003). 34. L. Liu, B. S. Hsiao, S. Ran, B. X. Fu., S. Toki, F. Zuo, A. H. Tsou, and B. Chu, Polymer, 47, 2884 (2006). 35. S. Datta, S. Srmanas, C. Y. Cheng, A. H. Tsou and D. Lohse, ACS Rubber Division Fall Meeting Proceeding, Paper #60, 2003. 36. S. Toki, I. Sics, C. Burger, D. Fang, L. Liu, B. S. Hsiao, S. Datta, and A. H. Tsou, Macromolecules, 39, 3588 (2006). 37. Y. Feng and J. N. Hay, Polymer, 39, 6589 (1998). 38. H. P. Wang, P. Ansems, S. P. Chum, A. Hiltner, and E. Baer, Macromolecules, 39, 1488 (2006). 39. Y. Zhao, A. S. Vaughan, S. J. Sutton, and S. G. Swingler, Polymer, 42, 6587 (2001). 40. Y. Zhao, A. S. Vaughan, S. J. Sutton, and S. G. Swingler, Polymer, 42, 6599 (2001). 41. J. Weng, R. H. Olley, D. C. Bassett, and P. Jaaaskelainen, J. Polym. Sci. Polym. Phys., 42, 3316 (2004). 42. Y. Feng and J. N. Hay, Polymer, 39, 5277 (1998). 43. S. Paul and D. D. Kale, J. Appl. Polym. Sci., 76, 1480 (2000). 44. S. Wang and D. Yang, Polymer, 45, 7711 (3004). 45. M. Yamaguchi, K. Nitta, H. Miyata, and T. Masuda, J. Appl. Polym. Sci., 63, 467 (1997). 46. K. Nitta, K. Okamoto, and M. Yamaguchi, Polymer, 39, 53 (1998) 47. M. Kijima, S. Yukimasa, and T. Tatsumi, Polym. Mater. Forum Prep. Jpn. 10, 89 (2001). 48. Y. Hu, M. T. Krejchi, C. D. Shah, C. L. Myers, and R. M. Waymouth, Macromolecules, 31, 6908 (1998). 49. A. Kelarakis, K. Yoon, I. Sics, R. H. Somani, B. S. Hsiao, and B. Chu, Polymer, 46 5103 (2005).

Chapter

9

Morphology and Mechanical Properties in Blends of Polypropylene and Polyolefin-Based Copolymers Koh-Hei Nitta1 and Masayuki Yamaguchi2

9.1 INTRODUCTION Isotactic polypropylene (iPP) is a typical semicrystalline polymer and iPP resins have been extensively used in various products such as automotive parts, electronics, medical applications, semiconductor parts, housing, housewares, woven and nonwoven fibers, and food containers. However, the application of iPP is limited owing to its poor impact toughness in particular at lower temperatures. The modification of mechanical properties of iPP has long been of very interests to scientists and engineers in polymeric materials. So far the toughness of iPP at lower temperatures has been improved by blending various rubbery polyolefin materials (1–8) such as ethylene–propylene random copolymers (EPR), styrene–ethylene butylene–styrene triblock copolymers (SEBS), butylrubbers, and ethylene–propylene–diene terpolymers (EPDM). In particular, binary blends of iPP and EPR have been extensively investigated from the commercial point of view. It is quite important to comprehend that the miscibility and/or compatibility in polyolefin blends dominate their material characteristics and basic properties. The lack of polarity, however, often leads to the phase-separated state in most polyolefin blends in spite of the similar chemical structure. For example, it is well known that 1 Department of Chemical Engineering, Graduate School of Material Sciences, Kanazawa University, Kanazawa 920-1192, Japan 2 School of Materials Science, Japan Advanced Institute of Science and Technology, Nomi, 923-1292 Japan

224

Chapter 9 Morphology and Mechanical Properties in Blends

225

iPP is immiscible with polyethylene (PE) and commercially available EPR. Since iPP has huge market and the demand is still growing at a high rate, it has been a major research topic for the last decade to improve the miscibility and/or compatibility with rubbery polyolefins. Although the control of miscibility and/or compatibility plays an important role in the improvement of mechanical properties of the iPP/EPR blends, the gross morphology in iPP/EPR blends was hard to control because the conventional rubbery copolymers essentially are incompatible with iPP (9). Meanwhile, recent development in the field of catalyst technology for polyolefin copolymerization enabled us to control the content of a-olefin as well as the stereoregularity of propylene or a-olefin sequence in the copolymers. Further, various types of random copolymers covering nearly entire composition range using metallocene-based catalyst systems have been capable of being prepared. Their mechanical properties and morphology essentially differ from those of the commercially available copolymers by Ziegler–Natta catalysts. The addition of these new types of copolymers is expected to be a powerful tool for improvement in mechanical toughness or drawability of iPP and to make it possible to produce a new type of iPPbased thermoplastics. We begin in Section 9.2 with the morphology in binary blends of iPP and various rubbery olefin copolymers where we remark the interrelation between the miscibility and dynamic mechanical properties. Section 9.3 describes the molecular orientation behavior under tensile deformation of iPP-based blends, and we compare the differences in deformation behavior between miscible and immiscible blends. Section 9.4 contains the solidification process in iPP-based blends where the effects of miscibility in the molten state on the crystallization of iPP matrix are discussed.

9.2 MORPHOLOGY AND DYNAMIC MECHANICAL PROPERTIES 9.2.1 Blends with Polyethylene or Poly(butene-1) According to the dynamic mechanical analysis by Piloz et al. (10), poly(butene-1) (PB) is miscible with iPP in the amorphous region. Boiteux et al. (11) later confirmed the miscibility between iPP and PB from dielectric relaxation measurements. It has been clarified that blending PB depresses the crystallization rate of iPP because PB molecules act as a diluent for iPP (12–15). This phenomenon implies that PB is miscible with iPP in the molten state. It has been reported that the morphology and the mechanical properties in iPP/PE blends depend on the species of PE. Dumoulin et al. demonstrated that linear low density polyethylene (L-LDPE) shows superior mechanical properties to low density polyethylene (LDPE) produced by radical polymerization under high pressure and high density polyethylene (HDPE) (16). Also, the interfacial tension between iPP and PE depends on the primary structure of PE; that is, the interfacial tension in iPP/LDPE and iPP/HDPE blends is higher than that in iPP/L-LDPE (17,18), indicating better adhesion between iPP and L-LDPE (17). Moreover, it was reported that the interfacial tension between iPP and L-LDPE

226

Polyolefin Blends

decreases with the increase in the a-olefin content in L-LDPE (19). These experimental results suggest that the miscibility with iPP depends on the primary structure or chemical composition of polyolefins.

9.2.2 Blends with Ethylene–a-Olefin Copolymers 9.2.2.1 Ethylene–Butene-1 Copolymer Considering that PB is miscible with iPP in the amorphous region, ethylene–butene-1 random copolymer (EBR) with large amount of butene-1 is expected to be miscible with iPP in the amorphous region. Yamaguchi et al. revealed that the EBRs with and 62 mol% of butene-1 are dissolved in the amorphous iPP region, whereas the EBRs with 36 and 45 mol% of butene-1 are immiscible with iPP (19). Weimann et al. also found that the EBRs with 73 wt% (57 mol%) and 90 wt% (81 mol%) of butene-1, which are obtained by hydrogenation of polybutadienes with various amounts of 1,2 addition, are miscible with iPP in the molten state by small-angle neutron scattering (SANS) (20). According to the electron microscopic observation of the morphology in iPP/EBR blends by Ma¨der et al. (21), there is no phase separation in the blends with the EBRs with 82 wt% (69 mol%) and 90 wt% (81 mol%) of butene-1 (21). The experimental results are consistent with the theoretical prediction proposed by Lohse et al. (22) that EBRs require at least 58 mol% of butene-1 to be miscible with iPP. Figure 9.1 shows the transmission electron micrograph (TEM) for the blends composed of iPP and EBRs with a different butene-1 content (iPP/EBR¼75w/25w), where the white region denotes the iPP-rich phase and the dark region is the EBRrich phase. Apparently, the blend with EBR45 (EBR with 45 mol% of butene-1) shows finer dispersion particles than iPP/EBR36 (that with 36 mol% of butene-1) does. The magnified picture of iPP/EBR45 (50/50) as shown in Fig. 9.2 demonstrates that the iPP lamellae are inserted into EB-rich amorphous region, resulting from the thick interfacial region in the molten state, that is, low interfacial tension between iPP and EBR45. On the contrary, the iPP/EBR56 exhibits no phase separation and fairly homogeneous morphology. Dynamic mechanical spectra of the iPP/EBR (75/25) blends are shown in Fig. 9.3. In the blends with phase-separated morphology; for example, iPP/ EBR36, double peaks appear in the tensile loss modulus E00 curve distinctly in the temperature range from 200 to 300K, which are assigned as b1 and b2 in the order of

Figure 9.1 TEM micrographs for iPP/EBR(75/25) stained by ruthenium tetraoxide. (From Reference 18 with permission from Society of surface science Japan.)

Chapter 9 Morphology and Mechanical Properties in Blends

227

Figure 9.2 TEM micrographs for iPP/EBR45(50/50) stained by ruthenium tetraoxide.

10

10

(b)

iPP/EBR36 (75/25) E'

2

10 Hz 1

9

10

β1

β

2

10-1

γ

1

10 E"

8

0

10

10

β

7

10

tan d 200

10-2

300

2

-1

10 α tan d

6

10 100

400

200

300

-2

400

10

Temperature, K

Temperature,K (c)

β 1

γ

α

106 100

E', E", Pa

10

tan d

0

10 Hz

10

E" 8

102

9

10

107

iPP/EBR45 (75/25) E'

10 E', E",Pa

1010

10

tan d

(a)

1010

iPP/EBR56 (75/25)

E'

102 10 Hz

9

1

10

10

0

10

10

tan d

E ', E ",Pa

E" 8

β -1

7

10

10 γ

α tan d

6

10

100

200

300

-2

10 400

Temperature, K

Figure 9.3 Temperature dependence of dynamic tensile moduli at 10 Hz for (a) iPP/EBR36 (75/25), (b) iPP/EBR45 (75/25), and (c) iPP/EBR56 (75/25).

228

Polyolefin Blends

the lower temperature, and the magnitude of the b2-relaxation decreases with the decrease in the domain size. However, the iPP/EBR56 shows only single peak in the temperature range. All blends show a broad relaxation in the temperature range from 320 to 380K, assigned as a-relaxation, which is associated with the molecular mobility in the crystalline phase of iPP. The storage modulus monotonously reduces with temperature in the a dispersion and falls off sharply at the melting point of iPP being independent of blending with EBR, suggesting that EBRs do not affect the crystalline region of iPP. Figure 9.4 shows the temperature dependence of E00 curves at around b-relaxation for iPP, EBR, iPP/EBR (50/50), and iPP/EBR (75/25). The higher (a)

9

10

(b)

iPP/EBR36 β1

109

iPP/EBR45 β1

10 Hz

β2

E ", Pa

E", Pa

β2 8

10

107 175 200 225

108

107 175

250 275 300

200

Temperature, K (c)

109

10 Hz

225 250 275 Temperature, K

300

iPP/EBR56

E ", Pa

10 Hz

108

107 175 200

225

250

275 300

Temperature, K

Figure 9.4 Temperature dependence of tensile loss modulus at 10 Hz for (a) iPP/EBR36, (b) iPP/

.

EBR45, and (c) iPP/EBR56: ( ) iPP, (#) iPP/EBR(75/25), (^) iPP/EBR(50/50), and (^) EBR. (From Reference 19 with permission from John Wiley & Sons, Inc.)

Chapter 9 Morphology and Mechanical Properties in Blends

229

relaxation peak b2 is ascribed to the Tg of the amorphous region of iPP, and the lower relaxation peak b1 is to the Tg of EBR. For the iPP/EBR45 blend, however, the b2 peak is slightly shifted to lower temperatures than Tg of iPP, whereas the b1 peak is slightly shifted to higher temperatures than Tg of EBR45. These results suggest that the amorphous region of iPP and EBR45 are partially dissolved each other in the blends. Details will be described in Section 9.4. Moreover, there is only single peak between Tgs of pure components for the blend of iPP with EBR56 (56 mol% of butene-1), suggesting that EBR56 molecules are completely incorporated in the amorphous region of iPP. 9.2.2.2

Ethylene–Hexene-1 Copolymer

Ethylene–hexene-1 random copolymers (EHRs) with more than 50 mol% of hexene-1 are miscible with iPP in the amorphous region (23,24). Figure 9.5a shows the master curves of dynamic shear moduli in the molten state for the binary blends composed of iPP and the EHRs with different hexene-1 contents ((a) 33 mol%, EHR33 and (b) 51 mol%, EHR51). The values of dynamic shear moduli of the iPP/EHR51(70/30) are intermediate between those of the pure components, whereas the iPP/EHR33(70/30) shows a higher shear storage modulus G0 in the low frequency regions, which is associated with the existence of a long-time relaxation. Figure 9.6 shows that the iPP/EHR33 has a longer relaxation time although EHR33 and EHR51 have similar relaxation spectra. The longest relaxation time of the iPP/EHR51 (70/30), evaluated by the Procedure X diagram, was 1.6 s, whereas that of iPP/EHR33 (70/30) was 4.3 s, which is greater than that of each pure component; that is, 1.9 s for iPP and 0.5 s for EHR33. The long-time

(a)

5

10

(b)

iPP/EHR33(70/30)

5

10

4

4

3

10

G', G",Pa

G', G",Pa

10

10

G" G'

2

10

103

G" G'

2

10

1

1

10

10

463 K 0

10 -2 10

iPP/EHR51 (70/30)

-1

10

0

1

10

ωa ,s T

10 -1

2

10

100 -2 10

463 K 10-1

100 ωaT,s

101

102

-1

Figure 9.5 Master curves of dynamic shear moduli at 463K for (a) iPP/EHR33 (70/30) and (b) iPP/ EHR51(70/30). (From Reference 24 with permission from John Wiley & Sons, Inc.)

230

Polyolefin Blends 104 463 K

G(t),Pa

103

102 iPP/EHR33 (70/30)

101 iPP/EHR51 (70/30) 0

10

0

5

10

15

t,s

Figure 9.6 Relaxation modulus for iPP/EHR (70/30) at 463K. (From Reference 24 with permission from John Wiley & Sons, Inc.)

relaxation component of the iPP/EHR33 blends can be considered to correspond to the geometrical relaxation of the dispersed phase owing to the interfacial tension. Miscibility in the molten state affects the lateral growth rate of secondary crystallization of iPP component. Figures 9.7a and b show the growth curves of the spherulite radius during isothermal crystallization at 403K for the iPP and blends. As seen in the figure, the spherulite radii of all samples increase linearly with time over the entire experimental range, suggesting that a constant blend ratio is maintained in the melt at the front of spherulites. In addition, blending EHR51 depresses the spherulite growth (a)

80

iPP/EHR33

(b)

80

iPP/EHR51

Tc : 403K

Tc : 403K 60

R,μm

R,μm

60

40

20

iPP 10 wt% EHR 30 wt% EHR

0 0

200

400

600

40

20

iPP 10 wt% EHR 20 wt% EHR 30 wt% EHR

0 0

t,s

200

400

600

t,s

Figure 9.7 Growth curves of spherulite radius R at 403K for (a) iPP/EHR33 and (b) iPP/EHR51. (From Reference 24 with permission from John Wiley & Sons, Inc.)

Chapter 9 Morphology and Mechanical Properties in Blends (a)

109

(b)

iPP/EHR33

109

iPP/EHR51

10 Hz β 108

1

β

231

β

10 Hz

2 8

10

α

E ",Pa

E ",Pa

α

107

7

iPP 10% EHR 20% EHR 30% EHR EHR33

106 100

200 300 400 Temperature, K

10

106

iPP 10% EHR 20% EHR 30% EHR EHR51

100

200 300 400 Temperature, K

Figure 9.8 Temperature dependence of dynamic tensile moduli at 10 Hz for (a) iPP/EHR33 and

.

(b) iPP/EHR51; ( ) iPP, (#) iPP/EHR (90/10), (^) iPP/EHR (80/20), (D) iPP/EHR (70/30), and (~) EHR. (From Reference 24 with permission from John Wiley & Sons, Inc.)

rate to a great degree, demonstrating that EHR51 molecules are dissolved into iPP matrix and act as a diluent. On the contrary, the growth rate of iPP spherulites in the immiscible iPP/EHR33 blends is independent of the addition of EHR33, suggesting that the immiscible blends show phase separation in the molten state. Although the crystallization condition affects the morphology to some degree, which will be discussed later, the miscibility in the molten state has strong influence on the morphology in the solid state, in particular, for quenched samples. The temperature dependence of dynamic tensile modulus in the blends quenched in ice-water bath is shown in Fig. 9.8. The level of the tensile storage modulus E0 of iPP/EHR51 is lower than that of iPP/EHR33. As seen in the figures, the iPP/EHR51 shows a single b-relaxation peak ascribed to the glass transition in the temperature region between Tgs of the pure components (200–300K). This suggests that most of EHR51 chains is incorporated in the amorphous region of iPP. On the contrary, the iPP/EHR33 shows two separated peaks in the temperature region and each peak is located at the Tg of each pure component (279K for iPP and 214K for EHR33). This is reasonable because EHR33 is immiscible with iPP in the molten state. As shown in Fig. 9.9, TEM pictures of the blends show fairly homogeneous morphology in the iPP/EHR51 at any blend compositions, whereas the iPP/EHR33 shows apparent phase separation, which corresponds to two separate Tg peaks in the dynamic mechanical spectra. The characteristics in the crystalline region of both the blends are summarized in Table 9.1. It is also found from WAXD measurements that the location of the 2u peaks is unchanged by blending EHRs, suggesting that EHRs are excluded from the crystalline region of iPP. Moreover, there is no significant difference in the melting point and crystalline form for all blends. However, the long period, defined as the distance between the centers of the two adjacent lamellae, increases with the blend ratio of EHR51. Assuming the two-layers model, the

232

Polyolefin Blends

Figure 9.9 TEM micrographs for (a) iPP/EHR33 and (b) iPP/EHR51. The samples were stained by ruthenium tetraoxide. (From Reference 24 with permission from John Wiley & Sons, Inc.)

Table 9.1 Sample iPP iPP/EHR33 90/10 80/20 70/30 0/100 iPP/EHR51 90/10 80/20 70/30 0/100

Characteristics of iPP and Blends. Melting temperature, K

Crystal form

Long period, nm

438

a

17.7

438 437 437 No detect

a a a No detect

18.8 19.6 20.5 No detect

438 438 438 No detect

a a a No detect

17.3 18.4 18.4 No detect

[From Reference 24 with permission from John Wiley & Sons, Inc.]

Chapter 9 Morphology and Mechanical Properties in Blends

233

LP

20

c

L , L , L , nm

30

P

a

Lc 10 La 0 0

10

20

30

40

EHR content, wt%

Figure 9.10 LP , Lc, and La plotted against the content of EHR51.

lamellar crystal thickness Lc and amorphous layer thickness La can be evaluated from the data of the volume fraction of crystallinity xv and SAXS long period Lp, employing the relation Lc ¼ xv Lp and La ¼ ð1 xv Þ Lp. In Fig. 9.10, Lc and La are plotted against the EHR content for the iPP/EHR51 blends. It is apparent that blending EHR51 thickens the amorphous layer. These experimental results lead us to conclude that EHR51 is miscible with iPP and the EHR51 is completely incorporated in the amorphous region between adjacent iPP lamellae. Consequently, iPP/EHR51 blends have larger long periods as shown in Table 9.1. The lamellar morphology of iPP/EHR51 is schematically illustrated in Fig. 9.11. It follows that the fraction of tie molecules, defined as the amorphous chains connected with neighbor lamellae, decreases with the increase in the EHR51 content. Since tie molecules are greatly responsible for the resistance to the applied stress for semicrystalline polymers, such types of blends show different mechanical properties from the immiscible blends having the same amount of rubbery polymer. The role of tie molecules in tensile properties is described in Section 9.3.

Figure 9.11 Schematic illustration of the structure for iPP/EHR51. Solid lines represent iPP chains and dotted lines denote EHR chains.

234

Polyolefin Blends

Figure 9.12 TEM micrographs for (a) iPP/EPR37 and (b) iPP/EPR67. The samples were stained by ruthenium tetraoxide. (From Reference 19 with permission from John Wiley & Sons, Inc.)

9.2.2.3 Ethylene–Propylene Copolymer and Ethylene–Octene-1 Copolymer Although the EPRs with a large amount of propylene units are also expected to be miscible with iPP, Lohse found that EPR with 92 wt% (88 mol%) of propylene is immiscible with iPP in the molten state by small-angle neutron scattering experiments (25). Yamaguchi et al. also showed phase-separated morphology in the blends with EPR37 and EPR67 (19) (see Fig. 9.12). The immiscible state in the amorphous region is also consistent with the dynamic mechanical spectra showing double Tg peaks as shown in Fig. 9.13.

(a) 9

10

(b)

iPP/EPR37 β

109

10 Hz

β1

10 Hz

1

β2

2

E", Pa

β E ", Pa

iPP/EPR67

108

107 175 200 225 250 275 300 Temperature, K

108

107 175

200

225

250

275

300

Temperature, K

Figure 9.13 Temperature dependence of tensile loss modulus at 10 Hz for (a) iPP/EPR36 and (b) iPP/ EPR67: ( ) iPP, (#) iPP/EPR (75/25), (^) iPP/EPR (50/50), and (^) EPR. (From Reference 19 with permission from John Wiley & Sons, Inc.)

.

Chapter 9 Morphology and Mechanical Properties in Blends

235

Carriere and Silvis studied the interfacial tension between iPP and ethylene– octene-1 copolymers (EORs) with various amounts of octene-1 (less than 14 mol%) (6). They found that the interfacial tension decreases monotonically from 1.5 mN m1 (2.4 mol % of octene-1) to 0.56 mN m1 (14 mol%). Yamaguchi and Nitta demonstrated that the EOR with 52 mol% of octene-1 is miscible with iPP in the amorphous region whereas the blend with EOR32 having 32 mol% of octene-1 shows phaseseparated morphology (26). 9.2.2.4

Blends of Syndiotactic Polypropylene

According to the theoretical approaches (22,27,28), the miscibility of polyolefin blends is decided by the packing size of molecules and/or the stiffness of molecular chain. Therefore, the stereoregularity of iPP reflecting the chain stiffness (29–31) has an influence on the miscibility with other polyolefins. Thomann et al. studied the morphology of the blends of iPP and syndiotactic PP (sPP) and found that both polymers crystallize in different unit cells and there is no evidence of the interdiffusion of both polymers in the amorphous region (32). Ma¨ier et al. showed that sPP is immiscible with iPP in the molten state while atactic PP (aPP) with M w ¼50 103 is miscible with iPP (33). According to Silvestri and Sgarzi (34), aPP with high Mw ð280 103 to 472 103 Þ was found to be also immiscible with iPP. Mattice et al. showed that sPP is immiscible with iPP and aPP (35). Yamaguchi and Miyata revealed that the miscibility with ethylene–a-olefin copolymers is dependent on the stereoregularity of iPP (31). Figure 9.14 shows the dynamic mechanical spectra in the iPP/EHR57(75/25) blends. The magnitude of E0 of the sPP/EHR57 is lower than that of the iPP/EHR57,

1010

E ', E '', Pa

109

10 Hz E'

108 E" 107 106 105

150 200 250 300 350 400 450 Temperature, K

Figure 9.14 Temperature dependence of dynamic tensile moduli at 10 Hz for iPP/EHR57 (75/25) (closed symbols) and sPP/EHR57 (75/25) (open symbols). (From Reference 31 with permission from American Chemical Society.)

236

Polyolefin Blends (a) 6

6

10

5

10

5

104

10

4

10

3

10

2

10

1

3

10

G', G", Pa

G', G", Pa

(b)

iPP/EHR

10

10

G"

2

10

1

10

G'

0

463K

10

-1

10 -3 10 10-2 10-1 100 101 102 103 -1

ωaT, s

sPP/EHR

G"

G'

100 10

463 K

-1 -3

10

10

-2

10

-1

10

0

10

1

10

2

10

3

-1

ωaT , s

Figure 9.15 Master curves of dynamic shear moduli at 463K for (a) iPP/EHR57 (75/25) and (b) sPP/ EHR57 (75/25). (From Reference 31 with permission from American Chemical Society.)

which is because of the less crystallizability in sPP. Furthermore, the iPP/EHR57 shows only single relaxation peak in the E00 curve in the temperature range between 200 and 330K, indicating that most of EHR chains is incorporated into the amorphous region of iPP. On the contrary, the sPP/EHR57 showing phase-separated morphology on TEM pictures shows two separate peaks in the E00 curve, ascribed to Tg (about 230K) of the pure EHR57 and Tg (about 280K) of the pure sPP (31). Figure 9.15 shows the master curve of dynamic shear moduli for the blends in the molten state at 463K. The slopes of G0 and G00 with frequency in the terminal zone are closed to 2 and 1, respectively. This is due to the fact that these materials have a narrow molecular weight distribution because they are produced by metallocene catalysts. Furthermore, the moduli of iPP/EHR57 are intermediate between those of the individual pure components, indicating that the longest relaxation mechanism is caused by the entanglement slippage in iPP/EHR57. On the contrary, the sPP/EHR57 shows the ‘‘second plateau’’ in the G0 curve at around 102 to 101 s1 , which is ascribed to the long-time relaxation mechanism related to the heterogeneous structure. This also implies that sPP is immiscible with EHR57 in the molten state.

9.2.3 Ethylene–Isotactic Propylene Copolymers Recently, a new technology that manipulates the propylene sequence in EP copolymers has been proposed using a Cr(acac)3/MgCl2-Et2AlCl-ethylbenzoate catalyst system (36). The novel EP random copolymers with high isotacticity in propylene sequence show essentially different properties from those of conventional EPRs (36). This section deals with morphology and dynamic mechanical properties in binary blends of the novel EP copolymers (referred as EP) with various propylene contents.

Chapter 9 Morphology and Mechanical Properties in Blends

237

Figure 9.16 TEM micrographs for the blend films: (a) iPP/EP89 (80/20) blend, (b) iPP/EP84 (80/20) blend, (c) iPP/EP77 (80/20) blend, and (d) iPP/EP53 (80/20) blend. (From Reference 37 with permission from Elsevier Ltd.)

The end numeral of EP used as sample code indicates the propylene content, that is, EP89 contains 89 mol% propylene unit. TEM micrographs of the iPP/EP blends are shown in Fig. 9.16. The iPP/ EP89(80/20) and the iPP/EP84(80/20) blends exhibit homogeneous morphology, suggesting that both EP89 and EP84 are miscible with iPP, whereas the iPP/EP77 (80/20) and iPP/EP52(80/20) films exhibit heterogeneous morphology. According to WAXD studies of the iPP and iPP/EP blends, the diffraction patterns of all the blend samples have a broad amorphous background superimposed upon five sharp diffraction lines ascribed to the a-form (monoclinic) (37). These results indicate that the blending of EP copolymers little affects the crystalline region of iPP irrespective of molecular composition in EP copolymers. The values of La and Lc, evaluated by SAXS and the volume fraction of crystallinity, are plotted against the propylene unit (P-unit) content in EP for the iPP/EP(80/20) blends in Fig. 9.17. The figure suggests that Lc of the miscible blends

Figure 9.17 The crystalline lamellar thickness and the amorphous layer thickness plotted against the propylene-unit content of EP copolymers in iPP/EP (80/20) blends. (From Reference 37 with permission from Elsevier Ltd.)

238

Polyolefin Blends

iPP Miscible

EP

Immiscible

Figure 9.18 Schematic illustration of the lamellar morphology in miscible iPP/EP and immiscible iPP/EP blends. The red line denotes the EP copolymers and the closed dots in the red chains indicate the ethylene units.

is slightly thinner than Lc of the iPP film and the amorphous layer thickness La is larger than Lc. On the contrary, both Lc and La of the immiscible blends are comparable with those of the iPP, being independent of the addition of EP copolymers. These results strongly suggest that in the miscible blends the EP copolymers are partially incorporated into the amorphous layers in iPP, while in the immiscible blends the EP copolymers little affect the crystalline morphology of iPP. As mentioned in Section 9.2.2.3, considering that the EPRs are essentially immiscible with iPP, the high miscibility of EP92, EP87, and EP84 copolymers will be due to their isotactic propylene sequence, which can be included in the iPP crystal lattice. Thus, the miscible state in iPP/EP copolymers is different from the miscible iPP/EHR or EBR copolymers. The morphological features of iPP/EP are illustrated in Fig. 9.18. Figure 9.19 shows the pictures of the spherulite morphology for the iPP and all blends under the optical microscope with crossed polars. As shown in the figure, there are a number of dark points within the spherulites of iPP/EP77 and iPP/EP53 blends (see Fig. 19e and f). Figure 9.20 compares the polarized optical micrographs (POM) of spherulite morphologies obtained by isothermal crystallization at 403K for miscible iPP/EP84 and immiscible iPP/EP77 blends with various blend ratios. Since EP84 and EP77 copolymers have no ability to crystallize at 403K, the micrographs of the blends represent the spherulite morphology of iPP component. It is found that all the iPP/EP84 blends show homogeneous spherulite morphology, whereas dark points are dispersed within the iPP spherulite for all the blends of EP77 and iPP. These spherulite features are independent of the isothermal crystallization temperatures. In addition to the data of TEM pictures, the dark points within the spherulites are associated with isolated domains of the EP copolymers (EP53 and EP77). It follows that the EP copolymers are incorporated in the intercrystalline regions of iPP spherulites in the miscible blends. This morphological feature is in agreement with that of miscible iPP and ethylene–a-olefin copolymer blends as described in Section 9.2.2 (24). In the dynamic mechanical spectra, there are three relaxation processes assigned as a, b, and g in the order of decreasing temperature: the a-process is ascribed to the

Chapter 9 Morphology and Mechanical Properties in Blends

239

Figure 9.19 Optical micrographs of spherulites during crystallization at 403K, (a) iPP, (b) iPP/EP92 (80/20) blend, (c) iPP/EP89 (80/20) blend, (d) iPP/EP84 (80/20) blend, (e) iPP/EP77 (80/20) blend, and (f) iPP/EP53 (80/20) blend. (From Reference 37 with permission from Elsevier Ltd.)

relaxation of crystals, the b-process is to glass transition (Tg) in amorphous parts, and g-process is to the segmental motion of ethylene sequence (36,38,39). As shown in Fig. 9.21a, the magnitude of the a-process of pure EP copolymers markedly decreases with the increase in the ethylene content, resulting from the fact that the a-process is ascribed to the relaxation of crystalline phase of iPP. The temperature of b-process, as the ethylene content increases, shifts from the Tg of iPP to the Tg of a completely amorphous EPR (22). As seen in Fig. 21b, the miscible blends such as iPP/EP92(80/20), iPP/EP89(80/20), and iPP/EP84(80/20) exhibit a single peak of b-relaxation ascribed to the glass transition Tg at the temperature region between Tgs of the iPP and EP copolymers. On the contrary, the iPP/EP77(80/20), iPP/EP70(80/ 20), and iPP/EP53(80/20), showing the heterogeneous morphology, exhibit two separate peaks in the b-relaxation temperature range from 200 to 300K, which are assigned Tb1 and Tb2 in the order of the lower temperature. The relaxation at the higher temperature Tb2 is ascribed to Tg of iPP and the relaxation at the lower temperature Tb1 is to that of the corresponding EP copolymers. These dynamic mechanical behaviors also demonstrated that the EP92, EP89, and EP84 molecules are incorporated in the amorphous region of iPP in the solid state, whereas the EP77,

240

Polyolefin Blends

Figure 9.20 Optical micrographs of spherulites during crystallization at 403K, (a) iPP/EP92 (50/50) blend, (b) iPP/EP89 (50/50) blend, (c) iPP/EP84 (20/80) blend, (d) iPP/EP77 (20/80) blend, (e) EP84, and (f) EP77. (From Reference 37 with permission from Elsevier Ltd.)

(a) 1011

(b)

iPP iPP/EP92(E" × 100)

E " x A , Pa

EP92(E" × 100)

10

9 EP89(E" × 10)

iPP/EP89(E" × 10)

EP84

10

iPP/EP84 iPP/EP77(E" × 0.1)

7

iPP/EP70(E" × 0.01)

10

EP77(E " × 0.1)

5

iPP/EP53(E" × 0.001)

EP70(E " × 0.01)

3

10 100

EP53(E " × 0.001)

200

300

400

Temperature, K

500 100

200

300

400

500

Temperature, K

Figure 9.21 Temperature dependence of mechanical loss modulus (E00 ) at 10 Hz, (a) iPP and EP copolymers and (b) iPP/EP (80/20) blends. (From Reference 37 with permission from Elsevier Ltd.)

Chapter 9 Morphology and Mechanical Properties in Blends

1010

(a) 10

9

10

8

E' / Pa

E' , Pa

10

(100/0)

10

7

(50/50)

10

(0/100)

6

10 10

(b)

9

8

(100/0)

7

(20/80)

(0/100)

10 9 10

(80/20)

10 10

8

10

10

7

10

E" / Pa

E" , Pa

10

(100/0) (50/50)

6

10

10 10

(20/80)

100

(20/80)

300

400

Temperature, K

(50/50)

9

8

7

(100/0) (80/20)

6 (0/100)

(80/20)

200

(80/20)

6

(0/100)

105

241

5

10 500 100

200

300

(20/80) (50/50)

400

500

Temperature, K

Figure 9.22 Temperature dependence of mechanical storage modulus (E0 ) at 10 Hz, (a) iPP and iPP/ EP84 blends and (b) iPP and iPP/EP77 blends. (From Reference 37 with permission from Elsevier Ltd.)

EP70, and EP53 copolymers are immiscible with the iPP phase and the immiscible blends show phase-separated morphology in the solid state. Figure 9.22a and b compares the dynamic mechanical spectra of miscible iPP/EP84 and immiscible iPP/EP77 blends. The iPP/EP84 blends show a single peak of b process at the temperature region between Tgs of the pure components (at 245–275K), being independent of the EP84 contents. This suggests that EP84 chains are incorporated in the amorphous region of iPP. On the contrary, all iPP/EP77 blends show two separated peaks or single peaks in the lower temperature regions; these peaks are located at the Tgs of the pure component, that is, Tg = 275K for the iPP and Tg = 238K for the EP77, being independent of the EP77 content. The EP84 molecules are incorporated into the iPP phase irrespective of the EP content, resulting in trapping the EP portions into the interlamellar regions of iPP at any blend ratio. In the iPP/EP77 blend, there are two separated Tg peaks corresponding to pure iPP and the EP copolymer and this immiscible feature in mechanical spectra is independent of the content of the EP copolymer. The morphological features in the TEM and POM observations support these dynamic mechanical behaviors in the iPP/EP binary blends.

9.3 TENSILE AND RHEO-OPTICAL PROPERTIES 9.3.1 Principles for Rheo-Optical Characterization Mechanical properties in the blends of polyolefin materials and elastomers are of considerable importance for engineering applications (40,41). Various elastomers

242

Polyolefin Blends

such as EPR, EPDM, and butylrubber are used widely for improving the mechanical brittleness of iPP materials (42–45). This section compares the tensile properties of miscible and immiscible blends of iPP and various polyolefin copolymers. In addition, in order to examine the deformation mechanism, we present the results on the rheo-optical measurements of these blend materials. Before going to main topics, the background information on rheo-optics needed for this section is briefly described below. Rheo-optical techniques (46–48) afford information on the strain dependence not only of stress but also of optical quantities associated directly with the structure or molecular morphology. The techniques were developed extensively for crystalline polymers to investigate the molecular deformation mechanism underlying the tensile elongation. In this part, the chain orientation behavior is characterized by infrared dichroism measured simultaneously with tensile deformation at a constant rate of elongation. A tensile tester was set in Fourier-transform infrared spectrometer in such a way so as to allow infrared beam through a film specimen mounted on the tensile tester. The tensile tester was specially designed for upper and lower clamps to symmetrically move from the central point of the film so that the beam spot remains at the initial position during a whole stretching. In this work, the intensities of 998 cm1 band, which is associated with CH3 rocking mode as coupled with C–CH3 stretching mode, were measured as a function of the elongation time every 10 s. Using the dichroic ratio of the band, we estimated the orientation function of crystal c-axis according to the literature (49).

9.3.2 Blends with Ethylene–a-Olefin Copolymers The stress–strain curves of iPP and iPP/EHR blends at room temperature, which is above Tg of all samples, are shown in Fig. 9.23. The tensile behavior of iPP is 50 298K

Stress, MPa

40

iPP iPP/EHR33(90/10) iPP/EHR33(70/30) iPP/EHR51(90/10) iPP/EHR51(70/30)

30

20

10

0 0

0. 2

0. 4 0. 6 Strain

0. 8

1. 0

Figure 9.23 Stress–strain curves for iPP and iPP/EHR blends at 298K.

Chapter 9 Morphology and Mechanical Properties in Blends

243

80 253K

Stress, MPa

60

40

20

iPP iPP/EHR33(90/10) iPP/EHR33(70/30) iPP/EHR51(90/10) iPP/EHR51(70/30)

0 0

0. 2

0. 4 0. 6 Strain

0. 8

1. 0

Figure 9.24 Stress–strain curves for iPP and iPP/EHR blends at 253K.

characterized as a well-defined yielding and neck formation with the stress whitening at the strain around 0.3. The immiscible blends such as iPP/EHR33 immediately show the intense stress whitening, whereas the miscible iPP/EHR51 blends show no stress whitening and more ductile behavior. The tensile curves measured at 253K are shown in Fig. 9.24. The temperature is above Tg of EHRs but is below Tgs of iPP and/or iPP phase in the immiscible blends. The pure iPP and immiscible iPP/ EHR33(90/10) exhibit brittle behavior at 253K, which is below Tg of iPP phase, while the miscible iPP/EHR51 blends is ductile because their Tgs are more than the Tg of the miscible blends. The iPP/EHR33(70/30) with high content of EHR shows the same manner as its tensile behavior at room temperature because of its large rubbery EHR domains. Figure 9.25 shows the stress–strain curves at 233K, which is 80 233K

Stress, MPa

60

40

20

iPP iPP/EHR33 (90/10) iPP/EHR33 (70/30) iPP/EHR51 (90/10) iPP/EHR51 (70/30)

0 0

0. 2

0. 4 0. 6 Strain

0. 8

1. 0

Figure 9.25 Stress–strain curves for iPP and iPP/EHR blends at 233K.

244

Polyolefin Blends

Table 9.2

Temperature Dependence of Tensile Behavior.

Temperature, k 233 243 253 273 298 323 353

iPP

iPP/EHR33 (90/10)

. . . . 4 4

. . . 4 4 4

iPP/EHR33 (70/30)

iPP/EHR51 (90/10)

.

. .

4 4 4 4 4

4 4 4

IPP/EHR51 (70/30)

.

. brittle with stress whitening. 4 ductile with stress whitening. ductile without stress whitening.

below Tgs of both phases. All the samples display brittle behavior with stress whitening. The temperature dependence of tensile behavior is summarized in Table 9.2. In this table, open circles represent ductile behavior without stress whitening, and closed circles brittle behavior with stress whitening. Bold lines in the table denote Tg. The table shows that the ductile-brittle transition in the immiscible blends is independent of temperature, suggesting that the fracture will be associated with the boundary separation at the interface between iPP matrix and EHR domains. However, the EHR51 material plays a role of a plasticizer for iPP because the addition of EHR51 lowers the Tg region. This will be because of the incorporation of the EHR molecules into the amorphous region of iPP. The similar results were obtained in the iPP/EBR blends (50). The orientation functions of chain axis as well as the stress are plotted against the strain in Fig. 9.26a–c. As seen in Fig. 9.26a, the orientation function of the iPP slightly decreases with increasing strain up to around the yield point that corresponds to the onset strain of the stress whitening. The orientation functions of both the 998 and 974 cm1 bands monotonously increase beyond the minimum point. The orientation function of both bands in the miscible blend (iPP/EHE51) shows a great decrease and then increases after passing a minimum of around 1.0 (see Fig. 9.26b). The depth and location of the iPP/EHR51 are considerably larger than those of the iPP. The yield peak is more diffuse and poorly defined. The negative orientation of chain axis (c-axis) in the yielding region can be explained by the affine deformation of spherulites (51,52): the deformation from sphere to ellipsoid causes a positive orientation of lamellae (a* or b-axis) within the deformed spherulites and this leads to a negative orientation of c-axis because the c-axis is perpendicular to the lamellar axis. Consequently, the great decrease of chain orientation in iPP/EHR51 suggests that the ellipsoidal deformation of spherulite occurs without plastic deformation, microvoids, and crazes due to the lamellar fragmentation. This result is also supported by the experiments of the

245

Chapter 9 Morphology and Mechanical Properties in Blends (a) 0.10

iPP

(b) iPP/EHR51 (70/30) 0.10

40

20 –1

973 cm 998 cm–1 Stress, MPa

Stress

10

F

F

20 Stress

0.05

0

0

5

10 –1

F –0.05

0

0. 5

Stress, MPa

15

30 0.05

973 cm 998 cm–1 1. 5

1.0

F 2. 0

-0.05

0

0

0. 5

1. 0

1. 5

0 2. 0

Strain

Strain

(c) iPP/EHR33 (70/30) 0.10

20 973 cm–1 998 cm–1

Stress

F

10 0

Stress, MPa

15 0.05

5 F -0.05 0

0. 5

1. 0 Strain

1. 5

0 2. 0

Figure 9.26 Strain dependence of orientation functions and stress for (a) iPP, (b) iPP/EHR51(70/30), and (c) iPP/EHR33 (70/30). The open and closed circles denote the orientation function of crystal and amorphous chains.

simultaneous measurements of tensile testing and small-angle light scattering. The light scattering pictures were taken using a photographic camera under the polarization condition in which polarization direction is vertical with an analyzer in horizontal direction (HV scattering). Figure 9.27 shows the HV scattering pattern of iPP/EHR51(70/30) during stretching, indicating the ellipsoidal deformation of spherulites (53). The lack of the plastic deformation, cracks and voids, is consistent with the fact that the iPP/EHR51 shows no stress whitening. This is plausible because the rubbery EHR chains are incorporated into the amorphous region of iPP, leading to decrease in Tg of amorphous iPP chains and reduction of the number of tie molecules

246

Polyolefin Blends

Figure 9.27 HV light scattering pattern during stretching for iPP/EHR51 (70/30) blends.

plays an important role in the fragmentation of lamellae. In contrast, the immiscible iPP/EHR33 blend shows no orientation of both crystal and amorphous chains as shown in Fig. 9.26c. These results suggest that macroscopic interfacial separation occurs between iPP matrix and EHR domains in the initial strain region. The interfacial separation can be caused by the difference in Poisson ratio between iPP and EHR33. Bedia et al. also studied the deformation mechanism of immiscible iPP and EHR32 blends by AFM and found that the specimens with 30 wt% of EHR32 were broken fractured at the interface between both phases at low strain region (54). Yamaguchi and Nitta revealed that the structural defects such as crazes and cracks occur significantly in the immiscible blends, which was quantitatively evaluated by the attenuation of ultrasonic wave (55). According to a statistical treatment proposed by Huang and Brown (56), the formation of tie molecules depends on whether the chain dimensions are more than the distance between adjacent lamellae with the amorphous region or not. Consequently, the tie-molecule fraction can be controlled by addition of a miscible EHR into an iPP because such molecules are completely dissolved in the interlamellar region of iPP. The yield stress of the miscible blends is found to increase linearly with the tiemolecule fraction (see Fig. 9.28). The linearity suggests that tie molecules act as stress transmitters and support the external force required for the lamellar fragmentation, which takes place on the yielding process. It is interesting to note that there exists extra tie molecules extrapolated to the yield stress s y = 0. The extra tie molecules, that is, inactive tie molecules, have no contribution to the stress transmitters and are considered to bind several crystalline lamellae together with the interlamellar amorphous layers to form lamellar clusters (57–59) and the remainder of tie molecules connects adjacent lamellar clusters. The external force through the intercluster links composed of active tie molecules connecting between adjacent clusters decomposes the lamellar clusters into their fragments at the yield point.

Chapter 9 Morphology and Mechanical Properties in Blends

247

Yield stress, MPa

40

30

20

10

0 0.1

0.2

0.3

Tie molecule fraction

Figure 9.28 Yield stress versus tie molecule fraction of iPP/EHR blends. (From Reference 57 with permission from Elsevier Ltd.)

9.3.3 Blends with Novel Ethylene–Isotactic Propylene Copolymers Binary blends of iPP and EPR copolymers have been extensively investigated from the commercial point of view. Although the control of compatibility plays an important role in the improvement of mechanical properties in the iPP/EPR blend systems, the strong incompatibility in iPP and EPR blends limits the modifications of morphology and mechanical properties (9). This section deals with the tensile properties in binary blends of iPP and the novel EP copolymer having high stereoregularity of propylene sequence described in Section 9.2.3. As mentioned before, EP92, EP89, and EP84 are miscible with iPP, while EP77, EP70, and EP53 are immiscible and the blends show phase-separated morphology. Figure 9.29a and b shows the yielding region of stress–strain curves of all the iPP/EP(80/20) blends and pure iPP at room temperature. The iPP sample shows a well-defined yield peak and formed a neck in the postyield region. The overall stress in the yield region for all blends decreases with the decrease in bulk density or crystallinity of the samples in the same manner with the previous mechanical data (60,61). Furthermore, the ductility is pronounced in the miscible blends, and the stress whitening occurs in a higher strain region as compared with iPP. The ductility will be due to the fact that EP copolymers are incorporated into the amorphous region, and they act as a plasticizer as shown in miscible iPP/EBR and iPP/EHR blends (60). On the contrary, for the immiscible blends the stress whitening occurred in a lower elongation region. This will be associated with the separation at the interface between matrix and dispersed domains (50). Figure 9.30 compares the effects of blend ratio on stress–strain curves of miscible iPP/EP84 and immiscible iPP/EP77 blends. The pure EP84 and EP77 samples show no yield peaks and no stress whitening in this experimental condition. The increase in the amount of EP84 and EP77 copolymers lowers the yield stress of

248

Polyolefin Blends

Figure 9.29 Stress–strain curves at 298K and 20 mm min1 : (a) iPP, iPP/EP92 (80/20) blend, iPP/ EP89(80/20) blend, and iPP/EP84(80/20) blend and (b) iPP/EP77 (80/20) blend, iPP/EP70 (80/20) blend, and iPP/EP53 (80/20) blend. (From Reference 37 with permission from Elsevier Ltd.)

the blends. However, all the iPP/EP84 blends show a yield peak and stress whitening at higher elongation region. On the contrary, the immiscible blends of iPP/EP77 (50/50) and (20/80) have no clear yield peaks. These results demonstrate that the miscibility somewhat affects the yield behavior. In Fig. 9.31, the yield energy or resilience of iPP/EP(80/20) blends is plotted against the propylene content of EP. The yield energy of these blends and iPP can be estimated from the area under the stress–strain plot from the origin to the stress drop. It was found that the yield energy of the miscible blends is greater than that of iPP and immiscible blends. Thus, the addition of EP92, EP89, and EP84 toughens the spherulite structure composed of iPP lamellar crystals. Conventional EPRs showing poor crystallizability are essentially immiscible with iPP, irrespective of P-unit content in the EPRs as demonstrated in Section 9.2.2.3. In the present EP copolymers, the length and number of crystallizable sequences progressively increases with the decrease in the ethylene-unit content according to

(a)

40

iPP

20

iPP/EP84 (20/80)

10

iPP

30

iPP/EP84 (80/20) iPP/EP84 (50/50)

Stress, MPa

Stress, MPa

30

(b)

40

iPP/EP77 (80/20) iPP/EP77 (80/20)

20

iPP/EP77 (80/20)

10

EP77

EP84

0

0

0.4

0.8

1.2

Strain

1.6

2

0

0

0.4

0.8 1.2 Strain

1.6

2

Figure 9.30 Stress–strain curves at 298K and 20 mm min1 : (a) iPP and iPP/EP84 blends and (b) iPP and iPP/EP77 blends. (From Reference 37 with permission from Elsevier Ltd.)

Chapter 9 Morphology and Mechanical Properties in Blends

249

Yield energy, MJ m–3

15 Miscible Immiscible 10

5

iPP

0 50

60

70

80

90

100

P-unit cont. in EP, mol % Figure 9.31 Yield energy of the blends versus the propylene-unit content (mol%) in EP copolymer:

.

(&) iPP; ( ) miscible blends such as iPP/EP92 (80/20) blend, iPP/EP89 (80/20) blend, iPP/EP84 (80/ 20) blend; (#) immiscible blends such as iPP/EP77 (80/20) blend, iPP/EP70 (80/20) blend, iPP/ EP53 (80/20) blend. (From Reference 37 with permission from Elsevier Ltd.)

Shin et al. (36). Therefore, in the miscible blends the isotactic propylene sequence in the EP copolymer chains having a relatively high P-unit contents such as EP92, EP89, and EP84 is considered to be capable of participating in the crystallization process of iPP during solidification, resulting in that the EP copolymer chains are incorporated partly in the crystal lattice and partly in the amorphous region. Consequently, the EP92, EP89, and EP84 molecules act as the additional tie molecules linking between adjacent lamellae and this would lead to the enhancement of the yield toughness of the spherulitic structure. The EP chain portions that are not available to participate in the crystallization process are trapped into the interlamellar region, leading to the reduction of Tg of iPP as well as the increment of amorphous layer thickness (see Table 9.3). Table 9.3 Structural Parameters of iPP and iPP/EP Blends. Blend samples

Tga, K

xvb, %

iPP 275,0 iPP/EP92(80/20) 269.8 iPP/EP89(80/20) 267 3 iPP/EP84(80/20) 263.7 iPP/EP77(80/20) 238.3(Tb1) 274.3(Tb2) iPP/EP70(80/20) 223.5(Tb1) 277.9(Tb2) iPP/EP53(80/20) 215.1(Tb1) 275.1 (Tb2)

56.3 45.3 45.7 47.9 54.6 54.9 56.3

LF c, nm

Lc d, nm La e, nm

10.4 11.46 11.49 11.52 11.29 11.34 11.36

[From Reference 37 with permission from Elsevier Ltd.] a

Glass-transition temperature determined by dynamic mechanical tests.

b c

d e

Degree of crystallinity in volume fraction estimated using density data.

Long period evaluated by SAXS. Crystalline lamellar thickness.

Amorphous layer thickness.

5.86 5.19 5.25 5.41 6.16 6.23 6.40

4.54 6.27 6.24 6.11 5.13 5.11 4.97

Density, kg m3 900 898 898 897 896 892 891

250

Polyolefin Blends

The morphological feature is in accordance with that of binary blends of iPP and true diblock iPP-b-EPR copolymers (63). For the immiscible blends showing phase separation, the EP domains have no ability to affect the lamellar or crystalline morphology of iPP. It follows that the strength of iPP lamellae is essentially independent of the addition of EP copolymers although the magnitude of bulk stress is reduced by the addition of EP copolymers with a lower modulus.

9.4 SOLIDIFICATION PROCESS AND FINAL MORPHOLOGY 9.4.1 Morphology Formation During Crystallization It is very important to study the solidification process in the miscible blends considering that the miscibility in the molten state dominantly affects the final morphology. In the actual processing, the solidification process, such as cooling condition and flow field, affects the structure in the solid state and thus the mechanical properties to a great extent. The ethylene–a-olefin copolymer chains that are miscible in the molten state also are excluded out from the crystalline region of iPP as described in Section 9.2. Therefore, the applied crystallization condition affects not only the characteristics of crystalline region such as spherulite texture, degree of crystallinity, and the defects in crystals but also the molecular aggregation state of iPP and the copolymers, which will play a central role in controlling the mechanical properties. In this part, the effect of crystallization temperature on the morphology is discussed by means of the dynamic mechanical properties (24,64) that are important informations on the characterization of injection-molded products as described in Section 9.4.2 (65). Figure 9.32 shows the temperature dependence of dynamic tensile moduli of miscible iPP/EHR51 and immiscible iPP/EHR33 isothermally crystallized at 403K for 1 h. The magnitude of E0 in the iPP/EHR33 is slightly larger than that in the iPP/ EHR51 with the same blend composition over the temperature range from 240 to 380K. This tendency becomes pronounced as the EHR fraction increases. The higher E0 of the iPP/EHR33 samples will be due to the continuity of iPP phase having a higher modulus. The iPP/EHR51 blends show broad and ambiguous double b-relaxation peaks in the E00 curves. The location of the b peak in the higher temperature, which corresponds to Tg of iPP component of the iPP/EHR51 blends slightly shifts to lower temperatures than that of the pure iPP, and the b peak in the lower temperature, which corresponds to Tg of EHR phase, shifts to higher temperatures than Tg of the pure EHR51. This result indicates that EHR51 molecules and iPP chains in amorphous region partially dissolve in each other. Furthermore, the dynamic mechanical spectra of the iPP/EHR51 prepared at 403K are quite different from that of the sample quenched at 273K showing a single peak between Tgs of the pure components (see Fig. 9.8b). The difference in the shape of b-relaxation peaks between the samples quenched in ice-water bath (at 273K), and the samples

251

Chapter 9 Morphology and Mechanical Properties in Blends (a)

(b)

(70/30)

10 iPP/EHR

10

2

10

1

10

0

10

-1

10

-2

10

(80/20)

10 Hz E'

10

β 8

10

α

tan d

E ', E", Pa

10

E'

1

10

α

10

0

10

–1

10

–2

E"

10

tan d

6

200

10 β

8

E" 7

10

2

10 Hz 9

E ', E ", Pa

9

10

tan d

10 iPP/EHR

10

300

400

7

10

tan d

6

10

200

Temperature, K

(c) 10

10

300

400

Temperature, K

iPP/EHR (70/30)

10

2

10

1

10

0

10

-1

10

-2

9

E', E", Pa

10

E' β

8

10

α

tan d

10 Hz

E" 7

10

tan d

6

10

200

300

400

Temperature, K

Figure 9.32 Variation of the dynamic tensile moduli with temperature for iPP/EHR crystallized at 403K. The blend ration of iPP/EHR (w/w) is (a) 90/10, (b) 80/20, and (c) 70/30. The open symbols denote iPP/EHR51 and filled symbols denote iPP/EHR33. (From Reference 64 with permission from Society of Rheology Japan.)

crystallized isothermally at 403K demonstrates that molecular aggregation state in the amorphous region for the iPP/EHR51 samples depends largely on the crystallization condition. On the contrary, iPP/EHR33 samples quenched at 273K show two separate peaks at temperatures between 200 and 300K in the Tg regions. The b relaxations in the iPP/EHR33 crystallized at 403K are identical with those of the samples quenched at 273K as shown in Fig. 9.8a. Compared to iPP/EHR51, the crystallization condition has less effect on the structure in iPP/EHR33.

252

Polyolefin Blends

The a-relaxation process in the iPP/EHR51 samples is located at higher temperatures than that for the iPP/EHR33, suggesting that blending of EHR51 affects the crystalline structure of iPP, despite the fact that no EHR51 molecules are incorporated into the crystal lattice. Details of a- and b-relaxation processes are discussed below. 9.4.1.1 a-Relaxation Process It is well known (66) that the a-relaxation process of crystalline polymers consists of at least two processes, referred to as a1 and a2 in the order of lower temperature, respectively. The a1-process (67–77) is pronounced in melt crystallized samples and is associated with the relaxation of grain boundaries, such as dislocation of lamellae with a frictional resistance related to disordered interface layers. The magnitude of the a1-process increases with the increase in the crystal defects. The a2-process (71,73,78–83) is pronounced in single crystal mats and is ascribed to incoherent oscillations of the chains about their equilibrium positions in the crystal lattice in which intermolecular potential suffers smearing out. The magnitude of the a2-process increases with the increase in the lamellar thickness and/or the degree of crystallization (39). Figure 9.33 shows the temperature dependence of tan d for the iPP samples crystallized at various temperatures. There is a broad peak ascribed to a-relaxation process in the temperature region between 300 and 430K. Furthermore, the location of a-relaxation shifts to higher temperatures and the magnitude reduces with the increase in the crystallization temperature. The increase in crystallization temperature causes well-organized crystals, which leads to significant lowering of the magnitude of the a1-process and enhances the a2-process located at higher temperature. iPP

tan d + A

0.20

Tc:413K A=0.125

10 Hz

Tc:408K A=0.1 Tc:403K A=0.075

0.15

Tc:398K A=0.05 Tc:393K A=0.025

0.10

Tc:388K A=0

0.05

300

350

400

450

500

Temperature, K

Figure 9.33 Temperature dependence of tan d around the a-relaxation process for iPP crystallized at various temperatures. (From Reference 64 with permission from Society of Rheology Japan.)

Chapter 9 Morphology and Mechanical Properties in Blends (b)

(a) iPP/EHR33 Tc 403K

0.20

0.25 EHR 30% A=0.075

EHR 20% A=0.05

0.15

EHR 10% A=0.025

0.10

i-PP A=0

Tc 403K

EHR 30% A=0.075

EHR 20% A=0.05

0.15 EHR 10% A=0.025

0.10

i PP A=0

0.05

0.05

0

iPP/EHR51

0.20

tan d + A

0.25

tan d + A

253

300

350 400 450 Temperature, K

500

0 300

350 400 450 Temperature, K

500

Figure 9.34 Temperature dependence of tan d around the a-relaxation process for the blends of iPP with (a) EHR33 and (b) EHR51 crystallized at 403K. (From Reference 64 with permission from Society of Rheology Japan.)

Figure 9.34 illustrates the compositional dependence of a-relaxation of both blends crystallized at 403K. There is no measurable shift of the location of a-process by blending of EHR33, indicating that EHR33 molecules have little influence on the crystallization of iPP. This result is plausible since EHR33 is immiscible with iPP even in the molten state as shown in Section 9.2.2.2 and the crystallization rate of iPP is independent of blending EHR33 as shown in Fig. 9.7. On the contrary, the temperature location of the a-relaxation in the iPP/EHR51 samples fairly shifts to higher temperatures with the increase in the EHR51 fraction. As the crystallization temperature increases, the peak of a-relaxation of iPP/EHR51(90/10) shifts to a higher temperature in the same manner with iPP (see Fig. 9.35), indicating that the a2-process is enhanced in the iPP/EHR51(90/10) crystallized at higher crystallization temperatures. The increase in crystallization temperature causes well-organized crystals and thicker crystalline lamellae, leading to the significant lowering in the magnitude of a1-process and to a higher temperature shift and enhancement of a2-process. The enhancement of the a2-process results from the well organization of crystalline lattice caused by the decrease in the crystallization rate. 9.4.1.2

b-Relaxation Process

Figure 9.36 shows the variation of E00 with temperature around the b-relaxation process for iPP, EHR51, and their blends crystallized at 403K. The peak temperatures of the iPP and the EHR51 are 278 and 222K, respectively. Furthermore, the E00 curves in the blends show ambiguous double peaks in the temperature between 200 and 300K. The peak located at lower temperature in the b-process, which is ascribed to Tg of EHR51-rich region, becomes clear and shifts to lower tempera-

254

Polyolefin Blends 0.25

iPP/EHR51 10 Hz Tc:408K A=0.075

tan d + A

0.20

Tc:403K A=0.05

0.15

Tc:398K A=0.025

0.10 Tc:393K A=0

0.05

0 300

350

400

450

500

Temperature, K

Figure 9.35 Temperature dependence of tan d around the a-relaxation process for iPP/EHR51 (90/10) crystallized at various temperatures. (From Reference 64 with permission from Society of Rheology Japan.)

tures as the EHR51 fraction increases. The peak located at higher temperature, ascribed to Tg of amorphous iPP-rich region, slightly shifts to higher temperatures as the iPP fraction increases. Figure 9.37 shows the temperature dependence of E00 for the iPP/EHR51 crystallized at various temperatures. As seen in Figure 9.37a, the iPP/EHR51(90/10) crystallized at 393K shows only single peak due to the glass transition between those of the pure components. This result well agrees with those of the samples quenched at 273K. On the contrary, there are broad and ambiguous iPP/EHR51

10

11

10

10

10

9

10

8

10 Hz

Tc 403K

A=5

4

E "×A, Pa

i-PP A=5

3

10% EHR 2

A=5 20% EHR A=5 30% EHR

EHR A=1

200

250

300

350

Temperature, K

Figure 9.36 Temperature dependence of E00 around the b-relaxation process for iPP/EHR51 with various blend ratio. The samples are crystallized at 403K. (From Reference 64 with permission from Society of Rheology Japan.)

255

Chapter 9 Morphology and Mechanical Properties in Blends (b)

(a)

10

11

iPP/EHR51(90/10)

11

10

iPP/EHR51(80/20) 10 Hz

10 Hz 10

10

10

10

A=53

A=5

10

9

A=52

Tc : 403K

9

10

8

10 A=1

A=1

Tc : 393K 7

7

150

A=5

8

Tc : 393K

10

A=5 2

Tc : 403K Tc : 398K

A=5

Tc : 398K

10

3

Tc : 408K

E"×A, Pa

E "×A, Pa

Tc : 408K

10

200

250

300

350

150

200

250

300

350

Temperature, K

Temperature, K

Figure 9.37 Temperature dependence of E00 around the b-relaxation process for iPP/EHR51 (a) 90/10 and (b) 80/20. (From Reference 64 with permission from Society of Rheology Japan.)

double peaks for the iPP/EHR51 (90/10) crystallized at higher temperatures and for the iPP/EHR51 (80/20). The results demonstrate that the blend composition and/or the crystallization temperature have significant influence on the molecular aggregation state in the amorphous region for iPP/EHR51 blends. According to Stein et al. (84), the diffusion range of the noncrystallizable rubbery chains during the crystallization of crystallizable chains for miscible blends with semicrystalline polymer matrix can be estimated using the parameter dP (72) as follows: dP ¼

Dz G

ð9:1Þ

where Dz is the z-average diffusion coefficient for noncrystallizable chains and G is the spherulite growth rate. The parameter dP represents a distance that the component segregated from the crystalline regions may move during the crystallization. The b-relaxation process shown in Figs. 9.36 and 9.37 for the iPP/EHR51 blends can be discussed in terms of the parameter dP and the long period, defined as the distance between the centers of two adjacent lamellae. The diffusion coefficient Dz in Equation 9.1 can be estimated by the following equation on the basis of the tube model (85): Dz ¼

L2 N 2 a2 ¼ 2 2 p t1 p t1

ð9:2Þ

where t1 is the reptation time that corresponds to the maximum relaxation time and L is the length of the primitive chain, which is given by the product of the number of

256

Polyolefin Blends

Table 9.4

Comparison of the Parameter dP and Long Period LP of iPP/EHR51 Blends. Crystallization temperature, K

iPP/EHR, w/w 90/10 80/20 70/30

dP , LP nm

393

398

403

408

dP LP dP LP dP LP

1.3 34 1.6 36 2.5 35

3.8 34 4.6 36 6.8 35

13 34 15 36 21 35

35 36 38 40 51 36

[From Reference 64 with permission from Society of Rheology Japan.]

entanglement points N and the distance between the neighbor ones a. Considering that EHR51 is miscible with iPP in the molten state and shows no specific interaction reducing the number of entanglement points (23,24), N2 and a2 can be expressed as the following equations (86): N 2 ¼ fPP

MwPP MzPP MwEHR MzEHR þ fEHR 2 MePP MeEHR 2

a2 ¼ fPP hR2 iPPMe þ fEHR hR2 iEHRMe

ð9:3Þ ð9:4Þ

where fi are the volume fraction, Mwi and Mzi are the weight and z-average molecular weights, Mei are the molecular weights between the entanglement points, and hR2 ii-Me are the mean-square distance between the entanglement points for the i-component. Dz can be calculated using the following experimental values; Me of the EHR is 1:0 104 (23), and the flow activation energy of the blends is 40 kJ mol1 (24). Furthermore, Me of iPP is taken to be 4:65 103 , which is the value of atactic polypropylene (87). hR2 ii-Me is calculated using the Discover (BIOSM) by the rotational isomeric state model. The values of the interlamellar distance and the parameter dP using experimental values (24,65) are summarized in Table 9.4. It is noted that the magnitude of dP and LP in the table have some error because the value of Me and LP in the blend samples cannot be determined precisely by the present experimental techniques. Figure 9.38 shows the values dP and LP plotted against the crystallization temperature. The values of dP greatly increase with the increase in the crystallization temperature and/or EHR51 fraction, whereas the long period is less affected by the crystallization temperature and/or EHR51 fraction. These results suggest that most EHR51 molecules will reside in the interlamellar region for the blends crystallized at low temperatures. This is consistent with the experimental results that these blends will show only single peak in the b-process. As the crystallization temperature increases and/or EHR51 fraction increases, EHR51 molecules have a time to diffuse out of the interlamellar region so that EHR51 molecules agglomerate within the space between lamellar stacks or lamellar clusters, in which the concentration of EHR51 is higher than the other amorphous region.

10

3

10

2

L 10

1

10

0

257

P

P

d , LP , nm

Chapter 9 Morphology and Mechanical Properties in Blends

d

P

90/10 80/20 70/30 10

–1

390

395

400

405

410

Tc, K

Figure 9.38 Relationship between crystallization temperature and the parameter dP and long periods LP. (From Reference 64 with permission from Society of Rheology Japan.)

These blends will show ambiguous double peaks in the b-process, which is qualitatively consistent with the results shown in Figs. 9.36 and 9.37.

9.4.2 Structure and Properties of Injection-Molded Products Flow field and cooling condition between injection molding and compression molding are essentially different. In particular, the flow-induced molecular orientation and/or the distorted shape of the dispersed phase have to be considered seriously in the injection molding. This part deals with the relation between morphology and mechanical properties in the injection-molded products for iPP and iPP/EHR blends (65). This is directly important for the industrial application. The characteristics of these samples are summarized in Table 9.5. The POM pictures of iPP, immiscible iPP/EHR30(70/30), and miscible iPP/EHR53(70/30) are shown in Fig. 9.39. These materials are produced by an Table 9.5 Characteristics of Injection-Molded Samples. Sample iPP iPP/EHR30 iPP/EHR53

K-value

Thickness of skin layer, mm

F-value

Tm in core layer, K

xw in core layer, %

4.4 3.3 3.3

0.36 0.36 0.38

0.15 0.10 0.17

436.6 436.7 437.9

49.7 50.0 52.6

[From Reference 65 with permission from John Wiley & Sons, Inc.]

258

Polyolefin Blends

Figure 9.39 Skin–core structure in the injection-molded products for the iPP and blends. (From Reference 65 with permission from John Wiley & Sons, Inc.)

injection-molding machine with a film-gated 70 150 2 mm plate mold. All sample specimens display distinct skin–core structure and the thickness of skin layer is almost the same among the samples (see Table 9.5). According to the TEM micrographs in the core region of the blend samples (see Fig. 9.40), the injection-molded iPP/EHR53 sheet shows homogeneous morphology, corresponding to the structure described in Section 9.2.2. On the contrary, the dispersed EHR domains with about 100–300 nm of diameter, stained by RuO4, are observed in the iPP/EHR30 blend. Furthermore, they also include thread-like domains of EHR, suggesting that a part of EHR phases is separated into thin domains because of the shear applied in the mold. The TEM micrographs of the skin layer show homogeneous morphology in the iPP/EHR53, whereas the thin layers of EHR phase orient to the flow direction in the iPP/EHR30 blend as shown in Fig. 9.41. The distribution of birefringence in the thickness direction is depicted in Fig. 9.42. The birefringence values in the skin layer of the iPP/EHR53 are slightly

Figure 9.40 Morphology in the core region for iPP/EHR30 and iPP/EHR53. The arrow indicates the flow direction. (From Reference 65 with permission from John Wiley & Sons, Inc.)

Chapter 9 Morphology and Mechanical Properties in Blends

259

Figure 9.41 Morphology in the skin region for iPP/EHR30 and iPP/EHR53. The arrow indicates the flow direction. (From Reference 65 with permission from John Wiley & Sons, Inc.)

higher than those of the iPP and the iPP/EHR30. Furthermore, the orientation functions F of c-axis of iPP crystallites determined from the infrared dichroism are shown in Table 9.5. The data obtained from the dichroism are consistent with the data from the birefringence, suggesting that the skin layer in the iPP/EHR53 exhibits a higher orientation of c-axis of iPP as compared with the iPP and the iPP/EHR33. The injection-molded sheets sometimes include the hexagonal b-crystal form. The fraction of b-form can be estimated from the relative intensity (K-value) of X-ray diffraction peaks of b-crystal according to Turner Jones et al. (88). As seen in

15

iPP/EHR53 –3

10

n x 10

iPP

iPP/EHR30

5

Core

Skin

0 0 Surface

0.1

0.2

0.3

0.4

0.5

Distance, mm

Figure 9.42 Distributions of birefringence (Dn) in the thickness direction for iPP (closed diamonds), iPP/EHR30 (closed circles), and iPP/EHR53 (open circles). (From Reference 65 with permission from John Wiley & Sons, Inc.)

260

Polyolefin Blends

Table 9.5, the K-values of blend samples are lower than that of iPP, indicating that blending of a rubbery EHR suppresses the formation of b-form crystals of iPP, irrespective of the miscibility in the molten state. The weight fraction of crystallinity xw of iPP phase in the core region of the iPP/EHR53 is larger than those of the iPP and the iPP/EHR30. The well-organized crystalline lattice proceeds to the core region owing to the enhancement of molecular mobility resulting from the cooperative motion of iPP and EHR53 chains during crystallization. The injection-molded sheets exhibit the mechanical anisotropy in the flow direction (MD) and its perpendicular direction (TD). Figure 9.43 compares the

(b)

10 iPP

2

10 iPP/EHR30(90/10)

8

0 E"

7

log [E ', Pa], log [E ", Pa]

1

MD TD

–1

E'

9

1

MD TD

8

0 [ E"

7

–1

tan d 6

–100

2 10 Hz

E'

9

log [tan δ ]

log [E ', Pa], log [E ", Pa]

10 Hz

0

tan d

–2 200

100

log [tan δ ]

(a)

6

Temperature, °C

–100

0

100

200

–2

Temperature, °C

(c) iPP/EHR53 (90/10) 10

2

E'

9

1

MD TD

8

0 E"

7

log [tan δ ]

log [E ', Pa], log [E ", Pa]

10 Hz

–1 tan d

6

–100

0

100

–2 200

Temperature, °C

Figure 9.43 Temperature dependence of dynamic tensile moduli for the MD (open symbols) and the TD (closed symboles): (a) iPP, (b) iPP/EHR30, and (c) iPP/EHR53. (From Reference 65 with permission from John Wiley & Sons, Inc.)

Chapter 9 Morphology and Mechanical Properties in Blends

261

Figure 9.44 Mechanical model composed of amorphous (A) and crystalline (C) regions. (From Reference 65 with permission from John Wiley & Sons, Inc.)

dynamic mechanical spectra between MD sample cut out from the parallel to the flow direction and TD sample cut out perpendicular to the flow direction. Moreover, the magnitude of E0 for the MD samples of iPP and immiscible blend is higher than that for the TD sample in the lower temperature region, but above the b-relaxation temperature the TD sample shows higher E0 . The crossing behavior in E0 is not observed in the miscible iPP/EHR53 blend. The cold drawn and annealed iPPs (89), and an injection-molded iPP (90) showed the crossing behavior in E0 . Takayanagi et al. (88) explained the crossing behavior in terms of a simple mechanical model composed of amorphous (A) and crystalline (C) regions, as shown in Fig. 9.44, in which (C) region has anisotropy of modulus, that is, a high modulus in mechanical direction MD, about 1011Pa ðEC k Þ, and a low modulus in transverse direction TD, about 109Pa ðEC ? Þ, according to molecular orientation in the MD direction. The parameter 1 fC expresses the amorphous fraction of parallel connection and the parameter 1 lC denotes the fraction of series connection; therefore, the ð1 lC Þð1 fC Þ becomes a volume fraction of amorphous region. As postulated by Hosemann’s structural model (90) and supported by X-ray scattering measurements (91), molecular chains in the (C) regions are bonded together enough to transfer the force along TD as well as MD. Typical transverse modulus is considerably higher than the modulus EA ð¼ about 107 PaÞ in (A) above the b-dispersion, because the modulus EC in (C) region keeps high until the melting point. When fC in Fig. 9.44 is quite larger than lC, the TD modulus of fC EC ? is higher than the MD modulus lC EC k above the b-dispersion. In other words, the crossing behavior in E0 occurs when the connection of (C) and (A) along the flow direction (series connection) is weaker than the connection in the perpendicular to the flow direction (parallel connection) (88). Therefore, it is necessary for the crossing behavior in E0 that neighbor crystallites aligned perpendicular to the flow direction are mechanically linked together, which was explained in detail by Takayanagi et al (89). Fujiyama et al. (90) suggested that the skin layer of an injection-molded iPP displays the shishkebab structure proposed by Keller and Machin (93), in which c-axis and a*-axis orient to the flow direction. Kalay and Bevis have also demonstrated the shish-kebab morphology by WAXD and TEM observation (94). Fig. 9.45a shows a schematic model of the shish-kebab structure in the skin layer of an injection-molded iPP. Most

262

Polyolefin Blends

Flow direction

Kebab Shish (a)

(b)

Figure 9.45 Schematic model for the skin layer: (a) iPP and (b) iPP/EHR53. The bold lines in (b) represent EHR chains. (From Reference 65 with permission from John Wiley & Sons, Inc.)

of crystallites (kebabs) fill space, and the fibrous crystals (shishes) are parallel to flow direction. Furthermore, some kebabs are bonded with neighbor kebabs aligned perpendicular to the flow direction. As a result, the transverse connection of crystalline region perpendicular to the flow direction is attained in a skin layer and the shishes act as a linkage along the flow direction. As seen in Fig. 9.43, however, the level of E0 in the MD sample for the iPP/EHR53 is higher than that in the TD sample, even above b-relaxation temperature. The fact is owing to the lack of the transverse connection of crystalline region. As discussed before, the EHR molecules are rejected from iPP crystals, although both polymers are mixed in the molten state. Consequently, some amounts of EHR chains will exist between neighbor crystalline fragments aligned perpendicular to the flow direction. This will be responsible for the poor mechanical anisotropy, regardless of the high degree of orientation in c-axis of iPP as shown in Table 9.5 and Fig. 9.42. The schematic model of the skin layer in iPP/EHR53 is shown in Fig. 9.45b. However, the iPP/EHR30 blend shows macroscopic phase separation in which EHR domains are dispersed in the iPP matrix in the molten state. Therefore, the blending of the EHR has little influence on the connection of crystalline kebabs. The stress–strain curves of the iPP and the iPP/EHR blends are shown in Fig. 9.46. The overall stress level in iPP is considerably greater than those of the blends. The iPP failures just beyond the yield point while the blend samples show a higher extensibility. In the case of TD samples, the yield stress of the iPP/EHR53 is higher than that of the iPP/EHR30, which is different from the results of the samples prepared by the compression-molding as shown in Fig. 9.23. Furthermore, the iPP/EHR53 shows less anisotropy of the yield stress. Table 9.6 summarizes the tensile properties, such as yield stress s Y, strain at yield point, and anisotropy of yield stress s Y(MD)/s Y(TD). The strain at the yield point for the iPP/EHR53 is larger than those for the iPP and the iPP/EHR30. Furthermore, it is also found during the tensile testing that many cracks appear on the surface of the TD samples before the yield

263

Chapter 9 Morphology and Mechanical Properties in Blends (a)

(b) 50

50

TD

MD 40 Stress, MPa

Stress, MPa

40

30

20

iPP iPP/EHR30 iPP/EHR53

10

30

20

iPP iPP/EHR30 iPP/EHR53

10

0

0 0

2

4

6

0

8

2

4

6

8

Strain, %

Strain, %

Figure 9.46 Stress–strain curves for iPP (solid line), iPP/EHR30 (dotted line), and iPP/EHR53 (dashed line): (a) the MD samples and (b) the TD samples. (From Reference 65 with permission from John Wiley & Sons, Inc.)

point of the iPP and the iPP/EHR30. Figure 9.47 shows the optical micrographs for the surface of the TD samples after removal of the stress at their yield points. The arrow in the figure indicates the stretching direction, that is, perpendicular to the flow direction. The direction of cracks appears to be parallel to the flow direction. Formation of many cracks on the surface leads to the brittle behavior of the skin layer. On the contrary, there are no cracks on the iPP/EHR53, although the yield stress and yield strain are larger than those of the iPP/EHR30. In the MD samples, however, there are no cracks on all samples before break in the iPP or the neck formation in the blends. As increasing the strains for the TD samples of the iPP and the iPP/EHR30, the lamellar fragmentation takes place even at small strains and the crystalline fragments segregated during deformation coalesce into the origin of the cracks. Therefore, the skin layer in TD samples for the iPP/EHR30 shows brittle behavior, leading to the low level of yield stress in the TD sample as well as the high

Table 9.6 Characteristics of Stress–Strain Behavior. Sample iPP iPP iPP/EHR30 SPP/EHR30 iPP/EHR53 iPP/EHR53

Yield stress, MPa MD TD MD TD MD TD

41.4 37.2 32.3 28.9 31.9 30.6

Strain at Yeild point, % 1.3 1.4 1.8 1.5 2.7 2.4

[From Reference 65 with permission from John Wiley & Sons, Inc.]

s Y(MD)/s Y(TD) 1.11 1.12 1.04

264

Polyolefin Blends

Figure 9.47 Optical micrographs for the surface of the TD samples: (a) iPP, (b) iPP/EHR30, and (c) iPP/EHR53. The samples were elongated by the yield point and then gold coated before the observation. The arrow indicates the stretching direction. (From Reference 65 with permission from John Wiley & Sons, Inc.)

mechanical anisotropy. On the contrary, the iPP/EHR53 show high extensibility in both MD and TD samples, which will be responsible for the less mechanical anisotropy.

9.5 CONCLUSIONS The miscibility of olefin copolymers such as ethylene–a-olefin copolymers was found to be controlled by the structural composition and the primary structure of the copolymers. Using these copolymers, binary blends with various compatibilities were prepared and the effects of compatibility on mechanical properties in the binary blends were investigated. The tensile properties in binary blends of iPP with rubbery olefin copolymers are considerably influenced by the miscibility between iPP and the copolymers. The miscibility of iPP with other polyolefins is described in detail based on the dynamic mechanical properties, morphology observation, and solidification process. It is found that EBR, EHR, and EOR having more than 50 mol% of a-olefin are miscible with iPP in the molten state. In the solid state, the miscible copolymers are dissolved in the amorphous region of iPP, although the copolymers are excluded from crystalline lattice of iPP. The isotactic propylene sequence in the EP copolymers with a propylene-unit content of more than 84 mol% participates in the crystallization process of iPP, resulting that a part of the EP copolymers is included in the crystalline lattice of iPP. The immiscible blends having phase-separated morphology show more brittle behavior accompanied by interfacial debonding between iPP matrix and rubber domains. In the case of miscible blends where rubbery copolymers are completely incorporated into the amorphous region of iPP, the spherulites are affinely deformed without plastic deformation such as crazes and cracks and the miscible blends exhibit significant high drawability. The miscible blends where the crystallizable parts are cocrystallized with iPP show a high resilience and become more toughened.

Chapter 9 Morphology and Mechanical Properties in Blends

265

The crystallization temperature Tc affects the dynamic mechanical properties for miscible iPP/EHR blends. The b or Tg process appears as broad ambiguous double peaks for the blend crystallized at higher Tc, indicating the change of molecular aggregation state in the melt during the crystallization. The location of a-process, the relaxation associated with crystalline region, shifts to higher temperatures with the EHR fraction, which is owing to the well-organized crystalline structure of iPP. Furthermore, the injection-molded products of the iPP/EHR blends include shishkebab structure and blending miscible copolymers depresses the mechanical anisotropy owing to the lack of the mechanical connection between neighbor crystalline kebabs aligned perpendicularly to the flow direction, which is qualitatively evaluated by dynamic mechanical analysis.

NOMENCLATURE aPP E0 E00 EBR EHR EOR EPR G0 G00 HDPE iPP LDPE L-LDPE Me PB POM SAXS sPP tan d TEM Tg WAXD

Atactic polypropylene Tensile storage modulus Tensile loss modulus Ethylene–butene-1 random copolymer Ethylene–hexene-1 random copolymer Ethylene–octene-1 random copolymer Ethylene–propylene random copolymer Shear storage modulus Shear loss modulus High density polyethylene Isotactic polypropylene Low density polyethylene Linear low density polyethylene Entanglement molecular weight Poly(butene-1) Polarized optical microscope Small-angle X-ray scattering Syndiotactic polypropylene Loss tangent Transmission electron microscope Glass transition temperature Wide-angle X-ray diffraction

REFERENCES 1. J. Karger-Kocsis, A. Kallo, A. Szafner, G. Bodor, and Z. Senyei, Polymer, 20, 37 (1979). 2. L. Dorazio, R. Greco, E. Martuscell, and G.. Ragosta, Polym. Eng. Sci., 23, 489 (1983). 3. P. Galli, S. Danesi, and T. Simonazzi, Polym. Eng. Sci., 24, 544 (1984). 4. C. S. Ha and S. C. Kim, J. Appl. Polym. Sci., 37, 317 (1989).

266

Polyolefin Blends

5. V. Choudhary, H. S. Varma, and I. K. Varma, Polymer, 32, 2534 (1991). 6. P. J. Carriere and H. C. Silvis, J. Appl. Polym. Sci., 66, 1175 (1997). 7. A. van der Wal, J. J. Mulder, J. Oderkerk, and R. J. Gaymans, Polymer, 26, 6781 (1998). 8. M. H. Kim, R. G. Alamo, and J. S. Lin, Polym. Eng. Sci., 39, 2117 (1999). 9. J. W. Teh, A. Rudin, and J. C. Keung, Adv. Polym. Technol., 13, 1 (1994). 10. A. Piloz, J. Y. Decroix, and J. F. May, Die Ange. Makromol. Chem., 54, 77 (1976). 11. G. Boiteux, J. C. Dalloz, A. Douillard, J. Guillet, and G. Seytre, Eur. Polym. J., 16, 489 (1980). 12. M. M. Dumoulin, P. J. Carreau, and L. A. Utracki, Polym. Eng. Sci., 27, 1627 (1987). 13. R. M. Gohil and J. Petermann, Macromol. Sci., Phys., B18, 217 (1980). 14. A. Siegmann, J. Appl. Polym. Sci., 27, 1053 (1982). 15. C. C. Hsu and P. H. Geil, Polym. Eng. Sci., 27, 1542 (1987). 16. P. M. Cham, T. H. Lee, and H. Murand, Macromolecules, 27, 4263 (1994). 17. M. Yamaguchi, J. Appl. Polym. Sci., 70, 457 (1998). 18. M. Yamaguchi, Surface Sci. Jpn., 21, 226 (2000). 19. M. Yamaguchi, H. Miyata, and K. Nitta, J. Appl. Polym. Sci., 62, 87 (1996). 20. P. A. Weimann, T. D. Jones, M. A. Hillmyer, F. S. Bates, J. D. Londono, Y. Melnichenko, G. D. Wignall, and K. Almdal, Macromolecules, 30, 3650 (1997). 21. D. Ma¨der, Y. Thomann, J. Suhm, and R. Mu¨lhaupt, J. Appl. Polym. Sci., 74, 838 (1999). 22. D. J. Lohse and W. W. Graessley, Thermodynamics of polyolefin blends, in: Polymer Blends, D. R. Paul and C. B. Bucknall (eds.), Wiley, New York, 1999, Chapter 8. 23. M. Yamaguchi, K. Nitta, H. Miyata, and T. Masuda, J. Appl. Polym. Sci., 63, 467 (1997). 24. M. Yamaguchi, H. Miyata, and K. Nitta, J. Polym. Sci., Polym. Phys., 35, 953 (1997). 25. D. J. Lohse, Polym. Eng. Sci., 26, 1500 (1986). 26. M. Yamaguchi and K. Nitta, Rep. Polym. Phys. Progr. Jpn. 39, 457 (1996). 27. F. S. Bates, M. F. Schulz, J. H. Rosedale, and K. Almdal, Macromolecules, 25, 5547 (1992). 28. K. F. Freed and J. Dudowicz, Macromolecules, 29, 625 (1996). 29. U. W. Suter and P. J. Flory, Macromolecules, 8, 765 (1975). 30. A. Eckstein, C. Friedrich, J. Suhm, C. Friedrich, R. D. Ma¨ier, J. Sassmannshausen, M. Bochmann, and R. Mulhau¨pt, Macromolecules, 31, 1335 (1998). 31. M. Yamaguchi and H. Miyata, Macromolecules, 32, 5911 (1999). 32. R. Thomann, J. Kressler, S. Setz, C. Wang, and R. Mu¨lhaupt, Polymer, 37, 2627 (1996). 33. R. D. Ma¨ier, R. Thomann, J. Kressler, R. Mu¨lhaupt, and B. Rudolf, J. Polym. Sci., Polym. Phys., 35, 1135 (1997). 34. R. Silvestri and P. Sgarzi, Polymer, 39, 5871 (1998). 35. W. L. Mattice, C. A. Helfer, S. S. Rane, E. D. von Meerwall, and B. L. Farmer, J. Polym. Sci., Polym. Phys., 43, 1271 (2005). 36. Y. W. Shin, T. Uozumi, M. Terano, and K. Nitta, Polymer, 42, 9611 (2001). 37. K. Nitta, Y. W. Shin, H. Hashiguchi, S. Tanimoto, and M. Terano, Polymer, 46, 965, (2005). 38. C. Bauwens, J. C. Bauwens, and G. J. Home´s, J. Polym. Sci. A2 22, 870 (1969). 39. K. Nitta and A. Tanaka, Polymer, 42, 1219 (2001). 40. J. Karger-Kocsis (ed.), Polypropylene: Structure, Blends and Composites, Chapman & Hall, London, 1995. 41. L. A. Utracki, Polymer Alloy and Blends, Carl Hanser, Munich/FRG, 1989. 42. P. Gulli, S. Daresi, and T. Simouzzi, Polym. Eng. Sci., 24, 544 (1984). 43. R. Greco, C. Mancarella, E. Martuscelli, F. Ragosta, and Y. Jinghua, Polymer, 28, 1929 (1987). 44. E. Martuscelli, Polym. Eng. Sci., 24, 563 (1984).

Chapter 9 Morphology and Mechanical Properties in Blends

267

45. W. Rlaris and Z. H. Stachurski, J. Appl. Polym. Sci., 62, 87 (1996). 46. R. S. Stein, in: New Methods of Polymer Characterization, B. Ke (ed.), Wiley, New York, 1968 p. 255. 47. G. L. Wilkes, Adv. Polym. Sci., 8, 91 (1971). 48. H. W. Siesler, Adv. Polym. Sci., 65, 1 (1984). 49. Y. Song, K. Nitta, and N. Nemoto, Macromolecules, 36, 1955 (2003). 50. K. Nitta, K. Okamoto, and M. Yamaguchi, Polymer, 39, 53 (1998). 51. S. Onogi, Y. Fukui, and T. Asada, Proceedings of 5th International Congress on Rheology, Vol. 4, S. Onogi (ed.), University of Tokyo Press, Tokyo, 1970, p. 87. 52. A. Tanaka, M. Fukuda, H. Nagai, and S. Onogi, J. Polym. Sci., Polym. Phys., 27, 2283 (1989). 53. M. Yamaguchi, H. Miyata, and K. Nitta, PPS-15 The Polymer Processing Society, ’sHertogenbosch, The Netherland, June 1999. 54. E. L. Bedia, S. Murakami, K. Senoo, and S. Kohjiya, Polymer, 43, 749 (2002). 55. M. Yamaguchi and K. Nitta, Polym. Eng. Sci., 39, 833 (1999). 56. Y. L. Huang and N. Brown, J. Polym. Sci., Polym. Phys., 29, 129 (1991). 57. K. Nitta and M. Takayanagi, J. Polym. Sci., Polym. Phys., 37, 357 (1999). 58. K. Nitta and M. Takayanagi, J. Polym. Sci., Polym. Phys., 38, 1037 (2000). 59. K. Nitta and M. Takayanagi, J. Macromol. Sci. B-Phys., B42, 107 (2003). 60. R. Popli and L. Mandelkern, J. Polym. Sci., Polym. Phys., 25, 441 (1987). 61. O. Darras and R. Se`gue`la, J. Polym. Sci., Polym. Phys., 31, 759 (1993). 62. K. Nitta, H. Ando, and T. Asami, e-Polymers 021 (2004). 63. K. Nitta, T. Kawada, M. Yamahiro, H. Mori, and M. Terano, Polymer, 41, 6765 (2000). 64. M. Yamaguchi, H. Miyata, and K. Nitta, Nihon Rheoroji Gakkaishi, 26, 163 (1998). 65. M. Yamaguchi, K. Suzuki, and H. Miyata, J. Polym. Sci., Polym. Phys., 37, 701 (1999). 66. H. Nakayasu, H. Markovitz, and D. J. Plazek, Trans. Soc. Rheol., 5, 261 (1961). 67. Y. Wada and K. Tsuge, J. Appl. Phys., Jpn., 1, 64 (1962). 68. N. G. McCrum, B. E. Read, and G. Williams, Anelastic and Dielectric Effects in Polymeric Solids, Wiley, London, 1967. 69. A. Tanaka, E. P. Chang, B. Delf, I. Kimura, and R. S. Stein, J. Polym. Sci., Polym. Phys., 9, 391 (1974). 70. T. Kajiyama, T. Okada, and M. Takayanagi, J. Macromol. Sci. Phys., B9, 391 (1974). 71. L. E. Nielsen, Mechanical Properties of Polymers and Composite, Marcel Dekker, New York, 1975. 72. H. D. Keith and F. J. Padden, J. Appl. Phys., 35, 1270 (1975). 73. J. D. Ferry, Viscoelastic Properties of Polymers, 3rd ed., Wiley, New York, 1980. 74. H. Kawai, T. Hashimoto, S. Suehiro, and K. Fujita, Polym. Eng. Sci., 24, 361 (1984). 75. R. H. Boyd, Polymer, 26, 1123 (1985). 76. H. D. Keith and F. J. Padden, J. Polym. Sci., Polym. Phys., 25, 229 (1987). 77. H. D. Keith and F. J. Padden, J. Polym. Sci., Polym. Phys., 25, 2371 (1987). 78. R. Hayakawa and Y. Wada, Rep. Progr. Polym. Phys. Jpn., 11, 215 (1968). 79. K. Okano, J. Polym. Sci., C15, 95 (1966). 80. N. G. McCrum, Polymer, 25, 299 (1984). 81. H. G. Olf and A. Peterlin, J. Polym. Sci., A2, 753 (1970). 82. K. Inohara, K. Imada, and M. Takayanagi, Polym. J., 4, 232 (1973). 83. T. Kajiyama, T. Okada, A. Sakoda, M. Takayanagi, J. Macromol. Sci. Phys., B7, 583 (1973). 84. S. Stein, F. B. Khambatta, F. P. Warner, T. R. Russell, A. Escala, and E. Balizer, J. Polym. Sci., Polym. Symposium, 63, 313 (1978).

268

Polyolefin Blends

85. M. Doi and S. F. Edwards, J. Chem. Soc., Faraday Trans. II 74, 1789, 1802, 1818 (1978); 75, 38 (1979). 86. S. Wu, J. Polym. Sci., Polym. Phys., 25, 2511 (1987). 87. D. L. Plazek and D. J. Plazek, Macromolecules, 16, 1469 (1983). 88. A. Turner Jones, J. M. Aizlewood, and D. R. Beckett, Makromol. Chem., 75, 134 (1964). 89. M. Takayanagi, K. Imada, and T. Kajiyama, J. Polym. Sci., C15, 263 (1966). 90. M. Fujiyama, T. Wakino, and Y. Kawasaki, J. Appl. Polym. Sci., 35, 29 (1988). 91. R. Hosemann, J. Appl. Phys., 34, 25 (1963). 92. M. Takayanagi, Pure Appl. Chem., 15, 555 (1967). 93. A. Keller and M. J. Machin, J. Macromol. Sci., 16, 176 (1976). 94. G. Kalay and M. J. Bevis, J. Polym. Sci., Polym. Phys., 35, 265 (1997).

Chapter

10

Functionalization of Olefinic Polymer and Copolymer Blends in the Melt Boleslaw Jurkowski1, Stepan S. Pesetskii2 and Yuri M. Krivoguz2

10.1 INTRODUCTION During the past 20 years, there has been much interest in understanding the grafting of polar vinyl monomers to polyolefins (PO). The grafting process can be performed in an inert solvent or in a PO melt. A direct grafting of a monomer to chains in molten PO that follows the free-radical mechanism appears more preferable, and it has been studied more widely. It is most often done by means of reactive extrusion (RE), where an extruder is used as a reactor of continuous action. This technology permits the production of a variety of functionalized PO (1–8). The major advantages of PO functionalization by RE method are the possibility of high and controlled yield of the final product, simplicity of its separation, high rate of the reaction, and the continuity of the process (1,5). The disadvantages and difficulties of performing the RE are caused by the necessity to thoroughly account for the effect of polyolefin structure, nature of the initiator used for radical transformations, the monomer grafted, technological parameters and design of the extrusion mixer reactor in the course of the major process (grafting of a monomer to chains) as well as side reactions (degradation

1 Division of Plastic and Rubber Processing, Institute of Material Technology, Poznan University of Technology, Piotrowo 3, 60-950 Poznan, Poland. 2

Laboratory of Chemical Technology of Polymeric Composite Materials, V.A. Belyi Metal-Polymer Research Institute of National Academy of Sciences of Belarus, 32a Kiroc Street, 246050 Gomel, Belarus.

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

269

270

Polyolefin Blends

and/or cross-linking of chains, monomer oligomerization, formation of grafted copolymers, etc.) (1–3,5,6). The experimental data on monomer grafting to homopolyolefins (polypropylene (PP), polyethylene (PE), copolymers of ethylene and propylene (EPR)) make it possible to predict the course of the main process and associated reactions depending on the factors mentioned. Considering that the grafting rate can be described by a kinetic equation of first order with respect to the monomer concentration (3), it seems possible to qualify the level of functionalization. It was found for homopolyolefins that competitive secondary reactions of crosslinking or degradation of the chains occur simultaneously with the grafting reaction (1–3,5,6,9). Their course depends on the PO nature: in the course of free-radical transformations, PE mostly undergoes cross-linking (1–3,5,6,9) while PP suffers degradation (10–12). In EPR, a decrease in propylene fraction leads to a change from degradation of chains to their cross-linking (1,3,10,11). The yield ratio of the major and secondary products strongly depends on both the reacting system’s composition and the reactive extrusion parameters—temperature, design of the extruder, shearing actions on the moving polymer melt, residence time for the reacting system in the reaction zone, and several others (5,6,9,13,14). A rather different situation can be observed in monomer grafting to PO blends or blends of POs with olefin copolymers. The data available suggest that freeradical grafting of polar monomers to PO blends will be followed by a complicated series of macromolecular transformations as compared with the processes taking place in grafting to homopolyolefins. At the same time, despite the clear actuality, the works on grafting of monomers to PO blends during reactive extrusion are not numerous. Of particular interest are the problems related to the behavior of freeradical reactions in heterogeneous melts of PO blends, selective nature of grafting and secondary reactions, the effect of thermodynamic affinity—the monomer, as well as the initiator, with the polymer components—the course of those reactions. Their analysis with due regard to the available literature data, and also new original results, is undoubtedly of interest for the development of general ideas about the mechanisms of macromolecular transformations and is, particularly, the main aim of this work.

10.2 SCOPE OF REVIEW Before analyzing the state of art in the functionalization of blends of both polymers and olefin copolymers, we first consider the most important points of grafting of polar monomers to homopolyolefins in melt by means of RE.

10.2.1 Free-Radical Grafting of Unsaturated Monomers to PO Chains A general scheme of probable reactions taking place in a polymer melt when monomers are the free radical grafted onto its chains (6) is shown in Fig. 10.1.

271

Figure 10.1 Scheme of free-radical grafting of monomers to polyolefin chains. (Adapted from Reference (6).

272

Polyolefin Blends

The stage of initiation is realized owing to the thermolysis of a peroxide initiator (ROOR)—reaction rate constant kt—and the formation of primary alkoxyl radicals RO*, which—because of fragmentation reactions—can give secondary free radicals.

`

R

CH

(

The free radicals thus formed, in the presence of a polymer (

CH 2

and

a monomer (M), can interact with each of these components. In this case, either a

`

R

macroradical (

CH 2

) (kH) or a monomer radical ROM* (ki) is formed.

C

*

The chain process can develop in different directions. In the case of reaction between a macroradical and a monomer molecule, the monomer is grafted (kgi) onto a chain in the form of a single monomer unit and forms a grafted macroradical. The latter can either interact with other monomer molecules (kgp) and form a longer grafted chain or participates in reactions of the chain transfer (kHi). During the chain R` CH

transfer, final grafted products

(

2

C MH

) are formed along with new

`

R

macroradicals

(

CH 2

C

*

) , which can initiate a new grafting cycle. The

chain propagation can be initiated not only by macroradicals but also by free radicals formed during peroxide’s decomposition. In this case, the monomer undergoes homopolymerization (kp) (RO(M)nM*) in polyolefin. So far as grafting onto macroradicals of oligomers and homopolymers of a monomer being grafted is less likely than grafting of a monomer because of thermodynamic factors, the reaction usually results in a mixture of the polyolefin with a grafted monomer and polymerization products of the monomer. The reactions of chain breakage and formation of stable products are quite diverse. Among them, the following are worth mentioning: recombination of similar macroradicals (kc), disproportionation (kd), and a combination of dissimilar chains. The latter reactions are typical of realization of grafting in polymer blends. Macroradicals can also suffer degradation at the expense of b-scission reaction (ks). All of the mentioned processes, except grafting of a monomer onto chains, are side ones and depending on the polymer origin, they cause changes in the molecular structure and properties of the final product (6,7). The analysis of the kinetic scheme of free-radical grafting showed that the grafting efficiency, as well as the final structure of grafted products, largely depends on the ratio of rate constants of a series of elementary reactions. The higher the ratio kH/ki, the higher the yield of macroradicals and the higher the grafting degree of the monomer reached. The more a monomer is grafted, the lower is the probability of its homopolymerization in polyolefin matrix. An important role should belong to the reactivity of free radicals that resulted from peroxide thermolysis and of importance is the mobility of hydrogen atoms in chains.

Chapter 10 Functionalization of Olefinic Polymer and Copolymer Blends in the Melt

273

When selection is made of a peroxide initiator, the capacity of its free radicals to detach hydrogen from the polymeric substrate must be taken into consideration. It is important for the monomer being grafted that the kgi/ki ratio be as high as possible. In other words, the monomer should have a higher power of interaction with macro-

`

R

radicals (

CH 2

C

*

) than with R–O* free radicals.

The ratio kHi/kgp is also an important parameter of grafting since it characterizes the relationship between the rate of intra- and interchain transfer of hydrogen from chains to the grafted monomer and the rate of monomer attachment to the grafted macroradical. The higher the kHi/kgp ratio, the shorter is the length of grafted fragments, and vice versa. Besides, owing to the reactions of intra- and intermolecular transfer of hydrogen between chains and grafted macroradicals, new macroradicals are formed and a new cycle of grafting is initiated. The kHi/kgp ratio depends much on the reactivity of the grafted macroradical and the mobility of hydrogen atoms present in the polymer chain (6,7). The final structure of a grafted product will depend on the ks/kgi, kc/kgi, and kd/kgi ratios. For PE macroradicals, high values of kc are typical and grafting is followed by cross-linking of the chains. For polypropylene, chain degradation is typical because of high ks values. In the case of ethylene–propylene rubber, there is observed a competition between the two side reactions. In spite of a fundamental difference in the mechanisms of the processes that occur at the grafting of monomers to chains of PE, PP, and copolymers of ethylene with different monomers (first of all, propylene), a number of important laws have been established that are of common nature and should be taken into account when developing particular formulations and technologies. It should be considered that at RE conditions when the process is fast in order to provide for a high yield of the grafted product, grafting initiators should be peroxides that show thermodynamic affinity with PO (12,15–18). A higher solubility of a peroxide initiator in a polymer (not in a monomer) leads to the fact that radicals— formed during thermal decomposition of the initiator—first interact with chains. Macroradicals formed initiate the reaction of grafting of the monomer to the chains. A stronger affinity of peroxide with a monomer leads to a lower efficiency of grafting and promotes monomer’s oligomerization (12,18). For PE, the relationship between grafting efficiency and cross-linking of chains can be controlled by varying the concentration ratio of the initiator to the monomer (3). For example, no cross-linking was detected when diethylmaleate (DEM) was grafted to PE as initiated by dicumyl peroxide (DCP) if the ratios (DCP):[DEM] < 0.09. A positive effect on grafting is caused by lower molecular weight of PO (lower melt viscosity) (3). The bond strength between hydrogen and carbon in PO chains weakens in the series of groups: methyl (–CH3), methylene (–CH2 –), methine (–CH–) group (6). That is, hydrogen gets detached more easily from a tertiary atom of carbon. In view of this, it can be assumed that a higher rate of formation of macroradicals should be

274

Polyolefin Blends

typical of PP and not of PE. It seems, therefore, that grafting should proceed with a greater yield when PP has been used. In practice, however, an opposite result is usually observed: the grafting efficiency of a monomer to PE is higher than to PP (10–12,16,17,19). It had been reported earlier (11) that 0.23–1.25 wt% of peroxide initiator provides for 0.15–1.08 wt% of maleic anhydride (MAH) grafted onto PP. The initiation efficiency is below 1 (unity), whereas for PE this parameter is much higher than unity (when MAH is grafted to PE, 0.5 mol of a peroxide is needed for 10 mol of a monomer) (11). This can be explained as follows. Reactivity of free radicals is inversely proportional to their rate of formation (20). In the case when tertiary macroradicals are formed, their efficiency of interaction with monomer molecules also decreases owing to steric hindrance created by the methyl groups. Besides, a certain role is played by the kinetic factor (11); b-scission of tertiary radicals results in an equivalent number of macroradicals with a reacting center of a secondary carbon atom (Fig. 10.1). However, no grafting have been detected along those centers (11). This fact can be explained by the rapid termination of secondary macroradicals of PP. Higher rates of b-scission of tertiary macroradicals and termination of secondary ones—in comparison with the attachment rate of the monomer (MAH)—can be one of the explanations for lower grafting efficiency of the monomer to PP as compared with PE. Because of different grafting efficiency to PP and PE, variations in the unit ratio of ethylene and propylene within the EPR structure lead to variations in the yield of the grafted product (3,21). It had been shown elsewhere (21) that the unit concentration of propylene in EPR, when increased from 28 up to 45 wt%, led to decreased grafting of dibutylmaleate initiated by DCP, while the degradation rate of the copolymer increased. It has been found that grafting efficiency rose by extending the –CH2 – sequences. The grafting efficiency of monomer to –CH2 – fragments grows as they become more distant from the tertiary macroradical (21). The reaction scheme described above (Fig. 10.1) has been suggested for freeradical grafting of unsaturated monomers to homopolyolefins. There is no proof to consider that the processes will differ much for multicomponent systems composed of several polyolefins. In the case of PO blends, a number of specific features will occur. It should be mentioned that a polymer blend contains chains that differ in conformations, stereoregularities, reactivity of hydrogen attached to carbon atoms having substitution of different degrees, and, consequently, reactivities in freeradical processes. Therefore, the concentration ratio of polymer components in the reacting blend must significantly affect the course of grafting reaction and determine its efficiency. In addition to this, blends of olefin polymers are thermodynamically incompatible systems. The incompatibility and the tendency of PO blends to create separate phases persist in their melts. The presence of two phases in a reacting system can affect variations in the favorable concentrations of low molecular weight reagents within the polymers owing to the specific absorption or their preferential dissolution in one of the polymers. Thus, in a grafting reaction of polar monomers within a heterophase system, the difference in solubilities of low molecular weight reactants

Chapter 10 Functionalization of Olefinic Polymer and Copolymer Blends in the Melt

275

in polymer components can be one of the parameters that influence its course. The above information indicates that free-radical grafting in polyolefin blends must have distinct selective nature. As for secondary processes that occur during grafting, their behavior in blends should also differ from that of homopolyolefins. In addition to side reactions typical of homopolymers, cross reactions between dissimilar chains are possible in blends.

10.2.2 The Use of Monomers and Initiators Of all the elementary reactions concurrent with functionalization (Fig. 10.1), the most important is the grafting of a monomer to macroradicals. A competitive reaction is the interaction of the monomer with primary free radicals that resulted from peroxide thermolysis (homopolymerization of the monomer). It is hardly possible that oligomers and polymers of the monomer can be grafted to macroradicals because of thermodynamic factors (6). This results in a product with low efficiency of grafting that includes impurities of oligomers as well as homopolymers of the monomer grafted. Therefore to obtain a functionalized polyolefin, it is advisable to use monomers that show a weak tendency to homopolymerization. Typical monomers that are most widely used for polyolefins functionalization are given in Table 10.1. As a rule, in their structure, monomers are monosubstituted (RCH CH2), 1,1disubstituted (R1R2C CH2), or 1,2-disubstituted ethylenes (R1CH CHR2). Most of monosubstituted and 1,1-disubstituted ethylenes are capable of homopolymerization and form high molecular weight compounds. The exceptions are 1,1-disubstituted ethylenes with huge substituents, which—because of steric obstacles—can form only low molecular weight oligomers. An example is 1,1-diarylethylenes that solely form dimers (22). Unlike monosubstituted and 1,1-disubstituted ethylenes, 1,2-disubstituted ethylenes, as a whole, cannot undergo homopolymerization and form high molecular weight polymers (22). Maleic anhydride (MAH) and its mono- and diesters make an example of such compounds. MAH is one of the monomers most often used for polyolefins functionalization. It is characterized by an extremely low capacity to homopolymerization, and this fact is explained by the steric features of its structure. The reactivity of MAH to macroradicals, however, is comparatively low. From the chemistry viewpoint, a steric hindrance and a lack of electron density in the double bond explain the low reactivity of MAH, which in MAH is symmetrical owing to the presence of two carbonyl groups. Attempts have repeatedly been made to work out procedures for increasing the chemical activity of MAH. Three methods have been proposed to activate the double bonds in MAH: (i) to perform a grafting reaction for MAH in presence of an electron-donating monomer, for example, styrene, which is capable of forming a charge transfer complex (CTC) with MAH; (ii) substitution

276

Polyolefin Blends

Table 10.1

Monomers Most Widely Used for Free-Radical Grafting to PO at RE.

No.

Monomers

1. 1.1. 1.2. 2.

1-Substituted ethylenes Vinyltrimethoxysilane Vinyltriethoxysilane 1,1-Disubstituted ethylenes

2.1.

2-Isopropenyl-2-oxazoline

2.2.

Itaconic acid

2.3.

Glycidyl methacrylate

2.4.

2-(Dymethylamino)ethyl methacrylate

2.5.

3-Isopropenyl-a; adimethylbenzene isocyanate

3.

1,2- Disubstituted ethylenes

3.1.

Maleic anhydride

3.2.

Fumaric acid

3.3.

3.4.

Structure RCH CH2 Si(OCH3)3 CH2 CH CH Si(OC2H5)3 CH2 R1R2C CH2

R1CH CHR2

CH

CH

O

C

C

C2H5

O CH

O C2H5 CH

O

C

C

Diethyl maleate

Dimethyl maleate

O

O CH3

CH3 O O CH CH 3.5. 4.

C

C

O

Maleimide

O

Macromonomers

NH Styrene derivatives, esters of (meth) acrylic acid, derivatives of maleimide

Chapter 10 Functionalization of Olefinic Polymer and Copolymer Blends in the Melt

277

of one of the MAH-hydrogen atoms for a strong electron-donating or an electronaccepting group, for example, bromine; (iii) addition of diene with the aim to form an adduct the decay of which results in the activation of the MAH double bond (23,24). According to the first approach, an interaction of the electron-donating (styrene) and the electron-accepting (MAH) monomers results in a charge transfer complex. It is a combination of radical ions formed by electron transfer from styrene to MAH. MAH, included into CTC, shows an increased reactivity toward macroradicals. Among the electron-donating monomers capable of forming CTC with MAH, besides styrene, are a-methylstyrene (MeSt), N-vinylpyridone (NVP), trimetoxyvinylsilane (TMVS), methylmethacrylate (MMA), methylacrylate (MAc), vinylcyclohexane (VCH) (6). By their electron-donating power, these can be arranged as follows: St> MeSt>>NVP > TMVS > MMA > MAc > VCH. The other approach is based on the double-bond activation in MAH by substitution of one of the MAH hydrogen atoms for a strong electron-donating or electronaccepting group. In this case, the MAH double bond undergoes polarization. It is known that the stronger the double bond is polarized, the more actively the monomer reacts with macroradicals. According to Al-Malaika (6), the substitution of one of the MAH hydrogen atoms for a halogen atom promotes the yield rise of the grafted product even more than that by CTC formation. Diels–Alder system, which includes MAH and dicyclopentadiene, showed increased grafting efficiency only in a mixer of intermittent action. No special advantages have been detected in this system in comparison with neat MAH when extruder reactors were used (24). Besides MAH, acryl monomers, for example, glycidyl methacrylate (GMA), have been widely used for polyolefins functionalization. In its chemical structure, GMA belongs to 1,1-disubstituted ethylenes and has a stronger tendency to homopolymerization. Because of this, its grafting power to polyolefin chains is still lower than that of MAH. In order to raise the reacting power of GMA, electron-donating monomers (styrene or others) are added to the reacting system. Unlike MAH, GMA cannot form CTC with electron-donating monomers. There is another mechanism of raising the grafting efficiency of GMA to PO. It has been reported (6) that styrene first interacts with macroradicals. Then grafted styrene fragments copolymerize with GMA. Here, GMA is not directly grafted to chains but grafted through styrene units. The so-called ‘‘comonomer concept’’ is effective in grafting several other monomers, for example, 3-isopropenyl-a, a-dimethylbenzene isocyanate or 1-heptadecafluorooctylethyl acrylate (25,26). Vainio et al. (27) have studied the grafting of ricinoloxazoline maleate to PP and reported that the use of styrene did not increase the reaction efficiency; in some instances it decreased the latter while the degradation of polypropylene decreased markedly. The copolymerization constant value could be the basic point in explaining the effect of styrene on the grafting of vinyl monomers to polyolefins and concurrent secondary reactions (27). This means that ‘‘comonomer concept’’ acts quite successfully if at least two basic conditions are fulfilled: (i) a high rate of interaction of a comonomer with

278

Polyolefin Blends

macroradicals, and (ii) a high rate of interaction of a comonomer macroradical with the monomer being grafted (6,23,24). Macromonomers used in RE include poly- or oligomers with a reactive double bond. Because of mostly steric factors, macromonomers usually display a weaker tendency to homopolymerization in comparison with monomers. A disadvantage of macromonomers used in RE is that owing to lower volatility the ungrafted portion cannot be removed from the melt during the stage of degassing. The tendency of a monomer to homopolymerization depends on the propagation rate of chain (kp) and on the condition of dynamic equilibrium between the break and the formation of the chains (Fig. 10.1). The equilibrium depends on the temperature at which grafting takes place. The temperature at which a dynamic equilibrium is reached between the formation and the decay of monomer macroradicals is called a ceiling temperature. For certain monomers, there are published ceiling temperatures, heats, and entropy of polymerization (28,29). Their values are, for example, 150 C for MAH, 200 C for methacrylate, 400 C for acrylate and styrene (28). It should be noted that these values are typical of reactions occurring at a constant (atmospheric) pressure and monomer concentration (usually 1 mol). The peak temperature rises with monomer concentration and pressure. That is why MAH was observed to homopolymerize at an extrusion temperature above 160 C (30). When grafting is done by RE, many factors must be accounted for that concern a specific influence of the monomer on the process kinetics as well as the yield of the grafted product (1,3,5), that is, monomer concentration, its solubility in the molten PO, thermal stability, method of introduction into the reacting system, reactivity toward the initiator and macroradicals, and tendency to homopolymerization. The role of monomer is in trapping of radicals which otherwise are spent on degradation or cross-linking of the chains. If a monomer concentration is too high, phase separation may occur and lead to a lower yield of the grafted product and a stronger probability of homopolymerization (1). In such a case, a higher level of grafting can be achieved by stepwise feeding of the monomer and the initiator (31). The target-oriented selection of an initiator should be done depending on the monomer. The most important factors are the distribution of the initiator between monomer and PO phases, and also monomer’s reactivity in comparison with PO in reactions with free radicals generated during the thermal decomposition of the initiator (1). The stage of initiation of free-radical reactions taking place during monomer grafting to PO, as shown in Fig. 10.1, is the first and the most important stage. With other equal conditions, the type of initiator determines not only the yield of grafted product and the course of secondary reactions, but also the performance qualities of final materials (presence of unreacted initiator or its products of thermal decomposition within the functionalized PO can lead to accelerated ageing of goods in the course of their service). The problems of selection of initiators, discussion of the mechanism of their action, description of free-radical reactions at RE have been treated in numerous

Chapter 10 Functionalization of Olefinic Polymer and Copolymer Blends in the Melt

279

publications (1,5–7,15–20,32,33). We shall deal with most important points concerning grafting to PO blends. Most suitable initiators are organic peroxides (Table 10.2). In certain cases, it is advisable to use azo compounds as initiators (33). It is particularly typical of grafting amine-containing monomers: peroxides cause oxidation of amino groups, which leads to a loss of commercial value of a grafted product. It should be remembered that most of the azo compounds are unsuitable as initiators not only because of half-life (t0:5 ) but also because cyanoalkyl radicals formed are inactive in abstraction reactions of hydrogen from chains. An exception is phenylazo-compounds (33) being a source of phenyl radicals that are most active in hydrogen abstraction reactions (Table 10.3). The important factors concerning initiators that are to be accounted for in grafting experiments at RE include as follows: half-life; reactivity of products of the thermal decomposition of the initiator with respect to PO, the monomer, and other components of a reacting blend; solubility and distribution factor for the initiator among the components of a reacting blend in the molten PO; thermal stability and volatility of the initiator, its physical state, method of introduction into a reacting system, concentration, and tendency to form secondary products. It is advisable that the initiator should undergo thermal breakdown entirely within the reaction zone of the extruder. As the transport time of PO melt through the cylinder of the extruder does not exceed 3–5 min as well as t 0:5 —value for peroxides (Table 10.2), it is clear that thermal breakdown of a peroxide can be controlled by varying the melt temperature. At T 200 C, most peroxides have quite low values of t 0:5 . It is reasonable to assume that they undergo thermal decomposition virtually in full. If a residence time for a reacting blend in the extruder exceeds five times the half-life, the peroxide consumption exceeds 97% (1,6). On the contrary, the use of initiators with a very short t0.5 can result in negative consequences. An initiator with a short half-life leads to an increased concentration of radicals in a polymer melt. This can increase the probability of macromolecular cross-linking through recombination reactions of macroradicals. Besides, the yield of grafted product can be reduced because of a limited diffusion rate of the monomer or macroradical into the reaction zone. The latter fact is especially important for heterogenous melts. Therefore, if peroxides with short t 0:5 are used, the factors like concentration and introduction method of initiator become more important. In order to ensure a high yield of grafted product, it is advisable to introduce the initiator repeatedly. The advantage of reactive compounding, when short half-life peroxides are used, is the possibility of functionalizing PO in the first stage of compounding and mixing it with the basic thermoplastic polymer in the second stage of mixing. The use of extruder with large screw lengths (large value of L=D) and initiators with short t 0:5 values is not reasonable because of the weak influence of the melt residence time in the extruder upon the grafting efficiency and probable thermomechanical degradation of the grafted product. It should be noted that t 0:5 values in Table 10.2 have been determined at conditions different from those typical of RE process. In case of model experiments,

280

2,5-Dimethyl-2,5di(t-butylperoxy) hexane, DTBH, L-101

a:a’-Di (t-butylperoxy) diisopropyl benzene, DIPB, Perk-14

Dicumyl peroxide, DCP

PE

PP

EPDM (70–80 wt% ethylene)

Type of polymer/ peroxide, trade name

H

CH2 CH2 CH2 CH2

C

CH 3

H

CH 3

CH 2 C CH 2

H

CH2 C CH2 CH2 CH2 CH2

CH3

Structure

H

CH2

CH3 C

CH3 H

C

t-Butoxyl alkoxy

Methyl alkyl

Methyl alkyl

Methyl

Cumyloxy

t-Butoxyl alkoxy

—

—

—

—

—

—

Liquid

Solid

Solid

Solid

Solid

Solid

11.03

9.0

5.92

—

—

—

176

192

202

—

—

—

15.5

16.4

17.4

16.1

16.6

17.2

TempSolubility Concenerature of parameter tration of complete at 25 Cc Forming radicals Physical active decomposi- (calc.), Primary Secondary state oxygen,% tion, Cb (J cm3 Þ1=2

Table 10.2 Peroxide Initiators Used in RE Technology.

11.6

12.8

14.6

—

15.1

16.6

14.0

13.0

9.2

—

—

—

0.30

0.31

0.25

—

—

—

Solubility parameter at 180 Cc t 0:5 , mina (calc.), (J cm3 Þ1=2 150 C 200 C

281

CH 3

CN

OCH3

CH CH 3 3 N C CH C CH 2 3

CH 3

OO C

C H3

From derivatographic analysis (heating rate, 5 C min).17,18

N

CH3

C

CH 3

[From References (1,17,18) with permission from Elsevier].

c

Ph

H3C

According to Moad (1).

From studies.1,17,18

b

a

2-Phenylazo-2.4dimethyl-4-methoxylvaleronitrile, V-19

Di-t-butyl peroxide, DTBP

t-Butylcumyl peroxide, TBCP

2,5-Dimethyl-2,5di(t-butylperoxy) hexyne-3, DTBHY, L-130

Phenylazo alkyl

Phenyl Liquid

Methyl Liquid

—

10.95

7.3

Methyl Liquid

Cumyloxy t-butoxyl t-Butoxyl

10.2

Methyl Liquid

t-Butoxyl alkoxy

a

—

125

—

195

—

15.3

16.2

15.0

—

—

13.8

11.5

30.0

18.0

—

45.0

1.5

0.35

—

0.74

282

Polyolefin Blends

Table 10.3 180 C.

Estimated Equilibrium Constants for Hydrogen Abstraction Reaction at

No.

Radical

Primary

1. 2. 3. 4. 5. 6.

Ph RO CH3 RCH2 (R)2CH (R)3C

6 106 1 103 8 102 1 4 102 4 104

Carbon type Secondary 2 108 7 104 2 104 3 10 1 1 102

Tertiary 1 1010 6 106 2 106 2 103 9 101 1

[From Reference (33) with permission from Plenum Press].

it is practically impossible to account for the effects of pressure, shear, and other factors on the thermal decomposition of peroxides. Turcsanyi (34–36) has suggested that in selecting an initiator one of the criteria should be the temperature (T*) at which a maximum decomposition rate of peroxide is reached, and not t0:5 value. The reactivity and specific behavior of free radicals produced during initiator’s thermal decomposition strongly depend on the type of the radicals formed, which is determined by the nature of peroxide (28,37). Table 10.2 lists primary and secondary radicals formed during the decomposition of an initiator, while Table 10.3 gives data on the activity of certain types of free radicals in abstraction reactions of hydrogen atoms from carbon (33). Primary radicals are formed directly at breakdown of an initiator molecule; secondary radicals result from transformations of primary radicals by a monomolecular mechanism. It is shown in Table 10.3 that phenyl (Ph*) and alkoxyl (RO*) radicals are most active in abstraction reactions of hydrogen from carbon atoms. Methyl radicals (CH3*) are less active in abstraction reactions of hydrogen from carbon (especially primary carbon). These are especially effective in addition reactions by double bonds (1). For higher alkyl radicals, this tendency is stronger (Table 10.3). In should also be noted that of importance is the selective interaction of radicals, generated by an initiator, with PO base. It is owing to the interaction of monomer molecules with macroradicals and chain behavior of the process that grafting occurs (Figure 10.1). Up to 20 monomer units can be grafted onto one radical generated (38,39). It has been mentioned already that the initiator’s solubility in PO–monomer systems is one of the most important factors in RE technology. When grafting a monomer onto blends of olefin polymers or copolymers, the role of initiator’s solubility grows since it influences the distribution factor of the initiator between individual phases in multiphase melts (1). The distribution factor influences the grafted product yield and grafting selectivity (12,17,22). The initiator solubility in the components of a reacting system can be predicted using calculated values of the solubility parameter (d).

Chapter 10 Functionalization of Olefinic Polymer and Copolymer Blends in the Melt

283

According to other authors (40,41) vﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ u P DE i u u d ¼ t iP ; Na DVi

ð10:1Þ

i

Values of d calculated by the procedure of group shares (40,41) from Equation 10.1 are in good agreement (the error does not exceed 10%) with the experimental values found by measurements of evaporation heats of the substances (32). It is stated in other studies (32,40) that calculated d-values, an absolute error of determination for which is 0:1ðJ cm3 Þ1=2 , are as a rule more precise than experimental ones. Values of d depend on the temperature and can be found (32) from Equation 10.2 if the temperature varies: ln dT ¼ ln d298 bkðT 298Þ

ð10:2Þ

When determining the mutual solubility of substances by d-values one should remember that in the absence of strong specific intermolecular interactions, thermodynamic miscibility between the substances mixed (complete mutual dissolution among them) has been detected if their solubility parameters differed by no more than 2 (J cm3 Þ1=2 (42). The thermal stability of the initiator influences not only the grafting course, but also the method for introducing an initiator into molten PO and the safety requirements for RE procedure (1,5). Although for RE method there are recommended solid and thermally stable—up to PO melting temperature in the reaction zone—initiators, some authors have reported high yields of the grafted product with the use of thermally unstable peroxides that undergo decomposition near the feeding zone of the extruder (12,17). This fact supports the opinion that the major role in grafting involves the interaction of monomers with the radicals generated in the reaction of PO chains with products of thermal decay of peroxides. At RE conditions, macroradicals have, probably, quite a long lifetime since the reaction zone of their interaction with a monomer can be shifted for some distance from the introduction zone of a peroxide (10–18). There are different versions of introducing an initiator into a reacting system (1,5), for example, together with PO and a monomer into the major feeding hopper; into molten PO together with a monomer, before or after the monomer; an initiator can be preliminarily absorbed into PO; by introducing an initiator into PO as a solution in a monomer or in a solvent; by addition of the whole of the initiator at once or by portions. The introduction method of initiator, as well as the determination of its required concentration, is usually selected in view of the requirement to obtain a maximum yield of the grafted product and suppress secondary reactions. If a monomer is grafted to PO blends, an initiator can be introduced locally in metered quantities into an individual polymer phase.

284

Polyolefin Blends

10.3 FUNCTIONALIZATION OF PP/PE BLENDS The problem of functionalization of PP/PE blends is of interest for two reasons. First, such a functionalization can be treated as a means of obtaining blend compositions with better—in comparison with homopolyolefins—adhesional, compatibilizing, technological and other qualities as well as price. Second, addition of one polyolefin to another during free-radical grafting can be a means of controlling the course of concurrent secondary reactions. In PP/PE blends, the degradation of chains and their cross-linking can, in fact, be balanced at the expense of cross-reactions of PP and PE macroradicals that lead to grafted copolymers such as PP-g-PE (43).

10.3.1 Effect of Reacting Blend Formulation on Grafting Efficiency and Rheological and High Elastic Properties of Melt of Functionalized PP/PE Blends In an earlier study (44), dedicated to functionalization of PP/PE blends, there was considered the effects of MAH and DCP concentration on the viscosity of the product (PE/PP)-g-MAH and on the MAH-grafted level. It was found that on increasing DCP concentration from 0.05 to 0.3 wt%, the viscosity of (PE/PP)-g-MAH melt—polymer components ratio being 90:10—varies but negligibly (Fig. 10.2a). Obviously, with increased concentration of the initiator, both degradation and cross-linking of the chains are initiated alike, which oppositely influence the polymer melt viscosity. An increase in MAH content in the PE/PP (90:10) blend causes some drop in (PE/PP)-g-MAH viscosity (Fig. 10.2b). The main reason for this is the degradation process in PP, which proceeds faster than the PE cross-linking (44). The dependence of grafting efficiency on the monomer and initiator concentration is indicative of a strong effect of PP additions on the yield of grafted product (Table 10.4). Irrespective of DCP and MAH ratio in a reacting system, grafting efficiency of PE/PP blends exceeds that of neat PE. According to Chaoqin Li et al. (44), this can be explained by a lower viscosity of PE/PP melt against neat PE, which makes diffusion of reagents easier and raises the homogeneity of the reacting system. On the basis of the IR spectral analysis of the grafted products, Chaoqin Li et al. (44) have concluded that in the PP/PE melt, MAH gets grafted to chains of both PE and PP. This conclusion is based on the fact that the values of characteristic band frequencies of carbonyl absorption of MAH grafted to PE/PP blends (1864.1 and 1785:6 cm1 ) are between frequency values for MAH grafted to PE (1865.2 and 1784:9 cm1 ) and to PP (1862.5 and 1785:8 cm1 ). However, if Chaoqin Li et al. (44) had separated the (PE/PP)-g-MAH blend into separate PE and PP fractions by some known methods used in polymer blend fractionation (e.g., temperature rising elution fractionation, TREF, or some others), then determination of grafting location by IR spectral analysis would have been much more certain and precise. Besides the study of MAH and DCP concentration effects, Chaoqin Li et al. (44) analyzed the effect of polymer components ratio on the parameters under

Chapter 10 Functionalization of Olefinic Polymer and Copolymer Blends in the Melt

285

Figure 10.2 Effect of DCP content (a) and MAH content (b) on melt viscosity of MAH grafted LDPE/PP blend. (From Reference (44) with Permission from Elsevier.)

Table 10.4 Effect of DCP and MAH on the Grafting Degree.

MAH

DCP

1.5 1.5 1.5 1.5 0.75 3.0

0.05 0.1 0.2 0.3 0.1 0.1

Grafting degree, % MAH-g-LDPE/PP MAH-g-LDPE (90/10 blend)

[From Reference (44) with permission from Elsevier].

0.42 0.69 0.71 0.68 0.51 0.90

0.47 0.71 0.74 0.79 0.54 0.98

286

Polyolefin Blends

Table 10.5 Dependence of Properties of Initial Polyolefins, Functionalized (PP/LDPE)g-IA Blends and Unmodified PP/LDPE Blends on the Ratio of Polymer Components. T, C 190 C Test material, wt%

a, %

PP PP-g-IA [99PP/1LDPE]-g-IA [95PP/5LDPE]-g-IA [75PP/25LDPE]-g-IA [50PP/50LDPE]-g-IA [25PP/75LDPE]-g-IA [5PP/95LDPE]-g-IA [1PP/99LDPE]-g-IA LDPE-g-IA LDPE 95PP/5LDPE 75PP/25LDPE 50PP/50LDPE 25PP/75LDPE 5PP/95LDPE

— 60.2 61.8 66.8 74.2 78.3 85.1 89.8 90.6 91.8 — — — — — —

230 C

Gel MFI, MFI, Ea, g/10 min g/10 min kJ mol1 content, % 4.9 15.6 16.9 17.3 16.6 5.7 5.5 0.1 0.2 0.3 7.4 5.2 6.1 6.6 6.8 7.1

11.4 16.2 17.0 17.4 16.2 17.1 16.5 0.5 0.7 1.6 18.0 13.3 15.5 15.7 16.0 17.1

40.8 1.5 0.3 0.3 7.0 53.6 53.2 77.8 62.0 81.0 43.0 45.4 45.1 42.0 45.1 42.6

— — — — — 3.4 5.7 27.4 22.1 16.7 — — — — — —

sm, kPa

Km, rel. unit

2.3 0.4 0.3 0.5 2.0 13.6 20.2 21.3 32.6 35.0 3.1 1.5 1.7 2.2 2.4 2.7

2.5 1.5 1.5 1.6 2.1 3.0 2.2 2.1 2.0 1.9 3.0 1.8 2.1 2.2 2.1 2.0

[From Reference (46) with permission from Wiley].

investigation. The PE/PP blend composition, however, was varied within a narrow range (100:0; 90:10; 80:20) with PE predomination. Therefore, the effects observed were negligible and the analysis of data from an earlier work (44) cannot allow general conclusions about mutual influence of polymer components on functionalization of these blends. The effect of PP and PE ratios has been studied earlier (45–47) when itaconic acid (IA) was grafted to their blends containing L-101 peroxide as the initiator. Grafting was done in the extruder reactor assembled on the base of Brabender plastograph (Duisburg, Germany) equipped with the dynamic mixer (48,49). The information showing the course of grafting reaction of IA onto PP, PE, and PP/PE blends is presented in Table 10.5 and Fig. 10.3. The ratios of PP and PE in the reacting mixture strongly influence the grafting efficiency (a values). The higher the PE concentration in PP/PE blends, the higher is the grafting efficiency, whereas a higher concentration of PP causes the a-value to decrease. The extent to which the addition of PE or PP influences a-values depends on their concentrations. For relatively low amounts (up to 25 wt%) of PE added, a-value increases at a higher rate and the relationship is nonlinear. For PE concentrations between 25 and 95 wt%, a values rise monotonically with the PE concentration. The increased grafting efficiency of IA, in contrast with the additive one, at PE below 25 wt% (the continuous phase in PP/PE blends is formed by the PP phase) can

Chapter 10 Functionalization of Olefinic Polymer and Copolymer Blends in the Melt

287

Figure 10.3 Effect of PE concentration in PP/PE Blends on IA grafting efficiency (a) (solid line). In Figs. 10.3 and 10.6 dotted lines stand for additive values of the tested variables calculated as the sum of the values for neat PP and PE (EPR) with account for their concentrations in blends under consideration. (From Reference (46) with permission from John Wiley & Sons.)

result from PE* macroradicals formed—as being initiated by PP* macroradicals. Judging by melt flow index (MFI) values PEs participate chiefly in the grafting reaction of IA and do not undergo recombination between themselves. It is quite probable, therefore, that at PE concentrations below 25 wt%, increased a-values, in contrast with additive ones, result from prevailing grafting of the monomer onto PE chains and not onto PP ones. The MFI values determined at T ¼ 190 C and P ¼ 5 kg were found to depend on blend composition in a more complicated manner than the grafting efficiency (Table 10.5). The data in Table 10.5 permitted distinguishing three groups of blends that showed different rheological behaviors depending on the ratio of PP to PE (45,46). The (PP/PE)-g-IA systems, for example, containing PP between 99 and 75 wt%—the continuous phase in PP/PE blends is formed by PP phase—showed higher MFI values (more than three times as high as MFI of the neat PP). Maximum magnitudes of MFI are typical of (95PP/5PE)-g-IA systems. Such blends, moreover, show a sharp decrease in the apparent activation energy of viscous flow Ea, up to its negative values for the (75PP/25PE)-g-IA system (Table 10.5). For PP/PE blends with polymer component ratios of 50:50 and 25:75, both components most likely formed the continuous phase in the blend and, as a result, the apparent viscosity was observed to rise sharply (MFI decreased). These blends have MFIs close to the values for the neat PP and PE. A further increase in the PE concentration (compositions of 5:95 and 1:99 in which PE makes the continuous phase) led to still much lower MFI values. At the same time, Ea was observed to grow (Table 10.5). It is noticeable that (PP/PE)-g-IA systems containing 1 and 5 wt% of PP have lower MFIs than PE-g-IA. The a values for these compositions are lower than additive ones (Fig. 10.3). The pattern of dependence of MFI (determined at T ¼ 230 C and P ¼ 5 kg) on the composition of (PP/PE)-g-IA systems is roughly the same as at T ¼ 190 C and

288

Polyolefin Blends

Figure 10.4 Scheme of probable chemical reactions occurring at free-radical grafting of IA onto PP/PE blends. (From Reference (46) with permission from John Wiley & Sons.)

P ¼ 5 kg (Table 10.5). The only difference is that a greater number of systems differing in composition show higher MFIs (several times exceeding MFI of the PP-g-IA and PE-g-IA), while the viscosity begins to increase sharply only with <25 wt% of PP. It is quite probable that at T ¼ 230 C, a low viscous PP-g-IA melt forms a continuous phase up to its concentration of 25 wt%. It is only after the phase inversion at PP <25 wt% that the viscosity of blended systems depends on the viscosity of PE-g-IA, which forms a continuous phase. The related viscosity of the materials varies in the inverse manner to MFI variations (Table 10.5). The variations observed when IA was grafted to PP/PE blends can be explained, at present, in terms of the concepts about the mechanisms of free-radical processes (initiated by organic peroxides) occurring in molten PP and PE (Fig. 10.4). A peroxide is used to produce macroradicals by abstracting a hydrogen atom from the polymer substrate when the substrate interacts with free radicals that have been formed by the thermolysis of the peroxide used. The macroradicals can participate in subsequent chemical reactions such as grafting of a monomer, crosslinking of the chains, their degradation, and so on (Fig. 10.4). A relatively low grafting efficiency that is typical of both the neat PP and the PP/PE blends (containing between 75 and 99 wt% of PP) can be explained by a higher rate of b-scission of PP chains than the IA-grafting rate (10,11). The fact of a higher rate of b-scission is indicated by a marked increase in MFI for PP-g-IA and (PP/PE)-g-IA systems with high PP content, in comparison with MFI for the neat PP (Table 10.5). During grafting in PP and PE, there occur opposing side reactions (degradation and crosslinking of the chains). It can be assumed, therefore, that certain quantities of

Chapter 10 Functionalization of Olefinic Polymer and Copolymer Blends in the Melt

289

one or another polymeric component introduced into a PP/PE blend can control the extent of macromolecular transformations and rheological behavior of polymer melts. However, it is evident from Table 10.5 that a small quantity of PP does not increase but, on the contrary, leads to a lower MFI for (PP/PE)-g-IA systems in comparison with PE-g-IA. In turn, a low quantity of PE, despite expectations, appears to raise MFI of functionalized PO blends against PP-g-IA. In our opinion, these facts seem to be of importance when attempts are made to understand general mechanisms of the transformations observed. It is obvious that, for (PP/PE)-g-IA blends containing PE below 25 wt% when PP is as a continuous phase, degradation of PP chains prevails over macroradicals recombination and the former proceeds faster in comparison with PP-g-IA. This can be explained as follows. First, it is quite probable that the rate of reaction between PP chains and free radicals (formed during the decomposition of the peroxide initiator) is higher for PP than that for PE. For PP, an H atom is abstracted from a tertiary carbon atom, and not from a secondary one, as is the case with PE. The formed radical would rather tend to undergo b-scission than to recombination. Most likely because of this, it is difficult to produce grafted copolymers such as PE-g-PP, the presence of which could influence the rheological properties of the materials. Another obstacle to obtaining PE-g-PP copolymers (43) could be the two-phase structure of PP/PE blends. This type of structure is also found in their melts. As a result, each of the polymers here may be involved in free-radical processes independently of other. For such a situation, however, additive variations in MFI should be anticipated. Obviously, with equal component concentrations (about 1:1 ratio) at which two concurrent continuous phases are likely to form, their effect on MFI depends on the conversion degree in each of the blended components. The MFI for (PP/PE)-g-IA systems containing small quantities of PP, when PE is as a continuous phase, against MFI for PE-g-IA could drop owing to a higher concentration of PE macroradicals that could be initiated by products of PP chain degradation through the radical mechanism. Besides, in the presence of PP there is a great probability that PE macroradicals be formed through mechanochemical reactions, which are also typical of the reactive extrusion process (5). The information on MFI and Ea values determined for the initial, that is, unfunctionalized blends of PP/PE (Table 10.5) evidences to the prevailing effect— on the values examined—of specific chemical transformations that occur during grafting of IA. When the ratio of the components in a mixture of PP/PE was varied, MFI and Ea values fell within the range typical of the initial components irrespective of the phase structure of the materials. This fact evidences to the absence of any specific interactions in PP/PE blended systems prepared at our experimental conditions. The above reasoning is logically supported by data on gel content in the initial PP/PE as well as functionalized (PP/PE)-g-IA blends (Table 10.5). It has been found that in functionalized blends with prevailing PP component, cross-linking of PE is completely suppressed. The functionalized blends with small additions of PP (between 1 and 5 wt%) are characterized by a higher gel content in comparison

290

Polyolefin Blends

with PE-g-IA. These data corroborate the above reasoning about a possibility of initiating grafting and cross-linking of PE at the expense of PP* macroradicals formed in the course of macromolecular transformations. It is worth mentioning that a maximum gel content (27.4%) was found in the (5PP/95PE)-g-IA. Table 10.5 shows that this proper composition has the lowest MFI (0.1 g/10 min) and the highest Ea (77.8 kJ mol1) for all the blends tested. As it should be expected, at the conditions of the experiment—when neat PP and PE are mixed—no cross-linked structures are formed and gel fraction is not detected (45,46). Chemical macromolecular changes at RE significantly affect the high elastic properties of (PP/PE)-g-IA melts, which are characterized by strength (s m ) and swell coefficient of melt jet (Bm). Postextrusion swelling and strength of polymer melts are of primary importance in the manufacture of films or hollow vessels by blow extrusion. PP, unlike PE, has low strength in melt; this fact reduces its application. There is little information about the effect of functionalization on s m and Bm for POs as well as their blends. It was of interest, therefore, to study changes in these factors caused by grafting IA to PP/PE blends. Table 10.5 shows a strength behavior of molten (PP/PE)-g-IA depending on the ratio of PP and PE. When comparing the systems containing dominant PP with those containing dominant PE, a considerable increase in the strength of the molten material was observed with higher PE contents. The (PP/PE)-g-IA systems containing high quantities of PP (between 75 and 100 wt%) show extremely low strength of the molten material in comparison with both neat PP and (PP/PE)-g-IA systems containing large quantities of PE. As the strength of the molten material mostly depends on the molecular characteristics of the polymers (molecular weight, molecular weight distribution (MWD), degree of branching, etc.), as well as on the intermolecular interactions in the blends and their phase structure, it is quite possible that a low strength of the molten (PP/PE)-g-IA systems containing higher PP concentrations results from its largely reduced molecular weight caused by degradation processes involved in IA grafting. The strength of the molten material increases for (PP/PE)-g-IA systems of equal polymer component ratios and for systems containing higher amounts of PE. This can be explained by a continuous phase developed by partially cross-linked PE (Table 10.5), which leads to a growth in the melt viscosity and, obviously, to intensified interactions of phases in the blends. Unlike the strength of the molten material, the dependence of swell index on the ratio of polymeric components in (PP/PE)-g-IA systems is of extreme nature (Table 10.5). The (PP/PE)-g-IA systems with the polymer component ratio 1:1 show a maximum swell index. The swelling effect depends on the elastic and relaxation properties of polymers. Those characterized by a wide MWD and a high melt viscosity usually have higher swell index (50,51). It is quite possible that blends with equal weight ratios of the polymers have a most wide MWD owing to reactions involved in IA grafting. On the basis of the experimental data presented, it can be understood that a reduction in molecular weight typical of blends with increased concentrations of PP, same as the presence of cross-linked structures typical of blends with an increased concentration of PE, leads to a reduced elasticity of the melt, which results in a reduced swell coefficient for such systems.

Chapter 10 Functionalization of Olefinic Polymer and Copolymer Blends in the Melt

291

It is evident that s m and Bm for unfunctionalized PP/PE blends (Table 10.5) are somewhat lower than the additive ones irrespective of the phase structure of the blends. An obvious reason for this is the absence of intensive interactions between the phases in PP/PE binary blends as well as in cross-linked structures. These results and also the results for the initial and functionalized blends (Table 10.5) prove the determinant role of chemical processes (especially secondary reactions of degradation and cross-linking)—occurring when IA is being grafted to PP/PE blends—in the rheological and high elastic properties of the molten PP/PE blends (47).

10.3.2 Structure and Mechanical Properties of Functionalized PP/PE Blends Table 10.6 shows results of DSC analysis for the materials tested. The heating and cooling data for (PP/PE)-g-IA systems show the phase transitions typical of the homopolymers, which indicates a lack of compatibility between the crystalline phases of the polymers. However, the temperature-dependent location of the peaks that describe the phase transitions in (PP/PE)-g-IA systems changes with the blend composition and does not usually coincide with respective values for the homopolymers. It should be underlined that the crystallization temperature of the polypropylene component in the blends is 5–11 C higher than that of the PP homopolymer, whereas the typical variations in the melting point of PP are less significant. The (PP/PE)-g-IA systems of 99:1 ratio show a maximum Tcr for the Table 10.6 Results of DSC Analysis of (PP/LDPE)-g-IA Systems and Unmodified PP/LDPE Blends. Polypropylene component Test material, wt%

Tm, C

Tcr, C

DIcr , Exp.

PP PP-g-IA [99PP/1LDPE]-g-IA [95PP/5LDPE]-g-IA [75PP/25LDPE]-g-IA [50PP/50LDPE]-g-IA [25PP/75LDPE]-g-IA [5PP/95LDPE]-g-IA [1PP/99LDPE]-g-IA LDPE-g-IA LDPE 75PP/25LDPE 50PP/50LDPE 25PP/75LDPE

163.0 167.0 163.0 162.0 164.0 163.0 163.0 — — — — 162.0 163.0 162.0

113.0 121.0 125.0 119.0 118.0 119.0 119.0 — — — — 111.0 110.0 111.0

1.0 1.1 1.03 1.1 1.04 0.9 0.5 — — — — 0.75 0.57 0.27

[From Reference (47) with permission from Wiley].

Polyethylene component

DIcr , Calc. Tm, C 1.0 1.0 0.99 0.95 0.75 0.5 0.25 — — — — 0.75 0.5 0.25

— — — — 106.0 106.0 106.0 103.0 107.0 107.0 106.0 106.0 107.0 107.0

Tcr, C — — — — 93.0 95.0 94.0 94.0 95.0 90.5 92.0 90.0 91.0 92.0

DIcr , DIcr , Exp. Calc. — — — — 0.5 0.7 0.5 0.8 0.9 0.86 1.0 0.2 0.5 0.7

— — — — 0.25 0.5 0.75 0.95 0.99 1.0 1.0 0.25 0.5 0.75

292

Polyolefin Blends

polypropylene component. The polyethylene component in the blends shows a higher crystallization temperature than that for the neat one. The melting point of the PE component in (PP/PE)-g-IA systems varies but extremely little in comparison with those of neat PE (47).The comparison of experimental and calculated values of DIcr ¼ DH=DHo (Table 10.6) revealed rather strong variations in crystallization kinetics of the components in (PP/PE)-g-IA systems. For all of the ratios investigated, DIcr exceeds its calculated values for the PP phase. If the PE concentration in the blend increases, the DIcr value grows and reaches twofold the level for (25PP/75PE)-g-IA systems. An obvious and probably the major cause of this increase may be PP degradation at IA grafting, which is followed by a reduced molecular weight of PP, thus making crystallization easier (47,52). Unlike the PP phase, PE crystallizes slower in (PP/PE)-g-IA systems, with its concentration up to 50 wt%, against the calculated data (Table 10.6). It is explained by the growing melt viscosity of PE during functionalization, which makes the process of crystallization kinetically more difficult. The systems that give low viscous melts crystallize much more easily; the values of DIcr , therefore, differ. The variations in Tm, Tcr, and DIcr , stated above for the components of (PP/PE)g-IA blends, are due to the specificity of chemical macromolecular transformations that occur during IA grafting and not due to the simple mutual influence of PP and PE during their melt mixing. For instance, a comparison of the results in Table 10.6 shows that, for unmodified PP/PE mixtures, Tcr of the PP phase not only increased but also, on the contrary, it dropped somewhat (crystallization is hindered) in comparison with the neat PP. The values of DIcr for PP as well as PE phases are close to the calculated ones. This fact can be explained by relatively weak mutual influence of PE and PP on their crystallizability in unmodified PP/PE mixtures and by the fact that neither PP nor PE undergoes substantial chemical changes. Thus, the DSC results allow making a presumption that on cooling (PP/PE)-g-IA systems, their components undergo crystallization without involving a foreign phase into the crystallites. There is incompatibility between the components on the level of crystalline phases. The variations in the rate of crystallization and in the crystallinity degree of the components, along with Tcr values, serve as indirect evidences for rather intensive interactions taking place between the phases in (PP/PE)-g-IA systems (47). Analyses of (PP/PE)-g-IA by the method of dynamic mechanical losses using the inverted torsion pendulum tester (47) show a lack of compatibility between amorphous phases of the polymeric components. Two glass transition peaks, related to the glass transition of the components, were observed in the tan d versus temperature plots for the (PP/PE)-g-IA systems. However, the Tg values of the (PP/PE)-g-IA systems approach one another in contrast to those of the initial components, indicating that in the (PP/PE)-g-IA systems interactions between PP and PE lead to partial miscibility (Table 10.7). The single b-relaxation peak found for these systems also indicates compatibility and that small structural units in the homopolymers are responsible for b-relaxation. The temperatures of relaxation transition for unmodified PP/PE blends correlate with those for (PP/PE)-g-IA (Table 10.7). However, the opposing shifting of Tg

Chapter 10 Functionalization of Olefinic Polymer and Copolymer Blends in the Melt

293

Table 10.7 Results of Relaxation Analysis of (PP/LDPE)-g-IA Systems and Unmodified PP/LDPE Blends. Polypropylene component Test material, wt% PP PP-g-IA [95PP/5LDPE]-g-IA [75PP/25LDPE]-g-IA [50PP/50LDPE]-g-IA [25PP/75LDPE]-g-IA [5PP/95LDPE]-g-IA LDPE-g-IA LDPE 25PP/75LDPE 50PP/50LDPE 75PP/25LDPE

Polyethylene component

b-transition, a-transition, b-transition, C C C 47.5 46.6 — — — — — — — — — —

7.3 7.4 7.4 5.1 2.4 1.8 — — — 3.6 4.3 5.6

— — — 125.5 130.6 131.6 132.4 131.5 132.7 128.6 129.3 129.0

a-transition, C

D Tg, C

— — — 25.2 25.1 25.2 25.6 28.2 29.8 24.6 24.8 25.8

— — — 30.3 27.5 27.0 — — — 28.2 29.1 31.4

[From Reference (47) with permission from Wiley].

values for PP/PE blends is somewhat less than for (PP/PE)-g-IA. Consequently, macromolecular transformations, including the possible formation of grafted copolymers, occurring during the stage of blend preparation in case of (PP/PE)-g-IA, are favorable for stimulated partial compatibility of the components. Figure 10.5a and b is stress–strain plots for the initial homopolymers, for PP-gIA, PE-g-IA, (PP/PE)-g-IA, and for binary PP/PE blends. It can be seen that the behavior of the plots changes significantly when initial homopolymers and their blends are compared with (PP/PE)-g-IA blends. It is interesting to note that most of (PP/PE)-g-IA blends and PP-g-IA, unlike initial homopolymers and binary PP/PE blends, lose their ability to large deformations (47). Figure 10.5a shows that even (PP/PE)-g-IA blends containing 75 wt% of PE display low relative elongation. A sharp rise in relative elongation occurs with 5:95 (PP:PE) ratio. Most probable reasons for this are a two-phase nature of (PP/PE)-g-IA samples, the imperfection of contact zones between the phases, and the prevailing degradation of PP chains that accompanies IA grafting onto PP and PP/PE blends (47). Thus, at grafting of polar monomers to PP/PE blends there are observed numerous specific features, which distinguish them from grafting to individual homopolyolefins. For example, when MAH was grafted, it was found that the melt viscosity of PE/PP blends varied but little with increasing the peroxide concentration because degradation and cross-linking took place simultaneously. Increased contents of MAH in the PE/PP (90:10) blend led to lower viscosity of (PP/PE)-g-MAH, though grafting of MAH to homopolyethylene gave an opposite result. The IR spectral analysis revealed that the monomer was grafted to the two phases simultaneously. When IAwas grafted to PP/PE blends, it was found that increased concentrations of PP in a blend reduced the yield of the grafted product. This can be explained by the

294

Polyolefin Blends

Figure 10.5 Tensile diagrams for (PP/PE)-g-IA systems (a) and unmodified PP/PE blends (b). (From Reference (47) with permission from John Wiley & Sons.)

tendency of PP to degrade during peroxide-initiated grafting when the formed macroradicals suffer degradation by the b-scission. With relatively low concentrations of PE in the PP/PE blends, a nonlinear rise (more visible in comparison with additive dependence) in the IA grafting efficiency is observed. For PE concentrations between 25 and 95 wt%, the dependency of grafting efficiency on composition approaches a linear pattern, that is, a rises monotonously with PE concentration. The data on the rheological behavior of molten (PP/PE)-g-IA systems suggest a complex nature of influence of the polymers on the course of free-radical chain transformations taking place during reactive extrusion. Small amounts of one or another polymeric component, introduced into a blend, cause some changes in the rheological properties of the molten (PP/PE)-g-IA systems. For instance, a low (below 25 wt%) quantity of PE, contrary to the expectations, would increase MFI while low amounts of PP would decrease MFI for (PP/PE)-g-IA systems against

Chapter 10 Functionalization of Olefinic Polymer and Copolymer Blends in the Melt

295

PE-g-IA ones. Here, blends with PE concentrations below 25 wt% are characterized by extremely low apparent activation energy of the viscous flow. For (75PP/25PE)-gIA systems, the activation energy can even be negative. The variations in the ratio of polymeric components in (PP/PE)-g-IA systems cause non-additive, complex changes in the swell index of a molten jet and in the strength of the molten blended materials under discussion. The results reported in this paper are characteristic not only of grafting of IA onto PP/PE blends carried out in the Brabender plastograph equipped with a dynamic mixer, but also of IA grafting performed in reactors of other types, particularly in a single-screw extruder.

10.4

FUNCTIONALIZATION OF PP/EPR BLENDS

The problem of PP/EPR functionalization is important from a practical standpoint because starting blends are a widely available commercial product: EPR is added to PP to increase impact strength and low temperature resistance (53). Functionalized PP/EPR blends have been used as modifiers of impact strength and rheological behavior of engineering thermoplastics melts (54). From the scientific viewpoint, the functionalization of PP/EPR blends attracts attention of researchers, similarly to PP/PE blends, because of the differences in the course of secondary reactions in certain POs that accompany grafting and because of the process selectivity. Mierau et al. (55) and Kamfjord et al. (56) have studied these problems experimentally while grafting MAH to PP/EPR blends. Mierau et al. (55) have studied MAH grafting to impact-resistant PP manufactured by BASF AG (Germany). This PP contains 45 wt% of EPR. MAH had been grafted in a quantity of 2 wt% in a twin-screw extruder (ZSK 30, Werner and Pfleiderer, Germany) at 180–210 C with 0.1 wt% of L-101. After the grafted product was fractionated, and each of the fractions analyzed, it was found (55) that MAH had been mostly grafted to EPR chains and not to PP. The rheological measurements revealed that grafting caused reduction in the molecular weight of both fractions. The EPR fraction, however, undergoes degradation to a greater extent than PP. One of the reasons for the events observed is, in Mierau et al.’s opinion (55), the higher dissolution of the low molecular weight reagents (MAH and L-101) in the EPR phase. This study proved the fact that the presence of two polymer phases in a reacting system leads to a variation in active concentrations of low molecular weight reagents within their volume. It has been mentioned already that dissolution of low molecular weight reagents in polymer media influences the course of reactive processes (1,17,31). Some authors (32,57) reported that in blends of PP with EPR and NBR, differences in solubility parameters of peroxides cause differences in the cross-linking degree of the rubbers. Another factor that determines the course of preferable grafting in one of the polymer phases is the so-called ‘‘structural factor.’’ This term implies effects of the chemical structure of polymer chains, molecular weight, conformation features, configuration, and mobility of hydrogen atoms in the chains on the reaction kinetics.

296

Polyolefin Blends

An attempt has been made (56) to understand the role of the structural factor when MAH was grafted to heterophase PP. The latter was a mixture of highly crystalline homopolypropylene, which made the matrix and copolymers dispersed in the PP matrix similar to rubber particles. During fractionation of the heterophase PP, three fractions were separated: PP (50 wt%), EPR with irregular distribution of ethylene units (43 wt%), and ethylene–propylene block copolymer (EPM) (2 wt%). Peroxide initiators were DTBP (di-tert-butyl peroxide), d ¼ 15:3ðJ cm3 Þ1=2 ; TBCP (tert-butyl cumyl peroxide), d ¼ 16:2ðJ cm3 Þ1=2 ; DCP (dicumyl peroxide), d ¼ 17:4ðJ cm3 Þ1=2 ; DTBPH (2,5-bis(tert-butyl peroxy)-2,5dimethylhexane), d ¼ 15:5ðJ cm3 Þ1=2 . The analysis—by titration procedures and IR spectroscopy—of each of the separated phases showed that MAH becomes mostly grafted to EPR chains, and a small portion of the monomer combines with the PP. No differences have been detected in the effect of peroxide initiators that show solubility values close to that of EPR or PP. As for the selectivity of the reaction, it should be underlined that the nature of the free radicals formed is of importance. Decomposition of DTBP, as well as DTBPH, mostly gives methyl and butoxyl radicals, while TBCP and DCP generate a 2-phenylisopropoxy radical. In reactivity, they can be arranged as follows: CH3* (CH3)3CO* < C6H5(CH3)CO*. The higher the reactivity of a radical, the lower is its selectivity as expressed by an approximately equal transfer rate of hydrogen atoms from different substrates. The same authors (56), however, did not report any effects of free-radical nature on grafting selectivity. When different peroxides initiated grafting, the efficiency of MAH grafted to the EPR phase was much higher than to PP. It is worth mentioning that EPR was breaking down more actively than PP. The grafting reaction of MAH to a heterophase PP can be sterically controlled (56). Hydrogen atoms are mostly released from EPR chains containing numerous ethylene fragments. Considering that the abstraction rate of hydrogen rises in an order of primary < secondary < tertiary hydrogen atoms, it is reasonable to assume that a free radical preferably attacks hydrogen atoms bonded with tertiary carbon located either between ethylene units or terminate with propylene blocks. These tertiary atoms are less sterically protected by neighboring methyl groups, thus easily available for free radicals and huge molecules of MAH. It is clear, therefore, why MAH becomes grafted to the EPR phase, whereas the PP phase remains modified but only slightly. It is noteworthy that a similar result—a higher level of grafting to the EPR component—had been obtained by other authors (58) who investigated plasmainitiated grafting of polystyrene to PP/EPR blends. This was explained (58) by a high degradation rate of PP, which causes fast ruin of reactive centers in the PP phase. It was of interest—by analogy with PP/PE blends—to investigate effects of polymer ratios in PP/EPR blends on the IA grafting. Numerous experiments were performed, using Brabender plastograph, following a method as described elsewhere (45). The concentrations of the monomer and initiator (peroxide L-101) were constant and made 1 and 0.3 wt%, respectively. The PP to EPR ratio in the blends

Chapter 10 Functionalization of Olefinic Polymer and Copolymer Blends in the Melt

297

Figure 10.6 Effect of PP/EPR ratio on grafting efficiency (a) and MFI of (PP/EPR)-g-IA (b).

was varied between 0 and 100 wt%. The propylene concentration in EPR was 50 wt%. The comparison of data in Figs. 10.3 and 10.6 shows that a nonadditive character of the a-concentration dependence for PP/EPR blends is more distinct than for the PP/PE. For the whole range of components ratios, the grafting values for IA to PP/EPR blends exceed additive values. Generally, the tendencies of a to change, with increased concentrations of both PE and EPR, are similar, which is indicative of generality of mechanisms of macromolecular changes in both types of the materials. The analysis of MFI values (Fig. 10.6b) shows that with a PP concentration up to 25 wt%, the blends have a low MFI value (the melt viscosity is high). Low additions of EPR, same as PE additions (Table 10.5), cause an increase in MFI values. Since propylene units belong to the molecular structure of EPR, it appears that at equal concentrations of PE and EPR in blends with PP, the concentration of hydrogen atoms, bonded to tertiary carbon atoms, is higher in the second case. That is why, specificities of IA-grafting reaction, related to the concentration and the reactivity of macroradicals, are more pronounced in PP/EPR than in PP/PE blends. Thus, during monomer grafting to PP/EPR blends, there is observed an unusual mode of variation in the yield of grafted product as well as in the rheological behavior of melts with variations in component ratio. When MAH is grafted to such blends, mostly the EPR phase becomes functionalized. In a wide range of the component ratio in PP/EPR blends, higher values of grafting efficiency are typical in comparison with individual polymer components making the blend. The explanation may be that there is an increased reactivity of the hydrogen atoms bonded to tertiary carbon that is present in polypropylene units of EPR.

10.5 FUNCTIONALIZATION FEATURES OF BLENDS: PE/EPR, PP(PE)/EOC, AND PP(PE)/STYRENE POLYMER Functionalized PE/EPR blends are attractive because of a possibility to control the rheological and high elastic properties of their melts by varying the components

298

Polyolefin Blends

Table 10.8

Grafting Efficiency and Rheological Properties of Molten PO/EOC Blends. T, C 190 C a

Test material, wt%

a, %

PP EOC PP-g-IA [85PP/15EOC]-g-IA [70PP/30EOC]-g-IA [50PP/50EOC]-g-IA EOC-g-IA PE PE-g-IA [85PE/15EOC]-g-IA [70PE/30EOC]-g-IA [50PE/50EOC]-g-IA

— — 61.4 77.9 76.4 81.3 82.0 — 85.4 85.4 88.7 86.3

MFI , g/10 min 4.7 7.6 14.5 16.6 17.7 15.9 0.6 5.1 0.5 0.9 1.0 1.3

230 C a

h, k Pas

MFI , g/10 min

h, kPa s

10.8 6.7 9.0 2.9 2.8 2.9 84.6 10.8 104.6 57.8 56.7 39.8

12.2 17.0 16.9 17.5 17.8 17.9 2.6 11.1 2.4 3.9 4.1 5.2

4.5 3.0 3.0 3.0 2.8 2.9 19.0 4.5 19.8 13.3 12.4 9.7

a

MFI was determined at P ¼ 50 N

ratios (54). It was found that both neat PE and PE/EPR blends undergo, predominantly, cross-linking during grafting. The cross-linking degree, as in the case of PE, can be controlled by varying the monomer-to-initiator ratio. An increase in the EPR concentration in a blend raises the melt viscosity and strength, which indirectly points at an intensified cross-linking reaction (55). In recent years, a number of studies have been published on grafting of monomers, for example, MAH (54,60), undecylenic acid (61), and silanes (62,63) to higher a-olefins, particularly, ethylene–octene copolymers (EOC). A possibility to obtain high yields of a grafted product when grafting was done by RE process was reported. It was of interest to estimate the possibility of functionalizing the abovementioned copolymers with homopolypropylene and homopolyethylene. The studies were done on IA grafting in presence of L-101 peroxide. The grafting experiments followed the same procedure as for PP/EPR blends described in Section 10.4, which allowed a comparison of the results obtained. For all of the experiments, the IA concentration was 1 wt% and that of L-101 was 0.25 wt%. As EOC, Engage 8454 was used as supplied by Du Pont Dow Elastomers Co. Ltd. (density 0:875 g cm3 ; MFI ¼ 3 g=10 min at T ¼ 190 C and P ¼ 21:6 N; Tg ¼ 53 C; Tm ¼ 67 C; average degree of crystallization, 20%). Table 10.8 shows that irrespective of the basic PO, an addition of EOC causes a rise in the grafting efficiency. At grafting, the two types of blends show some drop in the viscosity (higher MFI) in comparison with homopolyolefins. The lower melt viscosity can explain, to some

Chapter 10 Functionalization of Olefinic Polymer and Copolymer Blends in the Melt

299

degree, an increased yield of the grafted product considering the kinetics (1). Somewhat lower grafting efficiency of IA to EOC, against PE, results from the fact that in case of EOC, a greater portion of free radicals is consumed in intermolecular cross-linking reactions. This assumption has been supported by rheological measurements (Table 10.8). Therefore, an introduction of EOC into homopolyolefins is advantageous from a practical viewpoint because it is possible to increase the grafting efficiency and targeted influence on the rheological properties of molten blends. Functionalization of styrene polymers and their blends have been treated in several studies (3) dealing with the preparation of compatibilizers for blends containing engineering thermoplastics. It is worthy to note that in presence of peroxides (DCP), polystyrene (PS) suffers preferable degradation in the melt (64,65). The grafting power of MAH and its derivatives (diethyl maleate, dimethyl maleate) is much lower to PS than to PO, and this fact—in Xue et al.’s opinion (64)—indicates a low reactivity of benzyl macroradicals. A comparative study of MAH and diethyl maleate grafting—initiated by DCP—to styrene–(ethylene–butylene)–styrene triblock copolymer (SEBS) and PS blended with a random ethylene–1-butene copolymer has been described elsewhere (66). The weight ratio of PS to PO components in the block copolymer was the same as in the blend (20% PO and 80 wt% PS). The analysis of the functionalized PS/PO blend—after PS had been extracted in a solvent—showed that the monomer was grafted only to the PO component. Considering that the PS/ PO blend and SEBS have quite similar phase and molecular structures, it was concluded that in styrene-containing block copolymers only aliphatic blocks become functionalized. Thus, during functionalization of PO blends containing styrene polymers, the grafted functionalized groups concentrate within the aliphatic phase. The course of secondary reactions (cross-linking and degradation) is controlled, such as in the case of grafting to neat PO, by variations in the monomer-to-initiator ratios (3).

10.6

USE OF FUNCTIONALIZED POLYOLEFIN BLENDS

The major field of application of functionalized PO blends and functionalized homo-polyolefins is compatibilization of blends of condensation polymers. The use of functionalized PO blends, in place of homopolyolefins, has a number of advantages. First, by ratio variation of PO in the blend, it is possible to change the degree of microheterogeneity of the final product; this fact is of importance for impact strength (67). Second, the selective nature of grafting can be useful in controlling the functionalization power of one or another component in PO blends. Additions of different types of PO can be used to control the mechanical characteristics of the blends, the melt viscosity of functionalized PO blends, their high elastic properties, which is important for widening the technological opportunities and the properties of composites. Of no lesser importance is the fact that an introduction of a cheaper component into functionalized PO blends is economically

300

Polyolefin Blends

advantageous. Recently, published information has become available that supports the efficiency and the practical usefulness of functionalized PO blends in the manufacture of plastics for engineering applications (54,68–70). It can be expected that blends of functionalized POs will be more efficient compatibilizers for multicomponent blended systems containing homopolymers. For example, the g-PP/g-PE blend should a priori be more efficient than g-PP or g-PE introduced separately in the PA/PP/PE material. In addition to the above-mentioned fields of application, functionalization of blends of olefin polymers and copolymers has been widely employed in the manufacture of thermoplastic dynamic vulcanizates (71,72).

10.7 CONCLUSION The analysis of experimentally obtained data on free-radical grafting (at RE) of unsaturated polar monomers to blends of PP/PE, PP/EPR, PE/EPR, PP(PE)/ECO, and PP(PE)/styrene polymers has indicated that there is a number of features that distinguish this process from the functionalization of homopolyolefins. A selective nature of grafting should be underlined in such systems as PP/ EPR (in presence of DCP, MAH becomes grafted to the EPR phase in the most), and PO/styrene polymer (grafted functionalized groups are concentrated in the aliphatic phase). The grafting selectivity shows rather a general behavior. This can be explained by differences in the thermodynamic affinity of low molecular weight reagents (monomer and initiator) with individual components of a PO blend, along with the differences in the reactivity of unlike chains in free-radical reactions. To ensure satisfactory grafting to each of the components making a PO blend, it is reasonable to use peroxide initiators that have a stronger affinity with these components. In case of a great difference in the solubility parameters of the components, it seems reasonable to use a mixture of initiators that do not react chemically between themselves but are soluble in certain PO components. When monomers are chosen for grafting, 1,2-disubstituted ethylenes should be preferred because they are not disposed to homopolymerization and show a weak thermodynamic affinity with the peroxide initiator. A mutual influence of polymer components in PO blends at grafting can cause unusual changes in the rheological behavior of the melts. For example, when MAH was grafted to PP/PE (90:10) blends, it was found that increased concentrations of a peroxide initiator (DCP) led to negligible variations in the viscosity of polymer melts because of a competitive effect of side reactions (degradation and cross-linking). It is well known, however, that neat PE, being a basic component of the blend under discussion, shows an increased melt viscosity as well as higher contents of the gel fraction with increasing peroxide concentration. It was found for PP/PE blends functionalized by IA grafting, as initiated by L-101 peroxide, that low (up to 25 wt%) quantities of PE (which undergoes crosslinking during functionalization) caused a decrease in the melt viscosity of

Chapter 10 Functionalization of Olefinic Polymer and Copolymer Blends in the Melt

301

(PP/PE)-g-IA in comparison with PP-g-IA. Small amounts (1–5 wt%) of PP (which suffers degradation during functionalization) increased the melt viscosity of the (PP/ PE)- g-IA against the PE-g-IA. The data obtained are indicative of a strong mutual influence of polymer components on the course of the basic process (grafting of monomer) as well as concurrent reactions (degradation and cross-linking of the chains). This is supported by a non-additive behavior of variation in the yield of the grafted product, as well as the rheological properties of molten functionalized blends of PO. Functionalization of PO blends, as a rule, leads to an improved compatibility of the components and to variations in the parameters describing their semicrystalline structure. Blends of functionalized PO have mostly been used for compatibilization of blended systems based on engineering plastics. They have found applications in the development of thermoplastic dynamic vulcanizates.

NOMENCLATURE Bm CTC DSC Ea DEi DH DHo k DIcr MFI Na P T* Tm Tcr Tg DTg DVi a b d h sm t 0:5

Swell coefficient of melt jet Charge transfer complex Differential scanning calorimetry Apparent activation energy of viscous flow Contribution of every atom and type of the intermolecular interaction in the molar cohesion energy of the substance Crystallization heats of the polymers in a blend Crystallization heats of the polymers individually Coefficient (k 1 for PO; k 1:25 for peroxide initiator and monomer grafted) Crystallinity index Melt flow index Avogadro number Load Temperature at which a maximum decomposition rate of peroxide is reached Melting temperature Crystallization temperature Glass transition temperature Difference between Tg of PP component and Tg of PE component van der Waals volume of the atom Grafting efficiency Thermal (linear) expansion coefficient Solubility parameter Viscosity Melt strength Half-life of a free-radical initiator

302

Polyolefin Blends

REFERENCES 1. G. Moad, Prog. Polym. Sci., 24, 94 (1999). 2. S. B. Brown, Annu. Rev. Mater. Sci., 21, 409 (1991). 3. F. Ciardelli, M. Aglietto, M. B. Coltelli, E. Passaglia, G. Ruggeri, and S. Coiai, Functionalization of polyolefins in the melt, in: Modification and Blending of Synthetic and Natural Chains, F. Ciardelli and S. Penczek (eds.), Kluwer Academic Publishers, Dordrecht, 2003, Chapter 4. 4. N. C. Liu and W. E. Baker, Adv. in Polym. Tech., 11, 249 (1992). 5. M. Xanthos (ed.), Reactive Extrusion, Hanser, Munich, 1992. 6. S. Al-Malaika (ed.), Reactive Modifiers for Polymers, Chapman and Hall, London, 1996. 7. T. C. (Mike) Chung, Functionalization of Polyolefins, Academic Press, San Diego, 2002. 8. S. Datta and D. J. Lohse, Polymeric Compatibilizers: Uses and Benefits in Polymer Blends, Hanser/ Garder Publications Inc., Munich, 1996. 9. S. S. Pesetskii, B. Jurkowski, Y. M. Krivoguz, and R. Urbanowicz, J. Appl. Polym. Sci., 65, 1493 (1997). 10. A. V. Machado, Van M. Duin, and J. A. Covas, J. Polym. Sci. A Polym. Chem., 38, 3919 (2000). 11. A. V. Machado, J. A. Covas, and Van M. Duin, Polymer, 42, 3649 (2001). 12. S. S. Pesetskii, B. Jurkowski, and O. A. Makarenko, J. Appl. Polym. Sci., 86, 64 (2002). 13. Y. M. Krivoguz, S. S. Pesetskii, and Y. M. Pleskachevsky, Polym. Sci. Ser. A, 46, 698 (2004). 14. S. S. Pesetskii and B. Jurkowski, Reactive extrusion for processing of functionalized Polymers and Blended compositions, in: Proceedings of Polymer Composites-98 International Conference, Gomel, 1998, p. 36. 15. A. H. Hogt, J. Jelenic, and J. Meijer, Chemical modification of polypropylene by organic peroxides, in: Polymer Processing Society, European Meeting, 26–28 September, Stuttgart, 1995, p. 3–21. 16. S. S. Pesetskii, Y. M. Krivoguz, and A. P. Yuvchenko, Russ. J. Appl. Chem., 71, 1364 (1998). 17. S. S. Pesetskii, B. Jurkowski, Y. M. Krivoguz, and K. Kelar, Polymer, 42, 469 (2001). 18. S. S. Pesetskii and O. A. Makarenko, Russ. J. Appl. Chem., 75, 643 (2002). 19. L. F. Chen, B. Wong, and W. E. Baker, Polym. Eng. Sci., 36, 1596 (1996). 20. V. L. Antonovskii, Organic Peroxide Initiators, Khimia, Moscow, 1972 (in Russian). 21. R. Grego, G. Maglio, P. Musto, and G. Scarinzi, J. Appl. Polym. Sci., 37, 777 (1989). 22. A. D. Jekins and A. Ledwit, Reactive Mechanism and Structure in Polymer Chemistry, Wiley, London, 1977. 23. G.-H. Hu and M. Lambla, Materials Science and Technology—A comprehensive treatment, 18, 345 (1997). 24. M. Lambla, Macromol. Symp., 83, 37 (1994). 25. G.-H. Hu, H. Li, L. Feng, and L. A. Pessau, J. Appl. Polym. Sci., 88, 1799 (2003). 26. R.-A. Tan, W. Wan, G.-H. Hu, and A. S. Gomes, Eur. Polym. J., 35, 1979 (1999). 27. T. Vainio, G.-H. Hu, M. Lambla, and J. Seppala, J. Appl. Polym. Sci., 63, 883 (1997). 28. K. E. Russel, Prog. Polym. Sci., 27, 1007 (2002). 29. G. Moad and D. H. Solomon, The Chemistry of Free Radical Polymerization, Pergamon, Oxford, 1995. 30. B. De Roover, J. Devaux, and R. J. Legras, J. Polym. Sci. A Polym. Chem., 34, 1195 (1996). 31. T. Bray, S. Damiris, A. Grase, G. Moad, M. O’skea, E. Rizzardo, and G. Van Diepen, Macromol. Symp., 129, 109 (1998). 32. A. H. Hogt, Solubility Aspects of Peroxides in Modification of Polymers and Polymer Blends, in: 2nd Conference on Advances in Additives and Modifiers for Polymer Blends, Philadelphia, 1993, pp. 1–17.

Chapter 10 Functionalization of Olefinic Polymer and Copolymer Blends in the Melt

303

33. H.-Q. Xie and W. E. Baker, Novel azoinitiators for polyolefin modification and grafting, in: New Advances in Polyolefins, Chung T. G. (ed.), Plenum, New York, 1993, p. 101. 34. B. Turcsanyi, J. Appl. Polym. Sci., 76, 1976 (2000). 35. B. Turcsanyi, Polym. Bulletin, 30, 297 (1993). 36. B. Turcsanyi, Polym. Bull., 32, 713 (1994). 37. D. C. Nonhebel and J. C. Walton, Free-Radical Chemistry: Structure and Mechanism, University Press, Cambridge, 1974. 38. B. Wong and W. E. Baker, Annu. Tech. Conf. Soc. Plast. Eng., 54, 283,1996. 39. H. Hung and N. C. Liu, J. Appl. Polym. Sci., 67, 1957 (1998). 40. D. W. van Krevelen, Properties of Polymers: Correlations with Chemical Structure, Elsevier Publishing Co., Amsterdam, 1972. 41. A. A. Askadskii and Y. I. Matveev, Chemical Structure and Properties of Polymers, Khimia, Moscow, 1983 (in Russian). 42. M. M. Coleman, C. J. Serman, D. E. Bhagwagar, and P. C. Painter, Polymer, 31, 1187 (1990). 43. D. Braun, S. Richter, G. P. Hellmann, and M. J. Ra¨tzsch, J. Appl. Polym. Sci., 68, 2019 (1998). 44. Chaoqin Li, Yong Zhang, and Yinxi Zhang, Polym. Test., 22, 191 (2003). 45. Y. M. Krivoguz and S. S. Pesetskii, Russ. J. Appl. Chem., 78, 305 (2005). 46. S. S. Pesetskii, B. Jurkowski, Y. M. Krivoguz, T. Tomczyk, and O. A. Makarenko, J. Appl. Polym. Sci, 102, 5095 (2006). 47. Y. M. Krivoguz, S. S. Pesetskii, B. Jurkowski, and T. Tomczyk, J. Appl. Polym. Sci., 102, 1746 (2006). 48. S. S. Pesetskii, B. Jurkowski, Y. M. Krivoguz, and Y. A. Olkhov, J. Appl. Polym. Sci., 81, 3439 (2001). 49. B. Jurkowski and B. Jurkowska, Sporzadzanie Kompozycii Polymrowych. Elementy Teorii i Praktyki (Polymer Compounding, Elements of Theory and Practice) Wydawnictwa Naukowo-Techniczne, Warsaw, 1995. 50. A. C.-Y. Wong and V. H. K. Cheung, Materials Processing Technology, 67, 117 (1997). 51. A. H. Dekmezian, W. Weng, C. A. Garcia-Franco, and E. J. Markel, Polymer, 45, 5635 (2004). 52. B. Wunderlich, Macromolecular Physics, Vol. 1, Crystal Structure, Morphology, Defects, Academic Press, New York, 1973. 53. E. V. Prut and A. N. Zelenetskii, Russian Chem. Rev., 70, 72 (2001). 54. A. A. Bogoslavsky, Extrusion processed engineering materials based on blends of polyamide-6 and functionalized olefin polymers and copolymers: Dissertation thesis, Cand Sci. (Techn.) 05.02.01, V. A. Belyi Metal-Polymer Research Institute of NAS, Belarus, Gomel, 2004. 55. U. Mierau, D. Voigt, F. Bo¨hme, and E. Brauer, J. Appl. Polym. Sci., 63, 283 (1997). 56. T. Kamfjord and Aa. Stori, Polymer, 42, 2767 (2001). 57. K. Naskar and J. W. M. Noordermeer, Rubber Chem. Technol., 76, 1001 (2003). 58. I. Keen, G. A. George, and P. M. Fredericks, J. Appl. Polym. Sci., 88, 1643 (2003). 59. K. Premphet and S. Chalearmthitipa, J. Appl. Polym. Sci., 41(11), 1978–1986 (2001). 60. C. S. Wu, H. T. Liao, and S. M. Lai, Polym. Plast. Technol., 41, 645 (2002). 61. P. He, H. Huang, Y. Zhang, and N. C. Lin, React. Func. Polym., 62, 25 (2005). 62. K. Sirisinha and D. Meksawat, J. Appl. Polym. Sci., 93, 901 (2004). 63. C. Jiao, Z. Wang, Z. Gui, and Y. Hu, Eur. Polym. J., 41, 1204 (2005). 64. T. J. Xue, D. Jiang, and C. A. Wilkie, Grafting of vinyl monomers onto polymers containing styrene in: 212th ACS National Meeting, Orlando, FL 25–29 August 1996 (Conference Paper). 65. E. Passaglia, S. Coiai, I. Ricci, and F. Crardelli, Polystyrene modification by reactive processing in the bulk, in: Atti XVI Convegno di Science e Tecnologia delle Macromolecole, Pisa 22–25 September, 2003. 66. E. Passaglia, S. Ghetti, F. Picchioni, and G. Ruggeri, Polymer, 41, 4389–4400 (2000).

304

Polyolefin Blends

67. D. R. Paul and S. Newman (eds.), Polymer Blends, Academic Press, New York, 1978. 68. G. H. B. Wang and X. Zhon, Mater. Lett., 58, 3457 (2004). 69. S. C. Wong and Y. W. Mai, Polymer, 40, 1553 (1999). 70. C. P. Papadopoulou and N. K. Kalfoglou, Polymer, 41, 2543 (2000). 71. A. Y. Coran, Thermoplastic elastomers based on elastomer–thermoplastic blends dynamically vulcanized, in: Thermoplastic Elastomers—A Comprehensive Review, N. R. Legge, G. Holden, and H. E. Schroeder (eds.), Hanser Publishers, Munich, 1987. 72. J. George, K. T. Varughese, and S. Thomas, Polymer, 41, 1507 (2000).

Chapter

11

Deformation Behavior of b-Crystalline Phase Polypropylene and Its Rubber-Modified Blends Sie C. Tjong1

11.1 INTRODUCTION Isotactic polypropylene (iPP) is one of the commodity polymers that find widespread applications in domestic sectors due to its low cost, favorable performance, and ease of processability. However, iPP exhibits low impact resistance, particularly at low temperatures that limits its application as an engineering thermoplastic. iPP is known to exhibit several crystallographic forms, that is a, b, g, and liquid crystalline. The monoclinic a-form dominates in the crystallization of the PP under general molding conditions. The b modification exhibits a trigonal lattice with a frustrated chain packing, while the g modification possesses an orthorhombic lattice with molecular epitaxy by chain cross 80 or 100 . The formation of b-form can be assisted by crystallization in a temperature gradient (1), by shear-induced crystallization (2), and by the incorporation of appropriate b-nucleators with concentrations 0.1 wt% (3–10). The performance of PP depends mainly on the final crystal morphology and structure. The a-crystal has higher yield stress, higher density, greater stiffness, and hardness (11,12). The b-form shows better mechanical ductility and toughness in terms of higher values of tensile elongation and higher impact resistance (13–16).

1 Department of Physics and Materials Science, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

305

306

Polyolefin Blends

Thus, control of the growth of the b-form polymorph using appropriate nucleating agents is of great scientific and technological importance. The b-nucleators employed include g-quinacridone (permanent red E3B), calcium stearate, pimelic acid, and N,N0 -dicyclohexylnaphthalene-2,6-dicarboxamide (NJ Star NU-100; New Japan Chemicals) (3–10). A major problem of the compounding PP with g–quinacridone is the contamination of the final products with red pigment color. Other deficiency includes a relatively low amount of b-PP phase induced (3). In addition to organic nucleating agents, mineral fillers such as wollastonite and carbonate are used to produce various levels of b-PP phase (11,12). Liu et al. reported that the relative content of b-PP phase approaches only 37% when 17.7 vol% wollastonite is added. For the carbonate-filled PP, the amount of b-PP phase in these samples is very low (12%) even adding the mineral content up to 40 wt%. Therefore, efforts have been made by several researchers to increase the bPP phase content by using effective nucleating agents. For example, Shi and Zhang reported that a high purity b-PP can be developed by adding a specific bicomponent b-nucleator consisting of equal amounts of pimelic acid and calcium stearate (6,7). The relative b-phase content can reach as high as 94% by adding only 0.1 wt% of such b-nucleator. Varga et al. synthesized the b-nucleators from calcium salts of suberic and pimelic acids. Such b-nucleators are found to be thermally stable, highly active, and colorless (8). In some cases, commercially available b-nucleating agent NJ Star NU-100 has been reported to achieve the b-phase content higher than 90% as evidenced by the X-ray diffraction (XRD) analysis (9,10). The content of the b-form PP is generally determined from the intensity of XRD peaks of a-form and b-form PP using the following equation (13): Ið300Þb ð11:1Þ K¼ Ið110Þa þ Ið040Þa þ Ið130Þa þ Ið300Þb where Ið300Þb is the intensity of the characteristic b-form peak (300), Ið110Þa ; Ið040Þa , and Ið130Þa are the intensities of the strong a-form peaks (110), (040), and (130), respectively. Moreover, the order parameter (S) for the b-form PP can be determined from (14): Ið300Þb ð11:2Þ S¼ Ið300Þb þ Ið301Þb where Ið301Þb is the intensity of the characteristic b-form peak (301). The order parameter reflects the order of chain packing in the chain direction. The higher the S value, the higher the order of the b-phase is. Figure 11.1a and b shows the typical XRD patterns of the skin and core sections of injection-molded b-PP nucleated from pimelic acid and calcium stearate (15). For the purposes of comparison, the XRD pattern of a-PP is also presented. The K value determined from the skin section is 0.54, and it implies that the relative content of the b-form PP in the skin layer is 54%. The calculated K value for the core section is 0.92. Apart from the K value, the structure and morphology of the b-phase spherulites play important roles in enhancing the mechanical and fracture resistances of PP. The spherulites of the a and b modifications are very different. The spherulites of the a

Chapter 11 Deformation Behavior of b-Crystalline Phase

307

Figure 11.1 XRD patterns of the b-form polypropylene taken from (a) skin layer and (b) core region of the injection-molded specimen. For the purposes of comparison, the XRD pattern of the a-form PP is shown in (b). (From Reference 15 with permission from The Society of Plastics Engineers.)

modification possess very low birefringence, while those of the b modification have high birefringence. This is because the lamellar structures are crosshatched in the spherulites of the a modification (17), while no such crosshatched lamellae can be found in the spherulites of the b modification. The spherulites of b-form PP are characterized by well-known sheaflike structure in which the boundaries between the spherulites are hardly distinguishable (Fig. 11.2a and b).The bundles of lamellae of neighboring spherulites tend to cross each other, as shown in Fig. 11.2c (18). Such lamellae bundles are linked by the tie molecules of the amorphous phase. Varga and coworkers have studied the morphology of the b-phase spherulites grown from the melt extensively (8,19,20). They reported that hedrites were formed as a precursor to the formation of b-spherulites at high crystallization temperatures and denoted them

308

Polyolefin Blends

Figure 11.2 Low magnification SEM micrographs of b-spherulites (a and b). (c) Higher magnification view of b-spherulites showing sheaflike structure. (From Reference 18 with permission from Elsevier.)

as quasi-spherulites. Hedrites are polygonal formations consisting of clusters of folded chain lamellar crystallites and can be termed as axialites, hexagonites, and ovalites depending on the view and the maturity of the structure (20). More recently, Varga and coworkers studied the melting, crystallization characteristics as well as the structure of polymer blends based on the b-PP (21). They reported that b-form PP cannot form in the iPP/PA-6 and iPP/polyvinylidene fluoride (PVDF) blends even in the presence of a highly effective b-nucleator. This is because of the strong a-nucleating ability and higher crystallization temperature range of PA-6 and PVDF. However, b-spherulites can develop in the blends of iPP/random propylene

Chapter 11 Deformation Behavior of b-Crystalline Phase

309

(rPP) copolymers. This is because the components of iPP/rPP blends are compatible in the molten state. Because of the difference in the lamellar morphology between the a- and b-spherulites, the deformation and fracture behavior of a-PP and b-PP differ distinctly. The a-crystal with higher density exhibits larger yield stress, greater stiffness and hardness (22,23). The b-form with lower density shows better tensile ductility and impact toughness (15,18,24). The lamellae bundles of b-PP, held together by tie molecules, can be easily detached from one another upon mechanical loading. This leads to the formation of microcrazes and microfibrils (15,18,25–27). Aboulfaraj et al. reported that the monoclinic a-PP phase exhibits brittle behavior when subjected to tensile loading, while the b-PP phase deforms plastically up to high deformations (28). The interlocked structure of the a-spherulites makes the plastic glide of this phase very difficult. Upon loading, cavitation tends to occur at the aspherulite boundaries or at their equatorial region perpendicular to the tensile axis at the early stage of deformation. This can lead to the formation of cracks at the spherulitic boundaries. Some of the researchers attributed the superior tensile ductility of b-PP to the occurrence of b ! a phase transition during drawing (24–27). However, the mechanisms associated with this transition remain unclear. Therefore, understanding the structure–property relationship of the b-form PP enables us to elucidate the mechanisms responsible for the deformation and failure of this material under conditions of static tensile and dynamic impact loadings. As mentioned above, iPP exhibits a relatively low impact resistance, particularly at low temperatures or high speeds. To improve impact toughness of iPP, it is common practice to incorporate impact modifiers or elastomer particles such as ethylene–propylene copolymer (EPR), ethylene–propylene–diene copolymer (EPDM), and poly(styrene-b-ethylene-co-butylene-b-styrene) triblock copolymer (SEBS) (29,30). Among these, EPR is often introduced by varying the monomer composition during synthesis. The elastomers tend to trigger shear yielding of iPP matrix associated with cavitation of the rubber particles during deformation. Similarly, impact modifiers are also added to b-PP to further improve its impact performances (31,32).

11.2

DEFORMATION CHARACTERISTICS

It is generally known that the mechanical properties of semicrystalline polymers depends greatly on several intrinsic parameters such as crystalline structure, molecular weight and spherulite size, and extrinsic factors such as strain rate, temperature, and processing conditions. The effect of the crystalline modification on the mechanical properties of PP has attracted considerable attention of researchers in the past decade (15,18,25–28,33–36). The a-modification of PP exhibits high yield strength and stiffness as a result of its interlock structure with crosshatched lamellae. The b-PP having a sheaflike structure exhibits lower yield strength but higher tensile ductility as well as more fracture resistance than does the a-PP. The tensile drawing behavior of b-PP is of great technological importance. For example, formation of the

310

Polyolefin Blends

microvoids in b-PP during tensile drawing has led to the development of tough microporous films that find applications as materials for printing and gas exchange membrane purposes (37–39). Further development of such promising materials relies on a better understanding of the fundamentals of their deformation characteristics. The stiffness, yield strength, and elongation of break of the b-PP can be determined from the tensile tests at various crosshead speeds. It is recognized that the yield strain and elongation of break are not good indicators of the toughness of polymers. Large ductility under plane-stress condition can be achieved more readily in tensile specimens without notches. In some cases, polymers with high tensile ductility could exhibit poor impact performance. The impact tests are commonly used to characterize the fracture resistance of materials under dynamic loading. The impact behavior of polymers can be evaluated by Izod or Charpy pendulum impact (40) and by falling weight types of impact (41). The presence of a notch in the impact specimen results in the buildup of triaxial stress or plane-strain condition ahead of the notch tip. Conventional Izod and Charpy impact tests only measure the energy absorption of the materials in fracture subjected to impact of pendulum or dropweight-type strikers. These measurements provide simple impact ranking for different polymers under limited test conditions but do not provide insights into their fracture behavior. Instrumental impact tests are more informative regarding the fracture behavior of polymers because the impact events are displayed as the force–time or force–displacement response curves. Thus, the crack initiation and propagation energies can be readily obtained. Brittle failure of polymers is characterized by a triangular shape of the force–displacement curve with little or no propagation energy. The fracture surface is relatively smooth or mirror-like. Ductile failure is characterized by an inelastic deformation in which stable crack propagation takes place immediately after yielding. Stress whitening occurs readily as a result of multiple crazing or shear yielding.

11.2.1 Static Tensile Behavior Tjong et al. studied the tensile behavior of injection-molded a-PP and b-PP homopolymers at crosshead speeds ranging from 1 to 500 mm min1 (15). The b-nucleator used is a mixture of equal amounts of pimelic acid and calcium stearate. Tensile measurements indicate that the elongation of pure PP is enhanced greatly by adding 0.1 wt% b-nucleator. Stress whitening occurs readily during postyielding and necking stages. Figure 11.3 shows the typical stress–strain curves of the a-form and b-form PP specimens at a strain rate of 0.15 min1. As the strain at break of b-PP exceeds more than 6.6, this figure only shows the tensile behavior of b-PP up to the postyielding stage. The cold-drawn behavior of b-PP is distinct from that of the a-PP. For the a-PP, the stress drops suddenly and markedly after reaching the yielding point. However, the stress drop is considerably slower for the b-PP homopolymer after yielding. Large-scale plastic deformation of the b-phase crystals is expected during cold drawing. The unique structure of b-PP having many intermingled lamellae enables microdrawing of the b-PP proceeds mainly via interlamellar

Chapter 11 Deformation Behavior of b-Crystalline Phase

311

Figure 11.3 Stress–strain curves for the a-PP and b-PP specimens at a strain rate of 0.15 min1. (From Reference 15 with permission from The Society of Plastics Engineers.)

slippage and chain slip (42,43). However, some researchers considered that the b ! a phase transformation is responsible for the strain hardening observed in b-PP during tensile drawing (24,26–28). Figure 11.4 shows the variation of yield stress with strain rate for the a-PP and bPP specimens. Compared to a-PP, the incorporation of b-nucleator brings about distinct softening as evidenced by lower values of yield stress at various strain rates. The strain rate dependence of yield stress can be described by Erying equation given by s y DH R 2_e ¼ ð11:3Þ þ ln T VT V e0 where DH is the activation energy of the plastic flow, V is the activation volume of the element motion unit, e_ is the strain rate, and e0 is the pre-exponential factor. Thus, the

Figure 11.4 Strain rate dependence of yield stress for a-PP and b-PP specimens. (From Reference 15 with permission from The Society of Plastics Engineers.)

312

Polyolefin Blends

process involved in yielding is considered as an activated rate process. The activation volume can be determined from the slope of the plot of s y versus ln(_e), that is slope ¼ ðRTÞ=V. Li and Cheung found that the slope of the plot of s y versus ln(_e) for b-PP at high strain rates differs considerably from that at low strain rates (26). Thus, the yield process for b-PP is described by a two-stage Erying equation. They determined the activation volume of b-PP to be 3 nm3 at low strain rates regime. It is believed that the plastic flow is controlled by the cooperative motion of chain segments. Grein (32) also found a two-stage Erying behavior or a bilinear relation in the plot of s y versus ln(_e) for the b-PP/EPR and iPP/EPR blends (Fig. 11.5). The yield stress values of the b-PP/EPR

Figure 11.5 Strain rate dependence of yield stress for (a) a-PP and b-PP homopolymers, and (b) 15% toughened iPP/EPR and b-nucleated iPP/EPR blends at room temperature. (From Reference 31 with permission from Elsevier.)

Chapter 11 Deformation Behavior of b-Crystalline Phase

313

Figure 11.6 Dependence of yield stress and Young’s modulus on calcium carbonate content. (From Reference 34 with permission from The Society of Plastics Engineers.)

blend at various strain rates are considerably lower than those of the iPP/EPR, indicating higher ability to initiate plastic flow at lower stresses in the b-PP/EPR blend. As b-PP exhibits higher ductility than a-PP, increasing interest in the b-PP composites has occurred recently because tough b-PP is considered to be an ideal matrix for the polymer composites. In designing polymer composites with enhanced mechanical properties, several parameters such as strength, stiffness, and toughness must be taken into account. The incorporation of inorganic fillers into thermoplastics generally leads to a large reduction in their tensile ductility and impact toughness. In the case of b-PP composites, a balance between the stiffness and toughness can be achieved via proper control of the filler content. Figure 11.6 shows the effect of untreated calcium carbonate filler (2.7 mm) content on the yield strength and stiffness of b-PP composites. The Young’s modulus of b-PP composites tends to increase, whilst the yield strength appears to decrease with increase in the filler content. A similar behavior is observed for the a-PP composites. A decrease in yield strength is caused by the pullout of filler particles from the polymeric matrix due to a poor filler– matrix interaction. Such an interaction can be enhanced by treating the filler particles with appropriate surface-modifying agents. The coupling agents can also improve the dispersion of filler particles in the polymeric matrix. More recently, Kotek at al. studied the effects of untreated (denoted as C1) and stearate-treated calcium carbonate particles (denoted as C2) of 1.3 mm on the tensile properties of b-PP. Moreover, stearate-coated calcium carbonate particles of 0.075 mm (denoted as C3) are also used (44). The b-nucleating agent employed is based on an amide of dicarboxylic acid. The critical concentration of b-nucleator needed to achieve 61% b-PP is 0.03 wt% on the basis of the XRD analysis. With further increase in the b-nucleator content to 0.1 wt% (supercritical), the b-PP content increases, but very slightly, to 65%. Furthermore, the K value also depends

314

Polyolefin Blends Critically nucleated PP

Supercritically nucleated PP

2400

2410

2200

2200

2210

2000 1800 1600

C1 filler C2filler C3 filler

1400 1200 0

10 20 Filler content, wt%

30

1800 1600

1200 1000

1610

1210 1010

30

25

0

10

20

C3 filler 30

25

10 20 Filler content, wt%

300 200 100 0

300 200 100 0

0

10

20

Filler content, wt%

30

25

0

10 20 Filler content, wt%

30

500 C1 filler C2 filler C3 filler

400 Strain at break, %

C1 filler C2 filler C3 filler

C1 filler C2 filler C3 filler

30

30

500

400

30

20 0

500

10 20 Filler content, wt%

35

20

30

0

40

35 Yield stress, MPa

30

C1 filler C2 filler C3 filler

1410

C1 filler C2 filler

Filler content, wt%

Strain at break, %

10 20 Filler content, wt%

1810

40 C1 filler C2 filler C3 filler

35 Yield stress, MPa

0

2010

Yield stress, MPa

40

20

C1 filler C2 filler C3 filler

1400

0

10

20

Filler content, wt%

30

C1 filler C2 filler C3 filler

400 Strain at break, %

1000

2000

Young modulus, MPa

2400 Young modulus, MPa

Young modulus, MPa

Non nucleated PP

300 200 100 0 0

10 20 Filler content, wt%

30

Figure 11.7 The effects of untreated and surface treated filler loading on the tensile behavior of injection-molded PP specimens. (From Reference 44 with permission from Elsevier.)

on the types of filler used. A stearatecoated filler with very fine size (C3) exhibits the highest b-nucleation activity. Figure 11.7 summarizes the effects of filler content on the tensile behavior of nonnucleated PP (a-modification), critically and supercritically nucleated b-PP. The stiffness of nonnucleated PP, critically nucleated PP, and supercritically nucleated PP composites generally increases dramatically with increasing filler content (C1, C2, or C3) as expected. For critically nucleated PP, the strain at break of b-PP composites decreases slightly with the addition of surfacetreated fillers (C2 and C3) up to 20 wt%. However, the strain at break decreases linearly with untreated C1 filler content.

11.2.2 Strain-Induced b ! a Phase Transition In general, the b-PP is mechanically stable up to yield point but transforms to a-phase during necking (24,26,27,44–47). The strain-induced b ! a phase

Chapter 11 Deformation Behavior of b-Crystalline Phase

315

transformation of b-PP still remains controversial. The mechanism by which the transformation takes place is not fully understood. Some researchers proposed localized melting and recrystallization mechanism (26,27), while others suggested low temperature solid state transformation mechanism (48). Nevertheless, the straininduced b ! a phase transformation enhances the ductility of b-PP. Karger-Kocsis and coworkers reported that the phase transition from a less densely packed b-PP toward a denser a-PP crystalline could lead to a volume contraction and enhanced toughness of b-PP (24,25,45–47). This behavior is somewhat similar to the so-called phase transformation toughening (PTT), commonly observed in stabilized zirconia ceramics. However, it is accompanied by a volume expansion in zirconia (24,25). The volume contraction associated with the b ! a phase transition in b-PP promotes microvoiding and strain hardening due to its exothermic character. In the sheaf-like structure of b-PP, the lamellae bundles are held together by the tie molecules which can easily detach from one another on loading. The lamellae separation is accompanied by massive voiding and crazing. The lamellae involved in crazing deform and breakup by slippages. Moreover, those lamellae that are aligned along the loading direction may defold. The breakup and defolding of the lamellae triggers the b ! a phase transition in b-PP (45–47). Li et al. conducted XRD and differential scanning calorimetry (DSC) measurements on cold-drawn b-PP specimens to verify the occurrence of b ! a phase transition (26). Figure 11.8a–c shows the XRD patterns of undeformed b-PP and the b-PP yielded at 8% and 110% strain, respectively. The K value of undeformed b-PP

Figure 11.8 XRD patterns of injection-molded b-PP. (a) Before deformation, (b) yielded at 8% strain but without obvious necking, and (c) after cold drawing at 110% strain. (From Reference 26 with permission from Elsevier.)

316

Polyolefin Blends

Figure 11.9 DSC traces of undeformed and deformed material taken from different sections of the neck of injection-molded b-PP. (a) Undeformed material; (b) Yielded material with 8% strain; (c) Upper shoulder of neck with 22% strain; (d) lower shoulder of neck with 55% strain; (e) 65% strain; (f) 110% strain; (g) Cold-drawn section with 160% strain; and (h) 230% strain. (From Reference 26 with permission from Elsevier.)

is determined to be 0.73. The XRD pattern of b-PP deformed at 8% exhibits a similar feature as that of undeformed specimen, that is, appearance of a prominent b-phase peak at 16 . The K value of this deformed b-PP remains at 0.73. With further cold drawn to 110 %, the intensity of prominent b-phase peak at 16 decreases significantly, while the intensity of a-PP phase diffracting peaks at 14.1, 16.8, and 18.8 increases. In this case, the K value of cold-drawn b-PP reduces to 0.21, indicating some b-phase crystals have transformed into the a-PP phase. A similar behavior is observed in the DSC results. Figure 11.9 shows the DSC traces of the materials removed from undeformed and deformed necking regions of b-PP after tensile test. The undeformed material yields a b-fusion and a-fusion peaks at 152.5 and 166.9 C, respectively. The DSC traces show that the b-fusion peak intensity decreases with increase in drawing extent along the gage length of tensile specimen. The results are listed in Table 11.1. From these, they proposed that the b-PP crystals are transformed to a-PP phase during cold drawing due to the localized melting and crystallization (26). To get further insight into the deformation mechanisms of b-PP during drawing, Li et al. employed electron microscopy (scanning electron microscopy, SEM and transmission electron microscopy, TEM) to observe the lamellar structural changes

Chapter 11 Deformation Behavior of b-Crystalline Phase

317

Table 11.1 DSC Parameters of a- and b-Phase Crystals Over the Neck of b-PP. [From Reference 26 with permission from Elsevier]. baStarting Melting Melting Strain, fusion point, point, DHb ; Sample % C C C J g1 A B C D E F G H

0 8 22 55 65 110 160 230

106.0 105.5 106.3 113.0 114.0 119.0 120.5 121.0

152.5 152.5 152.3 151.2 150.9 148.8 148.0 147.0

166.9 166.7 166.6 166.5 166.8 166.8 167.2 167.2

61 61 56 36 30 13 9 2

DHa ; J g1 41 41 46 66 70 83 86 88

Xb , % Xa , % 36 36 33 21 18 8 5 1

23 23 26 37 39 46 48 50

X, %

fb , %

59 59 59 58 57 54 53 51

61 61 56 36 31 14 9 2

caused by tensile load (27). Because of a strong effect of the injection molding condition on the b-phase content, compression molding technique was adopted to prepare the tensile test specimens. The deformation of b-PP is observed to be highly inhomogeneous. In the early stage of deformation, horizontal lamellae were stretched to deformation, leading to the formation of deformation bands (Figure 11.10). On further deformation, these bands develop into crazes and cracks. Intralamella slipping takes place in the regions where the lamellae are aligned with the loading direction (Fig. 11.11). Moreover, the lamellae could be stretched to distortion and disintegration, leading to the formation of melting spots as marked

Figure 11.10 SEM micrograph of compression-molded b-PP sample at 3.6% strain and under a crosshead speed of 5 mm min1. In area A the lamellae are roughly along the loading direction (marked by the arrow). Many fine slits can be seen along horizontal lamellae near the lower left-hand side of micrograph. The arrow indicates the loading direction. (From Reference 27 with permission from Elsevier.)

318

Polyolefin Blends

Figure 11.11 TEM micrograph of yielded b-PP sample at 5.0% strain. (a) shear band caused by intralamella slipping and interlamella shear; (b) shear band primarily caused by interlamella shear; and (c) and (d) melt spots. The arrow indicates the loading direction. (From Reference 27 with permission from Elsevier.)

by C and D. Shear bands would develop in the vertical lamellae through combined interlamella shear and intralamella slipping as marked by A, or through interlamella shear as mark by B in Fig. 11.11. The shear band at B can be termed more precisely as a ‘‘kink’’ band. From these, they concluded that melting, mixed shearing, and crazing take place concurrently in b-PP during tensile drawing. However, the main cause of failure for the deformed sample is crazing (27). Coulon et al. reported that the plasticity of PP after yielding mainly results from the slip of crystalline phase. The major plastic deformation mechanism is the intralamellar slip, which is governed by the resistance to the motion of dislocations in the slip planes (49). The mechanism for formation of the melt spots as observed in the TEM micrograph during tensile drawing of b-PP remains unclear. This mechanism requires a conformation change of the chains from left-handed helices in b-PP to alternate helices in a-PP. Would localized heating during tensile drawing of b-PP at room temperature be high enough to cause the melting and disintegration of original

Chapter 11 Deformation Behavior of b-Crystalline Phase

319

b-phase crystals? The b-PP is known to undergo the b ! a phase transition during slow heating in the temperature range of 140–148 C. During heating process, the b-form first melts and then reorganizes into stable a-form PP through a phase transformation (50,51). Asano et al. investigated the plastic deformation of b-PP via hot rolling from 68 to 142 C (42,43). They reported that the b ! a phase transition occurs during hot rolling. The original b-phase crystal was destroyed through unfolding or melting and then recrystallized into a-phase. The amount of a-phase transformed tended to increase with increasing rolling temperature and draw ratio. It can be argued that high rolling temperatures employed would assist the melting of b-phase PP and its subsequent recrystallization to a new a-phase PP. From the infrared thermography, a low temperature rise of only 4 C in the vicinity of notch tip of b-PP is recorded during tensile deformation (24). Labour et al. pointed that the adiabatic heating resulting from the conversion of plastic work into thermal energy could lead to temperature increase in the plastic zone ahead of b-PP. Consequently, plastic drawing of the material into fibrils and crazes is more likely to occur if the b ! a phase transition does take place (52). From these, it is unlikely to yield a large temperature rise close to the melting during drawing b-PP at a slow tensile strain rate. More recently, Xu et al. indicated that the b ! a transition in b-PP could take place via a solid-state transformation rather than localized melting and recrystallization mechanism (48). Solid-state transformation involves the slips along the b-(110) and b-(120) planes during shear of the crystal lattice. This is somewhat similar to the so-called martensitic transformation in metals and other thermoplastics. Martensitic transformation is well recognized in metal community. The shear deformation of face-centered cubic austenitic lattice yields a diffusionless transformation of austenite into tetragonal martensite phase. For the polymers, martensitic-like solid-state transformation has been used to describe the transformation from orthorhombic to monoclinic polyethylene in which the molecular chains of the initial and final crystalline phases exhibit identical conformations (53). In other words, preservation of helical hands must be maintained during the transformation process. Reversal of helical hand is an unlikely event for most thermoplastics (54). However, molecular chains of the b-PP crystals have the same helical hand as those of the a-PP crystals packed in the form of right- or left handed. In this aspect, the reversal of helical hand is needed for solid-state transformation to take place at low temperature. Xu et al. introduced molecular chains with 120 helical jump defects to accommodate the conformation changes during molecular simulations of the b-PP lattice (48). The helical hand reversal achieved via conformation defects during the phase transformation has been found in polytetrafluoroethylene (PTFE) (55). PTFE exhibits phase II triclinic unit comprising 54/25 helical chains with opposite hands below 19 C at atmospheric pressure, and Phase IV hexagonal unit cell comprising 15/7 helical chains with identical helical hand. Phase transformation from Phase II to Phase IVoccurs when the temperature is increased between 19 and 30 C. Xu et al. performed uniaxial compression rather than the tensile tests to investigate the b ! a transition in b-PP (44). Figure 11.12a and b shows the typical true

320

Polyolefin Blends

Figure 11.12 True stress versus strain of (a) a-PP and (b) b-PP at various temperatures under a compressive strain rate of 0.01 s1. (From Reference 48 with permission from Elsevier.)

stress versus true strain curves of the a-PP and b-PP at various test temperatures for a compression strain larger than 1.5. As temperatures increase, both the modifications show decrease in modulus and yield stress, which is behavior typical of semicrystalline polymers. It is evident that b-PP shows larger strain hardening than a-PP at all test temperatures, with the exception at 135 C. Moreover, strain hardening rate of both the modifications decreases with increasing temperature. The strain hardening of semicrystalline polymer generally involves crystallographic slip mechanism and crystallographic texturing in addition to amorphous chain orientation. The evolution of microstructure with plastic deformation in b-PP during compression is examined by DSC and XRD techniques. Figure 11.13a and b shows the DSC traces of a-PP and b-PP specimens after deformation to various

Chapter 11 Deformation Behavior of b-Crystalline Phase

321

Figure 11.13 DSC traces of (a) a-PP and (b) b-PP at various compressive strains under a strain rate of 0.01 s1and at room temperature (25 C). (From Reference 48 with permission from Elsevier.)

compressive strains at room temperature. As a-PP is plastically deformed, the area under the endotherm decreases, indicating a reduction in the degree of crystallinity. From Fig. 11.13b, the endotherm peak at 165 C represents melting of the a-phase crystals. Undeformed b-PP shows a strong endotherm peak at 155 C together with a small endotherm at 165 C. The DSC traces clearly indicate that the intensity of endotherm b-PP decreases, while the a-peak increases with increasing compressive strains. This implies a continuous transformation of b-PP to a-PP with increasing inelastic deformation. The effect of temperatures on the b ! a phase transformation of b-PP subjected to a compressive strain of 1.6 is depicted in Fig. 11.14. For the purpose of comparison, the DSC traces of aPP deformed to a true strain of 1.6 at various temperatures is shown in this figure. From Fig. 11.14b, the b-PP melting peak intensity decreases, whilst the a-PP peak increases as the deformation temperature is increased. This implies that high deformation temperatures favor the b ! a phase transformation. The formation of a-PP from b-PP after deformation under various compressive strains at room temperature or under a high compressive strain at various temperatures is verified by the XRD method. The XRD patterns generally show the disappearance of b-PP (300) peak at 16 and the increase of a-PP (110) peak intensity at 14 with increasing strain at room temperature. Therefore, the induced a-PP crystals

322

Polyolefin Blends

Figure 11.14 DSC traces of (a) a-PP and (b) b-PP after deformation to true strain of 1.60 under a strain rate of 0.01 s1 at various temperatures. (From Reference 48 with permission from Elsevier.)

have their (111) planes oriented normal to the compression direction during the transformation at room temperature. At higher temperatures, the a-PP (040) peak intensity at 17 increases and predominates. This indicates strong alignment of (040) plane of a-PP with the plastic flow direction. Xu et al. proposed that the b ! a transition can take place in b-PP at low temperatures via solid-state transformation mechanism and without local heating and crystallization. As mentioned above, solid-state transformation from the b-PP to a-PP molecular lattice is feasible by inserting two 120 helical jumps defects near the central position of b-PP molecular chains (Fig. 11.15). The first defect changes the helical hand from left-handed to right-handed and the second reverses it back to left-handed. This results in the formation of alternate helical lattice of a-PP. At present, there is no conclusive experimental evidence for reversal change of the helical chains in b-PP. Keeping in mind that a small rise in temperature in b-PP due to deformation induced adiabatic heating favors low temperature solid-state transformation rather than melting and crystallization. The microkink band resulting from shear deformation of the b-PP lamella as observed in TEM micrographs shed light on the solid-state transformation (Fig. 11.11). The kink band is considered to be associated with the formation of alternate helical chains with the insertion of

Chapter 11 Deformation Behavior of b-Crystalline Phase

323

Figure 11.15 Molecular chain simulation of (a) perfect 31 PP helix and (b) 31 PP helix with two 120 helical jumps at the central position; the helical hand changes at every helical jump; propagation of the defect along the chain will reverse the helical hand. (From Reference 48 with permission from Elsevier.)

defects (55). The XRD and DSC techniques as well as strain hardening provide important clues on the b ! a phase transition in b-PP during tensile drawing and compression, further in-depth studies are needed to elucidate the mechanisms responsible for such phase transformation.

11.2.3 Impact Behavior Figure 11.16 shows the falling weight Charpy impact strength for a and b modifications as a function of testing speed. Interestingly, high impact energy

324

Polyolefin Blends

Figure 11.16 Falling weight Charpy impact strength of the a-PP and b-PP specimens impacted at different speeds. (From Reference 18 with permission from Elsevier.)

values are obtained in b-PP, particularly at a low impact speed of 0.76 ms1. The smaller sizes of b-PP spherulites also contribute to the toughness enhancement. Both a-PP and b-PP specimens fracture in a macroscopically brittle mode under impact loading, as indicated by the absence of crack propagation energy in the load–time curves. However, SEM can provide further information relating to the localized plastic deformation ahead of the notch tip of both PP modifications. For the a-PP, the fracture initiation zone (plastic zone) impacted at 0.76 m s1 is considerably larger than that developed at a speed of 4.8 m s1. Figure 11.17a shows the SEM micrograph of a-PP impacted at a speed of 0.76 m s1. These specimens experience little plastic deformation, as evidenced by the smooth surface and the presence of few microvoids. In contrast, high density of microfibrils and microvoids is observed in the plastic zone of b-PP (Fig. 11.17b). The density and length of microfibrils decrease dramatically with increase in the impact speed to 4.8 m s1. The fibrils are obtained by microdrawing the matrix and the tie molecules between the b-spherulites. The tie molecules tend to align to tie fibrils during the microdrawing process. It is believed that an increase in the impact strength of b-PP is associated with the formation of fibrillated zone ahead of the notch tip. The DSC measurements were performed on b-PP specimens subjected to various impact speeds. There is no evidence of the b ! a transition because the impact test time is relatively short for the phase transformation to occur. Karger-Kocsis et al. investigated the effect of melt flow indices (MFI) or molecular weight on the impact strength of the injection-molded a-PP and b-PP specimens via instrumented tensile impact (ITI) and instrumented falling weight techniques (46). The b-phase was induced by adding 0.1 wt% calcium salt of pimelic acid to isotactic PP with varying MFI. The b-phase content was found to decrease with increasing molecular weight (or decreasing MFI) due to the development of thicker skin layer composed of the a-modification. ITI tests were

Chapter 11 Deformation Behavior of b-Crystalline Phase

325

Figure 11.17 SEM micrographs showing the morphology of fracture initiation zone of (a) a-PP and (b) b-PP impacted at a speed of 0.76 m s1. (From Reference 18 with permission from Elsevier.)

performed on the dumbbell specimens at room temperature under a speed of 3.7 ms1. The Young’s modulus, elongation at break, and tensile impact energy were determined accordingly. Figure 11.18a–c shows the variations of Young’s modulus, elongation at break, and tensile impact energy versus MFI for the injection-molded a-PP and b-PP specimens. It can be seen that the stiffness of a-PP is higher than that of b-PP, particularly at low MFI. The elongation at break and tensile impact energy of a-PP are lower than those of b-PP, as expected. These results demonstrate that the MFI or molecular weight plays a key role in the toughness enhancement of b-PP in addition to the size of b-spherulites. KargerKocsis et al. attributed the outstanding impact fracture resistance of b-PP to the formation of fibrillar structure owing to a dense intraspherulitic and interspherulitic tie molecule density. From Fig. 11.18, it can be seen that the injection molding speed has little effect on the tensile properties of b-PP. By contrast, the molding speed has a profound effect on the stiffness, elongation at break, and impact energy of a-PP, particularly at low MFI. Increasing the injection molding speed generally results in the formation of a thinner skin layer.

Figure 11.18 (a) Young’s modulus, (b) elongation at break, and (c) tensile impact energy as derived from instrumented tensile impact tests for the a-PP and b-PP specimens molded at low and high injection speeds (vinj). Increasing vinj results in a thinner skin layer. (From Reference 46 with permission from John Wiley & Sons.)

326

Chapter 11 Deformation Behavior of b-Crystalline Phase

327

Grein et al. studied the effect of the molecular weight of dispersed phase of EPR on the impact behavior of iPP/EPR and b-nucleated iPP/EPR reactor blends containing 23 wt% impact modifier (32,56). They varied the intrinsic viscosity (IV) of EPR phase while keeping the matrix melt flow rate constant. They reported that the Charpy impact strength of the iPP/EPR blend increases with increase in the IV for the rubbery phase. This behavior is also observed for the b-PP/EPR blend (Fig. 11.19a). They found a large difference (delta) between the notched impact strength of b-PP/EPR and iPP/EPR blends at given IV, D(NIS), defined as DðNISÞ ¼

½NIS ðb-nucleatedÞ NIS ðnonnucleatedÞ NIS ðnonnucleatedÞ

ð11:4Þ

Figure 11.19 (a) Notched impact strength (NIS) at room temperature of a nonnucleated and a b-nucleated PP/EPR blends versus the IV of the rubber phase; (b) particle size (Dw) of the dispersed phase of the blends vs. the delta in NIS, at given IV, between the b-modified grade and its nonnucleated counterpart. (From Reference 32 with permission from Springer Science and Business Media.)

328

Polyolefin Blends

Figure 11.20 Instrumented load–time traces of the unnotched dart-impact test for two unfilled samples with different b-phase contents. (From Reference 52 with permission from John Wiley & Sons).

They correlated this difference with the size of the EPR particles in the systems as shown in Fig. 11.19b. This figure reveals that the smaller the diameter of dispersed EPR particles, the higher the beneficial effect of b-PP on improving the impact strength. Labour et al. (52) studied the effect of b-phase content on the impact crack initiation and propagation behavior of unfilled and stearate-coated CaCO3 (0.1 mm) filled PP using an instrumented dart impact test at a speed of 1 m s1. The unnotched rectangular bars having dimensions of 50 mm 10 mm 4 mm for impact test were cut from the compression-molded sheets. A special heat-treatment procedure was performed to induce various amounts of b-phase. A maximum value of 40% can be achieved in filled PP samples owing to the nucleating effect of stearate agent in contrast with about 10% for unfilled PP samples. Figure 11.20 shows the typical load–time traces for unfilled PP samples with different b-phase contents. Increasing the b-phase content in unfilled PP entends the plastic deformation capability and improves the resistance to crack propagation. The filled samples fracture before reaching the force threshold for crack propagation. From the force–time traces, the work for crack initiation (Wint ), the elastic component (Wint-e), and the plastic component (Wint-p) of the work for crack initiation for the unnotched PP and filled PP samples can be determined. Figures 11.21a–c illustrates Wint, Wint-e, and Wint-p versus the b-phase content for the unnotched PP and filled PP samples, respectively. Apparently, unfilled PP specimens display a linear increase of Wint with the b-phase content up to 10%. This is due to the greater plasticity of b-PP as compared to the a-PP. From the SEM observation of fractured samples, the plastic zone size of unfilled PP samples increases with increase in the b-phase content. By contrast, the Wint values of the filled PP show little changes over a large b-phase content range. From Fig. 11.21b, it is evident that Wint-e is insensitive to the b-phase content. However, Wint-p for unfilled PP displays a steady increase with the b-phase content. The Wint-p of filled PP samples remains nearly unchanged with

Figure 11.21 (a) Work for crack initiation, (b) elastic, and (c) plastic components of the work for crack initiation under impact test for filled and unfilled PP as a function of b-phase content. (From Reference 52 with permission from John Wiley & Sons.)

329

330

Polyolefin Blends

the b-phase content. This implies that the filled PP samples exhibit poor crack propagation resistance under impact, irrespective of the amount of b-phase in samples. These findings are in good agreement with those of Tjong et al. in which the addition of only 3% CaCO3 particles leads to a large drop in the impact strength of b-PP. Thereafter, the impact strength reduces slightly with further increase in the filler content. It is believed that the CaCO3 particles constraint the plastic deformation of b-PP matrix and promote easy crack propagation via the weak particle–matrix interface (34).

11.3 FRACTURE TOUGHNESS 11.3.1 General Aspects Toughness is an important parameter for materials design and engineering practice. Catastrophic failures would occur in engineering components during practical services when the critical crack has propagated in unstable mode. Reliable service performance of the polymers depends greatly on their tensile ductility and impact toughness. Conventional Izod and Charpy impact tests are commonly used to determine the impact fracture behavior of polymers due to their simplicity, speed, and cost effectiveness. These tests involve the determination of energy required for breaking a notched specimen with a specific notch geometry and a given fracture area. The results simply denote the energy absorption in the notched samples. Thus, they are inappropriate to describe the toughness of polymers, particularly at high speeds. In this regard, fracture mechanics are considered to be an effective tool to characterize the fracture toughness of polymers under tensile or impact-loading conditions. It allows the design and selection of materials that are resistant to fracture for industrial applications. Considerable effort has been devoted to the determination of fracture toughness for various thermoplastics as a function of strain rate or temperature. Linear elastic fracture mechanics (LEFM) has been used successfully for characterization of the toughness of brittle materials. The driving force of the crack advance is described by the parameters such as the stress intensity factor (K) and the strain energy release rate (G). Unstable crack propagates when the energy stored in the sample is larger than the work required for creation of two fracture surfaces. Thus, fracture occurs when the strain energy release rate exceeds the critical value. Mathematically, it can be written as KC2 ¼ E0 GC ¼ 2E0 g

ð11:5Þ

where g is the surface energy and E0 is the Young’s modulus under plane-stress condition. In plane strain, E0 ¼ E=ð1 n2 ), where n is the Poisson’s ratio. In some cases, the material fails macroscopically in a brittle mode, but smallscale plastic deformation or yielding takes place ahead of the crack tip as a consequence of stress concentration. The LEFM is valid provided that a term 2g p , that is, the plastic work required to extend the crack, is incorporated into

Chapter 11 Deformation Behavior of b-Crystalline Phase

331

Equation 11.5 (57,58). In this case, a circular plastic zone of a diameter 2rp is formed ahead of crack tip. The plastic zone radius rp is given by 1 KI 2 ð11:6Þ rp ¼ 2p s y where KI is the stress intensity factor under mode I. Taking the radius of plastic zone into consideration, the crack length, a, is corrected into effective crack length, aeff, accordingly: aeff ¼ a þ rp

ð11:7Þ

For the experimental determination of LEFM parameters, compact tension (CT) and single edge notched (SEN) bending specimens are commonly used. According to the European Structural Integrity Society (ESIS) test protocol (59) and standard practices (60,61), KI under mode I can be determined from the load–displacement curves according to the relationship: Fmax p ﬃﬃﬃﬃ ﬃ ð11:8Þ KIQ ¼ f ða=WÞ B W where a is the crack length, W and B are the width and thickness of the specimen, respectively, and f ða=WÞ is a geometric factor that depends on the ratio a=W. The validity of LEFM based on Equation 11.8 is established when a sharp crack of adequate size exists in a specimen. The strain energy release can be evaluated from the following equation: U ¼ GI BWF

ð11:9Þ

where U is the energy or area of the force–displacement curve in which maximum load (Fmax) is attained, and Fða=WÞ is a calibration factor depending on sample geometry. KIQ and GIQ approach their critical values when the LEFM validity criteria are satisfied.

11.3.2 Mode I LEFM Approach The mechanical behavior of polymers is well recognized to be rate dependent. Transitions from ductile to brittle mode can be induced by increasing the test speed. The isotactic PP homopolymer with high molecular weight is ductile at low speed tensile tests. It is brittle at tension under high test speeds at room temperature. Grein et al. (62) determined the variation of KIQ with test speed for the a-PP CT samples (Fig. 11.22). The force–displacement (F–d) curves and the schematic diagrams of the fracture surfaces of CT samples are presented in Fig. 11.23. At a very low test speed of 1 mm s1, the F–d curve exhibits a typical ductile behavior as expected. At 10 mm s1, the F–d curve still displays some nonlinearity before the load reaches its maximum value, but this is substantially suppressed as test speeds increase further. The samples fail in brittle mode at test speeds 500 mm s1. From Fig. 11.22, the KIQ values maintain at 3.2 MPa m1/2 at test velocities from 1 to

332

Polyolefin Blends

Figure 11.22 Variation of the stress intensity factor KIQ of isotactic PP (a-form) as a function of test velocity; mode of crack propagation as indicated. (From Reference 62 with permission from Elsevier.)

Figure 11.23 Schematic representation of the fracture surfaces of isotactic PP CT specimens as a function of test velocity along with the corresponding force–displacement curves. (From Reference 62 with permission from Elsevier.)

Chapter 11 Deformation Behavior of b-Crystalline Phase

333

10 mm s1. It then decreases to 2.5 MPa m1/2 at 103 mm s1, and further reduces to 2.0 MPa m1/2 when the test speed reaches 5 103 mm s1 . During stable crack propagation at the lowest speeds (0.1–1 mm s1), the fracture surfaces of CT specimen next to the notch are characterized by stress whitening together with pronounced shear lips. Stress-whitening is resulted from multicrazing process that takes place during deformation. At 10 mm s1, three zones can be identified next to the notch, that is stress-whitened, shear lips, and smooth. In the range 70–100 mm s1, the fracture surfaces are rough, unwhitened and uniform in appearance over the entire length. The rough surface zone tends to decrease, whilst the smooth surface zone appears to extend as test speeds increase further. They concluded that highly dissipative shear process dominates at speeds 1 mm s1 and multiple crazing at intermediate speeds (50–103 mm s1). At speeds 2 mm s1, crack tip deformation is limited to a single crack tip craze zone (62). Grein et al. further investigated the fracture toughness of a-PP and b-PP at different test speeds and temperatures (31). Figures 11.24 and 11.25 show the force– displacement curves for a-PP and b-PP CT specimens at 25 C (T > Tg ) and at 5 C

Figure 11.24 Force–displacement curves of (a) a-PP and (b) b-PP compact tension specimens tested at various speeds at 25 C. (From Reference 31 with permission from Elsevier.)

334

Polyolefin Blends

Figure 11.25 Force–displacement curves of (a) a-PP and (b) b-PP compact tension specimens tested at various speeds at 5 C. (From Reference 31 with permission from Elsevier.)

(T < Tg ), respectively. It is apparent that the a-PP (Mw ¼ 248:1 kg mol1 ) shows brittle behavior tested at various speeds (Fig. 11.24(a)). In contrast, the a-PP in Fig. 11.23 displays ductile behavior at a test speed of 1 mm s1 (0.001 m s1) due to its higher molecular weight (Mw ¼ 455 kg mol1 ). It can be seen from Fig. 11.24(b) that the b-PP CT specimens exhibit ductile and plastic yielding tested from 0.0001 to 0.1 m s1 with the exception of 1 m s1. At 5 C, both the a-PP and b-PP CT specimens fail in a brittle mode (Fig. 11.25). In mode I compact tension tests, there is a considerable increase in fracture toughness of b-PP over that of a-PP in certain ranges of temperature and test speed. Figure 11.26(a)–(d) show the variations of KIQ with the crack-tip loading rate, dK/dt, for the a-PP and b-PP CT specimens tested at various temperatures. The dK/dt is defined as dK dF=dt ¼ f ða=WÞ pﬃﬃﬃﬃﬃ ð11:10Þ dt B W which accounts implicitly for variations in specimen stiffness. For T < Tg, both a-PP and b-PP show brittle behavior. For T > Tg, The KIQ values remain nearly constant up to about 200 MPa m1/2 s1 but decrease sharply at higher rates. This implies that b-PP is ductile up to 200 MPa m1/2 s1 (0.1 m s1) at 25 C and up to 800 MPa m1/2 s1 (0.4 m s1) at 60 C. The KIQ values of b-PP are higher than those of a-PP at

Chapter 11 Deformation Behavior of b-Crystalline Phase

335

Figure 11.26 Variation of KIQ with the crack-tip loading rate, dK/dt, for a-PP and b-PP CT specimens tested at (a) 30 C, (b) 5 C, (c) 25 C, and (d) 60 C. The upper scale gives an indication of the test speed. The arrows indicate the ductile–brittle transition in b-PP. (From Reference 31 with permission from Elsevier.)

25 and 60 C. Similarly, formation of b-form PP phase also leads to substantially higher GIQ values at 25 and 60 C. The higher fracture toughness of b-form PP is related to the formation of stress-whitened zone ahead of crack tip. TEM images reveal that extensive microcavitation in the form of dense zone of craze-like structures are developed within the stress-whitened zone. As mentioned above, bPP is much more prone to crazing than a-PP due to its lower yield stress that promotes microvoiding in the amorphous layer (15). The bundled b-lamellar structure linked by the tie molecules can easy detach from one another on loading. This lamellar detachment is accompanied by massive voiding with the simultaneous onset of a craze-like microporous structure (15,18,25–27). In the case of iPP/EPR and b-PP/EPR blends containing 15 vol% impact modifier, Grein et al. indicated that the b-nucleation had little effect on the brittle– ductile transition at 30, 5, 25, and 60 C on the basis of the results of KIQ versus dK/dt plots. Thus, elastomer particles play a major role during deformation of

336

Polyolefin Blends

rubber-modified a-PP or b-PP blends. These particles act as stress concentrators, causing cavitation and crazing of the matrix ahead of the crack tip. The elastomeric particles are therefore inferred to control the initiation and propagation of the plastic zone ahead of the crack tip during fracture of the rubber-modified blends. From the compact tension results and SEM observations, they further developed deformation maps to illustrate the deformed regions of the iPP/EPR and b-PP/EPR blends at different temperatures and crack-tip loading rates (Fig. 11.27). Little difference is

Figure 11.27 Deformation maps of (a) a nonnucleated PP/EPR with 15% EPR and (b) its b-nucleated counterpart for different temperatures and crack-tip loading rates as deduced from the fracture surfaces of compact tension specimens. A rough indication of the test speed is provided by the upper scale. () shearing, (&) shearing and crazing, (D) multiple crazing, and (X) single craze. (From Reference 32 with permission from Springer Science and Business Media.)

Chapter 11 Deformation Behavior of b-Crystalline Phase

337

observed in the deformation modes of these two rubber-modified blends. The presence of b-phase extends ductile failure of the b-PP/EPR blend slightly to higher test speeds. It is obvious that the b-phase is more effective in improving the fracture resistance of PP homopolymer than of the rubber-modified blends. As b-PP shows ductile behavior at test velocities 1 m s1, crack-tip plasticity must be taken into account in the determination of fracture toughness. LEFM concept can still be applied to b-PP under conditions of small-scale yielding ahead of crack tip. In this regard, radius of the plastic zone must be determined and incorporated into the existing crack length (a) in solving Equation 11.8. Figure 28a and b shows the typical load–displacement curves of the b-PP compact tension specimens having different crack lengths loaded under velocities of 0.001 and 3 m s1 at room temperature. The b-PP specimens exhibit ductile mode tested at 0.001 m s1 and display brittle behavior at a high velocity of 3 m s1. These figures can be expressed in terms of Fmax versus BW1/2/f(a/w) (Fig. 11.29). The regression line of brittle b-PP specimens tested at 3 m s1 passes through the origin as expected. The slope of the line yields KIC. By contrast, the ordinate of the regression line of ductile b-PP specimens tested at 0.001 m s1 has a negative value. This is because plastic yielding takes place ahead of crack tip, and a plastic zone correction is needed to obtain intrinsic K values. The radius of plastic zone, rp, can be determined by numerical iteration method in which the f(a/W) term of Equation 11.8 is replaced by f ða þ rp Þ=W) such that all data points fall on a line through the origin as shown in Fig. 11.30a. The radius of the plastic zone is determined to be about 2:22 0:45 mm. The effective toughness (Keff) can be determined from the slope of this corrected line, that is, 4.9 MPa m1/2 (Fig. 11.30b). Similar procedures are adopted for the determination of Gc in which the effective crack length is expressed in terms of a þ dp , where dp is the diameter of plastic zone (63). Figure 11.31 shows the variations of Keff and Geff values with temperature for the b-PP specimens tested at 0.001 m s1. The Geff values are very low at low temperatures (30 and 5 C) but increase significantly at room temperature and above. The Keff values of the b-PP specimens at below and above room temperature are slightly lower than that at room temperature, reflecting temperature dependence of the yield stress. The temperature and strain-rate dependence of the mechanical behavior of polymers is well known. It is considered that in many cases the effect of an increase in temperature is similar to the effect of a decrease in strain rate. For the b-nucleated iPP/EPR blend containing 15% impact modifier, Grein and coworkers determined the radius of plastic zone at 30, 5, 23, and 60 C under 0.001 m s1 using the corrected LEFM approach to be 2:07 0:28, 2:17 0:28, 2:59 0:25, and 1:78 0:45 mm, respectively (63). The corresponding fracture toughness values of the b-nucleated iPP/EPR blend at these temperatures are 7:76 0:25, 6:63 0:18, 4:75 0:30, and 2:48 0:11 MPa m1=2 , respectively. Compared to the fracture toughness of b-PP homopolymer at 30, 5, 23, and 60 C (Fig. 11.30), the beneficial effect of EPR addition to b-PP can be only found at low temperatures, that is, below 23 C. This implies that the impact modifier improves the fracture toughness of b-PP at low temperatures only.

338

Polyolefin Blends

Figure 11.28 Force–displacement curves of b-PP compact tension specimens having different crack lengths loaded under velocities of (a) 0.001 m s1 and (b) 3 m s1 at room temperature. (From Reference 63 with permission from Elsevier.)

Chapter 11 Deformation Behavior of b-Crystalline Phase

339

Figure 11.29 Fmax versus BW1/2/f(a/W) plots for b-PP compact tension specimens loaded under velocities of 0.001 and 3 m s1 at room temperature. (From Reference 63 with permission from Elsevier.)

11.3.3

IMPACT FRACTURE TOUGHNESS

Tjong et al. have determined the impact fracture toughness of b-PP using the falling weight Charpy impact tests on SEN specimens at room temperature under a test speed of 1.2 m s1 (64). The b-PP was nucleated from an agent consisting 0.1 wt% mixture of pimelic acid and calcium stearate (K value ¼ 0:94). Figure 11.32 shows the typical impact force–time traces of a-PP and b-PP specimens. Apparently, the a-PP and b-PP specimens fracture in brittle mode as evidenced by nearly triangular shape of load–unloading trace. In this case, Equation 11.9 can be used to determine the critical strain energy release rate at high impact speed of 1.2 m s1. The GIC values for the a-PP and b-PP specimens are determined to be 5.26 and 6.71 kJ m2, respectively. The higher impact fracture value of b-PP is attributed to the formation of fibrillated zone next to the sharp notch region on the basis of SEM observation (64). Recently, Karger-Kocsis and coworkers also determined the Charpy impact fracture toughness of the a-PP and b-PP specimens at various test temperatures under an impact speed of 1.2 m s1 (47). The b-nucleant is of quinacridone type and the amount of b-PP phase is 9.49%. Figure 11.33a and 33b shows the impact force– time traces of a-PP and b-PP specimens at room temperature and 40 C, respectively. The a-PP homopolymer is brittle at both test temperatures as expected. The impact force–time curve of b-PP shows ductile yielding behavior, particularly at room temperature, as evidenced by the occurrence of crack propagation stage after reaching maximum load. The resulting Kc and Gc values are listed in Table 11.2. The results clearly indicate that the b-PP exhibits higher impact fracture toughness than their a-PP at room temperature and at T < Tg . It is noted that the Gc value of a-PP at room temperature is comparable to that of Tjong et al. However, the Gc value of b-PP at room temperature is higher than that reported by Tjong et al. This is due to higher molecular weight of b-PP employed by Karger-Kocsis and coworkers

340

Polyolefin Blends

Figure 11.30 (a) Determination of radius of plastic zone and effective toughness for the b-PP specimens tested at 0.001 m s1. (b) Geometry independence of Keff values. (From Reference 63 with permission from Elsevier.)

Chapter 11 Deformation Behavior of b-Crystalline Phase

341

Figure 11.31 Variations of Keff and Geff values with temperature for the b-PP specimens tested at 0.001 m s1. (From Reference 63 with permission from Elsevier.)

(Mw ¼ 103 kg mol ). The larger Gc value is derived from the formation of smallscale yielding ahead of the propagating crack tip. The improved impact fracture toughness of b-PP is related to several factors such as molecular weight and tie molecules density, lamellar arrangement, and b ! a phase transition (47). However, there are no solid experimental evidences for the b ! a phase transition in b-PP subjected to impact measurements, possibly due to short testing time periods (18).

11.3.4 Essential Work of Fracture The fracture criterion of LFEM based on the Kc is invalid when extensive plastic deformation occurs ahead of the crack tip. For ductile polymers and their blends, the

Figure 11.32 Falling weight impact load–time traces of the a-PP and b-PP SEN specimens. (From Reference 64 with permission from Elsevier.)

342

Polyolefin Blends

Figure 11.33 Characteristic load–time traces due to impact of the notched Charpy specimens (a=W ¼ 0:5) for the a-PP and b-PP homopolymers at (a) room temperature and (b) 40 C. (From Reference 47 with permission from Elsevier.)

plastic zones may even extend across the whole uncracked ligament of the specimens. In this case, nonlinear elastic fracture mechanics parameters should be adopted. The J-integral concept developed by Rice (65) and the essential work of fracture (EWF) approach proposed by Broberg (66) can be used to characterize the fracture behavior of ductile materials. The determination of a critical value of J-integral (Jc) is generally carried out through the construction of the resistance

Chapter 11 Deformation Behavior of b-Crystalline Phase

343

Table 11.2 Impact Fracture Toughness (Kc and Gc ) of a-PP and b-PP (Mw ¼ 103 kg mol1 ) at Various Temperatures. [From Reference 47 with permission from Elsevier].

40 C a-PP b-PP

Gc , kJ m2

Kc, MPa m1/2

Sample

2.4 4.2

20 C

0 C

2.6 4.2

2.5 3.8

RT

40 C 20 C

2.2 2.6

1.8 5.7

2.0 6.0

0 C

RT

2.1 5.9

6.2 9.3

curve J–Da, where Da is the advanced crack length. The Jc value is determined at the point of intersection between the crack growth resistance (J–R) curve and the blunting line (J ¼ 2s y Da, where s y is the yield stress). The process of determination of Jc for ductile polymer is rather tedious and controversial as the application of different standard practices can yield different critical values (67–69). In postyielding fracture mechanics, the EWF concept has been successfully employed to characterize the fracture toughness of ductile polymers and tough composites due to its simplicity over conventional J-integral analysis (70–75). The EWF approach involves the determination of the total fracture energy (Wf) of notched specimen. It can be divided into two components: Wf ¼ We þ Wp

ð11:11Þ

The first term in Equation 11.11 is the essential work of fracture (We), that is, the work required to create new surfaces in inner fracture surface zone and is surface related. The second term is referred to as nonessential work (Wp). It is associated with various energy dissipation mechanisms that occur in the outer plastic zone and is volume related. Thus, Wf can be written as Wf ¼ we LB þ bwp L2 B wf ¼

Wf ¼ we þ bwp L LB

ð11:12Þ ð11:13Þ

where wf is the specific total fracture work, we and wp are the specific essential fracture work and specific plastic work, respectively, L is the ligament length, B is the sample thickness, and b is a shape factor of the plastic zone. The validity of EWF approach under plane-stress condition requires the ligament to be fully yielded prior to crack propagation. The validity range of ligament L under plane-stress condition is given by ð3 5ÞB L minðW=3; 2rp Þ

ð11:14Þ

where W is the width of the specimen and 2rp is the size of plastic zone. The ESIS protocol (76) for EWF recommends the use of double edge notched tensile specimen. Apparently, EWF concept is a simple method that consists of testing specimens with different ligament lengths, recording the area under the load displacement curve

344

Polyolefin Blends

(Wf), plotting the wf versus L diagram, and evaluating the best fit linear regression line. The we value is then evaluated by extrapolation to zero ligament length. It has been demonstrated by Mai that the plane strain essential work of fracture, wIe is equivalent to JIc if the sample thickness meets the condition: B 25 (wIe =s y ) (77,78). Otherwise, wIe is considered a near-plane-strain fracture toughness and is dependent on B. Karger-Kocsis and Varga determined the plane stress we and bwp values of b-PP using SEN and DENT specimens with dimensions of 100 mm 35 mm 1 mm under tensile mode (24,25). Figure 11.34a and b shows typical load–displacement curves for the a-PP and b-PP SEN specimens with different ligament lengths. Selfsimilarity in the shape of the load–displacement curves for the a-PP is not maintained when L 15 mm. At this stage, the samples fail with fast fracture due to an incomplete development of the plastic zone. The EWF approach becomes invalid for the a-PP. By contrast, the b-PP specimens show more ductile behavior as evidenced by nonlinearity in the force–displacement curves. The specific total fracture work

Figure 11.34 Tensile load–displacement curves of the SEN specimens at various ligament lengths for (a) a-PP and (b) b-PP. (From Reference 24 with permission from John Wiley & Sons.)

Chapter 11 Deformation Behavior of b-Crystalline Phase

345

Figure 11.35 Specific total work of fracture vs. ligament length plots for the tensile SEN specimens of a-PP and b-PP. Note: The data related to fast fracture of a-PP at L 15 mm (filled circles) are excluded. (From Reference 24 with permission from John Wiley & Sons.)

versus ligament plots for the a-PP and b-PP specimens are shown in Fig. 11.35. It is apparent that both PP modifications exhibit the same we value of 34 kJ m2. However, the plastic work dissipation per unit volume (bwp) of b-PP is three times higher than that of the a-PP. This implies that more energy is spent in the enlarged plastic zone of b-PP, leading to the toughness improvement as revealed by the infrared thermography frames taken during the tensile loading of SEN specimens of a-PP and b-PP. The EWF measurements also yield similar we value of 32 kJ m2 for both the a-PP and b-PP using DENT specimens. The bwp values for the a-PP and b-PP DENT specimens are 2.8 and 10 MJ m3, respectively (25). Compared to a-PP, the plastic work term is 3.6 times higher for b-PP. It is noted that the same we value obtained for both the a-PP and b-PP using either SEN or DENT specimens is derived from a lack of the ligament yielding in a-PP. The accuracy of we value is doubtful despite the data related to fast fracture of a-PP at L 15 mm are neglected (Fig. 11.35). The b-PP is expected to exhibit larger we and bwp values than a-PP owing to its ductile nature. Recently, Tordjeman et al. used EWF approach to determine the fracture toughness of PP having different b-phase contents via three-point bending tests on SEN specimens with dimensions of 100 mm 10 mm 3 mm at 500 mm min1. The PP homopolymer with variable a/b phase contents but with constant crystallinity and constant spherulite size were controlled with proper heat treatment (79). Figure 11.36 is a typical plot showing the total work of fracture versus ligament length plot for the PP homopolymer containing 70% b-phase. The resulting near-plane-strain fracture toughness and plastic work versus b-phase content for PP homopolymers are depicted in Fig. 11.37a and b, respectively. It is apparent that wIe increases markedly with increasing b-phase content. It increases from 1.1 kJ m2 for a-PP specimen to more than 6 kJ m2 for PP containing 80% b-phase. The wIe value obtained for b-PP is somewhat much smaller than that reported by Karger-Kocsis

346

Polyolefin Blends

Figure 11.36 Total work of fracture versus ligament length for PP containing 70% b-phase. (From Reference 79 with permission from Springer Science and Business Media.)

and Varga (24,25). This is attributed to the difference in the thicknesses of the specimens employed. As mentioned above, a near-plane-strain fracture toughness is thickness dependent, whilst the plane strain wIe is independent of the specimen thickness (71).

Figure 11.37 (a) Near-plane-strain essential work of fracture and (b) plastic work dissipation per unit volume as a function of b-phase content. (From Reference 79 with permission from Springer Science and Business Media.)

Chapter 11 Deformation Behavior of b-Crystalline Phase

11.4

347

CONCLUSIONS

This chapter reviews the structure–property relationship, deformation and failure behavior of b-PP homopolymer and b-PP/EPR blends subjected to tensile and impact tests. The unique sheaflike lamellar morphology of the spherulites of b-PP renders it to possess superior tensile ductility and impact strength. The molecular weights of PP and dispersed elastomer phase (EPR) influence the impact strength of b-PP and its blends considerably. Compared to a-PP, the incorporation of b-nucleator brings about distinct softening, as evidenced by lower values of yield stress at various strain rates. This behavior is also observed in the b-nucleated iPP/EPR blends. The b-PP homopolymer exhibits large strain hardening and stress whitening during tensile drawing. This is considered to be resulted from the b ! a phase transformation on the basis of TEM, DSC, and XRD measurements. However, the exact mechanism by which the strain induced b ! a phase transition takes place remains unclear. Furthermore, there is no evidence for the b ! a phase transformation during impact as revealed by the DSC results. Large plastic zone is developed in b-PP and b-nucleated iPP/EPR blends during tensile test. Similarly, extended plastic zone is observed in the b-PP during impact loading; the size of plastic zone tends to increase with increasing b-phase content. The b-phase is more effective in improving the impact strength and fracture resistance of PP homopolymer than in the rubbermodified blends. In the later case, the elastomer particles mainly control the mechanical deformation behavior. The fracture toughness of b-PP homopolymer and b-nucleated iPP/EPR blends at low strain rates (mode I) can be determined by means of the corrected LEFM approach. This approach can be applied to characterize the toughness of ductile b-PP and b-nucleated iPP/EPR blends provided that the cracktip plasticity effect is taken into account. In this regard, a plastic zone correction is needed to obtain intrinsic fracture toughness values. Finally, the EWF concept is more adequate to evaluate the fracture toughness of ductile b-PP. The specific work of fracture of b-PP is found to increase markedly with increasing b-phase content.

NOMENCLATURE a B b e_ E f ða=WÞ Fmax Gc DH Ið110Þa Ið040Þa Ið130Þa

Crack length Sample thickness Shape factor of the plastic zone Strain rate Young’s modulus Geometric factor Maximum load Critical strain energy release rate Activation energy of the plastic flow Integrated area of (110) peak of a-PP Integrated area of (040) peak of a-PP Integrated area of (130) peak of a-PP

348

Polyolefin Blends

Ið300Þb Ið301Þb K KI L y Fða=WÞ 2rp R S t T V we wp Wf W

Integrated area of (300) peak of b-PP Integrated area of (301) peak of b-PP Content of b-PP phase Stress intensity under mode I Ligament length Poisson’s ratio Calibration factor depending on sample geometry Size of plastic zone Universal gas constant Order parameter Sample thickness Absolute temperature Activation volume Specific essential fracture work Specific plastic work Total fracture work Width of sample

REFERENCES 1. S. V. Meille, D. R. Ferro, S. Bruckner, A. J. Lovinger, and F. J. Padden, Macromolecules, 27, 2615 (1994). 2. J. Varga and J. Jarger-Kocsis, J. Polym. Sci. B: Polym. Phys., 34, 657 (1996). 3. H. J. Leugering, Makromol. Chem., 109, 204 (1967). 4. J. Garbarczyk and D. Paukszta, Colloid Polym. Sci., 263, 985 (1985). 5. P. Jacobi, B. H. Bersted, W. J. Kissel, and C. E. Smith, J. Polym. Sci. Polym. Phys. Edn., 24, 461 (1986). 6. G. Shi and Z. Zhang, US Patent 5231126 (1993). 7. G. Shi and X. Zhang, Thermochim. Acta, 205, 235 (1992). 8. J. Varga, I. Mudra, and G. W. Ehrenstein, J. Appl. Polym. Sci., 74, 2357 (1999). 9. F. Chu and Y. Kimura, Polymer, 37, 573 (1996). 10. M. Obadal, R. Cermak, M. Raab, V. Verney, S. Commereuc, and F. Fraisse, Polym. Degr. Stab., 91, 459 (2006). 11. J. Liu, Z. Wei, and Q. Guo, J. Appl. Polym. Sci., 41, 2829 (1990). 12. P. M. McGenity, J. J. Hooper, C. D. Paynter, A. M. Riley, C. Nutbeem, N. J. Elton, and J. M. Adams, Polymer, 33, 5215 (1992). 13. A. Turner Jones, J. M. Aizlewood, and D. R. Beckett, Makromol. Chem., 75, 134 (1964). 14. G. Zhou, Z. He, J. Yu, Z. Han, and G. Shi, Makromol. Chem., 187, 633 (1986). 15. S. C. Tjong, J. S. Shen, and R. K. Y. Li, Polym. Eng. Sci., 36, 100 (1996). 16. R. Cermak, M. Obadal, P. Ponizil, M. Polaskova, K. Stoklasa, and A. Lengalova, Eur. Polym. J., 41, 1838 (2005). 17. D. R. Norton and A. Keller, Polymer, 26, 704 (1985). 18. S. C. Tjong, J. S. Shen, and R. K. Y. Li, Polymer, 37, 2309 (1996). 19. J. Varga and G. W. Ehrenstein, Colloid Polym. Sci., 275, 511 (1997). 20. J. Varga, J. Macromol. Sci. B: Phys., 41, 1121 (2002).

Chapter 11 Deformation Behavior of b-Crystalline Phase

349

21. A. Menyhard, J. Varga, A. Liber, and G. Belina, Eur. Polym. J., 41, 669 (2005). 22. M. C. Cramez, M. J. Oliveira, and R. J. Crawford, J. Mater. Sci., 36, 2151 (2001). 23. T. Labour, L. Ferry, C. Gauthier, P. Hajji, and G. Vigier, J. Appl. Polym. Sci., 74, 195 (1999). 24. J. Karger-Kocsis and J. Varga, J. Appl. Polym. Sci., 62, 291 (1996). 25. J. Karger-Kocsis, Polym. Eng. Sci., 36, 203 (1996). 26. J. X. Li and W. L. Cheung, Polymer, 39, 6935 (1998). 27. J. X. Li, W. L. Cheung, and C. M. Chan, Polymer, 40, 2089(1999). 28. M. Aboulfaraj, C. G’Sell, B. Ulrich, and A. Dahoun, Polymer, 36, 731 (1995). 29. W. Jiang, S. C. Tjong, and R. K. Y. Li, Polymer, 41, 3479 (2000). 30. M. Denac, I. Smit, and V. Musil, Composites Part A, 36, 1094 (2005). 31. C. Grein, C. J. G. Plummer, H. H. Kausch, Y. Germain, and Ph. Beguelin, Polymer, 43, 3279 (2002). 32. C. Grein, Adv. Polym. Sci., 188, 43 (2005). 33. S. C. Tjong and R. K. Y. Li, J. Vinyl Technol., 3, 89 (1997). 34. S. C. Tjong, R. K. Y. Li, and T. Cheung, Polym. Eng. Sci., 37, 166 (1997). 35. H. B. Chen, J. Karger-Kocsis, J. S. Wu, and J. Varga, Polymer, 43, 6505 (2002). 36. M. Raab, J. Scudla, and J. Kolarik, Eur. Polym. J., 40, 1317 (2004). 37. F. Chu and Y. Kimura, Polymer, 37, 573 (1996). 38. G. Shi, X. Zhang, J. Zheng, and G. Zhou, Intern. Polym. Process., 10, 330 (1995). 39. F. Chu, T. Yamaoka, and Y. Kimura, Polymer, 36, 2523 (1995). 40. ASTM D256-05a: Standard Test Methods for Determining the Izod Pendulum Impact Resistance of Plastics, p. 1–20, American Society for Testing and Materials, Philadelphia, 2005. 41. ASTM D 5628-96: Standard Test Method for Impact Resistance of Flat, Rigid Plastic Specimens by Means of a Falling Dart (Tup or Falling Mass), pp. 1–10, American Society for Testing and Materials, Philadelphia, 1996. 42. T. Yoshida, Y. Fujiwara, and T. Asano, Polymer, 24, 925 (1983) 43. T. Asano, Y. Fujiwara, and T. Yoshida, Polym. J., 11, 383 (1979). 44. J. Kotek, I. Kelnar, J. Baldrian, and M. Raab, Eur. Polym. J., 40, 679 (2004). 45. J. Karger-Kocsis, J. Varga, and G. W. Ehrenstein, J. Appl. Polym. Sci., 64 (1997) 2057. 46. J. Karger-Kocsis, D. E. Mouzakis, G. W. Ehrenstein, and J. Varga, J. Appl. Polym. Sci., 73, 1205 (1999). 47. H. B. Chen, J. Karger-Kocsis, J. S. Wu, and J. Varga, Polymer, 43, 6505 (2002). 48. W. Xu, D. C. Martin, and E. M. Arruda, Polymer, 46, 455 (2005). 49. G. Coulon, G. Castelein, and C. G’Sell, Polymer, 40, 95(1998). 50. K. Cho, D. N. Saheb, J. Choi, and H. Yang, Polymer, 43, 1407 (2002). 51. K. Cho, D. N. Saheb, H. Yang, B. I. Kang, J. Kim, and S. S. Lee, Polymer, 44, 4053 (2003). 52. T. Labour, G. Vigier, R. Seguela, C. Gauthier, G. Orange, and Y. Bomal, J. Polym. Sci. B: Polym. Phys., 40, 31 (2002). 53. J. C. Wittmann and B. Lotz, Polymer, 30, 27 (1989). 54. B. Lotz, C. Mathieu, A. Thierry, A. J. Lovinger, C. De Rosa, O. Ruiz de Ballesteros, and F. Auriemma, Macromolecules, 31, 9253 (1998). 55. K. S. Macturk, R. K. Eby, and B. L. Farmer, Polymer, 37, 4999 (1996). 56. C. Grein, K. Bernreitner, A. Hauer, M. Gahleitner, and W. Neibl, J. Appl. Polym. Sci., 87, 1702 (2003). 57. G. R. Irwin, Fracture, in Encyclopedia of Physics, Vol. VI, Springer, Berlin, 1958. 58. G. R. Irwin, J. Appl. Mech., 24, 361 (1957). 59. A. Pavan, Fracture Mechanics Testing Methods for Polymers, Adhesives and Composites, in: ESIS Publication 28, D. R. Moore, A. Pavan, and J. D. Williams (eds.), Elsevier, Oxford, 2001, p. 27.

350

Polyolefin Blends

60. J. G. Williams and M. J. Cawood, Polym. Test., 9, 15 (1990). 61. ASTM Annual Book of Standards, American Society of Testing Materials and Philadelphia, PA, E-399, 1987, pp. 417–427. 62. R. Gensler, C. J. G. Plummer, C. Grein, and H. H. Kausch, Polymer, 41, 3809 (2000). 63. C. Grein, H. H. Kausch, and Ph. Beguelin, Polym. Test., 22, 733 (2003). 64. S. C. Tjong, J. S. Shen, and R. K. Y. Li, Scr. Metall. Mater., 33, 503 (1995). 65. J. R. Rice, J. Appl. Mech., 35, 379 (1968) 66. K. B. Broberg, Int. J. Fract., 4, 11 (1968). 67. ASTM E813–81: Standard Method for JIC, A Measure of Fracture Toughness, American Society for Testing and Materials, Philadelphia, 1981, p. 810. 68. ASTM E813–87: Standard Method for JIC, A Measure of Fracture Toughness, American Society for Testing and Materials, Philadelphia, 1987, p. 1968. 69. ASTM E813–89: Standard Method for JIC, A Measure of Fracture Toughness, American Society for Testing and Materials, Philadelphia, 1989, p. 700. 70. C. A. Paton and S. Hashemi, J. Mater. Sci., 27, 2279 (1992). 71. S. Hashemi, J. Mater. Sci., 28, 6178 (1993). 72. D. E. Mouzakis, F. Stricker, R. Mulhaupt, and J. Karger-Kocsis, J. Mater. Sci., 33, 2552 (1998). 73. S. C. Tjong, S. A. Xu, and R. K. Y. Li, J. Appl. Polym. Sci., 77, 2074(2000). 74. S. C. Tjong and S. P. Bao, J. Polym. Sci. B: Polym. Phys., 43, 585 (2005). 75. S. C. Tjong, S. P. Bao, and G. D. Liang, J. Polym. Sci. B: Polym. Phys., 43, 3112 (2005). 76. E. Clutton, Essential work of fracture, in: Fracture Mechanics Testing Methods for Polymers, Adhesives and Composites, ESIS Publication 28, D. R. Moore, A. Pavan, and J. D. Williams (eds.), Elsevier, Oxford, 2001 77. Y. W. Mai and B. Cotterell, Int. J. Fract., 32, 105 (1986). 78. J. S. Wu and Y. W. Mai, Polym. Eng. Sci., 36, 2275 (1996). 79. Ph. Tordjeman, C. Robert, G. Marin, and P. Gerard, Eur. Phys. J., 4, 459 (2001).

Chapter

12

Multiphase Polypropylene Copolymer Blends Francis M. Mirabella1

12.1 INTRODUCTION The most widely used multiphase polymer system is polypropylene impact copolymer. These copolymers are typically composed of isotactic polypropylene (iPP) and poly(ethylene–propylene), referred to as ethylene–propylene rubber or EPR. The world demand for all types of polypropylene is about 90 billion pounds per year. About 22 billion pounds per year of that total are impact polypropylene copolymers, referred to by various names such as high impact polypropylene (hiPP) and thermoplastic olefin (TPO). Growth is very strong at about 10% per year. These PP copolymers are primarily used in injection-molded parts for automotive, appliances, and other durable goods applications, as well as for extruded sheet and thermoforming. The wide range of physical and mechanical properties, relative ease of processing, and low density constitute these polypropylene copolymers as extremely attractive materials capable of competing with more expensive plastics in many demanding applications. The automotive industry has made TPO resins the primary choice for an increasing range of interior and exterior applications. Interior applications include instrument panels, consoles, door panels, and pillars. Exterior applications include bumpers, fascia, body side cladding, rocker panels, and cowl vent grilles. Over the last several decades, these TPO resins have progressively replaced other polymeric compositions in interior and exterior applications due to their desirable balance of properties and safety attributes. This chapter presents a discussion of the multiphase polypropylene copolymers with emphasis on the commercially important blends. The discussion will be more narrowly focused on the composition, structure, and phase morphology of these

1

Lyondell Chemical Co., Equistar Technology Center, Cincinnati, OH 45249, USA

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

351

352

Polyolefin Blends

commercial polymer systems to the exclusion of related systems of mainly academic interest.

12.1.1 Commercial Production Impact polypropylene copolymers are produced by various processes, but are generally characterized by the synthesis of iPP in the first reactor and EPR in the second reactor (1–3). Therefore, these systems are typically reactor blends. Postreactor blending is a common practice, but the starting material is most often the reactor blend polypropylene copolymer. In the minority are blends employing polypropylene as the starting material to which the rubber phase and other components are added. The iPP is typically the majority, that is, matrix, phase, while the EPR is the minority, that is, dispersed, phase. The EPR has low Tg and increases the impact strength, but lowers the stiffness of the system. The iPP has high crystallinity and stiffness and acts as a rigid matrix. The iPP also has high Tg . The EPR particles typically average about 0.5–1.0 mm in diameter with the entire distribution of particle sizes ranging from about 0.1 to 3 mm. The reactor grades of impact polypropylene copolymers are often compounded with other components, especially other toughening agents with low Tg, such as ethylene–propylene diene monomer (EPDM), metallocene ethylene–a-olefin copolymers, styrenic block copolymers, and so on. The compounded systems containing much higher rubber content relative to the base resin are often called thermoplastic elastomers (TPEs). In the TPEs, the rubbery components may constitute the major phase. However, TPEs include many other base resins, which are not polyolefins, such as polyurethanes, copolyamides (segmented block copolymers), copolyesters (segmented block copolymers), styrenics, and so on. Polypropylene homopolymer, compounded with EPDM in a dynamic melt-mixing/curing process, is often called thermoplastic vulcanized elastomer (TPV), which offers chemical cross-linking of the rubber phase. Paraffinic oils are always included in the melt-mixing process for viscosity control and cost. The entire range of systems in these categories is difficult to specify, since compounders have the option to blend an extremely diverse range of materials. The rigid hiPP and TPO grades account for the majority of PP copolymer demand, while the elastomeric TPE grades are in the minority. Total TPE demand is about 5.0 billion pounds per year. Other common additives to polypropylene copolymer compounds are talc, nucleating agents, clarifiers, other inorganic fillers, and so on. A recent movement toward the addition of nanofillers such as organoclay (especially oganically modified montmorillonite), TiO2, SiO2, and so on has begun to gain wider acceptance and utility.

12.1.2 Morphology of Commercial Impact PP Copolymers Excellent treatments of the compatibility, thermodynamics, phase separation (including kinetics), rheology, physical and mechanical properties, fracture phenomena, and so on have been given in numerous books, for example, the two-volume

Chapter 12 Multiphase Polypropylene Copolymer Blends

353

treatise by Paul and Newman (4). Quite a bit of attention in such treatments is inclined toward miscibility and phase separation. In the case of reactor grades or compounded polypropylene impact copolymers, however, these exhibit a phaseseparated morphology in the commercial reactors and undergo modification in that morphology in the postreactor compounding extruders. These systems are never one phase, but are already multiphasic, and remain so, when transferred into the melt state. It is the nature of the production of these copolymers, which yields systems that are always multiphasic. It should be noted, however, that the equilibrium concentrations of the components are not necessarily established until the already multiphasic system is transferred into the melt state at a specified temperature. Phase changes during this process are expected to be relatively small, but significant for the establishment of the true equilibrium concentrations of the components under the specified conditions in the melt. Therefore, considerations of miscibility, conditions for and kinetics of phase separation, and the like are not typically relevant to polypropylene impact copolymers. The development of the morphology of impact polypropylene copolymers follows a pattern from the commercial reactor to finished parts, such that the system is multiphasic throughout these conversions. However, the morphology exhibits some large changes in some steps. The form of the polymer after synthesis in the commercial reactors is typically powder, or particles in some newer processes. The powder or particles are in the range of 0.5–1.0 mm (flakes) and 1–2 mm in diameter, respectively. The morphology in the case of powder is extremely heterogeneous with some regions in the powder containing no observable EPR and other regions containing large ‘‘pools’’ of EPR in the rigid iPP matrix. This heterogeneous morphology is shown in atomic force microscopy (AFM) images of several representative powder flakes from a single lot of reactor powder in Figs. 12.1–12.4. Figure 12.1 is an example of a region in the powder containing no observable EPR, while Fig. 12.4 shows a region containing ‘‘pools’’ of EPR, which are roughly 5 mm in diameter and exhibit connectivity between EPR domains. Figure 12.2 shows a region with a very sparse population of (0.1–0.3 mm diameter spherical) EPR domains, while Fig. 12.3 shows interconnected (0.1–0.7 mm wide) channels of EPR. Note that AFM is performed on the powder by imbedding it in an adhesive, microtoming a flat surface, and imaging the surface with no further treatment, such as etching or staining used in scanning electron or transmission electron microscopy, respectively (5). In the AFM micrographs, the topographic image is on the left and phase contrast image is on the right. In the topographic image, dark areas are low and bright areas are high and in the phase contrast image dark areas are soft and bright areas are hard. Therefore, the EPR particles appear dark, since these are noncrystalline, in the phase contrast images and, in fact, also typically appear dark in the topographic images due to an artifact of the microtomy (5). The iPP matrix phase appears bright in the phase contrast image, since it is crystalline and, therefore, hard. Similar results were obtained in a study of impact polypropylene copolymer particles produced in the Basell Spheripol process (6) in which EPR concentration varied widely in particles of different sizes (7). The EPR concentration was found to decrease with increasing particle size. In a recent study, the detailed physical

354

Polyolefin Blends

Figure 12.1 Polypropylene TPO reactor powder morphology. AFM image of embedded and cryofaced-off Rx powder particle. Topographic image is on left and phase contrast image is on right. In the topographic image, dark areas are low and bright areas are high and in the phase contrast image dark areas are soft and bright areas are hard.

Figure 12.2 Polypropylene TPO reactor powder morphology. AFM image of embedded and cryofaced-off reactor powder particle.

Chapter 12 Multiphase Polypropylene Copolymer Blends

355

Figure 12.3 Polypropylene TPO reactor powder morphology. AFM image of embedded and cryofaced-off reactor powder particle.

structure of the hiPP particle produced in a two-stage laboratory process was studied, as well as the formation and distribution of the EPR (8). The commercially produced polymer is normally formed into pellets in the commercial plant by mechanical mixing in a pelletization extruder (9) for convenient shipment by railcar, truck, and so on. In order to produce a consistent material for

Figure 12.4 Polypropylene TPO reactor powder morphology. AFM image of embedded and cryofaced-off reactor powder particle.

356

Polyolefin Blends

Figure 12.5 Polypropylene TPO extruded pellet (made from reactor powder in Figs. 12.1–12.4) morphology. AFM image of embedded and cryo-faced-off reactor powder particle.

commercial sale, the pellets are required to exhibit a much more homogeneous dispersion of the EPR in the iPP matrix than exhibited in the commercial reactor product. This is accomplished by the input of a large amount of work in the form of shearing into the molten polymer in the extruder. The extruder design is judiciously done to maximize dispersion of the EPR phase, but minimize polymer degradation. The pellets in fact normally do exhibit homogeneous dispersion of the EPR, although particle size distribution is typically relatively broad. This is shown in Fig. 12.5 for pellets formed from the powder in Figs. 12.1–12.4. It may be observed in Fig. 12.5 that the EPR particle size distribution is broad and that particles are generally round (except for obvious coalescence of neighboring particles) indicating that EPR particles are spherical droplets in the molten state. The EPR particles in Fig. 12.5 have an included hard phase, which is due to crystalline polyethylene copolymers that segregate away from the iPP matrix due to their incompatibility (10,11). The fabrication of useful articles from the pellets is the usual next step (8). Injection molding is one of many processes that are used to fabricate articles. Injection molding is used to fabricate a variety of articles, such as automotive and appliance parts. Figure 12.6 shows AFM images of an injection-molded tensile bar used for mechanical property testing, which was molded from the commercial pellets in Fig. 12.5. The specimen was taken from the ‘‘core’’ of the bar, where residual stresses are minimal due to slow cooling (12). This core specimen of the injectionmolded bar in Fig. 12.6 exhibits very similar morphology to the pellets in Fig. 12.5, although, some evidence of EPR particle coarsening may be observed in Fig. 12.6. This coarsening is due to incorporation of material in smaller particles into larger particles, which will be discussed later. In some applications, the commercial plant

Chapter 12 Multiphase Polypropylene Copolymer Blends

357

Figure 12.6 Polypropylene TPO injection-molded bar (made from commercial pellets in Fig. 12.5) morphology. AFM image of cryo-faced-off section of core of injection-molded bar.

pellets are often compounded with additional components, as discussed in Section 12.1.1, and are repelletized for subsequent use in the fabrication process.

12.2

DISPERSIVE MIXING DURING PROCESSING

Figures 12.1–12.6 show the radical change in EPR particle morphology from reactor powder to pellets, but the relatively static morphology from pellets to fabricated articles. This is due to the great efficiency of commercial-scale corotating twin-screw pelletization extruders (8). The EPR phase is efficiently dispersed and attains the ‘‘stationary’’ value of particle size, as described by theoretical treatments of droplet breakup and coalescence (13–15). This droplet breakup and coalescence occurs in the molten state of the viscoelastic iPP and EPR, matrix and dispersed phases, in the extruder under a complex strain field, which is a combination of nonuniform, transient shear and elongational fields. Further, a variable temperature profile is used along the barrel of the extruder causing complex variation in the viscoelastic properties of these components. A set of empirical equations was obtained by Wu to describe the dispersed phase average particle size obtained after dispersive mixing in an extruder (13). The equations were based on the case of a Newtonian drop suspended in a Newtonian matrix, that is, Taylor’s theory (16,17) with an extension to the case of a viscoelastic drop in a viscoelastic matrix. The empirical data employed were for blends containing 15 wt% dispersed phase and 85 wt% matrix phase. The particle size was found to be critically dependent on the ratio of the dispersed phase to the matrix phase melt

358

Polyolefin Blends

viscosity (p). The equation for the case of higher melt viscosity of the dispersed phase than the matrix phase is an ¼

4gp0:84 Ghm

ðfor p > 1Þ

ð12:1Þ

where p ¼ hd =hm , hd is the melt viscosity of the dispersed phase, hm is the melt viscosity of the matrix phase, G is the effective shear rate, an is the number-average particle diameter, and g is the interfacial tension. The same equation is used for p < 1, except that the exponent is 0:84. The two equations for p > 1 and p < 1 predict a V-shaped curve with particle size at a minimum at p ¼ 1. This empirical treatment might be expected to be fairly realistic, due to its recognition of the nonNewtonian behavior of the multiphase systems and the finite dispersed phase concentration. In contrast and for comparison, the theoretical equation from Taylor’s theory (16, 17) for a Newtonian drop suspended in a Newtonian matrix with the concentration of the dispersed phase particle assumed to be vanishingly small is an ¼

4gðp þ 1Þ Ghm p 19 4 þ4

ðfor p < 2:5Þ

ð12:2Þ

The calculated particle diameters from Equation 12.2 may be considered a lower limit, that is, the ‘‘Taylor limit,’’ due to the assumption of Newtonian behavior of the system and vanishingly small concentration of the dispersed phase. Polymers exhibit non-Newtonian behavior, namely, the droplets elongate elastically before breaking. This behavior corresponds to an increase in interfacial tension, and therefore, particle size increases as predicted by Equation 12.1, over that predicted from Equation 12.2. (This is discussed below and can be seen in the last two columns of Table 12.3). A study was done to determine the effect of extruder conditions on dispersed phase particle size for an hiPP. Table 12.1 presents polymer data on the hiPP (assumed to be a binary iPP/EPR blend) studied. Table 12.2 shows the Berstorff twin-screw corotating extruder conditions. Extruder runs were done under low (0-series) and high (1-series) shear rate conditions and nitrogen purge. Temperature profiles were somewhat different along the extruder for the 0- and 1-series experiments. Table 12.3 presents particle size determined by scanning electron microscopy (SEM), along with the calculated values of an from Equations 12.1 and 12.2. The dn, dw, and dz values in Table 12.3 are the number, weight and Z-average diameters determined by SEM, respectively. The molecular weight of the EPR in the impact PP Table 12.1

PP/EPR Blend Data.

Polymer iPP/EPR blend iPP EPR (33 wt% C2)

wt%

Mw ð103 Þ

— 84 16

620 510 935

Mw =Mn 10.3 7.3 20.4

h @ 230 C @100 s1 (Pa s) 19000 9400 82600

g (mN/m1) 0.2 — —

Chapter 12 Multiphase Polypropylene Copolymer Blends Table 12.2 Run ID 0-1 0-2 0-3 1-1 1-2 1-3

359

Berstorff Twin-Screw Corotating Extruder Conditions. T ( F) zones 1–7

Residence time (s)

257–422 323–454 325–416 360–419 394–447 394–411

Shear rate (s1)

Feed rate (lbh1)

115 115 115 172 172 172

250 250 250 250 250 250

75 75 75 87 87 87

is approximately two times that of the iPP. Higher molecular weight of the EPR, relative to the iPP in impact PP resins, is often the case in hiPP resins. The melt viscosities at the operating shear rate of the extruder followed these trends; therefore, the values of p were much larger than 1 in this case (see Table 12.1). The concentration of the dispersed phase (EPR) was 16 wt%. The particle size observed (dx values in Table 12.3) was two to three orders of magnitude larger than that calculated (an values in Table 12.3) from Equations 12.1 and 12.2. The much larger particle sizes observed compared to those calculated are probably due to the very high values of hm and p, and, also, due to the effects of flow-induced coalescence, as has been clearly demonstrated for similar systems (18,19). In those studies it was found that the observed particle size for systems with higher dispersed phase concentration increased exponentially at high dispersed phase concentration. However, the values of hm and p were significantly smaller (950 Pa s and 0.9, respectively) in that study than those in the present study. Therefore, it is not surprising that the observed particle sizes in this study exceeded those calculated by Wu’s empirical equation and by Taylor’s theoretical equation. This indicates the very crude nature of the predictions and the necessity of experimental determination of particle morphology in dispersively mixed systems. Although, the 0-series was run at lower shear rate than the 1-series, the experimentally observed particles sizes were larger in the higher shear rate 1-series. The cause of this observation is not known, but might relate to the differing extruder temperature profiles.

Table 12.3 Particle Size Data for Dispersed Phase in hiPP a.

Run

dn ðmmÞ

0-1 0-2 0-3 1-1 1-2 1-3

0.49 0.58 0.51 0.57 0.57 0.61

a

dw ðmmÞ 0.50 0.59 0.53 0.61 0.58 0.62

dz ðmmÞ

dw =dn

dz =dw

Wu limit, an ðmmÞ

0.63 0.73 0.67 0.76 0.72 0.82

1.02 1.02 1.04 1.07 1.02 1.02

1.26 1.24 1.26 1.25 1.24 1.32

0.0046 0.0046 0.0046 0.0030 0.0030 0.0030

From Reference 28 with permission from John Wiley & Sons, Inc.

Taylor limit, an ðmmÞ 0.00015 0.00015 0.00015 0.00010 0.00010 0.00010

360

Polyolefin Blends

12.3 MOLECULAR STRUCTURE OF IMPACT PP COPOLYMERS The nominal components of impact PP that are targeted in commercial production are iPP and EPR containing about 25–50 wt% ethylene, so that the EPR is essentially noncrystalline. However, the composition of these copolymers is more complex, due to the broad composition distribution of the ethylene/propylene copolymers, typically made in the second reactor, and to a lesser extent due to the tacticity distribution of the iPP, typically made in the first reactor. The structure of a commercial impact PP was determined by preparative temperature rising elution fractionation (P-TREF), followed by analysis of the fractions by nuclear magnetic resonance (NMR) spectroscopy in order to determine chemical composition and tacticity (10,20). It was found that the hiPP (MFR ¼ 6) was composed of about 75 wt% iPP with average tacticity of about 97% and exhibiting a broad tacticity distribution extending to very low values, 17 wt% EPR containing an average of about 48 wt% ethylene and exhibiting a broad composition distribution of about 15–80 wt% ethylene and about 8 wt% of an ethylene-rich copolymer containing 92 wt% ethylene. It should be noted that, although these commercial PP copolymers are often called ‘‘block’’ copolymers, no evidence was found in this study to indicate that there was significant block copolymer formation. The ethylene–propylene copolymers made in the second reactor are random copolymers and no detectable block copolymers are formed between first and second reactor polymers. That is, the iPP is ‘‘dead’’ polymer upon arriving in the second reactor. This result has been found for a wide range of commercial PP copolymer resins. The compositional structure is further complicated by the molecular weight heterogeneity of the various components. For the case of this MFR ¼ 6 hiPP, which had Mw ﬃ 380,000 and Mw =Mn ¼ 6:3, the iPP had Mw ﬃ 382,000 and Mw =Mn ¼ 4:3 and the EPR had Mw ﬃ 342,000 and Mw =Mn ¼ 14:2. Therefore, theoretical calculations for the prediction of the properties of these systems are complicated by the heterogeneity in composition and molecular weight. The composition and molecular weight of the various components of the hiPP and TPO resins may vary widely. Although, this is the case, theoretical calculations are routinely done under the assumption that these systems are simple binary blends of homogeneous (composition and molecular weight) iPP and EPR.

12.4 COARSENING IN MULTIPHASE PP COPOLYMER SYSTEMS 12.4.1 Background It was argued above that phase separation from a one-phase melt was not typically relevant to systems based on impact polypropylenes. Further, it was shown how the initial particle size distribution is created in an extruder, due to efficient dispersive mixing. Also, it was shown that this initial particle size distribution did not tend to be greatly modified in subsequent processing steps, namely part fabrication.

Chapter 12 Multiphase Polypropylene Copolymer Blends

361

However, it was mentioned that coarsening of the dispersed phase particles is a potentially important effect in systems in which the two-phase melt experiences, at least partly, a quiescent state. In the case in which a polymer system begins as a one-phase melt, the sequence of events is as follows: (1) rapid liquid–liquid phase separation proceeding with continuous change in phase compositions and volume fractions, (2) attainment of ‘‘equilibrium,’’ that is, stationary values of phase compositions and volume fractions and particle size distribution, (3) the excess free energy of the system due to excess interfacial energy of the small dispersed phase droplets (or particles) may be reduced by the much slower process of coarsening of the droplets, during which process the phase compositions and volume fractions remain constant. The late-stage coarsening process occurs over a time scale many orders of magnitude larger than the early-stage phase separation process. The PP copolymers do not normally exist as a one-phase melt; therefore, only the coarsening process is of material interest for these systems. The demixing of the immiscible components in an initially one-phase melt proceeds rapidly due to a thermodynamic instability, this is spinodal decomposition, or a thermodynamic metastability (i.e., nucleation and growth of a new phase). However, this distinction of the character of the liquid–liquid phase transition is largely academic for the purposes of this work, because in either case the system rapidly produces droplets of a new phase (i.e., for off-critical mixtures). The demixing process itself was not studied in detail in this work. However, further details of this process may be found elsewhere (21). The production of droplets of a new phase in these immiscible systems is extremely rapid compared with the subsequent coarsening of the system to the final morphology. The coarsening of the phase-separated system can occur by two mechanisms. Particle diffusion, collision, and coalescence is one mechanism. Particle diffusion occurs in the quiescent melt by Brownian motion of the particles. Another mechanism is evaporation and condensation, called Ostwald ripening. Ostwald ripening occurs by molecular diffusion of the minor component, which primarily makes up the minor phase particles, through the matrix phase. This results in ‘‘evaporation’’ of particles smaller than a critical radius by diffusion of the minor component out of these and growth of particles larger than the critical radius by ‘‘condensation’’ of the diffusing molecules into these. A simplistic rationale for the differing timescales of phase separation, for example, by the nucleation and growth process, and the subsequent coarsening process may be developed as follows. At time zero, the molecules of a binary solution, with molecules A and B being the major and minor components, respectively, are intimately mixed at the molecular level (i.e., there is no macroscopic phase segregation). This system has an extremely high driving force for segregation of the two types of molecules, since the solution is highly supersaturated. Therefore, the molecules separate rapidly into two macroscopic phases. This results in two phases with the thermodynamic equilibrium concentrations of B in the A-rich phase and A in the B-rich phase. At this point of the very rapid process, the system still has significant excess of energy due to the large surface area of the small particles of the minor phase B. In the next step of the process, the particles of the minor phase, B,

362

Polyolefin Blends

grow relatively slowly by combining of the nuclei of B. This occurs by the diffusion of molecules of B through the dilute solution of B molecules in the A matrix. The driving force for this process is different from the first process. In this case, the B nuclei have extremely high surface area and, therefore, extremely high cumulative interfacial tension against the A matrix. The driving force then is to minimize the surface area, thereby minimizing the cumulative interfacial tension between A and B. This is done by simply increasing the particle size of B. The minimum of the interfacial tension is one particle of B in the A matrix. The slower coarsening process is of interest because it occurs on a timescale comparable to melt processing procedures, such as pelletization, injection molding, film blowing, and the like. Coarsening can occur by coalescence or Ostwald ripening, or by both simultaneously. The coarsening of a two-phase immiscible system has been kinetically modeled for the case of metal alloys. The classical derivation was done, following the theory of Ostwald ripening (22) for immiscible metals in which grains of a new phase grow from a matrix of the original supersaturated solution. Again, the first step is very rapid (i.e., nucleation and growth) and is not included in the kinetic scheme of the model. The second step treats the slower growth of the grains of the minor phase in order to reduce the interfacial tension between the grains and the matrix. The results of this derivation for metals are most compactly expressed by the Lifshitz–Slyozov equation for a binary system (23): r 3 ðtÞ ¼ r 3 ð0Þ þ Kt

ð12:3Þ

where rðtÞ and rð0Þ are the particle radii at time t and t ¼ 0 (where zero time is defined as the beginning of coarsening, or some arbitrary time during coarsening, in the longtime regime) and K is a constant, which is independent of the volume fraction of the coarsening phase f. Although the overall driving force of the coarsening process is generally accepted to be a reduction of the interfacial area (and concomitantly the system energy), the complex diffusional processes occurring during coarsening have undergone limited study. A central assumption in the theory leading to equations of the form of Equation 12.3 is that coarsening occurs for the case of infinitely separated, noninteracting spheres. Therefore, it is not certain that polymeric systems with high volume fraction of the coarsening phase will generally adhere to this theory developed for metals. An alternative mechanism for the longtime coarsening regime, which also follows a r t1=3 dependence, is diffusion and coalescence. Coalescence occurs by the movement of the dispersed phase particles through the matrix by Brownian motion (diffusion), collision, and formation of fewer, larger particles (24,25). Coalescence follows the same law as stated in Equation 12.3. The rate constant K in Equation 12.3 can be expressed for Ostwald ripening as KOR ¼

8 DVm gCea 9 RT

ð12:4Þ

where D is the molecular diffusion coefficient of the minor component in the matrix phase (usually the mutual diffusion coefficient), Vm is the molar volume of the minor

Chapter 12 Multiphase Polypropylene Copolymer Blends

363

component, g is the interfacial tension between the matrix and droplet phase, Cea is the equilibrium concentration (mole fraction) of the minor phase in the matrix phase, R is the gas constant, and T is the absolute temperature. As noted in Section 12.1.2 the equilibrium concentrations of the components are not necessarily established until the already multiphasic system is transferred into the melt state at a specified temperature. Although phase changes during this process are expected to be relatively small, these may be significant for the establishment of the true equilibrium concentrations of the components under the specified conditions in the melt, for example, Cea . The rate constant K in Equation 12.3 can be expressed for coalescence as (26) Kc ¼

DBr ¼

2kTf ph

ð12:5Þ

kT 6ph r

ð12:6Þ

Particle diameter

Long-time regime

log dw

Short-time regime

where k is the Boltzman constant, f is the volume fraction of the minor (droplet) phase, h is the melt viscosity of the major phase, and r is the particle radius. DBr is the droplet diffusion coefficient, due to Brownian motion (26). It may be noted that the coalescence mechanism depends on the reciprocal of the melt viscosity (Eq. 12.5). The two different timescales for phase separation and coarsening are shown schematically in Fig. 12.7. As may be seen in Fig. 12.7, particle diameter increases

log time dwo

0 Time

Figure 12.7 Schematic representation of the short-time and longtime coarsening regimes. The diameter at the beginning of the longtime regime (dw0) corresponds to zero time in Equation 12.3. Note that the lower time axis represents an unspecified time dependence for the two regimes. However, in the coarsening regime the dependence is specified as log time. (From Reference 28 with permission from John Wiley & Sons, Inc.)

364

Polyolefin Blends

on a steeply ascending line in the short-time demixing regime, while it increases on a much less steeply ascending line in the longtime coarsening regime. The changeover from the short-time phase demixing regime to the longtime coarsening regime has been interpreted as ‘‘pinning,’’ that is, cessation of morphology modification (27). However, the morphology of the system continues to change, that is, coarsens, in the longtime regime, but at a much slower rate of change as compared to the short-time regime (28). A qualitative consideration of the two mechanisms, according to Equations 12.4–12.6, logically results in the assumption that coalescence is only favored in systems with lower melt viscosity matrices, while Ostwald ripening is favored in higher melt viscosity matrices. Additionally, Ostwald ripening is only favored if the minor phase has finite solubility in the matrix phase. In fact, a quantitative comparison of theory to experiment was made in which binary blends of essentially monodisperse components were made by Crist and Nesarikar (29). That study showed that coalescence could be effectively suppressed by increasing the melt viscosity h of the matrix, thereby decreasing the diffusion DBr due to Brownian motion and, thereby, inhibiting collision of the droplet phase. Conversely, Ostwald ripening could be suppressed by decreasing either the rate of molecular diffusion D of the minor phase or the solubility Cea of the minor phase in the matrix phase. By the manipulation of these parameters, the coarsening mechanism was modified from nearly pure coalescence to equally coalescence and Ostwald ripening to nearly purely Ostwald ripening. Previous studies of the coarsening of PP copolymer systems were conducted in a way that the one-phase melt was artificially produced (for these normally multiphase systems) to establish a true ‘‘zero time’’ for the onset of late-stage coarsening (30). To ensure the homogeneity of the initial system the ex-reactor hiPP powder was dissolved in a solvent and precipitated by a nonsolvent to form a one-phase ‘‘molecular dispersion.’’ This artificially produced system was then stored in the quiescent, isothermal melt state for various times (5 s to 1 h) to observe the phase separation and subsequent coarsening processes. The system was subsequently quenched in ice water to produce the ‘‘frozen’’ liquid, which was assumed to be an unaltered record of the system in the melt state. The coarsening behavior of such systems would find application for the case of thick moldings of these multiphase polymers or for other unmixed melts for which coarsening of the EPR phase in the quiescent melt would play a significant role in the formation of the final two-phase morphology. Therefore, the study was done on the quiescent melt, although it was recognized that subsequent mechanical mixing of the system may have a profound effect on the phase morphology. For the typical commercial polypropylene copolymer systems the viscosity of the matrix phase is quite high, and the molecular diffusion and solubility of the minor phase component in the matrix phase are relatively high. These factors tend to favor the evaporation/condensation, that is, Ostwald ripening, mechanism and suppress the coalescence mechanism in these systems. Mirabella and coworkers studied a series of multiphase systems, including a hiPP (30), a high density polyethylene (HDPE)/ hydrogenated polybutadiene (HPB) blend, (31) and an unbranched PE molecular

Chapter 12 Multiphase Polypropylene Copolymer Blends

365

weight fraction/HPB blend (32), in order to compare the theoretical predictions of coarsening to the experimental observations.

12.4.2 Coarsening of High Impact Polypropylene The coarsening of the EPR-rich phase in the hiPP is shown in scanning electron micrographs in Fig. 12.8. Coarsening was monitored for the ‘‘molecular disper-

Figure 12.8 Scanning electron microscopy photomicrographs of quiescently coarsened and ice water quenched specimens of impact polypropylene copolymer (hiPP). Magnifications are given above each photomicrograph. (From Reference 28 with permission from John Wiley & Sons, Inc.)

366

Polyolefin Blends

sion’’ for storage times in the melt of 5 s to 60 min at 193 C. Each scanning electron micrograph in Fig. 12.8 corresponds to a different specimen of the same material stored for the specified time. The commercial ex-reactor hiPP powder that was studied (molecular structure provided in Reference 28), of course, contained iPP and EPR components with broad molecular weight distribution and in the case of the EPR broad composition distribution. Therefore, the calculation of theoretical rate constants for coalescence and Ostwald ripening was not possible, without extreme assumptions. However, this system exhibited the theoretically predicted temporal dependence of coarsening, that is, r t1=3 , as can be observed in Fig. 12.9. Figure 12.9 shows the weight-average particle diameters versus time of storage in the melt from the data as in Fig. 12.8. The slope of the linear least squares fitted line in Fig. 12.9 is 0.32 and the coefficient of determination is R2 ¼ 0:97. The predictions of the coarsening theory that are embodied in Equation 12.3 are as follows. The temporal exponent for the rate of radial growth of the particle is 1/3. Therefore, the radius should vary with t1=3 . It is assumed that the total volume (VTot ) of the coarsening phase is invariant with time after the first-order transition (i.e., phase segregation). Therefore, the number of particles is proportional to VTot =r 3 ðtÞ. According to Equation 12.3 then, the number of particles per unit volume, N, is proportional to t1 . The number of particles per unit area NA (as observed in scanning electron photomicrographs) is related to the particles per unit volume N as follows: NA ¼ Ndn

ð12:7Þ

The number of particles per unit volume N as a function of storage time in the melt for the ice water quenched (from data as in Fig. 12.8), as well as another set of more

0.1 0 –0.1 –0.2

log dw

–0.3 –0.4 –0.5 –0.6 –0.7 –0.8 –0.9 0

0.5

1

1.5

2

log t

2.5

3

3.5

4

Figure 12.9 The weight-average dispersed phase particle diameter (dw ; mm) versus time in the melt (t s) for quenched specimens of the hiPP. The line is the linear least squares fitted to the data. (From Reference 28 with permission from John Wiley & Sons, Inc.)

Chapter 12 Multiphase Polypropylene Copolymer Blends

367

3.5

log NA/dn

3

2.5

2

1.5

1 0.5

1

1.5

2

2.5

3

3.5

4

log t

Figure 12.10 The number of particles per unit volume N (where N ¼ NA =dn ) versus time in the melt (t s) for the (O) quenched and (&) bench-top cooled specimens. (From Reference 28 with permission from John Wiley & Sons, Inc.)

slowly cooled (bench-cooled) samples, is shown in Fig. 12.10, where NA =dn is plotted versus the time in the melt for these specimens (30). The initial slope of this plot is 0:16 and the slope at longer time is 1:02 (R2 ¼ 0:96). This appears to indicate that the expected N t1 behavior is not approached until longer time (100 s). It has been observed in this and other phase-segregated systems that the shortest time specimens (5 s in the melt) consistently exhibited larger particle diameters than expected from the t1=3 dependence. It may therefore be inferred that specimens that spend only 5 s in the melt have not yet entered the longtime coarsening regime. These short times may actually be in the short-time regime, that is, before coarsening, which is not well understood (33,34). The short-time regime may be complicated by simultaneous occurrence of the demixing process and the coarsening process in this time regime (21). The coarsening process is predicted to exhibit a self-similar nature at long time (33). That is, the particle size distribution is predicted to remain constant in the longtime regime. The particle size data for the ice water quenched specimens exhibited a trend of broadening particle size distribution from very narrow distributions at short time to slightly broader distributions at long time. The corresponding particle size distributions (PSDs) for the ice water quenched specimens are presented in Fig. 12.11 (30). The PSDs are plotted as reduced chord length versus frequency, where reduced chord length is the chord length (Li ) divided by the number-average chord length (Ln ). A trend of broadening PSD with increased time in the melt is clearly observed and may be due to the shift from the short-time regime to the longtime coarsening regime, as shown schematically in Fig. 12.7. Little is known about the nucleation and growth process and the PSD that is produced in the shorttime regime (33, 34). Further, the shift from the short-time regime to the asymptotic longtime regime, and its effect on PSD, has been little studied (33).

368

Polyolefin Blends

Figure 12.11 The particle size distributions for the quenched specimens plotted as frequency ½ f ðNÞ versus the reduced chord length ðLi =Ln ) for (—)5 s, () 45 s, (- - -) 2 min, (&) 8 min, and (o) 60 min storage time in the quiescent melt. (From Reference 28 with permission from John Wiley & Sons, Inc.)

12.4.3 Coarsening of Model Blends The coarsening of a model blend consisting of a hydrogenated poly(1,4-butadiene) with Mn ¼ 93,000,Mw ¼ 111,000, and Mw =Mn ¼ 1:2 and 100 ethyl branches/1000 total carbon atoms (equivalent to a poly(ethylene-butene-1) copolymer with mole fraction butene-1 of 0.205) and a commercial HDPE with Mn ¼ 22,000, Mw ¼ 159,000, and Mw =Mn ¼ 7:2 and 0 branches/1000 total carbon atoms was studied (31). As in the previous study (30), the one-phase melt was artificially produced (for this normally two-phase system) to establish a true ‘‘zero time’’ for the onset of latestage coarsening and this system exhibited the theoretically predicted temporal dependence of coarsening, that is, r t1=3 . Although this system was somewhat idealized, relative to the hiPP, it is still a multicomponent system, due to the broad molecular weight distribution of the HDPE. Using some approximations (described in Reference 29), the theoretical rate constant for Ostwald ripening was calculated from Equation 12.4 and was KOR ¼ 3:6 1018 cm3 s1 . The experimentally determined rate constant (K in Equation 12.3) was Kexp ¼ 4:8 1018 cm3 s1 . The agreement between the experimentally measured and the theoretically calculated rate constants is quite good. The ratio of these two rate constants is Kexp =KOR ¼ 1:3. This was taken as an indication that the coarsening occurred by Ostwald ripening, although the theoretical constant for coalescence was not calculated. However, the agreement may be somewhat fortuitous due to the uncertainty in the parameters used to calculate the theoretical rate constant. Nonetheless, the good agreement between the theoretical and experimental values strongly suggests that the theory is particularly applicable to this polymer blend system.

Chapter 12 Multiphase Polypropylene Copolymer Blends

369

Each of the previous two systems (30,31) was polydisperse in at least one component, thereby causing the systems to be multicomponent as opposed to binary. The fact that these systems were not simple binary systems made the comparison of the experimental data to the theory more difficult and uncertain. Notwithstanding this fact, the agreement between experiment and theory was remarkably good. In another study (32), a binary blend was formed using an HDPE fraction, containing 0 branching (NIST SRM 1484) and Mn ¼ 100,500, Mw ¼ 119,600, and an HPB containing 100 ethyl branches per 1000 C atoms and Mn ¼ 96,000, Mw ¼ 111,000. Again, the one-phase melt was artificially produced (for this normally two-phase system) to establish a true ‘‘zero time’’ for the onset of late-stage coarsening and this system exhibited the theoretically predicted temporal dependence of coarsening, that is r t1=3 . This system was assumed to be a binary blend, since molecular weight and composition distributions were very narrow. All constants (defined in Section 12.4.1) for this system are presented in Table 12.4. Also presented in Table 12.4 are the theoretical rate constants for Ostwald ripening and coalescence KOR ¼ 0:86 1018 cm3 s1 and KC ¼ 3:6 1020 cm3 s1 , respectively, from calculations using Equations 12.4–12.6. It may be observed that these theoretical rate constants (KOR and KC ) differ by about two orders of magnitude. The experimentally determined rate constant was Kexp ¼ 1:23 1018 cm3 s1 and was in fairly good agreement with that for Ostwald ripening. The ratio of these two rate constants is Kexp =KOR ¼ 1:4, which is similar to that obtained in the previous study (31). Therefore, this result indicates that this binary system coarsens by Ostwald ripening and that coalescence is negligible. The results for the above, binary system, lent credence to the conclusion that the coarsening in all these polymer–polymer blends (30–32) with relatively high melt viscosity (i.e., high molecular weight) matrices was due to Ostwald ripening. This was also supported by the work of Crist and Nesarikar (29), which showed that

Table 12.4 Parameters for Calculation of Theoretical Rate Constants for Coarsening by Ostwald Ripening (Evaporation/Condensation) and Coalescence for HDPE (Fraction)/HPB Blenda. Ostwald ripening (evaporation/condensation) D cm2 s1 1:91 1011

Vm (cm3 mol1)

g (erg cm2 )

Cea (mol fraction)

T (K)

KOR (cm3 s1)

1:44 105

1.20

0.011

450

0:86 1018

Coalescence DBr (cm2 s1 ) 2:00 1013 a

h (poise)

f (vol fraction)

T (K)

KC ðcm3 s1 Þ

1:10 105

0.10

450

3:60 1020

From Reference 30 with permission from John Wiley & Sons, Inc.

370

Polyolefin Blends

polymer systems with high molecular weight matrices coarsen mainly by Ostwald ripening. Further, the excellent agreement of the experimental rate constants with that for Ostwald ripening supported this conclusion. Ratios of Kexp =KOR ¼ 1:3 and 1.4 appear to strongly support Ostwald ripening. This author considers the case to be decisive. However, alternate views do exist. Fortelny and coworkers argued that the theory of coalescence of Smoluchowski (24) is based on dilute systems, derived a theory of coalescence for concentrated systems (35), and analyzed literature data according to their derived theory (36). They argued that phase-separated polymer systems in the literature, such as discussed above, do not satisfy the criteria of the Smoluchowski theory and that for systems with greater than 10% of the droplet phase the average distance between particles is substantially smaller than the droplet radius. They described a coalescence mechanism for such systems in which the forces between neighboring droplets were comparable to those from Brownian motion, leading to drainage of the matrix film between particles followed by droplet coalescence. Literature data were claimed to fit their theory of coalescence for concentrated systems better than other theories. The main thrust of the work of Fortelny and coworkers is toward an alternate mechanism to Ostwald ripening, that is, the mechanism of coalescence of neighboring droplets described above. No new experimental evidence was provided by Fortelny and coworkers to support the alternate mechanism they proposed.

12.4.4 Interfacial Effects in Polypropylene Copolymer Systems Interfacial effects on multiphase polymer systems have been of interest to polymer scientists. For example, Koberstein and coworkers demonstrated the compatibilizing effects of block copolymers in ternary blends of polystyrene/polybutadiene/ poly(styrene-block-butadiene) (37). These workers showed that the block copolymer produced a sharp decrease in interfacial tension. Torkelson and coworkers applied this technology to HDPE/polystyrene blends compatibilized with styrene/ethylene– butylene/styrene (SEBS) triblock copolymers blended in an intensive solidstate shear pulverization process (38). They showed that the SEBS acted as a compatibilizer, manifested by a decreased quiescent coarsening rate: 3.5, 5, and 10 wt% SEBS resulted in a 10-fold and 30-fold decrease in rate, and cessation of coarsening, respectively. Similar results were obtained by Cavanaugh and Nauman for polystyrene/polybutadiene and polystyrene/polyisoprene blends compatibilized with polystyrene/poly(styrene-random-butadiene) diblock copolymer (39). These are examples of marked deceleration of coarsening or true ‘‘pinning’’ of the dispersed phase droplets against coarsening by the use of interfacial agents. The application of this strategy to commercially important polyolefin multiphase systems to produce ‘‘tuned’’ and/or stabilized morphologies is certainly attractive. However, the required interfacial agents for this purpose have

Chapter 12 Multiphase Polypropylene Copolymer Blends

371

historically been unavailable or prohibitively expensive. Recently, this situation may have changed due to the development of polyolefin block copolymers by Dow Chemical Co. (40). It remains to be demonstrated that such olefin block copolymers (OBC), which Dow has introduced commercially as Infuse1 OBCs (41), may be effective in controlling the morphology of related multiphase polyolefin systems. The formation, optimization, and commercialization of polypropylene nanocomposites is presently a field of active development (42). A recent study of TPO

Figure 12.12 AFM phase images of (a) TPO-0 (0 wt% clay), (b) TPO-1 (0.6 wt% clay), (c) TPO-3 (2.3 wt% clay), (d) TPO-4 (3.3 wt% clay), and (e) TPO-6 (5.6 wt% clay). (From Reference 42 with permission from John Wiley & Sons, Inc.)

372

Polyolefin Blends

Figure 12.13 EPR number-average particle diameter versus weight percent clay in the TPO. (From Reference 42 with permission from John Wiley & Sons, Inc.)

organoclay nanocomposites by Mirabella and coworkers showed that the EPR particle morphology in the TPO exhibited a progressive breakup and decrease in particle size, as clay loading increased in the range from 0.6 to 5.6 wt% clay (43). The TPO was composed of about 70 wt% of iPP and about 30 wt% of EPR. The nanocomposites were prepared by blending the TPO with maleic anhydride (MA)grafted iPP with 1.0 wt% MA concentration, as a compatibilizer and Cloisite1 20A natural montmorillonite clay modified with a quaternary ammonium salt. Figure 12.12 shows AFM phase contrast micrographs of the EPR particle morphology as clay loading increased in the range from 0.6 to 5.6 wt% clay. The EPR particle size decreased with clay loading as shown in Fig. 12.13. The breakup of the EPR particles was suspected to be due to the increasing melt viscosity, which was observed as clay loading increased, and/or the interfacial activity of the accompanying chemical modifiers on the clay. Figure 12.14 shows transmission electron microscopy (TEM) micrographs of the same system as in Fig. 12.12. It may be observed in Fig. 12.14 that the EPR domains (elliptically shaped) are surrounded by clay platelets (dark rodlike structures) and that the clay platelets preferentially segregate to the EPR/iPP interface. This was taken as evidence for the proposal that the interfacial activity of the accompanying chemical modifiers on the clay reduces the interfacial tension with concomitant reduction in particle size. A study of a similar TPO/iPP-MA/organoclay system was reported in which AFM and TEM images of the EPR particle morphology revealed a systematic reduction in the EPR particles as clay loading increased (44). Further, these workers observed an increase in impact strength of the TPO nanocomposites with increasing clay loading, which is generally the opposite trend as that expected for such nanocomposites. They explained this unexpected behavior in terms of the reduction in particle size of the EPR elastomer. They suggested that the reduction in particle size was due to some

Chapter 12 Multiphase Polypropylene Copolymer Blends

373

Figure 12.14 Transmission electron micrographs (varying magnification as indicated by scale bars) of (a) TPO-0 (0 wt% clay), (b) TPO-1 (0.6 wt% clay), (c) TPO-3 (2.3 wt% clay), and (d) TPO-6 (5.6 wt% clay). EPR-rich rubbery domains (elliptically shaped) surrounded by clay platelets (dark rodlike structures) are clearly seen. (From Reference 42 with permission from John Wiley & Sons, Inc.)

combination of the melt viscosity increase and the interfacial effects of the clay, as clay loading increased.

12.5

CONCLUSIONS

Polypropylene copolymers are commercially produced for applications especially requiring high impact strength. However, many other properties are offered by

374

Polyolefin Blends

such resins. The typical processes involve the production of iPP in the first reactor, followed by the production of EPR in the second reactor to yield hiPP and TPOs. The dispersion of the rubbery EPR component especially serves to produce high impact strength, while high stiffness is contributed by the iPP matrix. These copolymers offer a balance of many other desirable physical and mechanical properties. These PP copolymers are primarily used in injectionmolded parts for automotive, appliances, and other durable goods applications, as well as for extruded sheet and thermoforming. The wide range of physical and mechanical properties, relative ease of processing, and low density constitute these polypropylene copolymers as extremely attractive materials capable of competing with more expensive plastics in many demanding applications. The automotive industry has made TPOs the primary choice for an increasing range of interior and exterior applications. These TPO resins continue to replace other polymeric compositions in interior and exterior applications, due to their desirable balance of properties and safety attributes. Interior applications include instrument panels, consoles, door panels, and pillars. Exterior applications include bumpers, fascia, body side cladding, rocker panels, and cowl vent grilles. The reactor grades of impact polypropylene copolymers are often compounded with other components, especially other toughening agents to extend their applications. A characteristic of hiPP and its compounds is that these are never miscible systems that undergo phase separation. The morphology of the polymer blends is multiphasic from the start. The morphology of the reactor product, typically powder, is extremely heterogeneous with some regions containing no observable EPR and other region containing large ‘‘pools’’ of EPR in the rigid iPP matrix. The reactor product with very heterogeneous morphology is typically converted to pellets for shipment with homogeneous morphology due to the great efficiency of commercialscale corotating twin-screw pelletization extruders. The heterogeneous reactor product undergoes homogenization by droplet breakup and coalescence in the pelletization extruder. This droplet breakup and coalescence occurs in the molten state of the viscoelastic iPP and EPR, matrix and dispersed phases, in the extruder under a complex strain field, which is a combination of nonuniform, transient shear and elongational fields. Further, a variable temperature profile is used along the barrel of the extruder causing complex variation in the viscoelastic properties of these components. The dispersive mixing theory covering droplet breakup and coalescence is generally a modified one, based on the case of a Newtonian drop suspended in a Newtonian matrix, that is, Taylor’s theory. Observed average particle size was 100 times that calculated from the dispersive mixing theory for an hiPP containing 16 wt% EPR. The larger particle size observed compared to that calculated is due to the effects of flow-induced coalescence. At practical concentrations of the dispersed phase of 10–20% in hiPP and TPO resins, observed particle size reported is 10–100 times above that predicted by theory. The very crude nature of the Taylor limit predictions is evident. Experimental determination of particle morphology is necessary in dispersively mixed systems.

Chapter 12 Multiphase Polypropylene Copolymer Blends

375

The structure of a commercial impact PP was determined by P-TREF, followed by analysis of the fractions by NMR in order to determine chemical composition and tacticity. It was found that the hiPP (MFR ¼ 6) was composed of about 75 wt% iPP with average tacticity of about 97% and exhibiting a broad tacticity distribution extending to very low values, 17 wt% EPR containing an average of about 48 wt% ethylene and exhibiting a broad composition distribution and about 8 wt% of an ethylene-rich copolymer containing 92 wt% ethylene. The hiPP had Mw ﬃ 380,000 and Mw =Mn ¼ 6:3, the iPP had Mw ﬃ 382,000 and Mw =Mn ¼ 4:3, and the EPR had Mw ﬃ 342,000 and Mw =Mn ¼ 14:2. Although, phase separation from a one-phase melt was not typically relevant to the polypropylene copolymer systems, coarsening of the dispersed phase particles is a potentially important effect in systems in which the two-phase melt experiences, at least partly, a quiescent state. The coarsening of the phase-separated system can occur by two mechanisms. Coalescence, by particle diffusion and collision, is one mechanism, while evaporation and condensation, called Ostwald ripening, is another. It was shown that these hiPP systems exhibited the theoretically predicted temporal dependence of coarsening, that is, r t1=3 , which holds for coalescence or Ostwald ripening. Coarsening of model blends exhibited the theoretically predicted temporal dependence, that is, r t1=3 , and theoretical rate constants for Ostwald ripening and coalescence were KOR ¼ 0:86 1018 cm3 s1 and KC ¼ 3:6 1020 cm3 s1 , respectively. The experimentally determined rate constant was Kexp ¼ 1:23 1018 cm3 s1 and was in fairly good agreement with that for Ostwald ripening. The ratios of the experimental to the theoretical rate constants were found to be Kexp =KOR ¼ 1:3–1.4 for two model blend systems and appear to strongly support Ostwald ripening as the coarsening mechanism. Interfacial effects on multiphase polymer systems have been of interest to polymer scientists. Application of this strategy to commercially important polyolefin multiphase systems to produce ‘‘tuned’’ and/or stabilized morphologies is certainly attractive. Recently, polyolefin block copolymers have been introduced commercially and may be effective in controlling the morphology of related multiphase polyolefin systems, such as hiPP and TPO. For the case of TPO/organoclay nanocomposites, which are attracting wide commercial interest, it was shown that the EPR particle morphology in the TPO exhibited a progressive breakup and decrease in particle size, as clay loading increased. The breakup of the EPR particles was suspected to be due to the increasing melt viscosity, which was observed as clay loading increased, and/or the interfacial activity of the accompanying chemical modifiers on the clay. The development of polypropylene copolymer multiphase systems is continuing at a robust pace. These polymer systems, based on simple and inexpensive polymer building blocks, are being improved for applications historically reserved for more expensive engineering thermoplastics. The understanding of the structure/property relationships in polypropylene copolymers will indeed be a driver for further innovation in the commercial application of these polymers.

376

Polyolefin Blends

NOMENCLATURE AFM an Cea Cloisite1 D DBr dn dw EPDM EPR hd hm G g HDPE hiPP HPB Infuse1 iPP Kc KOR Kraton Li MA MFR Mn Mw Mw/Mn N NA PP PSD f R rð0Þ rðtÞ SEBS T t TEM Tg TPE TPO

Atomic force microscopy Number-average particle diameter Equilibrium concentration (mole fraction) of the minor phase Montmorillonite clay modified with a quaternary ammonium salt Molecular diffusion coefficient Droplet diffusion coefficient due to Brownian motion Number-average particle diameter Weight-average particle diameter Ethylene–propylene diene monomer Ethylene–propylene rubber Melt viscosity of the dispersed phase Melt viscosity of the matrix phase Effective shear rate Interfacial tension High density polyethylene High impact polypropylene Hydrogenated polybutadiene OBC olefin block copolymer Isotactic polypropylene Coalescence rate constant Ostwald ripening rate constant Styrenic block copolymers Chord length Maleic anhydride Melt flow rate Number-average molecular weight Weight-average molecular weight Molecular weight polydispersity Number of particles per unit volume Number of particles per unit area Polypropylene Particle size distribution Volume fraction of the minor (droplet) phase Gas constant Particle radius at time t ¼ 0 Particle radius at time t Styrene/ethylene–butylene/styrene Absolute temperature Time Transmission electron microscopy Glass transition temperature Thermoplastic elastomer Thermoplastic olefin

Chapter 12 Multiphase Polypropylene Copolymer Blends

TPV Vm VTot

377

Thermoplastic vulcanized elastomer Molar volume of the minor component Total volume of the coarsening phase

REFERENCES 1. N. F. Brockmeier, Gas-phase polymerization, in: Encyclopedia of Polymer Science and Engineering, Vol. 7, H. Mark, N. Bikales, C. Overberger, and G. Minges (eds.), 1987, p. 480. 2. G. Schweier, Poly. Div., Abstracts, 214th ACS National Meeting, Las Vegas, NV, 1997, p. 450. 3. T. Simonazzi, G. Cecchin, and S. Mazzullo, Prog. Polym. Sci., 16, 303 (1991). 4. D. R. Paul and S. Newman (eds.), Polymer Blends, Academic Press, New York, 1978. 5. F. M. Mirabella, Polym. News, 30, 1 (2005). 6. P. Galli, T. Simonazzi, and D. Del Duca, Acta Polym., 39, 81–90 (1988). 7. I. Urdampilleta, A. Gonzalez, J. J. Iruin, J. C. de la Cal, and J. M. Asua, Ind. Eng. Chem. Res., 45, 4178 (2006). 8. Y. Chen, Y. Chen, W. Chen, and D. Yang, Polymer, 47, 6808 (2006). 9. S. Middleman, Fundamentals of Polymer Processing, McGraw-Hill, New York, 1977. 10. F. M. Mirabella, Polymer, 34, 1729 (1993). 11. F. M. Mirabella, Structure and thermodynamic aspects of phase segregation of commercial impact polypropylene copolymers, in: New Advances in Polyolefins, T. C. Chung (ed.), Plenum Press, New York, 1993. 12. J. J. Strebel, F. M. Mirabella, C. Blythe, and T. Pham, Polym. Mater. Sci. Eng., 44, 1588 (2004). 13. S. Wu, Polym. Eng. Sci., 27, 335 (1987). 14. H. Shariatpanahi, H. Nazokdast, B. Dabir, K. Sadaghiani, M. Hemmati, J. Appl. Polym. Sci., 86, 3148 (2002). 15. V. E. Ziegler and B. A. Wolf, Macromolecules, 38, 5826 (2005). 16. G. I. Taylor, Proc. R. Soc. (London) A, 138, 41 (1932). 17. G. I. Taylor, Proc. R. Soc. (London) A, 146, 501 (1934). 18. U. Sundararaj and C.W. Macoscko, Macromolecules, 28, 2647 (1995). 19. J. J. Elmendorp and A. K. Van Der Vegt, Polym. Eng. Sci., 26, 1332 (1986). 20. F. M. Mirabella, J. Polym. Sci.: Appl. Polym. Symp. 51, 117 (1992). 21. T. Hashimoto, Phase Transitions, 12, 47 (1988). 22. W. Ostwald, Z. Phys. Chem., 37, 85 (1901). 23. I. M. Lifshitz and V. V. Slyozov, J. Phys. Chem. Solids, 19, 35 (1961). 24. M. von Smoluchowski, Phys. Z., 17, 557, 585 (1916). 25. V. G. Levich, Physicochemical Hydrodynamics, Prentice-Hall, Englewood Cliffs, NJ, 1962, pp. 207–211. 26. E. D. Siggia, Phys. Rev. A, 20, 595 (1979). 27. T. Hashimoto, M. Takenaka, and T. J. Izumitani, Chem. Phys., 97, 679 (1992). 28. B. Crist, Macromolecules, 29, 7276 (1996). 29. B. Crist and A.R. Nesarikar, Macromolecules, 28, 890 (1995). 30. F. M. Mirabella, J. Polym. Sci. B Polym. Phys., 32, 1205 (1994). 31. F. M. Mirabella and J. S. Barley, J. Polym. Sci. B Polym. Phys., 32, 2187 (1994). 32. F. M. Mirabella, F. M., Barley, and J. S., J. Polym. Sci. B Polym. Phys., 33, 2281 (1995). 33. P. W. Voorhees, J. Stat. Phys., 38, 231 (1985).

378

Polyolefin Blends

34. G. Venzl, Ber. Bunsenges. Phys. Chem., 87, 318 (1983). 35. I. Fortelny and A. Zivny, Polymer, 21, 4113 (1995). 36. I. Fortelny, A. Zivny, and J. J. Juza, Polym. Sci. B Polym. Phys., 37, 181 (1999). 37. S. H. Anastasiadis, I. Gancarz, and J. T. Koberstein, Macromolecules, 22, 1449 (1989). 38. Y. Tao, A. H. Lebovitz, and J. M. Torkelson, Polymer, 46, 4753 (2005). 39. T. J. Cavanaugh and E. B. Nauman, Polym. Sci. B Polym. Phys., 36, 2191 (1998). 40. D. J. Arriola, E. M. Carnahan, P. D. Hustad, R. L. Kuhlman, and T.T. Wenzel, Science, 312, 714 (2006). 41. http://news.dow.com/prodbus/2006/20060620a.htm. 42. F. M. Mirabella, Polypropylene and thermoplastic olefin nanocomposites, in: Encyclopedia of Nanoscience and Nanotechnology, J. A. Schwarz, C. I. Contescu, and K. Putyera (eds.), Marcel Dekker, New York, 2004. 43. S. Mehta, F.M. Mirabella, K. Rufener, and A. Bafna, Appl. Polym. Sci., 92, 928 (2004). 44. H. Lee, P. D. Fasulo, W. R. Rodgers, and D. R. Paul, Polymer, 46, 11673 (2005).

Chapter

13

Heterogeneous Materials Based on Polypropylene Jesu´s Marı´a Garcı´a Martı´nez1, Susana Areso Capdepo´n1, Jesu´s Taranco Gonza´lez1, and Emilia Pe´rez Collar1

13.1 INTRODUCTION The long-range elasticity, high strength, and high viscosity, all of the self-defining macromolecular states are deeply influenced by the intermolecular forces—a direct consequence of the size and constitution of the covalent structures of macromolecules. For the thermoplastic polymers family polypropylene belongs to, the great number of atoms involved in the primary intrachain bonds above a given critical value—that is, mean size or molecular weight—are able to induce interchain secondary interactions at such a strong level that the matter becomes a material, with enough structural integrity to be useful. Since the beginning, organic polymers have been combined with another substances, mainly inorganic, to lower costs. However, the reinforcement effect in the polymer-based material caused by the presence of a second component was soon ascertained (1). In mid twentieth century, it was surmised that the contact regions between both components, that is the interphases, play a crucial role in almost all the heterogeneous materials. Numerous efforts, mainly in the development of light-weight structural applications based on thermosetting polymers, were devoted to improve these interfacial regions. Since then began a 20th second revolution on the basis of heterogeneous materials, after the so-called industrial revolution. Once their production

1 Department of Physics and Engineering of Polymers, Polymer Engineering Group, Institute of Science and Technology of Polymers, CSIC, CL Juan de la Cierva 3, 28006 Madrid, Spain

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

379

380

Polyolefin Blends

processes were consolidated on an industrial scale, organic-based materials have been called to be the key to the new era. After the pioneering work by Helfand and Tagami (2), the basis to the study of interphases from a perspective of the organic matter-based materials and to control the properties of any given multicomponent systems, was established. So the option of designing with materials began to be surpassed by the option of designing multiphase materials, with at least one organic polymer obtained in a polymerization reactor being one of their components (3).

13.2 THE INTERPHASE: DEFINITION From the early 1980s, it was well accepted that the desired performance of a polymer-based heterogeneous material would pass through a well-optimized interaction level across the interphases between the components (4). The interphase is defined as the dynamic and finite spatial region placed between the border of each of the two different phases where momentum, mass, and energy transport phenomena may occur. It follows from this definition that transport phenomena across the interphases are governed by the morphology and even topography of the two phases, with dimensions of the nanoscopic scale, in such a way that atoms or small groups of atoms located on the border surfaces determine the flows of momentum, mass, and/or energy balances between both phases. Because only in a few heterogeneous systems (mainly in the highly selective bioactive systems) is it possible to measure such flows, given the high complexity inherent in the macromolecular systems, the transport phenomena or the changes in any given property across the interphases can be evaluated by considering the interface model as displayed in Fig. 13.1. According to this figure, the interface would be obtained by projecting all volumes between both phases sensitive to the desired property onto a single and finite surface defined by a critical thickness, where any measured property would exhibit the sharpest possible change from one phase to another. So, different critical thicknesses for different properties or response functions can be expected.

13.3 MAGNITUDE ORDERS IN THE INTERPHASE For some of the recent enthusiasts of nanochemistry, it would be convenient to remind that macromolecules of thermoplastic polymer materials remain together due to the high intensity of the secondary interactions derived from the very high number of atoms present in their polymer backbones. It is important to note that the primary bonds located on the main chain typically require energy of bonding values from 400 to 800 kJ mol1. Meanwhile, the order of magnitude for the secondary forces ranges from 4 to 40 kJ mol1. The latter means distances of 0.3 and 0.4 nm between the atoms involved in the secondary forces, that is, three times or more larger than the 0.1

Chapter 13 Heterogeneous Materials Based on Polypropylene

381

Figure 13.1 The interface model.

to 0.15 nm existing between the atoms that form a primary bond. The reach and intensities of these intermolecular forces acutely depend on temperature and time, in a much more sensitive way than in any other classical materials, for example, metals, concrete, or ceramics. In consequence, several assumptions valid to explain the solidstate behavior of these materials can hardly be accepted in the case of organic thermoplastic materials. Because both the softening as well as the molten state of a polymer require the strain phenomenon to occur before the flow takes place, it is logical that to induce flow in any organic macromolecular system not only temperature but also pressure is required. The coupled viscous dissipation due to the self-friction of the macromolecular chain segments hardly may be neglected when undertaking any study. Even it may occur that while the strengths of those secondary forces in the macromolecular systems would be still strong enough to maintain the macromolecules together, some primary bonds would begin to break in a local adverse environment. Figure 13.2 compiles the main variables and parameters defining the polymerbased heterogeneous materials from the macro- to nanoscales, which determine the performance of the material as a whole.

382

Polyolefin Blends

Order of magnitude in multiphase materials

Parameters

Structure assembling Length at molding stages Width at molding stages Length Width Fiber length in finished parts Fiber length at molding stages Fiber diameter Surface roughness Crystal aggregates Polymer crystals Chain length Macromolecular diameter –12

–10

–8

–6

–4

–2

Twelve orders of magnitude Interfacial interactions

0

Multiphase material

Reinforcement and/or dispersed phase

Polymer

2 4 Log units, m Macroscopical properties

Figure 13.2 Orders of magnitude at the interphase studies on polymer-based heterogeneous materials.

13.3.1 The Dispersed Phase The most frequent disposal of most of the polymer-based heterogeneous materials family takes place by the dispersed phase/matrix mode. So the dispersed phase components may be identified by the finite size of each of their domains, being surrounded by the continuous matrix. Both the size and the geometry of the particles featuring the dispersed phase together with their surface properties govern the transport phenomenon across the interphase between the dispersed particles and the continuous matrix. According to the interface approach defined in the previous section, it is obvious that the domain size and its distribution confine the interfacial volume available for effective transport flows between the matrix and the disperse phase. To control this, parameters of the highest relevance for studies of interfacial phenomena have been set, so that particle size and distribution remain constant, or almost constant, all along the experimental work. When the dispersed phase is constituted by rigid particles (5), the breakdown caused by external stresses may occur, yielding significant changes in both the mean particle size and the size distribution. However, if reinforcing particles do not break down during processing steps, these variables remain under control all along the experimental work. Hence, interfacial modifications are able to change the breakdown behavior of the solid particles dragged by the matrix flowing streams, for example, by lubrication, and introduce an additional experimental variability by way of

Chapter 13 Heterogeneous Materials Based on Polypropylene

383

noncontrolled effects at the different stages of the experimental runs. Consequently, misleading or wrong conclusions on the reach of the interfacial changes may be obtained from such works. Otherwise, in the case of a deformable dispersed phase, that is, polyblends, when the polymer matrix flows, it occurs that the interfacial volume available for the transport flows between both phases could change in a way much more complex than when the dispersed particles are not deformable, not only in terms of changes in mean size and size distribution all along the processing steps but also in shape and preferential geometry of the dispersed phase domains. These changes would evolve as the primary function of the viscosity ratio between the components, and of the surface tension balance when interfacial modifiers are present (6). Even, if primary bonds connecting dispersed and continuous phases are formed through the interphase in whatever processing step, they can generate stable morphologies that remain intact even after other postreactive processing steps, yielding a controlled distribution of the dispersed phase into the matrix (7).

13.3.2 The Matrix The design criteria based on a well-defined performance of a material, with emphasis on its mechanical behavior, have greatly increased the application spectrum of the organic polymeric materials. The application varies from the neat thermosetting matrices for engineering devices to both the development of engineering thermoplastics and the composites based on them incorporating discrete reinforcements, where the polymer matrix properties have been designed to satisfy specific requirements (1,3,8). Not only have mechanical performance considerations and other specific properties of the material led to this situation, but also the economic benefits derived from faster processing operations for thermoplastics compared to those for thermosetting polymers, as well as environmental considerations dealing with easier recycling possibilities of thermoplastic matrices. This has emerged as the driving force for a change in designing criteria. Figure 13.3 compiles the main variables and parameters that define a semicrystalline thermoplastic polymer matrix that, in combination with the processing steps, optimize thermophysical and mechanical behavior of the material to be useful for any purpose. This is the well-known structure–processing–properties dynamic triangle.

13.3.3 The Interphase: Designing the Interface From the previous sections it follows that when designing the interface, starting on the desired performance of the heterogeneous material at the macroscopical scale, of 12 magnitude orders, may be the dimensional gap required to obtain the optimum material (9). The classification of primary or secondary bonds as a power function of the distance between each pair of atoms involved for each bond and the way the

384

Polyolefin Blends Rheology, processing steps

Molecular weight

Intermolecular parameters: chain volume diameter and length (end to end distance)

Intramolecular parameters: interaction parameters (space/volume ratio)

Semicrystalline polymers: ❖ Amorphous/crystal ratio ❖ Unit cell ❖ Lamellar thickness distribution ❖ Spaced and long spacing ❖ Crystalline aggregates

Tortuosity parameter (Length/2*radius of gyration) Matrix morphology Thermophysical: ❖ Glass transition temerature ❖ Melting temperatures ❖ Melting and crystallization heats ❖ Specific heat ❖ Thermal conductivity ❖ Coefficient of thermal expansion ❖ Moisture absorption

Properties

Mechanical: ❖ Strength and module (tensile, flexural and shear modes) ❖ Poisson coefficient ❖ Impact behavior ❖ Hardness ❖ Friction coefficient

Others required by performance

Figure 13.3 Structure–processing–properties relationship for designing materials based on organic thermoplastic matrices.

intensities of these forces decrease with distance is of the highest relevance at the nanoscale level the interfacial phenomena occur (10). In particular for the mechanical responses across the interphase, it would be obvious that long distance forces let higher strain levels across the interphase than those from short distances, giving rise to brittle materials. Moreover, permanent dipole–dipole interactions yield attractive forces that decrease as the third power of the distance; meanwhile, induced dipole–dipole interactions yield attractive forces that decrease as the seventh power of the distance between the atoms involved. Dispersive forces due to electronic displacements mean attractive interactions decreasing as the sixth power of the distance. For an ideal amorphous polymer, the specific (referred to as chain segment) free volume should vary only with temperature. According to the basic principle of ideal elasticity, the internal energy of a system remains constant during the mechanical energy absorption/dissipation processes. From here it follows that the specific free volume for an ideal amorphous polymer would be an index of the available local space for the dissipation of the absorbed mechanical energy through the segmental rotational motions of the macromolecular chains (11). When the polymer has the potential to reach the three-dimensional order (12), that is, the crystalline state, the fracture behavior becomes more complex. It is well known that the distance between the macromolecular chain segments in the crystal are shorter than those located in the amorphous phase. Hence, when the material is required to dissipate any kind of energy, this would be preferably dissipated through the amorphous phase and, even more, through the zone where the chain segments are significantly constrained, as it is an amorphous/crystal interphase.

Chapter 13 Heterogeneous Materials Based on Polypropylene

385

Then, the latter forms the weakest region in the material as a whole, the region where the fracture phenomenon would first start. This phenomenon takes place at a so called nanoscale. Meanwhile, studies on bioactive species, ranging from living things to biochemistry as a concept, have been conducted to get a hierarchical range of responses, passing through the different organized systems such as organs, functional tissues, expertise cells and so on with their different inner devices based on macromolecules. In the study of materials, However there exists a deep lack of knowledge of the macroscopic performance of these materials as a whole and their relationship with their macro-, micro-, meso-, and nanoscale morphologies, each of them defined by the emerging property that determines their functionality at each scale (13). In the case of an organic thermoplastic polymer, it must be mentioned that the generated morphologies and the supramolecular organization possibilities of their molecular structures are developed under external fields, mainly thermal and stress/strain fields, acting during the processing steps. This appears to be very sensitive to any other external field acting over the polymer when becoming a final part. These unknown hierarchical response levels and the ways they are interconnected are undoubtedly the key when designing the interphase between the components of any organic polymer-based heterogeneous material. It is then obvious that establishing the above mentioned hierarchical range of the interconnection parameters between the macro-, micro-, meso-, and nanoscales is necessary before a great effort in the study for an efficient control of the ultimate properties of the material is made.

13.4 INTERFACIAL MODIFICATION OF HETEROGENEOUS MATERIALS BASED ON POLYPROPYLENE It is well accepted that the good properties of the isotactic polypropylene as an engineering polymer matrix in thermoplastic composite materials and engineering blends are seriously affected by the inability of this polymer to develop an adequate level of interfacial interaction with polar components such as mineral fillers (calcium carbonate) and reinforcements (talc, mica, wollastonite), synthetic reinforcements (glass fibers, carbon fibers, and nanotubes), or engineering polymers such as polyamide, aliphatic polyesters, and so on. In order to preserve the affinity with the nonpolar polypropylene matrix, the substitution of a small fraction of the matrix by another polypropylene with just a few polar groups grafted onto its backbone has been proven, in the last 20 years, to be a very effective strategy to promote isotactic polypropylene for engineering polymer applications. Among the different polar groups chosen to be grafted into polypropylene, maleic anhydride has proven to be of the highest efficiency as follows from the

386

Polyolefin Blends

reviews by Xu and Lin (14) (299 references), Jois and Harrison (15) (129 references), Naqui and Choudhari (16) (118 references), and Moad (17) (188 references). From the point of view of designing at the interface level, there is not yet available any mathematical model based on the physicochemistry of this spatial region for general application to develop any composite material based on organic matrices like polypropylene. At this point, it must be noted that the main goal, when building a mathematical model for materials design, is to transform a physical event connected to a real system into a mathematical structure that can be used to describe and understand such event. Furthermore, the model must be able to predict the magnitude of the changes in the system responses under new external stimuli, as well as to obtain the desired performance. Aris (18) defines a mathematical model as a system of equations formulated to express the laws of a prototype system, including in a much more general system and whose prototype is characterized by a set of specific features of main interest in the system under study. Hence, it follows that to build a model that is able to include the overall aspects of a real system while maintaining its accuracy and robustness is hardly possible. The complex nature of the organic macromolecular systems because of the nonlinear regime of responses of these materials appears as the best example of these considerations, especially when faced with classical materials such as ceramics, metals, and inorganics. The first critical step toward developing an efficient model begins with the identification of the key aspects to be considered and those to be discarded in the process, while looking for obtaining a model simple enough to be managed. However, discarding any essential feature of a real problem while looking for an easier resolution of an algorithm model would lead to forecasts of little use and sense. When building a model to predict the mechanical behavior of heterogeneous materials based on polypropylene, the basic principle of fracture mechanics should be considered that fracture is induced in a material when an energy threshold on the weakest zones is reached. It is also well accepted that such energy threshold consists of at least two components: necessary energy to initiate a craze and the necessary energy to propagate it. From the preceding sections it may be deduced that the generation of a craze would need much more energy in the case of a well balanced interphase than the unbalanced interphase. Hence, both craze growth and propagation across a much more continuous interphase are expected to be much faster than in a less continuous interphase. Indeed, the craze progress would be a critical factor for forecasting the failure of an initially strong interphase. Then not only would it be necessary to take into account the numerical values of the macroscopic mechanical properties, moduli, or strength, but also their hierarchical relationship with different mechanical energy dissipation possibilities at macro-, micro-, meso-, and nanoscales should be taken into consideration when building a general model to design the interface. They are necessary heuristic models in which any deficiencies once evidenced would be successively removed. These models would allow the researchers to build models

Chapter 13 Heterogeneous Materials Based on Polypropylene

387

that would enable control of the behavior of the organic polymer heterogeneous materials on the basis of their optimal balance of interactions through the interphases between their components.

13.4.1 Composites: When the Dispersed Phase is Rigid Mathematical models may be classified into deterministic and stochastic models. For deterministic models, knowledge of the relationship between dependent and independent variables is necessary. Consequently, the complex nature of polymer-based heterogeneous materials is rather incompatible with such requirements. Hence, stochastic models become necessary either when the existing knowledge about the stimulus-response behavior of a system is not enough as to ascertain its behavior or when it is not possible to build an efficient deterministic model able to score the system response. Data, collected by random experimental runs using probability and statistical models, may be a good way of approaching the responses of the system inside the experimental space scanned (19). Then, this would be a preliminary consideration when undertaking a model for a disperse phase/continuous matrix systems where the dispersed phase domains are rigid. Once assured that reinforcing particles do not change their size by breaking during the processing steps, the design of a surface treatment for the rigid particles would be twofold. On the one hand, it would tend to avoid any agglomeration of the solid particles and on the other, it would tend to improve their physicochemical interaction level with the polymer matrix. From the early 1980s, attempts were to avoid the negative effects of hydroxyl groups located on the surface of the inorganic particles by treating them with the socalled coupling agents such as silanes, titanates, and so on (20–23). The substitution of hydroxyl groups by chlorine or amine groups onto the lamellar talc surface, finally yielding polypropylene/talc composites with good mechanical performance, was studied by various groups (24–26).

13.4.2 Blends: When the Dispersed Phase is Flexible As it is well known, for most of the polymer/polymer pairs because of their low values of mixing entropy, a multiphase system is always generated. The dispersed phase morphology is the result of the balance between the applied shear forces and the counteracting interfacial tension forces. As a consequence, the long-range morphological regularity that may be observed on block copolymers, sometimes claimed to be model polymer blends, is hardly obtained by mixing two or more different polymers. Large shear rates enhance deformation capabilities of the dispersed phase domains generally as droplets, flowing with the matrix during the mixing and further

388

Polyolefin Blends

processing steps. Evolution of the morphology may be followed as a function of the mixing time and also the mixing conditions affecting drop dimensions and stability (6,7). These may be expressed by the Weber number given by We ¼

ðg hm Dn Þ ð2GÞ

ð13:1Þ

where g is the shear rate, hm is the matrix viscosity, Dn is the dispersed phase size, and G is the interfacial tension. From Equation 13.1, the existence of a critical value of the Weber number emerges, and then a critical particle size for the dispersed phase below which there is no particle deformation. Furthermore, if only physical interactions exist between components, the remaining dynamic coalescence mechanism could change on further processing steps such as a nonstable postprocessing morphology. Any model capable of predicting the emerging morphologies and their stability must take these elemental principles into consideration. The understanding of the relationship all along the 12 magnitude orders between the structural response levels at the macroscopic scale, and the driving forces acting at the interfacial scale also at the molten state, lead us to approach the hierarchical sequence of possibilities of matter organization from macro-, to micro-, to meso-, and to nanoscales at different windows of response, where each of them controls the behavior of the material either as a whole or when overlapped their effects (13).

13.4.3 The Role of the Interfacial Modifiers from the Matrix Side From above discussion, it becomes obvious that interfacial modifications may be driven either by the dispersed phase or from the matrix perspective, or even both. But driving forces in both cases must be well balanced to achieve the desired performance of the material at the macroscopic scale. Nevertheless, because of their affinity with the matrix, an increase in the flow across the interphase due to the presence of the interfacial modifiers from the matrix side is always observed. Besides, a decrease in the critical particle size for the polymer/polymer systems could also be obtained due to the decrease in the Weber number values, caused by an increase in the interfacial area available. A saturation level at the interaction plane across the interphase and in the concentration of the interfacial modifiers also emerges from the finite dimensions of the interphase. Above a critical concentration, the interfacial modifier could form its own phase, and then a nth phase ought to be considered in the studies rather than a model based on modified interfaces. Following sections show some examples of the role of the interfacial modifiers from the matrix side for both rigid and nondeformable dispersed phase polypropylene/mineral reinforcement system and polymer/polymer binary system based on polypropylene and polyamide 6.

Chapter 13 Heterogeneous Materials Based on Polypropylene

13.4.3.1

389

Polypropylene/Mineral Systems

Depending on the number of grafted moieties as well as on the place where these grafted groups are located all along the polymer backbone, and even in the chain ends, it is possible to obtain many different functionalized polyolefins as a function of the reaction conditions chosen to obtain them. The effectiveness of the functionalized polypropylenes as interfacial modifiers would be associated with the increase of the second-order interactions between the matrix and the dispersed phase, as well as with the higher mobility of polar groups bonded at the chain ends. These would increase the interaction possibilities with the surface of the discrete phase particles across the interface, which is especially useful to reduce the interfacial tension between the components of the heterogeneous system. The presence of mineral reinforcements such as talc or mica, as foreign solid particles embedded into a polypropylene matrix, usually induces a nucleation effect. A significant increase in the crystalline content of the polymer is evidenced if compared with the neat polymer when processed at the same setup conditions that are necessary to ensure a good accommodation of the solid particles into the amorphous phase of the polymer in order to obtain a material with a good mechanical performance (27). The comparison between PP/mica and PP/talc composites in terms of their mechanical behavior under dynamic conditions in the solid state agrees with the morphological features derived from their chemical structures of both minerals (28). In fact, if the width and thickness of a lamellar particle are defined as a fraction of its longest and main dimension, lr , it is easy to define the surface/volume ratio (S/V) for each particle as follows: S ¼ V

1 1 1þ þ n1 n2

þ

2 lr

ð13:2Þ

where n1 and n2 are the fractional numbers for, respectively, the width and the thickness of the particle relative to its longest dimension, lr . Because the S/V ratio is proportional to the specific surface of the mineral and lr being higher for mica than for talc, it follows that specific surface would always be lower for mica than for talc particles. Then for the same crystalline amount of the polypropylene matrix, a higher fraction of amorphous phase involved in the coating of talc particles than in the coating of mica particles would be expected. So, if the interfacial regions have to be considered, we must differentiate the amorphous/crystal polymer interface from the amorphous polymer/mineral interface (29). By DMA measurements, a decrease in Tg matrix from 7 C for the PP/talc composites and also for the neat PP processed under similar conditions while a decrease upto 13 C was found for PP/mica composites. A higher fraction of ‘‘free amorphous phase’’ on the PP/mica system than on the PP/talc composites was evidenced. This ‘‘free amorphous phase’’ appeared to participate in the cooperative segmental free-rotation motion, well accepted (30) to be responsible for glass transition for the polymer matrix as fully discussed in Reference 29.

390

Polyolefin Blends

Such kind of considerations about the finite dimensions of the interfacial area at the polymer/mineral interphase must be taken into account when we try to modify the interface by substituting a part (frequently a little part is enough) of the polypropylene matrix with a chemically modified polypropylene by grafting of polar groups. Also, the structural characteristics of such chemically modified polypropylene must be considered, mainly its molecular weight, if crystalline or not, and the nature and number of polar grafts. An example of the relationship between macroscopic responses of the composite material as a whole and the micro-, meso-, and nanomorphological scales involved can be found when a classical tensile loading test of this kind of material above Tg of the polymer matrix portion of the composite is performed. Consequently, two effects are involved in the evolution of the elastic modulus and the tensile strength at yield and at break too. On the one hand, the higher number of crystallites due to the nucleation effects that act as cross-linking regions between the interconnecting chain segments that belong to the so-called tie molecules. On the other hand, the self-contribution of those new crystallites to increase the elastic modulus because of the higher stiffness of the crystals compared to that of the amorphous phase. Strength at yield usually evolves in the same way as the elastic modulus does, while for strength at break, and depending on the amorphous phase present in the material, many differences can be expected at the ultimate stages of the overall strain process. Figure 13.4 shows the variations between the tensile test parameters for the PP/interfacial modifier/talc composites as a function of either the grafted group, succinic anhydride (SA) or succinyl-fluorescein (SF) attached to an atactic polypropylene, as well as the differences on tensile strength and strain levels at yield or at break points, depending on the amorphous or semicrystalline nature of the interfacial modifier as it was fully discussed elsewhere (29,31). Figure 13.5 compares the differences between the evolution of the elastic modulus and tan d from the DMA tests with both the kind of attached group at the atactic polypropylene and the level of grafting of the interfacial modifier present in the composites. The different effect of each interfacial agent is clearly concluded as discussed elsewhere (33,56). The capability of the succinyl-fluorescein grafted polypropylene used as interfacial modifier in the PP/talc composites to saturate the interfacial area available at the interphase is well observed on the left-hand plots of Fig. 13.6 where it can be seen how the amount of molten material during the second heating cycle in dynamical tests by DSC, compared to that previously ordered during the cooling step, evolves as a min–max curve in a very similar way that strength at break does (34,35). Right-hand plots in the same Fig. 13.6 allows to conclude the well-optimized material performance from the macroscopic up to the mesoscales by the correspondence between the elastic modulus component obtained from the tensile tests and the elastic component of complex modulus from the DMA measurements performed at room temperature and which ratio results to be independent of the grafting level in the interfacial modifier (32,55).

Chapter 13 Heterogeneous Materials Based on Polypropylene

391

Figure 13.4 Macroscopic responses by changes at the interfacial level. Variations on tensile test parameters with the amount and the structure of the interfacial modifiers for the PP/talc system. (From Reference 29 with permission of Elsevier.)

13.4.3.2

Polypropylene/Polyamide 6 System

Both from an applied and from an academic point of view, the polypropylene/ polyamide 6 system is indeed one of the most studied polymer/polymer binary systems, together with the interfacial modification possibilities induced in the morphology of the blend either by mixing during the processing steps or by the presence of interfacial modifiers. The immiscibility between both polymers is an important drawback that obliges us to look for a well-balanced set of properties, namely, mechanical, vapor, and gas permeability for barrier applications. Most applications are as multilayer, or sandwich sheet structures (polymer A/adhesive layer/polymer B), obtained by coextrusion, lamination or coating operations. The high operation costs associated with these processing techniques and the inability to obtain complex shapes that can expand

392

Polyolefin Blends

Figure 13.5 Mechanical and dynamic mechanical responses on PP/mica system. Changes with the grafting level and the structure of the interfacial modifier. (From References 33, 56 with permission of John Wiley & Sons, Inc.).

application to markets are the main driving forces behind the study of these blends from a technological point of view. From an academic perspective, this family of blends appears as an excellent model of the most general polycondensate/polyolefin polymeric systems; focusing on the development and control of post-reactive processing stable morphologies can yield a material with the desired performance, that is, good processability, low water absorption and liquid and vapor permeability, improved dimensional stability, good impact strength, and improved chemical resistance to alcohol and glycols. In a recent paper (9), a micromechanical model is proposed that is mainly characterized by a shape factor, w, given by w¼

ðn1 n2 Þ ðn1 þ n2 Þ

ð13:3Þ

Chapter 13 Heterogeneous Materials Based on Polypropylene

393

Figure 13.6 Examples of correlation between responses from different scales when the interfacial modifier is changed: Left, tensile strength at break point (up) and relative crystalline variation for the polypropylene matrix (down) versus the grafting level in the interfacial modifier; right, elastic moduli (tensile/DMA) ratio (up) and components of the complex modulus from DMA tests (down) versus the grafting level in the interfacial modifier. (From References 32 and 55 with permission of John Wiley & Sons, Inc. and Elsevier, respectively.)

where n1 and n2 are the fraction numbers that refer, respectively, to the width and the thickness of the dispersed phase particles to their longest dimension. Because of the deformability of the dispersed phase, this shape factor is associated with the processing history and allows us to define an approach to the morphology– mechanical properties relationship, starting on the two limit values for the shape factor, always between zero and one as proposed earlier (9). Optical microscopy on phase contrast mode allows observation of the different morphologies obtained for each PP/interfacial modifier/PA6 blend. By image analysis techniques, it is possible to carry out statistical field measurements not only of the mean number of particles on the dispersed phase but also of their preferential geometry, mean size, and size distribution.

394

Polyolefin Blends

Figure 13.7 PP/PA 6 binary system with modified interface. Example of morphological changes in the 85/15 w/w ratio blend with the amount (3% or 15% over PP phase) and with the grafted group (succinic anhydride, SA, or succinyl-fluorescein, SF) attached to the atactic polypropylene used as interfacial modifier.

By following the Box-Wilson experimental design methodology, we tried to fit all these different parameters looking for correlations and emerging functions between the response levels of the system with those obtained by mechanical (9), thermal (34), and diffraction (35) techniques. Figure 13.7 displays, as an example, the morphology observed by optical microscopy on negative phase contrast mode for the PP/interfacial modifier/PA6 system after compression molded on 100 mm thick sheets previously Banbury batch mixed. The phase contrast micrograph on top shows the 85/15 unmodified blend. Because the refractive index for the polyamide 6 (1.53) is higher than that for the polypropylene (1.49), the negative contrast condition leads us to observe the polypropylene matrix as a dark field, while the disperse domains of polyamide appear as bright regions. When comparing the micrographs for the modified blends with that for the unmodified one, in Fig. 13.7, it is easy to observe a sharp difference between the modified blends morphologies as a function not only of the amount (3% or 15% w/w)

Chapter 13 Heterogeneous Materials Based on Polypropylene

395

of the interfacial modifier at the interphase, but also depending on the attached group (SA or SF) onto the atactic polypropylene used as interfacial modifier. Indeed, the polypropylene matrix appears as a more uniform dark field because of the lower sizes of the domains of polyamide 6. The changes in the preferential geometry of the dispersed phase domains in the modified blends are also clearly observable as well as their higher brightness that should agree with a much more clear interphase in the modified blends. This is because the first-order bonds formed at several points of the PP/PA6 interphase are able to improve the optical light path on the microscope. The stable postreactive processing morphology of some of these blends was preliminary discussed by authors over their dynamical thermal responses (34,36). Moreover, the correspondence with the solid-state behavior of these blends can be observed in Figs. 13.8 and 13.9. The latter showing the differences between both a-PP-SA and a-PP-SF as interfacial modifiers for the PP/PA6 system at, respectively, the macroscopic mechanical response level, Fig. 13.8, and at the

Figure 13.8 Variations in the tensile strength values at yield (s F ) and at break (s R ) for the 85/15 and the 15/85 composition ratio of the PP/PA6 binary system with modified interface.

396

Polyolefin Blends

Figure 13.9 Evolution of long spacing of the crystalline phase of the polypropylene all along the PP/ PA 6 binary system with modified interfaces.

nanoscopic diffraction response level taking place over the crystalline phase of polypropylene, Fig. 13.9. For example, in Fig. 13.8, strength values at yield and at break from tensile tests for the PP/PA6 blends 85/15 and the 15/85 w/w ratio, have been displayed over a square plot (9). For the blends modified with a-PP-SA and for both PP/PA6 ratios, the linear increase of yield and break strength values when the aPP-SA amount increases up to the 9% is clearly observable. Above such value, both tensile strength values decrease if the a-PP-SA increases. In the same plots, it also may be observed that the tendency of higher strength at yield than at break for the neat PP remains in those blends where polypropylene acts as the matrix. Meanwhile, in those blends where polyamide 6 is the matrix (and so polypropylene the dispersed phase), it exhibits much closer values for strength at yield and at break, just as the neat polyamide 6 does. The different roles played by each interfacial modifier when located at the interphase are evident when we observe the two opposite effects caused by the aPP-SF/SA in the 85/15 and the 15/85 PP/PA6 modified blends. For the former, where the polypropylene is the matrix, the presence of the lowest amount of a-PP-SF/SA is enough to increase the tensile strength values with respect to the unmodified blend, while both strength at yield and at break decrease if the a-PP-SF/SA amount continues to increase. When polyamide becomes the matrix in the blend, both tensile strength values improve with respect to the unmodified blend by the presence of a-PP-SF/SA and also with an increase in a-PP-SF/SA amount.

Chapter 13 Heterogeneous Materials Based on Polypropylene

397

As observed for blends modified by a-PP-SA, the tendency of higher strength at yield than at break for the neat PP remains for the blends modified by a-PP-SF/SA where the polypropylene is the matrix. However, closer tensile strength at yield and at break values obtained for the neat polyamide 6 may be also observed in the modified blends by a-PP-SF/SA where the polyamide 6 acts as the matrix. These results would be coherent with those plotted in Fig. 13.9 that shows the higher long spaced values (L) of the crystalline phase of polypropylene present in the blends modified by the succinyl-fluoresceine grafted atactic polypropylene, than L for the blends modified by the succinic anhydride grafted atactic polypropylene. This would agree with the higher spatial volume of the former and its lower reorganizing possibilities at the highly constrained amorphous/crystal interphase (35). Consequently, the overall responses of the thermoplastic organic materials related to their applications are sharply affected by the thermal and strain/stress fields as they underwent before reaching the solid state, that is, the top-down approach to develop ways to control the performance of these materials seems indeed promising.

13.5 INTERFACIAL MODIFIERS BASED ON POLYPROPYLENE Chemical modification of polymers affords an opportunity to modify undesired properties that would limit or even invalidate the potential usefulness of these materials. Polypropylene polymers are a good example, because their applications depend on their stereoregularity. Atactic homopolymer, a by-product from industrial polymerization reactors, lacks good material properties. However, its chemical modification by grafting with polar groups, one can convert it into a useful new material, as suggested by Natta (37). As it has been demonstrated in the previous sections and also in numerous works available in the literature, the efficiency of the succinic anhydride grafted groups onto polypropylene backbone as interfacial modifiers on blends and composites based on polypropylene is indeed proved. In spite of its very high economic relevance, the process to obtain these new high value-added polypropylenes is far away to be realized. Indeed, the chemical modification of the polypropylene by grafting of polar groups is not well understood because of the complex nature of the process from the point of view of the chemical engineering. It means one needs to take into consideration the reactor where the process takes place, as well as the nature of the macromolecules as reactants or coreactants.

13.5.1 The Kinetic Approach: Basic Aspects Due to the complex nature of the macromolecular systems, theoretical models of the reaction mechanism are very difficult to build. In many cases, empirical models based on the experimental results obtained in any given conditions are implemented

398

Polyolefin Blends

for a specific process but they cannot be generalized. From this, semiempirical models based on theoretical considerations and supported by experimental data would be the most interesting way to approach the problem solution. Before building a semiempirical model, several basic aspects must be observed. First of all, for classical low molecular weight reactants the same reactivity for each individual molecule can be assumed because they must be almost identical. In the case of macromolecular reactants, like polypropylene, the existence of a molecular weight distribution makes this assumption invalid (38). However, the conversion in terms of modified and unreacted or unmodified macromolecules is not a trivial measurement. From here the analysis of the chemically modified polymer fraction and the variation of the modification level, that is, the yield of the desired reaction, as a function of the process conditions is not indeed an easy task, and the quality of the experimental results and procedures would determine the kind of kinetic studies. Regarding the radical mechanism that goes ahead through the reaction, it is important to note that the termination rates are proportional to the square of the radical concentration, while the relative contribution to the reaction yielding increases along with the radical concentration. A good control of the kinetic process requires fast initiation, looking for the same reaction probabilities for any active site, fast exchange possibilities between sites with different activity, and a low contribution of chain breaking reactions or any other undesired secondary processes. Next section describes the main findings by the present authors following these fundamental principles.

13.5.2 Chemical Modification of Polypropylenes by Grafting of Polar Monomers After preliminary studies conducted on isotactic polypropylene and maleic anhydride in the presence of dicumyl peroxide as initiator, we concluded, in agreement with several others, the existence of a maximum grafting level, or ‘‘ceiling,’’ whatever was the initial concentration of maleic anhydride and peroxide in the reaction media. By using a statistical design method for experimental runs, it was put into evidence that results needed to be carefully evaluated in order to find the true evolution of the process. According to initial findings on batch reactors, both in solution and in the molten state, the role of the reaction time in overall yield of the process was too far to be neglected (39) (Fig. 13.10). In order to check that point and also the role stereospecificity could play in the reaction extension, that is the molecular motion possibilities of the macromolecular coreactant, atactic polypropylene was chosen to be modified by grafting maleic anhydride in the same mode as before for isotactic polypropylene. The influence of the temperature of the process had also to be checked because the atactic polymer could not be processed at the same temperature as the isotactic one due to its nature. Results were fully discussed elsewhere (40). Figure 13.11 exhibits the dynamic and unsteady character of the grafting process.

Chapter 13 Heterogeneous Materials Based on Polypropylene

399

Figure 13.10 Grafting level (w/w) on polypropylenes chemically modified by maleic anhydride on batch processes. The role of the reaction time. (From Reference 39 with permission of John Wiley & Sons, Inc.)

This was in good agreement with the chemisorption mechanism over a changing surface, which had been proposed in Reference (39) to be responsible for the desired reactions. As shown in Fig. 13.11 the highest yield on grafting over the atactic polymer is also observable. If the assumption that polyolefin was responsible for grafting reaction were true, then by changing the polar monomer to be bonded, the grafting results ought to be showing similar tendencies (41,42).

Figure 13.11 The influence of the stereoregularity and the reaction time on the w/w grafting level of polypropylenes chemically modified by maleic anhydride. (From Reference 40 with permission of John Wiley & Sons, Inc.)

400

Polyolefin Blends

Figure 13.12 Grafting level (w/w) on isotactic polypropylene chemically modified by p-phenylenbis-maleamic acid on batch processes. The role of the reaction time. (From Reference 41 with permission of Elsevier.)

Figure 13.12 exhibits the unsteady evolution of the p-phenylen-bis-maleamic acid grafts on isotactic polypropylene, while plots in Fig. 13.13 confirm the same on atactic polypropylene. One of the oldest aspects of discussion about the topic under consideration was the possibility of grafting of polysuccinic anhydride sequences into the polypropylene backbone. The condensation reaction occurring between resorcinol and single succinic anhydride groups, previously bonded to the macromolecule, following Scheme 13.1 leads to the conclusion of the nonexistence of such polysuccinic anhydride grafts, as well as to the finding of a new family of fluorescent additives

Figure 13.13 Grafting level (w/w) on atactic polypropylene chemically modified by p-phenylene-bismaleamic acid on batch processes. The role of the reaction time. (From Reference 42 with permission of John Wiley & Sons, Inc.)

Chapter 13 Heterogeneous Materials Based on Polypropylene

401

Scheme 13.1 The reaction between resorcine and succinic anhydride grafted polypropylene to obtain the succinyl-fluorescein grafted polypropylene.

that has proven to be efficient as interfacial modifiers either in blends (9,34–36) or in polypropylene based composites (31–33,35,43). Taking into account the general chemisorption reaction scheme proposed in Reference 39 as reproduced in Fig. 13.14, new experimental results were obtained.

Figure 13.14 The Chemisorption process for maleic anhydride grafting in polypropylene. (From Reference 39 with permission of John Wiley & Sons, Inc.)

402

Polyolefin Blends

Recently a general balance of species for the unsteady grafting process of maleic anhydride into polypropylene in solution has been fully discussed in a recent paper (41), giving rise to the following general equation.

Scheme 13.2 General species balance involved in the reaction between maleic anhydride and polypropylene by a peroxide initiator in the presence of a solvent.

The pathway based on the general scheme for the classical consecutive reaction model, as shown in Fig. 13.15, contains all the relevant species involved in the chemical modification of polypropylene by a polar monomer, and agrees with a series of remarks associated with findings in previous studies (39–43,45). This implies the generation of free succinic anhydride (SA) as a reaction byproduct, and obtaining of the grafted polymer in a ratio of only 2/3 with respect to the remaining radicals active in the polymer bulk (unable to yield grafting reactions and obliged to lose their activity by b-scission processes leading to degradation of the polymer).

Figure 13.15 The consecutive reaction model for the unsteady chemisorption process of grafting of maleic anhydride in polypropylenes. (From References 44 and 46 with permission of John Wiley & Sons, Inc.)

Chapter 13 Heterogeneous Materials Based on Polypropylene

403

The latter is confirmed by the fact that the reaction proceeds with no loss of efficiency when performed in the presence of radical traps, or the thermal stabilizers usually used in polyolefin processing operations (38–45). The low grafting yield was obtained, both in terms of the number of bonded polar groups and of conversion levels. However, the experimental evidence reveals higher grafting yields obtained when the process occur in the melt rather than in solution (38,40,45). Scheme 13.2 once particularized for the molten state reaction media (46) drive us to the following.

Scheme 13.3 General reaction scheme between maleic anhydride and polypropylene by a peroxide initiator in the molten state.

Here it can be seen that for each SA group grafted onto the PP backbone, a second group is attached but involving some other sequence of the ‘‘solvent’’ PP. Further, every two SA groups grafted onto a-PP-SA require that three PP radicals remain active, that is, transferring radical activity from the reactant mass, which will continue with increasing reaction time until it achieves full deactivation. These three active radicals are, by definition, unable to graft MAH. They therefore manifest their activity through, for example, degradation processes such as chain scissions (in the case of PP) or by reacting with other species in the reaction mass (with the exception of MAH) such as the thermal stabilizer molecules. This can slow down or even stop the degradation processes. The two grafted SA units obtained in the species balance for the batch solution process in Scheme 13.1 may be considered grafted onto equivalent reactive sites (propylene sequences on the PP backbone). However, they must be essentially different from the two grafted SA groups obtained in the molten state, whose reaction surroundings are different due to the absence of solvent molecules. Scheme 13.3 provides a coherent explanation for earlier findings (38,45) regarding the higher grafting levels of i-PP-SA obtained with the molten state process than with the batch solution process (always with the same concentration of peroxide and MAH with respect to polypropylene). These findings were confirmed when the macromolecular coreactant was a-PP, as discussed elsewhere (39,40,43). As Scheme 13.3 shows, the presence of three active PP radicals for each two grafted SA groups at any given reaction time is coherent with the unsteady character of the reaction proposed (39). The ability of resorcine to act as a molecular probe, able to stabilize the SA groups on the polypropylene backbone according to the condensation reaction displayed in Scheme 13.1, was used in Reference 46 to validate the proposed

404

Polyolefin Blends

Figure 13.16 Experimental validation of the intermediate reaction on the consecutive reaction model describing the grafting of maleic anhydride into polypropylene. (From Reference 46 with permission of John Wiley & Sons, Inc.)

pathway. Figure 13.16 clearly shows the very high risk of neglecting the early stages of the process, a problem of existing kinetic models for the grafting of polar monomers onto polyolefines. Indeed, both reactions in Schemes 13.2 and 13.3 shows that in the balances of the species involved in the solution and molten state processes, three polypropylene radicals remain active for every two grafted SA groups. These become involved in the termination of the radical step that could proceed, as it is well accepted by disproportion and recombination, as the two major processes involved in the termination step of the classical three-step radical processes. In light of experiment results fully discussed in a recent paper (46), termination by disproportion would seem to be the dominant mechanism for the grafted species obtained in molten state, while recombination between radical species yielding SA groups trapped between two polypropylene sequences may be responsible of some 40% of the SA units being unable to react with the resorcine molecules. These could not undergo the structural rearrangement shown in Scheme 13.1 and would remain as grafted SA units when the condensation step is complete. From these findings, it follows for both the solution and the molten state reactions, and due to the unsteady character of the process that a dynamic distribution of grafted groups all along the polypropylene backbone takes place. These groups would increase in number until an optimum reaction time is reached. Once this has passed, an increasing number of grafted groups would form end chains. If the reaction is allowed to continue, a critical time would be reached, after which degradation processes would be the only means of deactivating the radical activity

Chapter 13 Heterogeneous Materials Based on Polypropylene

405

still remaining in the bulk of the reactant mass. In fact, as fully discussed by us (46), three different grafting population distributions for each of the a-PP-G samples produced by the batch solution and molten state processes were identified as welldefined fractions obtained by selective extraction in boiling n-heptane. The first was a nonsoluble fraction of a-PP-G unable to leave the Soxhlet cartridge; the second fraction from the ‘‘macroscopic’’ a-PP-G sample corresponded to soluble grafted chains able to leave the Soxhlet cartridge and recovered by classic precipitation in a nonsolvent such as methanol; and the third, soluble fraction—termed the soluble ad infinitum fraction—was recovered after the evaporation of the solvent–nonsolvent mixture and vacuum drying. While the insoluble species were characterized by the highest grafting levels obtained for a single a-PP-G sample close to the theoretically highest possible (according to sterical hindrance) for any given a-PP segment, the precipitant fraction showed the lowest grafting levels, far below the average macroscopic graft value; and the soluble ad infinitum fraction was characterized by grafting values similar to those of the macroscopic average, and almost identical to those of the a-PP-G samples obtained by the batch solution process. According to the experiment results for each of the solution batches, as displayed in Fig. 13.17, the three fractions showed the decreasing weight sequence soluble>insoluble>soluble1, while those produced by the molten state

Figure 13.17 The three different grafted species population distributions for the a-PP-G samples as revealed by solvent extraction in boiling n-heptane.

406

Polyolefin Blends

batches (with reaction conditions close to the optimum) followed the order soluble1>soluble>insoluble. Otherwise, if the initial concentration of maleic anhydride was far from optimum, either by default or because of an excess of the optimum coating on the chemisorption process (39), an inversion occurs between the weight populations of the two soluble fractions. The sequence then becomes soluble>soluble1o insoluble. The very low weight population of the insoluble fraction in all cases is in good agreement with a nonoptimum initial concentration of MAH. In the light of these findings, it may be hypothesized that the shortest chains, that is, those coming from disproportion termination reactions plus others produced by chain scissions processes (if there are any) should contribute to the (soluble)1 fraction. The inversion between the amounts of the different sequences in fractions

Figure 13.18 Modelization of grafted polar groups sequences on polypropylene for succinic anhydride (SA), succinyl-fluoresceine (SF), and p-phenylen-bis-maleamic acid ( p-PBM): HOMO (highest occupied molecular orbital) and LUMO (lowest unoccupied molecular orbital) draws and the potential surfaces for each group.

Chapter 13 Heterogeneous Materials Based on Polypropylene

407

of samples produced by the solution process (the lowest weight population) and in those obtained by the molten state process when close to optimal conditions (the highest weight populations) should agree with the recombination termination processes proposed for the solution reactions, in contrast with the disproportion reactions that should be the dominant termination pathway when the graft processes are carried out in the molten state. The way the properties of the original unmodified polypropylenes are affected by the presence of the different polar groups bonded to their macromolecular structure remains an open line of work not only to the need characterization of the modified polyolefin, but to take the full control of the modification processes in terms of the number and localization of the sites where the polar groups are attached to the macromolecule. Further studies on modified polymer structure are in progress now in our laboratories following those about their molecular weights, the thermal behavior of the modified polymers in terms of the first- and second-order thermal transitions, respectively, assigned to the crystalline and to the amorphous phases (47–49), and those dealing with both their radiation behavior, that is, vibrational (50,51) and NMR (52) spectroscopies as well as their molecular modeling (53). Main goal would be to develop correlations between the experimental and theoretical structural data of the chemically modified polypropylenes and their different interfacial modification capabilities on polymer blends and composites, not only by the chemical structure of the attached polar group (Fig. 13.18), but also by the number of those attached groups and where they are alocated all along the polymer backbone (54–63).

13.6

CONCLUSIONS

As the twenty-first century goes ahead, the organic materials remain on the vanguard of the development of new and advanced materials and trends to approach to the Nature. Furthermore, it is evident that the new technological frontiers to be crossed lay on the development and understanding of the fundamentals that govern the properties of the material combinations, rather than on the deepest understanding of the application of a particular material. Environmental considerations and sustainable development criteria when applied to the materials design field should favor the applications based on the organic thermoplastic polymers (64). The high versatility of polypropylene has placed it beyond the border of commodities to become an engineering polymer. Indeed, new developments based on this polyolefin would be the key in the strategic area of the heterogeneous materials based on organic polymers, both from an academic as well as technical point of view. Because of the increasing demand on the full control over the material properties from macro up to the nanoscales, the top-down approach in the modeling and development of new, advanced materials makes it necessary to improve analysis and characterization tools together by interdisciplinary research teams, always faithful to the rigors of scientific inquiry.

408

Polyolefin Blends

ACKNOWLEDGMENT Results compiled and referred to in this chapter were partially financed and developed under the following Spanish Research Public Projects: MAT93-0115, MAT960386, and MAT2000–1499.

NOMENCLATURE a-PP-G b lr Dn d hm w g n1 n2 PP PA 6 S SA SF sF sR Tg V We G

Chemically modified atactic polypropylene by grafting of the polar group G Carbon atom onto polypropylene backbone where the scission process goes ahead Longest dimension in a particle Mean size of the particles in a dispersed phase Loss phase angle on DMA tests Matrix viscosity in a polymer blend Shape factor Shear rate Width/longest dimension ratio in a particle Thickness/longest dimension ratio in a particle Polypropylene Polyamide 6 Particle surface Succinic anhydride Succinyl-fluoresceine Tensile strength at yield Tensile strength at break Glass transition temperature Particle volume Weber number Interfacial tension between the components of a polymer blend

REFERENCES 1. F. N. Cogswell, Thermoplastics Polymer Composites, Butterworth-Heinemamm, Oxford, 1987. 2. E. Helfand and Y. Tagami, J. Polym. Sci., Polym. Lett., 9, 741 (1971). 3. D. V. Rosato, D. P. DiMattia, and D. V. Rosato, Designing with Plastics and Composites, Van Nostrand-Reinhold, New York, 1991. 4. L. E. Nielsen, Mechanical Properties of Polymer and Composites, Vols I and II, Marcel Dekker, New York, 1974. 5. H. S. Katz and J. V. Milewski, Handbook of Fillers and Reinforcements for Plastics, Van NostrandReinhold, New York, 1978. 6. L. A. Utracki, Polymer Alloys and Blends, Hanser, Munich, 1989. 7. W. Baker, C. Scott, and G. H. Hu (eds.), Reactive Polymer Blending, Hanser, Munich, 2001.

Chapter 13 Heterogeneous Materials Based on Polypropylene

409

8. L. Mascia, Thermoplastics Materials Engineering, 2nd edition, Elsevier, London, 1989. 9. J. Mª. Garcı´a-Martı´nez, S. Areso, and E. P. Collar, J. Macromol. Sci. B Phys., B40, 387 (2001). 10. R. Toth, A. Coslanich, M. Ferrone, M. Fermeglia, S. Pricl, S. Mierturs, and E. Chiellini, Polymer, 45, 8075 (2004). 11. B. Wunderlich, Macromolecular Physics, Vols 1 and 2, Academic Press, New York, 1973. 12. J. M. Schultz, Polym. Eng. Sci., 24, 770 (1984). 13. P. W. Anderson, Science, 177, 393 (1972). 14. G. Xu and S. J. Lin, J. Macromol. Sci. Rev. Macromol. Chem. Phys., C34, 555 (1994). 15. Y. H. R. Jois and J. B. Harrison, J. Macromol. Sci. Rev. Macromol. Chem. Phys., C36, 433 (1996). 16. M. K. Naqui and M. S. Choudhari, J. Macromol. Sci. Rev. Macromol. Chem. Phys., C36, 601 (1996). 17. G. Moad, Prog. Polym. Sci., 22, 81 (1999). 18. R. Aris, Mathematical Modeling Techniques, Pittman, San. Francisco, 1978. 19. P. K. Freakley and G. A. W. Murray, J. Appl. Polym. Sci., Appl. Polym. Symp., 50, 150 (1992). 20. E. P. Plueddemann, Interfaces in Polymer Matrix Composites, Academic Press, New York, 1974. 21. A. Whelan and J. L. Craft, Developments in Plastics Technology, Vol. 2, Elsevier, London, 1985. 22. H. Ishida, Proc. NATO ASI Series, Akovali, G. (ed), Appl. Sci., 230, 169 (1993). 23. T. F. Cooke, J. Polym. Eng., 7, 197 (1987). 24. J. Taranco, O. Laguna, and E. P. Collar, J. Polym. Eng., 11, 359 (1992). 25. J. Taranco, J. M. Garcı´a-Martı´nez, O. Laguna, and E. P. Collar, J. Polym. Eng., 13, 287 (1994). 26. E. P. Collar, J. M. Garcı´a-Martı´nez, O. Laguna, and J. Taranco, J. Polym. Mater., 13, 111 (1996). 27. J. Mª. Garcı´a-Martı´nez, O. Laguna, and E. P. Collar, J. Polym. Mater., 15, 127 (1998). 28. N. H. Ray, Inorganic Polymers, Academic Press, London, 1978. 29. E. P. Collar, O. Laguna, S. Areso, and J. M. Garcı´a-Martı´nez, Eur. Polym. J., 39, 157 (2003). 30. D. W. Van Krevelen, Properties of Polymers, 3rd edition, Elsevier, Amsterdam, 1990. 31. J. Mª. Garcı´a-Martı´nez, O. Laguna, and E. P. Collar, J. Polym. Eng., 17, 269 (1998). 32. J. Mª. Garcı´a-Martı´nez, O. Laguna, S. Areso, and E. P. Collar, J. Polym. Sci. Polym. Phys., 40, 1371 (2002). 33. J. Mª. Garcı´a-Martı´nez, O. Laguna, S. Areso, and E. P. Collar, J. Polym. Sci. Polym. Phys., 38, 1554 (2000). 34. E. P. Collar, C. Marco, S. Areso, and J. M. Garcı´a-Martı´nez, J. Macromol. Sci. B Phys., B40, 369 (2001). 35. J. Mª. Garcı´a-Martı´nez, S. Areso, M. A. Go´mez, C. Marco, and E. P. Collar, Hasylab Annual Report, Part I, Digital Realease, 2003. 36. C. Marco, S. Areso, E. P. Collar, and J. M Garcı´a-Martı´nez, J. Polym. Sci. Polym. Phys., 40, 1307 (2002). 37. G. Natta, F. Severini, M. Pegoraro, and C. Tavazzani, Makromol. Chem., 119, 201 (1968). 38. J. Mª. Garcı´a-Martı´nez, J. Taranco, O. Laguna, and E. P. Collar, Int. Polym. Proc., 9, 246 (1994). 39. J. Mª. Garcı´a-Martı´nez, O. Laguna, and E. P. Collar, J. Appl. Polym. Sci., 65, 1333 (1997). 40. J. Mª. Garcı´a-Martı´nez, O. Laguna, and E. P. Collar, J. Appl. Polym. Sci., 68, 483 (1998). 41. J. Mª. Garcı´a-Martı´nez, A. G. Cofrades, O. Laguna, S. Areso, and E. P. Collar, Eur. Polym. J., 36, 2253 (2000). 42. J. Mª. Garcı´a-Martı´nez, A. G. Cofrades, S. Areso, and E. P. Collar, J. Appl. Polym. Sci., 88, 2202 (2003). 43. J. Mª. Garcı´a-Martı´nez, O. Laguna, S. Areso, and E. P. Collar, J. Appl. Polym. Sci., 70, 689 (1998). 44. J. Mª. Garcı´a-Martı´nez, S. Areso, and E. P. Collar, J. Appl. Polym. Sci., 102, 1182 (2006). 45. J. Mª. Garcı´a-Martı´nez, J. Taranco, O. Laguna, and E. P. Collar, Int. Polym. Proc., 9, 346 (1994).

410

Polyolefin Blends

46. J. Mª. Garcı´a-Martı´nez, S. Areso, and E. P. Collar, J. Appl. Polym. Sci. 104, 345(2007). 47. E. P. Collar, S. Areso, O. Laguna, and J. M. Garcı´a-Martı´nez, J. Therm. Anal. Calorim., 58, 541 (1999). 48. E. P. Collar, A. G. Cofrades, O. Laguna, S. Areso, and J. Mª. Garcı´a-Martı´nez, Eur. Polym. J., 36, 2265 (2000). 49. E. P. Collar, S. Areso, O. Laguna, and J. M. Garcı´a-Martı´nez, Eur. Polym. J., 35, 1861 (1999). 50. J. Mª. Garcı´a-Martı´nez, S. Areso, M. A. Go´mez, C. Marco, and E. P. Collar, Hasylab Annu. Rep., Part I, 845, 2002. 51. J. Mª. Garcı´a-Martı´nez, O. Laguna, S. Areso, and E. P. Collar, J. Appl. Polym. Sci., 73, 2837 (1999). 52. L. Garrido, Ma. M. Pe´rez-Me´ndez, J. Aisa, J. M.a Garcia-Martinez, C. Marco, G. Ellis, S. Areso, E. P. Collar, Final Report MAT 1499-2000, Spanish Public Research Project, unpublished results, 2003. 53. Mª. M. Pe´rez-Me´ndez, L. Garrido, J. Aisa, J. M.a Garcia-Martinez, C. Marco, G. Ellis, S. Areso, and E. P. Collar, Final Report MAT 1499-2000, Spanish Public Research Project, unpublished results, 2003. 54. G. Ellis, C. Marco, M. A. Go´mez, E. P. Collar, and J. M. Garcı´a-Martı´nez, J. Macromol. Sci. B Phys., B43, 251 (2004). 55. J. Mª. Garcı´a-Martı´nez, O. Laguna, S. Areso, and E. P. Collar, Eur. Polym. J., 38, 1583 (2002). 56. J. Mª. Garcı´a-Martı´nez, O. Laguna, S. Areso, and E. P. Collar, J. Appl. Polym. Sci., 81, 625 (2001). 57. E. P. Collar, S. Areso, O. Laguna, and J. Mª. Garcı´a-Martı´nez, J. Polym. Mater., 15, 363 (1998). 58. E. P. Collar, S. Areso, O. Laguna, and J. Mª. Garcı´a-Martı´nez, J. Polym. Mater., 15, 355 (1998). 59. E. P. Collar, S. Areso, O. Laguna, and J. Mª. Garcı´a-Martı´nez, J. Polym. Mater., 15, 237 (1998). 60. J. Taranco, O. Laguna, and E. P. Collar, J. Polym. Eng., 11, 345 (1992). 61. J. Taranco, O. Laguna, and E. P. Collar, J. Polym. Eng., 11, 335 (1992). 62. J. Taranco, O. Laguna, and E. P. Collar, J. Polym. Eng., 11, 325 (1992). 63. J. Taranco, O. Laguna, and E. P. Collar, J. Polym. Eng., 11, 315 (1992). 64. Opinion, Nature, 419, 543 (2002).

Chapter

14

Polypropylene/Ethylene– Propylene–Diene Terpolymer Blends Chang-Sik Ha,1 Subhendu Ray Chowdhury,2 Gue-Hyun Kim,3 and Il Kim1

14.1 INTRODUCTION Different polymers have different unique properties. To combine these unique properties of component polymers blending is an attractive means. There are a few methods to make polymer/polymer blends: solution blending, melt extrusion, in situ polymerization, among others. Compatibility usually plays a major role in the development of properties. The blends prepared by melt mixing of thermoplastic materials and rubbers have met industrial needs in recent years. Thermoplastic elastomeric materials have many important applications including cable and wire especially in mineral, electronic equipment, and automobile industries. The most commonly used method of obtaining thermoplastic elastomer in materials is to toughen plastics by blending rubbers and plastics. Among the most versatile polymer matrices, polyolefins such as polypropylene (PP) are the most widely used thermoplastics because of their well-balanced physical and mechanical properties and their easy processability at a relatively low cost that makes them a versatile material. PP has the disadvantage of becoming brittle at low temperature, however, because of its high transition temperature and high crystallinity. The best way to improve its impact strength is to blend PP with elastomers 1

Department of Polymer Science and Engineering, Pusan National University, Busan 609-735, Korea

2

Department of Materials Science and Engineering, Pennsylvania State University, University Park, PA

16802, USA 3

Division of Applied Bioengineering, Dongseo University, Busan 617-716, Korea

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

411

412

Polyolefin Blends

Figure 14.1 Young’s modulus of PP/EPDM blends plotted against EPDM content (blends were prepared with 20 rpm of a rotor speed of an internal mixer at 220 C). (From Reference 4 with permission from Elsevier Science Ltd.)

particularly ethylene–propylene copolymers and terpolymers (EPM and EPDM, respectively). The blend of PP and EPDM has been prepared by different means, melt mixing, dynamic radiation curing, and ultrasonic curing, among others. These PP/EPDM blends have been widely studied from different angles; that is, structure–properties relationship, morphology, mechanical properties, rheology, thermal properties, among others. In this sense, some monographs and reviews have been published on the PP/EPDM blends and related materials for these two decades (1–3). Thus, readers should refer to those review articles and book chapters to get general guidelines on the preparation and properties of PP/EPDM blends. In this chapter, we do not reproduce those general aspects of the PP/EPDM blends. Instead, we wish to review recent interesting reports on the PP/EPDM blends in terms of thermoplastic polyolefins (TPO) and thermoplastic vulcanizates (TPVs) as commercially important products. We also briefly touch the applications of the PP/EPDM blends for readers. A typical example of the effect of EPDM addition on the mechanical properties of PP can be seen in the work of DA Silva and Coutinho (4). They described the effect of EPDM amount and also processing condition on the mechanical properties of PP/EPDM blends. As EPDM contents increases, the impact strength of PP/EPDM blends increases but the tensile strength and Young’s modulus decrease and the elongation at break increases. Figure 14.1, for instance, illustrates the effect of EPDM contents on the Young’s modulus of PP.

14.2 PP/EPDM BLENDS 14.2.1 Toughness and Crystallization Behaviors of PP/EPDM Blends The toughness of PP/EPDM blend was investigated by Huang et al. (5) over a wide range of EPDM content and temperature. It is seen in Fig. 14.2 that the Izod impact strength of PP increases with increasing EPDM contents and temperature. It is

Chapter 14 Polypropylene/Ethylene–Propylene–Diene Terpolymer Blends

413

Figure 14.2 Variation of notched Izod impact strength with temperature for various EPDM content (weight fraction). (From Reference 5 with permission from Elsevier Science Ltd.)

worth noting that both toughness and brittle–ductile transition (BDT) of the blends were a function of the notch radius (R) for the Izod notch impact tests. (In Fig. 14.2, a–c three different notch tips are shown; Notch A is a 45 V-shaped notch with the tip radius (R) of 0.25 mm, notch B is a 45 V-shaped notch with the R of 1.0 mm, and notch C is a rectangular notch.) At test temperature, the toughness tended to decrease with increasing 1/R for various PP/EPDM blends. The brittle–ductile transition temperature (TBT) increased with increasing 1/R, whereas the critical interparticle distance (IDc) reduced with increasing 1/R. The different curve of IDc versus test temperature ðTÞ for notches reduces down to a M M –T, where TBT is the TBT of PP itself for master curve if plotting IDc versus TBT M a given notch, indicating that TBT –T is a more universal parameter that determined the BDT of polymers. The thermal and morphological behaviors of PP/EPDM blends were studied by Da Silva and Coutinho (6) using differential scanning calorimetry (DSC) and polarized optical microscopy (POM), respectively. Crystallization kinetics of PP/ EPDM blends were found similar. Ten to twenty weight percent addition of EPDM resulted in increasing of spherulite size (Fig. 14.3). Heat of fusion and crystallinity degree of PP/EPDM systems decreased when EPDM contents were increased.

414

Polyolefin Blends

Figure 14.3 PP spherulite size vs. EPDM content. (From Reference 6 with permission from John Wiley & Sons, Inc.)

14.2.2 Compatibilization of PP/EPDM Blends Sometimes, the compatibilization of PP/EPDM blends has been the key issue to improve the properties of the blends. Lopez-Manchado’s group (7) studied the effect of grafted PP on the compatibility and properties of PP-EPDM thermoplastic elastomer blends. They functionalized the isotactic PP (iPP) through grafting in Brabender plasticorder with two itaconate. The functionalization of iPP was performed by melt blending through grafting with two itaconic acid derivatives, monomethyl itaconate and dimethyl itaconate (MMI and DMI, respectively). Grafting was performed at 190 C using two different initial monomer concentration and 2,5 dimethyl 2,5 bis (tert-butyl peroxy) hexane as radical initiator. Some PP/EPDM blends were made from unmodified PP and modified PP (i.e., grafted PP with MMI or DMI) with EPDM. The flow properties analyzed by torque values, melt index, and rheological studies showed that the blends made with grafted PP have better processability, show lower viscosity with this effect being more significant in DMI-modified PP. Good interaction between two phases are evident from a dynamic mechanical analysis (DMA) and tensile properties. The functional polar monomer acted as a compatibilizer as well as nucleating agent for PP crystallization. It shows a substantial decrease in the halftime of crystallization, which is attributed to the presence of a greater number of nuclei in the crystallization process. Su et al. (8) studied the mechanical properties and morphological structure relationship of blends based on sulfated EPDM ionomer and PP. They synthesized Zn2þ neutralized low degree sulfated EPDM (Zn-SEPDM) ionomer and PP blends and studied their mechanical properties. They found that Zn2þ neutralized low degree sulfated EPDM ionomer and PP blends have better mechanical properties than those of PP/EPDM blend, as shown in Fig. 14.4. They explained the reason why mechanical properties are higher for Zn-SEPDM and PP than for PP and EPDM using scanning electron microscopy (SEM) (Fig. 14.5). Finer dispersed phase size and the shorter interparticle distances are the main reasons for the improved mechanical properties of the PP/EPDM blend.

Chapter 14 Polypropylene/Ethylene–Propylene–Diene Terpolymer Blends

415

Figure 14.4 Stress–strain curve of neat PP, PP/EPDM, and PP/Zn-SEPDM blends (a) neat PP, (b) PP/EPDM, (c) PP/0.03 mol% Zn-SEPDM, (d) PP/0.06 mol% Zn-SEPDM. (From Reference 8 with permission from John Wiley & Sons, Inc.)

Figure 14.5 SEM micrographs of fractured etched PP/EPDM blends: (a) PP/EPDM, (b) PP/0.03 mol% Zn-SEPDM, (c) PP/0.06 mol% Zn-SEPDM. (From Reference 8 with permission from John Wiley

416

Polyolefin Blends

Figure 14.6 The plot of Tg vs. blend composition in the m-PP/Zn-EPDM blends. (From Reference 9 with permission from Wiley Interscience.)

Similarly, Ha et al. (9) blended Zn-SEPDM with maleic anhydride grafted PP (PP-gMAH). Using light scattering, DMA, and FT-IR spectroscopy, they found that the compatibility of PP/EPDM blend was significantly improved by the use of both EPDM ionomer and maleic anhydride grafted PP. In Fig. 14.6, a sigmoidal trend in Tg of the PP-g-MAH/Zn-SEPDM blends as a function of a blend composition suggests that there exists strong interaction between the maleic anhydride grafted PP and the ionomeric EPDM and compatibilization was achieved between PP and EPDM.

14.2.3 Ternary Blends and Composites from PP/EPDM Blends In order to improve properties and compatibility of PP/EPDM blends, ternary blends and composites are sometimes prepared from the PP/EPDM blends. For instance, Sanchez et al. (10) prepared ternary blends of PP, high density polyethylene and EPDM with several blending ratios and investigated the melt rheological behaviors. They discussed the effect of the shear rate on the viscosity and flow curve in terms of the exponent of low power for a non-Newtonian liquid. They showed that addition of an elastomer to the polyolefin blends changes the shape of the viscosity–composition curve, and they discussed it in terms of the possible morphology of the blend. Similar works have been also reported by Ha et al. (11,12).

Chapter 14 Polypropylene/Ethylene–Propylene–Diene Terpolymer Blends

417

14.2.4 Application of Radiation A series of gamma-radiation cross-linking of polymer blends containing fresh or waste EPDM and PP was prepared by Zaharescu et al. (13,14). They figured out the optimal dose range for the efficient cross-linking of all EPDM/PP blends to be 40–180 kGy. Thermal stability of the studied mixtures was assessed in order to state the contribution of the components to the radiation compatibilization of investigated polymers. Figures 14.7 and 14.8 illustrate typical results. They investigated the effect of radiation on tensile strength and elongation at break of EPDM/PP blend. Fresh and waste PPs were separately compounded with EPDM. In spite of the improvement in their gel content at 150 kGy, the degraded component causes alternation in mechanical properties. The effect of ionizing radiation on thermal oxidation of PP/EPDM blend over the range of total gamma doses up to 250 kGy was also studied by Zaharescu and

Figure 14.7 Change of the elongation at break of (a) EPDM/fresh PP and (b) EPDM/waste PP samples (&,^) unirradiated blends, (&,*) irradiate blend (150 kGy). (From Reference 14 with permission from Elsevier Science Ltd.)

418

Polyolefin Blends

Figure 14.8 Change in tensile strength of (a) EPDM/(fresh) PP and (b) EPDM/(waste) PP samples, () unirradiated blends; (c) irradiated blends (150 kGy). (From Reference 14 with permission from Elsevier Science Ltd.)

Budrugeac (15). They studied the influence of irradiation dose on oxidation induction periods by oxygen uptake and thermal analysis on polymer samples containing various concentrations of components (100/0, 80/20, 60/40,40/60, 20/80, and 100/ 0 w/w). Drastic decrease in oxidation induction time was observed for low doses. The competition between cross-linking and scission has been examined on the basis of radical recombination on postirradiation time. Effect of specimen formulation on oxidation induction time has been discussed considering the antagonistic processes, such as crosslinking, and oxidative degradation. The effect of ultrasonic irradiation on the mechanical property, morphology, and crystal structure of PP/EPDM blends were examined by Chen and Li (16). Appropriate ultrasonic intensity can increase the toughness of PP/EPDM blends noticeably. SEM showed that with ultrasonic irradiation, the morphology of a well-dispersed EPDM phase is formed in the PP/EPDM blend. Ultrasonic irradiation interestingly

Chapter 14 Polypropylene/Ethylene–Propylene–Diene Terpolymer Blends

419

Figure 14.9 SEM micrographs of fractured surface at the temperature of liquefacient nitrogen and treated by xylene at 20 C of PP/EPDM ¼ 70/30 (A: 0þ0, B: 150þ0, C:150 þ 100). (From Reference 16 with permission from Wiley Interscience.)

brought the Tgs of PP and EPDM closer. Ultrasonic irradiation increases crystallization PP/EPDM blend and b-crystal of PP form in PP/EPDM blend, which is proven by X-ray diffraction. The authors again studied the morphology and compatibility of PP/EPDM blend change by ultrasound irradiation. Stable morphology with reduced dispersed phase size is obtained. Dynamic rheological analysis indicated more homogeneous internal structure of PP/EPDM blend. If the ultrasonic irradiation time is increased, the blend becomes more homogeneous due to the reduction of interfacial tension between PP and EPDM decreased (Fig. 14.9). Zaharescu (17) studied the compatibility of PP/EPDM blends prepared by gamma irradiation. He found that the composition of polymer specimens does not influence the relative contribution of constituents to the calculated heat capacity ðCp Þ values for tested blends exposed to the same dose.

14.3 DYNAMICALLY VULCANIZED PP/EPDM BLENDS (OR THERMOPLASTIC VULCANIZATES (TPVs)) The blends of cross-linked EPDM and PP can be prepared in a roll mill or extruder by the ‘‘dynamic vulcanization (DV) (or curing)’’ method where EPDM is vulcanized under shear with curing agent. The dynamically vulcanized thermoplastic elastomers (TPVs) have been widely used in the plastic industry because of their technical advantages in processing as well as their versatile end-use properties (18–22). The blends have important technical advantages in processing because of the thermoplastic nature of the melt, even though they contain a vulcanized elastomer as one

420

Polyolefin Blends

component. In general, vulcanized rubber cannot be reprocessed because of the formation of network structure. Such thermoplastic nature of the TPV might be ascribed to the fact that the network structures are formed in small rubber particles dispersed in the uncross-linked thermoplastic polymer matrix. They have a number of practical advantages over conventional rubber: a short mixing and processing cycle and low energy consumption; the scrap can be recycled; and properties can be easily manipulated by changing the ratio of the components. Unvulcanized EPDM/PP blends cannot replace thermoset rubber because of poor resistance to compression or tension set at elevated temperature or under prolonged deformation and poor oil resistance. Compared with unvulcanized EPDM/PP blends, the following properties are improved in EPDM/PP TPVs: oil resistance, permanent set, ultimate mechanical properties, fatigue resistance, heat deformation, melt strength, among others (3). As a result, EPDM/PP TPVs are quite adequate for most vulcanized rubber applications.

14.3.1 Effect of Cross-linking on the Properties of PP/EPDM TPVs Ishikawa et al. (23) examined the toughening mechanism of PP/EPDM blend before and after crosslinking, as shown in Figs. 14.10 and 14.11. They evaluated yield stress, strength of craze, and density of void, which are dominant factors for enhancing toughness in PP blends. The fracture and deformation mechanism were discussed. Ha et al. (24) studied the structure–property relationship of EPDM/PP blend. EPDM was cured with PP with dicumyl peroxide (DCP) at different shear conditions.

Figure 14.10 Izod impact strength of PP blended with EPDM following selective cross-linking in comparison with that of PP blend before cross-link. (From Reference 23 with permission from John Wiley & Sons, Inc.)

Chapter 14 Polypropylene/Ethylene–Propylene–Diene Terpolymer Blends

421

Figure 14.11 Variation of blending moment–displacement curve of the round notched bar of PP blended with EPDM with increase in EPDM content. (From Reference 23 with permission from John Wiley & Sons, Inc.)

They also cured EPDM without PP and then blended with PP. The effect of DCP concentration, intensity of the shear mixing, and rubber/plastic composition were studied. They found in blend cure that the melt viscosity increased with increasing DCP concentration in the blends of 75% EPDM and 25% PP but decreased with increasing DCP concentration in blends of 75% PP and 25% EPDM. Melt viscosity increased with increasing DCP concentration for all compositions in cure blend. With increasing intensity of the shear mixing, the melt viscosity decreased. Figures 14.12 and 14.13 show the results. In dynamically vulcanized EPDM/PP blends, the melt viscosity increased with increasing DCP concentration in blends of 75% EPDM and 25% PP but decreased with increasing DCP concentration in blends of 75% PP and 25% EPDM. It is clear that increasing DCP concentration for EPDM/PP (75/25) blend leads to the more molecular restrictions by the chemical cross-links in EPDM. The decrease in viscosity with DCP concentration for EPDM/PP (25/75) blend is attributed to the accelerated mechanochemical degradation of PP in the presence of peroxide during the dynamic vulcanization. It seems that with compositions of PP greater than 50%, mechanochemical degradation of PP becomes predominant, whereas at EPDM/PP (75/25) composition, the cross-linking effect of EPDM is dominant. Organic peroxide, sulfur, and phenolic resin system are generally used for the preparation of TPVs as cross-linking agents. However, all cross-linking agents have their own disadvantages: Organic peroxide leads to the degradation of PP greatly, sulfur causes the odor, and considerable concentrations of phenolic resin and accelerator required for the effective cross-linking deteriorate the impact strength of TPVs. Recently developed dynamically photocross-linked PP/EPDM blends showed improved mechanical properties (25). Especially impact strength was enhanced dramatically due to improved compatibility between PP and EPDM. This improved

422

Polyolefin Blends

Figure 14.12 Shear stress-shear rate curves for EPDM/PP linear blends at 200 C: (*) EPDM 100; (&) EL75PP25; (~) EL50PP50; (&) EL25PP75; (*) PP100. (From Reference 24 with permission from John Wiley & Sons, Inc.)

compatibility was ascribed to the possible existence of grafted chains of PP and EPDM. Jain et al. (26) studied the effect of dynamic cross-linking on the melt rheological property of PP/EPDM rubber blends. Rheological properties of vulcanized and unvulcanized EPDM/PP blends were reported. The melt viscosity increases with increasing EPDM concentration and decreased with increasing intensity of the shear mixing for all compositions. Vulcanized blend displays highly pseudoplastic behavior, which provides unique processability in injection molding and extrusion. The high viscosity at low shear rate provides the intensity of the extruders during extrusion and the low viscosity at high shear rate enables low injection pressure and less injection time. They also explained the property difference for vulcanizates with the help of morphology study. Cure characteristics of EPDM/PP blends were investigated by Sengupta and Konar (27). They calculated the state of cure in blends containing conventional sulfur curing system under variable time–temperature conditions. They found that the activation energy for the cross-linking is almost similar for the virgin EPDM and EPDM/PP mixtures. Cross-link densities in TPVs can be analyzed by swollen-state

Chapter 14 Polypropylene/Ethylene–Propylene–Diene Terpolymer Blends

423

Figure 14.13 Effect of DCP concentration (phr) on viscosity for EB75PP25 blend cure at 200 C (mixed speed 60 rpm): (*) 0.00; (*) 0.33; (~) 0.67; (&) 1.00; (~) 1.33. (From Reference 24 with permission from John Wiley & Sons, Inc.)

nuclear magnetic resonance (NMR) spectroscopy with a network visualization techniques (28,29).

14.3.2 Microstructure of PP/EPDM TPV The enhanced properties of TPV are ascribed to their specific morphology that consists of a continuous PP matrix with tiny cured EPDM particles dispersed throughout the matrix. As the EPDM particle size decreases, the ultimate elongation and tensile strength increase. Therefore, the physical properties depend on the morphology of the EPDM particle size. Their morphological behavior is different from the unvulcanized EPDM/PP blends. According to Abdou-Sabet et al. (30), before dynamic vulcanization, PP was the dispersed phase in the EPDM matrix for 80/20 EPDM/PP composition. However, as dynamic vulcanization progresses, the continuous rubber phase becomes elongated further and further and finally breaks up into rubber droplets. As a result, the PP becomes the continuous phase. The detailed mechanism of the morphological development was suggested for the dynamically vulcanized EPDM/PP (60/40 w/w) blends (31,32). Hot xylene

424

Polyolefin Blends

extraction revealed the formation of agglomerate structure by the rubber particles. As the vulcanization of the rubber proceeds, melted PP molecules adsorb onto the surface of semicrosslinked rubber particles through segmental interdiffusion mechanism. This would lead to the formation of an immobilized PP shell attached to the surface of the rubber particles. When the shelled particles are close enough, they would form an agglomerate structure. When the melt viscosity of PP is matched with the viscosity of EPDM, the formation of higher number but smaller size rubber particle aggregates was observed. This would result in better mechanical properties of the TPV. Figure 14.14 shows the SEM micrograph of dynamically vulcanized EPDM/PP (75/25) blend fracture surfaces etched by hot xylene vapor. The microdomains of EPDM have the shape of dumbbell-like microgel of about 0.8–1.0 mm in size, where the dark portions represent the PP phase extracted out by hot xylene vapor. The morphology of the microgel domain of EPDM reveals the reason why the dynamically vulcanized blend can be processed and the dynamic vulcanization prevents the cross-linking of EPDM phase from truly continuous network (24). PP/EPDM blends including dynamically vulcanized one show a higher rate of crystallization and higher Tc than those of PP homopolymer (33). It was reported that

Figure 14.14 SEM micrograph of EB75PP25 blend-cure with DCP concentration of 0.67 phr (dynamically cured at 60 rpm). (From Reference 24 with permission from John Wiley & Sons, Inc.)

Chapter 14 Polypropylene/Ethylene–Propylene–Diene Terpolymer Blends

425

such crystallization behavior was attributed to the role of EPDM to selectively extract the defective molecules within PP crystals and also increase the mobility of neighboring PP chains by the reduction of glass transition temperature (34).

14.3.3 PP/EPDM/Ionomer TPVs Kim et al. (35) reported on the control of miscibility for iPP and EPDM by adding polyethylene-co-methacrylic acid ionomer and by applying dynamic vulcanization (DV). Blending and curing were performed simultaneously, that is, EPDM was vulcanized with DCP in the presence of PP/ionomer. Addition of ionomer and the application of the dynamic vulcanization were effective in enhancing the miscibility of PP and EPDM. It was found that the addition of ionomer and application of dynamic vulcanization were effective in enhancing the miscibility of PP/EPDM binary blend. This was due to the formation of the thermoplastic interpenetrating polymer network (IPN) of the ternary blends (36). The structure and properties of the ternary blends differed depending on the types and contents of ionomer, that is, the ternary blend containing Na-neutralized ionomer did not show a thermoplastic IPN structure, even if the blends were prepared by dynamic vulcanization. When the contents of ionomer and DCP were 15 parts or 10 parts, respectively, the ternary blend containing Zn-neutralized ionomer clearly showed the behavior of a thermoplastic IPN. Although the Na-neutralized ionomer can form an interpenetrating network between PP and EPDM, the possibility is less than that of Zn-neutralized ionomer because of the monovalent nature of Naþ. The crystallization rate of ternary blends is slower than that of the binary blends, and the ternary blends, which include Zn-neutralized ionomer, showed slower crystallization rate than ternary blends that included Na-neutralized ionomer (35). According to the literature, IPNs that possess physical interlocking at interfaces, strongly restrict crystallization (16,37). This IPN structure is postulated for the dynamically vulcanized EPDM and ionomers, especially for the blends containing Zn-neutralized ionomer. When the ionomer content was 5 wt%, the PP and EPDM blends are incompatible, that is, their phases are separated and the domain of EPDM was peeled off from the continuous matrix of PP (35). For the dynamically vulcanized EPDM and PP/ionomer ternary blends with 15 wt% ionomer, compatibilization was achieved between the PP and EPDM phases. The Zn-neutralized ionomer showed a much better compatibilizing effect than Na-neutralized ionomer. The dynamically vulcanized EPDM and PP binary blends showed somewhat quasicleavage fracture topology, regardless of the DCP contents (38). Figure 14.15 shows SEM micrographs of the fractured surfaces taken around the crack-tip for the dynamically vulcanized EPDM and PP/ionomer ternary blends having 20 wt% ionomers when the DCP content is low. The ionomer-added ternary blend shows no clear fracture surface topology of tough materials even the ionomer contents are 20 wt%, regardless of ionomer types. The ionomer-added ternary blends with high DCP content, however, showed different fracture surface topologies. In Fig. 14.16, SEM micrographs of the fractured surface taken around the crack-tip are shown for the 5 and 20 wt% ionomer-added dynamically vulcanized

426

Polyolefin Blends

Figure 14.15 SEM microfractographs of lightly vulcanized ternary blends: (a) EPDM / PP / Naneutralized ionomer (40/40/20) and (b) EPDM /PP/ Zn-neutralized ionomer (40/40/20). The initial crack length was 8 mm and DCP content is 0.33 phr. (From Reference 38 with permission from Springer.)

ternary blends when the DCP content is high. The ionomer-added ternary blends show typical fracture surface topology of tough materials irrespective of ionomer types, and the trend is clearer when ionomer contents are higher. The micrographs reveal well dimple-ruptured topologies, which are usually observed in tough materials. Careful inspection of Fig. 14.16 shows that the fracture surface of 20 wt% zinc-neutralized ionomer-added dynamically vulcanized ternary blends has most clear dimple fracture topology, exhibiting the toughest characteristics among the blends. The application of dynamic vulcanization and the addition of

Chapter 14 Polypropylene/Ethylene–Propylene–Diene Terpolymer Blends

427

Figure 14.16 SEM microfractographs of highly vulcanized ternary blends: (a) EPDM/PP/Naneutralized ionomer (47.5/47.5/5), (b) EPDM/PP/Na-neutralized ionomer (40/40/20), (c) EPDM/PP/Znneutralized ionomer (47.5/47.5/5), and (d) EPDM/PP/Zn-neutralized ionomer (40/40/20). The initial crack length was 8 mm and DCP content is 1.00 phr (From Reference 38 with permission from Springer.)

428

Polyolefin Blends

ionomer played a synergistic role to enhance the fracture toughness of PP or PP/EPDM blend.

14.3.4 Mechanical and Rheological Properties With increasing cross-link density, a narrowing of EPDM domain size and an increase in the PP ligament thickness were observed by Ellul et al. (29). EPDM domain size is a very important factor for the mechanical properties of EPDM/PP TPV. Smaller EPDM domain provides higher strength and elongation (3). Tensile strength and tension set improves with higher cross-link density. Gupta et al. (39) has studied the effect of dynamic cross-linking on tensile yield behavior of PP/EPDM rubber blends. They prepared blends of PP/EPDM in internal mixer by simultaneous blending and dynamic vulcanization. Dimethyl phenolic resin vulcanized PP/EPDM blends showed higher yield stress and modulus than unvulcanized PP/EPDM blend (Fig. 14.17 and Table 14.1). They found the increase in

Figure 14.17 Change in shape of the stress–strain curves for unvulcanized PP/EPDM blends. (From Reference 39 with permission from John Wiley & Sons, Inc.)

429

Sample notation

33.25 23.50 20.45 16.20 13.00

24.65 20.75 17.80 13.94

00 10 20 30 40

10 20 30 40

25.00 21.60 18.00 14.50

35.00 24.95 23.00 18.02 14.15

Tensile Stress at yield MPa wt % EPDM (YTS) rubber Unaged Aged

24.65 20.75 17.84 13.90

33.25 23.50 20.45 16.20 13.00 25.00 21.60 18.00 14.50

35.00 24.95 23.00 18.02 14.15 21.15 18.03 16.80 13.40

23.95 20.85 18.25 14.60 12.45 22.00 19.40 18.00 13.40

24.62 21.35 20.06 17.25 13.60

Tensile Strength Tesnsile strength at max at break stress, MPa MPa Unaged Aged Unaged Aged

1164 1008 787 690

1240 1050 858 695 576

1195 1082 802 680

1320 1125 924 740 622

56 76 140 350

44 52 66 122 285

Tensile modulus, modulus, (Young’s), Ultimate MPa elongation, Unaged Aged % (UEL)

24 56 99 143

18 20 33 49 100

Area under yield, peak, au

Values of Tensile Parameters for Polypropylene (PP)/EPDM Blends (from Reference 39 with permission from John Wiley & Sons)

Unvulcanized Control blends PP100 PP90EL10 PP80EL20 PP70EL30 PP60EL40 Vulcanized blends PP90EB10 PP80EB20 PP70EB30 PP60EB40

Blend systems

Table 14.1

430

Polyolefin Blends

interfacial adhesion by three-dimensional network is the cause of improvement in mechanical property. Shapes and sizes of EPDM phase were studied by SEM. According to Jain et al. (40), the impact strength of PP increased dramatically when the optimum amount of EPDM was blended. The cross-linking of the EPDM reduced the optimum amount of EPDM. Increased interfacial adhesion due to crosslinking leaded to the reduced EPDM particle sizes and a more uniform particle size distribution. Compared with unvulcanized PP/EPDM blend, the tensile properties of TPV were significantly improved (39,40). It was reported that the deformation is concentrated on the equatorial PP region between the rubber domains perpendicular to the extension direction, and polar ligament PP region between adjacent rubber domains along the extension direction is undeformed (41). As a result, the stress passes through the polar ligament PP region and concentrates on the EPDM domains, and EPDM domains are highly deformed. The high drawability of EPDM domain is the origin of rubber elasticity with ductile thermoplastic matrix. There has been the question why the TPV materials with ductile thermoplastic matrix display rubber elasticity. Several models have been suggested to answer this question (41–47). Inoue group first analyzed the origin of rubber elasticity in TPVs (43). They constructed a two-dimensional model with four EPDM rubber inclusions in ductile PP matrix and carried out the elastic–plastic analysis on the deformation mechanism of the two-phase system by finite-element method (FEM). The FEM analysis revealed that, even at highly deformed states at which almost the whole matrix has been yielded by the stress concentration, the ligament matrix between rubber inclusions in the stretching direction is locally preserved within an elastic limit and it acts as an ‘‘in-situ formed adhesive’’ for interconnecting rubber particles. In a series of works on the TPVs, Boyce et al. also analyzed such unusual elasticity of TPVs by deformation analysis using stress–strain curve of TPVs with a constitutive model and/or simulation (44–46). In particular, Boyce et al. reported the important role of matrix ligament thickness in controlling the initial stiffness and flow stress of the TPVs (45); thinner ligaments lead to earlier matrix yielding and thus earlier formation of the pseudocontinuous rubber phase. Upon formation of the pseudocontinuous rubber phase, the matrix material is seen to accommodate the large straining of the rubber phase by nearly rigid body motion (rotation and translation) of the bulky domains of the matrix; the rubber phase is seen to undergo large contortions as it attempts to deform as an almost continuous network around the ‘‘rigid’’ domains of matrix material. Furthermore, the asymmetry together with the thin matrix ligaments greatly aids the recovery of the material during unloading. Upon unloading, the rubber phase attempts recovery in a rubber-like manner. The bulkier regions of matrix material simply rotate and translate with the recovering rubber domains. The thin ligaments also rotate, but eventually also undergo bending and buckling, which enables the large amount of recovery observed in TPVs. Similar elastic behavior was also reported for nylon 6/EPDM TPVs as well as PP/EPDM TPVs by Oderkerk et al. (47). Hardness and tensile set are the properties usually specified to choose the grade to be used for its end-use properties. TPV hardness can be easily adjusted by

Chapter 14 Polypropylene/Ethylene–Propylene–Diene Terpolymer Blends

431

Figure 14.18 Notched–Izod impact strength versus EPDM content of uncross-linked and dynamically photocross-linked PP/EPDM blends measured at 25 and 30 C: (From Reference 25 with permission from John Wiley & Sons, Inc.)

changing the PP/EPDM composition. The rubber-rich blends can be used as thermoplastic elastomers, and the plastic-rich blends can be applied as rubber-toughened plastics. The mutual interaction and bulk properties are dependent on the composition. Tang et al. (48) prepared a dynamically photocross-linked PP/EPDM rubber thermoplastic elastomer by exposing the elastomer to UV light while melt mixing in the presence of photoinitiation as well as a cross-linking agent. The effect of dynamic photocross-linking and blend composition on the mechanical properties, morphological structure, and thermal behavior of PP/EPDM blends were investigated. Tensile strength, modulus of elasticity, and elongation at break improved slightly after photocrosslinking (Figs. 14.18 and 14.19). The notched Izod impact strength was also enhanced compared to uncrosslinked blend. From SEM, it is found that for uncross-linked PP/EPDM blends, the cavitation of EPDM particles was the main toughening mechanism, whereas for dynamically photocross-linked blends, shear yielding of matrix became the main energy adsorption mechanism. DSC showed a new smaller melting peak at about 150 C together with a main melting peak at about 166 C for each dynamically photocross-linked PP/EPDM blend. Compatibility between EPDM and PP was improved by dynamic photocross-linking. This was studied by dynamic mechanical thermal analyzer (DMTA). Goharpey et al. (49) studied the relationship between the rheology and morphology of dynamically vulcanized PP/EPDM blends. They performed

432

Polyolefin Blends

Figure 14.19 Notched–Izod impact strength versus photocrosslinking time for PP/EPDM (70/30) measured at 25 and 30 C: (From Reference 25 with permission from John Wiley & Sons, Inc.)

morphological study by SEM on cryogenically fractured samples (Fig. 14.20). Rheological behavior and melt viscoelastic properties of the samples were studied by rheometric mechanical spectrometry (RMS) at a temperature of 220 C. Sample showed a significant viscosity upturn and a strong storage modulus that tended to plateau at low shear rates, for 60% EPDM containing sample (Fig. 14.21). These structure findings were attributed to a network structure resulting from agglomerates formed between the cured rubber particles, as evidenced by the morphological features of the samples. Xiao et al. (50) studied the miscibility of PP and EPDM by DMTA, transmission electron microscopy (TEM), and DSC. The result showed that a decrease in the PP content and an increase in the cross-linking density of EPDM in the EPDM/PP blends caused an increase in the glass transition temperature of EPDM, although there is no change of Tg of PP (Fig. 14.22). The degree of crystallinity is decreased (Table 14.2). They found that mechanical properties of blends prepared by single screw extruder is higher than that made by open mill. From TEM, PP phase is seen as a bright phase and EPDM as a dark phase. With the increase in cross-linking density, the interface between the EPDM and PP becomes less defined and EPDM gradually dispersed in the PP phase became a continuous phase. Two different ways of processing to improve interfacial adhesion of PP and EPDM by introducing MAH were attempted by Ha et al. (51); In one way, the in situ

Chapter 14 Polypropylene/Ethylene–Propylene–Diene Terpolymer Blends

433

Figure 14.20 Scanning electron microscopy of the dynamically crosslinked EPDM/PP blend samples of different composites: (a) 20/80; (b)40/60; (c) 60/40, ww. (From Reference 49 with permission from Wiley Interscience.)

434

Polyolefin Blends

Figure 14.21 Results of dynamic (&) viscosity h, storage modulus G0 , versus angular frequency v, of the dynamically crosslinked EPDM/PP blend samples of different compositions; 20/80 (^); 40/60 (!); 60/40 (*) w:w. (From Reference 49 with permission from Wiley Interscience.)

grafting and dynamic vulcanization (ISGV) were performed simultaneously from PP and EPDM with MAH in the presence of DCP in an intensive mixer. In another way, PP was first grafted with MAH and then the PP-g-MAH was blended with EPDM in the intensive mixer in the presence of DCP by the dynamic vulcanization. It was found that the glass transition temperatures (Tg) of both PP and EPDM phases were shifted to higher temperature as the EPDM content increased for the blends prepared by both IGSVand DV methods, mainly due to the cross-linking of EPDM. The higher Tgs and larger storage moduli were observed for the blends prepared by the ISGV

Table 14.2 Physical Properties of EPDM/PP Blends (from Reference 50 with permission from John Wiley & Sons). Rubber/plastica

Amount of cure agentsb

Properties

70/30

60/40

50/50

0/100

0.03

0.12

0.21

0.30

Tc, C Tm, C DHf, Jg1 Xc, %

137 161 24 11

134 161 30 14

132 163 33 16

122 164 92 44

139 159 36 17

137 160 34 16

135 160 32 15

134 161 30 14

a

with 5 phr curing agents EPDM60PP40

b

Chapter 14 Polypropylene/Ethylene–Propylene–Diene Terpolymer Blends

435

Figure 14.22 DMTA curves of EPDM/PP blends with 5 phr of phenolic resin. The mass ratios of rubber to plastic were (-.-) 0/100, (-) 50/50, (—) 60/40, and (. . .) 70/30. (From Reference 50 with permission from John Wiley & Sons, Inc.)

method than those prepared by the DV method, while the morphology showed that the size reduction of dispersed particles in latter blends was larger than that of the former blends. Wang and Cakmak (52) studied the development of structure hierarchy in tubular film blown dynamically vulcanized PP/EPDM blend. The blown films were found to exhibit an unusual asymmetric structure. The PP phase was found to fibrillate at all the outside surface, while the inner surface remained relatively featureless. This was attributed to disproportionally rapid cooling of the outside surface by the air steam blown externally onto the film being extruded. This, in turn, resulted in solidification of very thin PP surface layers that caused their fibrillation under the heavy stress they had to endure. Increase in the blow-up ratio was found to expand this web-like surface texture. As a result of this fibrillation mechanism, the increase of both the blow-up ratio and draw-down ratios was found to reduce the mechanical properties.

436

Polyolefin Blends

14.4 APPLICATIONS OF PP/EPDM BLENDS Because of low cost, high heat deflection temperature (HDT (104 C)), notched impact resistance, improved low temperature impact and flexibility, weather resistance, flame retardancy, and impact resistance, the PP/EPDM blend has got a widespread applications. These unique characteristics of this thermoplastic elastomer blend make it an attractive alternative to conventional elastomer in a variety of markets such as automobile industries, wire, cable insulator, automobile bumpers and fascia, hose, gaskets, seals, weather stripping, among others. The market of PP/EPDM blends has grown dramatically because of its recycling ability and processability by conventional thermoplastic processing equipment. The unique characteristics of thermoplastic elastomer made it an attractive alternative to conventional elastomers in a variety of markets. Liu et al. showed from the experimental blends (53) that materials cost reduction of between 30% to 50% is possible in comparison to commercial products if one applies the PP/EPDM blends to the construction of a basketball court, a tennis court, and a roller hockey rink, which were estimated around $7000, $14,000, and $40,000, respectively. The cost comparison took into account the percentage of rubber or PP used in experimental blend, the exponential factor for a scale-up process and the overall surface area of the specific applications. Among many possible application of this blend two readily feasible applications are roofing and flooring. It was reported that the strategies used by several producers to design commercial TPV formulations are very similar (54). All the EPDMs used in several commercial TPV formulations showed very similar Tg and similar EPDM: oil ratio. They are also believed to have ethylene content of about 60% as supported by the endothermic peaks right after the EPDM-rich phase Tg. Considerable amounts of oil were found in all the formulations, and oil is believed to be preferentially located in EPDM. Oil improves TPV processability, especially in grades with higher crosslinked EPDM parts. Some oil may also remain in the amorphous region of the PP phase, thus improving the elasticity of the blend. The Ellul group reported the shear flow behavior and oil distribution between phases in TPVs (55). The distribution of the high temperature oil between the PP melt and the EPDM was a key parameter because this affected the viscosity of the PP/oil medium. Several PP/oil mixtures were prepared and their viscosity curves were correlated with the neat PP melt viscosity curves by means of shift factors varying with oil concentration. The oil distribution between the PP and EPDM phases was estimated from TEM micrographs of the TPV blends. It was found that the PPs are mixed with oil in different proportions in different TPVs and that the viscosity curves of these mixtures exhibit the same trends in magnitude as the corresponding TPV viscosity curves. Hence, the shear flow of TPVs could be understood more readily in terms of the effective PP/oil medium flow behavior than in terms of the neat PP melt flow. Ellul also reported two important works on the plasticization of TPVs (56,57). She reported that since only the amorphous component of PP is plasticized, the crystalline fraction is not much affected and the upper service temperature range is maintained (56). The plasticized TPVs have an excellent balance of engineering

Chapter 14 Polypropylene/Ethylene–Propylene–Diene Terpolymer Blends

437

properties to 125 C and are much more elastic than unplasticized TPVs due to the suppression of yielding behavior. In another report (57), she found that the use of a PP phase with a high degree of long-chain branching can improve the elasticity of the TPVs with the condition that the branching index at molecular weight greater than 1,000,000 should be less than about 0.6. It was postulated that in the melt and at low frequencies, the long-chain branched PP behaves as a network. In the melt, the dynamically vulcanized alloy behaves as a dual network material: one network being the chemically cross-linked rubber phase and the other being the physical network arising from the high level of long-chain branching in PP. In the solid state, the cocontinuous morphology arising from the choice of longchain branched PP contributes to the enhanced elasticity of the TPVs. PP/EPDM TPV for automotive applications include rack and pinion bellows, air ducts, underhood tubing and connector, underhood hose, plugs, bumpers, grommets, prop rod shaft cover, radiator air deflector, air bag door covers and skins, grips, seals, mats and cupholders, windshield wiper motor cover, fuel line hose, suspension bellows, and weatherseals. EPDM/PP TPV was also used for construction such as door sweep, expansion joints, and window and door seals. Electronic applications include scroll wheel, printer roller, scanner lid, grip, access panel door, speaker surrounds, and holders and bumpers. Also, TPVs are widely used in business machines, power tools, motor mounts, and many other fields. Because of their unique properties and widespread applications, PP/EPDM blends have also been recently subjected to thermoplastic elastomer nanocomposites. Different kinds of nanofiller have been used to prepare nanocomposites like nanoclay, spherical nanoparticles, carbon nanotube, among others (58–72). In addition, the development of environment-friendly polymers is one of recent important issues. In this sense, the use of recycled blends is also an important task. PP/EPDM blends are no exception. For example, Pfaendner et al. (73) studied mechanical recycling of thermoplastics for high value applications, which is directly associated with restabilization. They described the state of the art in processing, long-term heat and light stabilization of recyclates of PP, PP/EPDM blends with examples from packaging, distribution, and automobile and construction industry. They found that because of predamage and impurities, recyclates degrade faster and differently compared with virgin polymers, and therefore specially designed stabilizer systems are required according to previous damage and subsequent application.

14.5

CONCLUSIONS

Thermoplastic elastomeric materials have many important applications including cable and wire especially in mineral, electronic equipment, and automobile industries. The most commonly used method of obtaining thermoplastic elastomer in materials is to toughen plastics by blending rubbers and plastics. Among the most versatile polymer matrices, polyolefins such as PP are the most widely used thermoplastics because of their well-balanced physical and mechanical properties and their easy processability at a relatively low cost, which makes them a versatile material. PP has the disadvantage of

438

Polyolefin Blends

becoming brittle at low temperature, however, because of its high transition temperature and high crystallinity. The best way to improve its impact strength is to blend PP with elastomers particularly EPDM. The blend of PP and EPDM has been prepared by different means, melt mixing, dynamic radiation curing, ultrasonic curing, among others. These PP/EPDM blends have been widely studied from different angles; that is, structure–properties relationship, morphology, mechanical properties, rheology, thermal properties, among others. In this chapter, therefore, we reviewed recent reports on the PP/EPDM blends in terms of thermoplastic polyolefins (TPO) and thermoplastic vulcanizates (TPVs) as commercially important products as well as environmentfriendly use of recycled EPDM or PP for the PP/EPDM blends. The advances made in the plasticization of PP by low Tg diluents through the TPO and/or TPVs, thus changing its low temperature brittleness characteristics, are also worth noting (74). The market of PP/EPDM blends has grown dramatically because of its recycling ability and processability by conventional thermoplastic processing equipment. The unique characteristics of thermoplastic elastomer have made them an attractive alternative to conventional elastomers in a variety of markets. Thus, we introduced some recent practical applications of PP/EPDM blends.

ACKNOWLEDGMENTS The work was supported by the Korea Science and Engineering Foundation (KOSEF) through the National Research Laboratory Program funded by the Ministry of Science and Technology (MOST) (No. M10300000369-06J0000-36910), the SRC/ERC of MOST/KOSEF program (grant #R11-2000-070-080020), and the Brain Korea 21 Project.

NOMENCLATURE BDT DCP DMA DMI DMTA DSC DV EPDM IDc IPN iPP ISGV MAH MMI POM PP PP-g-MAH

Brittle–ductile transition Dicumyl peroxide Dynamic mechanical analysis Dimethyl itaconate Dynamic mechanical thermal analyzer Differential scanning calorimetry Dynamic vulcanization Ethylene-propylene-diene terpolymer Critical interparticle distance Interpenetrating polymer network Isotactic polypropylene In situ grafting and dynamic vulcanization Maleic anhydride Monomethyl itaconate Polarized optical microscopy Polypropylene Maleic anhydride grafted PP

Chapter 14 Polypropylene/Ethylene–Propylene–Diene Terpolymer Blends

R RMS SEM TBT TBTM Tg TEM TPO TPV Zn-SEPDM

439

Tip radius of a notch for impact strength measurement Rheometric mechanical spectrometry Scanning electron microscopy Brittle–ductile transition temperature TBT of PP itself for a given notch Glass transition temperature Transmission electron microscopy Thermoplastic polyolefin Thermoplastic vulcanizate Zn2þ neutralized sulfonated EPDM

REFERENCES 1. E. N. Kresge, Rubber Chem. Technol., 64, 469 (1991). 2. M. T. Payne and C. P. Rader, Thermoplastic elastomers: A rising star, in Elastomer Technology Handbook, N. P. Cheremisinoff (ed.), CRC Press, New York, 1993, pp. 571–582. 3. A. Y. Coran and R. P. Patel, Thermoplastic elastomers based on dynamically vulcanized elastomer— Thermoplastic blends, in Thermoplastic Elastomer, G. Holden, N. R. Legge, R. Quirk, and H. E. Schroeder (eds.), Hanser, New York, 1996, p. 153, Chapter 7. 4. A. L. N. Da Silva and M. B. Coutinho, Polym. Test., 15, 45 (1996). 5. L. Huang, Q. Pei, Q. Yuan, H. Li, and W. Jiang, Polymer, 44, 3125 (2003). 6. N. Da Silva and E. M. B. Coutinho, J. Appl. Polym. Sci., 81, 3530 (2001). 7. M. A. Lopez-Manchado, J. M. Kenny, and M. Y. Pedram, Macromol. Chem. Phys., 202, 1909 (2001). 8. Z. Su, P. Jiang, Q. Li, P. Wei, G. Wang, and Y. Zhang, J. Appl. Polym. Sci., 94, 1504 (2004). 9. C. S. Ha, Y. W. Cho, W. J. Cho, Y. Kim, and T. Inoue, Polym. Eng. Sci., 40, 1816 (2000). 10. F. H. Sanchez, R. Olayo, and A. Manzur, Polym. Bull., 42, 481 (1999). 11. C. S. Ha and S. C. Kim, J. Appl. Polym. Sci., 35, 2211 (1988). 12. C. S. Ha, J. Kor. Inst. Rubber Ind., 25, 203 (1990). 13. T. Zaharescu, R. Setnescu, S. Jipa, and T. Setnescu, J. Appl. Polym. Sci., 77, 982 (2000). 14. T. Zaharescu, M. Chipara, and M. Postolache, Polym. Deg. Stab., 66, 5 (1999). 15. T. Zaharescu and P. Budrugeac, Polym. Bull., 49, 297 (2002). 16. Y. Chen and H. Li, Polym. Eng. Sci., 44, 1509 (2004). 17. T. Zaharescu, Polym. Deg. Stab., 73, 113 (2001). 18. W. K. Fischer, US Patent 3,758,643 (1973). 19. W. K. Fischer, US Patent 3,862,106 (1975). 20. A. Y. Coran and R. Patel, Rubber Chem. Technol., 54, 892 (1981). 21. A. Y. Coran and R. Patel, Rubber Chem. Technol., 56, 210 (1983). 22. L. A. Coettler, J. R. Richwine, and F. J. Wille, Rubber Chem. Technol., 55, 1448 (1982). 23. M. Ishikawa, M. Sugimoto, and T. Inoue, J. Appl. Polym. Sci., 62, 1495 (1996). 24. C. S. Ha, D. J. Ihm, and S. C. Kim, J. Appl. Polym. Sci., 32, 6281 (1986). 25. L. Tang, B. Qu, and X. J. Shen, J. Appl. Polym. Sci., 92, 3371 (2004). 26. A. K. Jain, N. K. Gupta, and A. K. Nagraz, J. Appl. Polym. Sci., 77, 1488 (2000). 27. A. Sengupta and B. B. Konar, J. Appl. Polym. Sci., 66, 1231 (1997). 28. M. D. Ellul, J. Patel, and A. J. Tinker, Rubber Chem. Technol., 68(4), 583 (1995). 29. M. D. Ellul, A. H. Tsou, and W. G. Hu, Polymer, 45(10), 3351 (2004).

440

Polyolefin Blends

30. S. Abdou-Sabet, R. Puydak, and C. Rader, Rubber Chem. Technol., 69, 476 (1996). 31. F. A. Goharpey, A. Katbab, and H. Nazockdast, J. Appl. Polym. Sci., 81, 2531 (2001). 32. H. Nazockdast, F. Goharpey, and A. Katbab, Rubber Chem. Technol., 76, 239 (2003). 33. Y. Yang, T. Chiba, H. Saito, and T. Inoue, Polymer, 39, 3365 (1998). 34. E. Martsuscelli, C. Silvestre, and G. Abate, Polymer, 23, 229 (1982). 35. Y. Kim, W. J. Cho, C. S Ha, and W. H. Kim, Polym. Eng. Sci., 35, 1592 (1996). 36. S. C. Kim, D. Klempner, K. S. Frisch, W. Radigan, and H. L. Frisch, Macromolecules, 9, 258 (1976). 37. Y. Kim, W. J. Cho, and C. S. Ha, Polymer(Korea), 18, 737 (1994). 38. Y. Kim, W. J. Cho, and C. S. Ha, J. Mater. Sci., 31, 2917 (1996). 39. N. K. Gupta, A. K. Jain, R. Singhlal, and A. K. Nagpal, J. Appl. Polym. Sci., 78, 2104 (2000). 40. A. K. Jain, A. K. Nagpal, R. Singhal, and N. K. Gupta, J. Appl. Polym. Sci., 78, 2089 (2000). 41. T. Asami and K. Nitta, Polymer, 45, 5301 (2004). 42. C. S. Ha and S. C. Kim, J. Appl. Polym. Sci., 37, 317 (1989). 43. Y. Kikuchi, T. Fukui, T. Okada, and T. Inoue, Polym. Eng. Sci., 31(14), 1029 (1991). 44. M. C. Boyce, K. Kear, S. Socrate, and K. Shaw, J. Mech. Phys. Sol., 49 (95), 1073 (2001). 45. M. C. Boyce, S. Socrate, K. Kear, O. Yeh, and K. Shaw, J. Mech. Phys. Sol., 49(6), 1323 (2001). 46. M. C. Boyce, O. Yeh, S. Socrate, K. Kear, and K. Shaw, J. Mech. Phys. Sol., 49(6), 1343 (2001). 47. J. Oderkerk, G. Groeninckx, and M. Soliman, Macromolecules, 35(10), 2002. 48. L. Tang, B. Qu, and X. Shen, J. Appl. Polym. Sci., 92, 3371 (2004). 49. F. Goharpey, N. Nazockdast, and A. Katbab, Polym. Eng. Sci., 201, 84 (2005). 50. H. W. Xiao, S. Q. Huang, T. Jiang, and S. Y. Cheng, J. Appl. Polym. Sci., 83, 315 (2002). 51. C. S. Ha, Y. Cho, J. H. Go, and W. J. Cho, J. Appl. Polym. Sci., 77, 2777 (2000). 52. M. D. Wang and L. Cakmak, Rubber Chem. Technol., 74, 761 (2001). 53. http//www.sperecycling.org/PDF%20Files/0157.PDF. 54. M. Montoya, J. P. Tomba, J. M. Carella, and M. I. Gobernado-Mitric, Eur. Polym. J., 40, 2757 (2004). 55. K. Jayaraman, V. G. Kolli, S. Y. Kang, S. Kumar, and M. D. Ellul, J. Appl. Polym. Sci., 93(1), 113 (2004). 56. M. D. Ellul, Rubber Chem. Technol., 71(2), 244 (1998). 57. M. D. Ellul, Rubber Chem. Technol., 76(1), 202 (2003). 58. E. P. Giannelis, Adv. Mater., 8, 29 (1996). 59. S. S. Ray and M. Okamoto, Prog. Polym. Sci., 28, 1539 (2003). 60. X. Li, T. K. Kang, W. J. Cho, J. K. Lee, and C. S. Ha, Macromol. Rapid Commun., 22, 1306 (2001). 61. S. R. Lee, H. M. Park, H. T. Lim, T. K. Kang, X. Li, W. J. Cho, and C. S. Ha Polymer, 43, 2495 (2002). 62. H. M. Park, X. Li, C. Z. Jin, C. Y. Park, W. J. Cho, and C. S. Ha, Macromol. Mater. Eng., 287, 553 (2002). 63. J. K. Mishra, I. Kim, and C. S. Ha, Rubber Chem. Technol., 78, 42 (2005). 64. G. J. Hwang, J. W. Park, I. Kim, C. S. Ha, and G. H. Kim, Macromol. Res., 14, 179 (2006). 65. H. M. Jeong, M. Y. Choi, and Y. T. Ahn, Macromol. Res., 14, 312 (2006). 66. S. M. Garces, D. J. Moll, T. Bicerano, R. Fibiger, and D. G. McLeod, Adv. Mater., 12, 1835 (2000). 67. K. Y. Lee and L. A. Goettler, Polym. Eng. Sci., 44, 1103 (2004). 68. J. K. Mishra, I. Kim, G. H. Kim, and C. S. Ha, J. Polym. Sci. Part B Polym. Phys., 42, 2900 (2004). 69. S. Mehta, F. M. Mirabella, K. Rufener, and A. Befna, J. Appl. Polym. Sci., 92, 928 (2004). 70. B. B. Khatua, D. J. Lee, H. Y. Kim, and J. K. Kim, Macromolecules, 37, 2454 (2004). 71. J. K. Mishra, I. Kim, and C. S. Ha, Polymer, 46, 1995 (2005). 72. L. Valentine, J. M. Kenny, and M. A. Lopez-Manchado, J. Appl. Polym. Sci., 89, 2657 (2003). 73. R. Pfaendner, H. Herbst, K. Haffmann, and T. Sitek, Angew. Makromol. Chem., 232, 4140 (1995). 74. M. D. Ellul, Plastic Rubber Compos. Process. Appl., 26(3), 137 (1997).

Chapter

15

Ethylene–Propylene–Diene Rubber/Natural Rubber Blends Soney C. George1 and Sabu Thomas2

15.1 INTRODUCTION The scientific and commercial progress in the area of polymer blends during the past decades has been tremendous and was driven by the realization that, by blending, new material can be developed and can be implemented more rapidly and economically. Blending of polymers is technological way of providing materials with full set of desired specific properties at the lowest price. Blending also benefits the manufacturer by offering improved processability, product uniformity, quick formulation changes, plant flexibility, and high productivity. Many elastomers that are dissimilar in chemical structure are blended to improve processability, performance, durability, and physical properties. The elastomer blends generally exhibit poor mechanical properties due to incompatibility and gross phase separation (1,2). In addition to the poor interfacial adhesion caused by the thermodynamic incompatibility, these blends usually present cure rate incompatibility because of the differences between the reactivity of the elastomers with the curing agents and/or differences in solubility of the curatives in each elastomer phase (3,4). Extensive research work has been carried out in the field of elastomer blends for the past few decades (5–10). NR/EPDM blends are one of the elastomer blends, which gained lot of commercial interest during these days due to their excellent properties. 1 Department of Basic Science, Amal Jyothi College of Engineering, Koovapally, Kottayam 686518, Kerala, India 2 School of Chemical Sciences, Mahatma Gandhi University, Priyadarshini Hills, Kottayam 686560, Kerala, India

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

441

442

Polyolefin Blends

Natural rubber vulcanizates exhibit good elasticity and good tensile strength besides high resilience and excellent wear resistance. On the contrary, natural rubber is not recommended for outdoor applications where maximum resistance to sunlight, ozone, oxygen, or heat aging are major factors. EPDM exhibits excellent resistance to ozone, oxidants, and severe weather conditions, thereby making it an outstanding material for outdoor applications. But the resilience and tensile properties of EPDM are lower than that of natural rubber. Blending a suitable amount of low unsaturated ethylene–propylene–terpolymer (EPDM) into a diene rubber has been found to improve both heat and ozone resistance (11–15) besides its improvement in chemical resistance, mechanical properties, and building tack properties along with decrease in compression set (16). Apart from property benefits, the blends of EPDM and NR are also attractive from an economic point of view. This is due to the relatively high expense of EPDM. However, the difference in olefin concentration of EPDM and natural rubber results in a cure-rate misbatch leading to an incompatible blend. This has been recognized to cause both inferior static and dynamic mechanical properties such as poor tensile strength, fatigue resistance, and high hysteresis in the rubber blend (17). The chapter will review the preparation, properties, and application of NR– EPDM blends. Several issues related to cure-rate mismatch, compatibility problems, morphology control by introduction of compatibilizers will be discussed.

15.2 MISCIBILITY, COMPATIBILITY, AND THERMODYNAMICS OF POLYMER BLENDING Elastomers with similar polarities and solubility characteristics can be easily combined to produce miscible polyblend (18). Miscible polymer blend is a polymer blend, which is homogeneous down to the molecular level and associated with the negative value of the free energy of mixing and the domain size is comparable to the dimensions of the macromolecular statistical segment. Complete miscibility in a mixture of two polymers requires that the following condition be fulfilled (19): DGm ¼ DHm TDSm < 0

ð15:1Þ

where DGm , DHm , and DSm are the Gibb’s free energy, the enthalpy and entropy of mixing at temperature T, respectively. The value of TDSm is always positive since there is an increase in the entropy on mixing. Therefore, the sign of DGm always depends on the value of the enthalpy of mixing DHm . The polymer pairs mix to form a single phase only if the entropic contribution to free energy exceeds the enthalpic contribution, that is, DHm < TDSm

ð15:2Þ

The miscibility may be achieved through specific interaction, for example, (i) repulsive, (ii) dipole–dipole, (iii) ion–dipole, (iv) ion–ion, hydrogen bonding, and (v) chemical reaction between active blend constituents (20). The commercially useful polymer–polymer combination is linked by intermolecular forces such as

Chapter 15 Ethylene–Propylene–Diene Rubber/Natural Rubber Blends

443

van der Waals forces or dipole moments and exhibit sufficient thermodynamic compatibility to prevent the polymer phases from separating during melt processing (21). Blends of elastomers having similar polarity and cure rate exhibit almost additive properties, whereas dissimilar elastomers result in blends with inferior properties (22). This property failure has been ascribed mainly to three types of incompatibilities that exist between dissimilar elastomers: (1) thermodynamic incompatibility involving phase separation on molecular scale (23,24), (2) viscosity mismatch causing delay or even preventing the formation of coherent blends (25), and (3) cure-rate mismatch due to imbalance in unsaturation levels of the elastomers. Among these, viscosity mismatch can be improved through proper blending processes by adjusting the raw polymer viscosities, extender oil, and filler concentrations. Thermodynamic incompatibility can be alleviated to some extent by reducing the interfacial energy through the creation of microdomains and subsequent adhesion between the elastomeric phases or by cross-linking the phases across the interfaces during vulcanization (26). Several attempts have been made to minimize phase separation and to increase interfacial adhesion. The addition of a third component, that is, a compatibilizer, which is able to act at the interface between the phases is also a good approach to obtain polymer blends with a more homogeneous morphology and improved mechanical properties (27). The compatibilization of elastomer blends may be successfully performed by using a functionalized polymer as a reactive compatibilizer (28–30).

15.3

BLEND PREPARATION

The basic materials used for the preparation of blends are EPDM and natural rubber. The structure and general characteristics of EPDM and NR are given in Fig. 15.1 (31) and Table 15.1. In general, the masticated NR and EPDM were mixed together with other ingredients including compatibilizers and homogenizing agents in an internal mixer or open mills. The rubber compound is cured in an electrically heated press at 160 C for optimum cure time, which is determined by rheometer. In order to study the

Figure 15.1 Molecular structure of (a) ethylene–propylene rubber and (b) natural rubber. (From Reference 31 with permission from Elsevier Science Ltd.)

444

Polyolefin Blends Table 15.1 General Characteristics of EPDM and Natural Rubber. Properties Tensile strength,ps: Tg C Specific gravity Tear resistance Adhesion to metal Abrasion resistance Compression set Rebound cold Rebound hot Permeability to gases Dielectric strength Electrical insulation Acid resistance Heat resistance Sunlight resistance Ozone resistance

EPDM

Natural rubber

Over 3000 0.58 0.85 Good Good Good Good Good Good Poor Excellent Good Good Excellent Excellent Excellent

Over 3000 0.75 0.93 Good Excellent Excellent Good Excellent Excellent Fair Excellent Excellent Fair Good Poor Fair

cure-rate mismatch the vulcanizates are reported to be prepared in two different methods (32–33). These include one-stage vulcanization and two-stage vulcanization processes. In the one-stage vulcanization process, NR and EPDM are first masticated separately and then mixed with each other. Additives such as ZnO, stearic acid, carbon black (or silica), and process oil are added. The mix thus obtained is allowed to cool to room temperature. Finally, coupling agent known as DIPDIS and sulfur are added to the mix on the cooled mill. The stocks are finally cured under pressure at 160 C (32–33). In the two-stage process, NR and EPDM are first masticated separately. Then, additives such as ZnO, stearic acid, DIPDIS, and sulfur are incorporated in the EPDM. The compounded EPDM mix is then heated at 160 C in the hydraulic press for the predetermined time to yield the grossly undercured mix. The undercured mix is then blended with NR to the required blend ratio. The blend compound is finally vulcanized to the optimum cure time values (32–33).

15.4 COVULCANIZATION Cure-rate mismatch is extreme when the blends constitute high unsaturated diene rubbers like NR, and low unsaturation rubbers like EPDM. It involves the migration of polar curatives from the low unsaturation phase to a more polar high unsaturation phase, further undercuring the low unsaturation phase (11–13,34) and it has been shown that unvulcanized EPDM exists in the vulcanized blend with NR (35). Several approaches have been made to obtain a cocured blend vulcanizate of NR–EPDM without sacrificing the physical properties by (1) increasing the unsaturation of EPDM elastomer so that the cure rate becomes at par with NR or other diene rubbers

Chapter 15 Ethylene–Propylene–Diene Rubber/Natural Rubber Blends

445

(36–38); (2) curing with peroxide and polysulfide (38); (3) halogenating the rubber (39); (4) effecting partial prevulcanization (36,40); (5) using accelerators with long hydrocarbon chains (41,42); and (6) grafting accelerators or sulfur donors ( 43,44) or polydiene chains (45,46) in EPDM. Incorporation of lead dithiocarbamate into EPDM, before it is mixed with NBR, has been reported to yield an improved blend (11,13) vulcanizate. EPDM has been reported to react with N-chlorothioamides (47) to produce a macromolecular cure retarder, making it compatible with NR. EPDM modified with maleic anhydride (48) has been observed to produce blend vulcanizate of improved physical properties. The use of polyoctanomer (TOR) (27,49) and halobutyl rubber (50) as the compatibilizer for NR–EPDM has also been reported. It is also reported that proton NMR spectroscopy is a good tool in examining the curative migration and cross-link density within the NR phase of the NR/EPDM blends (51). The spectral line widths are found to increase smoothly with cross-link density. Analysis of this signal leads to a measure of line broadens H%. H% increases with the increase in cross-link density (52) and H% of any of the blends was found to be greater than the H% of the appropriate NR control vulcanizate. This analysis showed that NR components of the blends are utilizing 84–92% of the curatives in the compound, that is, about 80% of the curatives initially located in the EPDM phase are diffusing to the NR phase during vulcanization. Similarly, the swollen state FT– NMR spectroscopic method of blend analysis (53) has been reported to be effective in analyzing the cross-linking density in NR/EPDM blends. This study shows that the presence of chemical modification in the EPDM has a dramatic effect on the crosslinking in the EPDM phase, but only a minor on that in the NR phase. But the overall cross-link density in the blend is increased. Despite these changes, there remains a large imbalance in the cross-link distribution in favor of the NR phase in both modified blends, yet the physical properties are good. Ghosh et al. (32) had developed a new vulcanization technique to mitigate the cure-rate mismatch of NR/EPDM blends by introducing a multifunctional additive, namely, DIPDIS. In order to control the curative diffusion from the nonpolar to polar elastomer, EPDM has made more polar through its reaction with DIPDIS. Although EPDM used in the investigation has a low unsaturation content (5%), it will react with DIPDIS and yield rubber-bound intermediates as shown in Fig. 15.2 (32). It was

Figure 15.2 Reactions of bis(diisopropyl)-thiophosphoryl disulfide (DIPDIS) with ethylene– propylene–diene rubber (EPDM) and zinc oxide. (From Reference 32 with permission from John Wiley & Sons.)

446

Polyolefin Blends

reported that DIPDIS is capable of raising the maximum rheometric torque values of both NR and EPDM as well as the physical properties of the blend vulcanizate. The properties are further enhanced by the two-stage vulcanization. The grossly undercured material obtained in the first stage favorably contains a higher amount of reactive fragments compared with that in one-stage vulcanization. The intermediates thus formed are expected to combine with NR in the second stage of the procedure, the resultant effect being the generation of more interrubber linkages and formation of novel rubber blends of significantly improved physical properties. There is significant improvement in modulus, tensile strength, elongation at break, and cross-linking value over those obtained in one-stage vulcanization of the corresponding blends.There is more coherency and homogeneity in the blend composition of two-stage vulcanizates. The cure–rate mismatch problem could thus be solved through the formation of rubber bound intermediates with a DIPDIS, thereby restricting the curative migration from EPDM to NR. The blend morphology as revealed by SEM studies (Fig. 15.3) accounts for significant improvement in physical properties, particularly in two-stage vulcanizates.

Figure 15.3 Scanning electron micrographs of tensile fractured surfaces of the vulcanizates cured at 160 C; (a) 75:25 unsaturated natural rubber–ethylene–propylene–diene (NR–EPDM) blend (one-stage) at 500 ; (b) 75:25 NR–EPDM blend (two-stage) at 500 ; (c) 50:50 NR–EPDM blend (one-stage) at 750 ; (d) 50:50 NR–EPDM blend (two-stage) at 750 . (From Reference 32 with permission from John Wiley & Sons.)

Chapter 15 Ethylene–Propylene–Diene Rubber/Natural Rubber Blends

15.5

447

FILLER DISTRIBUTION IN NR/EPDM BLENDS

The distribution of filler is a major problem in NR/EPDM blends due to its polarity difference. The effect of carbon black and silica on the physicomechanical properties of the covulcanized NR–EPDM rubber blends has been reported (33). The blends rich in NR content exhibit comparatively better results due to inefficient cocure by curative migration and filler transfer (33) as well as the lower unsaturation of EPDM compared with the high unsaturation of NR. All these factors lead to further weakening of the already weakly reinforced EPDM phase by the dearth of polar curatives. Evidently, vulcanizates with poor physicochemical properties are being produced. As discussed earlier, in two-stage vulcanization EPDM is more polar due to the pendant moieties of DIPDIS (Fig. 15.2). Incorporation of carbon black into that modified EPDM matrix restricts the transfer of filler to the NR phase. In all the cases, significant improvement in tensile strength and elongation at break are observed, compared with one-stage vulcanization; however, modulus and cross-linking values show an opposite trend because of the longer cure time used for one-stage samples. Though the cross-linking density is more in one-stage vulcanizates, the properties such as tensile strength and elongation at break values decrease due to the absence of interfacial cross-linking between the components of the blend. The blend vulcanizates exhibit highest modulus, tensile strength, elongation at break, and cross-linking values, and least weight loss in the swelling experiment. These better properties are as a result of the curative fixation on the EPDM backbone. This fixation, in fact, lowers filler transfer from EPDM to NR. Silica-filled samples exhibit faster cure rate compared with gum (32) and black-filled vulcanizates of similar composition. This difference is due to the reaction between the fragments of DIPDIS and the silanol (Si-OH) groups present at the silica surfaces (Fig. 15.4). The isopropyl moiety (-OR) in DIPDIS reacts with the silanol group of silica through the elimination of isopropyl alcohol (54) facilitating the rubber–filler interaction and thus resulting in faster curing. It is evident that inefficient interfacial cross-linking between NR and EPDM occurs in the case of silica-filled systems subjected to one-stage

Figure 15.4 Reaction of silica with pendant DIPDIS fragments. (From Reference 33 with permission from John Wiley & Sons.)

448

Polyolefin Blends

vulcanization, where both curative and filler migration from EPDM to NR phase occurs during vulcanization. Distribution of silica in the NR matrix is facilitated by the interaction of the protein component of NR with the hydroxylated surface of the precipitated silica. As the mixing progresses, the viscosity of the NR phase is reduced gradually. Thus, highly polar silica opts for its migration from high viscosity EPDM to low viscosity NR phase. In the early stage of vulcanization, pendant DIPDIS fragments might react with silica to form large EPDM–DIPDIS (fragment)–SiO2 aggregates, which in turn facilitates the reaction with NR. Improvement in the physical properties of the vulcanizates is thus the outcome of the interfacial cross-linking. Addition of silica in the modified EPDM makes the latter substantially polar and in this way restricts the filler transfer to the NR phase. All these facts lend support to the formation of coherent and homogeneous blend systems of practical importance. The SEM micrographs indicate interfacial chemical bridges in the blend vulcanizate and thus corroborates the results obtained.

15.6 MORPHOLOGY OF NR/EPDM BLENDS The morphology and properties of polymer alloys and blends can be controlled via phase separation or phase dissolution during cure (55). Phase dissolution in elastomer blending is the key factor for achieving optimal morphology having a synergistic effect in the polymer blends. The morphology and properties of NR/EPDM blends could be controlled through compatibilization and covulcanization to promote phase dissolution during cure. The morphology of NR/EPDM blends depends upon the type of compatibilizer added. Hsich reported (55) that a synergistic result in properties of blends was achieved by having interconnected phase morphology. The domain sizes of the blends in this study are 3–5 mm before cure and 1–2 mm after cure. These blends showed superior mechanical properties than those of the individual NR or EPDM, after a long term of high temperature aging. Figure 15.5a and b represents the phase morphology of NR/EPDM blends of 70/30 wt% without compatibilizer and with 2.5% EPDMSH compatibilizer. Figure 15.5c and d represents NR/EPDM blends 60/40 wt% without and with compatibilizer 2.5% EPDMSH (56). The light region corresponds to the NR phase, and the dark points are related to the unstained EPDM phase. In the case of NR/EPDM (60:40 wt%) blend, the presence of EPDMSH resulted in a more homogeneous morphology, but the differences are marginal. On the contrary, the addition of TOR reduced the domain size (27) of the dispersed EPDM particles and it becomes uniform and spherical (Fig. 15.6 b) compared to the uncompatibilized, that is, in the absence of TOR, the dispersed particles are generally large, the size distribution is broad, and the shape of particles is very irregular (Fig. 15.6a). The irregular large particles over 20 mm coexist with spherical small particles below 1 mm. The influence of various compatibilizers on the morphology of 50/50 NR/EPDM blends is shown in Fig. 15.7 (57).

Chapter 15 Ethylene–Propylene–Diene Rubber/Natural Rubber Blends

449

Figure 15.5 SEM micrographs of NR/EPDM (a and b) 70/30 wt% and (c and d) 60/40 wt% blends. Micrographs (a) and (b) are related to compatibilized blends with 2.5 phr of EPDMSH. (From Reference 56 with permission from John Wiley & Sons.)

Figure 15.7a shows that the two phases are with irregular domain sizes and shapes. This indicates that the NR/EPDM blends were completely immiscible, large EPDM domains being dispersed in the NR matrix. The average domain size of the dispersed phase was 4.1 mm. The compatibility of the NR/EPDM system was improved by the addition of a compatibilizer, as can be seen in Fig. 15.7b–g; the treatment resulted in noticeable surface hardening, and the physical changes in the surface were expected to influence physically both the deformation and adhesion of the two rubbers, that is, the compatibilizers improved both the morphology and compatibility of the blends because of the reduction in the interfacial tension between EPDM and NR rubbers. The size of the dispersed phase (EPDM) domain decreased with the addition of compatibilizers, and no gross phase separation was present in the blends (Fig. 15.7). For NR/BR/EPDM, the domain size was approximately 3.8–1.26 mm; NR/PVC/EPDM, 2.7–0.75 mm; NR/chlorosulfonated PE/EPDM, 2–0.75 mm; NR/g-radiation/EPDM 4–1.5 mm; and NR/MAH/EPDM.1– 0.25 mm. These results are in agreement with the observations of Anastasiadas and Koberstein (58) and Meier (59), who reported that compatibilizers reduced the phase domain size.

450

Polyolefin Blends

Figure 15.6 SEM micrographs of the NR/EPDM/TOR blends: (a) 70/30/0; (b) 70/30/10. (From Reference 27 with permission from John Wiley & Sons.)

15.7 COMPATIBILIZATION OF NR/EPDM BLENDS The properties of NR/EPDM blends are not very good due to their incompatibility (60). Therefore, EPDM was modified by several modifying agents such as, TOR (27), EPDMTA, and EPDMSH (56,61), g-rays, EPDM-g-MAH, polybutadiene rubber (BR), chlorinated rubber, chlorosulfonated polyethylene (SPE), and poly(vinyl chloride) (PVC), (57,62) in order to make it more compatible with natural rubber. The effect of concentration of trans-polyoctene rubber as a compatibilizer in various properties of NR/EPDM was investigated (27). The blend composition was fixed at NR/EPDM 70/30 with varying concentration of compatibilizer as 0,5, 10,15, and 20. As TOR having much lower viscosity compared to NR and EPDM, it will locate at the interface of NR and EPDM phase. So the TOR at the interface region reduces the interfacial tension between the incompatible rubbers, which will facilitate the mixing of the rubber blends, thereby improvement in compatibility. The morphology of the compatibilized and uncompatibilized blends (NR/EPDM: 70/30) shown in Fig. 15.6 is in agreement with this observation. Compatibilization is more effective in this case

Chapter 15 Ethylene–Propylene–Diene Rubber/Natural Rubber Blends

451

Figure 15.7 SEM micrographs: (a) 50/50 NR/EPDM without a compatibilizer, (b) NR/BR/EPDM, (c) NR/PVC/EPDM, (d) NR/chlorinated rubber/EPDM, (e) NR/chlorosulfanted PE/EPDM, (f) NR/gradiation/EPDM, and (g) NR/MAH/EPDM. (From Reference 57 with permission from John Wiley & Sons.)

452

Polyolefin Blends

compared to the use of EPDMSH (56). By using EPDMSH for compatibilization purpose, more homogeneous morphology was observed. The tensile modulus of the blends is increased gradually, with increasing TOR content; however, tensile strength and elongation at break of the blends are decreased in the presence of TOR. In the case of NR/EPDM blends with 2.5% EPDMSH compatibilizer, tensile strength, and elongation at break are improving in all compositions but in 70/30 composition blends with 5% of TOR having higher tensile strength. The effect of mercaptoand thioacetate-modified EPDM (61) on the curing parameters and mechanical properties of natural rubber/EPDM blends were reported. In this study, 2,20 dithiobisbenzothiazole (MBTS) was employed as a conventional accelerator, whose proportion in the rubber formulation has been varied, in order to observe the performance of these functionalized copolymers, not only as a compatibilizing agent but also as a secondary accelerator (61). Both EPDMTA and EPDMSH resulted in increase of the ultimate tensile strength of the vulcanized blends. Nevertheless, the efficiency of EPDMSH was superior, which was attributed to the combination of several factors: (a) increase of the cross-link density, (b) the vulcanization of some fraction of the EPDM phase, thus promoting the covulcanization in some extent, and (c) the reactive compatibilization, as a consequence of chemical reactions between the mercapto groups and the rubber matrix. EPDMTA was not so effective in improving the mechanical performance or cross-link density of the blends because of the lower reactivity of the thioacetate groups. However, it was efficient in increasing the aging resistance of the corresponding blends. g Rays at radiation doses of 6 and 8 k gray were found to be suitable for creating cross-links necessary for the mixing homogeneity of NR and EPDM (57). Similarly, great improvement in the homogeneity of the blends was achieved by the use of 10 phr of compatibilizers like EPDM-g-MAH, polybutadiene rubber, chlorinated rubber, chlorosulfonated polyethylene and poly(vinyl chloride) (57). This is also evident from the morphology of the compatibilized blends shown in Fig. 15.7a–g. EPDM was modified with maleic anhydride and blended with natural rubber. A concentration of 10 phr of MA-g-EPDM improved the compatibility of NR/EPDM blends and thereby led to finer morphology (63). The degree of compatibility of NR/ EPDM blend was improved by adding homopolymers and copolymers of acrylonitrile and N-(4-chlorophenyl) acrylamide (64). It was found that a fairly good compatibility had been achieved by using polyacrylonitrile. NR/EPDM compatibility was improved by blending EPDM with epoxidized natural rubber (65).

15.8 MECHANICAL AND VISCOELASTIC PROPERTIES 15.8.1 Mechanical Properties 15.8.1.1

Effect of Blend Ratio

Blend composition has a strong influence on the mechanical properties of NR/EPDM blends (66). Figure 15.8 shows that tensile strength; stress at 100% and yield strain is increasing with the increase in the concentration of EPDM in the blend. This is attributed to the reinforcing effect of EPDM domains in the NR matrix. The

Chapter 15 Ethylene–Propylene–Diene Rubber/Natural Rubber Blends

453

Figure 15.8 The effect of blend composition on the tensile and yield strain of NR/EPDM blends. (From Reference 66 with permission from Sage Publications Ltd.)

properties such as fatigue life and Young’s modulus increased while strain energy and cross-link density decreased (Table 15.2). 15.8.1.2

Effect of Compatibilization

Compatibilizers have a key role in improving the compatibility between the blend components and thereby enhancing the mechanical properties. Several reports are available in connection with improving the properties of NR/EPDM blends by compatibilization. George et al. (67) has reported on the improvement of mechanical properties of NR/EPDM blends by precuring EPDM prior to blending. Blends of Table 15.2 Mechanical Properties of Uncompatibilized NR/EPDM Blends. NR/EPDM blend ratio Properties Stress at 100% strain, MPa Tensile strength, MPa Rupture strength, MPa Yield strain,% Rupture strain% Young’s modulus, N mm2 Equilibrium swelling,% Soluble fraction,% Cross-link density 104, g mol1 Strain energy, MJ m3 No. of fatigue cycles 102

100/0 0.3 2.75 2.73 334 345 0.81 304 3.4 1.18 1.28 32

75/25 0.33 3.48 3.45 405 363 0.97 420 3.8 0.629 1.05 40

[From Reference 66 with permission from Sage Publications Ltd.]

50/50 0.35 3.92 3.89 495 379 1.05 431 4.2 0.517 0.89 45

25/75 0.4 4.35 4.31 515 394 1.12 450 4.8 0.365 0.79 49

100/0 0.53 3.32 3.28 315 359 0.94 392 3.7 0.9 1.12 43

454

Polyolefin Blends

Table 15.3 Mechanical Properties of NR/EPDM Blends as a Function of Composition and Compatibilization s aB;MPa NR/EPDM 100/0 80/20 70/30 60/40 0/100

ebB;%

c

d

c

d

11.5 10.3 10.5 9.4 3.4

9.6 13.0 13.5 15.0 5.5

750 850 900 600 100

780 830 840 700 200

a

Ultimate tensile strength.

b

Elongation at break.

c

Blends without compatibilizer.

d

Blends with 2.5 phr of EPDMSH.

[From Reference 56 with permission from John Wiley & Sons.]

NR/modified EPDM, in which EPDM was modified by pendant sulfur, exhibited improved endurance to repeated stress over that of covulcanized EPDM–NR rubber blends (68). The effects of ethylene and diene contents in EPDM, blend ratio, dicumyl peroxide curing system on the physical properties, interfacial adhesion force, and dynamic crack growth were examined (69). As the ethylene and diene contents in EPDM increased, the physical properties, such as dynamic cut growth, adhesion to other component were also increased. The mechanical properties of the blends are compared to those of the pure components in Table 15.3 (56). The ultimate tensile strength of noncompatibilized blends is lower than that of pure NR, as expected since these blends are incompatible. The addition of 2.5 phr of EPDMSH resulted in an improvement of this property for NR/EPDM blends. The values found for the compatibilized blends were even higher than that observed for pure NR, indicating a synergistic behavior with the compatibilization. Concerning the elongation at break, the compatibilization did not affect substantially this property, except for the system containing 100% of EPDM, where a significant increase was observed. The improvement of the tensile strength of the blends with the addition of EPDMSH is attributed to the interfacial action of this component associated to an increase of the cross-link degree. Abou-Helal and El-Sabbagh (66) found that compatibilizers like EPDM-g-MAH, chlorinated rubber, polybutadiene rubber, chlorosulfonated polyethylene, and so on improved the mechanical properties of NR/EPDM. A regression analysis was employed to correlate the mechanical properties, fatigue life, and strain energy with respect to blend ratio. It was found to obey the formula Y ¼ A þ BX, where Y is the mechanical properties, fatigue life, or strain energy, X is the blend ratio, and A and B are material constants. In the presence of compatibilizers, equation becomes Y ¼ AX B , where Y represents mechanical properties, and X is the blend ratio, and A and B are material constants. The stress strain curves of 50/50 NR/EPDM blends without and with compatibilizers are shown in Fig. 15.9. The behavior goes from a polymer that almost

Chapter 15 Ethylene–Propylene–Diene Rubber/Natural Rubber Blends

455

Figure 15.9 Stress–Strain curves for 50/50 NR–EPDM blend without and with different compatibilizers. (From Reference 66 with permission from Sage Publications Ltd.)

yields with maximum and present strain hardening to polymer that nearly yields and brakes at deformations giving a large stress for different compatibilizers. 15.8.1.3

Effect of Homogenizing Agent

The concentrations of Ultrablend 4000 homogenizing agent of 0, 3, 5, and 7 phr were added into NR/EPDM: 70:30 blends (70). It was found that the tensile strength of blends was increased when a small amount of Ultrablend 4000 (3 phr) was added. It can be seen that tensile strength increases in the blends containing 3 and 5 phr Ultrablend 4000. However, this decreased when more Ultrablend 4000 (7 phr) was added. This could be explained in terms of the compatibility gained by the addition of Ultrablend 4000, which improves the compatibility between the matrix phase (NR) and dispersed phase (EPDM). When the sample was stretched in the tester, the stress was transferred from the matrix phase (NR) to the dispersed phase (EPDM) through the homogenizing agent. The mechanical properties such as tensile strength and elongation at break were gradually improved when 3 and 5 phr of Ultrablend 4000 were added. On the contrary, when Ultrablend 4000 was further added to 7 phr, tensile strength was poor. This could be due to the agglomeration of the excess amount of Ultrablend 4000 to become another phase. This new phase induces slippage or weak points between the matrix and dispersed phases, yielding a lower tensile strength. The high value of tensile strength and elongation at break at 5 phr of homogenizing agent shows a good adhesion between the phases of NR and EPDM at this particular concentration.

456

Polyolefin Blends 40

Tear strength, N mm–1

Unaged 30 Aged 20

10

0 With Si 30 phr

With CB 3 phr

With Si 30+CB 3 phr

70:30 NR/EPDM blend with filters

Figure 15.10 Comparison of tear strength of the NR/EPDM blends containing carbon black (3%), silica (30%), and carbon black/silica (3:30). (From Reference 70 with the permission from John Wiley & Sons.)

15.8.1.4

Effects of Fillers

Addition of silica was reported to improve the mechanical properties, such as tear strength and hardness of NR/EPDM blends (Fig. 15.10) (70). Blends with 30 phr of silica show higher tear strength and hardness compared with blends with 3 phr of carbon black. Synergistic improvement in tear strength and hardness is observed in presence of both carbon black (3 phr) and silica (30 phr).When the amount of Ultrablend 4000 was kept constant at 5 phr, it was found that the silica particles absorbed a greater concentration of Ultrablend 4000 than did the carbon black particles, resulting from their higher surface area and thereby better properties.

15.8.2 Dynamic Mechanical Properties The dynamic mechanical properties of NR/EPDM/TOR (27) and NR/EPDM/ EPDMSH (56) are shown in Fig. 15.11a and b. In Fig. 15.11a, it is clear that by the addition of TOR, dynamic elastic modulus (E0 ) of a blend over the measured temperature range is increased whereas the hysteresis (tan d) is not increased. A single tan d peak is observed, despite the phase-separated structure, which is due to the similarity of the glass-transition temperature of the blend components. This indicates that TOR increases rigidity of blend vulcanizates without substantial change in heat buildup under dynamic stress. In Fig. 15.11b (A), noncompatible blend displays only transition at a temperature. This is because of the presence of noncross-linked rubber mainly constituted by EPDM phase due to its lower cure rate as compared to NR phase. Considering the higher chain mobility of the noncross-linked EPDM, the glass transition temperature is expected to be shifted toward lower temperatures when compared to the same vulcanized rubber. Therefore, the single tan d peak in this blend can be attributed to the vulcanized NR

Chapter 15 Ethylene–Propylene–Diene Rubber/Natural Rubber Blends

457

Figure 15.11 (a) Dynamic storage moduli and tan delta for he NR/EPDM/TOR blends containing different concentrations of TOR. (b) Dynamic mechanical properties of vulcanized NR/EPDM (70:30 wt%) blends; (a) noncomptibilized blend and compatibilized with 2.5 phr of EPDMSH. (From References 27,56 with permission from John Wiley & Sons.)

phase together with the nonvulcanized EPDM phase whose transition occurs at similar temperature. But compatibilized blend exhibits a lower Tg value related to the NR phase and a lower damping (Fig. 15.11b (B)). These results are due to the effective compatibilization of this functionalized copolymer. The strong interactions between the mercapto groups along with the EPDMSH backbone and the NR phase decrease the mobility of this phase, giving rise to a decrease of the damping. The blend compatibilized with EPDMSH displays a second transition with low damping at temperature. The proportion of the vulcanized EPDM phase in

458

Polyolefin Blends

the NR/EPDM blend compatibilized with EPDMSH may be responsible for the second transition at higher temperature.

15.9 RHEOLOGICAL PROPERTIES The log–log plots of apparent shear stress versus apparent shear rate for STR5L/ EPDM and STR5L/BEPDM blends with various blend compositions are shown in Figs. 15.12 and 15.13, respectively (71). Flow curves of all the blends show reasonably straight lines, whose intercept K and slope n correspond to the power law equation (the Ostwald–de Waele equation) (72). Table 15.4 shows the power law index and the consistency of flow of STR5L/EPDM and STR5L/BEPDM blends. The values of nðn < 1Þ indicate the pseudoplastic nature of STR5L, EPDM, BEPDM, and their blends. Hence, the apparent viscosity of the two sets of blends decreased as the shear rate increased. It can also be seen that for the pure rubbers, EPDM had the lowest n value and STR5L has the highest n value. This accounts for the high pseudoplasticity, the highly shear thinning fluid in the modified BEPDM, and the more plug-like profile (73) consequently, the blends of STR5L/BEPDM tended to have a lower n value at a given blend composition, which increased with increasing levels of STR5L. It is observed that the modified EPDM by bromination reaction affects the

Figure 15.12 Effect of apparent shear rate on the apparent shear stress of STR5L/EPDM blends at various blend compositions. (From Reference 71 with permission from John Wiley & Sons.)

Chapter 15 Ethylene–Propylene–Diene Rubber/Natural Rubber Blends

459

Figure 15.13 Effect of apparent shear rate on the apparent shear stress of STR5L/EPDM blends at various blend compositions. (From Reference 71 with permission from John Wiley & Sons.)

shear flow property. That is, at a given shear rate, a higher apparent shear stress of pure BEPDM compared to that of EPDM and STR5L was found. The highest shear viscosity of BEPDM was therefore observed at a given apparent shear rate. It indicates that the bromine substituents on the rubber main chain may increase the chain rigidity of the rubber, consequently increasing the ability to resist flow, whereas STR5L gave the lowest apparent shear viscosity because of its ease in molecular weight breakdown with mastication during sample preparation and with shear force during the capillary flow test. Generally, the true shear viscosity of a polymeric blend follows the log additive rule (74–76). For the miscible blends, rheological properties (e.g., viscosity and die swell) show a positive deviation from their additive values; whereas the immiscible blends give a negative deviation in rheological properties (76). Blends in all blend compositions were evaluated and found to be negative deviations relating to their additive values. It is therefore indicated that the blends of STR5L/EPDM and STR5L–BEPDM were the immiscible blends. It means that there is no specific interaction between the two components of both blends. This may be attributed to the dissimilar, low unsaturated structure of EPDM, and the polarity

Table 15.4 The Power Law Index (n) and the Consistency of Flow (K) for Various Blend Compositions. NR/EPDM blend 0/100 25/75 50/50 75/25 100/0

n

K, KPa

NR/EPDM blend

n

K, KPa

0.14 0.15 0.20 0.21 0.22

293.0 149.9 88.7 85.1 86.8

0/100 25/75 50/50 75/25 100/0

0.10 0.14 0.16 0.20 0.22

444.6 169.0 124.5 81.8 86.8

n and K are power law index and consistency constant, respectively. [From Reference 71 with permission from John Wiley & Sons.]

460

Polyolefin Blends

of the bromine substituents on the BEPDM chains compared to the unsaturated nonpolar structure of natural rubber. In presence of homogenizing agent (Ultrablend 4000), the shear rate and shear viscosity are measured and plotted against content of homogenizing agent (70). A positive deviation of the blends was observed only at 5 phr of the homogenizing agent and which confirmed the blend compatibility of NR and EPDM. Pseudoplastic behavior was observed in the flow of blends. That is, the apparent shear viscosity decreased with an increase in the apparent shear rate. Therefore, the shear stress produced became smaller when the rate of shear increased. This flow behavior indicates a pseudoplastic fluid or shear thinning behavior of the blends.

15.10 THERMAL PROPERTIES The thermal characteristics of NR, EPDM, and their blend (50/50 NR/EPDM) have been examined with a DSC technique from 100 to 50 C (57,62). The glasstransition temperatures (Tg s) of NR, EPDM, and their blend are examined and are given in Table 15.5. It is found that the 50/50 (w/w) NR/EPDM blend having two distinct glass transitions; the lower glass transition was due to the EPDM phase and the higher glass transition was due to the NR phase. The large melting endoderm was attributable to the high crystallinity of EPDM. It is found that the specific heat capacity depends on the type of rubber, the compatibilizer, and the concentration of the blend. A careful inspection of Table 15.5 shows that the mean Tg value of pure NR was 63 C, and this changed to 64 C in the blend; the Tg value of pure EPDM was 37 C, and this changed to 45 C in the blend. This may be due to some interaction between NR and EPDM at the boundaries of their phases forming a third phase. DSC thermographs show the compatibilizing effects of BR, PVC, EPDM-g-MAH, and gradiation on NR/EPDM. For each component in the blend, Tg showed a higher shift Table 15.5 DSC Results Obtained for NR, EPDM, and 50/50 NR/EPDM Without and With Compatibilizers. Contribution from NR Sample NR EPDM NR/EPDM (control) NR/BR/EPDM NR/PVC/EPDM NR/EPDM-g-MAH/EPDM NR/g-radiation/EPDM NR/chlorinated rubber/EPDM NR/chlorosulfonated polyethylene/EPDM

Contribution from EPDM

Tg C

Shift in NR Tg C

Tg C

63 — 64 58 62 60 59.6 48

— — — þ6 þ2 þ4 þ4.4 þ12

— 37 45 44 2.5 40 45.6 48

60

þ4

[From Reference 57 with permission from John Wiley & Sons.]

60

Shift in NR Tg C — — — þ1 þ2.5 þ5 .6 3 14

Chapter 15 Ethylene–Propylene–Diene Rubber/Natural Rubber Blends

461

than that observed in the following order: NR/BR/ EPDM > NR/g-radiation/EPDM >NR/ MAH /EPDM>NR/PVC/EPDM. Only one Tg was detected after the addition of chlorinated rubber or chlorosulfonated PE to NR/EPDM blends, and this indicated the improved compatibility or dominance of these phases. However, when the compatibilizers were added to the blends, the glass transition became less distinct, and this indicated improved compatibility. By using O’Neill’s method (77), the specific heat capacity of NR/EPDM blends was determined (31). It is found that the law of reciprocal affinity, the linear contribution of components to the specific heat capacity is followed in EPDM/NR blends.

15.11

ELECTRICAL PROPERTIES

NR/EPDM blends with various carbon black concentrations (0–30 phr) were analyzed in terms of electrical resistivity, dielectric breakdown voltage testing, and physical properties so that these blends could be used for the high insulation iron cross arms (70). It is essential that the carbon black concentration used in these applications is investigated to obtain an appropriate reinforced and insulated rubber for coating on an iron cross arm. Usually, the materials for iron cross arm coating must have a volume resistivity higher than 1011 ohm cm. Thus, the carbon black concentration is critical, which indicates that at most 10 phr are required for a suitable insulation compound. The volume resistivities of natural rubber and EPDM are 1015 and 106 ohm cm, respectively. As a consequence, the concentrations of carbon black at 0–10 phr were studied along with high concentration and are given in Table 15.6. Electrical properties such as volume resistivity and surface resistivity of blend samples are furnished. Blend with 30 phr carbon black was found to be overloaded in terms of both volume resistivity and surface resistivity. That is, the resistivity of these formulations was lower than the limit of the testing equipment. This is an effect of the quasigraphitic microstructure of the carbon black; this makes the blend more electrically conductive. The higher the surface/volume resistivity, the lower the leakage current and the less conductive the material is. The major application of carbon black Table 15.6

Effect of Concentration of Carbon Black on Electrical Properties.

Properties Volume resistivity 105, ohm cm) Surface resistivity 105, ohm

NEC0

NEC3

NEC5

NEC7

NEC10

NEC20

NEC30

2.7

2.2

3.8

3.5

3.3

1.8

b

7.5

6.8

1.6

1.2

4.2

5.5

b

NEC0, NEC3, NEC5, NEC7, NEC10, NEC20, and NEC30 are 70:30 NR/EPDM blends with carbon black content 0, 3, 5, 7, 10, 20, and 30 phr, respectively. b—the resistivities were lower than the limit of the testing equipment. [From Reference 70 with permission from John Wiley & Sons.]

462

Polyolefin Blends

in low concentration is for antistatic protection. The small amount of carbon black used in this application is to achieve a suitable resistivity before a significant drop in resistivity is noted. After applying the required high voltage at 10 kVac for 1 min to the specimens, the NR/EPDM samples do not exhibit any electrical burn-through, punctures, or electrical discharges. The results of dielectric breakdown voltage tests indicated that the appropriate concentration of carbon black is lower than 10 phr for insulation compounds. Effect of curing systems like sulfur and peroxide on the electrical properties of NR/EPDM blends were carried out (78). There is an improvement in the volume resistivity of blend samples, which could be used for insulation of wires and high voltage cables. The dielectric properties of NR/EPDM blends were conducted in varying proportions with a reinforcement of semireinforcing furnace carbon black (79). The permittivity and dielectric loss were determined. The permittivity and dielectric loss of NR/EPDM blends were determined with different compatibilizers at different frequencies (66). By changing the compatibilizer, the permittivity increases for BR, SPE, and chlorinated rubber are much more than that for PVC. An abrupt increase in the permittivity and dielectric loss was observed (Fig. 15.14) at a concentration of 6 phr EPDM-g-MAH in all blend ratios.

15.12 AGING PROPERTIES 15.12.1 Thermal Aging The effects of silica, carbon black, and a mixture of silica and carbon black on the mechanical properties of aged NR/EPDM sample are reported (70). The presence of silica in the blends produces a marked increase in mechanical strength after aging by 120%. The main effect of silica additive alone in NR/EPDM is very attractive. More interestingly, the synergistic improvement in tensile strength of the blend was obviously seen in the presence of both carbon black (3 phr) and silica (30 phr), after aging, with another increase of tensile strength by 38%. The silica filler contains the hydroxyl functional group on its surface. The intermolecular bonding between the hydroxyl group in silica and the NR/EPDM might be able to take place. The network structure further developed when carbon black was added. It is obvious that silica enhances the tensile strength of NR/EPDM blends after thermal aging, because silica improves heat resistance of the material, and it does not promote curing. The synergistic effect of carbon black and silica is likewise seen in the tear strength and hardness. This effect enhances the blend heat resistance by a substantial increase in tear strength and hardness. Figure 15.15 presents the retention of the tensile properties of the blends after treatment in an air-circulating oven at 70 C for 72 h. The best aging resistance has been achieved with the addition of EPDMTA, except in terms of elongation at break in blends vulcanized with higher amount of MBTS (61). Concerning blends compatibilized with EPDMSH, the aging resistance decreases as the concentration of the accelerator in the blend increases. The aging resistance of EPDMSH compatibilized blends is lower than noncompatibilized blends for accelerated-sulfur curing system.

Chapter 15 Ethylene–Propylene–Diene Rubber/Natural Rubber Blends

463

Figure 15.14 (a) The permittivity (e0 ) versus the frequency (log f) for EPDM containing different 00

concentrations of EPDM-g-MAH as compatibilizer; (b) dielectric loss (e ) versus the frequency (log f) for EPDM containing different concentrations of EPDM-g-MAH as compatibilizer. (From Reference 66 with permission from Sage publications Ltd.)

15.12.2 Ozone Resistance NR/EPDM blends were extensively analyzed for its ozone resistance property (70). The blend ratios of NR/EPDM used in this study were 100/0, 80/20, 70/30, 60/40, and 0/100. To improve the NR/EPDM blends for ozone resistance, the amount of EPDM

464

Polyolefin Blends

Figure 15.15 Retention of the tensile properties of NR/EPDM blends with thermal aging: (a) noncompatibilized; compatibilized with (b) EPDMTA and (c) EPDMSH. (From Reference 61 with permission from Elsevier Ltd.)

cannot exceed 40% by weight. The natural rubber specimens could not withstand the ozone gas, which is the nature of the NR polymer. One can see cracks in the vulcanized rubber after tests. When 20 phr of EPDM and higher concentrations were blended, the specimens could withstand the ozone gas. The specimens did not crack. The dispersed EPDM domains reduced the crack length and increased the critical energy for macroscopic cracks. The EPDM domains function as crack dissipation centers, which delays the crack appearance. In the case of silica-filled blends, they enhance both mechanical properties (as reinforcement filler) and static ozone resistance. The silicafilled 80/20 (NR/EPDM) blends could withstand ozone gas of 50 ppm. Losses in tensile strength after aging in unfilled blends were higher than 50%, which is a very good indicator that silica can also withstand heat deterioration.

Chapter 15 Ethylene–Propylene–Diene Rubber/Natural Rubber Blends

465

Table 15.7 Ozone Resistance of NR/EPDM Blends. Property Critical stress, MPa Critical strain, % Critical stored elastic energy kJ m3

1

2

3

4

0.142 10.15

0.180 11.55

0.210 12.55

0.240 11.25

7.350

10.00

12.65

13.06

1, 2, 3, and 4 are 70/30:NR/EPDM blends with TOR content of 0, 5, 10, and 20 phr, respectively. [From Reference 27 with permission from John Wiley & Sons.]

The ozone resistance of the blend was determined quantitatively in terms of critical stress–strain parameters and the values are given in Table 15.7 (27). The ozone resistance of the NR/EPDM blend is increased by the amount of 80% upon the addition of 10 phr of compatibilizer, TOR. The improvement in ozone resistance for the TOR containing blend is attributed to the better dispersion of the EPDM particles in the NR matrix, which is aided by TOR. That is, more finely dispersed EPDM particles prohibit the growth of ozone cracks initiated in the NR matrix before the crack grows over the critical length. Once the ozone crack grows over the critical size, crack propagation cannot be stopped by EPDM particles.There are studies available in the literature with special reference to the 60:40 ratio of NR: EPDM blend, which can provide excellent ozone resistance even in the absence of any antiozonant (80). This ratio was found good in maintaining complete ozone protection in the blend while maintaining the good set and elasticity performance. Absolutely, no crack was observed when the extruded product was subjected to ozone exposure at 40 C for 72 h.

15.13

TRANSPORT PROPERTIES

The swelling behavior of NR/EPDM blends in motor oil under compression strain was investigated (81). The weight uptake percentage was determined and plotted against the square root of the swelling exposure time (minutes) in motor oil (Fig. 15.16). The weight uptake increased as the exposure time increased, and equilibrium swelling was achieved after 192 h of immersion in oil. The lowest weight uptake was recorded with a mix containing EPDM polymer only (0/100), whereas the 25/75 NR/EPDM blend ratio recorded a lower weight uptake value. Generally, diffusion curves show the same pattern irrespective of the compression applied. It is also noted that the curve levels depended on the degree of compression applied. At a short exposure time in motor oil, the weight uptake was independent of the applied compression, whereas at longer exposure times, a decrease in the weight uptake with the applied compression was much more pronounced. At low compression (3%), the compression recovery percentage for all blend ratios was positive. At high compression, the recovery had negative values. The highest recovery value was recorded for the 25/75 NR/ EPDM blend. However, the lowest recovery was observed with the EPDM vulcanizate (0/100).

466

Polyolefin Blends

Figure 15.16 Weight uptake (%) of motor oil for NR/EPDM vulcanizates versus the exposure time (t1=2 ) at 100 C and 35% compression. (From Reference 81 with permission from John Wiley & Sons.)

The compression recovery decreased with the increase in applied compression for all exposure periods. The 25/75 NR/EPDM blend was advantageous because of its high degree of elastic recovery and its lower weight uptake of motor oil. Interaction of EPDM/NR blends with aromatic penetrates (82) aldehydes and ketones (83) and chlorinated penetrants (84) were studied. The sorptivity, diffusivity, and permeability values in nitrobenzene, cholrobenzene, and bromobenzene are lower compared to other NR-based polymer blends discussed (82). This is attributed to the tightly packed structure of EPDM blend and exhibiting both toughness of the plastic and elasticity of the gum elastomer phase. The mechanism of transport in NR/EPDM blends is Fickian irrespective of the penetrant used. The values of D, S, and P are much higher in chlorohydrocarbons than the other systems discussed. This is because of the strong interaction between the chlorinated hydrocarbons and blends.

15.14 APPLICATIONS The technology of making the NR/modified EPDM blends has been shown to be suitable for a number of applications such as extruded profile weather strips

Chapter 15 Ethylene–Propylene–Diene Rubber/Natural Rubber Blends

467

Figure 15.17 Applications of NR/EPDM blends.

(Fig. 15.17a) for automotive and molded products like grommet (Fig. 15.17b) and washing machine gasket (Fig. 15.17c). These products typically use 100% EPDM, hence, the inclusion of a certain percentage of NR would reduce the products cost without affecting the products quality and performance. NR:EPDM with a ratio of 60:40 is also widely used for making extruded profile weather strip shown in Fig. 15.17a. Examples of such extruded profiles are door liners of car and back window seal of a van. Another major utility of these blends are for making diving suits. A few examples are shown in Fig. 15.18. Pro-Am (Fig. 15.18a) is a rubber suit made from NR/EPDM blend. The main features are good stretch characteristics for comfort and three layer construction for durability. It is ideal for sports, military, rescue, and light commercial applications. The Pro-hd (Fig. 15.18b) is tough NR/ EPDM blend rubber suit made to endure the harshest conditions. USIA (Fig. 15.18c) is also a vulcanized rubber suit made from NR/EPDM blend and it is ideal for sports, military, rescue, and light commercial applications. NR/EPDM blends are also used for making rubber boots for agriculture. These boots will have more resistance against cracks caused by ozone. A ratio of 25:75 to 40:60 parts by wt of NR/EPDM is used for preparing these kinds of boots.

468

Polyolefin Blends

Figure 15.18 Applications of NR/EPDM blends.

Chapter 15 Ethylene–Propylene–Diene Rubber/Natural Rubber Blends

469

NR/EPDM-based sidewalls of radial tires are also prepared to get more durability and appearance. These blends with an EPDM content varying from 30 to 70 phr are used for making cables and conductors.

15.15

CONCLUSIONS

This chapter brings out the recent development in the field of EPDM/NR blends.The development of blends of NR with EPDM combines the superior physical properties and competitive price of NR with the excellent resistance to weathering, in particular, attack by ozone of EPDM. However, NR and EPDM blends are heterogeneous dispersion of fast curing NR phase and a slow curing EPDM phase. The overall result is that the blend vulcanizate will be composed of overcured NR and undercured EPDM. This would adversely affect the properties of the blend. By the judicious selection of NR-to-EPDM ratio and the concentration of DIPDIS in the compound, one can improve the physical properties of the vulcanizates. These properties can further be improved by two-stage vulcanization. The incorporation of compatibilizers into the NR/EPDM blends greatly enhances their compatibility and greatly improves the overall properties of the blends. The compatibilizers are able to create a well-dispersed bicontinuous phase. The Tg s from the DSC analysis indicate that both NR/EPDM and NR/BEPDM blends are thermodynamically incompatible. NR/ EPDM blends with 3 phr of carbon black are suitable for cable applications since it exhibits high tensile strength and suitable volume resistivity. EPDM-g-MAH was found to be an effective compatibilizer, which brings together the two incompatible components into the compatible level. A new method known as ‘‘reactive mixing’’ has developed recently to increase the cure rate of EPDM by modifying the EPDM phase to make it more reactive toward curatives, using commercially available sulfur donors such as bis-alkylphenoldisulphide (BAPD), in combination with dithiocaprolactam (DTDC) and/or dithiomorpholine (DTDM). The refinement of reactive mixing process with cost effective sulfur donors is one of the challenges in the maximum utilization of these elastomer blends.

NOMENCLATURE BAPD BEPDM BR DIPDIS DTDC DTDM EPDM EPDMTA EPDMSH

Bis-alkylphenoldisulphide Bromianted EPDM Polybutadiene rubber Bis(diisopropyl)-thiophosphoryl disulfide Dithiocaprolactam Dithiomorpholine Ethylene–propylene–diene rubber Thioacetate-modified EPDM Mercapto-modified EPDM

470

Polyolefin Blends

EPDM-g-MAH MBTS NR STR5L PVC TOR DGm DHm DSm

A graft copolymerization of EPDM with maleic anhydride 2, 20 -Dithiobisbenzothiazole Natural rubber Natural rubber from Thailand Poly(vinyl chloride) Trans-polyoctene rubber Gibb’s free energy Enthalpy of mixing Entropy of mixing

REFERENCES 1. J. A. Pomposo, E. Calahorra, L. Eguiazbal, and M. Cortazar, Macromolecules, 26, 2104 (1993). 2. J. W. Barlow and D. R. Paul, Polym. Eng. Sci., 27, 1482 (1987). 3. E. H. Andrews, Rubber Chem. Technol., 40, 435 (1967). 4. W. M. Hess, C. R. Herd, and P. C. Vegvari, Rubber Chem. Technol., 66, 329 (1993). 5. T. Johnson and S. Thomas, J. Mater. Sci., 34, 3221 (1999). 6. H. Varghese, S. S. Bhagawan, and S. Thomas, J. Polym. Sci. B Polym Phys, 37, 1815 (1999). 7. S. C. George, K. N. Ninan, G. Groeninckx, and S. Thomas, J. Appl. Polym. Sci., 78, 280 (2000). 8. S. Sadhu and A. K. Bhowmick, Rubber Chem. Technol., 78, 321 (2004). 9. S. Sadhu and A. K. Bhowmick, Rubber Chem. Technol., 76, 860 (2003). 10. S. Sadhu and A. K. Bhowmick, J. Appl. Polym. Sci., 92, 698 (2004). 11. J. A. Davidson and M. E. Woods, Rubber Chem. Technol., 49, 112 (1976). 12. V. A. Shershnev, Rubber Chem. Technol., 55, 537 (1982). 13. W. H. Whittington, Rubber Ind., 9, 151 (1976). 14. E. H. Andrews, Rubber Chem. Technol., 40, 635 (1967). 15. L. Spenadel and R. L. Sutphin, Rubber Age., 102, 55 (1970). 16. A. Y. Coran, Rubber Chem. Technol., 64, 801 (1992). 17. K. C. Baranwal and P. N. Son, Rubber Chem. Technol., 47, 88 (1974). 18. G. N. Ghebrehiwet and S. R. Stephanie, Rubber World, 227, 26 (2003). 19. D. J. Walsh, Polymer blends, in: Comprehensive Polymer Science, Vol. 2, Collin Booth and Collin Price (eds.), Pergamon Press, Oxford, 1989, chapter 5. 20. R. N. Jana and G. B. Nando, J. Elastom. Plasti., 34, 349 (2002). 21. R. L. Jalbert (ed.), Modern Plastics Encyclopedia, McGraw-Hill, NewYork, 1984. 22. G. Kerrutt, H. Blumel, and H. Weber, Kautsch. Gummi. Kunst., 22, 413 (1969). 23. O. Olabisi, L. M. Roberson, and M. T. Shaw, Polymer–Polymer Miscibility, Academic Press, New York, 1979. 24. D. R. Paul and J. W. Barlow, in: Multiphase Polymers, S. L. Cooper and G. M. Estes (eds.), American Chemical Society, Washington, DC, 1979. 25. G. N. Avgeropoulos, F. C. Weissert, P. H Biddison, and G. G. A. Bohm, Rubber Chem. Technol., 49, 83 (1976). 26. R. F. Bauer and E. A. Dudley, Rubber Chem. Technol., 50, 35 (1977). 27. Y. –W. Chang, Y. -S. Shin, H. Chun, and C. Nah, J. Appl. Polym. Sci., 73, 749 (1999). 28. B. Boutevin, E. Fleury, J. P. Parisi, and Y. Pietrasnta, Makromol. Chem., 190, 2363 (1989).

Chapter 15 Ethylene–Propylene–Diene Rubber/Natural Rubber Blends

471

29. F. Romani, E. Passaglia, M. Aglietto, and G. Ruggeri, Macromol. Chem. Phys., 200, 524 (1999). 30. U. Gorski, K. Maenz, and D. Stadermann, Angew. Makromol. Chem., 253, 51 (1997). 31. T. Zaharescua, V. Meltzer, and R, Vilcu, Polym. Degrad. Stabi., 70, 341 (2000). 32. A. K. Ghosh, S. C. Debnath, N. Naskar, and D. K. Basu, J. Appl. Polym. Sci., 81, 800 (2001). 33. A. K. Ghosh and D. K. Basu, J. Appl. Polym. Sci., 84, 1001 (2002). 34. F. Guillaumond, Rubber Chem. Technol., 49, 105 (1976). 35. M. vanDuin, J. C. J. Krans, and J. Smedinga, Kautsch. Gummi. Kunst., 46, 455 (1993). 36. K. H. Wirth, U.S. Patent, 3, 492, 370 (1970). 37. C. B. Shulman, Rubber Chem. Technol., 59, 180 (1986). 38. K. Hashimoto et al., Nippon Gomu Kyokaishi, 43, 652 (1970). 39. R. T. Morrissey, Rubber Chem. Technol., 44, 1025 (1971). 40. Japan Patent, 3967 (1968). 41. F. P. Baldwin and G. Ver Strate, Rubber Chem. Technol., 45, 709 (1972). 42. F. Itsuro and M. Masao, Sumitomo Chemical Co. Ltd., Ger Offen., 2,045,574 (1971). 43. A. J. Tinker, Proceedings of International Rubber Conference, Moscow, Russia, MRPRA Publication, 1511, 1994, p. 180. 44. MRPRA. Res. Discl., 362, 308 (1994). 45. Chemische Werke Huls, A. G. Netherlands, 7, 31, 1958 (1974). 46. S. Yasui, M. Hirooka, and T. Oshima, Sumitomo Chemical Co., U.S. Patent, 3,649,573 (1972). 47. R. J. Hopper, Rubber Chem. Technol., 49, 341 (1976). 48. A. Y. Coran, Rubber Chem. Technol., 61, 281 (1988). 49. J. Lohmar, in: Proceedings of International Rubber Conference, Stuttgart, Germany, 1985, p 91. 50. D. G.Young, E. N. Kresge, and A. J. Wallace, Rubber Chem. Technol., 55, 428 (1982). 51. P. S. Brown and A. J. Tinker, J. Natl. Rubber Res., 5, 157 (1990). 52. M. J. R. Loadman and A. J. Tinker, Rubber Chem. Technol., 62, 234 (1989). 53. P. S. Brown and A. J. Tinker, J. Natl. Rubber Res., 11, 227 (1996). 54. S. K. Mandal, R. N. Datta, P. K. Das, and D. K. Basu, J. Appl. Polym. Sci., 35, 987 (1988). 55. H. S.-Y. Hsich, Adv. Polym. Technol., 10,185 (1991). 56. A. S. Sirqueira and B.G. Soares, J. Appl. Polym. Sci., 83, 2892 (2002). 57. S. H. El Sabbagh, J. Appl. Polym. Sci., 90, 1 (2003). 58. G. J. Anastasiadas and J. K. Koberstein, Macromolecules, 22, 1449 (1989). 59. D. J. Meier, Hetero-Phase Polymer Systems, American Chemical Society, Washington, DC, 1990. 60. R. Joseph, K. E. George, and J. D. Francis, Int. J. Polym. Matter, 11, 205 (1986). 61. A. S. Sirqueira and B.G. Soares, Eur. Polym. J., 39, 2283 (2003). 62. S. H. El Sabbagh, Polym. Test., 22, 93 (2003). 63. S. H. Botros, Poly-Plast. Tech. Eng., 41, 341 (2002). 64. A. B. Shehata, H. Afifi, N. A. Darwish, and A. M. El Syed, Poly-Plast. Tech. Eng., 45, 165 (2006). 65. A. M. El Sayed and H. Afifi, J. Appl. Polym. Sci., 86, 2816 (2002). 66. M. O. Abou-Helal and S. H. El Sabbagh, J. Elast. Plast., 37, 319 (2005). 67. N. Suma, R. Joseph, and K. E. George, J. Appl. Polym. Sci., 42, 2329 (1993). 68. H. Kenjiro, M. Minoru, M.Takahide, and O. Harunori, Nippon Gomu Kyokaishi. 49(3), 246 (1976). 69. G. Jin-Hwan and P. S. Soo, Komu Hakhoechi., 29, 121 (1994). 70. S. Kiatkamjornwong and K. Pairpisit, J. Appl. Polym. Sci., 92, 3401 (2004). 71. C. Lewis, S. Bunyung, and S. Kiatkamjornwong, J. Appl. Polym. Sci., 89, 837 (2003). 72. J. A. Brydson, Flow Properties of Polymer Melts, Plastics Institute, London, 1970, p. 12.

472

Polyolefin Blends

73. M. G. McCrum, C. P. Buckley, and C. B. Bucknall, Principles of Polymer Engineering, Oxford University Press, New York, 1997, p. 308. 74. L. A. Utracki, Polym. Eng. Sci., 22, 96 (1982). 75. P. P. Kundu, A. K. Bhattacharya, and D. K. Tripathy, J. Appl. Polym. Sci., 66, 1759 (1997). 76. P. P. Kundu, D. K. Tripathy, and B. R. Gupta, J. Appl. Polym. Sci., 63, 187 (1997). 77. M. G. O’Neill, Anal. Chem., 36, 1238 (1964). 78. P. Radivoj, T. Milan, G. Ivan, and S. Latinka, Plastika i Guma., 10(1), 23 (1990). 79. A. M. Ghoneim and M. N. Ismail, Polym. Plasti. Tech. Eng., 38, 979, (1999). 80. M. E. Samuels, in: Ethylene Propylene Rubber, R. O. Babbitt (ed.), R. T. Vanderbilt Company Inc, Norwalk, Connecticut, 1978, p. 147, chapter 5. 81. S. H. Botros and A. M. El Sayed, J. Appl. Polym. Sci., 82, 3052 (2001). 82. Siddaramaiah, S. Roopa, and U. Premkumar, Polymer, 39, 3925 (1998). 83. Siddaramaiah, S. Roopa, U. Premkumar, and A. Varadarajulu, J. Appl. Polym. Sci., 67, 101 (1998). 84. Siddaramaiah, S. Roopa and K. H. Guruprasd, J. Appl. Poly. Sci., 88, 1366 (2003).

Chapter

16

Phase Field Approach to Thermodynamics and Dynamics of Phase Separation and Crystallization of Polypropylene Isomers and Ethylene–Propylene–Diene Terpolymer Blends Rushikesh A. Matkar1 and Thein Kyu1

16.1 INTRODUCTION Polyolefins, one of the largest commodity polymeric plastics in the market place, have been widely studied over six decades covering synthesis, structural, physical, as well as mechanical properties point of views. However, the study on blends of polyolefins is rather scarce relative to their neat forms. One of the polyolefin blends that gained considerable attention is thermoplastic polyolefins (TPO) due to the enhanced impact strength and toughness of polyolefins. A typical example is a blend of polypropylene (PP) and ethylene–propylene rubber (EPR). EPR has been incorporated into PP through reaction in batch reactors or physical blending. The PP/EPR

1

Department of Polymer Engineering, University of Akron, Akron, OH 44325, USA

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

473

474

Polyolefin Blends

blends formed by mixing in the batch reactor are already phase separated and thus thermodynamics may not play a role, but the emerged structure and properties of physically blended ones may be affected by thermodynamic phase diagrams as well as by dynamics of phase separation. These rubber-modified polyolefins greatly improve the toughness and impact strength of the composites due to the rubber inclusion, whereas polyolefin constituents afford good tensile properties of the composites and also melt processability. A certain functional group may be introduced to EPR to provide chemical sites for cross-linking. Such cross-linking reaction further affords rubber-like network properties, but often it occurs at the expense of the reduction in melt processability. To circumvent such short comings, mineral or paraffinic oil has been added as a mean of improving processability and controlling swelling properties of the blends. Depending on the chemical structure of the functional groups, such reactive polyolefin blends are often known as thermoplastic elastomers (TPE) or thermoplastic vulcanizates (TPV). A classical example of TPE is the blend of polypropylene (PP) and ethylene propylene diene monomer (EPDM) (1–5). Despite a slight reduction in the rigidity or stiffness, these PP/EPDM blends exhibit enhanced toughness and impact strength, good resistance to ozone, and UV radiation without losing flow properties. The addition of small amount of PP raises the modulus and tensile stress of iPP/EPDM as compared with neat EPDM, and thus it has been regarded as a stiffness modifier. However, adding small amounts of impact modifier such as EPDM improves the toughness and impact strength of PP/EPDM at a marginal loss in tensile strength of the PP. As can be expected, the mechanical and physical properties of PP/EPDM blends are intimately related to the internal phase separated domain structures of the constituent phases. The emerged morphologies vary from a sea-and-island type to a bicontinuous structure, which may be governed by thermodynamic phase diagram, if it exists, as well as kinetics of thermally induced or reaction-induced phase separation. In addition, the crystalline phase of PP in the blends has to be addressed in evaluating the blend performance that has been ignored in the literature (1–5). In practice, the melt blends of commercial grade PP and EPDM were perceived to be immiscible, which may be a consequence of melt-blending conditions in given mixing equipment or driven by chemical reaction. To investigate the miscibility of a polymer blend, solution blending is preferred although it is usually not a customary practice in most industrial settings. This immiscibility perception has changed recently when a lower critical solution temperature (LCST) was first reported for the iPP/EPDM solution blends; this LCST was located very close to the melting temperature of the neat iPP (6). Moreover, the LCST phase diagram of iPP/EPDM blend was intervened by the melting transition of iPP, thereby implicating the phase behavior. In order to decouple LCST and Tm of iPP, Ramanujam et al (7) switched iPP to syndiotactic polypropylene (sPP) because sPP is known to have a lower crystalmelting temperature relative to that of iPP. These authors found that the existence of combined LCST and upper critical solution temperature (UCST) phase diagrams in the sPP/EPDM blend in which the melting transition of neat sPP is located in between the LCST and the UCST.

Chapter 16 Phase Field Approach to Thermodynamics

475

The motivation of the present article is (i) to elucidate the governing mechanisms of the spatio-temporal development of blend morphology involving the competition between the phase separation dynamics and kinetics of crystallization and (ii) to reconcile the different opinions of the complete immiscibility perception and the aforementioned complex liquid–solid phase diagrams of PP/EPDM thermoplastic elastomer systems. This chapter describes theoretical modeling and simulation on establishment of thermodynamic phase diagrams of PP isomers/EPDM and dynamics of thermal quench induced-phase separation and morphology development during crystallization of PP isomers.

16.2

EXPERIMENTAL PHASE DIAGRAMS

16.2.1 Cloud Point Phase Diagram of iPP/EPDM Blends Polypropylene may be classified in three isomeric forms based on the tacticity, viz., isotactic, syndiotactic, and atactic. Both isotactic and syndiotactic forms are highly crystalline, whereas the atactic PP is completely amorphous. On the contrary, EPDM is an amorphous copolymer comprised of equal parts of ethylene and propylene with 45% of ethylidene norbonene to afford dynamic vulcanization. We briefly review the miscibility studies of solvent cast iPP/EPDM blends primarily based on the contributions of Kyu’s group (6,7) and references therein. The solvent cast blend films were prepared by first dissolving iPP powder in xylene at 130 C, then adding EPDM after lowering the temperature to 100 C and stirred thoroughly for about 90 min to assure thorough mixing. The film specimens were prepared by solvent casting at ambient temperatures in a fume hood and dried in a vacuum oven for 48 h at room temperature. The average thickness of these blend films was approximately 1020 mm. The solvent cast films appear turbid to the naked eye, which might be attributed to the phase separation of the blends and/or the crystallization of iPP. It is difficult, if not impossible, to decouple the crystal-melting and liquid–liquid phase separation especially when the melting temperature of iPP and the LCST coexistence curve of the iPP/EPDM blend are in close proximity or intersecting each other. That is to say, iPP molecules may have gained sufficient mobility during premelting, and thus it is possible that phase separation could start before the crystal melting is completed. In order to circumvent the aforementioned problem, the cloud point measurement was performed based on light scattering by first melting the iPP crystal phase at 170 C and rapidly cooling it to 140 C within 2 min. Subsequently, the sample was reheated to 190 C at a rate of 0.5 C min1. The scattered intensity was measured at an approximate 2u angle of 20 during this reheating cycle. It was found that the blend at 140 C showed no scattering of light, suggestive of the homogenous character of the blend. However, the intensity increased rapidly as liquid–liquid phase separation commences at about 155 C. This procedure is thermally reversible so long as the phase separation process has not advanced significantly or the blend is not degraded. On the basis of this methodology, the phase diagram for this iPP/EPDM blend was

476

Polyolefin Blends

Figure 16.1 Experimental phase diagram of the iPP/EPDM blend as obtained by light scattering and DSC, showing the intersection of LCST and the crystal-melt phase transition. (From Reference (5) with permission from Elsevier.)

established by Chen et al. (6) as depicted in Fig. 16.1. The observed phase diagram is characterized by a lower critical solution temperature (LCST) type, which is intersected by the melting transition of iPP crystals. This phase diagram is reminiscent of an inverted teapot phase diagram of a polymer/liquid crystal system, exhibiting a liquid–liquid coexistence region, a narrow solid–liquid coexistence region, and neat crystal region bound by the solidus and liquid lines (Fig. 16.1).

16.2.2 Cloud Point Phase Diagram of sPP/EPDM Blends To alleviate the complex phase diagram of iPP/EPDM, Ramanujam et al. (7) investigated the sPP/EPDM blend as a complementary study. The advantage of its choice of sPP is that the melting point of sPP is significantly lower than that of iPP, which could effectively decouple the mutual interference of crystallization versus liquid–liquid phase separation. In Fig. 16.2 is shown the entire phase diagram of the solvent cast sPP/EPDM blends as established by a combination of differential scanning calorimetry (DSC) and cloud point determination. In descending order of temperature, the observed phase diagram exhibits an LCST curve, followed by a crystal–liquid transition, and subsequently UCST that lies underneath it. The coexistence lines have been drawn to guide the eyes. There exists a small miscible gap between the LCST and the UCST. The thermal reversibility of the LCST can be

Chapter 16 Phase Field Approach to Thermodynamics

477

Figure 16.2 Experimental phase diagram of the sPP/EPDM blend as determined by a combination of DSC and light scattering techniques, exhibiting the combined LCST and UCST together with the melting-point depression. The UCST curve was determined after the blends were homogenized in the single phase below the Tm , but above the crystallization temperatures. The symbols represent the experimentally determined points and the lines are drawn by hand or polynomial fits to guide the eyes. (From Reference (6) with permission from Elsevier.)

confirmed easily. However, the interference of crystal melting makes the confirmation of UCST more difficult at least experimentally, and thus it was cautioned that the UCST should be regarded as tentative. To substantiate the significance of these phase diagrams of iPP/EPDM and sPP/EPDM blends, we have developed a self-consistent theory for describing a crystalline–amorphous polymer blend in order to predict all possible phase diagram topologies and to compare some of these predictions with the observed cloud point phase diagrams. It may be anticipated that the present theoretical approach is capable of reconciling the discrepancy between the above phase diagrams and the perceived immiscibility of the PP/EPDM blend.

478

Polyolefin Blends

16.3 THERMODYNAMIC FREE ENERGY DESCRIPTION OF CRYSTALLINE POLYMER BLENDS 16.3.1 Flory–Huggins Free Energy of Amorphous–Amorphous Blends The Flory–Huggins (FH) lattice theory (8,9) has been customarily employed to establish phase diagram of binary amorphous–amorphous polymer mixtures based on the incompressible assumption that reads, DGm ðfÞ f ð1 fÞ ¼ lnðfÞ þ lnð1 fÞ þ xaa fð1 fÞ N A kB T r1 r2

ð16:1Þ

where DGðfÞ is the free energy of mixing. The volume fraction f may be expressed as f ¼ n1 r1 =ðn1 r1 þ n2 r2 Þ, where n1 and n2 are moles of each polymer having characteristic segment lengths of r1 andr2 , respectively. The total number of polymer chain segments is given as n ¼ ðn1 r1 þ n2 r2 Þ. Moreover, xaa is the FH amorphous– amorphous interaction parameter determining the enthalpy contribution toward mixing (8,10), which is proportional to the net interchange energy, but it is inversely proportional to absolute temperature. In order to account for free volume effects involving nonideality and noncombinatorial mixing, it is customary to expressxaa in the context of an empirical expression in what follows: xaa ¼ ðA þ B=T þ C ln TÞðDf þ Ef2 þ Ff3 þ Þ

ð16:2Þ

where the first bracket represents athermal and thermal dependencies, whereas the second bracket term accounts for concentration dependencies (11). This modified model has been used extensively to determine a variety of phase diagram topologies such as UCST, LCST, a combined UCST–LCST, a closed loop, and/or an hour-glass phase diagrams. Establishment of phase diagrams can be accomplished by applying a common tangent algorithm via free energy minimization of the mixture. This algorithm determines the coexistence line that satisfies the equal chemical potential principle and the free energy minimization which is the solution to the equations (12), @f ðfÞ @f ðfÞ f ðfa Þ f ðfb Þ ¼ ¼ @f fa @f fb fa fb

ð16:3Þ

where f ðfÞ is the free energy of the mixture at a given composition f. Although the FH theory may be adequate for elucidating the empirical phase diagrams of amorphous–amorphous, its extension to the crystalline–amorphous or crystalline–crystalline polymer blends requires a considerable modification by incorporating a solidification potential of the solid (crystal)–liquid (melt) transition (i.e., crystallization) of the crystalline constituent(s).

Chapter 16 Phase Field Approach to Thermodynamics

479

16.3.2 Extension of the FH Theory to Crystal–Amorphous Blends Existing theories of crystalline polymer blends have been predominantly dealt with the lowering of the pure crystalline constituent due to its miscibility with the polymeric diluent. The Flory diluent theory is a classical example of such a theory (8,11,13,14). One shortcoming of the Flory diluent theory is the assumption involved regarding the complete rejection of the solvent from the crystal phase, namely, the crystal phase in the blend is completely pure. This assumption has its own merit because equating the chemical potential of the crystalline component in the liquid phase to its chemical potential in the pure crystal phase has led to the analytically tractable solution, known as the Flory melting point depression equation. However, the Flory diluent theory encounters several limitations (15) when applied to partially miscible systems that exhibit phase separation in the vicinity of the melting temperatures such as the present iPP/EPDM blend. Also for systems that exhibit complex coupling between crystallization and phase separation (16–22) such as the above iPP/EPDM blend (6), it is imperative to take into account the crystal– amorphous interaction in addition to the conventional amorphous–amorphous FH interaction (23). Such a modification may make the free energy equation analytically intractable. However, with the advent of high speed computing capability of desktop machines, it is no longer necessary to adopt ‘‘a priori’’ for the neat crystal phase in the blends. In order to elucidate the thermodynamic phase diagram and kinetics of morphology evolution of crystalline microstructures, we shall introduce a phase field model of solidification for crystalline polymer blends undergoing the solid (crystal)– liquid (melt) phase transition. We then seek the self-consistent solutions to establish the coexistence phase boundaries of a hypothetical phase diagram of the crystal– amorphous blend. Polymer crystallization has been described in the framework of a phase field free energy pertaining to a crystal order parameter c in which c ¼ 0 defines the melt and assumes finite values close to unity in the metastable crystal phase, but c ¼ 1 at the equilibrium limit (23–25). The crystal phase order parameter (c) may be defined as the ratio of the lamellar thickness (l) to the lamellar thickness of a perfect polymer crystal (l0 ), i.e., c ¼ l=l0 , and thus it represents the linear crystallinity, that is, the crystallinity in one dimension. The free energy density of a polymer blend containing one crystalline component may be expressed as f ðc; fÞ ¼ ff ðcÞ þ

f ð1 fÞ lnðfÞ þ lnð1 fÞ þ fxaa þ xca c2 gfð1 fÞ ð16:4Þ r1 r2

where f ðcÞ is the free energy of crystallization of the crystalline component expressed as a Landau expansion in c that is weighted by its concentration or volume fraction in the mixture. This weighting ensures the free energy of the mixture to approach its pure crystal limit, that is, f ðc; fÞ ! f ðcÞ when f ! 1. The natural log terms represent the entropic contribution, whereas xaa fð1 fÞ is the enthalpic contribution to the Flory–Huggins free energy of liquid–liquid demixing. The

480

Polyolefin Blends

quantityxca c2 fð1 fÞ representing the solid–liquid interaction is complimentary to the amorphous–amorphous interaction term xaa fð1 fÞ. The xca parameter is called the crystal–amorphous interaction parameter that is repulsive and may be evaluated from the heat of fusion of the crystalline constituent as will be discussed later. When the linear crystallinity c is multiplied by its volume fraction, their product fc signifies the bulk crystallinity. Moreover, ð1 fÞc represents the amount of the noncrystalline component interacting with the crystal phase. The physical interpretation of xca c2 fð1 fÞ would therefore be the crystal–amorphous interaction term. It should be noted that the present modified theory reverts to the original Flory diluent theory in the extreme limit of complete rejection of the polymeric solvent from the crystal phase, that is, when the repulsive interaction between the crystal solute and amorphous solvent, xca becomes very strong or a neat crystal phase is formed in the blend. The free energy density of the crystal solidification may be expressed in the context of the asymmetric Landau expansion, f ðcÞ as FðcÞ zðTÞz0 ðTm Þ 2 zðTÞ þ z0 ðTm Þ 3 1 4 ð16:5Þ ¼W c c þ c f ðcÞ ¼ kB T 2 3 4 where W is a coefficient representing the energy cost for the system to overcome the nucleation barrier zðTÞ and z0 ðTm Þ represents, the location of the nucleation hump on the c axis and the solidification potential both of which are melting temperature dependent. This kind of asymmetric Landau potential has been utilized in the phase field model for the elucidation of solidification phenomena such as metal alloys or polymer crystallization (25). It should be cautioned that the coefficient of the cubic order must be nonzero in order to apply the Landau potential to the first-order phase transition; otherwise, the potential is applicable only to a second-order phase transition or at equilibrium where the two minima are equivalent (Fig. 16.3). The uniqueness of the present theory of polymer solidification is that these model parameters W, z, and z0 can be related to the material properties of the individual components and the experimental conditions (25). This Landau-type free energy of solidification has been successfully applied to describing the spatial temporal emergence of polymer single crystals, dendrite growth patterns, and dense lamellar branching in spherulites (25–29). The establishment of phase diagrams using the phase field model of crystallization is accomplished by minimizing the free energy of the mixture with respect to the nonconserved order parameters and finding the minimum free energy of the mixture at each composition. We then minimize f ðc; fÞ with respect to c in order to find the roots, leading to

or

@f ðc; fÞ ¼ Wðc zÞðc z0 Þ þ 2xca ð1 fÞ ¼ 0 @c Wðc zÞðz0 cÞ ¼ 2xca ð1 fÞ

ð16:6Þ

For the neat crystal, that is, f ¼ 1, the lower right-hand side of equation 16.6 becomes zero, thus cc ¼ z signifies the unstable potential hump, and cc ¼ z0

Chapter 16 Phase Field Approach to Thermodynamics

481

Figure 16.3 The variation of free energy of crystallization as a function of crystal order parameter, c of a pure homopolymer, showing a symmetric double well at equilibrium between c ¼ 0 and c ¼ 1 representing the melt and the solid phase, respectively. During supercooling to various temperatures, the shape of free-energy transforms to asymmetric double wells having the crystal order parameter at the solidification potential less than unity, reflecting the imperfect crystal (i.e., crystallinity of less than 1).

represents the solidification potential well of the neat crystal. In the blend, that is, f < 1, the solidification potential minimum cc ¼ cmin will shift away from z0 . At a given temperature of crystallization and a given concentration, W representing the penalty for overcoming the unstable potential hump. Since z and z0 are known, xca can be estimated analytically from the heat of fusion, DHu of PP crystals in the blends in accordance with the relationship of equation 16.6, that is, xca / W ¼ 6½ðDHu =kB TÞð1 T=Tm0 Þð1=2 zÞ1 . Alternatively, the self-consistent solution to equation 16.6 can be obtained by means of the steepest descent method with a tolerance of 1e 7. Subsequent to this minimization, we further calculate the coexistence curves based on the common tangent algorithm in what follows: @f ðc; fÞ @f ðc; fÞ f ðcmin ; fa Þ f ð0; fb Þ ¼ ¼ @f c¼cmin ;fa @f c¼0;fb fa f b

ð16:7Þ

16.3.3 Prediction of Phase Diagram Topologies In this section, we shall describe the various topologies of possible phase diagrams using the proposed thermodynamic model for a crystal–amorphous blend. Most polyolefin blends exhibit either an LCST or a UCST or a combination of both. In order to model the combined LCST–UCST or hour-glass phase diagram, the

482

Polyolefin Blends

empirical expression of xaa in equation 16.2 may be treated only as a function of temperature, viz., xaa ¼ A þ B=T þ C ln T

ð16:8Þ

Applying the conditions that xaa ¼ xcrit at the critical temperature of the UCST and LCST, one can treat these coefficients to be a function of only a single adjustable parameter A, that is, 2 3 1 1 6 7 lnðTUCST Þ lnðTLCST Þ 7 B ¼ ðxcrit AÞ6 ð16:9Þ 4 5 1 1 TUCST lnðTUCST Þ TLCST lnðTLCST Þ and C ¼ ðxcrit AÞ

TUCST TLCST TUCST lnðTUCST Þ TLCST lnðTLCST Þ

ð16:10Þ

where xcrit

1 1 1 2 ¼ pﬃﬃﬃﬃ þ pﬃﬃﬃﬃ 2 r1 r2

ð16:11Þ

On the basis of Equations 16.8–16.11, we have solved self-consistently for various phase diagram topologies of a hypothetical crystal–amorphous polymer blend having a critical LCST temperature at 252 C and a critical UCST temperature at 202 C, but the melting transition temperatures and crystal–amorphous interaction parameters vary. Figure 16.4 exhibits the influence of melting point on the phase diagram in columns from the top to the bottom as well as the effect of the crystal– amorphous interaction energy on the phase diagram in rows from left to right. As can be witnessed in Fig. 16.4a, the coexistence of the LCST and UCST can be established in which the UCST is intersected by the crystal solid–liquid transition, displaying liquid–liquid, crystal–liquid coexistence regions, and the neat crystal gap shown by the solidus line at the high crystalline polymer concentrations. With increasing repulsive crystal–amorphous interaction parameter, i.e., the solidus line moves toward the pure crystal component axis and concurrently the neat crystal gap become narrower suggesting that more solvent is rejected out from the solidus phase (Fig. 16.4a–c). On the contrary, upon raising the melting transition temperature of the crystalline polymer constituent, the solidus region increases as the solidus line moves up (Fig. 16.4a and d). With continued increase of the melting point, the crystal-melt transition is now intersecting with the LCST, thereby widening the crystal–liquid coexistence region like an hour-glass phase diagram (Fig. 16.4g). This kind of hour–glass phase diagram further transforms to almost completely immiscible

Chapter 16 Phase Field Approach to Thermodynamics

483

Figure 16.4 Hypothetical phase diagrams for crystal–amorphous polymer blends exhibiting a combined LCST and UCST intersected by the crystal–melt transition gap bound by the liquidus and solidus lines, showing (a) effect of melting temperature of the pure crystal component increasing from top to bottom (182, 223, and 262 C) and (b) the effect of repulsive crystal–amorphous interaction energy increasing from left to right. xca parameter is varied from 0.01, 0.1, and 1 at the melting temperature by setting the UCST temperature to 182 C and the LCST temperature to 262 C. The melting enthalpy of the constituent crystal was taken as 1500 cal mol1.

crystal solid—amorphous liquid (i.e., tree-trunk) gap with increasing the repulsive crystal–amorphous interaction (Fig. 16.4i). These theoretical predictions indicate clearly that it is possible to discern intricate phase diagram topologies encompassing the LCST coexistence curve coupled with the crystal–melt transitions, the combined LCST/UCST phase diagram and the solid–liquid coexistence region bound by the solidus and liquidus lines, all the way to the complete immiscibility of various PP/EPDM blends. The present calculation strongly suggests that the intricate phase diagrams of the iPP/EPDM reported by Chen et al. (6) and that of sPP/EPDM by Ramanujam et al. (7) are indeed possible and also the immiscibility perception for the melt blends of commercial PP/EPDM reported in literature is consistent with the present predictions.

484

Polyolefin Blends

16.3.4 Comparison with Experimental Phase Diagrams of PP/EPDM Blends It is encouraging that diverse phase diagram topologies predicted by the present theory for the hypothetical crystal–amorphous blend captures the observed trends of the phase diagrams of PP isomers/EPDM blends, it is essential to test directly with the experimental phase diagrams reported by Chen (6) for the iPP/EPDM blend (Fig. 16.1) and the sPP/EPDM blend of Ramanujam et al. (7) (Fig. 16.2) by utilizing the actual material parameters and the experimental conditions of iPP and sPP and their blends with EPDM. The materials’ parameters utilized were the enthalpy of fusion of iPP, DHiPP ¼ 2110 cal mol1 (30), the statistical segment length, riPP ¼ 1800, and rEPDM ¼ 1000, respectively. The equilibrium melting point of iPP is taken as 162.5 C and the LCST was calculated by setting A ¼ 0:01. The xca value was estimated to be 0.8 based on the heat of fusion of iPP crystal via the analytical expression of xca / W described earlier. Note that the density of the iPP crystal is approximated as unity so that the volume fraction roughly corresponds to the weight fraction of the PP. The conversion of the material parameters to the model parameters may be found elsewhere (23). There are two possible scenarios to draw the phase diagram depending on the value of xca as demonstrated in Fig. 16.4g and i. With xca ¼ 0:1, the solidus and liquidus lines are coincided, which intersected with the LCST that fitted reasonably well with the way the melting transition points and the experimental cloud point phase diagram were drawn in the original paper. However, in view of the estimated xca ¼ 0:8 value, the phase diagram seems more like that in Fig. 16.4i with the solidus line being located right on the pure iPP ordinate. Figure 16.5 shows the comparison between the self-consistently solved coexistence curves and the experimental cloud points of the iPP/EPDM blends and the melting points of iPP in the blends. This observation suggests that some cloud point data (UCST) falling below the liquidus line may be already in the solid crystal–amorphous liquid coexistence region suggesting the nonequilibrium nature of the cloud point determination methodology adopted by Chen et al. (6), especially below the crystal–melt transition. In addition, this kind of theoretical knowledge was not available at the time of their cloud point experiments (6). To alleviate such a complex interplay between the liquid–liquid phase separation and the crystallization, one idea that was developed was to replace iPP with sPP by virtue of the lower melting temperature of sPP relative to that of iPP. Upon replacing iPP with sPP, Ramanujam et al (7) were able to decouple the LCST of the sPP/EPDM with a minimum at around 150 C from the melting transitions of the sPP crystalline constituent located at 127 C (Fig. 16.2). However, the sPP/EPDM blend exhibited the UCST peak at around 100 C, which was buried under the melting point depression curve of sPP. The self-consistent solution gives the combined LCST/UCST phase diagram for this system by determining the model parameters using the material parameters for the sPP/EPDM blend. The material parameters utilized were the enthalpy of fusion of sPP (DHsPP ¼ 1912 cal mol1 ), the equilibrium melting temperature for sPP was approximated as 127 C, and the

Chapter 16 Phase Field Approach to Thermodynamics

485

Figure 16.5 Comparison between the self-consistently solved coexistence line and the cloud points (filled triangle) of iPP/EPDM and the melting points (filled circles). The solidus and liquidus lines are virtually overlapped (dots), but the existence of both lines is manifested by the kink in the LCST coexistence line. The phase diagram was calculated using the material parameters, DHiPP ¼ 2110 cal mol1 , Tm ¼ 162:5 C, riPP ¼ 1800, rEPDM ¼ 1000, and xca ¼ 0:8 at Tm .

statistical segment length, rsPP ¼ 1800 and rEPDM ¼ 1000 that roughly correspond to the molecular weights of the constituent polymers. The value was calculated through Equations 16.8–16.11 by setting A ¼ 0:01. Again the density of the sPP crystal is taken as unity so that the volume fraction in the theoretical description and the weight fraction of the experiment can be used interchangeably. The model parameter xca is estimated to be around 0.8 at 127 C from the relation of xca / W ¼ 6ðDHu =kB TÞð1 T=Tm0 Þð1=2 zÞ1 given above, which fits reasonably well to the experimental cloud point curves (Fig. 16.6). As shown in Fig. 16.6, the coexistence line afforded by the self-consistent solution shows a good fit with the LCST cloud points. The calculated liquidus line also closely matches the crystal-melting curve that is found to situate between the LCST coexistence curve and the UCST spinodal gap. The solidus line is situated right on the pure sPP crystal axis. However, the UCST coexistence line cannot be discerned since it has merged with the liquid and solidus lines, leaving the spinodal gap representing the unstable liquid envelope that was buried underneath the liquidus line. The tail end of the liquidus line curving downward asymptotically to the EPDM axis manifests the influence of the UCST, which in turn indicates that the UCST is merged with the liquidus line. At 67 C, it can be noticed that the buried spinodal spans all the way to 10 wt% of sPP. This probably explains the occasional observation

486

Polyolefin Blends

Figure 16.6 Comparison between the self-consistently solved coexistence curves and experimentally observed LCST (open diamonds) and UCST spinodal gap (open diamonds) showing the liquidus (denoted by open circles) and solidus lines on the pure sPP axis. The material parameters utilized were the enthalpy of fusion of sPP DHsPP ¼ 1912 cal mol1, Tm ¼ 123 C, rsPP ¼ 1800, rEPDM ¼ 1000, and xca ¼ 0:8 at the melting temperature.

of the spinodal type phase separated domains found in the 10/90 sPP/EPDM blend that experience deep quenches at around 27–67 C.

16.4 PHASE FIELD MODELING ON POLYMER PHASE TRANSITIONS 16.4.1 Theory on Phase Separation Dynamics and Morphology Evolution To elucidate the spatiotemporal emergence of crystalline structure and liquid–liquid phase separation in these polyolefin blends, we employ the time dependent Ginzburg–Landau (TDGL) equations pertaining to the conserved concentration order parameter and the nonconserved crystal order parameter. The spatiotemporal evolution of the nonconserved order parameter c, known as TDGL model-A equation (31,32), may be expressed as @c df ¼ Gc @t dc

ð16:12Þ

Chapter 16 Phase Field Approach to Thermodynamics

487

where Gc is the rotational mobility that is related to the viscosity of the system with a unit of frequency (s1 ) and df =dc is the functional derivative of the free energy density with respect to the order parameter, which is analogous to the pseudopotential. The product of the mobility and the potential represents the flux. For the case of the conserved order parameter such as volume fraction or concentration order parameter (f), we use the TDGL model B equation also known as the Cahn–Hilliard equation (33) @f df ¼ rGf r @t df

ð16:13Þ

where Gf is the mobility of the system, which is related to the self-diffusion coefficients of each constituent that obeys the Onsager reciprocal relation having a unit of m2 s1 . The functional derivative, which is analogous to the chemical potential pertaining to an arbitrary order parameter &, is defined as d=d& ¼ @=@& rð@=@r&Þ. This functional derivative is particularly important for a system whose total free energy of the system, f , comprises both the local free energy as well as the nonlocal free energy contributions that accounts for the moving interfaces. The nonlocal free energy may be given as fnonlocal ¼ f

kc kf jrcj2 þ jrfj2 2 2

ð16:14Þ

where kc and kf represent the coefficients of the interface gradient of respective order parameters. Then the total free energy of the system f may be represented by the total sum of local, nonlocal, and the coupling free energy terms in what follows: h i f kc ð1 fÞ lnð1 fÞ þ xaa fð1 fÞ f ¼ f f ðcÞ þ jrcj2 þ lnðfÞ þ 2 r1 r2 ð16:15Þ kf þ jrfj2 þxca c2 fð1 fÞ 2 The functional derivative of the free energy with respect to individual order parameters can thus be calculated, that is, df ¼ f½WðcÞðc zÞðc z0 Þ þ 2xca ð1 fÞc kc r2 c dc df 1 þ ln f 1 þ lnð1 fÞ þ xaa ð1 2fÞ kf r2 f ¼ df r1 r2 zðTÞz0 ðTm Þ 2 zðTÞ þ z0 ðTm Þ 3 1 4 c c þ c þW 2 3 4 kc 2 2 þ jrcj þxca ð1 2fÞc 2

ð16:16Þ

ð16:17Þ

488

Polyolefin Blends

Inserting Equations 16.16 and 16.17 into Equations 16.12 and 16.13, respectively, and subsequently nondimensionalizing them, the following equations of motion can be deducted, @c ~ c ff½WðcÞðc zÞðc z Þ þ 2x ð1 fÞc ~kc r~2 cg ¼ G 0 ca @t 9 8 1 þ ln f 1 þ lnð1 fÞ > 2 > > ~ þ xaa ð1 2fÞ ~kf r f > > > > > > > r r 1 2 > > > > = < @f zðTÞz ðT Þ zðTÞ þ z ðT Þ 1 2 m m 0 0 ~ 2 3 4 ¼r þW þ c c c > > @t 2 3 4 > > > > > > > > > > ~ k 2 c > > 2 ~ ; : þ rc þxca ð1 2fÞc 2

ð16:18Þ

ð16:19Þ

where the nondimensionalization was carried out by introducing dimensionless quantities for time and space coordinates, viz., t ¼ Gf t=‘2 and ~x ¼ x=‘. This allows ~ c ¼ Gc ‘2 =Gf , ~kc ¼ kc =‘2 and ~kf ¼ kf =‘2 us to obtain dimensionless variables, G that govern the spatio temporal growth of domain morphology. The advantage of the present model is that these phase field model parameters can be related to the material parameters that themselves are all supercooling dependent.

16.4.2 Dynamics of Crystal Growth in a Phase Separating System: iPP/EPDM Blends In view of the complex phase diagram of iPP/EPDM blend, it can be anticipated that crystallization kinetics and morphology development in iPP/EPDM blends would be complicated by the aforementioned phase separation. However, the thermodynamic phase diagram depicted in Fig. 16.1 certainly serves as guidance for mimicking the trajectory of thermal jump experiments. The 2D simulation has been carried out at the 50/50 iPP/EPDM blend composition (i.e., a near critical composition) following a temperature jump from a single phase temperature of 142 C to a temperature of 155 C, which is below the crystalmelting temperature, but it is above the LCST. The thermodynamic parameters used for the iPP/EPDM simulation are W ¼ 10, z ¼ 0:1, z0 ¼ 0:98, xca ¼ 0:8, ~kc 1, ~kf 1, r1 ¼ 18, r2 ¼ 10, and xaa ¼ 0:25. We have chosen the kinetic parameter ~ c 10 to reflect the fast crystallization rate as seen in the iPP melt crystallization G studies. The numerical simulations were carried out on several grid sizes of 128 128, 256 256, and 512 512 with varying time steps of 0.001, 0.0005, and 0.0001 to ensure the stability of the simulation. The results of the 512 512 simulations are shown in Fig. 16.7 exhibiting the emerged bicontinuous structure that is seemingly driven by liquid–liquid phase separation through spinodal decomposition in the composition field (upper row). The spinodal process is known to be spontaneous, thus any unstable fluctuation can grow very rapidly as this SD process does not required any energy to overcome. As can be seen in the bottom row, the nucleation (solid–liquid) of spherulite develops rather late relative to the phase separation, but it catches up very rapidly and eventually outgrows the SD domains.

Chapter 16 Phase Field Approach to Thermodynamics

489

Figure 16.7 Temporal evolution of the crystalline microstructure in the 50/50 iPP/EPDM blend, following a T-quench from the isotropic melt to a supercooled temperature below both the UCST spinodal gap, showing the growth of spherulitic front in the concentration field, but the overgrowth of this spherulitic boundary on the bicontinuous SD domain structures can be seen clearly only in the enlarged version.

The spherulitic boundary can be discerned in the enlarged picture, which is outgrowing over the interconnected SD domains. Although the simulated growth dynamics show interesting behavior, we did not have such knowledge at the time of the experiments performed some 10 years ago. Although the experiment was not ideal to quantitatively test with the present theoretical simulation regarding the competition between the phase separation and crystallization, it is still worthwhile to revisit what was observed experimentally at that time. Figure 16.8 shows the development of spherulitic morphology in the 50/50 iPP/ EPDM blend (6) upon cooling from 230 C to ambient at a slow cooling rate of 0.5 C min1. As depicted in Fig. 16.8a, the polarized optical micrograph under the cross

490

Polyolefin Blends

Figure 16.8 Morphology development in a 50/50 iPP/EPDM during cooling from 230 C to room temperature at a slow cooling rate of 0.5 C min1. (a) Polarized optical micrograph under cross polarizer’s clearly showing the maltese cross pattern indicative of the spherulite structure. (b) Four-lobe clover leaf pattern in SALS in the Hv configuration confirming the existence of the spherulite texture. (c) Polarized optical micrographs under parallel polarizer’s showing the phase-separated morphology and (d) the ring pattern in the Vv configuration dominated by the concentration fluctuations of the phaseseparated domains.

polarizers clearly reveals the maltese-cross pattern indicative of the iPP spherulite structure. The corresponding light scattering study exhibits a four-lobe clover pattern in the horizontal–vertical (Hv) configuration, which further confirms the existence of the spherulite texture (Fig. 16.8b). In the unpolarized configuration, the structure in the optical micrograph (Fig. 16.8c) is reminiscent of the phase-separated interconnected SD morphology. The corresponding light scattering under the vertical– vertical (Vv) configuration in Fig. 16.8d reveals a large scattering halo suggestive of domination by the concentration fluctuations of the phase-separated domains over the orientation fluctuations. That is to say, the orientation fluctuation can be attributed to the emerged crystalline spherulitic morphology, the size of which is too large that it primarily contributes to the main beam. Nevertheless, the observed overgrowth

Chapter 16 Phase Field Approach to Thermodynamics

491

of spherulitic structure on the existing interconnected SD structures can be identified, which accords very well with the above simulated structure. Hashimoto et al. (34) have also shown a similar growth trend in the PP/EPR blends exhibiting the spherulites outgrowing the SD domains upon quenching the system from a high temperature two-phase state to a temperature that is lower than the crystal-melting temperature yet above the LCST temperature of the blend.

16.4.3 Dynamics of Crystal Growth in a Phase Separating System: sPP/EPDM Blends In Fig. 16.9, the morphology development in a 50/50 sPP/EPDM blend isothermally quenched from 128 C (a single phase) to 100 C (two phases under the UCST) for a prolonged period is shown (7). The samples were analyzed under both unpolarized and cross-polarized configurations as time progressed. Even though phase separation occurs at this temperature, the cross-polar micrographs show the occurrence of crystallization, that is, the PP-rich domains are stretched along the bicontinuous regions dictated by the SD structure. A similar temperature quench has been carried out on the 70/30 sPP/EPDM blend that shows a different trend (7). As evident in Fig. 16.10, the EPDM-rich phase is dispersed in the form of globular droplets within the matrix of the PP-rich region. Thus the sPP crystals grow along channels by meandering around these EPDM-rich droplet domains, forming EPDM islands in the sea of sPP crystalline continuum. It may be inferred that in both the 50/50 and 70/30 blends, the crystal growth of sPP is confined within the phase-separated microstructure. This correlation between the crystalline and phase separated structures may be attributed to the strong crystal amorphous interaction that was predicted in the theoretical phase diagram. A similar result has also been reported by Crist and Hill (35) during thermal quenching of the blends of polyethylene and hydrogenated polybutadiene near the critical concentration. The 2D simulation of the kinetics of phase separation and morphology evolution in sPP/EPDM blends confirms this conjecture. Initially the system was set at a temperature (127 C) that is slightly higher than the melt temperature, but it is lower than the LCST temperature of the blend. First, a thermal jump was carried out from 127 to 157 C that lies inside the spinodal envelope of the LCST. No crystallization of the PP component was involved during such a jump beyond the melting temperature. The morphology evolution is seemingly governed by the simple liquid–liquid phase separation. As shown in Fig. 16.11a, the formation of the spinodally decomposed structures can be witnessed in the concentration order parameter field. As phase separation continues, the coarsening of the phases takes place as evident by the increase in periodic wavelength of the phase-separated bicontinuous domains (Fig. 16.11b). Another experiment was carried out by quenching the 50/50 blend from the initial temperature of 127 C (i.e., the single phase) to a temperature of 77 C that lies underneath the liquidus line and the UCST spinodal. The same set of parameters for the iPP/EPDM blend was employed except for the thermodynamic parameterxca ¼ 0:8 to account for the buried UCST immiscibility gap and the kinetic

492

Polyolefin Blends

Figure 16.9 Optical micrographs obtained for a 50/50 sPP/EPDM blend isothermally quenched at 100 C. Left column depicts the evolution of phase separation under the unpolarized condition and right column indicates the growth of crystals under the cross-polarized condition.

~ c 0:1 to reflect the slower crystallization of sPP relative to iPP. parameter G Nucleation was triggered by thermal noise in the crystal order-parameter field. During the initial stages, the formation of the spinodal phase-separated structures can be witnessed in the concentration field, but the crystals have yet to emerge in the crystal order-parameter field (Fig. 16.12). With the progression of time, some nuclei formed in the PP-rich regions because the probability of the nuclei to survive is much

Chapter 16 Phase Field Approach to Thermodynamics

493

Figure 16.10 Optical micrographs obtained for a 70/30 sPP/EPDM blend isothermally quenched at 100 C. Left column depicts the evolution of phase separation under the unpolarized condition and right column indicates the growth of crystals under the cross-polarized condition.

greater than those in the PP-poor regions. It can be envisaged that the growing crystalline regions are strongly dictated by the spinodal template, thereby loosing the radial growth habit such as spherulitic growth. This observation is indeed what was observed in the actual experiment of the sPP/EPDM blends (Fig. 16.9). In Fig. 16.13, the simulated structures are shown for the 70/30 sPP/EPDM mixture under the same conditions as described in the preceding case. The asymmetry in the composition results in the change of mechanism of phase separation from the spinodal to the nucleation-growth mechanism. Now the minority EPDM-rich phase forms the droplets in the continuum of sPP-rich phase. As can be anticipated, the PP crystallization lags behind the liquid–liquid phase separation because phase separation must occur first in order for the sPP phase to reach or exceed its critical

494

Polyolefin Blends

Figure 16.11 Emerged bicontinuous structure following a temperature jump from the isotropic melt between the LCST and the melting transition into the LCST gap, which is presumably driven by liquid– liquid phase separation through spinodal decomposition in the 50/50 sPP/EPDM mixture: (a) 1000 s and (b) 3000 s.

concentration so that the crystal nucleation can occur. The crystallization of sPP in the blends proceeds by weaving around the discrete EPDM domains. On the contrary, the sPP forms the droplets in the continuum of EPDM in the case of 30/70 sPP/EPDM. It is striking to discern that sPP crystals are strictly confined in the sPP-rich droplets (Fig. 16.14). The crystallization of sPP lags behind the liquid–liquid

Figure 16.12 Competition between the liquid–liquid phase separation through spinodal decomposition and the crystalline structure formation in the 50/50 sPP/EPDM blend. The crystallization occurs with the preformed SD networks. The top and bottom rows represent the temporal evolutions of the concentration field and the corresponding crystal order field.

Chapter 16 Phase Field Approach to Thermodynamics

495

Figure 16.13 Competition between the liquid–liquid phase separation through nucleation and growth showing the droplet domains and the crystalline structure formation in the 70/30 sPP/EPDM blend. The crystalline sPP component being the major phase, the sPP crystallization occurs in the matrix by weaving around the EPDM domains. The top and bottom rows represent the temporal evolutions of the concentration field and the corresponding crystal order field.

Figure 16.14 Competition between the liquid–liquid phase separation through nucleation and growth and the crystalline structure formation in the 30/70 sPP/EPDM blend. The crystallization of sPP is confined to the sPP-rich droplets. The top and bottom rows represent the temporal evolutions of the concentration field and the corresponding crystal order field.

496

Polyolefin Blends

phase separation occurring through the nucleation and growth mechanism except that the former is governed by the solid–liquid phase transition as opposed to the latter case of liquid–liquid phase separation. Again, the sPP concentration must reach the threshold value for the nucleation to occur that can only be achieved through phase separation.

16.5 CONCLUSIONS We have demonstrated the interplay of solid–liquid phase separation and liquid–liquid phase separation in the blends of iPP/EPDM and sPP/EPDM showing the influence of PP tacticity on phase diagrams. In the establishment of the experimental phase diagrams for the iPP/EPDM blend, we observed the intersection of the solid–liquid coexistence curves with the liquid–liquid coexistence curves that prompted the study of the sPP/EPDM blend to decouple the two competing processes. We have developed a thermodynamic model based on the crystal–amorphous interaction in addition to the conventional amorphous–amorphous interaction of the FH theory. With this modification, one can predict various phase diagrams of crystalline–amorphous polymer blends exhibiting both LCST and UCST behaviors coupled with the melting transition of one of the components. The crystal–liquid gap was bound by the solidus and liquidus lines that cannot be achieved by the original Flory diluent theory. On the basis of the present modified free energy, the interplay between phase separation and crystallization is explicable in the frame work of the phase field approach based on the TDGL-Model C equations of motion for the concentration and crystal order parameters. Various morphological features have been simulated that are consistent with the observed crystal morphologies of iPP/EPDM as well as of sPP/EPDM blends.

NOMENCLATURE c f z kc Gc Gf kf z0 xaa xca r1 ; r2 Tm W

Crystal order parameter corresponding to linear crystallinity Concentration order parameter or volume fraction Peak position of the energy barrier for solidification Coefficient tensor of the interface gradient of the c field Mobility related to propagation of crystal-melt interface of the c field Mutual diffusion coefficient related to translational mobility of the constituents Coefficient of the concentration gradient of the f field Crystal order parameter at the equilibrium solidification potential Amorphous–amorphous interaction parameter Crystal-amorphous interaction parameter Statistical segment lengths of components 1 and 2 Melting point at the extrapolated zero heating rate Coefficient representing the free energy penalty to overcome the unstable potential hump pertaining to c

Note: Symbols with tilde signify the dimensionless quantities

Chapter 16 Phase Field Approach to Thermodynamics

497

REFERENCES 1. A. Y. Coran, Handbook of Elastomers: New Developments and Technology, A. K. Bhowmick and H. L. Stephens (eds.), Marcel Dekker, New York, 1988, p. 249. 2. L. A. Goettler, J. R. Richwine, and F. J. Wille, Rubber Chem. Technol, 55, 1558 (1982). 3. A. Y. Coran and R. Patel, Rubber Chem. Technol. 53, 141 (1980); Rubber Chem. Technol. 53, 781 (1980); Rubber Chem. Technol. 54, 892 (1981); Rubber Chem. Technol. 56, 210 (1983). 4. A. Y. Coran and R. Patel, US Patent 4,297,453 (1981); US Patent 4,310,638 (1982); US Patent 4,350,470 (1982). 5. D. J. Lohse and W. W. Graessley, in: Polymer Blends, Vo.. 1, D. R. Paul and C. B. Bucknall (eds.), 2000, pp. 219–237, chapter 8. 6. C. Y. Chen, W. Md. Z. Yunus, H. -W. Chiu, and T. Kyu, Polymer 38, 4433 (1997); M.S. Thesis, University of Akron, Akron, Ohio, 1993. 7. A. Ramanujam, K. J. Kim, and T. Kyu, Polymer 41, 5375 (2000); A. Ramanujam, M.S. Thesis, University of Akron, Akron, Ohio, 1998. 8. P. J. Flory, Principles of Polymer Chemistry, Cornell University Press, Ithaca, NY, 1953. 9. P. J. Flory, J. Chem. Phys., 17, 223 (1949). 10. J. M. Prausnitz, R. N. Lichtenthaler, and E. G. de Azevedo, Molecular Thermodynamics of FluidPhase Equilibria, Prentice-Hall, NY, 1999. 11. R. Konigsveld and W. H. Stockmayer, Polymer Phase Diagrams, Oxford University Press, Oxford, 2001. 12. T. Kyu and H. -W. Chiu, Phys. Rev E 53, 3618 (1996). 13. T. Nishi and T. T. Wang, Macromolecules 8, 909 (1975). 14. W. R. Burghardt, Macromolecules 22, 2482 (1989). 15. J. C. Canalda, Th. Hoffmann, and J. Martinez-Salazar, Polymer 36, 981 (1995). 16. Y. W. Cheung and R. S. Stein, Macromolecules 27, 2512 (1994). 17. Y. W. Cheung, R. S. Stein, J. S. Lin, and G. D. Wignall, Macromolecules 27, 520 (1994). 18. P. M. Cham, T. H. Lee, and H. Marand, Macromolecules 27, 4263 (1994). 19. J. P. Penning and R. St. J. Manley, Macromolecules 29, 77 (1996). 20. J. P. Penning and R. St. J. Manley, Macromolecules 29, 84 (1996). 21. K. Fujita, T. Kyu, and R. St. J. Manley, Macromolecules 29, 91 (1996). 22. H. Tanaka and T. Nishi, Phys. Rev. A, 39, 783 (1989). 23. R. A. Matkar and T. Kyu, J. Phys. Chem. B. 110, 12728 (2006). 24. J. D. Gunton, M. San Miguel, and P. S. Sahni, in: Phase Transitions and Critical Phenomena, Vol. 8 C. Domb and J. L. Lebowitz (eds.), 1983, p. 269, chapter 3. 25. H. Xu, R. A. Matkar, and T. Kyu, Phys. Rev., E 72, 011804 (2005). 26. H. Xu, W. Keawwattana, and T. Kyu, J. Chem. Phys., 123, 124908 (2005). 27. R. Kobayashi, Physica D 63, 410 (1993). 28. A. Wheeler, W. J. Boettinger, and G. B. McFadden, Phys. Rev. A, 45, 7424 (1992). 29. T. Kyu, R. Mehta, and H. -W. Chiu, Phys. Rev. E, 61, 4161 (2000). 30. S. Brandrup and E. H. Imergut, Polymer Handbook, Vol. 5, Interscience, New York, 1975, p. 24. 31. S. -K. Chan, J. Chem. Phys. 67, 5755 (1977). 32. P. R. Harrowell and D. W. Oxtoby, J. Chem. Phys., 86, 2932 (1987). 33. J. W. Cahn and J. E. Hilliard, J. Chem. Phys., 28, 258 (1958). 34. N. Inaba, T. Yamada, S. Suzuki, and T. Hashimoto, Macromolecules 21, 407 (1988). 35. B. Crist and M. J. Hill, J. Poly. Sci. Phys., 35, 2329 (1997).

Part III

Polyolefin/Nonpolyolefin Blends

Chapter

17

Compatibilization and Crystallization of Blends of Polyolefins with a Semiflexible Liquid Crystalline Polymer Liliya Minkova1

17.1 BLENDS OF POLYOLEFINS (HIGH DENSITY POLYETHYLENE AND ISOTACTIC POLYPROPYLENE) WITH A SEMIFLEXIBLE LIQUID CRYSTALLINE POLYMER 17.1.1 Introduction Much attention has been paid recently to blends of commercial thermoplastics with liquid crystalline polymers (LCPs) (1). The main benefits expected from the use of LCPs as blend components are (1) the pronounced reduction of the melt viscosity, with consequent improvement of processability, and (2) the reinforcing effect granted by the immiscible LCP particles, which can attain oriented fibrillar morphology when the blend is processed under elongational flow conditions. In situ composites (2) have been prepared by the addition of LCPs into a great variety of flexible resins, such as polyamides, polyesters, polycarbonates, and so on, and their morphology–processing–property relationships have been studied. For the particular case of polyolefin/LCP blends the most available thermotropic LCPs, which belong to the classes of wholly aromatic copolyesters

1

Institute of Polymers, Bulgarian Academy of Sciences, Acad. G. Bonchev str. Bl.103A, 1113 Sofia, Bulgaria Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

501

502

Polyolefin Blends

and copolyesteramides, show very poor compatibility and interphase adhesion toward polyolefins. Moreover, the processing temperatures of common aromatic LCPs are in the 300 C range and are, therefore, much higher than those used for polyolefin processing. In fact, the reinforcement of polypropylene (PP) with different LCPs has normally led to blends with no improvement in the tensile strength, although some modulus enhancements were sometimes observed (3–5). As for the blends of polyethylene (PE) with LCPs, the studies confirm that the poor compatibility of the two polymers, coupled with the strong differences between their melting/processing temperatures, prevents the attainment of good morphologies and enhanced mechanical properties (6–11). It has been assumed that a semiflexible LCP, containing aliphatic spacer in the main change and having lower melting point, will be more suitable to be blended with polyolefins. In fact, slightly better results were obtained by blending linear low density polyethylene (LLDPE) with a semiflexible liquid crystalline polymer: SBH 112 by Eniricerche, Milan, synthesized from sebacic acid (S), 4,40 -dihydroxybiphenyl (B), and 4-hydroxybenzoic acid (H) in the mole ratio 1:1:2 (12). For these blends, the phase dispersion was good and the size distribution of the LCP droplets was fairly narrow. Moreover, the LCP phase was shown to play a nucleating effect for the crystallization of the LLDPE matrix (12,13). The mechanical characterization showed no reduction of the tensile strength and a 50% modulus increase over the LCP concentration range 0–20%, whereas the elongation to break decreased markedly only for SBH contents higher than 10%.

17.1.2 Blends of High Density Polyethylene (HDPE) with LCP The possibility of reinforcing HDPE by blending it with an LCP is based on the successful improvement of phase compatibility and interfacial adhesion of these two structurally unlike polymers. The addition of compatibilizing agents into intrinsically immiscible polymer blends can have a substantial merit to solving the problems of poor dispersion and low adhesion. Among the substances exhibiting compatibilizing activity, block or graft copolymers made up of chain segments having chemical structure and/or solubility parameters similar to those of the polymers being blended are most promising. The approach that has been considered in our laboratories consists of the synthesis of PE–LCP block or graft copolymers and of their use as compatibilizing agents for PE/LCP blends. Two main routes have been used for the PE-g-LCP synthesis. The first one is melt polycondensation of sebacic acid (S), 4,40 dihydroxybiphenyl (B), and 4-hydroxybenzoic acid (H), carried out at temperatures up to 280 C in the presence of an oxidized low molar mass PE sample containing free carboxylic groups (PEox) (14). The second one is reactive blending of PEox and a semiflexible liquid crystalline polyester (SBH 1:1:2) (50/50 w/w), at 240 C in a Brabender mixer, in the presence of Ti(OBu), catalyst, for different mixing times (15, 60, and 120 min) (15). The formation of PE-g-SBH copolymers is shown schematically in Fig. 17.1a for the first route and in Fig. 17.1b for the second route.

503

Chapter 17 Compatibilization and Crystallization of Blends + S, B, H

+ CH 3COOH

COOH

COO (SBH) HOCO (CH 2) 8 COOH

CH 3 COO

=S =B

OCOCH3

CH3COO

COOH

=H

(a)

H 2O COOH

(SBH)n =

+ (SBH) n

CO (CH 2)8 CO

+

COO

n

O

O

n

(SBH) n – m

O

CH 3COOH (SBH) m

CO

n

(b)

Figure 17.1 (a) Reaction scheme for the production of COP. (b) Reaction scheme for the production of the reactive blend. (From References 14 and 15 with permission from John Wiley & Sons, Inc.)

A physical blend between PEox and SBH 112 50/50 w/w has also been prepared in a Brabender apparatus at 240 C for 6 min without a catalyst for comparison. The polycondensation product COP, the reactive blends COP15, COP60, and COP120, and the physical blend MIX have been extracted with PE solvents with increasing solubility parameters and boiling points (toluene and xylene), in order to gain information on the presence of a true copolymer in the different fractions. The soluble fractions and the residues have been analyzed by Fourier transform infrared (FTIR) spectroscopy, (nuclear magnetic resonance) NMR spectroscopy, thermogravimetry (TG and DTG), differential scanning calorimetry (DSC), and scanning electron microscopy (SEM). All analytical procedures concordantly show that PE-g-SBH copolymers have, in fact, been obtained. For COP, both PEox and SBH chain segments are present, with different relative ratios, in all fractions of the polycondensate. Moreover, IR analysis has shown that a fairly quantitative esterification of the PEox carboxyl groups takes place under the adopted conditions. For the reactive blends, PE-g-SBH copolymers with different compositions, arising from differences of either the number of PEox carboxylic groups entering the transesterification or the length of grafted SBH branches, are formed as a result of blending. Qualitative IR analyses and quantitative TG measurements have shown that the amount of copolymers increases strongly with the mixing time of the reactive blends. The calorimetric data in Table 17.1 also demonstrate that the soluble and insoluble fractions of MIX consist essentially of neat PEox and SBH, respectively,

504

Polyolefin Blends

Table 17.1 Temperatures and Enthalpies of Fusion/Crystallization of PE Phase of MIX, COP, COP120, and Their Fractionsa. Cooling Tc , C Sample MIX 115.1 MIX toluene 114.1 soluble MIX xylene — insoluble COP 107.4 COP toluene 108.2 soluble COP xylene 105.0 insoluble COP120 111.2 COP120 toluene 113.1 soluble COP120 xylene 103.2 insoluble

DHc , J g1 87.3 186.6

Heating Tm , C 131.0 131.1

SBH, wt% SBH, wt% DHm , J g1 (from DSC) (from DTG) 90.3 45–55 53 186.1 — 0

—

—

—

—

100

82.4 167.9

126.5 128.6

81.0 172.4

50–57 5–10

49 0

29.4

120.1

26.7

70–75

75

73.8 180.9

128.1 128.1

73.5 177.2

— —

51 0

36.0

116.3

33.2

—

73

a

From References 14 and 15 with permission from John Wiley & Sons, Inc.

whereas those of the polycondensation product and reactive blends contain significant amounts of the other component, too. The contents of PE chains in the insoluble residues, calculated on the basis of the melting/crystallization enthalpies, are in fair agreement with those found by TG (Table 17.1) (14,15). The SBH content from TG measurements has been calculated by the weight losses corresponding to the degradation processes of both components, which take place at different temperature intervals. The SEM micrographs of MIX and COP60 are shown in Fig. 17.2, together with those of their insoluble residues. Both blends have a two-phase morphology. The SEM micrographs of the insoluble fractions of MIX and COP60 (Fig. 17.2c and d) reflect their different compositions. The morphology of the xylene-insoluble fraction of MIX appears homogeneous and fibrous, as expected for neat SBH, and that of COP60 xylene-insoluble fraction, on the contrary, shows a two-phase morphology. It could be assumed that the matrix is made of practically pure SBH and the minor phase consists of PE-g-SBH copolymer. Better understanding of the crystalline structure and morphology of the PEg-SBH copolymers is necessary to confirm their potential of compatibilizers. In cases of different polymer blends with crystallizable components, the X-ray diffraction method is reliable to find the degree of order of the LCP phase in the blend and its influence on the crystallinity of the matrix (16,17) as well to examine the crystallizability of the copolymer segments (18). The X-ray diffraction patterns of PE-g-SBH copolymers obtained via both procedures (19) consist of reflections typical for the orthorhombic crystalline lattice of PE and the single reflection of the solid LCP. The lack of dhkl variations with respect to those of neat PEox and SBH

Chapter 17 Compatibilization and Crystallization of Blends

505

Figure 17.2 SEM micrographs of MIX (a), COP60 (b), xylene-insoluble fraction of MIX (c), and xylene-insoluble fraction of COP60 (d). (From Reference 15 with permission from John Wiley & Sons, Inc.)

indicates the absence of interactions in the crystalline phase or that of cocrystallization phenomena between the components of the PE-g-SBH copolymers. The analysis of the crystallinity degree and normalized amorphous and crystalline contributions to the diffraction patterns of the products suggests that both copolymer components are partly miscible in the amorphous phase. The extent of miscibility depends on the copolymer structure, namely on the length of PE segments and SBH grafts. PE segments in PE-g-SBH copolymers obtained by the reactive blending are longer and exhibit a higher crystallizability than those obtained via melt polycondensation. SBH grafts of the copolymers obtained by reactive blending are also longer than those in the products obtained via melt polycondensation (19). To prove the compatibilizing efficiency of the ad hoc synthesized PE-g-SBH copolymers, the latter have been employed as compatibilizing additives for blends of HDPE with SBH (14,15,20). The rheological and mechanical properties, as well as the morphology of the compatibilized blends, have been compared with the uncompatibilized ones. A commercial sample of high density polyethylene grafted with maleic anhydride has also been used as compatibilizer for the same blends for comparison (20). In Fig. 17.3, the SEM micrograph of 76/8/16 HDPE/COP120/ SBH ternary blend is compared to that of the 80/20 HDPE/SBH binary blend. The improvement of both phase dispersion and interfacial adhesion brought about by the addition of COP120 is evident. Similar results have also been found by the addition of the PE-g-SBH copolymer prepared by polycondensation. The melt viscosity of the compatibilized blends is, in general, slightly higher than that of the corresponding uncompatibilized ones, due the increased interfacial

506

Polyolefin Blends

Figure 17.3 SEM micrographs of 80/20 HDPE/SBH (a) and 76/8/16 HDPE/COP120/SBH (b and c) blends. (From Reference 15 with permission from John Wiley & Sons, Inc.)

Chapter 17 Compatibilization and Crystallization of Blends

507

Figure 17.4 (a) Viscosity curves of the pure components: PE (HDPE), SBH, HDM (maleic anhydridegrafted PE), COPR (COP), and COPM (COP120). (b) Viscosity curves of the 80/20 PE/SBH blends with 5 wt% of compatibilizing agents. (From Reference 20 with permission from John Wiley & Sons, Inc.)

adhesion, especially in the low shear rate range (Fig. 17.4a and b) (20). Despite this, the viscosity of the blends always remains lower than that of the pure polyolefin (Fig. 17.4a and b). Thus, polyolefin processability can actually be enhanced by the addition of LCP, even in the presence of compatibilizers. However, the effect of the addition of COP120 or COP on the mechanical properties of the HDPE/SBH blend is fairly modest. The tensile modulus is slightly decreased, while the elongation at break is increased almost two times (20). It should be noted that the commercial compatibilizer does not improve the elongation at break. This behavior confirms that

508

Polyolefin Blends

whereas the compatibilizing role of the MA-grafted PE is very modest, the potential of the COP120 and COP is certainly higher. The obtained results may be rationalized considering that the PE-g-LCP copolymers used by us consist of fairly short PE backbones with LCP grafts of various length. The molecules with longer LCP branches are thought to become mixed at the surface of the LCP particles and to give rise to fairly weak interaction with the PE matrix. It is argued that new PE-g-LCP copolymers synthesized from functionalized PE samples of higher molar mass might show much better compatibilizing performance (20). The dependence of the components’ molar mass and of the mixing conditions on the compatibilizing efficiency of PE-g-SBH copolymers has also been studied (21). The results indicate that the PE-g-SBH copolymers do, in fact, compatibilize the HDPE/SBH blends and that the effect is more pronounced with the lower molar mass PE matrix and with the SBH sample having lower viscosity. Moreover, the compatibilizing ability of the graft copolymer is improved, if the latter is first blended with either of the two main components (21). An attempt has been made to synthesize copolymers with better compatibilizing efficiency (22) using functionalized PE of a higher molar mass. A novel graft copolymer (PE-g-LCP) consisting of PE backbones and LCP branches was synthesized via reactive blending of an acrylic acid-functionalized PE (Escor 5000 by Exxon) with a semiflexible LCP (SBH 1:1:2 by Eniricerche S.p.A.) (23). The investigations of the fractions soluble in boiling toluene and xylene and of the residue have shown that crude product (COP-AA) contains unreacted Escor and SBH, together with the graft copolymer forming the interphase. The morphology of the copolymers is not homogeneous and reveals two types of structures characteristic for both copolymer segments, namely tiny PE spherulites and small and large LCP domains (24). That means LCP does not lose its mesophase behavior when bonded to the PE backbone. The morphological observations also confirm that the distribution of SBH grafts along the PE backbone is random (24). The two segments in the copolymers form separate crystalline phases typical for the neat components. Similar to PE-g-SBH copolymers obtained previously, there are no interactions in the crystal phase or cocrystallization between the copolymer segments. The addition of relatively small amounts of COP-AA into binary HDPE/SBH blends was shown to cause a considerable reduction of the dimensions of the SBH particles and an improvement of the adhesion between the phases (23). A slight enhancement of interfacial adhesion was also shown to take place in the molten state as well. The compatibilizing efficiency of this product is considerably higher than that of pure Escor and, supposedly, of other commercially available functionalized polyolefins. With respect to the latter compatibilizers, COP-AA has apparently the advantage of a strong affinity to the SBH domains of the blends. The results also confirm previous findings indicating that optimum compatibility between the PE segments of the PE-g-SBH copolymer and the PE matrix is realized when the latter has an appropriate molar mass. Melt spinning tests demonstrated that deformation of the SBH droplets into highly oriented fibrils can be obtained for the blends of lower

Chapter 17 Compatibilization and Crystallization of Blends

509

Figure 17.5 SEM micrograph of the fracture surface of the fiber of the 78/4/18 HDPE/COP-AA/SBH blend. (From Reference 23 with permission from John Wiley & Sons, Inc.)

molar mass PE, compatibilized with small amounts of the novel PE-g-SBH copolymer (23) (Fig. 17.5). The assumption has been made that the compatibilization activity of these copolymers toward HDPE/SBH blends is due not only to the identical chemical structure of the copolymer segments and the corresponding blend components but also to the similarity of their crystalline structure, crystallization behavior, and morphology.

17.1.3 Blends of Isotactic Polypropylene with LCP Reinforcement of PP with several LCPs has been attempted by Baird and coworkers (3,4) and by others (5,25,26). Several investigations have dealt with the preparation and characterization of PP/LCP blends containing different commercially available compatibilizers, such as maleic anhydride-grafted PP (PP-g-MA) (27–29) or an ethylene-based reactive terpolymer (30). The latter additives have been shown to improve the phase dispersion and the interfacial adhesion, but the enhancement of tensile modulus, tensile strength, and surface finish was generally modest (27–29). Miller and coworkers (31–33) have used an acrylic acid-functionalized PP (PPAA) and a PPAA-based graft copolymer for the compatibilization of PP/LCP blends and have observed an improvement of the tensile properties, thermal stability, and crystallinity of the fibers produced therefrom. An ethylene–glycidyl methacrylate copolymer (EGMA) (34) and a terpolymer of ethylene, ethylacrylate, and glycidyl methacrylate (EEAGMA) (35) have also been investigated as reactive compatibilizers for PP/LCP blends. The reaction between the epoxy groups of the compatibilizers and the LCP end groups has been found to lower the dimensions of the LCP domains and to improve the impact strength (34,35). However, this positive effect has been shown to be accompanied by a substantial reduction of the PP degree

510

Polyolefin Blends

of crystallinity and of the tensile modulus (34). It is well known that, among compatibilizing agents, block or graft copolymers made up of segments whose chemical structure and solubility parameters are similar to those of the polymers being blended appear best suited to the scope (36). These compatibilizers can migrate to the interphase and reduce the interfacial energy between matrix and dispersed phase, thus causing a reduction of the minor phase dimensions and a stabilization of polymer blend morphology. New graft copolymers consisting of PP backbones and LCP branches, to be used as compatibilizing agents for PP/LCP blends, have been synthesized (37). The PP-g-LCP copolymers have been prepared by polycondensation of the monomers of a semiflexible liquid crystalline polyester (SBH 1:1:2), carried out in the presence of appropriate amounts of a commercial acrylic acid-functionalized polypropylene. The polycondensation products, referred to as COPP50 and COPP70, having calculated PPAA concentrations of 50 and 70 wt%, respectively, have been fractionated with boiling toluene and xylene, and the soluble and insoluble fractions have been characterized by Fourier transform infrared and nuclear magnetic resonance spectroscopy, scanning electron microscopy, differential scanning calorimetry, and X-ray diffraction. All analytical characterizations have concordantly shown that the products are formed by intricate mixtures of unreacted PPAA and SBH together with PP-g-SBH copolymers of different composition. An example for the analytical characterization of the samples has been demonstrated by the solid-state 13C NMR spectra (Fig. 17.6) and the average values of proton spin-lattice relaxation time of the materials (Table 17.2) (37). The observed behavior indicates that the grafting reaction has actually occurred between the two components and that these components are bonded chemically in the copolymers and form separated domains with linear dimensions larger than several hundred angstroms (37). Exploratory experiments carried out by adding small amounts of COPP50 or COPP70 into binary mixtures of isotactic polypropylene (iPP) and SBH while blending have demonstrated an appreciable improvement of the dispersion of the minor LCP phase (Fig. 17.7), as well as an increase in the crystallization rate of iPP. The uncompatibilized blend is characterized by a dispersed SBH phase appearing as large (20–70 mm) unequal droplets whose smooth surface denounces the complete incompatibility of the two phases. As it is clearly demonstrated, the addition of either COPP always leads to a significantly finer dispersion of the LCP droplets (5–7 mm). COPP50 seems to develop its maximum efficiency at concentration of about 5%, whereas higher concentrations are needed for COPP70 (37). The calorimetric characteristics of iPP phase of the uncompatibilized blends show that the presence of the SBH dispersed phase leads to a slight increase of the PP temperature of crystallization (Fig. 17.8) (37,38). This result can be interpreted by a slightly increased nucleation rate of PP phase in the presence of SBH dispersed particles. As seen in Fig. 17.8, iPP temperature of crystallization increases drastically in the presence of COPP70 and COPP50. The analysis of the X-ray patterns (38) (Fig. 17.9a and b) and calculated parameters (dhkl values, degree of crystallinity, crystallite size, and intensity ratio) allows the assumption that PP segments of

Chapter 17 Compatibilization and Crystallization of Blends

Figure 17.6

511

13

C NMR spectra of mechanical blends MIXP70 and MIXP50, the polycondensation products COPP70 and COPP50, and the residue of COPP50 (RXCP50). (From Reference 37 with permission from John Wiley & Sons, Inc.)

PP-g-SBH copolymers cocrystallize completely with bulk iPP, which increases PP crystallinity degree; part of the SBH component (SBH grafts of PP-g-SBH copolymers and bulk SBH) enters the mutual amorphous phase of the blends, leading to a decrease of the intensity of the SBH peak. That means each part of the compatibilizer is miscible with the corresponding bulk phase of the blend. The compatibilized iPP/LCP blends display improved crystallization kinetics, enhanced degree of crystallinity, and improved interphase adhesion (37,38). Consequently, an improvement of the mechanical characteristics should be expected for these blends. In fact, the investigation of the Vickers microhardness of uncompatibilized and

512

Polyolefin Blends Table 17.2 Average Values of Proton Spin-Lattice Relaxation Time for the PPAA and SBH Components of Different Materialsa. Sample

TH1 (s) (SBH carbons)

PPAA SBH MIXP70 MIXP50 COPP70 COPP50 RXCP50

— 0.75 1.15 1.15 0.71 0.31 0.30

a

TH1 (s) (PPAA carbons) 0.9 — 0.95 0.95 0.92 0.89 0.79

From Reference 37 with permission from John Wiley & Sons, Inc.

compatibilized iPP/SBH blends shows that the addition of 2.5, 5, or 10 wt% of PP-gSBH compatibilizer leads to an increase in the microhardness of the blends (39). The strong positive deviation of the experimental hardness values from the additive ones has been interpreted by the increase in the degree of crystallinity of PP crystals and by the decrease in the surface free energy of PP crystals and of SBH LC domains (Table 17.3). Moreover, it is well known that there is a correlation between the microhardness (H) and the modulus of elasticity (E) (40). On the grounds of the mechanism of hardness indentation, namely that the loading cycle is elastic–plastic and the unloading cycle is elastic, a plot has been derived (41), which is an indicator

Figure 17.7 SEM micrographs of the fracture surfaces of (a) 80/20 iPP/SBH blend, (b) the same with 2.5% COPP50, (c) the same with 5% COPP50, and (d) the same with 10% COPP50. (From Reference 37 with permission from John Wiley & Sons, Inc.)

513

120

0.65

115

0.60

110

0.55

105

Crystallinity

o

Tc, C

Chapter 17 Compatibilization and Crystallization of Blends

0.50 COPP70 COPP50

100

0.45 0

2

4 6 Compatibilizer, wt %

8

10

Figure 17.8 Dependence of iPP crystallization temperature (Tc ) and degree of crystallinity on the content of the two compatibilizers in iPP/SBH 80/20 w/w blends. (From Reference 38 with permission from John Wiley & Sons, Inc.)

of the position taken by different materials in the elastic–plastic spectrum. According to the H/E values obtained for the blends, the compatibilized blend iPP/COPP50/SBH 77.5/5/17.5 (H ¼ 94 MPa, E ¼ 950 MPa, H=E ¼ 0:100) is positioned in the elastic– plastic spectrum much closer to elastic material than the uncompatibilized blend iPP/ SBH 80/20 (H ¼ 78 MPa, E ¼ 890 MPa, H=E ¼ 0:087). This is in perfect agreement with the values of elongation at break, which for the uncompatibilized blend is 9%, while for the compatibilized blend this value is 15%. These results confirm the compatibilizing efficiency of COPP, leading to enhanced adhesion between the blend components.

17.2 CRYSTALLIZATION BEHAVIOR OF BLENDS OF POLYOLEFINS WITH A SEMIFLEXIBLE LIQUID CRYSTALLINE POLYMER 17.2.1 Isothermal and Nonisothermal Crystallization of Blends of Linear Low Density Polyethylene with a Semiflexible Liquid Crystalline Polymer The LCP dispersed phase may sometimes act as a nucleating agent, thereby enhancing the rate and/or the extent of crystallization of crystallizable matrices such as

514

Polyolefin Blends

Figure 17.9 (a) X-ray diffraction patterns of uncompatibilized iPP/SBH 80/20 w/w blend and of the blends compatibilized with 2.5, 5, and 10 wt% COPP50 (data from Reference 38). (b) Fitted profile of iPP/SBH 80/20 w/w X-ray pattern. (From Reference 38 with permission from John Wiley & Sons, Inc.)

Chapter 17 Compatibilization and Crystallization of Blends

515

Table 17.3 The Experimental Microhardness Values (Hexp ), the Calculated Microhardness Values According to the Additivity Law (Hcal ), PP Degree of Crystallinity Derived from WAXS (a), the Experimental Values of Melting Temperatures (Tm ), the Values of b Parameter, and the Values of the Surface Free Energy of Neat PP and PP Component in Uncompatibilized and Compatibilized Blends. PP-g-SBH copolymer

COPP70 COPP70 COPP50 COPP70 COPP70 COPP70 COPP50 COPP50 COPP50 a

Composition of blends iPP/COPP/SBH 100/0/0 0/0/100 90/0/10 88.2/2.5/9.2 86.5/5/8.5 87.5/5/7.5 80/0/20 78.2/2.5/19.2 76.5/5/18.5 73/10/17 78.7/2.5/18.7 77.5/5/17.5 75/10/15

Hcal , Hexp , MPa MPa 89 — 85 87 88 89 76 — — — 79 82 82

89 27 83 92 96 95 78 89 92 92 88 94 89

$

a

Tm , K

b , s PP , ˚ erg cm2 A

0.50 — 0.52 0.54 0.55 0.56 0.49 — — — 0.53 0.55 0.55

431.0 — 430.6 433.4 433.4 434.3 430.8 431.8 432.8 431.2 433.3 432.2 432.8

8.1 — 8.1 7.5 7.5 7.1 8.1 7.8 7.6 8.0 7.5 7.7 7.6

79.6 — 79.6 73.5 73.5 70.0 79.6 77.2 74.7 78.4 73.5 75.9 74.7

From Reference 39 with permission from Springer.

poly(ethylene terephthalate) (42–45), poly(butylene terephthalate) (46, 47), aliphatic polyamides (48), poly(phenylene sulfide) (49,50), and so on. In other systems, a depression of the crystallization rate induced by the added LCP was found (51). These effects depend critically on the interactions between the blend components in the molten state. The crystallization kinetics of LLDPE/SBH blends has been investigated under nonisothermal and isothermal conditions, and the results have been discussed with reference to the blends morphology (52). A very good dispersion of minute SBH droplets has been observed for LCP concentrations up to 10%, whereas for the 80/20 LLDPE/SBH blend, extensive droplet coalescence with formation of large LCP domains has been found to occur. It has been demonstrated that the morphology of the blends influences their crystallization behavior. In fact, both nonisothermal and isothermal studies have shown that the dispersed LCP phase plays a nucleating role whose efficiency is maximum for the 90/10 blend. Apparently, the nucleating effect depends on the interfacial surface available. The latter increases with the SBH concentration in the 0–10% range, and then decreases when the droplet coalescence prevails. The nonisothermal crystallization rate has been evaluated by the crystallization rate coefficient (CRC) proposed by Khanna (53). If the cooling rate is plotted against the crystallization temperature, the slopes of the straight lines drawn through the experimental points can be taken as CRCs (Fig. 17.10) (52). While the CRC value of neat LLDPE is 87 h1 , the CRCs of the blends are all about 110 h1 , independent of the LCP concentration. The overall nonisothermal crystallization kinetics has been studied by Harnisch and Muschik’s method (54). The Avrami

516

Polyolefin Blends 25

LLDPE 95/5 90/10 80/20

15

o

Cooling rate, C min–1

20

10

5

0 80

85

90

95 100 105 110 115 120 125 130 o

T cr, C Figure 17.10 Dependence of the cooling rate on the crystallization temperature of LLDPE/SBH blends. (From Reference 52 with permission from Springer.)

exponents n have been determined according to the following equation, which is valid at T ¼ Tc : h i y1 y2 ln 1x2 ln 1 x1 n¼1þ ð17:1Þ f2 ln f1 where xi is the crystalline fraction calculated by integration of the DSC endotherm, yi is the derivative of xi , and fi is the cooling rate. The values of the Avrami exponents (n ¼ 2:5–2.6) suggest a spherulitic three-dimensional growth process controlled by heterogeneous nucleation (52). The isothermal crystallization kinetics has been analyzed by means of the Avrami equation: Xt ¼ 1 expðKn tn Þ

ð17:2Þ

Chapter 17 Compatibilization and Crystallization of Blends

517

where Xt is the fractional crystallization occurred at time t, Kn is the kinetic constant, and n is the Avrami exponent, which depends on the type of nucleation and on the crystal growth geometry. The Avrami exponents of the samples are close to 3, which is an indication of heterogeneous nucleation with threedimensional growth, and drop to 1–2 during secondary crystallization (52). The values of the Avrami exponents found from isothermal measurements are slightly higher than those determined under nonisothermal conditions. This has been attributed to the formation of sheaflike superstructures not fully developed into spherulites, and to spherulites impingement effects (55) during nonisothermal crystallization. The interpretation of the nonisothermal and isothermal data in terms of the Avrami exponents indicated that the nucleation mechanism is not altered by the presence of a dispersed SBH phase. The values of the kinetic constants for the blends especially that with 10% SBH are appreciably higher than that measured for neat LLDPE (52). The nucleation effect of SBH dispersed phase has been confirmed by the decrease of LLDPE spherulite’s size in the blends by 50% (52). The average dimensions of the spherulites were determined from Hv patterns of small-angle light scattering by means of the equation (56) Rsph ¼

2:05l np sin Qm

ð17:3Þ

where Rsph is the spherulite radius, l=n is the wavelength of the light in a medium of refractive index n, and Qm is the angle of the incident and scattered beams corresponding to the maximum pattern intensity (52). The determination of the thermodynamic equilibrium melting point (Tm ) of the crystallizable component of the blends by means of the dependence of melting temperatures of isothermally crystallized samples (Tm ) on the temperatures of crystallization (Tc ) has shown that Tm ¼ 130 C for neat LLDPE and LLDPE/SBH blends (52).

17.2.2 Crystallization Behavior of PE-g-LCP Copolymers The valuable properties of block and graft copolymer compatibilizers, containing crystalline components, come from their ability to undergo microphase separation (57, 58). There is no simple relation between the crystallization characteristics of the copolymer components and their molecular structure parameters. But, it is well known that in the block or graft copolymers the transition temperatures and enthalpies of the crystallizable polymer segments decrease in comparison to those of the corresponding homopolymer (59). The nucleation and crystal growth of the corresponding phases can be hindered since the components of the copolymers are covalently bonded (60–62). The crystallization kinetics of PE-g-SBH copolymers obtained by polycondensation (COP) or reactive blending (COP120) has been investigated under

518

Polyolefin Blends 1

2

2 1

2

1

1

0

ln[y/(1– x)]

–1

MIX

COP120

COP

–2

–3

–4

–5 108

110

112

114

116

118

120

122

124

o

Temperature, C Figure 17.11 The Harnisch plots of MIX, COP120, and COP. Curve 1—2.5 C min1 cooling rate; curve 2—5 C min1 cooling rate. (From Reference 63 with permission from Springer.)

nonisothermal conditions (63). The crystallization temperature (Tc ) of the PE component of COP and COP120 decreases steadily upon increasing the concentration of the SBH grafts (63). It was found that COP120 crystallizes at slightly higher Tc than COP and at a higher rate, confirmed by the determination of the crystallization rate coefficients (63). The results have been interpreted by the fact that the PE crystallizable segments and SBH grafts of COP120 are longer than those of COP. The overall nonisothermal crystallization kinetics has been studied by the Harnisch and Muschik equation (54). The Harnisch plots of MIX, COP120, and COP are presented in Fig. 17.11. Values of n ¼ 3:3–3.4 were obtained for the PE component of MIX, COP120, and COP (63). These results show that the mechanism of the crystallization of PE phase does not change; it changes (n ¼ 1:5–1.6) only when the SBH content overruns about 50%, due to the decrease of both nucleation and crystal growth rates. The decrease of the crystal growth rate of the copolymers in comparison to that of the corresponding homopolymers could be explained by the lower mobility of the crystallizable segments of the copolymers (64). The results have shown that the crystallization behavior and morphology of COP120 are more similar to those of the neat components of PE and SBH. This explains the fact that COP120 is more effective as a compatibilizer for PE/SBH blends than COP.

Chapter 17 Compatibilization and Crystallization of Blends

519

17.2.3 Effect of PP-g-LCP Compatibilizer on the Morphology and Crystallization of PP/LCP Blends Concerning compatibilized blends, the interfacial behavior of the compatibilizer has an important effect upon the crystallization of the blended components as it was shown for crystalline/crystalline polymer blends (60,65–67) and for crystalline polymer/LCP blends (32,37,38,68). For the latter blends, the enhanced phase interactions and improved interfacial adhesion could increase the abovementioned nucleation activity of the LCP toward the crystallizable matrices. In the particular case of using polyolefin-g-LCP copolymer compatibilizer, the crystallization of the two blend phases might have a reverse effect upon the compatibilizing activity. Moreover, the miscibility (69,70) and/or cocrystallization (60) between the bulk homopolymers and corresponding segments of the copolymer could strongly influence the crystallization behavior of the blends. The nonisothermal crystallization behavior and morphology of iPP/SBH blends compatibilized with 2.5, 5, or 10 wt% PP-g-SBH copolymers have been investigated by differential scanning calorimetry and optical microscopy (38). The migration of the compatibilizer to the interphase leads to a reduction of the interfacial tension and to a strong reduction of the dimensions of dispersed SBH phase. This leads to an increase of the iPP nucleation rate (or crystallization rate) (Fig. 17.8) and to a decrease of the iPP spherulite dimensions (Fig. 17.12), but the iPP crystal growth mechanism remains unchanged (38). The effect of the compatibilization leads to an increase of the iPP degree of crystallinity (Fig 17.8).

17.2.4 Isothermal Crystallization Kinetics of Compatibilized Blends of Polyolefins with a Semiflexible LCP The crystallization behavior and kinetics under isothermal conditions of iPP/SBH and HDPE/SBH blends, compatibilized with PP-g-SBH and PE-g-SBH copolymers, respectively, have been investigated (71). It has been established that the LCP dispersed phase in the blends plays a nucleation role for the polyolefin matrix crystallization. This effect is more pronounced in the polypropylene matrix than in the polyethylene matrix, due to the lower crystallization rate of the former. The addition of PP-g-SBH copolymers (2.5–10 wt%) to 90/10 and 80/20 iPP/SBH blends provokes a drastic increase of the overall crystallization rate of the iPP matrix and of the degree of crystallinity. Table 17.4 collects the isothermal crystallization parameters for uncompatibilized and compatibilized iPP/SBH blends (71). On the contrary, the addition of PE-g-SBH copolymers (COP or COP120) (2.5–8 wt%) to 80/20 HDPE/SBH blends almost does not change or only slightly decreases the PE overall crystallization rate (71). This is due to some difference in the compatibilization mechanism and efficiency of both types of graft copolymers (PP-g-SBH and PEg-SBH). The two polyolefin-g-SBH copolymers migrate to blend interfaces and

520

Polyolefin Blends

Figure 17.12 Optical micrographs, taken at room temperature after cooling of the samples, of iPP/ SBH 90/10 w/w blend (a) and of compatibilized blends with 5 wt% COPP50 (b) or with 5 wt% COPP70 (c). Magnification 200. (From Reference 38 with permission from John Wiley & Sons, Inc.)

Chapter 17 Compatibilization and Crystallization of Blends

521

Table 17.4 Isothermal Crystallization Parameters for Uncompatibilized and Compatibilized iPP/SBH Blends. PP-g-SBH copolymer

Composition of blends iPP/COPP/SBH

Tc ( C)

t0:5 (s)

n

100/0/0

127 124 121 130 127 124 130 127 124 130 127 124 130 127 124 130 127 124 130 127 124 130 127 124 130 127 124 130 127 124 130 127 124 130 127 124

574 293 153 583 245 144 91 46 29 57 47 23 93 51 32 190 118 56 78 61 35 82 46 26 74 39 26 108 53 30 114 65 32 86 47 28

2.1 2.1 2.1 2.4 2.3 2.1 2.2 2.2 2.1 2.1 2.1 2.0 2.5 2.4 2.1 2.3 2.3 2.2 2.2 2.2 2.0 2.2 2.0 2.1 2.2 2.1 2.0 2.3 2.4 2.3 2.1 2.1 2.1 2.4 2.4 2.1

90/0/10

COPP70

88.2/2.5/9.2

COPP70

86.5/5/8.5

COPP50

87.5/5/7.5

80/0/20

COPP70

78.2/2.5/19.2

COPP70

76.5/5/18.5

COPP70

73/10/17

COPP50

78.7/2.5/18.7

COPP50

77.5/5/17.5

COPP50

75/10/15

a

Kn (sn ) 1:2 106 4:0 106 1:1 105 1:2 107 3:1 106 2:2 105 3:0 105 1:6 104 7:2 104 1:6 104 3:2 104 1:4 103 9:7 106 5:4 105 4:7 104 5:5 106 1:1 105 1:1 104 8:0 105 2:0 104 5:4 104 4:7 105 2:9 104 7:6 104 6:3 105 3:5 104 1:2 103 1:2 105 4:3 105 2:5 104 3:3 105 1:1 104 6:8 104 1:4 105 7:8 105 6:6 104

From Reference 71 with permission from John Wiley & Sons, Inc.

miscibility between SBH grafts and bulk SBH is realized. But while there is a complete cocrystallization between PP segments of PP-g-SBH copolymers and bulk iPP, only some interactions occur between PE segments of PE-g-SBH copolymers and bulk HDPE (71).

522

Polyolefin Blends

17.2.5 Crystallization and Morphology of Fibers Prepared from Compatibilized Blends of Polyethylene with a Liquid Crystalline Polymer The previous results have shown that the addition of PE-g-SBH copolymer compatibilizer to HDPE/SBH blends in a nonoriented state does not lead to an increase of the crystallization rate (71). However, the overall PE crystallization rate under

Figure 17.13 Optical micrographs of the samples: (a, b) uncompatibilized fiber of the blend high molar mass HDPE/SBH; (c, d) compatibilized with 4% Escor; (e–h) compatibilized with 4% COP-AA. Parallel polaroids (a, c, e, g) and crossed polaroids (b, d, f, h). The width of the micrographs corresponds to 150 mm. (From Reference 72 with permission from John Wiley & Sons, Inc.)

Chapter 17 Compatibilization and Crystallization of Blends

523

nonisothermal and isothermal conditions increases for fibers prepared by melt spinning of compatibilized HDPE/SBH blends (72). The dispersed LCP phase in the latter fibers appears as fibrils with high aspect ratio due to the enhanced interfacial adhesion in the presence of the COP-AA (Figs. 17.5 and 17.13). As seen (Fig. 17.13e–h) the LCP domains in the blend fiber compatibilized with 4% COP-AA (draw ratio 28) appear as long SBH fibrils with high aspect ratio (10–50). The results have been interpreted by an increase in the interfacial area between the SBH fibrils and the matrix phase, which enlarges the number of heterogeneous nuclei and increases the PE nucleation rate. The results from isothermal crystallization experiments of lower molar mass HDPE/SBH blends and the analyses of the kinetic parameters by Avrami equation have shown that the dispersed LCP fibrils have a nucleating role in the presence of the compatibilizers in good agreement with nonisothermal crystallization data (72).

17.3

CONCLUSIONS

The investigation of the compatibilization and crystallization of blends of polyolefins with a semiflexible LCP leads to the following conclusions: the compatibilization of polyolefin/LCP blends has been realized successfully by the addition of ad hoc synthesized polyolefin-g-LCP copolymers. The compatibilization results into materials, characterized by a stabilized morphology, improved crystallization kinetics under nonisothermal and isothermal conditions, and enhanced mechanical properties. Moreover, polyolefin processability has been enhanced by the addition of LCP, even in the presence of compatibilizers. New high quality materials with improved processability have been produced by technologies, which are economic, friendly to the environment, and socially acceptable.

ACKNOWLEDGMENT The author is grateful to Professor P.L. Magagnini from University of Pisa, Italy, for the years of fruitful collaboration in the field of the herein presented studies.

NOMENCLATURE COP COP15, COP60, COP120 COP-AA COPP50, COPP70

PE-g-SBH copolymer prepared by polycondensation of SBH monomers in the presence of PEox PE-g-SBH copolymers prepared by reactive blending of PEox with SBH PE-g-SBH copolymer prepared by reactive blending of an acrylic acid-functionalized PE with SBH PP-g-SBH copolymers prepared by polycondensation of the monomers of an SBH, carried out in the presence of an acrylic acid-functionalized PP

524

Polyolefin Blends

CRC DSC DTG E fi FTIR H HDPE Kn LCP LLDPE l=n MIX n NMR PE-g-LCP PEox iPP PP-g-LCP Qm Rsph SBH

SEM t Tc TG Tm Tm xi yi Xt

Crystallization rate coefficient Differential scanning calorimetry Differential thermogravimetry Modulus of elasticity Cooling rate Fourier transform infrared spectroscopy Microhardness High density polyethylene Kinetic constant Liquid crystalline polymer Linear low density polyethylene Wavelength of the light in a medium of refractive index n Physical blend between PEox and SBH Avrami exponent Nuclear magnetic resonance spectroscopy Graft copolymer between polyethylene and LCP Oxidized low molar mass PE sample containing free carboxylic groups Isotactic polypropylene Graft copolymer between polypropylene and LCP Angle of the incident and scattered beams corresponding to the maximum pattern intensity Spherulite radius Semiflexible liquid crystalline polymer, synthesized from sebacic acid (S), 4,40 -dihydroxybiphenyl (B), and 4-hydroxybenzoic acid (H) in the mole ratio 1:1:2 Scanning electron microscopy Time Temperature of crystallization Thermogravimetry Melting temperature Thermodynamic equilibrium melting temperature Crystalline fraction Derivative of xi Fractional crystallization

REFERENCES 1. F. P. La Mantia (ed.), Thermotropic Liquid Crystal Polymer Blends, Technomic, Lancaster, 1993. 2. G. Kiss, Polym. Eng. Sci., 27, 410, (1987). 3. D. Done, A. M. Sukhadia, A. Datta, and D. G. Baird, SPE Technol. Paper, 48, 1857 (1990). 4. A. Datta, A. M. Sukhadia, J. P. Desouza, and D. G. Baird, SPE Technol. Paper, 49, 913 (1991).

Chapter 17 Compatibilization and Crystallization of Blends

525

5. Y. Yongcheng, F. P. La Mantia, A. Valenza, V. Citta, U. Pedretti, and A. Roggero, Eur. Polym. J., 27, 723 (1991). 6. T. C. Hsu, A. M. Lichkus, and I. R. Harrison, Polym. Eng. Sci., 33, 860 (1993). 7. T. Harada, K. Tomari, A. Hamamoto, S. Tonogai, K. Sakaura, S. Nagai, and K. Yamaoka, SPE Antec92, 376 (1992). 8. A. M. Lichkus and I. R. Harrison, SPE Antec-92, 2257 (1992). 9. T. C. Hsu and I. R. Harrison, SPE-93, 1183 (1993). 10. U. Pedretti, A. Roggero, F.P, La Mantia, and P. L. Magagnini, SPE Antec-93, 1706 (1993). 11. E. G. Fernandes, I. Giolito, and E. Chilellini, Thermochim. Acta, 235, 67 (1994). 12. F. P. La Mantia, C. Geraci, M. Vinci, U. Pedretti, A. Roggero, L. I. Minkova, and P. L. Magagnini, J. Appl. Polym. Sci., 58, 911 (1995) 13. L. Minkova and P. L. Magagnini, Colloid Polym. Sci., 274, 34 (1996). 14. P. L. Magagnini, M. Paci, L. I. Minkova, T. Miteva, D. Sek, J. Grobelny, and B. Kaczmarczyk, J. Appl. Polym. Sci., 60, 1665 (1996). 15. L. I. Minkova, T. Miteva, D. Sek, B. Kaczmarczyk, P. L. Magagnini, M. Paci, F. P. La Mantia, and R. Scaffaro, J. Appl. Polym. Sci., 62, 1613 (1996). 16. C. U. Ko, G. L. Wilkes, and C. P. Wong, J. Appl. Polym. Sci., 37, 3063 (1989). 17. M. Pracella, D. Dainelli, G. C. Galli, and E. Chiellini, Makromol. Chem., 187, 2387 (1986). 18. B. Wunderlich (ed.), Macromolecular Physics, Vol. 1, Academic Press, New York, 1973. 19. Ts. Miteva and L. Minkova, Colloid Polym. Sci., 275, 38 (1997). 20. F. P. La Mantia, R. Scaffaro, P. L. Magagnini, M. Paci, C. Chiezzi, D. Sek, L. I. Minkova, and Ts. Miteva, Polym. Eng. Sci., 37, 1164 (1997). 21. F. P. La Mantia, R. Scaffaro, P. L. Magagnini, M. Paci, L. I. Minkova, and Ts. Miteva, J. Appl. Polym. Sci., 71, 603 (1999). 22. Y. Lyatskaya, D. Gersappe, N. A. Gross, and A. Balazs, J. Phys. Chem., 100, 1449 (1996). 23. L. I. Minkova, M. Velcheva, M. Paci, P. L. Magagnini, F. P. La Mantia, and D. Sek, J. Appl. Polym. Sci., 73, 2069 (1999). 24. L. Minkova and P. L. Magagnini, Macromol. Chem. Phys., 200, 2551 (1999). 25. C. Federic, G. Attalla, and L. Chapoy, Eur. Patent 0,340,655 A2, (1989). 26. S. C. Tjong, S. L. Liu, and R. K. Y. Li , J. Mater. Sci., 31, 479 (1996). 27. A. Datta, H. H. Chen, and D. G. Baird, Polymer, 34, 759 (1993). 28. A. Datta and D. G. Baird, Polymer, 36, 505 (1995). 29. H. J. O’Donnel and D. G. Baird, Polymer, 36, 3113 (1995). 30. M. T. Heino and V. Seppala, J. Appl. Polym. Sci., 8, 1677 (1993). 31. M. M. Miller, D. L. Brydon, J. M. G. Cowie, and R. Mather, Macromol. Rapid Commun., 15, 857 (1994). 32. M. M. Miller, J. M. G. Cowie, J. G. Tait, D. L. Brydon, and R. Mather, Polymer, 36, 3107 (1995). 33. Y. Qin, M. M. Miller, D. L. Brydon, J. M. G. Cowie, R. R. Mather, and H. Wardman, in: Liquid Crystalline Polymer Systems, Technological Advances, A. I. Isayev, T. Kyu, S. Z. D. Cheng (eds.), American Chemical Society, Washington, DC, 1996, p. 98. 34. Y. P. Chiou, K. C. Chiou, and F. C. Chang, Polymer, 37, 4099 (1996). 35. R. M. Holsti-Miettinen, M. T. Heino, and J. V. Seppala, J. Appl. Polym. Sci., 57, 573 (1995). 36. X. Jin and W. Li, J. Macromol. Sci. Rev. Macromol. Chem. Phys., C35, 1 (1995). 37. P. L. Magagnini, M. Pracella, L. I. Minkova, Ts. Miteva, D. Sek, J. Grobelny, F. P. La Mantia, and R. Scaffaro, J. Appl. Polym. Sci., 69, 391 (1998). 38. Ts. Miteva and L. Minkova, Macromol. Chem. Phys., 199, 597 (1998). 39. L. Minkova, H. Yordanov, G. Zamfirova, and P. L. Magagnini, Colloid Polym. Sci., 280, 358 (2002).

526

Polyolefin Blends

40. F. J. Balta Calleja, Adv. Polym. Sci., 66, 117 (1985). 41. B. Lawn and V. R. Howes, J. Mater. Sci., 16, 2745 (1981). 42. E. G. Joseph, G. L. Wilkes, and D. G. Baird, Am. Chem. Soc. Div. Polym. Chem. Polym. Prepr., 24, 304 (1983). 43. E. G. Joseph, G. L. Wilkes, and D. G. Baird, in: Polymer Liquid Crystals, A. Blumstein (ed.), Plenum Press, New York, 1984. 44. S. K. Battacharya, A. Tendolkar, and A. Misra, Mol. Cryst. Liq. Cryst., 153, 501 (1987). 45. S. K. Sharma, A. Tendolkar, and A. Misra, Mol. Cryst. Liq. Cryst., 157, 597 (1988). 46. M. Kimura, R. S. Porter, and G. Salee, J. Polym. Sci. Polym. Phys. Ed., 21, 376 (1983). 47. M. Paci, C. Barone, and P. L. Magagnini, J. Polym. Sci. Polym. Phys. Ed., 25, 1595 (1987). 48. M. Takayanagi, T. Ogata, M. Morikawa, and T. Kai, J. Macromol. Sci. Phys., B27, 591 (1980). 49. L. I. Minkova, M. Paci, M. Pracella, and P. L. Magagnini, Polym. Eng. Sci., 32, 57 (1992). 50. A. Valenza, F. P. La Mantia, L. I. Minkova, S. De Petris, M. Paci, and P. L. Magagnini, J. Appl. Polym. Sci., 52, 1653 (1994). 51. M. Pracella, E. Chiellini, and D. Dainelli, Makromol. Chem., 190, 175 (1989). 52. L. Minkova and P. L. Magagnini, Colloid Polym. Sci., 274, 34 (1996). 53. Y. P. Khanna, Polym. Eng. Sci., 30, 1615 (1990). 54. K. Harnisch and H. Muschik, Colloid Polym. Sci., 261, 908 (1983). 55. L. C. Lopez and G. L. Wilkes Polymer, 30, 882 (1989). 56. P. H. Geil, Polymer Single Crystals, Wiley, New York, 1968. 57. J. A. Manson and L. H. Sperling, Polymer Blends and Composites, Plenum Press, New York, 1976. 58. A. Noshay and J. E. McGrath, Block Copolymers—Overview and Critical Survey, Academic Press, New York, 1977. 59. B. F. Mathot and B. F. Vincent, Calorimetry and Thermal Analysis of Polymers, Hanser Publishers, Munich, 1994. 60. T. Tang and B. Huang, J. Polym. Sci., B 32, 1991 (1994). 61. S. Datta and D. Lohse, Macromolecules, 26, 2064 (1993). 62. P. Jannasch and B. Wesslen, J. Polym. Sci., A33, 1465 (1995). 63. L. Minkova and Ts. Miteva, P. L. Magagnini, Colloid Polym. Sci., 275, 520 (1997). 64. B. Wunderlich, Macromolecular Physics, Vol. 2, Mir, Moscow, 1979, p. 344. 65. V. Flaris, A. Wasiak, and W. Wenig, J. Mater. Sci., 28, 1685 (1993). 66. U. Plawky and W. Wenig, J. Mater. Sci. Lett., 13, 863 (1994). 67. T. Tang, H. Li, and B. Nuang, Macromol. Chem. Phys., 195, 2931 (1994). 68. G. Poli, M. Paci, P. L. Magagnini, R. Scaffaro, and F. P. La Mantia, Polym. Eng. Sci., 36, 1244 (1996). 69. J Huang and H. Marand, Macromolecules, 30, 1069 (1997). 70. L. Z. Liu, W. Xu, H. Li, F. Su, and E. Zhou, Macromolecules, 30, 1363 (1997). 71. Ts. Miteva, L. Minkova, and P. Magagnini, Macromol. Chem. Phys., 199, 1519 (1998). 72. L. Minkova, M. Velcheva, and P. Magagnini, Macromol. Mater. Eng., 280/281, 7 (2000).

Chapter

18

Functionalized Polyolefins and Aliphatic Polyamide Blends: Interphase Interactions, Rheology, and High Elastic Properties of Melts Boleslaw Jurkowski1 and Stepan S. Pesetskii2

18.1 INTRODUCTION Blends of thermoplastic polymers combine best characteristics of their components in a single material, eliminate demerits, and often possess a set of properties unattainable for homopolymers while the range of new products widens quickly and economically favorably. Polyamide (PA) blends for engineering applications appeared much earlier in 1948, when composites were developed based on PA66 and polyvinyl acetate (2). However, a wide-range development of new materials by means of blending existing homopolymers and copolymers was initiated in the 1970s (1–5); beginning with the work of Ide and Hasegawa (6), there were expanded investigations related to the technology of reactive compounding of PA blends (7). In 1975, Du Pont presented an ultra impact strength PA alloy, Zytel-ST, produced by reactive processing of PA66 with maleinated EPDM (copolymer of ethylene, propylene, and 2-ethylidene-5-norbornene) (2). Then followed investigations of 1

Division of Plastic and Rubber Processing, Institute of Material Technology, Poznan University of Technology, Piotrowo 3, 60-950 Poznan, Poland 2 Laboratory of Chemical Technology of Polymeric Composite Materials, V.A. Belyi Metal-Polymer Research Institute of National Academy of Sciences of Belarus, 32a Kirov Street, 246050 Gomel, Belarus Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

527

528

Polyolefin Blends

PA/polyolefin (PA/PO) blends, the main advantages of these in comparison with homopolyamides being increased impact strength, reduced water absorption, improved processability, and dimensional stability at varying humidity (9–11). Considering that among aliphatic PAs PA6 and PA66 are most abundant (in EU countries they account for 92% of all PAs produced), blended systems have been developed mostly based on them. Without making use of the reactive processing technology, it would in fact be impossible to create valuable engineering products based on PA/PO blends. Because of high density of cohesive energy (23.2,23.3,21.9, and 20.3 ðJ cm3 Þ0:5 for PA6, PA66, PA610, and PA12, respectively), aliphatic PAs cannot form compatible blends with polyolefins whose density of cohesive energy is considerably lower (16.1 ðJ cm3 Þ0:5 for PE and 16.3 ðJ cm3 Þ0:5 for PP). That is why PA/PO blends are characterized by a high surface tension (14–18 mN m1 ), weak interphase adhesion, and distinct phase separation (8,9). The properties of such blends—like those of many other systems with separated phases—strongly depend on interphase interactions and phase morphology. As the same is true for properties such as viscosity, strength, and high elastic properties of the blends, which are the subject of our further consideration, here is a brief analysis of certain aspects of the PA/PO interface and contemporary methods of reactive compatibilization for materials of this type.

18.2 COMPOUNDING AND INTERPHASE PHENOMENA IN PA/PO BLENDS In order to create morphology that is stable during use, the free energy of the blend (DGbl ) must be negative (4): DGbl ¼ DHbl TDSbl

ð18:1Þ

In Equation 18.1, the role of the entropy factor is negligible (DSbl ! 0), so a negative value of DGbl can only be reached if the process of blending is exothermic (DHbl < 0). In other words, the process must proceed with heat liberation, which may be caused by specific interactions between the blend’s components. Interactions that lead to strong covalent or ionic (base–acid) binding and to relatively weak bonds like hydrogen, ion–dipole, dipole–dipole, or donor–acceptor ones are quite probable (4,14,18). As in PA/PO blends mostly Van der Waals forces are feasible, these blends are always heterogeneous and show distinct phase boundaries. The following morphologies are most typical: dispersion of one polymer within the matrix of another one and cocontinuous two-phase morphology, and depend on the character of components, viscosity of their melts at blending, composition, and conditions of blending. The majority of investigations on PA/PO blends and their other types are dedicated to formation particularities of dispersed structures in them. The particularities of continuous morphologies formed and properties of materials containing coexisting continuous phases have been investigated to a lesser degree (19). Because

Chapter 18 Functionalized Polyolefins and Aliphatic Polyamide Blends

529

Figure 18.1 Methods for preparing PA/PO compatibilized blends.

of weak interphase adhesion in PA/PO blends, domains of the dispersed phase undergo aggregation and form spherical or elliptical micron-sized particles that are dispersed within the matrix of the basic polymer. When preparing specimens for morphological studies by ultramicrotoming under liquid nitrogen, some domains are torn out of the polymer matrix, which is a consequence of low adhesion between the phases (20). The key trend of creating valuable engineering PA/PO blends is compatibilization, which is a combination of procedures intended to improve the compatibility of the components either by specific means of blending or by introducing compatibilizers into the blend. Irrespective of the means of compatibilization, the blend shows a high dispersion degree of the dispersed phase, adhesion between the phases, resistance toward coalescence, and improved processing characteristics (2,13,14). The major methods of preparing compatibilized PA blends are given in Fig. 18.1 (9). Compatibilization of PA/PO blends is based on mechanochemical reactions that take place in a polymer melt during compounding; these reactions follow the radical mechanism and result in grafted or block copolymers (1,15–18). The regimes of blending play a decisive role in the development of phase morphology. An increased shear rate of the melt leads to smaller particles of the disperse phase. The particle size is inversely proportional to the shear stress applied. The ratio of viscosities of the molten components influences dispersion of components more than changes in shear stresses. For example, in the case of PA6/PE blends extruded through a shaping die, an increase of the shear stress from 19 up to 29 kPa causes a change in the morphology of PA6 particles from a spherical to a wavelike one (21). Application of the compatibilization method by means of melt blending is not decided as yet because of inadequate understanding of and difficulty in realization of real technologies, which require complex apparatus.

530

Polyolefin Blends

The effect of compatibilization of grafted or block copolymers as controllers of interphase interaction has been well known (1,28). Marosi and Bertlan (29), for example, described the compatibilizing effect of polybutylene terephthalate– polytetramethylene oxide block copolymers on PA6/HDPE blends. The addition of a block copolymer prevents separation in the blends during processing and increases the impact strength of the material several times. According to Paul and Newman (1), even 2 wt% of an A–B-type copolymer is sufficient for compatibilization of blends of A and B polymers if its molecular weight is 104. The block copolymers reduce the surface tension and raise the adhesion between phases. The latter may rise up to 100 times against the adhesion between the components (7). Despite the efficiency of the compatibilization with the help of block copolymers, their use in commercial blends has been on a rather limited scale since no profitable ways are available to synthesize such copolymers. The present authors are unaware whether the market offers any PA/PO blends compatibilized with block or grafted copolymers prepared in advance. A more advantageous alternative is to prepare compatibilized PA/PO blends, like many other types of blends, by reactive processing that implies creation of composites with a required level of interphase interaction in situ during compounding (18). Grafted polymers, whose macromolecules contain necessary functional groups, are most often used as compatibilizers. Reactive polymers, during blending, undergo chemical reactions similar to those in low molecular weight organic substances. Reactivity of functional groups depends little on the molecular weight (18,23–25). However, it must be taken into account that steric obstacles created by the main polymer chain, as well as restricted diffusion, reduce the reaction rate. Therefore, to realize reactions leading to compatibilization in a short period of blending (usually a few minutes), the grafted polymer (compatibilizer) must contain sufficient quantity of functional groups; the reaction must be fast and selective; the conditions of mixing must ensure a maximum contact surface between the interacting components (9). Reactions such as amidization, imidization, etherification, aminolysis, amide– ester exchange, ring opening, ionic bonding, and neutralization of carboxyl groups proceed fast enough in the polymer melts at the regimes of reactive compounding of PA/PO blends (9,23,25,26). Numerous compatibilizers can be produced by grafting monomers (containing some kind of functional groups) in melt onto homopolymers and copolymers of olefins or their blends (25). Reactive extrusion (26) is a basic process for this when the twin-screw extruder is used as a reactor of continuous action (27). When developing multicomponent blends, it should be taken into consideration that the greater the number of components (n), the greater the number of interfaces (N) between them: N ¼ nðn 1Þ=2. Interfaces are potential sources of initiation of damage spots (7,28,29). Therefore, the most important strategic trend in creating multicomponent blends is the introduction into a system of, at least, one ingredient with reactive groups, which can interact with the components of the blend and lead to its compatibilization.

Chapter 18 Functionalized Polyolefins and Aliphatic Polyamide Blends

531

It is worth mentioning that interphase interactions in blends proceed only within mesophases whose thickness for incompatible polymers is between 2 and 50 nm depending on the thermodynamic interactions of the phases, temperature, regimes of mixing, and some other factors (7,11,29,30). The mesophase thickness depends on pﬃﬃﬃﬃﬃﬃ the miscibility of the components and in first approximation (7) is x12 , where x12 is the parameter of thermodynamic interaction. It is one of the most significant factors that determines mechanical characteristics of the blends, especially the energy of impact failure (30,31). A mesophase of considerable thickness (>6 nm) constructed by the hierarchical method (28,29) promotes energy dissipation of impact failure and leads to microcrack healing. Intensified interphase interactions lead to mesophases of great thickness and restricted molecular mobility (29). This fact reduces the entropy; in the case of short chains, however, this fact is of a secondary importance. That is why polymer systems contain, in zones of interphase contact, extra number of shorter chain segments (32) or low molecular weight substances (32–37). This situation can be easily understood while counting the quantity of chain conformations; this quantity is higher if the chain terminal is on the interface rather than on the averaged statistical segment. It is clear, therefore, that polymer–polymer interphase boundary represents the region for most intensive interactions in systems with separated phases. The interface becomes a reaction zone. The major concentration of low molecular weight substances in the mesophase is used to promote compatibilization of PA/PO blends. For example, compatibility is improved in a PA/PE system if compatibilizer was alkylmaleic monoamide, alkylmaleic monoester (38), or stearic acid (39). Low molecular weight compatibilizers being reactive surfactants (29) allow controlling the morphology of PA/PO blends. Their favorable influence on the mechanical characteristics of the materials, however, is weaker than that of high molecular weight additives. One of the explanations can be the formation of weak mesophases if low molecular weight compatibilizers concentrate in zones of interphase contract (7,28,29). Because of this, PA/PO blends with reactive compatibilizers of macromolecular nature, so-called in situ compatibilized materials, have wider applications (7,11,40,41). As mentioned above, they are prepared by reactive processing. The properties of such materials depend much on the interactions on the interfaces occurring, as a rule, through the acid–base mechanism and depending, in their turn, on the nature and concentration of functional groups included in the compatibilizer chains, and on the nature of the olefin polymer or copolymer containing functional groups (7,29). The sulfoacidic group interacts stronger with PA than the carboxyl group (42). The anhydride group is also more reactive than the carboxyl one (7). It should also be considered that amide groups in PA are less reactive than terminal primary amino groups (24,41). However, as the concentration of amide groups is significantly higher than that of amino groups, the contribution of interactions with participation of amide groups can be quite important. The reaction between amide and anhydride groups results in PA chain opening and splitting off water molecules (41). The latter can cause hydrolytic splitting of the chains (24). That is why the region of interphase contact becomes enriched with PA chains of a lower molecular weight in comparison with the volume.

532

Polyolefin Blends

Figure 18.2 Strength dependence of adhesion bonds for film specimens of PA6, HDPE (1), and HDPE-g-MAH containing 1 wt% of grafted maleic anhydride (2) on formation time of adhesion contact at 220 C.

Reactions between anhydride and amino groups predominate in the formation of grafted copolymers that are obtained during reactive processing in the presence of anhydride-containing compatibilizers (41,43–45). Reactions between anhydride groups and PA chains are most often used in obtaining compatibilized blends of PA with PO. Their major feature is high rate in PA melt, which is important for the reactive compounding process. The consequence of such reactions is the increased interphase adhesion observed even in an early moment of contacting between the reagents. This fact can be visually illustrated using data on the kinetics of adhesional contact formed between PA6, neat HDPE, and HDPE grafted with maleic anhydride (MAH) (Fig. 18.2). It is evident that MAH grafted onto HDPE promotes a sharp rise in adhesion between the phases. The kinetics of adhesional contact formed in a polymer blend can be shown still more clearly using the comparative analysis of values of the conditional growth rate of adhesion between the phases (A ) (46,47): dA A ¼ lim t!0 dt

ð18:2Þ

where t is the time of creation of an adhesional contact. As the A value characterizes, in fact, the adhesional strength of bonding at a momentary contact between the polymers (t ! 0), this parameter does not account for the contribution of diffusion events in the adhesional interaction. The A value takes into account only energy interactions of unlike macromolecules in zones of interphase contact (47).

Q5 Chapter 18 Functionalized Polyolefins and Aliphatic Polyamide Blends

533

Figure 18.3 Effect of maleic anhydride grafted onto HDPE on Conditional growth of adhesion strength (A ) for joints of PA6/HDPE-g-MAH; temperature of adhesion contact is 220 C.

It can be seen in Fig. 18.3 that 1.2 wt% of MAH introduced into the HDPE structure raises A value more than 30 times, implying an extremely high rate of interaction between anhydride groups and PA6. Thus, grafting of MAH onto PE leads actually to momentary generation of an interphase contact zone. Similar results have been obtained by Yukioka and Inoue (49), who investigated the generation of interfaces in blends of amorphous PA and copolymer of styrene and acrylonitrile (SAN). The mesophase thickness was about 29 nm. The grafted copolymer was concentrated in this zone and its thickness was in fact independent of the time at 200 C. The extreme pattern of concentration dependence of A (Fig. 18.3) can be explained by the fact that 1.2 wt% of MAH is sufficient for bonding of all nonassociated amino and amide groups of PA6 in zones of interphase contact. A reduction in A at MAH concentration of >1.2 wt% can be caused by restriction of molecular mobility of the components and resultant deteriorated microrheological conditions of formation of an adhesional contact, as well as by accumulation—in the adhesional contact zone—of low molecular weight products of the reaction; these products give weak mesophases and weaken adhesion between the phases. In most blend systems of PA/PO compatibilized with olefin polymers and copolymers, are grafted with MAH, therefore, the concentration of the latter does not exceed 1–3 wt%. The main aim of grafting is (8) to ensure compatibility of polyolefins with polar polymers, which can be reached at the level of PO grafting of 0.1–0.5 moles of polar groups per 100 monomer units. A high rate of formation of adhesional contact between the phases in a PA/PO-gMAH system is very important from the technological viewpoint, because fast interphase interactions during compounding of materials depend little on subsequent processing of the blends.

534

Polyolefin Blends

When choosing compatibilizers for PA blends, the reactivity of amine and amide groups, as well as terminal carboxyl groups, in PA should be considered (24,50). Therefore, despite predominant use of polyamide–anhydride reaction for compatibilization, the compatibility and related engineering properties of the blends can be improved by introducing into macromolecules PO carboxyl, neutralized carboxyl (55,56), ester (24,52), oxazoline (58,59), epoxy (24,60,61), isocyanate (62), or other functional groups (24). Insufficient reactivity of some functional groups, in comparison with anhydride ones, can be compensated at the expense of their higher concentration. The concentration of carboxyl groups is sometimes increased upto 10 wt% or even upto 20 wt%.

18.3 RHEOLOGICAL AND HIGH ELASTIC PROPERTIES OF PA/PO MELTS The knowledge of rheological and high elastic properties of PA/PO melts is rather important for optimization of regimes of their processing, and for obtaining information on the flow mechanism and its influence on the morphology and engineering properties of the materials. Of particular interest are insufficiently studied blends, which show high viscosity and strength in melts. The fact is that aliphatic PAs, while having narrow molecular weight distribution and low viscosity of melt (MFI of PA6, e.g., is usually within the range of 10–25 g/10 min at T ¼ 250 C and P ¼ 21:6 N), are unsuitable for extrusion technologies. However, POs characterized by increased viscosity and strength of melts have been widely used for processing by blow extrusion. It seems promising that introduction of highly viscous PO into PA can lead to composites with rheological and high elastic properties typical of the extruded materials. In a general case, the melt viscosity of a polymer blend (hbl ) depends on the melt viscosities of the blended components and blend’s composition (1,65). The character of mutual influence between the components in a polymer blend on hbl can be described using the logarithmic rule (18): log hbl ¼ Si ui log hi

ð18:3Þ

where ui and hi are, respectively, the volume share and melt viscosity of an ith component. There are four types of the polymer blends: (i) Additive blends whose melt viscosity follows Equation 18.3, (ii) Blends with a positive deviation of hbl from Equation 18.3. These include blends with strong interphase interactions. (iii) Blends with a negative deviation from the logarithmic additivity, which is typical of incompatible components with weak interphase interactions. (iv) Blends that show both positive and negative deviations of hbl from the additive values (such a relationship is typical of materials in which structural changes take place during flowing). The viscosity of blends varies not only with the composition but with flow conditions as well, which depend on the temperature and shear rate (20,41,65). A decisive effect on PA/PO rheology is caused by the chemistry of interphase processes (41). Blends of PA with ungrafted PO, or uncompatibilized blends, are characterized by a decreased melt viscosity in comparison with the additive

Chapter 18 Functionalized Polyolefins and Aliphatic Polyamide Blends

535

value of hbl determined from Equation 18.3. In another study (20), this fact was clearly demonstrated for the PA6/PP blend. On increasing the shear rate g from 101 up to 103 s1 , the deviation of melt viscosity for PA6/PP blend from its additive values became stronger over the whole range of composites. This finding supports a lack of compatibilization and active interactions between the phases in PA6/PP uncompatibilized blends. The influence of a shear rate on hbl can be explained by variations in the morphology of the dispersed phase during flowing. On rising g or shear stress, spherical particles of the dispersed phase become elongated and oriented in the flow direction. At a low g , particles of the dispersed phase undergo minor deformation, and then hbl depends mainly on the size of drops and interaction between the phases. Blending of PA with grafted PO (g-PO), or compatibilization of PA/PO blends by addition of g-PO, changes the dependence pattern of hbl with component ratio and technological factors. As PA/PO compatibilized blends have a finer dispersed phase with a developed interface and show more intense interphase interactions, flowing disturbs a little the phase morphology, and hbl often becomes higher than the additive values and viscosity of any of the blend’s components (6,20,66–69). The viscosity of PA blends containing PP-g-MAH correlates with the volume of the mesophase falling at a volume unit of the blend (70). It is presumed that a decisive influence on hbl is caused by the concentration of the grafted copolymer, which becomes immobile within the mesophase formed during blending. When assessing the rheological behavior of PA/PO blends, a strong effect of shear forces upon hbl should be considered. The reason is a qualitative difference between the flow curves for PO and PA. Aliphatic PAs show an extended Newtonian plateau typical of polymers with a narrow MWD (71). PA6, for instance, can retain the Newtonian pattern of flow (72) up to a shear rate of 103 s1 . The curve describing the relationship of h versus g for PO is typical of polymers with a wide MWD: the anomaly in viscosity (h decreases with increase in the shear rate) was observed at a much lower shear rate of 102 s1 . That is why the effects of viscosity’s growth—in the case of PA6/PO compatibilized blends—manifest themselves to the utmost at relatively low shear rates, upto 102 s1 . Such shear rates are typical of extrusion of polymer materials (72). Reactive compounding of PA6 with HDPE compatibilized with a mixture of HDPE/copolymer of octene and ethylene with grafted MAH (Fig. 18.4) gives an increased viscosity of the melts especially at low shear rates (66). An increased concentration of the compatibilizer in a blend between 15 and 30 wt% causes the viscosity to rise. Unlike pure PA6, PA6/g-PE blends show 10-fold or higher increase in viscosity at low shear rates. Consequently, to create PA/PO blends with a high melt viscosity it is advisable to use fully functionalized polyolefins. It can be expected that proper dispersion in melt of PA blended with high viscosity g-PO can yield composites with satisfactory structural homogeneity and high melt viscosity. This possibility is based on the above data showing that contact between the phases in a PA/g-PO blend is created instantaneously, and the development of contact zones between the polymeric phases depends much on the degree of mechanical dispersion of the components and not on the diffusion processes taking place in the mesophase.

536

Polyolefin Blends

Figure 18.4 Effect of shear rate on melt viscosity at 275 C of PA6, PE, and PA6/PE blend compatibilized by PE-g-MAH containing 0.27 wt% of grafted MAH. Ciphers on curves stand for compatibilizer concentration in wt%, PE stands for LDPE/LHDPE (88:12) blend; LHDPE is copolymer of ethylene and octene. Reproduced from Reference (66) with permission from John Wiley & Sons. Inc.

We proved this presumption on blends of PA6 with LDPE and HDPE and on a blend of PE with ethylene–propylene rubber (EPDM). The polyolefin components were functionalized by grafting 1 wt% of itaconic acid (IA) (51,53). Functionalization and preparation of the blends were performed in a static mixer assembled onto the Brabender plastograph (51,73). Grafting with simultaneous cross-linking of PO resulted in a sharp reduction of MFI values and h of their melt. It can be seen in Table 18.1 that g-POs possess much higher melt viscosities than PA6. The viscosity of LDPE exceeds h of PA6 by two decimal orders. Ea of PA6 is twice as low as that of g-PO. The values of hMFI and Ea in Table 18.1 were calculated from MFI values (74): hMFI ¼

P rc r tc þ n rc Þ MFI

2D2p ðLc

ð18:4Þ

where P is the load; rc is the capillary radius; r is the density of polymer melt at a definition temperature of MFI, r ¼ 4m=pD2p L; tc ¼ 600, it is constant; Dp is the piston diameter; Lc is the capillary length; n is the input correction factor (its variations have not been accounted for in the calculations) (74); L is the stroke distance of the piston; and m is the volume of the extruded melt. Ea ¼

R ln k T1 T2 T2 T1

ð18:5Þ

Chapter 18 Functionalized Polyolefins and Aliphatic Polyamide Blends

537

Table 18.1 Properties of Blend Components. Component

Tm , C

PA6

222

LDPE

106

g-LDPE

106

HDPE

126

g-HDPE

126

g-LDPE/g-EPDM (50/50) g-HDPE/g-EPDM (50/50) PP

105*

124* 165

g-PP

165

MFI, g=10 min at T C 5.3 (230 C) 7.7 (250 C) 5.2 (230 C) 7.6 (250 C) No flow (230 C) 0.14 (250 C) 8.5 (230 C) 11.1 (250 C) No flow (230 C) 0.6 (250 C) No flow (230 C) 0.21 (250 C) No flow (230 C) 3.1 (230 C) 4.4 (250 C) 15.3 (230 C) 15.5 (250 C)

hMFI 103 , Pa s

Ea , kJ mol1

2.7

40.7

6.26

41.5

286.8 4.26

98.2 29.2

86.9

93.0

135.8

91.0

126.5 8.11

90.3 38.3

2.19

2.8

Notes: Tm is the melting temperature; r is the density; hMFI and Ea are the melt viscosity at T ¼ 250 C and the effective activation energy of viscous flow evaluated from MFI values; asterisks denote values of Tm for the PE phase.

where R is the universal gas constant; T1 and T2 are the temperatures at which MFI is determined (T2 > T1 ; P ¼ constant); and k ¼ MFI1 =MFI2 (at P1 ¼ P2 ; T1 < T2 ; the residence time in the melting cylinder being constant). Determined from Equation 18.4, hMFI is non-Newtonian. Despite a wide range of g values observed during determination of MFI (approximately between 0.1 and 15 s1 ) (75), it does not fall at the region of intensive development of viscosity anomaly for PO, the more so for PA. It is clear, therefore, that variations in g — during determination of MFI—for different composites cannot be the cause of differences in the values of this property. The concentration dependences of MFI are represented in Fig. 18.5. For all types of the blends there is a sharp decrease in MFI at g-PO 30 wt%. For blends containing 30–40 wt% of g-PO, the value of MFI is between 0.2 and 0.8 g/10 min, which is 10–26 times lower than that of pure PA6; the level of values corresponds with the requirements imposed on extrusion-processed materials (74). It is of interest that irrespective of g-PO type, at a concentration of 30–40 wt%, the MFI values of PA blends are quite similar despite a great difference in melt viscosities of g-POs used for blend preparation (Table 18.1). This can probably be explained by the fact that the continuous phase, during melt flowing of a blend, is formed by lesser viscous PA6 (76), which dominates in the blend. The decisive influence on the flow development of such blends comes from interphase interactions that are alike for all

538

Polyolefin Blends

Figure 18.5 Concentration dependence of MFI for binary PA6/g-PE (a) and ternary PA6/(g-PE/gEPDM) (b) blends; P ¼ 21:6 N; T ¼ 250 C.

types of the blends as all of g-POs used have been functionalized similarly. The higher the g-PO concentration, the more intensive are interactions between the phases and the more the high viscosity g-PO phase gets involved in the flow process, which results in a sharp rise of the total melt viscosity of the blend. PA6 and PA6/g-PO melt viscosities (Table 18.2) varied with shear stresses acting within melts during their movement through a capillary. The studies have been done using the capillary viscosimeter of Instron 1115 machine (capillary diameter: 1.225 mm; length; 5 mm). The analysis of data in Table 18.2 shows that melts of pure PA6 and its blends with g-PE behave like typical non-Newtonian liquids. The melt viscosity of the blends containing 30–40 wt% of g-PO is so high that at a shear stress of 0.5 MPa it is impossible to press them through the viscosimeter’s capillary. Some information for comparison of rheological behavior of PA6 blends with pure and grafted PO is given in Table 18.3. Values of MFI <1 g/10 min for PA blends

539

Chapter 18 Functionalized Polyolefins and Aliphatic Polyamide Blends

Table 18.2 Effect of Shear Stress on Melt Viscosity of PA6 and its Blends with g-PE (T ¼ 235 C). h 103 (Pa s), against shear stress in MPa Material, weight shares

0.5

0.75

1.0

2.0

3.0

PA6 PA6/g-LDPE (85/15) PA6/g-LDPE (70/30)

6.51 9.32 Melt would not pass through the capillary

5.40 7.57

4.22 4.7

0.80 2.26

— 0.97

16.96

13.66

7.05

3.95

As above As above

12.18 14.70

7.27 11.57

4.92 5.39

2.61 2.18

As above

20.10

10.21

5.11

2.13

PA6/g-LDPE (60/40) PA6/g-HDPE (70/30) PA6/(g-HDPE/g-EPDM)a (70/30) a

50 wt% g-HDPE/50 wt% g-EPDM.

can only be obtained using high viscosity g-PO. Thus, functionalized PE or g-PE/gEPDM blends used for PA6 modification allow obtaining the blended materials with high melt viscosities suitable for processing by extrusion technologies (74). The acceptability of polymer materials for processing by extrusion technologies is determined not only by their melt viscosity but also, largely, by viscoelastic properties. Such materials are characterized, first of all, by strength and swelling of the melt jet (74,77). An extrudate jet swells owing to normal stresses generated in a moving melt (Weissenberg effect) under shear stresses (71,74). Sheared flow of polymer melts under shearing forces is accompanied by forced changes in macromolecule conformations as compared with the equilibrium condition. This results in creation Table 18.3 Effect of PO Type on Rheological Behavior of PA6/PO Blends as Obtained from Analysis of MFI (Concentration of PO ¼ 30 wt%, P ¼ 21:6 N). Material PA6/LDPE PA6/g-LDPE PA6/(g-LDPE/ g-EPDM)a PA6/HDPE PA6/g-HDPE PA6/g-HDPE/ g-EPDM)b PA6/g-PP

T, C

MFI (g/10 min)

hMFI 103 , Pa s

Ea , kJ mol1

250 250

11.7 0.8

2.21 32.6

71.2 107.2

250 250 250

2.2 15.4 0.75

11.85 1.69 26.45

86.2 39.8 126.1

250 230

1.5 15.7

19.33 2.99

104.4 2.63

a

50 wt% g-LDPE/50 wt% g-EPDM.

b

50 wt% g-HDPE/50 wt% g-EPDM.

540

Polyolefin Blends

and accumulation of highly elastic (reversible) deformation. At the die outlet (narrow ring-shaped gaps, slots, or small orifices), after the shear stresses disappear, the heat-induced motion of segments tries to bring back macromolecules, elongated in the flow direction, into an equilibrium coiled state. As a consequence, the melt stream swells (the cross-sectional area of the extrudate exceeds that of the die) and the longitudinal dimensions get reduced. A final dimensional variation in the extrudate stream (jet) can only be reached at a certain distance from the outlet of the die channel as the high elastic deformation undergoes relaxation during some length of time. The extrudate swelling usually grows with shear rate when the melt flows through the die and when the temperature falls (74). The melt jet becomes less swollen while passing through the capillary, if the relative channel length L=D is increased, and reaches a minimum value at a definite L=D, remaining in fact constant with further increase in L=D (1,74,76). It is believed that at L=D 20 the melt stream becomes formed (steady state): some of the highly elastic deformation of the chains becomes relaxed in the channel so that at its outlet the effect of jet swelling is minimal (1,76). At the regime of steady flowing, all of the external work is spent on overcoming the resistance to viscous flow, that is, on the development of required deformation of the flow. For the case of underdeveloped (unsteady flow) stream (L=D < 20), the swell effect of the jet is most pronounced. The latter situation is typical of blow extrusion, because the length of a ring-shaped channel in the extruding head is much lesser than its diameter (74,76). The swell effect of a jet is also influenced by channels’ ratio (D=d, i.e., relationship between the larger and smaller diameters of the linking channels (74,78). The strength of a polymer melt (s m ) belongs to important characteristics of extrusion- and blow-molding-processed materials. This property depends much on the viscosity, high elastic properties of the melt, molecular weight of the polymer, temperature, and other technological parameters (74,76). It is very important for making large-sized blown articles for which tubular billets (workpieces) of great size and weight are required. Information on the swell effect of a melt jet and the strength of melt is quite important for the selection of a design of molding instruments and technological parameters of processing that ensure manufacturing of high quality products. Available information on the rules of swelling and strength of a polymer melt concerns mostly polyolefins and their blends (79–87). Data on polyamide-based materials are limited and refer to homopolymers (74). Polyolefin blends always show a raised swell of the extrudate jet in comparison with homopolymers (83). This was shown earlier (88) for LDPE/LLDPE blends. Some researchers reported (89,90) a significant influence of the molecular structure on properties of the polymer melts: the polymers whose chains contain starlike branches show increased sensitivity toward shear stresses applied onto their melts. The presence of comblike segments increases the melt strength. For example, the addition to LLDPE up to 10 wt% of a comblike polymer (copolymer of a polyethylene macromonomer, Mn 8000 g mol1 , and ethylene and hexene) allows raising the strength of polymer melt almost by one decimal order (86). The MFI of the blend varies less substantially. The melt strength of LLDPE blends can be

Chapter 18 Functionalized Polyolefins and Aliphatic Polyamide Blends

541

controlled by varying the concentration of comblike polymer additions and also by the macromolecular branching of this polymer (86). The combination of LLDPE with simultaneously introduced additions of comblike and starlike polymers in a polyethylene blend allows controlling efficiently the melt strength and viscosity of the blends (86). The role of the flow mechanism for molten LLDPE in its swelling and strength properties is important (87). Provision of melt slippage with respect to the capillary walls (this can be achieved by means of coatings with low surface energy (91–93) or by introduction into the polymer of an additive of a fluoropolymer that is immiscible with polyethylene and migrates during flowing toward the capillary wall (94,95)) restricting swelling of the melt jet. The introduction, into the LLDPE, of 0.1 wt% of a fluoropolymer additive ensures melt slippage relative to the capillary wall (87). As a result, the swell extent of the extrudate decreases and its strength and ability to elongation rises. This is explained by the fact that melt slippage relative to the capillary wall reduces the shear stress in the entire polymer and consequently decreases molecular orientation in the flow direction. Swelling is highly sensitive to the molecular weight and MWD in the LDPE melt. An increase in the molecular weight along with polydispersity index leads to greater swelling; variations in the shear rate of the melt influence not so much the jet swell as variations in MWD do. When a jet of molten LDPE is extruded into oil heated above the melting temperature of the polymer, the extrudate continues to swell for a few minutes reaching some maximal value (80). Several works (96–98) describe analytically swelling of polymer melts during extrusion through short channels. The obtained relationships have been chiefly tested on rubber blends and they do not take into consideration specificity of a structure formation in semicrystalline thermoplastics. For studies of swelling, the extrudate prepared at given values of technological parameters is cut into sections of definite lengths and cooled without damaging the shape. The extrudate diameter is measured with a digital micrometer (83). To prevent the melt jet from distortion while it touches the hard surface of the device base, extrudate sections are placed into a liquid, for example, heated water, from which the specimens are removed after the polymer has solidified (99). The swelling of a melt jet can be measured, at some distance from the capillary, using a special optical device fitted to the rheometer (87), which excludes errors caused by polymer shrinkage at cooling. To determine the strength of a polymer melt, the capillary rheometer is provided with a special device (Goettfert Rheoteus) (85,86). The principle of measurements is that a strand of a polymer melt extruded from a polymer is caught between two, rotating in opposite directions, rolls at an accelerating speed. The maximal force required to disrupt the jet is taken as the melt strength. It may be useful to exercise the procedure because after leaving the capillary of a particular geometry at certain shear and heat effects upon the melt, its jet, after reaching a certain length, breaks under its own weight (74,99). The melt strength (s m ) there is determined as the ratio of an extrudate weight, at which the jet starts to break (the extrudate cross section decreases), to the cross-sectional area of the capillary. In this case, when s m

542

Polyolefin Blends

Figure 18.6 Concentration dependence of melt jet swelling and strength for binary PA6/g-PE (a) and ternary PA6/(g-PE/g-EPDM) (b) blends of PA. Melt temperature, 250 C.

is being determined, any errors that can result from the difficulty to measure the true cross section of an extrudate at the moment when the jet starts to break off are excluded (99). Experimentally obtained data on concentration dependences of BðB ¼ DE =DC , where DE and DC are, respectively, extrudate diameter and capillary diameter) and s m values for binary and ternary blends of PA are given in Fig. 18.6. The values of the parameters were determined by the procedure (99) where the capillary’s D ¼ 1:45 mm; L=D ¼ 1:5. The melt flow rate from the capillary was constant and provided for melt consumption at 5 g min1 . Unlike the pure PA6, its melts of binary blends with g-PE are characterized by 2.3 times increase in swelling and 50 times increase in strength (Fig. 18.6a). The values of B for binary and ternary PA blends are approximately the same (Fig. 18.6a and b). It should be noted that at increased concentrations of g-PO (40 wt%) s m values of molten ternary blends are 1.1–1.7 times as large as the melt strength of PA6/g-PE blends (Fig. 18.6a and b). Considering that ternary blends, unlike binary ones, show a lower melt viscosity

Chapter 18 Functionalized Polyolefins and Aliphatic Polyamide Blends

543

Table 18.4 Swelling and Strength of PA6 Molten Blends Containing Neat and Functionalized PE. PA blend composition, wt% PA6/LDPE (70/30) PA6/LDPE (60/40) PA6/g-LDPE (70/30) PA6/g-LDPE (60/40) PA6/HDPE (70/30) PA6/HDPE (60/40) PA-6/g-HDPE (70/30) PA6/g-HDPE (60/40) a

Ba, rel. unit

s m , kPa

3.6 4.3 1.8 2.0 3.5 4.0 2.2 2.3

3.1 3.6 7.3 8.0 1.8 2.4 8.3 10.8

B ¼ DE =DC , where DE and DC are, respectively, extrudate diameter and capillary diameter.

(higher MFI) and activation energy of the viscous flow, it can be assumed that the rheological properties of melts do not critically affect the B and s m values. It is reasonable to think that these properties must greatly depend on the energetics of interphase interactions in melts of binary and multiphase blends. This assumption has been supported by the results of comparative analysis of jet swelling and melt strength for binary blends of PA6 with pure PE as well as grafted PE (Table 18.4). It is clear that blends containing pure PE show 1.6–2.0 times increase in jet swelling and 1.7–3.5 times reduction in the melt strength against g-PE-based blends. Taking into consideration that the latter are capable of stronger interphase adhesion against PA6/PE blends (9,55,66,101), one can assume that interphase interactions in molten blends of immiscible polymers (1) cause some decisive effect on their high elasticity qualities and strength. Thus, blending of PA6 with g-POs giving melts of high viscosity results in composites, which show melt strength that, under the experimental regimes, exceeds 50 times that of the pure PA6. This allows considering PA6/g-PO blend systems as potentially suitable for processing by ordinary and blow extrusion methods. Molten blends of PA6 containing functionalized polymers, as well as ethylene copolymers, have a higher (upto three times) strength against PA6/PE blends. The main reason for this is stronger interaction between the phases. The interphase interaction intensified by functionalization of PO is accompanied by reduced swell of the extrudate when it passes through the capillary. It is quite probable that variations in B coefficient and s m for PA/PO blends have a similar pattern and are typical of mixtures of immiscible thermoplastics especially of systems with continuous phases.

18.4 MORPHOLOGY AND IMPACT STRENGTH OF PA/G-PO BLENDS THAT GIVE MELTS OF HIGH VISCOSITY AND STRENGTH A number of studies have been dedicated to investigation of PA/PO blends’ morphology and structurization of their components (1,20,41,55,56,58,60,61,66,

544

Polyolefin Blends

102–107). This interest can be explained by correlation between the structure and mechanical properties (among them, impact strength) of such blends. Most of the studies performed were directed at understanding of generation of structures dispersed in a polymer matrix. The conditions for continuous morphologies have been understood to a lesser degree. At concentrations of immiscible components approaching a ratio of 1:1, melts of the cooled and solidified blend materials have morphology like ‘‘a droplet in a matrix’’ or ‘‘layers in a matrix’’ depending on the type of components, ratio of their viscosities, and level of interphase interaction. The layered (ribbonlike) distribution is most typical (1). The electron optical method may not detect a dispersed phase in blends of PA with g-PO, so morphologically such materials seem single-phased. Similar situations were observed, for example, with PA6/polypropylene blends compatibilized by PP-g-MAH (20). Figure 18.7 gives scanning electron micrographs of cuts made under liquid nitrogen of PA6 blends containing 30 wt% of pure LDPE (Fig. 18.7a) and high viscous LDPE-g-IA (Fig. 18.7b) as well as HDPE-g-IA (Fig. 18.7c). The effect of functionalization onto blend morphology is rather strong. In blends with pure PE, the latter aggregates into droplike particles of up to 5 mm in diameter. They have a smooth surface. As mentioned above, similar morphology is typical of blended systems with distinct immiscibility of polymer components and weak adhesion between the phases (1,9,20,66). When PA6 is mixed with g-PE, the boundary between the phases smears and polyolefin is uniformly distributed in the polyamide phase despite a high melt viscosity of PE (Fig. 18.7b and c). No great differences have been observed in the morphology of blends containing g-LDPE or g-HDPE. Morphologically, both blends seem single-phased. Similar structures are characteristic of blends with high interphase adhesion and morphology resistant to mechanical fields or other service factors (1,9,19). According to DSC, WAXS, as well as relaxation spectrometry data, the investigated blends have a clear two-phase structure with PA6 crystallites free of a foreign material. A qualitative estimation of crystallinity by DSC data (Table 18.5) shows a marked interference of the blended components in their crystallization. The degree of this influence depends on both the nature of g-PE and its concentration in the blend. The main particularity is that in PA6/g-LDPE blends the crystallinity of PA6 phase grows against its calculated (additive) value while that of the g-LDPE phase drops. In PA6/g-HDPE blends, the crystallinity of both components is noticeably lower than the calculated amount (Table 18.5). Probably owing to a lack of compatibility of the components, the effects observed are explained by the specific character of interphase events in the blends, which depend on the molecular mobility as well as on the specific interactions of grafted carboxyl groups in g-PE and functional groups of PA6, and also on the melting and crystallization temperatures of blended components, and on thermal properties of their melts.

Chapter 18 Functionalized Polyolefins and Aliphatic Polyamide Blends

545

Figure 18.7 SEM micrographs of blends: (a) PA6/LDPE; (b) PA6/g-LDPE; and (c) PA6/g-HDPE; LDPE and g-PE concentration, 30 wt%.

It has been revealed by relaxation spectrometry technique (108) that the glass transition temperature for PA6 amorphous phase in blends remains in fact unchanged, being 55–58 C irrespective of the type of g-PE or its concentration. The temperature of b-transition (T b), being 60 C for pure PA6, shifts to the high temperature region in blends with g-PE. The peak of b-relaxation for PA6 is generalized to glass transition peak for g-PE. The shift extent of T bPA6 depends on g-PE concentration in the blend and on the type of g-PE. For 15 wt% of g-LDPE this shift is 4–8 C, while for 30 wt%, it is 15 C (108). For blends containing g-HDPE, the shifts are, respectively, 5 and 10 C. Obvious reasons for this are interactions between the phases in a

546

Polyolefin Blends

Crystallinity of Components in PA6/g-PE Binary Blends Determined by DSC. DIcr , rel. unit, for blend components PA6 g-PE Concentration of g-PE in blend, wt% Experimental Calculated Experimental Calculated PA6/g-LDPE 0 1.00 1.00 15 1.57 0.85 0.07 0.15 30 1.17 0.70 0.13 0.30 40 0.82 0.60 0.29 0.40 100 — — 1.00 1.00 PA6/g-HDPE 15 0.75 0.85 0.11 0.15 30 0.50 0.70 0.19 0.30 40 0.38 0.60 0.33 0.40 100 — — 1.00 1.00 Table 18.5

Note: The crystallinity index DIcr was found from the expression DIcr ¼ DH=DHo , where DH and DHo are, respectively, heats of crystallization of a polymer in blend and separately.

blend, which occur probably with participation of segments, found in chains of the polyolefin component, and finer structural units responsible for b-relaxation in PA6 (108,109). Thus, PA6/g-PE binary blends, whose melts are highly viscous and strong, are characterized, despite a lack of thermodynamic miscibility, by stable morphology; marked mutual influence of the components on their crystallinity and relaxation behavior results from interactions between the phases. Like binary blends, PA6/(g-PE/g-EPDM) composites are two-phased. Irrespective of the components’ ratio, their individual effects of melting and crystallization in blends are shown on the DSC curve. The comparative analysis of structure– morphology features for ternary blends, PA6/(g-PE/g-EPDM), and binary blends indicated that g-EPDM added to PA6/g-PE affects crystallization of the polyolefin as well as polyamide components. The results of DSC on the two types of blends, PA6/ (g-LDPE/g-EPDM) and PA6/(g-HDPE/g-EPDM), show increased crystallinity in g-PE against a binary blend (108,109). This effect results, most likely, from easier crystallization of g-HDPE in the presence of g-EPDM owing to raised molecular mobility because of plasticization by the elastomeric phase. It was noticed that PA6 crystallizes more fully in blends with g-LDPE/g-EPDM (108,109). In the case of PA6/(g-HDPE/g-EPDM), DIcr for PA6 is 1.2–1.6 times as low as the calculated values (108,109). This finding can be caused by specificity of interphase interactions in polymer melts. The scanning electron microscopy (SEM) revealed that both ternary and binary blends have a phase morphology typical of blend composites with active interphase interaction between polymer components. At equal concentrations of g-PO in blends, the ternary systems look more homogeneous than binary ones. It may be presumed,

Chapter 18 Functionalized Polyolefins and Aliphatic Polyamide Blends

547

therefore, that an elastomeric phase introduced in a blend promotes stabilization of the structure at different operation factors (1,15,109). As it was found by relaxation spectrometry, Tb PA6 shifts by 5–15 C toward the lower temperature zone for blends with g-PO against the pure polyamide (109). The extent of this shift increases with g-EPDM concentration in the blends. This finding witnesses a possibility of increasing low temperature resistance of the blends if g-EPDM is introduced. Blend morphology is a unique parameter (1,7) that determines the impact strength of blends. A reduction in the particle size of the dispersed phase up to a submicron or micron level is usually accompanied by an increased impact strength. The temperature of change from viscous to brittle failure also depends on the average size of particles and falls with its decrease. It was also noticed that an extremely high concentration of a compatibilizer in a blend might result in embrittlement of the latter (7). The rubbery state of at least one of the blended components, besides the structural microheterogeneity of polymer blends, is necessary to ensure their high resistance to impact failure (1,47,110,111). It was shown earlier (111) that polycarbonate/polybutylene terephthalate (PC/PBT) as well as polycarbonate/polyethylene terephthalate (PC/PET) blends, known as materials with a low impact strength, remain such only over a certain temperature range. At temperatures between Tg PET , Tg PBT , and Tg PC , the impact strength of their blends may be rather high. The effect of increased impact strength for PC/polyalkylene terephthalate (PC/PAT) blends is common for polymer blends having different glass transition temperatures (111). A large number of studies (1,11,12,66,101,104,106–109,112–119) and patents (120–129) have been dedicated to the development of technologies, understanding of structural features and properties of PA blends, and development of particular formulations for impact-resistant materials. The strategy of producing impact-resistant PA materials by extrusion is based on the introduction into the PA matrix of an elastomer in the presence of low molecular weight additives that allow to control interphase interactions as well as compatibility (120,121) such as rubbers functionalized by grafting of polar monomers or containing polar functional groups (122,123), polymers and copolymers (block and grafted copolymers) of olefins with dienes, whose macromolecules contain fragments of different functionalities (117–119,124–127), rigid-chain polymers (polyphenylene oxide and aromatic polyamides) with elevated temperature of the glass transition in combination with elastomers and compatibilizing additives (128,129). There are but a few studies that consider, in an interrelated manner, the problems of impact strength for PA and rheological as well as high elasticity properties of their melts (101,108,109). Paul and Newman reported (1) that g-EPDM added to PA6/g-PE binary blends, irrespective of g-PE used, raises the impact strength over a wide temperature range (Fig. 18.8). The extent of growth of the impact strength depends on both the g-PO concentration in the blend and the g-EPDM content in it (108,109). At optimal concentrations of the components, the impact strength at 23 C is over 40 kJ m2.

548

Polyolefin Blends

Figure 18.8 Temperature dependences of Charpy impact strength (specimens with sharp notches) for PA6 and its blends with g-PE (a) and (g-PE/g-EPDM) (b); g-PO concentration, 30 wt%.

A variation in the ratio of g-PE to g-EPDM causes a marked variation in the impact strength of PA blends. For the two types of blends, a maximum was observed at 50–70 wt% of g-EPDM in g-PE/g-EPDM blends. A rather high level of impact strength is found at negative temperatures (about 25 kJ m2 at 40 C and 18 kJ m2 at 60 C, Fig. 18.8b). Ternary blends, impact failed at 23 and 40 C, showed that whiting—typical of materials that fail by multiple crazing mechanism (1,47,110,111)—spreads over the whole failed surface. The SEM technique detected a developed system of pores whose size is between some fractions of a micrometer and several micrometers (Fig. 18.9). Evidently, pore generation is a consequence of severe crazing (1). As during modification of the PA6/g-PE binary blends by ethylene–propylene copolymer

Chapter 18 Functionalized Polyolefins and Aliphatic Polyamide Blends

549

Figure 18.9 SEM micrographs of impact-failed surfaces at þ23 and 40 C for specimens of PA6/(gPE/g-EPDM)—30 wt%.

the mechanical properties decrease somewhat (108,109) along with the energy spent on crazing initiation, it can be presumed that increased impact strength of materials results from increased energy spent on propagation of microcracks owing to crazing and creation of a developed surface of failure (111). An introduction of functionalized ethylene–propylene rubber into the PA6/g-PE binary system results in blends of ultrahigh strength while their melts remain highly viscous and strong. At impact stresses, PA6/(g-PE/g-EPDM) composites fail through multiple crazing; their high impact strength results from the developed surface of failure owing to severe crazing.

550

Polyolefin Blends

18.5 NEW APPLICATIONS OF PA/g-PO BLENDS The combination of high viscosity and strength of melts, as well as impact strength over a wide temperature range, predetermines practical use of PA/g-PO blends. The most important are articles manufactured by blow molding and extrusion. A number of developments related to PA blends, intended to be processed by extrusion methods (130–137), are available. Most of them are based on blend composites of PA6 or PA66 with PO or olefin copolymers functionalized by grafting polymer monomers (131–133). Some studies employed modification methods for PA melts based on cross-linking of certain low molecular weight compounds: cyanate monomers (134) or acrylphosphozides (136). Considering difficulties in controlling the chemical interactions of polymers with cross-linking substances (8,24–26), for practical applications the problems of interphase interactions and control of rheological, high elastic, and other properties, as well as impact strength, are solved in the course of reactive processing and selecting of g-PO (e.g., g-PE/g-EPDM) with a required functionality and a ratio of polyolefin-to-elastomer components. At 30–40 wt% of g-PO in a blend, polyamide develops a continuous medium. That is why, PA/g-PO blends have qualities typical of aliphatic polyamides, that is, resistance to petroleum products and abrasion, high level of mechanical properties, and increased thermal stability. In the level of impact strength and stress–strain properties, PA6/g-PO blends containing 30 wt% of g-PO are close to polyamide-11 (Rilsan). However, the blends are much cheaper than this polyamide. Rather promising, therefore, are the fields in which PA11 dominates, namely, flexible hoses and pipes for fuel systems and pneumatic systems for automatic blocking of brakes in car and tractor machinery, tanks for petroleum products, and others (101,108). Testing of such materials, when separators are being made for gasoline vapors, in cars has shown perspectives for aforementioned applications (137). Tests performed in VT SHED chamber following a procedure of OAO ‘‘Avtovaz,’’ Russia, on gasoline permeability of separators made from PA6/g-PO-based composites correspond with Euro-3 requirements (138).

18.6 CONCLUSIONS Reactive processing, in melt, of polyamide-6 with polyethylene functionalized by grafting of oxygen-containing monomers and partially chemically cross-linked polyethylene and its blend with ethylene–propylene copolymer allows to produce technologically compatible blend composites with viscosity upto 20 times and melt strength upto 50 times as high as those of the pure polyamide. The decisive role of interphase interactions in achieving these values is underlined: PA6 blends containing pure (unfunctionalized) PE show the melt strength upto three times as low as those of PA6/g-PE blends. A sharply raised viscosity, along with melt strength, for PA6/g-PO has been observed at g-PO 30 wt% concentration in a blend, that is, when g-PO approaches the generation of a continuous phase.

Chapter 18 Functionalized Polyolefins and Aliphatic Polyamide Blends

551

Owing to immiscibility of components, PA6/g-PE blends have a two-phase structure and are characterized by smeared zones of contact between phases (morphologically, the blends look quasihomogeneous); a common relaxation transition within the temperature region of PA6 (b-relaxation) and of g-PE (a-relaxation); and components’ crystallinity differ from the calculated (additive) values. These effects are caused by interphase interaction that occurs with the participation of oxygen-containing functional groups present in the monomer grafted onto PA. PA6/g-PO blends of high melt viscosity and strength are promising for products manufactured by extrusion and blow molding.

NOMENCLATURE A A a B Ea DGbl ; DHbl ; DSbl DIcr MFI P T g hbl hMFI sm t x12

Adhesional strength Conditional growth rate of adhesional strength Charpy impact strength for sharply notched specimens Swell of extrudate Effective activation energy of viscous flow Free energy, enthalpy, and entropy of a blend, respectively Crystallinity index Melt flow index Load on piston when MFI is determined Temperature Shear rate when molten polymer flows Melt viscosity of polymer blend Viscosity determined from MFI values Melt strength of polymer Time Parameter of thermodynamic interaction

REFERENCES 1. D. R. Paul and S. Newman (eds.), Polymer Blends, Academic Press, New York, 1978. 2. L. A. Utracki, Polym. Eng. Sci., 35, 2 (1995). 3. L. A. Utracki, Commercial Polymer Blends, Chapman and Hall, London, 1998. 4. O. Olabisi, L.M. Roberson, and M.T. Shaw, Polymer–Polymer Miscibility, Academic Press, New York, 1979. 5. D. R. Paul and J.W. Barlow, J. Macromol. Sci. Rev. Macromol. Chem., C18, 109 (1980). 6. F. Ide and A. Hasegawa, J. Appl. Polym. Sci., 18, 963 (1974). 7. T. Imone and P. Marechal, Reactive processing of polymer blends: polymer–polymer interface aspects, from Processing of Polymers, Vol. 18, H. E. H. Meijer (ed.), in: Reprint from Materials Science and Technology, R. W. Cahn, P. Haasen, and E. J. Kramer (eds.), VCH Verlagsgesellschaft mbH, Weinheim, 1997, Chapter 8. 8. C. Harrats and G. Groeninckx, Reactive processing of polymer blend using reactive compatibilisation and dynamic crosslinking: phase morphology control and microstructure–property reactions,

552

Polyolefin Blends in: Modification and Blending of Synthetic and Natural Macromolecules, F. Ciardelli and S. Penczek (eds.), Kluwer Academic Publishers, Dordrecht, 2003, Chapter 9.

9. S. S. Pesetskii and A. A. Bogoslavsky, Mater. Technol. Instrum., 2, 27 (1999). 10. P. Domininghaus, Gummi Fasern Kunststoffe, 45(8), 408 (1992). 11. D. R. Paul, High performance engineering thermoplastics via reactive compatibilization, in: Modification and Blending of Synthetic and Natural Macromolecules, F. Ciardelli and S. Penczek (eds.), Kluwer Academic Publishers, Dordrecht, 2003, Chapter 14. 12. H. Keskkula and D. R. Paul, Toughened nylons, in: Nylon Plastics Handbook, M. I. Kohan (ed.), Hanser Publishers, Munich, 1995, Chapter 11.6. 13. L. A. Utracki, Polymer Blends and Alloys, Hanser Press, New York, 1989. 14. S. Datta and D. J. Lohse, Polymeric Compatibilizers: Uses and Benefits in: Polymer Blends, Hanser/ Gardner Publications Inc., Munich, 1996. 15. B. Jurkowski, K. Kelar, and D. Cieselska, J. Appl. Polym. Sci., 70, 1641 (1998). 16. B. Jurkowski, K. Kelar, D. Cieselska, and R. Urbanowicz, Kautschuk Gummi Kunststoffe, 47(9), 642 (1994). 17. B. Jurkowski and Y. A. Olkhov, J. Appl. Polym. Sci., 65, 1807 (1997). 18. C. Koning, M. V. Duin, C. Pagnoulle, and R. Jerome, Prog. Polym. Sci., 23, 707 (1998). 19. A. D. Pomogailo, Uspekhi khimii (Russ. J. Adv. Chem.), 71, 1 (2002). 20. S. Y. Park, B. K. Kim, and H. M. Jeong, Eur. Polym. J., 26, 131 (1990). 21. R. Conzalez-Nunez, D. Dekee, and B. D. Favis, Polymer, 37, 4689 (1996). 22. S. S. Pesetskii, V. D. Fedorov, M. B. Kaplan, and I. P. Storozhuk, Izvestia AN BSSR (Trans. Natl. Acad. Belarus), Phys Eng. Ser., 4, 27 (1991). 23. A. O. Baranov, A. V. Kotova, A. N. Zelenetsky, and E. V. Prut, Uspekhi khimii (Russ. J. Adv. Chem.), 66, 972 (1997). 24. G. Moad, Prog. Polym. Sci. 24, 84 (1999). 25. S. Al-Malaika, Reactive Modifiers for Polymers, Chapman and Hall, London, 1996. 26. M. Xanthos, Reactive Extrusion, Hanser, Munich, 1992. 27. W. Michaeli and A. Grefenstein, Polym. Eng. Sci. 35, 1485 (1995). 28. A. Ajji and L. A. Utracki, Polym. Eng. Sci., 36, 1574 (1996). 29. G. J. Marosi and G. Bertlan. Role of interfaces in multicomponent polymer system and biocomposites, in: Modification and Blending of Synthetic and Natural Macromolecules, F. Ciardelli and S. Penczec (eds.), Kluwer Academic Publishers, Dordrecht, 2003, Chapter 8. 30. E. P. Pludemann, Interfaces in Polymer Matrix Composites, Academic Press, New York, 1974. 31. K. L. Foster and R. P. Wool. Macromolecules, 24, 1397 (1991). 32. M. Staim and D. W. Schubert, Annu. Rev. Mater. Sci., 25, 325 (1995). 33. E. Helfand and Y. Tagami, J. Chem. Phys., 56, 3592 (1972). 34. E. Helfand and A.-M. Sapse, J. Chem. Phys., 62, 1327 (1975). 35. E. Helfand, J. Chem. Phys., 62, 999 (1975). 36. J. Kressler, Macromolecules, 27, 2448 (1994). 37. S. K. Kumar, Macromolecules, 27, 260 (1994). 38. J. Reiter, G. Zifferer, and O. F. Olaj, Macromolecules, 23, 224 (1990). 39. R. Li and R. Zhang, Chin. J. Polym. Sci. 15, 205 (1997). 40. C. C. Chen and J. I. White, Polym. Eng. Sci., 33, 923 (1993). 41. M. V. Duin and R. D. M. Borggreve, Blends of polyamides and maleic-anhydride-containing polymers: interfacial chemistry and properties, in: Reactive Modifiers for Polymers, S. Al. Malaika (ed.), Chapman and Hall, London, 1997, p. 133. 42. A. Molmar and A. Eisenberg, Polym. Commun., 32, 370 (1991).

Chapter 18 Functionalized Polyolefins and Aliphatic Polyamide Blends

553

43. B. De Roover, J. Devaux, and R. Legras, J. Polym. Sci. A Polym. Chem., 35, 901 (1997). 44. B. De Roover, J. Devaux, and R. Legras, J. Polym. Sci. A Polym. Chem., 35, 1313 (1997). 45. B. De Roover, J. Devaux, and R. Legras, J. Polym. Sci. A Polym. Chem., 35, 917 (1997). 46. S. S. Pesetskii, Izvestia Akademii Nauk Belarusi (Trans. Nat. Acad. Belarus), Chem. Ser., 1, 105 (1992). 47. S. S. Pesetskii, B. Jurkowski, and V. N. Koval, J. Appl. Polym. Sci, 78, 858 (2000). 48. A. A. Tager and V. S. Blinov, Uspekhi khimii (Russ. J. Adv. Chem.), 56, 1004 (1987). 49. S. Yukioka and T. Inoue, Polymer, 35, 1182 (1994). 50. P. Mzrechal, G. Coppens, R. Legras, and J. M. Dekoninck, J. Polym. Sci. A Polym. Chem., 33, 757 (1995). 51. S. S. Pesetskii, B. Jurkowski, Y. M. Krivoguz, and R. Urbanowicz, J. Appl. Polym. Sci., 65, 1493 (1997). 52. S. S. Pesetskii, B. Jurkowski, and O. A. Makarenko, J. Appl. Polym. Sci., 81, 64 (2002). 53. S. S. Pesetskii, B. Jurkowski, Y. M. Krivoguz, and K. Kelar, Polymer, 42, 469–475 (2001). 54. M. Lu, H. Keskkula, and D. R. Paul, Polym. Eng. Sci., 34, 33 (1994). 55. Y. M. Krivoguz, S. S. Pesetskii, and Y. M. Pleskachevskii, Polym. Sci. Ser. A, 46, 698 (2004). 56. S. S. Pesetskii, B. Jurkowski, and Y. M. Krivoguz, J. Appl. Polym. Sci., 92, (2004). 57. D. L. Beltrame, A. Castelli, M. Di Pasguantonio, M. Canetti, and A. Seves, J. Appl. Polym. Sci., 60, 579 (1996). 58. P. Hiefaoya, M. Heino, T. Vainio, and Y. Seppa¨la¨, Polym. Bull., 37, 353 (1996). 59. N. C. Liu and W. E. Bacer, Polymer, 35, 988 (1994). 60. L. Pan, T. Chiba, and T. Inoue, Polymer, 42, 8825 (2001). 61. E. G. Koulouri, A. X. Georgaki, and J. K. Kaltitsis, Polymer, 38, 4185 (1997). 62. K. Y. Park, S. H. Park, and K.-D. Suh, J. Appl. Polym. Sci., 66, 2183 (1997). 63. R. D. Deanin, S. A Orroth, and R. I. Bhagat, Polym.: Plast. Technol. Eng., 29, 289 (1990). 64. B. D. Favis and J. M. Willis, Polym. Sci. B Polym. Phys., 28, 2259 (1990). 65. J. L. White, Rheological behavior of molten polymer blends and particle-filled melts, in: Polymer Compatibility and Incompatibility, Principles and Practices, K. Solc (ed.), Harwood Academic Publishers, Chur, 1980. 66. A. R. Padwa, Polym. Eng. Sci., 32, 22, 1703 (1992). 67. V. J. Triacca, S. Ziace, S. Barlow, H. Keskkula, and D.R Paul, Polymer, 32, 1401 (1991). 68. I. Park, J. W. Barlow, and D. R. Paul, J. Polym. Sci. Phys., B30, 1021 (1992). 69. R. Vankan, P. Dege´e, P. Je´ro´me, and P. Teyssie´, Polym. Bull., 33, 221 (1994). 70. B. De Roover, Etude des Me´lauges Constiue´s de Poly(m-xylyle´ne Adipamide) et de Polypropylene Fonctionnalise Parla´nkydride male´ique, PhD Thesis, Universite´ Catholique de Louvain, 1994. 71. G. V. Vinogradov and A. Ya. Malkin, Rheology of Polymers, Khimia, Moscow, 1977. 72. V. E. Starzhynsky, A. M. Farberov, S. S. Pesetskii, S. A. Osipenko, and V. A. Braginsky, Precise Plastic Articles and Technology for Their Manufacture, Vysshaya Shkola Publ., Minsk, 1992 (in Russian). 73. B. Jurkowski and B. Jurkowska, Manufacturing of Polymer Composites: Elements of Theory and Practice, Wydawnictwa Naukowo-Techniczne, Warsaw, 1995 (in Polish). 74. E. L. Kalinchev and M. B. Sakovtseva, Properties and Processing of Thermoplastics, Khimia, Leningrad, 1983 (in Russian). 75. S. S. Pesetskii, B. Jurkowski, Y. M. Krivoguz, T. Tomczyk, and O. A Makarenko, J. Appl. Polym. Sci. 102, 5095 (2006). 76. V. N. Kuleznev. Polymer Blends, Khimia, Moscow, 1980 (in Russian). 77. D. V. Rosato and D. V. Rosato, Blow Molding Handbook, Hanser Publishers, 1989.

554

Polyolefin Blends

78. C. D. Chang, Rheology in Polymer Processing, Khimia, Moscow, 1979, 438 pp. (in Russian). 79. R. J. Koopmans, Polym. Eng. Sci., 32, 1741 (1992). 80. R. J. Koopmans, Polym. Eng. Sci., 32, 1750 (1992). 81. R. J. Koopmans, Polym. Eng. Sci., 32, 1755 (1992). 82. A. C.-Y. Wong and V. H. K. Cheung, Mater. Process. Technol., 67, 117 (1997). 83. A. C.-Y. Wong and J. Z. Liang, Chem. Eng. Sci., 52, 18, 3219 (1997). 84. A. C.-Y. Wong, Mater. Process. Technol., 79, 163 (1998). 85. A. Bernnar, M. H. Wagner, and C. K. Chai, Int. Polym. Process., 15, 268 (2000). 86. A. H. Dekmezian, W. Weng, C. A. Gargia-Franco, and E. J. Markel, Polymer, 45, 5635, 2004. 87. T. J. Guadarrama-Medina, J. Petez-Gonzalez, and L. Vargas, Rheol. Acta, 44, (2005). 88. D. Acierno, D. Curto, F. P. La Mantia, and A. Valenza, Polym. Eng. Sci., 26, 28 (1986). 89. N. Hadjichristidis, M. Pitsikalis, S. Pispas, and H. Iatrou, Chem. Rev., 101, 377 (2001). 90. A. Ajji, P. Sammut and N. A. Huneault, J. Appl. Polym. Sci., 88, 3070 (2003). 91. J. M. Pian, N. Kissi, F. Toussaint, and A. Mezgmani, Rheol. Acta., 34, 40 (1995). 92. S. Q. Wang, Adv. Polym. Sci., 138, 239 (1999). 93. K. B. Migler, C. Lavallee, M. P. Dillon, S. S. Woods, and C. L. Gettinger, J. Rheol., 45, 565 (2001). 94. S. G. Hatzikiriakos and J. M. Dealy, Int. Polym. Proc., 8, 36 (1993) 95. S. B. Kharchenco, P. M. MeGuiggan, and K. B. Migler, J. Rheol, 47, 1523 (2003). 96. D. Huang and J. L. White, Polym. Eng. Sci., 20, 182 (1980). 97. J. Z. Liang, Plast. Rubber Compos. Process. Appl., 23, 93 (1995). 98. J. Z. Liang, Plast. Rubber Compos. Process. Appl., 15, 75 (1991). 99. Y. M. Krivoguz and S. S. Pesetskii, J. Appl. Polym. Chem. (Russ. Acad. Sci)., 78, 310 (2005) (in Rusian). 100. L. A. Utracki, Z. Bakerdjian, and M. R. Kamal, J. Appl. Polym. Sci., 19, 48 (1975). 101. S. S. Pesetskii, B. Jurkowski, O. A. Makarenko, and A. A. Bogoslavsky, Polyfunctional compatibilized blends based on polyamide 6 and functional polyolefins, in: World Polymer Congress IUPAC MACRO 2000, 38th Macromolecular Symposium, Poland, Warsaw, 2000, p. 1197. 102. I. Campoy, J. M. Arribas, M. A. M. Zaporta, C. Marco, M. A. Gomez, and J. G. Fatou, Eur. Polym. J., 31, 475 (1995). 103. M. Psarski, M. Pracella, and A. Galeski, Polymer, 41, 4923 (2000). 104. G.-H. B. Wang and X. Zhon, Mater. Lett., 58. 3457 (2004).) 105. L. Minkova, Hr. Yordanov, and S. Filippi, Polymer, 43, 6195 (2002). 106. A. Hallden, M.-J. Deriss, and B. Wasslen, Polymer, 42, 8743 (2001). 107. S.-C. Wong, Y.-W. Mai, Polymer, 40, 1553 (1999). 108. S. S. Pesetskii, A. A. Bogoslavsky, B. Jurkowski, and O. A. Makarenko, in: Proceedings of VII Scientific Conference Rydzyna—1998. Poland, 1998, p. 231. 109. S. S. Pesetskii, A. A. Bogoslavsky, and O. A. Makarenko, Investigation of high viscous blends of polyamide 6 with grafted polymers and olefin copolymers, in: Polymer Processing Society, PPS-14, Yukohama, Japan, 1998, p. 15. 110. S. S. Pesetskii, B. Jurkowski, V. N. Koval, and I. P. Storoszuk, J. Appl. Polym. Sci., 73, 1823 (1999). 111. S. S. Pesetskii, B. Jurkowski, and V. N. Koval, J. Appl. Polym. Sci., 84, 1277, (2002). 112. T. K. Tsotsis, Polym. Compos., 17, 362 (1996). 113. M. Abbate, V. Di Liello, E. Martuscelli, P. Musto, G. Ragosta, and G. Searinzi Polymer, 33, 14, 2940 (1992). 114. L. L. Ban, M. J. Doyle, M. M. Disko, and G. R. Smith, Polym. Commun., 29, 163. (1988). 115. L. D’Orazio, C. Mancarella, and E. Martuscelli, J. Mater. Sci., 23, 161 (1988).

Chapter 18 Functionalized Polyolefins and Aliphatic Polyamide Blends

555

116. G. Bline, H. G. Dorst, and L. C. Plachetta, Kunststoffe, 78, 7, 612 (1988). 117. A. J. Oshinski, M. Keskkula, and D. R. Paul, Polymer, 33, 268 (1992). 118. A. J. Oshinski, M. Keskkula, and D. R. Paul, Polymer, 33, 2, 284 (1992). 119. S. Cartasegna and W. Heider, Gummi Fasern Kunststoffe. 41, 3, 110, 112, 117, 118 (1988). 120. T. Ohmae, Y. Toyoshima, K. Mashita, N. Yamguchi, and J. Nambu, U.S. Patent 5,010,136 (1991). 121. T. Inoue, U.S. Patent 5,310,792 (1994). 122. W. Nielinger, C. Linder, U. Grigo, R. Binsack, F. Fahnler, and B. Brassat, GFR Patent 3,025,606 (1982). 123. G. Plachetta, G. E. Mckee, H. Reimann, and R. Pflu¨ger, GFR Patent 3,842,618 (1990). 124. O. Nahodima and M. Idzawa, Japan Patent 61–204,267 (1986). 125. R. Gelles, W. P. Gergen, and R. G. Lutz, U.S. Patent 5,371,141 (1994). 126. J. Gerecke, K. Gruber, and U. Schu¨tz, GDR Patent 296,947 (1991). 127. G. D. Mason, J. A. Joung, and Y. C. Haylock, U.S. Patent 4,945,129 (1990). 128. Y. Ota and O. Nikadzima, Japan Patent 2–184,49 (1990). 129. H. C. Silvis, D. J. Murray, T. R. Fiske, S. R. Batso, and R. R. Turley, U.S. Patent 5,681,997 (1997). 130. P. M. Subramanian, World Patent 90/075,56 (1990). 131. P. M. Subramanian, U.S. Patent 5,126,407 (1992). 132. P. M. Subramanian, Can. Patent 1,335,219 (1995). 133. T. Ohmae, Y. Toyoshima, K. Mastia, and N. Yamaguchi, U.S. Patent 5,155,159 (1992). 134. J. P. Khanna, R. Kumar, and J. Sibilia, U.S. Patent 5,403,896 (1995). 135. V. Heppert, Pat. Application, Germany Patent 19,525,198 (1996). 136. H. R. Bhattacharjee and J. P. Khanna, U.S. Patent 4,906,708 (1990). 137. S. S. Pesetskii, S. A. Leonov, V. N. Koval, and Y. M. Krivoguz, Russ. Patent 2,278,787 (2006). 138. E. Kh. Ziganshina, S. A. Leonov, Y. A. Kutilin, S. S. Pesetskii, S. A. Lunin, I. P. Aizinson, and E. I. Shurshalina, in: Composite Materials in Industry: 24th Scientific Conference ‘‘Slavpolicom’’, Yalta, 2004, p. 262.

Chapter

19

Plastic Deformation and Damage Mechanisms of Ternary PP/PA6/POE Polymer Blends Shu-Lin Bai,1 Christian G’Sell,2 Gong-Tao Wang,3 Jean-Marie Hiver,2 and Min Wang1

19.1 INTRODUCTION The ternary polymer blends based on PP, PA6 (or PA66), and elastomers have been studied worldwide during last several decades. Many researchers tried to well understand the compatibility, morphology, and mechanical behavior of such ternary polymer blends. The blend systems studied include PP/PA66/SEBS-g-MA (1–5), PP/PA6/EPR-g-MA (3–5), PP/PA6/PP-g-MA (6), PP/PA6/EPDM-g-MA (7), and so on. Since several years, a research group has begun research work on PP/PA6/ POE-g-MA blends (8–12). It is well known that neat polypropylene (PP) is incompatible with neat polyamide. Therefore, in order to improve the compatibility between them, the maleated polypropylene is often added into the blends, which has proved to be very effective. Other attempts concentrate mainly on the improvement of impact toughness, which is generally realized by adding elastomer into the blends. The elastomer is either 1 Centre for Advanced Composite Materials (CACM), Department of Advanced Materials and Nanotechnology, College of Engineering, Peking University, 100871 Beijing, China 2 Laboratoire de Physique des Mate´riaux, Ecole des Mines de Nancy, Parc de Saurupt, 54042 Nancy Cedex, France 3 School of Aerospace, Mechanical & Mechatronic Engineering, The University of Sydney, Sydney, NSW 2006, Australia

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

556

Chapter 19

Plastic Deformation and Damage Mechanisms

557

independently dispersed in the matrix in the form of spherical phase, or encapsulates the dispersed PA6 or PP phase forming core–shell structures. The morphology of such blends with three constituents is complicated and depends on the proportion of the constituents. The continuous matrix phase is that with maximum content, and the second and third phases distribute in the matrix in the form of spherical or ellipsoidal shape. If the content of second phase is close to the first one, it may exist in irregular shape and sometimes contain small-sized matrix phase inside it. A transition of first to second phase will occur if the order of content is inversed. The study on the mechanical behavior of PP/PA blends was conducted principally on the reinforcing and toughening effects. A domain that is not explored widely and deeply is the plastic deformation and corresponding damage mechanisms of polymer blends. Early works on the necking phenomenon in tensile tests are extensively documented in the literature (13). The plastic deformation can be produced by two mechanisms: one is shear banding and another is cavitation. The latter results in a volume expansion of the samples deformed. Therefore, the volume change is considered to be a phenomenon accompanied with the plastic deformation and damage in particle-filled polymer or polymer blends. Bucknall et al. (14) have studied the dilatation of high impact polysterene (HIPS) under creep tests by measuring the dilatation with a special extensometer. They found that crazing is an important form of damage in brittle matrix blends. Gent and Park (15) revealed the process of cavity formation for a rigid inclusion-filled elastomer under uniaxial tension. G’Sell et al. (16) working on a rubber-toughened polymethylmethacrylate (PMMA) showed that large damage rate was attributed to the formation and the growth of voids in the rubber shell of the toughening particles. Another damage mechanism concerns the rigid particle-filled polymer, such as glass bead or CaCO3-filled polypropylene (PP) and high density polyethylene (HDPE), and so on. The decohesion at particle/matrix interface was observed and the decohesion stress became smaller with the decreasing bonding strength at the interface (17). Meddad and Fisa (18) showed that PP/glass bead system undergoes a high damage rate due to the decohesion of the beads. They found damage rate of 0.75 for a typical concentration of beads equal to 40 vol%. The micrographs obtained by scanning electron microscopy (SEM) after cryofracture unambiguously show the bare surface of the debonded beads. Pukanszky et al. (19) studied the CaCO3-filled polypropylene. They found that the dominating deformation process is debonding at PP/CaCO3 interface, and the initiation stress for void formation is close to yield stress of the material. The interface debonding was also observed for PP/PA6/POE-g-MA blends in our previous investigation (8). Many contributions to plastic deformation of polymer blends were fulfilled in condition of monotonic loading. The information obtained represents the final state of plastic deformation, so the deformation process and corresponding microstructural evolution are unknown. The accumulation of plastic damage under cyclic or fatigue load is seldom studied. Recently, Meyer and Pruitt (20) studied the cyclic loading effect on the morphology, structure, and relaxation of ultrahigh molecular weight polyethylene. The crazing, microvoid formation, and lamellae alignment were found to vary with increasing amount of true strain and number of cycles. The

558

Polyolefin Blends

plastic deformation and damage behavior of PP/PA6/POE-g-MA blends under cyclic tension was investigated by Bai and Wang (10) by measuring the volume variation and revealing the debonding and cavitation mechanisms. In another study, Bai et al. (21) found that for glass beads-filled HDPE the damage under cyclic loading– unloading tests can be estimated with the decreasing amount of elastic modulus. A model to the elastoplastic response under loading–unloading tests was proposed by Drozdov and Christiansen (22). Isotactic polypropylene was used as a target material. The rate of plastic strain was found to increase linearly with the maximum plastic strain, which was caused by the coarse slip and the fragmentation of lamellae at unloading. In this work, the plastic deformation and the damage of ternary PP/PA6/POE were studied as completely as possible, and the mechanisms of plastic deformation were revealed by microscopic observation. The results obtained are original and will be helpful to well understand the mechanical behavior of polymer blends under large deformation.

19.2 MATERIALS PRESENTATION AND EXPERIMENTAL METHODS 19.2.1 Materials Presentation The compositions of the PP/PA6/POE blends under investigation are listed in Table 19.1. The total alloying (PA6 þ POE) varies from 0 wt% (noted PP for short) to 60 wt% (noted BD16). In all materials, the percentage of PA6 is twice as large as the percentage of POE-g-MA. The polypropylene was obtained from Liaoyang Petrochemical Corp., P. R. China (ref. 401), and the polyamide 6 (PA6) from Shanghai Plastics Production Factory No. 18, China. The polyethylene–octene (POE) modifier, obtained from Dow Chemical Co (octene content of 9.5% and melt flow index of 3.5 g/10 min), was grafted in the laboratory with maleic anhydride at a ratio of about 1% in weight. All materials were dry blended together in a high speed blender following the predesigned composition ratios. Then a corotating twin-screw extruder (SHJ-30, diameter 30 mm) was employed at a screw speed of 110 rpm and barrel temperatures Table 19.1

Compositions of the Blends Studied.

Blend

PP, wt%

PA6, wt%

POE-g-MA, wt%

Total alloying, wt%

PP BD13 BD14 BD15 BD16

100 85 70 55 40

0 10 20 30 40

0 5 10 15 20

0 15 30 45 60

[From Reference 9 with permission from Elsevier.]

Chapter 19

Plastic Deformation and Damage Mechanisms

559

Figure 19.1 Method to cut sections for SEM observation. (From Reference 9 with permission from Elsevier Inc.)

of 190–200–210–200 C. The pelletized blends were dried and injection molded into standard ASTM specimens in an injection molding machine (SZ-160/80 NB) for mechanical tests.

19.2.2 Morphological Study under SEM and TEM The morphological study includes the intrinsic microstructures of the blends and the morphological characteristics of the fractured surfaces after mechanical tests. For the first objective, the cryofractured surfaces were prepared and then observed with an Amary-1910FE scanning electron microscope. In order to characterize the different phases present in the blends, one series of cryofractured surfaces was etched with formic acid for 24 h to remove the PA6 phase. The surface was then coated with gold. The size distribution of the PA6 phase in blends was determined by measurements of approximately 300 domains from sets of cryofractured SEM micrographs. The observation of fractured surfaces after mechanical tests was carried out with the same SEM. Figure 19.1 shows the method to cut the tensile samples in view of obtaining cross sections and longitudinal sections for SEM observation. Moreover, a Hitachi H-800 transmission electron microscope (TEM) was also used to reveal the details of dispersed phases and the interface. Ultrathin sections having minimum thickness of 60 nm were cut using a LKB Ultratome V ultramicrotome under the condition of sample temperature 80 C and knife temperature 70 C. The sections were stained with RuO4 vapor for 20 min and OsO4 for 40 min, respectively, in order to enhance the contrast.

19.2.3 Video-controlled Tensile System In most polymers, a marked necking phenomenon occurs very early after the yield point. This is the reason why it is not possible, in the range of large deformations, to determine strains in a large representative volume element (RVE). Consequently none of the dilatometers utilized to date can be used, except in a very restricted strain range. The latter statement concerns dual clip gage extensometers (axial þ transversal) and also liquid displacement dilatometers (23). Once plastic instability has

560

Polyolefin Blends

Figure 19.2 Configuration of the seven markers in the video-controlled tensile testing system: (a) schematic diagram and (b) micrograph of a sample deformed. (From Reference 10 with permission from Elsevier).

been initiated, and provided strain localization is axisymmetric (no oblique shear band), the RVE should be chosen as a very thin slice located at the center of the neck (Fig. 19.2a). Although a maximum of hydrostatic pressure develops in the core of the specimen, the distribution of strains is nearly homogeneous in this slice, as shown by early models (24) and recent finite-element simulations (25). The method utilized for determining the axial and transverse components of the strain tensor in the above-specified RVE has been patented recently (26) as a novel option of the video-controlled materials testing system, Vide´oTractionß, designed by two of the authors (C.G’s. and J.M.H.) and licensed to the Apollor Co. (Vandoeuvre, France). We recall below briefly the main features of the method. In order to localize the necking process at a predetermined location in the center of the parallelepiped specimen, a small geometric defect was machined on both lateral faces in the center of the calibrated length over a length of 10 mm, corresponding to a local reduction in width with a rounded profile. In this case, the minimum cross section is a rectangle of 8 4 mm2. Seven markers are printed on the front surfaces of the sample prior to deformation. In the present case the markers are black, nearly round, with a diameter of about 0.4 mm. The diagram in Fig. 2b illustrates their configuration after necking has developed. The five markers (A, B, C, D, and E) aligned along the tensile axis x3 are at relative distances of about 1 mm. The three markers (F, C, G) aligned along the transverse axis x1, are more widely separated, in such a way that they occupy a major fraction of the total width of the specimen. They are situated as close as possible to the plane of minimum cross section. The marker set is optically investigated in real time by a video camera interfaced with a microcomputer (Fig. 19.3) according to specifications indicated elsewhere (23,26). The in-plane transverse true strain e1 is

Chapter 19

Plastic Deformation and Damage Mechanisms

561

Figure 19.3 General diagram of the Vide´oTractionß system. (From Reference 11 with permission from Elsevier Inc.)

automatically computed from the distance between the centers of gravity of markers F and G with reference to their initial distance. The through-thickness transverse true strain e2 was checked to be equal to e1 (equal thickness and width reduction at the end of the test) due to the transverse isotropy of the specimen. As for the axial true strain e3, it is obtained from the doublets AB, BC, CD, and DE, the four corresponding values giving access through polynomial interpolation to the axial strain at the RVE. Finally, volume strain at the RVE is simply obtained by ev ¼ 2e1 þ e3. In its present state of development, the Vide´oTractionß system is capable of controlling a tensile test with full analysis of at least 50 images per second. The final precision is about 2% on axial stress, 0.5% on volume strain, and 0.1% on axial strain (16). Tests can be run at any temperature compatible with dot painting resistance and thermal radiation perturbations (typically up to about 250 C). In order to reveal the mechanisms of plastic damage, a series of monotonic tensile tests were undertaken with a small universal testing stage that had the maximum load of 2 kN. The loading was interrupted when the true strain reached a given value, which was chosen as 0.2–1.2 with the increment of 0.2.

19.3

MICROSTRUCTURE

In Figure 19.4 are shown the SEM micrographs of cryofractured cross sections etched with formic acid for the four blends. The large number of pits (or holes) observed correspond to PA6 particles removed by the formic acid etching. Since formic acid does not react with POE, the PA6 particles coated with POE should not be removed, and the observed pits are presumably of three types: (i) PA6 particles not coated with POE, (ii) PA6 particles partially coated with POE, and (iii) PA6 particles

562

Polyolefin Blends

Figure 19.4 Cross sections etched with formic acid (a) BD13, (b) BD14, (c) BD15, and (d) BD16 (all black pits are left by PA6 particles). (From Reference 9 with permission from Elsevier Inc.)

with debonded PA6/POE interface. It is known that the succinic anhydride group of the maleic anhydride grafted POE is able to react with PA6 amine terminal groups to form POE-co-PA6 copolymer that strongly tends to concentrate at the PP/PA6 interfaces during melt processing. The histograms in Fig. 19.5 give an approximately normal distribution of PA6 particle diameters and the graph in Fig. 19.6 shows the influence of total alloying on the ‘‘peak diameter’’ (most probable diameter in the distributions). The most striking feature is that the particles with highest peak diameter are found in the blend with 15% alloying (BD13), while particles are 70% finer in the blends with higher alloying contents. Also, it is seen that the diameter distribution is narrower in the blends at 15% and 30% alloying contents and wider in the other two blends. Subsequently, the blend BD14 (20% PA6 þ 10% POE) combines the finest PA6 particles (peak diameter about 0.4 mm) and the narrowest size distribution (diameter within the range from 0.1 to 1.1 mm). The SEM micrographs in Fig. 19.7 show the morphology of longitudinal sections of the blends that are subsequently etched with formic acid. Since the exposed surface is parallel to the flow direction of the injection-molded samples, the micrographs are likely to reveal some process-induced orientation in the morphology.

Chapter 19

Plastic Deformation and Damage Mechanisms

563

Figure 19.5 PA6 particle-sized distribution. (a) BD13, (b) BD14, (c) BD15, and (d) BD16 (From Reference 9 with permission from Elsevier).

Figure 19.6 Mean diameter of PA6 particles versus total alloying. (From Reference 9 with permission from Elsevier Inc.)

564

Polyolefin Blends

Figure 19.7 Longitudinal section etched with formic acid. (a) BD13, (b) BD14, (c) BD15, and (d) BD16. (From Reference 9 with permission from Elsevier Inc.)

Actually, one notes in Fig. 19.7 that many particles exhibit ellipsoidal shapes, whose major axis is parallel to the injection direction. The aspect ratio of the ellipsoids is variable from one specimen to another. While most particles are quasispherical in blends BD13 and BD14, they are markedly elongated in BD15 and BD16. The series of TEM micrographs in Fig. 19.8 shows longitudinal sections stained with RuO4 at medium magnification. RuO4 stains all phases in the blend but with different colors. The light grayish and continuous phase is the PP matrix. As for the large black ellipsoids, they correspond to PA6 particles. The TEM observation confirms that their aspect ratio depends on the alloying content: the most regular spheres are observed in BD14 and the most elongated ellipsoids in BD15 and BD16. The magnification of the TEM micrographs being 10 times larger than that of the SEM images of Fig. 19.7, finer particles are now distinguishable with a size smaller than 0.2 mm. A few ones are black and correspond to the lower wing of the size histogram of PA6 particles. Other ones, more grayish, are identified as isolated POE particles. The above investigation confirms that the BD14 shows the most regular particle morphology, in terms of size, distribution, and isotropy, among the blends of the PP/PA6/POE series. Another feature revealed by the TEM micrographs is the

Chapter 19

Plastic Deformation and Damage Mechanisms

565

Figure 19.8 TEM micrographs of the sections stained with RuO4 (a) BD13, (b) BD14, (c) BD15, (d) BD16. (From Reference 9 with permission from Elsevier Inc.)

presence of bright inclusions in the interior of the large PA6 particles in the BD15 and BD16 blends. Owing to their similar contrast with the PP matrix, these species are considered to be the PP particles trapped within the PA6 like in ‘‘salami’’ morphology. If the PA6 content is further increased, the morphology should become a continuous PA6 matrix with embedded PP particles. The big and irregular PA6 ‘‘patches’’ in BD16 announce this transition, but in this blend the PP phase is still continuous. The TEM micrographs in Fig. 19.9 were obtained at higher magnification. Here the staining agent is OsO4; since it cannot react with PP and PA6, only POE phases are stained in dark. The POE interlayer around the PA6 particles appears as a thin and black circle. The thickness of POE interlayer increases from 10 to 100 nm as alloying content is increased.

566

Polyolefin Blends

Figure 19.9 TEM micrographs of the sections stained with OsO4 (a) BD13, (b) BD14, (c) BD15, (d) BD16. (From Reference 9 with permission from Elsevier Inc.)

19.4 GENERAL MECHANICAL PROPERTIES 19.4.1 Dynamic Mechanical Thermal Analysis The curves in Fig. 19.10a and b, obtained by dynamic mechanical thermal analysis, show the variation of storage modulus and loss tangent for PP and the blends as a function of temperature. It is noted that blending PP with PA6 and POE has small but significant influence on viscoelastic properties. The storage modulus decreases with the alloying content, while the loss tangent increases. Two distinct transition temperatures are recorded for PP: one at about 18 C that corresponds to the b-transition and the other at about 80 C representing the a-relaxation. Most blends have two peaks just as PP (with greater tan d owing to the presence of POE), except for BD16 that does not show any b-transition. This phenomenon is possibly related to the

Chapter 19

Plastic Deformation and Damage Mechanisms

567

Figure 19.10 (a) Storage modulus E’ of PP and blends versus temperature and (b) loss tangent tan d of PP and blends versus temperature. (From Reference 9 with permission from Elsevier Inc.)

balanced concentration in PA6 and PP. Since PA6 has a higher transition temperature than PP (about 60 C), the concurrence between PA6 and PP is responsible for the disappearance of PP peak at b-transition for BD16 that has equal content of PP and PA6. Additionally, the blends show a low temperature peak at 50 C, which corresponds to the glass transition of POE.

19.4.2 Toughness by Impact Loading and Yield Stress by Tension The most interesting property of the PP/PA6/POE blends is their improved resistance to impact. The graph in Fig. 19.11 gives the notched Izod impact strength as a

568

Polyolefin Blends

Figure 19.11 Notched Izod impact strength and yielding of PP and blends. (From Reference 9 with permission from Elsevier Inc.)

function of alloying content. It is seen that the impact strength of the blends increases considerably with PA6/POE content, reaching a value of 158 m1 for the BD16 blend, which 4.7 times greater than that of PP. These data contrast with the results obtained previously (8), which showed that (i) the addition of PA6 to PP merely results in a small increase of impact strength and, (ii) the use of PP-g-MA as a compatibilizer causes a decrease of the impact strength. Therefore, the large toughening observed in the systems investigated in this work is certainly related to the effect of the POE compatibilizer. Concerning the yield stress, it follows a slight decrease with total alloying content as shown in Fig. 19.11. In fact, the combined influence of PA6 and POE can greatly change the behavior of PP. It could be envisaged that PA6 and POE particles have opposite effects on yield stress. However, yield stress is also controlled by the characteristics of particles (volume fraction, average diameter, size distribution, interfacial adhesion, etc.) SEM micrographs of impact-fractured surfaces of blends are shown in Fig. 19.12. The spherical particles are mainly PA6 and the black pits correspond to sites where PA6 particles were extracted from the PP matrix. Less and less PA6 particles are observed on the fractured surface as the PA6 content increases. The interlayer thickness is rather small when the POE content is as low as 5–10%, so that failure probably occurs at the PA6/POE interface in such a way that the PA6 surface is exposed. When the POE content increases, the interlayer becomes thicker and the failure occurs consequently at the PP/POE interface. This means that PA6 particles remain covered by POE after failure. On the basis of the experimental facts reported above, it is now possible to describe, at least schematically, the essential toughening mechanisms of the PP/PA6/ POE blends. The dissipation of impact energy in the blends is probably due to the following factors: (i) the isolated elastomer particles play a small but significant role

Chapter 19

Plastic Deformation and Damage Mechanisms

569

Figure 19.12 Micrographs of impact-fractured surfaces of blends. (a) BD13, (b) BD14, (c) BD15, and (d) BD16. (From Reference 9 with permission from Elsevier Inc.)

in either arresting the cracks or at least reducing their propagation rate, (ii) the high adhesion of the interfacial layer avoids early decohesion at the POE interphase between the PP matrix and the PA6 particles, and is later capable of cavitation, and (iii) the ellipsoidal geometry of the PA6 particles improves somehow the impact resistance thanks to its favorable orientation perpendicular to the crack propagation direction. Figure 19.13 gives a schematic representation of toughening mechanisms in the blends. For blends with low alloying content, such as BD13 and BD14, the POE interlayer is very thin and PA6 particles are mostly spherical. The crack propagates easily across the section and leaves a smooth fractured surface as shown by the micrographs in Fig. 19.13a and b. By contrast, for blends with high alloying content, such as BD15 and BD16, the POE interlayer is thick and PA6 particles are highly elongated, which result in rough fractured surfaces as shown by Fig. 19.13c and d.

19.5 PLASTIC DEFORMATION UNDER UNIAXIAL TENSION 19.5.1 Definition of Volume Strain In the case of uniaxial tension, early authors introduced the ‘‘nominal’’ definition of strains (fully Lagrangian strains) for describing elastic deformation (27),

570

Polyolefin Blends

Figure 19.13 Failure mechanisms under impact loading of the blends (a) low alloying content with thin POE interlayer and (b) high alloying content with thick POE interlayer. (from Reference 9 with permission from Elsevier Inc.)

subsequently applied to large strains and testing standards (28). As such, nominal strains are written as eN 1 ¼ ðL1 L10 Þ=L10 eN 2 ¼ ðL2 L20 Þ=L20 eN 3

ð19:1Þ

¼ ðL3 L30 Þ=L30

where L10 ; L20 ; L30 and L1 ; L2 ; L3 represent the initial and current dimensions of the RVE under investigation in the tensile sample. Here axis x3 is the tensile direction, x1 the direction across width, and x2 the direction across thickness. According to the above definitions, volume strain ev is related to the dilatation (or contraction) by the following equation: N N N eN v ¼ ð1 þ e1 Þð1 þ e2 Þð1 þ e3 Þ 1 ¼ ðV V0 Þ=V0

ð19:2Þ

where V0 and V are the initial and the current volume of the RVE. The main problem with nominal strains is that they do not fulfill associativity. For successive

Chapter 19

Plastic Deformation and Damage Mechanisms

571

deformation steps from state S1 to state S2 and then from state S2 to state S3, a total nominal strain increment DeN i ðS1 ! S3 Þ is not equal to the sum of partial strain N ðS ! S Þ þ De increments DeN 1 2 i i ðS2 ! S3 Þ. The Hencky definition of strains (also called ‘‘true’’ or ‘‘natural’’ or ‘‘logarithmic’’) was used by G’Sell et al. (29,30) to alleviate the above difficulty. It gives the expressions e1 ¼ lnðL1 =L10 Þ; e2 ¼ lnðL2 =L20 Þ, and e3 ¼ lnðL3 =L30 Þ for the three principal strains. Volume strain can also be written in Hencky formalism as ev ¼ e1 þ e2 þ e3 ¼ lnðV=V0 Þ

ð19:3Þ

where V and V0 concern the same RVE as previously. The strain components and the volume strain are associative. The two definitions introduced above are obviously equivalent at infinitesimal deformation, but their difference increases rapidly with strain. For eN 3 ¼ 20%, the relative discrepancy is already as large as 9%. Consequently, Hencky definitions should always be employed for describing the stretching properties of polymers. N In the elastic range, the Poisson’s ratio, nel ¼ eN 1 =e3 , can be used to relate the amplitude of transverse constriction with axial strain. The volume strain depends on N this material coefficient by eN v ¼ ð1 2nel Þe3 . Similarly, for nonlinear materials, the ‘‘tangent Poisson’s ratio’’ nT ¼ de1 =de3 may be introduced at large strain, so that coefficient is also obtained from the instantaneous slope of the volume strain versus axial strain from: ð1 2nT Þ ¼ dev =de3 . In this paper, the latter slope will be called ‘‘dilatation rate’’(or ‘‘damage rate’’) and denoted by the variable D. By adapting an analysis of Bucknall (31) to the Hencky formalism, it was recently shown by one of the authors (32) that volume strain could be usefully decomposed into three contributions: pl ca ev ¼ eel v þ ev þ ev

ð19:4Þ

The first term represents the elastic component that is related to Poisson’s effect by el el the equation eel v ¼ ð1 2nel Þe3 introduced above, where e3 is the elastic component , corresponding to plastic shear, is usually of axial true strain. The second term, epl v considered to be zero in metals, but we showed elsewhere (32) that it can be slightly negative in some polymers, due to the compaction of macromolecular chains subjected to strain-induced orientation. The last term eca v , measures the contribution of cavitation and/or crazing to the macroscopic volume change of the tensile specimen (33).

19.5.2 True Axial Stress–Strain Relation The true stress versus true strain curves obtained at room temperature under uniaxial tension with all the materials are displayed in Fig. 19.14. By means of the videocontrolled system, the true strain rate was regulated at a constant value of 103 s1 , until specimen became stretched in the range from e3 0:9ðeN 3 146%Þ for PA6 up 395%Þ for BD13. to e3 1:6ðeN 3

572

Polyolefin Blends

Figure 19.14 Tensile true stress–true strain curves at room temperature for a constant true strain rate of 103 s1 . (From Reference 11 with permission from Elsevier Inc.)

After the yield point at sy ¼ 29 MPa, the neat PP (51 wt% crystalline in this case) exhibits a small but significant true stress drop followed, for strain larger than about 0.3, by a progressive (although moderate) hardening. The first result concerning the PP/PA6/POE blends is the weak dependence of the yield stress on the alloying content. Fig. 19.14 shows that, with respect to neat PP, yield stress decreases by only 17% for the BD16 blend. Furthermore, the plot of Figure 19.14 displays the true yield strain and corresponding volume strain at yield point. It is noted that both decrease on increasing alloying content. This indicates that the onset of plastic deformation of blends is earlier for higher alloying contents. The above results are presumably due to the lower elastic energy required to activate plastic shear nuclei in the vicinity of the POE elastomer phase.

Figure 19.15 Yield strain of PP and blends versus alloying content. (From Reference 11 with permission from Elsevier Inc.)

Chapter 19

Plastic Deformation and Damage Mechanisms

573

As for the neat PA6, it experiences a characteristic double yield (probably due to residual humidity) and starts its steady-state plastic regime at an upper yield stress of about 50 MPa. The shape of the stress–strain curve in the steady-state plastic regime (Fig. 19.14) is seen to be very sensitive to material composition. Neat PP exhibits moderate softening after yield in a strain range of about 0.5, before the steep hardening stage that continues up to rupture. It is interesting to remark that, apart from the BD13 that appears somewhat singular, the blends show early hardening. In the BD16 blend, for example, hardening begins as soon as yield point is passed and increases gradually up to rupture. It is worthy to note the very high hardening of PA6 at large strains, so that flow stress attains nearly 100 MPa at failure. This feature is probably the major cause of the hardening increase in the PP/PA6/POE blends. Whatever the aspect of the stress–strain curves, it was observed that all specimens did undergo necking during their stretching. Only was it noticed that the neck shape was less abrupt for the materials with an early and rapid hardening, as predicted by the classical models of plastic instability in uniaxial tension (13). It should be noticed, however, that the softening/hardening behavior displayed in Fig. 19.15 is not a mere artifact caused by necking, since stress and strain are measured within the center of the neck and local strain rate is regulated at a strictly constant value. The features of the curves presented in this section really reflect the intrinsic behavior of the materials.

19.5.3 Volume Strain The variations of volume strain with axial true strain are displayed in Fig. 19.16 for the different materials. It is evident from this graph that nonisochoric effects are very

Figure 19.16 Volume strain versus true axial strain. (From Reference 11 with permission from Elsevier Inc.)

574

Polyolefin Blends Table 19.2 Poisson’s Ratio in the Elastic Stage [From Reference 11 with permission from Elsevier.) Materials PP BD13 BD14 BD15 BD16 PA6

Poisson’s ratio, nel 0.35 0.38 0.40 0.38 0.40 0.40

1.4% 5.2% 1.6% 5.5% 5.6% 6.0%

important for all materials, ev being typically in a range between 0.1 and 0.5 for e3 ¼ 1.0. Also, it is interesting to remark that not only each curve shows different stages for successive strain ranges, but also the amplitude of volume strain differs greatly according to materials. We will now analyze these variations in further details. The elastic portion of the curves was investigated from the recorded data in order to determine precisely the initial volume strain variations. Since we showed that lime3 !0 ðdev =de3 Þ ¼ ð1 2nel Þ, the initial Poisson’s ratio nel could be determined with a precision of a few percent. The values thus obtained are displayed in Table 19.2. It is seen, by comparison with neat PP, that the elastic Poisson’s ratio of the blends increases slightly with the alloying content. This is probably due: (i) to the contribution of PA6 that exhibits a Poisson’s ratio significantly higher than that of PP and (ii) to the contribution of the elastomer (POE) for which nel is near to 0.5 like most rubber-like materials. After the yield point is passed, most curves in Fig. 19.16 show a nearly immediate increase of the volume strain. Conversely, to neat PP that exhibits continuous increase in volume strain (although this evolution tends to saturate), all the blends, at the exception of BD13, show a marked decrease in the volume strain after a rounded maximum at a given strain threshold. Conversely, the volume strain of neat PA6 shows a monotonous and relatively small increase up to ev 0:2 for e3 ¼ 1:0, while it reaches ev 0:5 for PP. The plots in Fig. 19.17 display the evolution of the damage rate D ¼ (dev / de3) versus e3 for PP and for the blends. From the definition, the damage rate represents, in fact, the rate of volume dilatation per unit strain applied. The damage rate or volume dilatation rate first increases with the applied strain. As soon as the maximum value of damage rate is reached, the volume dilates more and more slowly even though the volume continues to expand. One can check on this graph that D presents a transient increase at small strains with a maximum at 0.1–0.25 according to materials, during which D increases to a maximum value. This peak in D is very much influenced by the microstructure, since it reaches nearly 0.7 for PP or BD13, and only 0.4 for BD16. The strain at which the damage rate becomes negative decreases with the alloying content. For example, D becomes null at e3 ¼ 1.2 for the BD14 blend and at e3 ¼ 0.9 for the BD16.

Chapter 19

Plastic Deformation and Damage Mechanisms

575

Figure 19.17 Evolution of damage rate versus true axial strain. (From Reference 11 with permission from Elsevier Inc.)

19.5.4 Under Cyclic Tension 19.5.4.1 True Axial Stress versus Strain Relation The cyclic tensile curves of studied materials are shown in Fig. 19.18. A significant decrease of yield stress with alloying content is also remarked. The strain softening phenomenon became less important with increasing alloying content, even disappeared for the blends of BD15 and BD16 with high PA6 or POE content (PA6 ¼ 30–40 wt% and POE ¼ 15–20 wt%). In fact, the yielding point for BD16 is difficult to recognize because the strain hardening has taken place

Figure 19.18 True stress–strain relationship under cyclic tension. (From Reference 10 with permission from Elsevier Inc.)

576

Polyolefin Blends

Figure 19.19 Residual strain versus applied true strain for each cycle. (From Reference 10 with permission from Elsevier Inc.)

immediately after the yielding without the occurrence of strain softening. It is known that the strain softening of a material is often caused by the nucleation and the growth of microvoids, while the strain hardening is produced by massive shear banding of the matrix. This means that the mechanisms of plastic deformation or damage for BD13 and BD14 differ from that of BD15 and BD16. In addition, true stress–strain curve of neat PP is located in the middle of all the five curves, which seems to give the information that the deformation mechanisms of neat PP are probably controlled by both microvoids formation and matrix shear banding. The hysteresis loops are similar in shape for all materials. The viscous characteristics of the blends are manifested by the nonlinear unloading curves. After each unloading to zero load, the strain does not return to zero. The residual strain, just after unloading, depends on the level of applied strain as shown in Fig. 19.19. The difference of residual strain among the blends is small. 19.5.4.2 Volume Strain The volume strain versus axial true strain curves are displayed in Fig. 19.20 for all blends and neat polypropylene. On the whole, the volume strain increases progressively with the applied true strain, but a saturation of volume strain is reached. The saturation strain, that is, applied true strain at the saturation of volume strain seems to decrease with the alloying content, from about 1.3 for BD13 to 0.8 for BD16. Connecting together the saturation points of all materials, a straight line can be drawn as indicated in Fig. 19.20. It is noted that, at a given axial strain, the volume strain decreases with the alloying content. The volume strain of neat PP has the same trend of variation as with the blends but drops among the values for the blends. This phenomenon corresponds well to the tensile curves in Fig. 19.18. During the unloading process, the volume strain decreases. The quantity of recovered volume

Chapter 19

Plastic Deformation and Damage Mechanisms

577

Figure 19.20 Volume strain versus true strain under cyclic tension. (From Reference 10 with permission from Elsevier Inc.)

strain increases with the cycle or the applied true strain. This quantity also decreases with the increasing alloying content, that is, relatively larger quantity of recovered volume strain for BD13 than for BD16. 19.5.4.3 Energy Dissipated during Cyclic Tension In Fig. 19.21 is shown the definition of energy dissipation during each cycle of loading and unloading. Here, the term ‘‘energy’’ is the energy density, that is, energy per unit of volume. The whole energy is composed of elastic energy Ee, represented by the area of shadow region and plastic energy Ep, represented by the area of white region, neglecting the thermal loss. The elastic energy is released after unloading to zero load. The plastic energy is further divided into two parts: one from matrix shear

Figure 19.21 Definition of energy dissipated (Ep) and elastic one (Ee). (From Reference 10 with permission from Elsevier Inc.)

578

Polyolefin Blends

Figure 19.22 Total energy density versus applied true strain for cyclic tension. (From Reference 10 with permission from Elsevier Inc.)

banding and another from the cavitation. From the experimental results, it is unfortunately impossible to separate these two parts of plastic energy. The curves in Fig. 19.22 represent the whole energy dissipated during the loading–unloading cycles. At early stage of deformation with true strain below 0.4, the total energy absorbed is not much different for the materials studied. However, with the increasing true strain, the increase of total energy for neat PP, BD13, and BD14 with low alloying content becomes slower compared to BD15 and BD16 with high alloying content. The aberration of the curves between the blends with low and high alloying content is larger and larger as the true strain continues to increase. This attendance of variation signifies also that only when the deformation reaches a certain level, the contribution of different deformation mechanisms to the energy dissipation can be identified qualitatively. If we consider only the energy dissipated plastically, the elastic energy being removed (which takes about 10–20% of total energy), we have the curves as shown in Fig. 19.23. The curves in Fig. 19.23 are similar to those in Fig. 19.22. This means also that the elastic energy for all materials is not much different. From the results given in Figs. 19.22 and 19.23, the energy dissipation for neat PP is always medium, compared with the blends. Now, let us discuss the relationship between the volume strain and the energy dissipated, based on the curves presented in Figs. 19.20 and 19.23. It is noted that the great volume strain corresponds to small energy dissipation, such as in the case of BD13 and BD14. Large volume strain comes from the abundant nucleation and growth of cavities inside the samples. The small energy value with large volume strain means that the cavitation is not the main source of energy dissipation. The polymer scientists accept the point of view that, for polymer blends, the main source of energy dissipation is the matrix shear deformation. It is stated previously that shear deformation cannot result in volume change. Therefore, we have the reason to

Chapter 19

Plastic Deformation and Damage Mechanisms

579

Figure 19.23 Energy dissipated versus applied true strain for cyclic tension. (From Reference 10 with permission from Elsevier Inc.)

conclude that for the blends of BD15 and BD16 with high alloying content, shear deformation of matrix is the controlling factor for energy dissipation, without causing volume change. The small amount of volume strain for BD15 and BD16 may be contributed by the interfacial debonding of PA6 phases whose content is not high.

19.6 MECHANISMS OF PLASTIC DEFORMATION AND DAMAGE 19.6.1 Damage Mechanisms in Polymer Blends The development of volume strain in the polymer blends was expected a priori since many authors previously observed the initiation of different kinds of voids in such materials at a microscopic scale. Crazing is a very important form of damage in blends whose matrix is brittle. Recently, one of the present authors (34) verified with the Vide´oTractionß technique that volume strain of HIPS at room temperature was nearly as high as the tensile strain itself (D ¼ dev =e3 ¼ 0:98 is almost constant up to the rupture strain eR3 1:0). Numbers of authors have clearly identified by TEM the formation of crazes at the equatorial periphery of the salami-type inclusions, where a high stress concentration appears under tension. The crazes in HIPS can even be detected indirectly by the naked eye through the diffuse whitening that they produce in the deformed samples. Crazing of the matrix was observed in a variety of other blends (35), and, for all these materials, it was stated that the multiplication of crazes dissipates considerable amounts of mechanical energy and hence increases the toughness of the matrix (36). The decohesion of the particles from the matrix was frequently observed in blends with ductile matrix and poorly adherent particles. Typical cases for such a

580

Polyolefin Blends

process are glass bead and CaCO 3 reinforced thermoplastics. Generally speaking, increasing the adhesion and/or thickness of the elastomer interphase around the particles tends to reduce the tendency for decohesion and favor the cavitation process. Cavitation was also identified as an active mechanism in systems where a rubberlike phase (particles or interphase) is susceptible to ‘‘implode’’ under the effect of the hydrostatic stress induced by the applied tension. Fond (36) has recently revisited the critical conditions under which this form of damage becomes energetically favorable. In the case of core–shell rubber-toughened PMMA, he ascribed the extensive whitening under tension at room temperature to the profuse formation of voids in the rubber shell of the toughening particles. The three mechanisms described above briefly contribute to the overall dilatation of the material under tension, and it is not easy to assign to each process its relative importance in the recorded damage rate D. However, some authors like Keskkula and Schwarz (37) for HIPS, showed from detailed morphological observation that the crazing in the PS matrix is not the only active source of damage, but that the decohesion at the PS/PB interface and cavitation in the PB nodules play a significant role as well.

19.6.2 Influence of Damage Type Apart from the identification of the damage mechanisms involved in our blends, it is essential to interpret the unexpected decrease of the damage rate with the alloying content in the PP/PA/POE blends. It would be interesting to understand why, for the blend of lowest concentration (BD13), this global evolution is not strictly obeyed. Before all, we will discuss the origin of plastic damage in the neat PP; this polymer undergoes a volume strain ev 0:5 for a tensile strain e3 1:5. Only a few authors have acknowledged the importance of plastic damage in semicrystalline polymers since G’Sell et al. (38) measured considerable loss of density in PP after stretching and unloading. More recently, extensive damage was measured in several semicrystalline polymers under tension by means of the novel Vide´oTractionß system (the present work on PP, G’Sell et al. (39) on POM, and Addiego et al. (40) on HDPE). Microscopic observation in deformed samples of neat PP (41) and HDPE (40) showed that crazing occurred very early within the spherulites between the equatorial lamellae and then along the interspherulitic borders. Later, when crystallite fragmentation begins, large cavities are observed among the fibrillated ‘‘shish-kebab’’ structures. Consequently, one should banish the simplistic scheme that nodules in a polymer necessarily induce damage in a medium that otherwise would exhibit isochoric deformation by plastic shear. On the contrary, (i) damage processes provide an important contribution to the plastic deformation of most polymers and (ii) dispersed particles often decrease the original damage rate of polymeric matrices, except those that are coarse and poorly adhesive such as glass beads.

Chapter 19

Plastic Deformation and Damage Mechanisms

581

The decrease of the damage rate at high alloying content in the PP/PA6/POE blends can be explained from the evolution of the microstructure itself. The POE elastomer tends to form more and more isolated nodules at high alloying content with respect to the proportion in the PP/PA6 interphase. As such, the contribution of these rubber-like nodules increases correlatively. On the basis of the literature quoted above, it is probable that the two types of particles induce different mechanisms. In the case of decohesion of the matrix from a hard nodule and/or cavitation in the rubbery interphase, large voids develop by stretching the matrix in the vicinity of the particle poles, while the matrix is unable to contract near the equator because of the rigidity of the particle. This induces locally a large axial strain, but very small transverse strains, so that ðe1 þ e2 Þ e3 . It results in volume strain that is very large ðD 1Þ. This is typically the case for the PP/glass bead system (18) and, to a certain extent, for our blends with low alloying content (BD13, BD14). By contrast, if cavitation occurs in rubberlike particles, it appears schematically that the material is free to contract transversally and that ðe1 þ e2 Þ e3 . It results in volume strain that is small ðD 1Þ. This could be the case for our blends with high alloying content (BD15, BD16). Of course, these extreme cases are schematic and intermediate cases should be envisaged.

19.6.3 Damage-induced Shear Banding Most authors have considered that shear banding is a mechanism concurrent to damage (14). Alternatively, we consider that shear banding is synergistic with damage. Whatever the damage mechanism acting, the formation of microscopic ‘‘voids’’ in a matrix (even a brittle polymer) is likely to locally favor the formation of shear bands. This statement is supported by some observations and simulations. Even with simple optical microscopy, we have observed in PET (42) that shear bands were preferably nucleated at the tip of crazes. Also in polycarbonate (PC), Smit et al.(43) have shown by finite-element computation that randomly distributed nanovoids make plastic deformation much easier than in neat PC (yield drop is suppressed) and initiate a profusion of shear bands joining close void doublets. It is probable that this effect is largely responsible for the influence of interparticular ligament length on the blend toughening, which was introduced by Wu (44). The detailed micromechanical features of this process are still in progress (45), but it seems that preferential activation of plastic shear from crazes and cavities is likely to play a major role in toughening polymers. Consequently, in the classical decomposition of the total volume strain into a sum of terms representing elastic, plastic shear and damage mechanisms (18,32,46,47), it should be envisaged to introduce a dependence of the plastic shear term on the damage term. In our blends, it is probable that the lower damage rate of BD15 to BD16 with respect to neat PP is for a significant part due to the enhancement of microscopic shear banding from the voids (decohesion and/or cavitation) formed early after the yield point. Also, the activation of shear yielding in our blends probably contributes to the increase of the impact energy with alloying content.

582

Polyolefin Blends

19.6.4 Microscopic Observation under SEM Above analysis on the deformation mechanisms is deduced from the tensile curves without solid proofs. Now, in situ tensile tests aimed to reveal the real mechanisms of plastic deformation and damage were carried out with a small tensile stage. The samples were machined from original plate into desired shape. Their total length was 40 mm, gage width and length were 10 mm and 4 mm, respectively, and the thickness was 3 mm. The pictures taken after each test are gathered together in Fig. 19.24. Each sample underwent different deformation magnitude from 0.2 to 1.0. The figures written on the left end of each sample represent the final true strain values applied to the sample, while those on the right are the number of material serial. From the color deepness, the transparency became more evident with the increasing PA6 or POE content. From the appearance of the samples deformed, two features are noted. First, the greater the PA6 or POE content, the smaller the extension or the necking of the samples. Second, the whitening phenomenon is more outstanding for the blends with less alloying elements (BD13 and BD14) than that with more alloying elements (BD15 and BD16). The whitening is generally produced by the crazes and cavities created during the loading. The microstructural characteristics of plastic deformation and damage are analyzed with the help of micrographs of the cryofractured surfaces taken from

Figure 19.24 In situ test samples deformed under different true strains. (a) BD13, (b) BD14, (c) BD15 and (d) BD16. (From Reference 10 with permission from Elsevier Inc.)

Chapter 19

Plastic Deformation and Damage Mechanisms

583

the samples deformed and along the loading direction. Detailed analysis is stated below. 19.6.4.1 Blend BD13 (Fig. 19.25) At 0.2 true strain, the matrix deformation is small, which is proved by the even and smooth cryofractured surfaces. However, the interfacial debonding has taken place, especially at the interface of large PA6 particles. The interfacial crack at one pole of the particle is in the form of crescent moon. As the applied strain increases further to 0.4, the stress drops slightly according to the tensile curves. The microstructure manifests two evolutional characteristics: interfacial crack growing and transforming into microvoid and crack propagating horizontally through several PA6 particles. The evolution of the microstructural morphology continues with the strain increasing to 0.6–1.0. The coalescence of the microvoids by the horizontal propagation of

Figure 19.25 Cryofractured surfaces of BD13 (figures on the micrograph indicating the level of applied true strain). Vertical loading. (From Reference 10 with permission from Elsevier Inc.)

584

Polyolefin Blends

interfacial cracks has taken place, as shown by the micrograph at 0.8 true strain. Some large PA6 particles are completely debonded from the matrix. Besides, the matrix ligaments lying in the loading direction are observed on the surface as shown by the micrographs at 0.6 and 1.0 true strain. From the above observation, it follows that the volume increase comes from the microvoids formed at two poles of particles and the cracks vertical to the loading direction. 19.6.4.2 Blend BD14 (Fig. 19.26) Compared with BD13, at low strain of 0.2, the matrix appearance is similar, but less particles are debonded from the matrix. At 0.4 true strain, the matrix deformation and microvoids formation are less important than BD13 as well. Large matrix deformation and large quantity of microvoids appear when the true strain reaches 0.6. The microfilaments across the interface can still be seen for some small size particles. Even though the microvoid formation and growth are developed greatly, the horizontal

Figure 19.26 Cryofractured surfaces of BD14 (figures on the micrograph indicating the level of applied true strain). Vertical loading. (From Reference 10 with permission from Elsevier Inc.)

Chapter 19

Plastic Deformation and Damage Mechanisms

585

cracks are seldom observed. At large true strain of 0.8, the microvoids formed at two poles of PA6 particles grow so greatly that transverse cracks are created and propagated. Abundant microvoids are engendered at 1.0 true strain, forming the cavity bands, that contain a large quantity of small PA6 particles. The matrix ligaments between two close PA6 particles are stretched greatly along the loading direction. The main reasons of plastic deformation and damage are same as BD13, but the volume increase of BD14 is smaller than that of BD13. 19.6.4.3 Blend BD15 (Fig. 19.27) As to BD15, a remarkable character is that the dispersed PA6 particles have the ellipsoidal shape, instead of spherical one as in the case of BD13 and BD14. The long axis of spheroid coincides with the loading direction (also the length direction of molded samples). At average, the size of PA6 spheroid is bigger than spheres. Under low strain magnitude of 0.2, the interfacial debonding can still be seen for small

Figure 19.27 Cryofractured surfaces of BD15 (figures on the micrograph indicating the level of applied true strain). Vertical loading. (From Reference 10 with permission from Elsevier Inc.)

586

Polyolefin Blends

spherical (existing seldom) and large ellipsoidal PA6 particles. The growth of interfacial cracks is not as important as in the case of BD13 and BD14 with increasing true strain. It seems that not only at the interface but also in the matrix far from the particles the cavitation has appeared. The latter is probably caused by the dispersed POE phase. However, the proof for this assumption should be provided in the future work. No horizontal crack is created. Interfacial debonding is still the main mechanism of volume expansion. But, the number of dispersed particles is less than that in BD13 and BD14, so the resulted volume strain is smaller. The shearing deformation of matrix ligament between two particles is considered to contribute much to the toughening effect. 19.6.4.4 Blend BD16 (Fig. 19.28) As to BD16, different characters on the morphology of cryofractured surfaces are remarked. The interfacial debonding can be seen only when the true strain reaches a

Figure 19.28 Cryofractured surfaces of BD16 (figures on the micrograph indicating the level of applied true strain). Vertical loading. (From Reference 10 with permission from Elsevier Inc.)

Chapter 19

Plastic Deformation and Damage Mechanisms

587

certain level (0.4 and 0.6). A certain quantity of microcavities in matrix is created with the increasing true strain. We consider that the dispersed POE phases may be the resources of the cavitation in matrix. At high strain values of 0.8 and 1.0, the interfacial debonding became important and even turned into microcavities by expanding and growing into matrix. However, the number of dispersed PA6 particles is so small that the cavitation at the interface cannot result in a great volume increase. The high toughness of BD16 comes evidently from matrix shearing deformation, as well as the cavitation and deformation of POE phase, which exists probably both at the interface and in the matrix. 19.6.4.5 PP (Fig. 19.29) In the case of a semicrystalline polymer such as PP, the microstructural features are likely to appear at the scale of the spherulites (typically 5–100 mm in diameter) or even closer at the scale of the long period of the lamellar stacks (10–100 nm). In order to accede to the latter details, it was shown previously (48) that etching of the polished surface with oxidizing acids engraves the amorphous interstices and let the crystalline morphology appear: lamellae, or at least stacks of lamellae, become visible. Figure 19.29a shows a micrograph of neat PP obtained in the undeformed state. It confirms the spherulitic nature of the morphology in this PP whose crystalline fraction (51 wt% measured by DSC) is totally composed of monoclinic a modification. One spherulite at the center of the micrograph is outlined with a very regular polyhedral shape and an average diameter of about 20 mm. Although the individual lamellae are hardly resolved at that scale, the radial growth direction appears clearly. Figure 19.29(b) was obtained with the same protocol but after a tensile test interrupted at a strain e3 1:2 that corresponds to a stretch ratio l ¼ expðe3 Þ of about 3.3. It shows essentially one spherulite elongated along the tensile direction (horizontal). The major feature visible is a network of microcracks within the spherulite. The microcracks are oriented not only in the direction perpendicular to the tensile axis, but also along various radial directions of crystal lamellae. Also, it seems that they stopped growing when approaching the boundary with an adjacent spherulite. No interspherulitic decohesion was observed in this polymer. The above results are in good agreement with those of Aboulfaraj et al. (41) who found that, under tension, the crazes in the a-PP appear preferentially at the center of the spherulites, grow parallel to radial directions, and eventually reorient their propagation direction perpendicular to the tensile axis. Others have shown (49) that the early stage of spherulite damage corresponds to the fragmentation of semicrystalline stacks after shear and bending under the constraint of tie molecules. Figure 19.29c corresponds to the same strain but at a much lesser magnification. Now it is no longer possible to resolve any crystalline morphology. Instead, one observes highly elongated cracks in the tensile direction. The size of the cracks is of the order of 10 mm in width and more than 100 mm in length. It seems that these defects represent the ultimate evolution of the spherulite microcracks under the effect of large plastic deformation. Consequently, the highly stretched polymer can be

588

Polyolefin Blends

Figure 19.29 SEM observation of neat PP etched with an acid solution. (a) evidence of spherulitic morphology at e3 0, (b) microcracks in the core of a spherulite at e3 1:2, and (c) elongated crack observed at low magnification at e3 1:2 (arrows indicate the tensile direction). (From Reference 12 with permission from Elsevier Inc.)

Chapter 19

Plastic Deformation and Damage Mechanisms

589

regarded as an anisotropic porous structure composed of ultraoriented fibrils separating such oriented cracks. As the volume strain measured with the Vide´oTractionß system at e3 1:2 being equal to ev 0:42, the void fraction in RVE of the stretched specimen is ðV V0 Þ=V0 ¼ expðev Þ 52%.

19.6.5 Microscopic Observation under TEM The TEM observation after staining with RuO4 provides us with a high definition view of the microscopic mechanisms and reveals the specific role of the POE phase. In the BD13 blend, illustrated in Fig. 19.30a, three phenomena are visible: (i) interfacial debonding at the poles of PA6 particles, (ii) high elongation of hyperelastic POE particles, and (iii) cavitation in dispersed POE droplets. On the whole, since the size and number of POE particles are smaller than that of PA6 particles, the former mechanism is more frequently observed and seemed to have a leading influence on the overall volume strain. For BD14, interfacial debonding in Fig. 19.30b represents the main damage factor such as in BD13. It occurs preferentially around the largest PA6 particles. Some smaller PA6 particles seem intact, even though their POE shell has been strongly stretched in the loading direction. Again the volume increase comes preferentially from the debonding of PA6 particles. However, it is interesting to note that cavitation inside the PA6 phase is never observed. For BD15, with an alloying content of 45 wt%, as shown in Fig. 19.30c, the cavitation at the poles of PA6 particles seems to result more from the cavitation of the POE interphase than from a simple interfacial debonding (although it is not always simple to make the difference). Also, the cavitation in dispersed POE droplets is very active. Furthermore, in some micrographs a new type of salami structure is present: large POE nodule containing small PA6 inclusions. Under uniaxial tension, the POE of these salami structures deforms easily, while the PA6 particles inside seem unaffected. At last for BD16 (Fig. 19.30d), in which PP is in minority, cavitation occurs more likely in POE interphase at the poles of large PA6 particles and in large isolated POE particles (that are more numerous in this blend). However, conservative deformation mechanisms become predominant and are commonly identified: (i) at smaller POE droplets, (ii) at salami structures, and (iii) in the PP matrix.

19.6.6 Discussion The above morphological characterization helps us to understand the influence of composition on the dilatation of the materials under uniaxial tension. For neat PP, the crazing mechanisms have been the object of many previous papers (41, 49, 50). It was shown that crazes resulting from early fragmentation of the crystalline stacks propagate in the amorphous phase between the lamellae along the radial directions as shown in Fig. 19.31. At higher strain, the cracks are rotated and become nearly parallel to the tensile direction. It is remarkable that although such

590

Polyolefin Blends

Figure 19.30 TEM micrographs of deformed blends after staining with RuO4 vapor: (a) BD13, (b) BD14, (c) BD15, (c) BD16 (same strains as in Fig. 19.29). (From Reference 12 with permission from Elsevier Inc.)

Chapter 19

Plastic Deformation and Damage Mechanisms

591

Figure 19.31 Schematic evolution of cracks in a PP spherulite for moderate and large applied strain in uniaxial tension. (From Reference 12 with permission from Elsevier Inc.)

damage phenomena were discovered earlier, they did not convince previous authors of the importance of volume strain in PP. Consequently, in the mechanical engineering community, finite element codes applied to the prediction of large deformation in PP parts (e.g., for the optimization of shock absorbers) implicitly make the assumption that the plastic response is isochoric ðev ¼ 0Þ as soon as the yield point is passed. As such, the quantitative dilatation measurements related to the microscopic observation establish the statement that the important volume increase undergone by neat PP under tension is the direct consequence of the development of crazes and the elongation of voids in the spherulitic structure. In the case of the ternary blends, the difficulty for modeling the behavior comes essentially from the multiplicity of phase configurations. In the case of PP/PA6/POE blends, the three components are not chemically compatible but the maleic anhydride grafted on POE elastomer chains provides with a certain affinity for the PA6. Consequently, most of the POE migrates spontaneously during the mixing sequence toward the interface of the PA6 nodules where it forms a peripheral interphase like in a ‘‘core–shell’’ particle with a stiff interior (PA6) and a thin rubbery envelope (POE). However, at high alloying content, the thickness of the interphase becomes too large and the compatibilizer in excess precipitates into isolated particles. As we saw in different micrographs, more complex configurations are found, notably salami-type nodules, with PP droplets inside the core of PA6–POE particles or with PA6 droplets within a big POE particle. Also, in the case of the blend with the highest alloying content, special ‘‘clusters’’ of agglomerated PA6 and POE particles were identified. These different configurations are summarized schematically in Fig. 19.32. When the blends are subjected to tensile testing, a certain fraction of the overall strain is accommodated by conservative deformation of the material. In the PP matrix, deformation results from the combination of amorphous phase hyperelasticity and crystal plasticity, as discussed earlier (50). The PA6 phase is also capable of deforming plastically, but its flow stress in the plastic stage is much higher than that of PP. Consequently, in the PP/PA6 blends the isolated PA6 particles exhibit less

592

Polyolefin Blends

Figure 19.32 Schematic representation of the different particles in the PP/PA6/POE blends. (From Reference 12 with permission from Elsevier Inc.)

deformation than the PP matrix, leading to interfacial stress concentrations. By contrast, the isolated POE nodules deform easily due to the rubberlike properties of the compatibilizer. It is thus logical to observe highly elongated POE particles in the stretched blend samples. At last for the POE–PA6 salami nodules, they deform

Chapter 19

Plastic Deformation and Damage Mechanisms

593

Figure 19.33 Schematic presentation of different deformation and damage mechanisms. (From Reference 12 with permission from Elsevier).

easily but the PA6 inclusions inside remain nearly undeformed. The deformed configurations are illustrated in Fig. 19.33. As revealed by the ev versus e3 curves obtained by means of the Vide´oTractionß system, cavitation plays an important role in the deformation of the blends investigated here. The micrographs shown above revealed the various mechanisms of creating voids for the different morphological configurations. The major ones (nonexhaustively) are illustrated in Fig. 19.34. One class of damage mechanism corresponds to interfacial debonding. Owing to the contrast in the mechanical properties of adjacent materials, the induced stresses break the weak adhesion leading to arc crack preferentially developing at the poles. Interfacial debonding is principally active for isolated PA6 particles, less for isolated POE particles. As for the PA6–POE core–shell nodules, decohesion seems to affect primarily the POE/PP interface. Although decohesion is the most active damage mechanism for blends with low alloying content (BD13 and BD14), another process appears as the size of the POE domains increases. It has been shown by many authors (51,52) that a rubberlike

594

Polyolefin Blends

Figure 19.34 Schematic presentation of different cavitation mechanisms. (From Reference 12 with permission from Elsevier Inc.)

inclusion embedded in a deformed matrix undergoes high hydrostatic stress and eventually ‘‘explodes’’ by nucleation and growth of an internal crack. This process releases significant amount of stored energy and reorganizes the stress field in the matrix with higher deviatoric stress component. This is the reason why cavitation is often observed in the POE phase of the BD15 and BD16 blends. According to configurations, voids are formed either in the isolated POE particles, in the POE interphase constituting the shell of the PA6–POE nodules or in the POE phase of the complex salami structures. Of course, when analyzing the micrographs, one should

Chapter 19

Plastic Deformation and Damage Mechanisms

595

take care in distinguishing the cavitation from the subsequent distortion induced by large deformation. Apart from the qualitative aspects of deformation and cavitation processes developed above, it is now important to interpret the quantitative importance of dilatation versus the conservative deformation contribution. Volume strain is high for neat PP and small for neat PA6. As for the PP/PA6/POE blends, they exhibit lower and lower volume strain as the total alloying content increases (except BD13 that is slightly over neat PP). This evolution is somewhat unexpected since we could have envisaged, in a simplistic approach, that the activation of additional cavitation mechanisms in the blends (decohesion and cavitation) should have increased more and more the overall dilatation. Since experiments prove that this point of view is not true, the problem should be reconsidered on more correct bases. A recent paper (53) have treated this question in the case of viscoelastic materials and led to the conclusion that the influence of cavitation on the overall response of the material depends on many factors such as applied strain rate, particle-size distribution, and relaxation time of matrix. Obviously developing a model valid for large strain plasticity would be much more complex and is beyond the scope of this study. Consequently, we will just analyze here some arguments that may explain the observed facts. The first point concerns the intrinsic properties of the materials constituting the blends. It is evident that, for the blends under consideration, we have mixed PP that suffers very high cavitation with increasing amounts of PA6 that deforms plastically with particularly low volume strain. Although the PA6 phase does not support the same amount of strain in the blends as the PP matrix (PA6 particles has shown little deformation in the micrographs), it is presumable that the lower tendency to cavitation of the blend is partly controlled by the more isochoric nature of the PA6 component. The role of the POE phase is more complex. As many elastomers, it undergoes extensive cavitation when it is hydrostatically loaded within the interphase layers and the isolated nodules. However, it is often seen in the micrograph (especially after stress triaxiality has been released by void nucleation) that POE is capable to undergo considerable stretching without further damage. The net effect could thus be to reduce volume strain. Last but not least, it should be envisaged that the nucleation of voids could have a positive effect on the plastic regime of the PP matrix. It has been shown before by many authors (and recently by van Melick et al. (45) using finite-element calculation) that shear plasticity in a solid containing a random array of nanometric voids is considerably facilitated with respect to the same material in its compact form. Shear bands form between pairs of neighboring voids whose distance vector is inclined at nearly 45 on the tensile axis. As such, a diffuse network of percolating shear bands forms in the matrix causing easy plasticity. The ‘‘ligament’’ model of Wu [44] is presented with slightly different words but is globally based on the same arguments. It is probable that plasticity enhancement due to the presence of voids is also relevant to our blends where cavitation is early and finely dispersed. Unfortunately, the microscopic techniques used in this work do not make visible the shear plasticity mechanisms per se and are not capable of providing information on whether the

596

Polyolefin Blends

above model is correct or not. It could be worthwhile to apply the atomic force microscopy technique to our samples in investigating this question more finely. The experimental facts shown in this paper and the deformation scheme presented above are in line with a model previously proposed by Lazzeri et al. (54,55) to explain the interparticle distance effect on the basis of the stabilization effect of dilatational band propagation exerted by stretched rubber particles. Although dilatational bands were not formally identified in our specimens after deformation, the toughening effect of POE particles is definitely important in the PP/PA6/POE blends under investigation here. Not only the above arguments help in understanding the physics of deformation in the blends under consideration, but also they can be utilized to guide the optimization of technical grades. As we showed it previously (9), increasing the (PA6 þ POE) content of the blend is highly favorable for mechanical engineering since the resistance to impact (notched Izod impact strength) is increased by a factor of 5 by passing from BD13 to BD16, while the elastic modulus and yield stress are decreased by 20% only. Although many modifications are available in the preparation process, the qualitative model proposed above to explain microstructural and mechanical features constitutes a valuable contribution to increase further the performances of the blends.

19.7 CONCLUSIONS Plastic damage in PP/PA6/POE blends was studied experimentally under uniaxial tension by using a novel version of the Vide´oTractionß system. Four compositions were investigated with increasing values of (PA6 þ POE), in addition to the neat PP and a reference PA6 grade considered as a generic reference. It was found that, for all systems, volume strain significantly contributes to the overall plastic strain at large deformation. Consequently, plastic deformation of neat polymers and blends should not be considered a priori as an isochoric process. Also, it was found that plastic strain hardening gradually increases with the alloying content. The present investigation reveals the unexpected decrease of macroscopic dilatation rate with increasing alloying content, which is due to several factors. First, we showed that the different damage mechanisms do not have equivalent influences on the overall dilatation. Decohesion and cavitation at the rubber-like interphase around the PA6 nodules participate much more to the macroscopic volume strain than cavitation within the isolated POE particles. The fact that the latter are more numerous for high alloying content explains a part of the damage evolution. Second, we gave elements indicating that the multiplication of shear bands is highly facilitated in the ligaments between voids, so that isochoric shear plasticity plays an increasing role in the blends with higher alloying content. According to the detailed examination of the cryofractured surfaces under SEM and TEM, the volume strain mainly comes from the interfacial debonding, which transformed into microcavities under large applied strain. However, large volume strain does not mean high energy dissipation. Conversely, the energy dissipation

Chapter 19

Plastic Deformation and Damage Mechanisms

597

decreases with increasing volume strain because the main source for energy dissipation is matrix shearing deformation, instead of cavitation that results in volume change. The relationship between the volume strain and the energy dissipation can well explain the damage mechanisms and toughening effect of polymer blends. As to neat PP, large volume strain comes from interlamellar crazes nucleated by crystal fragmentation and propagated along the radial direction in the spherulites. Under large strain, the cracks are turned toward the loading direction, forming large parallel ‘‘empty channels.’’

NOMENCLATURE E0 : E e: E p: Li0: Li: PA6: POE: PP: RVE: tan d: V 0: V: ei: eN i : e v: eca v : eN v: eel v: epl v : eR3 : l: s y: yel: yT: D: DeN i :

Storage modulus Elastic energy Plastic energy Initial dimensions of the RVE Current dimensions of the RVE Polyamide 6 Polyethylene–octene Polypropylene Representative volume element Loss tangent Initial volume of the RVE Current volume of the RVE Applied true strain along i direction Nominal engineering strain along i direction Volume strain Damage volume strain Nominal volume strain Elastic volume strain Plastic volume strain Rupture strain Stretch ratio Yield stress Poisson’s ratio Tangent Poisson’s ratio Damage rate Nominal strain increment

REFERENCES 1. S. C. Wong and Y. M. Mai, Polymer, 41, 5471 (2000). 2. S. C. Wong and Y. M. Mai, Polymer, 40, 1553 (1999). 3. A. Gonzales-Montiel, H. Keskkula, and D. R. Paul, Polymer, 36, 4587 (1995).

598

Polyolefin Blends

4. A. Gonzales-Montiel, H. Keskkula, and D. R. Paul, Polymer, 36, 4605 (1995). 5. A. Gonzales-Montiel, H. Keskkula, and D. R. Paul, J. Polym. Sci. B Polym. Phys., 33, 1751 (1995). 6. J. Roeder, R. V. B. DeOliveira, M. C. Goncalves, V. Soldi, and A. T. N. Pires, Polyn. Test., 21, 815 (2002). 7. R. N. Darie, M. Brebu, C. Vasile, and M. Kozlowski, Polym. Degrad. Stab., 80, 551 (2003). 8. N. Zeng, S. L. Bai, K. Cao, C. G’Sell, and Y. W. Mai, Polym. Int., 51, 1439 (2002). 9. S. L. Bai, G. T. Wang, J.-M. Hiver, and C. G’Sell, Polymer, 45, 3063 (2004). 10. S. L. Bai and M. Wang, Polymer, 44, 6637 (2003). 11. C. G’Sell, S. L. Bai, and J.-M. Hiver, Polymer, 45, 5785 (2004). 12. S. L. Bai, C. G’Sell, and J.-M. Hiver, Polymer, 46, 6437 (2005). 13. C. G’Sell C, N. A. Aly-Helal, and J. J. Jonas, J. Mater. Sci., 18, 1731 (1983). 14. C. B. Bucknall, D. Clayton, and W. E. Keast, J. Mater. Sci., 7, 1443 (1972). 15. A. N. Gent and B. Park, J. Mater. Sci., 19, 1947 (1984). 16. C. G’Sell, J.-M. Hiver, and F. Gehin, Real-time quantitative determination of volume variations in polymers under plastic strain, in: 13th Proc. On Deformation, Yield and Fracture of Polymers, Cambridge, UK, The Institute of Materials, London, 2000, p. 371. 17. S. L. Bai, J. K. Chen, Z. P. Huang, and Z. Z. Yu, J. Mater. Sci. Lett., 19, 1587 (2000). 18. A. Meddad and B. Fisa, J. Appl. Polym. Sci., 64, 653 (1997). 19. B. Pukanszky, M. Van Es, F. H. J. Maurer, and D. Voros, J. Mater. Sci., 29, 2350 (1994). 20. R. W. Meyer and L. A. Pruitt, Polymer, 42, 5293 (2001). 21. S. L. Bai, Z. D. Liu, and Y. Ju, Polym. Int., 50, 973 (2001). 22. A. D. Drozdov and J. Christiansen, Eur. Polym. J., 39, 21 (2003). 23. A. Gonzales-Montiel, H. Keskkula, and D. R. Paul, Polymer, 36, 4621 (1995). 24. P. W. Bridgman, Studies in Large Plastic Flow and Fracture, McGraw-Hill, New York, 1952. 25. H. G. H. van Melick, L. E. Govaert, and H. E. H. Meijer, Polymer, 44, 457 (2003). 26. C. G’Sell and J.-M. Hiver, French Patent 2,823,849 (2002). 27. A. E. H. Love, A Treatise on the Mathematical Theory of Elasticity, 4th edition, Dover, New York, 1944. 28. D638-02 Standard Test Method for Tensile Properties of Plastics, ATSM, West Conshohocken, PA, USA, 2002. 29. C. G’Sell and J. J. Jonas, J. Mater. Sci., 14, 583 (1979). 30. C. G’Sell, J.-M. Hiver, and A. Dahoun, Int. J. Solids Struct., 39, 3857 (2002). 31. C. B. Bucknall and D. Clayton, J. Mater. Sci., 7, 202 (1972). 32. C. G’Sell, A. Dahoun, and J.-M. Hiver, Compe´tition des me´canismes de cisaillement plastique et d’endommagement dans les polyme`res solides en traction uniaxiale, in: Proc. Conf. ‘‘Materials 2002,’’ Tours, France, 21–25 October 2002, Communication CM 03003-1-4 CA. 33. S. Elkoun, C. G’Sell, L. Cangemi, and Y. J. Meimon, J. Polym. Sci. B Polym. Phys., 40, 1754 (2002). 34. C. G’Sell, J.-M. Hiver, and A. Dahoun, Int. J. Solids Struct., 39, 3857 (2002). 35. C. B. Bucknall and T. Br. Toshii, Polym. J., 10, 53 (1978). 36. C. Fond and C. G’Sell, Me´canique Industries, 3, 431 (2002). 37. H. Keskkula and P. D. R.Schwarz, Polymer, 27, 211 (1986). 38. C. G’Sell, A. Dahoun, V. Favier, J.-M. Hiver, M. J. Philippe, and G. R. Canova, Polym. Eng. Sci., 37, 1702 (1997). 39. J. Mohanraj, D. C. Barton, I.M. Ward, A. Dahoun, J. M. Hiver, and C. G’Sell, Polymer, 47, 5852 (2006). 40. F. Addiego, A. Dahou, C. G’Sell, and J. M. Hiver, Polymer, 47, 4387 (2006). 41. M. Aboulfaraj, C. G’Sell, B. Ulrich, and A. Dahoun, Polymer, 36, 731 (1995).

Chapter 19

Plastic Deformation and Damage Mechanisms

599

42. C. G’Sell, J. Phys. IV, 23, 1085 (1998). 43. R. J. M. Smit, W. A. M. Brekelmans, and H. E. H. Meijer, J. Mech. Phys. Solids, 47, 201 (1999). 44. S. Wu, Polymer, 26, 1855 (1985). 45. H. G. H. van Melick, L. E. Govaert, and H. E. H. Meijer, Polymer, 44, 3579 (2003). 46. D. Heikens, S. Dirk Sjoerdsma, and W. Jan Coumans, J. Mater. Sci., 16, 429 (1981). 47. O. Frank and J. C. Lehmann, Polym. Sci., 264, 473 (1986). 48. M. Aboulfaraj, B. Ulrich, A. Dahoun, and C. G’Sell, Polymer, 34, 4817 (1983). 49. G. Coulon, G. Castelein, and C. G’Sell, Polymer, 40, 95 (1998). 50. A. Lazzeri and C. B. Bucknall, Polymer, 36, 2895 (1995). 51. C. Fond, A. Lobbrecht, and R. Schirrer, Int. J. Fract., 77, 141 (1996). 52. C. B. Bucknall, P. S. Heather, and A. Lazzeri, J. Mater. Sci., 16, 2255 (1989). 53. J. K. Chen, Z. P. Huang, and Y. W. Mai, Acta Mater., 51, 3375 (2003). 54. A. Lazzeri, The kinetics of dilatational bands and the interparticle distance effect in rubber toughened polymers, in: Proc. 10th International Conference on Deformation, Yield and Fracture of Polymers, Cambridge, UK, The Institute of Materials, London, 1997, p. 442. 55. S. M. Zebarjad, R. Bagheri, S. M. Seyed Reihani, and A. Lazzeri, J. Appl. Polym. Sci., 90, 3767 (2003).

Chapter

20

Reactive Compatibilization of Binary and Ternary Blends Based on PE, PP, and PS Mo´nica F. Dı´az,1 Silvia E. Barbosa,1 and Numa J. Capiati1

20.1 INTRODUCTION Multicomponent materials (fiber-reinforced composites, particulate-filled polymers, nanocomposites, polymer blends, biorelated materials, etc.) are used in increasing quantities in all fields of the economy and are expected to grow in the future. They are present in tools, utensils, and devices used every day at home, in offices, or in industrial plants. Particularly, polymer blends of existing materials offer the ability to produce materials with combinations of properties superior to those of single polymers. Thermoplastic blends have considerable potential for applications to important industrial fields such as automotive, electronic, packaging, medical, building, and so on (1–4). A most recent and less studied application of blending is the municipal plastic waste disposal (mainly composed of PE, PP, PS, and PVC) by commingled plastic recycling. However, due to the low compatibility between the components, the direct mechanical blending leads to materials with poor properties and phase segregation (5,6). Polymer blends always consist of several phases giving place to interphases between them. Interaction of phases across the interphase is one of the factors determining the properties of these materials; thus, the study and modification of interfacial interactions are of utmost importance for their further development. This key zone between the blend phases, which has thickness and properties different from those of the components, needs to be compatibilized in order to improve the blend properties. Such an interphase is 1 Planta Piloto de Ingenierı´a Quı´mica, PLAPIQUI (UNS-CONICET), Camino La Carrindanga km. 7 (8000), Bahı´a Blanca, Argentina

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

600

Chapter 20 Reactive Compatibilization of Binary and Ternary Blends

601

formed by interdiffusion of the blend components, behaving as an extra phase with well-defined dimensions and structure. The interactions between components are the dominating factor in the polymer blends. Interactions, or the lack of them, determine miscibility, phase structure, and properties of the blends. The phase compatibilization of blends is an important issue both in academia (3,7,8) and in industry (1,9,10). Indeed, the earliest and still a relevant example of blend compatibilization is the use of PS-graft-polybutadiene copolymers to enhance the properties of a PS/polybutadiene blend to produce high impact polystyrene (11). The compatibilization processes are based on the improvement of adhesion between the phases, the reduction of interfacial tension, and the phase stabilization by the inhibition of the coalescence of sized droplet in subsequent processing operations (2,12,13). Mechanical testing and morphological analysis were mainly used to assess the compatibilization efficiency. The most widely applied compatibilization methods are addition and reactive ones; in both a compatibilizer is located at the interphase, which provides physicochemical affinity to the components. In the addition method, previously synthesized copolymer is added to the blend, while in reactive compatibilization, the compatibilizer is generated in situ during compounding. The latter is preferred because the compatibilized products seem to be more stable than those obtained by addition (14,15). An attractive route to compatibilize thermoplastic blends, like polyolefin–PS, is the Friedel–Crafts alkylation (F-C). By this reaction, a hydrocarbon chain can be chemically bonded to the PS benzene ring through an aromatic electrophilic substitution. The graft copolymer formed (polyolefin-g-PS), situated at the interphase, will behave as an in situ compatibilizer. The present work discusses the binary (PE/PS, PP/PS) and ternary (PE/PP/PS) blends’ compatibilization by F-C reactions. Detailed studies were performed on the F-C and side reaction characterization, blend morphological aspects, and final mechanical properties.

20.2

FRIEDEL–CRAFTS ALKYLATION REACTION

In 1877, Charles Friedel and James M. Crafts discovered new methods for the preparation of alkylbenzenes, known as Friedel–Crafts alkylation reactions. The mechanism includes an electrophilic aromatic substitution whereby a carbocation is generated as the electrophile in the presence of a Lewis acid catalyst. The general scheme of F-C alkylation reaction is (16) as follows: AlCl

3 ! C6 H5 R þ HX C6 H6 þ R-X

The benzene ring is susceptible to electrophilic attack primarily because of its exposed p electrons. The reaction occurs in three steps: (1) a carbocation is formed by reaction of a halogenated alkane with aluminum chloride or another Lewis acid, (2) the carbocation (acting as an electrophile) hits the benzene ring to form an arenium ion, and (3) the arenium ion loses a proton to produce the alkylated benzene.

602

Polyolefin Blends

In the field of polymers, F-C alkylation reactions were previously proposed to compatibilize PE/PS blends. Earlier studies were carried out by Carrick in 1970 (17), who worked with LDPE and PS in a boiling cyclohexane solution catalyzed by aluminum chloride. He observed the formation of PE-g-PS graft copolymer before the starting of homopolymer degradation. About 20% of the PS added appeared in the graft copolymer. Baker et al. (18,19) also studied F-C compatibilization reactions of PE/PS blends processed in both a batch mixer and a twinscrew extruder. These authors used aluminum chloride as catalyst and styrene as cocatalyst and reported high conversion rates to copolymer. They also suggested possible occurrence of PE and PS chain scission reactions, but no details about process mechanism, morphological, and interfacial strength aspects were treated. The F-C reaction scheme proposed for polymers is

R +

CH

R: PE; PP

CH2

AlCl3/St (molten state)

CH

CH2

R

Although the mechanism of this reaction has not been properly explained for polymers yet, previous studies (17,18) proposed a three-step process: a low molecular weight (MW) carbocation is first formed from an AlCl3 ionic complex. Then, the carbocation hits the PE molecule to yield a macrocarbocation (electrophile). Finally, this macrocarbocation produces, by electrophilic attack on the polystyrene benzene ring, ‘‘brushlike’’ molecules of graft copolymer. On the contrary, the reactive process occurs at the interphase region, where chain ends and short chains are preferentially located (20–22). In consequence, the copolymer molecules formed will contain the shorter chains. Also, since copolymer is made out of the same phase homopolymers, a high compatibility can be expected because of physicochemical affinity. As the reactive process progresses, the shortest homopolymer molecules at the interphase are consumed, then graft copolymer is generated and new homopolymer molecules migrate to the interphase to react. In this dynamic process, molecules are continuously exchanged, causing a shear flow in and out of the interphase. The ‘‘brushlike’’ copolymer molecules are pulled out from their ‘‘hairs’’ by virtue of this shear flow. The longer the hairs, the higher the possibility of being pulled out (23–25). The polyolefin ‘‘hairs’’ of the brushlike copolymer, being capable of entangling with the homopolymer phase, play an essential role in the compatibilization process. Then, a great amount of macrocarbocations should be generated in order to get a large grafting density. The reaction yield, directly related to macrocarbocation concentration, can be increased by using a cation extra source, which should be added as a cocatalyst. Among different cocatalysts, the use of styrene seems to optimize the reaction yield (18).

Chapter 20 Reactive Compatibilization of Binary and Ternary Blends

603

Table 20.1 Homopolymer Molecular Weights and Process Conditions Used in Reactive Compatibilization of Binary Blends. PE/PS 1

Mw ; g mol Temperature, C Speed, rpm Mixing time, min Range of catalyst, %

20.3

50,700/260,000 190 30 12 0.1–1.5

PP/PS 303,000/260,000 200 60 24 0.1–1.0

BINARY BLENDS

The 80%/20% binary blends PE/PS and PP/PS were subjected to F-C reaction for compatibilization performed under nitrogen atmosphere in a Banbury mixer. Different concentrations of catalyst (AlCl3) and 0.3% of cocatalyst (styrene) were added to the completely melted and mixed physical blends. The blends and catalyst concentrations are weight based. High MW commercial grades of linear low density polyethylene (LLDPE), and injection-grade polypropylene and polystyrene were used as homopolymers. The compatibilization conditions and MW of the homopolymers are given in Table 20.1. Blend names are listed in nomenclature. In order to evaluate the level of compatibilization reached and the possible occurrence of side reactions, the detailed studies on chemical and morphological aspects as well as mechanical behavior of the blends were carried out. The PS (minority phase) particle size variation was measured with imaging analyzer software from scanning electron microscopy (SEM) micrographs, and their average particle diameters (Dp ) were determined for each blend. Possible reaction byproducts were also investigated using size exclusion chromatography (SEC) and Fourier transform infrared spectroscopy (FTIR). It is worth noting that in SEC chromatograms, peak polarity depends on the value of the sample refractive index with respect to the pure solvent one. This particular characteristic allowed a qualitative analysis of polyolefins (PE, PP) and PS contributions to the blend SEC chromatograms by signal subtraction. Polyolefin and PS peaks are opposite (PE, PP: positive; PS: negative). Then for any PE/PS, PP/PS, or PE/PP/PS blends, the PE or PP contribution will increase the peak height, while the PS one will reduce it (26).

20.3.1 PE/PS Blends 20.3.1.1

Chemical Aspects

In order to confirm the occurrence of reaction, a SEC characterization and comparison between chromatograms of the physical (PB) and reactive blends (RB) was performed for all samples prepared. The zone of low retention time corresponds to high MW, and high retention times correspond to low MW. Then, PB and RB can be differentiated by shift in retention times due to MW changes. Figure 20.1 shows the

604

Polyolefin Blends

Figure 20.1 SEC chromatograms of PE/PS physical and reactive blend with 0.3% catalyst. (From Reference 26 with permission from Elsevier.)

chromatograms for the PB and RB blends, the latter prepared with 0.3% AlCl3. Both chromatograms presented a single positive peak. The RB peak is shifted to the right, with respect to the PB peak, at both the high and low MW regions. In principle, the shift at high MW region could be attributed to (1) the desired copolymerization reaction leading to PE grafted onto PS (PE-g-PS) having higher MW and/or (2) a cross-linking reaction leading to homopolymers with higher MW than the initial ones. On the contrary, at low MW region, the relative reduction in the amount of short length molecules can be attributed to molecule consumption by F-C and/or crosslinking reactions. In order to distinguish the effect of F-C alkylation on blend MW from the effects of cross-linking and chain scission, the reaction was carried out onto PE and PS homopolymers. The same processing conditions and catalyst concentrations as in the reactive blends were used. These reactive homopolymers were also analyzed by SEC, and the results are summarized in Table 20.2 for different AlCl3 concentrations. Table 20.2

Molecular Weight of Neat and Reactive Homopolymers.a,b

AlCl3 concentration, %

Mw RPE, g mol1

0 (neat) 0.1 0.3 0.5 0.7 1.0 1.5

Mw RPS, g mol1

50,700 50,200 50,800 49,900 50,000 49,400 50,500

a

From Reference 27 with permission from John Wiley & Sons.

b

0.3% styrene, 190 C, 30 rpm, 12 min.

c

Out of detection range of the SEC columns used.

260,000 257,000 248,000 243,000 194,000 25,700 <20,000c

Chapter 20 Reactive Compatibilization of Binary and Ternary Blends

605

The MW of reactive PS (RPS) remains constant up to 0.5% AlCl3, indicating that neither cross-linking nor chain scission occurs at low catalyst concentrations. However, for AlCl3 contents greater than 0.5%, PS chain scission seems to occur as indicated by the corresponding MW reduction. Moreover, at high catalyst content (1.0–1.5%) MW dramatically drops down to 10 times less than the original value. In contrast, Table 20.2 shows that reactive PE (RPE) MW does not change along the range of catalyst content analyzed, indicating that neither cross-linking nor chain scission reactions have taken place. Considering these results, the behavior in Fig. 20.1 can now be explained. Since no cross-linking occurs at less than 0.5% catalyst concentration, the observed RB (0.3% catalyst) shift has to be related to the desired copolymerization reaction and the corresponding molecule consumption. In order to quantify the amount of copolymer formed, a novel technique, which was based on high pressure, high temperature (near critical) selective solvent extraction, was developed (27). The RB and the corresponding PB extractions (by n-heptane) were carried out in parallel. In this way, the results for reactive and physical blends can be compared under the same conditions. An important point in using this method is that PE phase extraction is complete. Then, for an RB the soluble mass extracted will be larger than that for the PB. The mass excess corresponds to the PS contained in the copolymer molecules, which is extracted along with PE. The longer PE chains, in the copolymer, drag the entire copolymer molecules through the hydrocarbon solvent. The amount of copolymer formed can be quantified from the PS contained in it. Figure 20.2 shows the amount of PS (within the copolymer) as a function of the catalyst content. This amount increases, as expected, with the catalyst concentration. A sharp increase is observed in the region of 0.7–1.0% catalyst. It implies a considerable growth in conversion: at 0.7% AlCl3, only 25% of the PS reacts, while at 1.0% AlCl3 more than 75% of the total PS is within the copolymer molecules.

Figure 20.2 PS amount within the copolymer (percentage of the initial PS mass) as a function of catalyst content.

606

Polyolefin Blends

Figure 20.3 SEC chromatograms of the insoluble phases of near-critical extraction from PS and PE/ PS physical and reactive blends with 0.1%, 0.3%, 0.7%, 1.0%, and 1.5% catalyst. (From Reference 27 with permission from John Wiley & Sons.)

In order to justify the sharp increase in the blend insoluble phases, they were characterized by SEC. Figure 20.3 shows these chromatograms along with that corresponding to neat PS. The insoluble phase contains only PS, as indicated by the negative peaks. Two curve sets appear for concentrations below and above 0.7% AlCl3. The chain length of the insoluble phase molecules decreases with the amount of catalyst used. This is evidenced by either the curve shifting (Fig. 20.3) or the MW variation (Table 20.3). The increase in the amount of copolymer formed, in the range 0.7–1.0% catalyst, can be explained by considering both, shorter molecules are more reactive (21,22) and chain scission is important for 0.7% catalyst. Tables 20.2 and 20.3 show that RPS and insoluble fraction MW are close to each other up to 0.7% AlCl3. In contrast, at higher catalyst concentration, where chain scission is more important, MW of RPS is considerably lower than that for insoluble fraction. This fact can be explained as follows: for RPS samples, AlCl3 is consumed Table 20.3 Molecular Weights of the Insoluble Phases From Near-Critical Solvent Extraction.a AlCl3 concentration, (%) 0 (neat) 0.1 0.3 0.5 0.7 1.0 1.5 a

Mw ; g mol1 260,000 260,000 254,400 245,000 200,000 95,700 90,000

From Reference 27 with permission from John Wiley & Sons.

Chapter 20 Reactive Compatibilization of Binary and Ternary Blends

607

Figure 20.4 Emulsification curve for PE/PS blends: effect of the catalyst concentration on the average particle diameter (Dp ). Solid line corresponds to exponential fit.

only by the chain scission reaction, while for RB, when the F-C reaction is present, the catalyst is consumed in macrocarbocation formation as well as in chain scission. 20.3.1.2

Morphological Aspects

As mentioned earlier, F-C reaction generates a graft copolymer (PE-g-PS) at the blend interphase. This copolymer exhibits a high physicochemical affinity with the blend components because it is produced from the same homopolymers that need to be compatibilized. Therefore, it will perform as a tailor-made compatibilizer for the blend. The effect of this affinity is indicated by a considerable interfacial tension reduction of the RB with respect to PB. This is shown by decrease of dispersed phase Dp as well as its dispersion homogeneity (28). Figure 20.4 shows the emulsification curve for the complete set of blends studied. The curve exhibits a similar shape as reported by other authors for compatibilization of immiscible polymer blends (29–31). The copolymer formed shows a remarkable emulsifying effect, reaching important Dp drops from a very low catalyst concentration. The content of catalyst, at which Dp is leveled, is called a critical micelle concentration (cmc). At this point (0.3% AlCl3 in Fig. 20.4), the copolymer saturates the interphase. In the region below cmc, the copolymer formation is apparent from the reduction of Dp . However, beyond cmc the increase in reaction yield, as shown in Fig. 20.2, is not evident from this curve since Dp remains constant. In fact, after the saturation, the copolymer in excess migrates to the bulk, forming micelles, and does not take part in the compatibilization process (2,25,32). The blend morphology was studied from SEM images of cryogenic fractured surfaces of the blends. Dispersed phase particle shape and homogeneity, as well as interphase adhesion, were observed from the RB and PB micrographs. Figure 20.5 shows the morphological features of PB and RB. PS particles and the holes left by them during the fracture process can be observed for PB (Fig. 20.5a). The fracture is clearly interparticle and the borders are detached showing the poor adhesion between

608

Polyolefin Blends

Figure 20.5 SEM micrographs for PE/PS blends; (a) PB, (b) RB0.3, and (c) RB1.0.

PS and PE when the blends are noncompatibilized. In contrast, the compatibilized blends exhibit a considerably different morphology. The micrographs of RB with 0.3% and 1.0% catalyst (Fig. 20.5b and c) show a reduction in particle size and a different type of fracture, as compared with PB. The particles fracture along the same plane as the matrix (transparticle fracture) corresponding to an increase in adhesion between the phases. The interphase strength appears to be at least as high as the phases themselves. The adhesion increment is more pronounced for RB with 1.0% of catalyst (Fig. 20.5c). Here, a significant reduction in particle size dispersion is also observed, which is in agreement with the emulsifying effect presented before. Figure 20.5c also shows the effect of some elastic shrinking of PE chains after fracture, leading to ring-shaped edges around the particles. These edges, which are more prominent for RB1.0 than that for RB0.3, possibly contain graft copolymer material elongated during the fracture process and retracted afterward. The appearing of these ring edges in RB micrographs can be related to the cmc and the length of the grafted PE chains as follows. As it is known, the shorter molecules of homopolymer components are located preferably at the interphase and, due to their mobility, they are the most reactive ones (20–22). As the F-C reaction progresses, the homopolymer molecules are spent in generating copolymer. When the copolymer concentration at the interphase reaches cmc, a renewal process starts to occur. Copolymer molecules migrate to the bulk and new homopolymer molecules to the interphase, so that the last species consumed and formed are larger than the initial

Chapter 20 Reactive Compatibilization of Binary and Ternary Blends

609

ones (33). Now, for RB1.0 the cmc have been largely exceeded and interfacial copolymer molecules are expected to have longer PE graft chains than the ones for RB0.3; therefore, more marked ring edges appear. 20.3.1.3

Mechanical Behavior

The tensile test behavior of the physical and reactive blends, as well as their corresponding homopolymers, is analyzed. In assessing the compatibilization effect on the tensile properties, the elongation at break (eb ) was selected as a representative property. It involves large deformations, and then it is related to the interfacial adhesion and gives information about interfacial strength (34,35). The elastic modulus is less sensitive to compatibilization effects because it is a quasizero strain property and depends more on the inherent composition of the blend. The yield behavior could give qualitative information on the state of compatibilization, particularly when the changes in adhesion between compatibilized and PB are large. For blends based on crystallizable matrices, the mechanical properties also depend on the crystalline state (crystallinity degree, distribution of crystal sizes, morphology of crystalline structures, etc.). So, in order to properly compare the compatibilization effect through mechanical properties, the possible changes of the crystalline parameters must be assessed. Figure 20.6 shows the differential scanning calorimetry (DSC) thermograms of neat PE and the physical and reactive PE/PS blends studied. The PE and blend peak melting temperatures are very close to each other, only a minor reduction in peak broadness of the blends with respect to the pure PE is observed. These results reasonably allow assuming that the crystalline structure of the blends studied is not affected by the presence of the compatibilizer. A previous

Figure 20.6 DSC thermograms of PE and PE/PS physical and reactive blends with 0.3%, 0.7%, 1.0%, and 1.5% catalyst.

610

Polyolefin Blends

Figure 20.7 Tensile stress–strain results for PE and PE/PS physical and reactive blends with 0.3%, 0.7%, 1.0%, and 1.5% catalyst.

work, on similar PE/PS blend (36), showed that the compatibilizer only reduces the dispersed phase particle size, but does not have any effect on the crystalline structure. Figure 20.7 shows the tensile response of neat PE, PB, and RB. As expected, no effect of compatibilization on elastic modulus is detected; there is a perfect curve superposition for all the blends. However, the blend moduli are slightly higher than that of PE due to the presence of stiffer PS. The compatibilization effect is not negligible on yield behavior. The PB has yield strength approximately 20% less than neat PE, whereas reactive blends present slightly greater yield strengths than the matrix. This can be explained assuming that PE and PS are immiscible, so there is no continuity between the phases (Fig. 20.5a). Under this condition, during the tensile test, only the continuous PE phase (which is about 80% in volume) resists the load. Consequently, the PB yield strength is approximately 20% less than the neat PE one. On the contrary, the behavior of reactive blends is different. First, all the blends tested were at or above the cmc. It means that the PS particle surface is saturated by copolymer molecules and therefore compatibilized. The copolymer molecules are entangled with both phases, anchoring them. Now, during the tensile test, both phases contribute to the strength, resulting in a slightly higher yield point than that for neat PE. eb of RB shows a significant increase in catalyst content. It is clear, from Fig. 20.7, that eb for the blend with high content of catalyst (i.e., RB1.5) is close to eb of PE, and nearly three times larger than that for PB. It is interesting to note that eb increases even for blends that are far above the cmc condition. Taking into account that the crystalline structure remains unchanged, and neither chain scission nor cross-linking reactions occur, the possible causes for this behavior should be assigned to interphase modifications. It can be explained by considering the dynamics of the interfacial reaction process. As mentioned above, the shorter molecules tend to move to the interphase and react in the first place. Now, when the saturation concentration is exceeded, the copolymer is forced to leave the interphase driven by both thermodynamic and shear

Chapter 20 Reactive Compatibilization of Binary and Ternary Blends

611

forces (37,38). Afterward, new copolymer will be produced from longer homopolymer molecules, whenever catalyst remains at the interphase. So, RB1.5 should have higher MW copolymer at the interphase than RB0.3. In consequence, the entanglements in RB1.5 will be stronger than in RB0.3, leading to greater elongation at break.

20.3.2 PP/PS Blends 20.3.2.1

Chemical Aspects

In this binary blend, both PP and PS are susceptible to suffer degradation by chain scission. The presence of tertiary and benzyl carbon atoms in PP and PS, respectively, likely allows the event of b-cleavage reactions that lead to chain scission (39,40). Consequently, certain amount of low MW PP and PS can be expected as by-products of the F-C reaction. In order to determine the possible occurrence of chain scission and/or cross-linking side reactions, a systematic study was performed in the same way as it was done for PE/PS in Section 20.3.1.1. Table 20.4 contains the MW of neat and reactive homopolymers. It is noticed that reacted homopolymers exhibit MW considerably lower than those of unreacted ones, especially for catalyst concentrations of 0.7% and 1.0%. The MW reduction is a clear indication of chain scission degradation produced by the F-C catalyst. However, no evidence of cross-linking was detected. Also, there is no oxidative degradation of RPP, RPS, RB, and PB as was reported elsewhere from FTIR spectra and SEC chromatograms (26,33). The occurrence of the F-C reaction was checked by a combination of SEC characterization and selective solvent extraction. The resulting chromatograms from PB, RB, and their corresponding insoluble and soluble fractions were analyzed. Figure 20.8 shows that the SEC trace for RB0.7 is narrower than that for PB. The direct comparison of curves does not give a straightforward evidence of F-C reaction event because both homopolymers suffer the chain scission. Nevertheless, considering that shorter homopolymer molecules are located at the interphase and take part in the grafting reaction (see Section 20.3.1.1), a

Table 20.4 Molecular Weight of the Neat and Reactive Homopolymers.a AlCl3 concentration, % 0 (neat) 0.1 0.3 0.5 0.7 1.0

Mw RPP, g mol1 303,000 303,000 298,000 240,000 165,000 56,000

a

0.3% styrene, 200 C, 60 rpm, 24 min.

b

Out of detection range of the SEC columns employed.

Mw RPS, g mol1 260,000 275,000 260,000 228,000 <20,000b <20,000b

612

Polyolefin Blends

Figure 20.8 SEC chromatograms of PP/PS physical and reactive blend with 0.7% catalyst. (From Reference 41 with permission from John Wiley & Sons, Inc.)

hypothesis can be proposed. The MW reduction in the high MW region is due to the chain scission mainly of PS, while in the low MW extreme the reduction can be attributed to short length molecule consumption by grafting. To confirm this hypothesis, the PB and RB were subjected to extraction with a good solvent for PS (tetrahydrofuran) and the resulting soluble and insoluble fractions were analyzed by SEC. Using this solvent, it is expected that soluble fractions are rich in PS. On the contrary, since THF dissolves neither PP nor PP-g-PS, the insoluble fractions of RB must contain only PP and the copolymer molecules (if grafting has occurred). Table 20.5 shows that the extracted PS from soluble fraction of PB exhibits virtually the same MW as the neat PS. In contrast, the soluble fraction of reactive blend shows a considerable reduction in the MW of PS indicating PS chain scission. Table 20.5 Molecular Weights of Soluble Fractions of Physical and Reactive Blends and PP and PS, Neat and Reacted (0.7% catalyst). Sample

Mw, g mol1

PP PS RPP0.7 RPS0.7 SFPB SFRB

303,000 260,000 164,000 <20,000a 240,000 51,000

a

Out of detection range of the SEC columns employed.

Chapter 20 Reactive Compatibilization of Binary and Ternary Blends

613

Figure 20.9 SEC chromatograms for insoluble fractions of PP/PS physical and reactive blend with 0.7% catalyst. (From Reference 41 with permission from John Wiley & Sons, Inc.)

Figure 20.9 shows the chromatograms for the insoluble fractions of PB and RB. The insoluble fraction of RB curve is shifted to higher MW, relative to the insoluble fraction of PB trace. In consequence, since no cross-linking appeared, it can be inferred that this high MW fraction corresponds to PP-g-PS molecules, of greater MW than their homopolymers, formed by consumption of short PP and PS molecules. Considering that both PS and PP are subject to chain scission but not to crosslinking (Tables 20.4 and 20.5), and also that copolymer is formed (Fig. 20.9), the hypothesis related to Fig. 20.8 is confirmed. 20.3.2.2

Morphological Aspects

The morphology of PP/PS blends was studied following the emulsification behavior as explained in Section 20.3.1.2. Figure 20.10 shows that the emulsification curve follows a typical trace, which was frequently reported for compatibilization of immiscible blends (28–30). It is clear that after a significant drop in particle size, an equilibrium value is reached at about 0.7% AlCl3. This value has been taken as the cmc condition. It has to be remarked that the particle size decreases to one third of its initial value, reaching an equilibrium diameter of about 0.5 mm. Also, the particle size homogeneity increases with the catalyst content. It is shown by the decrease in error bars in Fig. 20.10. From these results, it is foreseen that the copolymer formed by the F-C reaction behaves as an efficient in situ compatibilizer for the PP/PS blend. Figure 20.11 shows micrographs of the fracture surfaces corresponding to PB, 0.3%, 0.7%, and 1.0% catalyst RB. An improvement in adhesion is clearly observed along with decrease in particle size, as the copolymer content increases. Particularly, it appears that starting from cmc (0.7% catalyst) the interphase becomes almost

614

Polyolefin Blends

Figure 20.10 Emulsification curve for PP/PS. Effect of the catalyst concentration on the average particle diameter (Dp ). Solid line corresponds to exponential fit. (From Reference 41 with permission from John Wiley & Sons, Inc.)

Figure 20.11 SEM micrographs from PP/PS blends. (a) PB, (b) RB0.3, (c) RB0.7, and (d) RB1.0. (From Reference 41 with permission from John Wiley & Sons, Inc.)

Chapter 20 Reactive Compatibilization of Binary and Ternary Blends

615

Figure 20.12 Variation of elongation at break and yield strength with decatalyst content for PP/PS blends under tensile test. (From Reference 42 with permission from Elsevier.)

undiscernible, indicating a very good adhesion that confirms the compatibilizing effect. 20.3.2.3

Mechanical Behavior

The properties of PP/PS blends were studied with the same considerations as for PE/PS blends (see Section 20.3.1.3). Figure 20.12 shows the elongation at break (eb ) and the yield strength as a function of the amount of catalyst used. eb shows a slight increase up to 0.3% catalyst. However, with 0.5% AlCl3 eb is about three times higher than that for PB. The maximum, which amounts to five times the reference value, is reached for 0.7%. Afterward, at 1.0% catalyst, eb drops steeply. The sharp increase in eb is consistent with the adhesion improvement shown in the micrographs up to 0.7% AlCl3 (Fig. 20.11). At higher catalyst contents, although the adhesion remains high, as shown in Fig. 20.11d, eb falls possibly due to the matrix degradation by chain scission (Table 20.4). The yield strength does not change with catalyst content. In spite of Dp decay to one third of the initial value for 1.0% catalyst (Fig. 20.10), it remained constant. The relatively low deformation, involved in the yield process, makes this property less sensitive to molecular disentanglement (and therefore to matrix degradation) than elongation at break.

20.4

TERNARY BLENDS: PE/PP/PS

Ternary blends of PE, PP, and PS were compatibilized by F-C in situ reaction. The compatibilization of this kind of immiscible blends is complex. There are three

616

Polyolefin Blends Table 20.6 Process Conditions Used in Reactive Compatibilization of Ternary Blends.

Temperature, C Speed, rpm Time, min Range of catalyst, %

PE/PP/PS

PE-b-PP/PP/PS

210 30 30 0.1–1.3

210 30 30 0.1–1.5

interphases PE–PS, PP–PS, and PP–PE to compatibilize. This reaction, which requires an aromatic ring, is applicable only to the interphases that contain PS. Consequently, the presence of an additional compatibilizer for PP–PE, that is, a copolymer having PE and PP chain ends, is needed. The F-C reaction was applied to both ternary PE/PP/PS and PE-b-PP/PP/PS blends with 70/20/10 composition. It was carried out in the molten state using the catalytic system AlCl3/styrene. Ternary physical blends (TPBs) were prepared by melt mixing (Banbury mixer), while ternary reactive blends (TRBs) were obtained by performing the F-C reaction on the TPB in the same mixer. The processing conditions were set up taking into account the particle size reduction and homogeneity of both minority phases. The blend morphological features were assessed from SEM images of cryogenic fractured samples. eb under tensile test was studied and the occurrence of reaction was verified through the emulsifying effect variation with the catalyst concentration. The blend preparation conditions are shown in Table 20.6. The PS phase was identified by comparing fracture surface SEM images of the PE/PP/PS physical blend with and without solvent extraction of PS. The respective micrographs in Fig. 20.13 show the PS particles and the empty sites, originally occupied by PS, which remained after extraction. The other particles must be PP. Figure 20.13 reveals that blend morphology is organized as a continuous PE and two dispersed (PP and PS) phases. Also, the PP–PS interphase is not observed. The PE/PP/PS physical blend presents a clear transparticle fracture mode of PP particles,

Figure 20.13 SEM micrographs of fracture surface of PE/PP/PS physical blend: (a) without extraction and (b) PS phase extracted with THF.

Chapter 20 Reactive Compatibilization of Binary and Ternary Blends

617

Figure 20.14 SEM micrographs of PE/PP/PS fracture surfaces of (a) TPB, (b) TRB0.1, (c) TRB0.5, and (d) TRB1.

while for PS particles both inter- and transparticle modes are observed. It appears, from these features, that PE–PP interphase would have better adhesion than PE–PS. Consistently, other authors report higher adhesion for PE and PP due to mechanical interlocking (13). Figure 20.14 shows the fracture surfaces of PE/PP/PS physical and reactive blends without extraction. A gradual increase in interfacial adhesion is observed from 0.1% catalyst and up (Fig. 20.14). The reduction in particle size is a clear indication that the F-C reaction has occurred between PE and PS. The PS average particle diameter in the TRB decreased six times with respect to the TPB. A reduction in PP particle size is also observed, particularly for the 1.0% catalyst blend (Fig. 20.14d), even though it could not be precisely measured. The PE–PP interphase does not present appreciable changes that allow its evaluation in these micrographs. Figure 20.15 shows eb of PE/PP/PS blends as a function of catalyst concentration. The TPB reveals a fragile behavior, with only 4% of eb . To assess the effect of PP in these TPBs, and taking into account their relative component amounts, this ternary blend can be considered as a binary 80/20 PE/PS blend with a small amount of PP. As was shown before (Fig. 20.7), the PE/PS physical blend exhibits eb of about 120%, much greater than the value for the ternary blend. It is clear that the low ductility of PE/PP/PS blend must be related to the PE–PP interphase.

618

Polyolefin Blends

Figure 20.15 Variation of elongation at break with decatalyst content for PE/PP/PS blends under tensile test.

The PE/PP/PS reactive blends show eb up to three times higher than that of TPB (Fig. 20.15). Taking into account that F-C reaction is not possible between PE and PP, and PP–PS interphase could not be detected (Fig. 20.13), the increase in ductility can be attributed to the compatibilization of the PE–PS interphase. Moreover, such behavior was also observed for the binary PE/PS blend (Fig. 20.7). The incidence of PP in the low ductility of ternary blends was confirmed by analyzing the behavior of binary 80/20 PE/PP blends. The results in Table 20.7 indicate that eb is strongly reduced by the presence of PP, even if the micrographs show transparticle fracture mode (Fig. 20.13). eb for PE/PP blend is 9%, comparable to 4% exhibited by the TPB, in which PE and PP are in a ratio close to 80/20. In order to improve the compatibility of PE–PP interphase the F-C reaction was applied to PE-b-PP/PP/PS physical blend, where the PP–PE interphase is already compatibilized by a block copolymer EPR (PE-b-PP). Figure 20.16 shows fractured surfaces of these modified blends before and after the F-C reaction. It has to be noted that in these PE-b-PP/PP/PS blends the continuous phase is PP. In Fig. 20.16a, for the TPB, the PS particles appear unbounded from the matrix and the fracture mode is interparticle indicating, as was expected, a poor adhesion at the PP–PS interphase. Table 20.7

Elongation at Break of PE, PP, and PE/PP Blends.

PE concentration, %

eb %

100 90 80 70 0

450 43 9 5 470

Chapter 20 Reactive Compatibilization of Binary and Ternary Blends

619

Figure 20.16 SEM micrographs of PE-b-PP/PP/PS fracture surfaces of (a) TPB and (b) TRB1.0.

The PE particles were fractured along with the PP matrix in a transparticle mode, consistently with PE–PP compatibilization via EPR copolymer. As the catalyst content increases, the PS particle diameter decreases and the fracture becomes progressively transparticle (Fig. 20.16b). This indicates that the PP–PS and PE–PS are being compatibilized. The results of ductility are presented in Fig. 20.17 as a function of catalyst concentration. eb of PE-b-PP/PP/PS reactive blends shows a maximum for 0.5% AlCl3, which is three times greater than the value for the corresponding TPB, and then a drop at higher catalyst concentrations. This drop can be assigned to PP chain scission in the same way as above for the PP/PS blends (Fig. 20.12).

Figure 20.17 Variation of elongation at break with decatalyst content for PE-b-PP/PP/PS blends under tensile test.

620

Polyolefin Blends

eb of PE-b-PP/PP/PS with 0.5% catalyst, when all the interphases are compatibilized, reaches about 160%, while the PE/PP/PS physical blend showed only 4%. Then the F-C reaction compatibilization considerably improves adhesion between the phases and the mechanical performance at large deformations. Finally, it has to be remarked that all the interphases must be compatibilized and, for the application of F-C alkylation to reactive compatibilization, the interphases must contain species with benzene rings.

20.5 CONCLUSIONS The compatibilization of binary PE/PS and PP/PS blends by the F-C reaction was shown to be simple and effective. This route offers a proper combination of one-step and solvent-free process that uses a low amount of low cost reactants. The copolymer formed by the F-C reaction, even with catalyst concentrations as low as 0.3%, promotes a marked emulsifying effect and different levels of improvement in elongation at break. On the contrary, at high catalyst concentrations, PP and PS may undergo chain scission impairing the mechanical behavior improvement. Nevertheless, a considerable degree of compatibilization is reached for these blends before chain scission occurs, indicating that an optimal catalyst concentration should be selected to achieve the best performance of these blends. Ternary blends showed a considerable improvement in mechanical behavior only when all the interphases were compatibilized. Other alternatives are currently being investigated to apply the F-C reaction to compatibilize the different phases in ternary blends.

NOMENCLATURE eb AlCl3 cmc Dp DSC EPR F-C FTIR IFPB IFRB LLDPE MW Mw PB

Elongation at break Aluminum trichloride Critical micelle concentration Average particle diameter Differential scanning calorimetry Ethylene–propylene rubber Friedel–Crafts alkylation reaction Fourier transformed infrared spectroscopy Insoluble fraction resulting from solvent extraction of physical blend Insoluble fraction resulting from solvent extraction of reactive blend Linear low density polyethylene Molecular weight Weight-average molecular weight Physical blends of polyolefin/PS

Chapter 20 Reactive Compatibilization of Binary and Ternary Blends

PE/PP/PS PE/PS PE PE-b-PP/PP/PS PE-b-PP PE-g-PS PE–PP PE–PS PP/PS PP PP-g-PS PP–PS PS PVC RI RB RB0.1 RB0.3 RB0.5 RB0.7 RB1.0 RB1.5 RPE RPP RPS SEC THF TPB TRB TRB0.1 TRB0.5 TRB1.0

621

PE/PP/PS ternary physical blend PE/PS binary blend Polyethylene Ternary physical blend Ethylene–propylene block copolymer Polyethylene–polystyrene graft copolymer Interphase between PE and PP Interphase between PE and PS PP/PS binary blend Polypropylene Polypropylene–polystyrene graft copolymer Interphase between PP and PS Polystyrene Polyvinyl chloride Refractive index Reactive blends of polyolefin/PS (80/20) subject to F-C reaction RB with 0.1% AlCl3 and 0.3% styrene RB with 0.3% AlCl3 and 0.3% styrene RB with 0.5% AlCl3 and 0.3% styrene RB with 0.7% AlCl3 and 0.3% styrene RB with 1.0% AlCl3 and 0.3% styrene RB with 1.5% AlCl3 and 0.3% styrene Reactive polyethylene. PE subject to F-C reaction Reactive polypropylene. PP subject to F-C reaction Reactive polystyrene. PS subject to F-C reaction Size exclusion chromatography Tetrahydrofuran (good solvent for PS) Ternary physical blend Ternary reactive blend. TPB subject to F-C reaction TRB with 0.1% AlCl3 and 0.3% styrene TRB with 0.5% AlCl3 and 0.3% styrene TRB with 1.0% AlCl3 and 0.3% styrene

REFERENCES 1. R. Hudson, Commodity Plastics—As Engineering Materials?, Rapra Tech. Ltd, Shawbury, 1995. 2. L. A. Utracki, Polymer Alloys and Blends, Hanser, Munich, 1989. 3. D. R. Paul and C.B. Bucknall (eds.), Polymer Blends, Wiley, New York, 2000. 4. B. Pukansky, Eur. Polym. J., 41, 645 (2005). 5. A. T. Bisio and M. Xantos, How to Manage Plastics Waste: Technology and Market Opportunities, Hanser, Munich, 1995. 6. J. Brandrup, M. Bittner, W. Michaeli, and G. Menges, Recycling and Recovery of Plastics, Hanser, Munich, 1996. 7. C. Konig, M. Van Duin, C. Pagnoulle, and R. Jerome, Prog. Polym. Sci., 23, 707 (1998).

622

Polyolefin Blends

8. L. A. Utracki, Introduction, in: Polymer Blends Handbook, L. A. Utracki (ed.), Kluwer Academic Publishers, Netherlands, 2003. 9. J. Markarian, Plast. Adit. Compd., 6, 22 (2004). 10. L. A. Utracki, Commercial Polymer Blends, Chapman and Hall, Munich, 1998. 11. A. Echte, Rubber-toughened plastics, in: Advances in Chemistry Series, Vol. 222, C. K. Riew (ed.), American Chemical Society, Washington, DC 1989, p. 15. 12. S. A. Datta and D. J. Lohse, Polymeric Compatibilizers. Uses and Benefits in Polymer Blends, Hanser, Munich, 1996. 13. L. H. Sperling, Polymeric Multicomponent Materials, Wiley, New York, 1997. 14. B. Majundar and D. R. Paul, Reactive compatibilization, in: Polymer Blends, D. R. Paul and C. B. Bucknall (eds.), Wiley, New York, 2000, Chapter 17. 15. U. Sundararaj and C. Makosco, Macromolecules, 28, 2647 (1995). 16. G. Solomons, Organic Chemistry, 5th edition, Wiley, New York, 1992. 17. W. L. Carrick, J. Polym. Sci. Chem, 8, 215 (1970). 18. Y. Sun and W. E. Baker, J. Appl. Polym. Sci., 65, 385 (1997). 19. Y. Sun, R. Willemse, T. Liu, and W. Baker, Polymer, 39, 2201 (1998). 20. H. Helfand and K. Tagami. Polym. Lett., 9, 741 (1971). 21. E. J. Kramer, Isr. J. Chem., 35, 49 (1995). 22. B. O’Shaughnessy and U. Sawhney, Macromolecules, 29, 7230 (1996). 23. P. Charoensirisomboon, T. Chiba, S. I. Solomko, T. Inoue, and M. Weber, Polymer, 40, 6803 (1999). 24. P. Charoensirisomboon, T. Inoue, and M. Weber, Polymer, 41, 4483 (2000). 25. P. Charoensirisomboon, T. Inoue, and M. Weber, Polymer, 41, 6907 (2000). 26. M. Dı´az, S. Barbosa, and N. Capiati, Polymer, 43, 4851 (2002). 27. S. Barbosa, M. Dı´az, G. Mabe, N. Capiati, and E. Brignole, J. Polym. Sci. Polym. Phys, 43, 2361 (2005). 28. T. Tang and B. Huang, Polymer, 35, 281 (1994). 29. J. Li and B. D. Favis, Polymer, 43, 4935 (2002); 30. P. Macau´bas and N. Demarquette, Polymer, 42, 2543 (2001). 31. B. D. Favis, Factors influencing the morphology of immiscible polymer blends in melt processing, in: Polymer Blends, D. R. Paul and C. B. Bucknall (eds.), Wiley, New York, 2000, Chapter 16. 32. S. D. Hudson and A. M. Jamieson, Morphology and properties of blends containing block copolymers, in: Polymer Blends, D. R. Paul and C. B. Bucknall (eds.), Wiley, New York, 2000, Chapter 15. 33. M. Dı´az, Compatibilizacio´n de Mezclas de PE, PP y PS. Aplicacio´n de la Reaccio´n de Alquilacio´n de Friedel–Crafts, PhD thesis, Universidad Nacional del Sur (UNS), Argentina, 2004. 34. I. M. Ward and D. W. Hadley, An Introduction to the Mechanical Properties of Solid Polymers, Wiley, New York, 1993. 35. L. E. Nielsen and R. F. Landel, Mechanical Properties of Polymers and Composites, Dekker, New York, 1994. 36. T. Kim, D. Kim, W. Kim, T. Lee, and K. Suh, J. Polym. Sci. Polym. Phys., 42, 2813 (2004). 37. J. Noolandi, Polym. Eng. Sci., 24, 70 (1984). 38. L. Pan, T. Inoue, H. Hayami, and S. Nishikama, Polymer, 43, 337 (2002). 39. R. Constable, Chemical coupling agents for filled and glass reinforced polypropylene composites, in: Handbook of Polypropylene and Polypropylene Composites, H. G. Karian (ed.), Dekker, New York, 1999. 40. B. Pukanszky, J. Kennedy, T. Kelen, and F. Tu¨do˜s, Polym. Bull., 5, 469 (1981). 41. M. Dı´az, S. Barbosa, and N. Capiati, J. Polym. Sci. B Polym. Phys., 42, 452 (2004). 42. M. Dı´az, S. Barbosa, and N. Capiati, Polymer, 46, 6096 (2005).

Chapter

21

Polyolefin/Epoxy Resin Blends Bejoy Francis1 and Sabu Thomas2

21.1 INTRODUCTION Epoxy resins are commercially important thermosetting polymers available as liquids or as tack-free solids. Epoxy resins have several advantages over other thermosetting resins. For example, in the case of epoxy resin, cured products with excellent mechanical properties and chemical resistance can be obtained without the evolution of volatile products. Epoxide group can react with a variety of substrates at various reaction conditions to give a range of properties. Thus, it is possible to tune the ultimate properties by proper selection of curing agent and curing conditions (1). All polymeric materials are prone to one or the other disadvantages. Likewise, epoxy resins are prone to brittle failure. Therefore, the brittleness of epoxy resin has to be reduced to use them in structural applications. Research in this direction led to the development of several methods to decrease the brittleness of epoxy resin. Elastomers like carboxyl and amine terminated butadiene acrylonitrile rubber (CTBN and ATBN) (2–9), hydroxyl terminated butadiene acrylonitrile random copolymer (HTBN) (10), epoxidized hydroxyl terminated polybutadiene (EHTPB) (11), carboxyl terminated polybutadiene (CTPB) (12), hydroxyl terminated polybutadiene (HTPB) (8), polysiloxanes (13,14), triethoxysilyl terminated polycaprolactone elastomer (PCL-TESi) (15) and so on were successfully used to improve the toughness of epoxy resin. A disadvantage of these blend systems is the reduction in the thermomechanical properties of the cured product compared to neat resin. Of late thermoplastics like polysulfones (16–20), polyether ether ketones (21,22), polyetherimides (23–26), polyesters (27–32), polyethylene oxide (33–35), polymethylmethacrylate (36,37), polycarbonate (38–40), poly ecaprolactone (41–44), polyethernitrile (45), or modifications of these polymers were 1 Department of Chemistry and Biochemistry, Laurentian University, 935 Ramsey Lake Road, Sudbury, Ontario, P3E 2C6, Canada 2 School of Chemical Sciences, Mahatma Gandhi University, Priyadarshini Hills, Kottayam 686560, Kerala, India

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

623

624

Polyolefin Blends

extensively used to modify epoxy resin. In-depth studies on the miscibility, cure kinetics, phase morphology, and thermomechanical properties were carried out during the past two decades. Thermoplastic modification of epoxy resin resulted in tough systems without reduction in thermal stability or modulus. Polyolefin modified epoxy resin had been the subject of many investigations in the past few years (46–48). Polyolefins like polyethylene (PE), polypropylene (PP), or their modifications were used to blend with epoxy resin. Semicrystalline materials like polyethylene or polypropylene were selected because these materials do not adversely affect the stiffness and strength of the cured epoxy resin. Contrary to amorphous thermoplastics, these polymers can be precipitated in the matrix by crystallization below their melting points. Also polyolefins are softer than other thermoplastics like PET and PBT. Hence, these are expected to show better improvement in fracture toughness. The morphology of the blends had a profound effect on the ultimate properties of the system. A two-phase morphology is essential for improving the fracture toughness. A dispersed phase with no interaction or a completely miscible system deteriorates the fracture toughness, since the second component is unable to dissipate the energy. Therefore, a two-phase system with some interaction between the component polymers is preferred. This is achieved by the functionalization of the dispersed phase. The morphology was dependent on the composition of the blends. By adjusting the composition, dispersed, cocontinuous, or phase-inverted systems with different properties can be generated. Thus, it is possible to derive a system with specific properties depending on the final use. These possibilities are explored for epoxy/polyolefin blends. This chapter focuses on the preparation, morphology, miscibility, and properties of epoxy resin modified with polyolefins.

21.2 BLEND PREPARATION Melt-mixing technique is often used for the preparation of epoxy/polyolefin blends. Slight variations exist between the processing techniques of various blend systems. For example, functionalized polyethylene was mixed with epoxy resin at 115 C, the curing agent hexahydrophthalic anhydride (HHPA) or methyl nadic anhydride (MNA) was dissolved in the blend at the same temperature and the catalyst N,Ndimethylbezylamine (BDMA) was added to the blends. The blends were cured and postcured at various temperature cycles depending on the amount of catalyst (46). Polypropylene was mixed with epoxy resin in a Haake RC90 rheometer at 190 C and 50 rpm. Curing agent, 2-ethylene-4-methane imidazole (EMI-24) was added and the blends were finally compression molded. The amount of epoxy resin was less than that of polypropylene in the blends and PP-g-MA was added as a reactive compatibilizer (48). The carbon black filled polypropylene/epoxy blends was also reported to be prepared in a similar manner. In the case of polypropylene carbonate, the following methodology was used. Polypropylene carbonate was dissolved in epoxy resin at 80 C followed by the curing agent methyltetrahydrophthalic anhydride (MTHPA) and the accelerator 2,4,6-tris(dimethylaminomethyl)phenol (DMP-30). The blends were cured at 80 C for 17 h and 110 C for 28 h in a teflon mold (47).

Chapter 21

21.3

Polyolefin/Epoxy Resin Blends

625

MISCIBILITY STUDIES AND PHASE DIAGRAMS

An efficient modifier for epoxy resin usually forms a homogeneous blend with epoxy resin before curing and phase separates upon curing as a result of the increase in molecular weight of epoxy resin. This process is referred to as reaction-induced phase separation (RIPS) in epoxy resin technology. Semicrystalline thermoplastics like polyethylene and polypropylene are immiscible in epoxy resin. Convenient methods to improve the miscibility are to reduce the molecular weight of the polymer and to introduce functional group or bulky pendent groups on the polymer chain, which will alter the semicrystalline behavior of the polymer. A typical example is polyether–ether ketone (PEEK). Amine or hydroxyl terminated PEEK or phenolphthalein PEEK were found to form homogeneous blends with epoxy resin and curing with various curing agents resulted in reaction-induced phase separation (49–52). In a similar fashion, a series of low molecular weight functionalized polyethylene copolymers were synthesized and used to modify epoxy resin (46). They are (i) oxidized homopolymer with an acid functionality of 0.26 (AC 6702), (ii) two ethylene acrylic acid copolymers with functionalities approximately 1.2 and 2.0 (AC 540 and AC 5120), (iii) an ethylene acrylic acid–vinyl acetate terpolymer (AC 1450), having an acid functionality of approximately 2 and an equivalent weight of carboxylic groups equal to 1200, and (iv) an ethylene acrylic acid–vinyl acetate– vinyl alcohol terpolymer (AC 80), containing both hydroxyl and carboxyl groups with an average acid functionality equal to the average OH functionality, both approximately equal to 1 (the equivalent weight of OH groups was 485–750, while the equivalent weight of COOH was 1194). Some of the modified polyethylenes used for blending with epoxy resin and its properties are summarized in Table 21.1 and the chemical structures are given in Fig. 21.1 (46). Several other modification of the abovementioned polymers were also made to improve the processability. The acrylic acid copolymer AC 5120 was reacted with e-caprolactone to produce side group extension while maintaining the original acid group. Second modification of AC 5120 was made by reacting the acid groups with p-t-butyl phenol–glycidyl ether (Eurepox RVP). This reaction produced hydroxyl groups in the b position adjacent to the ester groups at the Table 21.1 Chemical and Physical Properties of Low Molecular Weight Functionalized Polyethylenea. Acid no. AC 6702 Ac 540 AC 5120 AC 1450 AC 80

15 40 120 35 42–46

f 0.26 1.2 2.1 2.0 1

Hydroxyl no.

f0

T m, C

Mw

Mn

Mw/Mn

— — — — 75–115

0 0 0 0 1

63 105 89 88 55

2140 4560 3325

970 1710 1005 a a

2.2 2.6 3.3

f ¼ acid functionality, calculated from Mn, f 0 ¼ OH functionality. a Number-average molecular weight (Mn) is expected to be in the region of 2000–4000 for AC 1450 and 1000–2000 for AC80. [From Reference 46 with permission from Wiley Interscience.]

626

Polyolefin Blends (CH2

CH2 )n (CH2

CH )m COOH

AC 540 and 5120

COOCH3 (CH2

CH2 )n (CH2

CH )m (CH

CH2 )p

COOH AC 1450

COOCH3 (CH

CH2 )

(CH2

CH2 )n (CH2

OH

CH )m (CH

CH2 )p

COOH AC 80

Figure 21.1 Chemical structure of functionalized polyethylenes. (From Reference 46 with permission from Wiley Interscience.)

junctions with the ethylene oligomer chains. To increase the reactivity of the ethylene–acrylic acid copolymer AC 5120 modified with p-t-butyl phenol-glycidyl ether (RVP) toward the epoxy resin, the hydroxyl groups produced in the above step were converted again to acid groups by reaction with a 1:1 molar ratio (plus a 10 mol% excess) of hexahydrophthalic anhydride (HHPA). RVP and p-t-butyl phenol-glycidyl ether were also used to enhance the solubility of functionalized terpolymers, AC 1450 and AC 80, by reacting RVP with AC 1450 at a molar ratio of 2:1, and with AC 80 at a molar ratio of 1:1, respectively. In a second step, the adducts of AC 1450 and AC 80 with RVP were reacted further with chlorendic anhydride; 1,4,5,6,7,7-hexachloro-5-norborene-2,3-dicarboxylic anhydride (CA). A further step in the modification of the oligomers to enhance their miscibility with epoxy resins was the subsequent reaction with e-caprolactone, to produce AC 80-RVP-CA-CP and AC 1450-RVP-CA-CP. The miscibilities of these polymers in 50/50 blend with a liquid diglycidyl ether of bisphenol-A (DGEBA), epoxy resin (epoxide equivalent weight 184–190), and a cycloaliphatic resin (epoxide equivalent weight 135) were given in Table 21.2. The polyethylenes with higher functionality were soluble in epoxy resin and required lower temperature and time for forming homogeneous blend systems. The miscibility of the polymers was dependent on the type of epoxy resin also. Cycloaliphatic epoxy resin showed more miscibility with the polymers compared to DGEBA resin and phase separation occurred in these blend systems as a result of crystallization of PE. Upper critical solution temperature (UCST) behavior was

Chapter 21

627

Polyolefin/Epoxy Resin Blends

Table 21.2 Solubilization Temperature on Heating (Ts) and Phase-Separation Temperature on Cooling (Tp) for 50/50 Mixtures of AC 6702, AC 540, Unmodified, and Modified AC 5120, AC 1450, and AC 80, with Unmodified and Modified DGEBA Resin (Epikote 828) or Cycloaliphatic Resin (Araldite CY 179). Formulations AC 6702/Epikote 828 AC 540/Epikote 828 AC 5120/Epikote 828 (AC 5120-g-caprolactone 50/5)/Epikote 828 (AC 5120-g-caprolactone 50/20)/Epikote 828 (AC 5120-g-RVP 69.8/30.2)/Epikote 828 (AC 5120-g-RVP/TPP 69.8/30.2/1)/Epikote 828 AC 5120/(Epikote 828-g-montanic acid 6.6/1) (AC 5120-g-caprolactone 50/5)/(Epikote 828-g-montanic acid 6.6/1) (AC 5120-g-caprolactone 50/20)/(Epikote 828-g-montanic acid 6.6/1) AC 540/Araldite CY 179 AC 5120/Araldite CY 179 (AC 5120-g-RVP/TPP 69.8/30.2/1)/Araldite CY 179 AC 1450/Epikote 828 (AC 1450-g-RVP/TPP 83/17/1)/Epikote 828 AC 80/Epikote 828 (AC 80-g-RVP/TPP 83/17/1)/Epikote 828 [(AC 80-g-RVP/TPP)-g-CA 70/30]/Epikote 828 [(AC 80-g-RVP/TPP-g-CA)-g-caprolactone 76/24)/Epikote 828 (AC 80-g-CA/TPP 63/37/1)/Epikote 828

Ts, C

Tp, C

273 236 190 160 198 168 146 158 213 238 100 88 87 211 160 167 151 142 90 128

246 188 74a 78a 72a 82a 86a 80a 74a 205 95a 79a 76 110 102 70a 70 112 72 67a

a

Phase separation occurs concomitantly with crystallization. (From Reference 46 with permission from Wiley Interscience.)

observed in blend systems comprising acrylic acid copolymer and DGEBA resin. The acrylic acid copolymer AC 5120 and its modification with RVP were the most effective in improving the miscibility of PE in epoxy resin. The RVP grafted polymer was miscible in epoxy resin at 146 and 86 C and at 87 and 76 C, respectively, in DGEBA and cycloaliphatic resin and reappeared on cooling. The presence of hardener hexahydrophthalic anhydride (HHPA) in the blends completely changed the miscibility behavior of the blends. All the blends became heterogeneous upon adding HHPA. The phase behavior of the ternary blends was evident from the phase diagrams given in Fig. 21.2. Comparing the phase diagrams of AC 540 modified DGEBA epoxy resin and hardener HHPA showed that the oxidized homopolymer with an acid functionality of 2.0 (AC 5120) was more miscible in epoxy resin and the RVP grafted AC 5120 had comparable miscibility as that of AC 5120. Similarly, ethylene acrylic acid–vinyl acetate terpolymer (AC 1450) was less miscible in epoxy resin compared to AC 5120 due to less chemical interaction with epoxy resin. Thus, the miscibility of the blends was reduced by the addition of HHPA to various blend systems.

628

Polyolefin Blends

Figure 21.2 Miscibility diagrams for (a) AC 540/Epikote 828/HHPA, (b) AC 5120/Epikote 828/ HHPA, and (c) for AC 5120 grafted with RVP/Epikote 828/HHPA mixtures. (From Reference 46 with permission from Wiley Interscience.)

Chapter 21

21.4

Polyolefin/Epoxy Resin Blends

629

CURE KINETICS

The cure kinetics of diglycidyl ether of bisphenol-A (DGEBA) epoxy resin modified with polypropylene carbonate (PPC) and cured with methyltetrahydrophthalic anhydride (MTHPA) was studied using the traditional differential scanning calorimetric (DSC) and by Fourier transform infrared spectroscopic (FTIR) method (47). The advantage of FTIR in the kinetic studies is that side reactions, if any, during curing could be followed, which is not possible in DSC studies since it measures the overall heat change in the system. The DSC thermograms of DGEBA/PPC/MTHPA system at different heating rates are shown in Fig. 21.3. The activation energy for the reaction calculated using Kissinger equation was 52.9 kJ mol1 and the order of reaction (n) was 0.92. The peculiarity of this system is that two reactions, namely, the reaction between epoxy resin and polypropylene carbonate or epoxy resin and MTHPA shown in Scheme 21.1a and b, respectively, could occur. In the FTIR method, the extent of reaction was determined by monitoring the intensity of the absorption at 1855 cm1 corresponding to the vibration of carbonyl group in MTHPA and the reference band was 1745 cm1 corresponding to the vibration of carbonyl group on the PPC chain. The ratio of the intensities 1855/ 1754 was used to characterize the fraction reacted. For the reaction between epoxy resin and MTHPA, the detected bands were 919 cm1 corresponding to epoxy group and the one at 833 cm1 corresponding to the ortho hydrogen on the aromatic rings of the epoxy resin and the ratio 919/833 was calculated. The reaction rate constant and the activation energies from the FTIR studies are summarized in Table 21.3. From the

Figure 21.3 DSC thermogram for uncured PPC/EP/MTHPA mixtures at different scanning rates (molecular weight of PPC: 4900). (From Reference 47 with permission from Wiley Interscience.)

630

Polyolefin Blends

(a)

CH

O

OH

O CH

CH2

EP

CH

O CH2

n

EP

CH

C

CH2 + CH3

O

O

C O

O

CH

CH

CH2

EP

CH

CH2

n

EP

CH

CH2

O

CH3

OH

C

O

C

CH2

O

CH

CH2

EP

O

CH

CH2

n

EP

CH

CH2

OH

CH3 EP =

O

CH

O

CH2

CH3

(b) O

O

HO

( PPC )

C

OH +

C CH3

O CH3

O

C

OH

C

O

PPC

OH

O

Scheme 21.1 (a) Reaction between epoxy resin and MTHPA and (b) reaction between epoxy resin and PPC. (From Reference 47 with permission from Wiley Interscience.)

Table 21.3 Reaction Rate Constant and Activation Energies for PPC/MTHPA, EP/ MTHPA, and PPC/EP/MTHPA Systems. Method FTIR

System

T, C

k, min1

PPC/MTHPA

80 90 70 80 90 100

7:40 104 2:25 104 3:24 104 6:00 104 1:20 104 1:97 104

EP/MTHPA

DSC

PPC/EP/MTHPA

[From Reference 47 with permission from Wiley Interscience.]

Ea, kJ mol1 115.8

66.5 52.9

Chapter 21

Polyolefin/Epoxy Resin Blends

631

activation energy values, the epoxy curing agent reaction was found to occur faster and have lower activation energy than the PPC/MTHPA reaction. Comparing the FTIR and DSC kinetics, it was confirmed that the curing reaction proceeded by the reaction between epoxy and MTHPA, but it is likely that some of the polypropylene carbonate reacted with anhydride as well.

21.5

CRYSTALLIZATION BEHAVIOR

The crystallization behavior of polypropylene modified epoxy resin was studied extensively. The crystallization of polypropylene was influenced by the addition of epoxy resin. The first cooling and second heating DSC thermograms of polypropylene, polypropylene/epoxy, polypropylene/MAH-g-PP, polypropylene/MAH-gPP/epoxy, and dynamically cured polypropylene/epoxy blends are shown in Figs. 21.4 and 21.5 and the data obtained from DSC studies are summarized in Table 21.4 (48). The crystallization peak temperature increased when epoxy resin was added to polypropylene, that is, the uncured and cured epoxy particles acted as effective nucleating agents and accelerated the crystallization of PP in the blends. When cured dynamically, the smaller epoxy particles in the blends resulted in the increase in the number of nucleating agents and hence accelerated the crystallization of polypropylene. Blending polypropylene with epoxy resin resulted in the decrease of crystallinity of polypropylene and increased the melting temperature (Tm) of polypropylene than those of pure polypropylene. The polarizing optical micrographs of polypropylene, polypropylene/epoxy (70/30), PP/MAH-g-PP/epoxy (60/10/30), and dynamically cured blends (60/10/ 30/1.2) are shown in Fig. 21.6. The spherulites of polypropylene crystals (80 mm) were larger than those of polypropylene/epoxy blends. This is because epoxy resin particles acted as nucleating sites, increased the number and decreased the size of PP spherulites in the blends. The normalized dynamic DSC cooling curves of polypropylene/epoxy blends are shown in Fig. 21.7 and the data from the DSC studies are summarized in Table 21.5. All the systems studied showed a single crystallization peak (Tp) and the Tps increased with the addition of epoxy resin to polypropylene (53). The Tp of dynamically cured blends varied slightly with epoxy resin content in the blends. Tonset Tp values were the highest for pure polypropylene than the corresponding values from the blends indicating the lowest crystallization rate. Dynamically cured blends had lower Tonset Tp values compared to other blends. Therefore, the overall crystallization rate in dynamically cured blends was greater than that of the other polypropylene/epoxy blends. The rate of nucleation (Si) was highest for dynamically cured blends due to smaller size of the epoxy-dispersed phase. The crystallite size distribution (DW) was the smallest for dynamically cured samples compared to polypropylene and polypropylene/epoxy blends. The

632

Polyolefin Blends

Figure 21.4 DSC thermograms at the first cooling of (a) PP, PP/MAH-g-PP, PP/epoxy, and PP/MAHg-PP/epoxy and (b) dynamically cured PP/epoxy blends. (From Reference 48 with permission from Wiley Interscience.)

isothermal DSC curves polypropylene and polypropylene/epoxy blends are shown in Fig. 21.8. The isothermal crystallization kinetics of the dynamically cured blends was analyzed using Avrami’s equation as given below. logf ln½1 XðtÞg ¼ n log t þ log KðTÞ

ð21:1Þ

Chapter 21

Polyolefin/Epoxy Resin Blends

633

Figure 21.5 DSC thermograms at the second heating of (a) PP, PP/MAH-g-PP, PP/epoxy, and PP/MAH-g-PP/epoxy and (b) dynamically cured PP/epoxy blends. (From Reference 48 with permission from Wiley Interscience.)

where X(t) is the relative crystallinity at different crystallization times, n is a constant depending on the mechanism of nucleation and the form of crystal growth and K(T) is a constant related to nucleation and growth parameters. The kinetic parameters of dynamically cured blends are given in Table 21.6. K(T) was higher for the blends and maximum was observed for 20% epoxy containing blend. The slightly lower value for 30% blend showed that epoxy-rich particles restricted the mobility of PP segments above 20% epoxy content. The value of K(T) of the dynamically cured blends suggested that the crystallization was more dependent on the temperature.

634

Polyolefin Blends

Table 21.4

Data Obtained from DSC Analysis.

Composition

Tm, C

Tc, C

PP 60/10 PP/MAH-g-PP 70/30 PP/epoxy 60/10/30 PP/MAH-g-PP/epoxy 85/10/5/0.2 PP/MAH-g-PP/epoxy/EMI-2,4 80/10/10/0.4 PP/MAH-g-PP/epoxy/EMI-2,4 70/10/20/0.8 PP/MAH-g-PP/epoxy/EMI-2,4 60/10/30/1.2 PP/MAH-g-PP/epoxy/EMI-2,4 40/10/50/2 PP/MAH-g-PP/epoxy/EMI-2,4

162.3 162.6 163.5 164.9 163.7 164.8 165.4 165.0 164.8

113.1 114.2 123.2 131.3 125.8 127.3 130.0 128.6 129.5

Xc wt % 56.1 56.5 53.3 50.7 46.7 50.1 47.8 49.5 39.5

Tm, peak melting temperature; Tc, peak crystallization temperature. [From Reference 48 with permission from Wiley Interscience.]

The value of n depends on the nucleation process and geometry of the growing crystals (54,55). The value of n increased corresponding to an increase in crystallization temperature (Tc). The increase in n was due to the change from instantaneous nucleation to sporadic nucleation. The n value remained close to 3 at all compositions

Figure 21.6 Polarized optical micrographs of PP, PP/epoxy, PP/MAH-g-PP/epoxy, and dynamically cured PP/epoxy blends at 136 C for 30 min: (a) PP, (b) 70/30 PP/epoxy, (c) 60/10/30 PP/MAH-g-PP/ epoxy, and (d) 60/10/30/1.2 PP/MAH-g-PP/epoxy/EMI-2,4. (From Reference 48 with permission from Wiley Interscience.)

Chapter 21

Polyolefin/Epoxy Resin Blends

635

Figure 21.7 Normalized DSC thermograms for PP and PP/epoxy blends prepared by nonisothermal crystallization at a cooling rate of 10 C min1 : (a) PP, (b) 70/30 PP/epoxy, (c) 60/10/30 PP/MAH-g-PP/ epoxy, (d) 80/10/10/0.4 PP/MAH-g-PP/epoxy/EMI-2,4, (e) 70/10/20/0.8 PP/MAH-g-PP/epoxy/EMI-2,4, and (f) 60/10/30/1.2 PP/MAH-g-PP/epoxy/EMI-2,4. (From Reference 53 with permission from Wiley Interscience.)

suggesting a three-dimensional spherical growth in the spherical form (56). The halftime for polypropylene crystallization (t1/2) decreased considerably in the dynamically cured blends. This was due to the nucleating effect of epoxy particles on polypropylene crystallization and at a given blend composition t1/2 increased with increase in Tc. The crystallization thermodynamics and kinetics of the samples were analyzed based on the secondary nucleation theory of Hoffman and Lauritzen

Table 21.5 Various Parameters of PP and PP/Epoxy Blends Obtained from the Nonisothermal Crystallization Exotherm at a Cooling Rate of 10 C min1 . Composition PP PP/epoxy (70/30) PP/MAH-g-PP/epoxy (60/10/30) PP/MAH-g-PP/epoxy/EMI-2,4 (80/10/10/0.4) PP/MAH-g-PP/epoxy/EMI-2,4 (70/10/20/0.8) PP/MAH-g-PP/epoxy/EMI-2,4 (60/10/30/1.2)

Tp, C

Tonset, C

Tonset Tp , C

Si

DW

112.1 124.4

120.8 128.8

8.7 4.4

0.51 1.48

4.0 3.7

127.3

131.3

4.0

2.05

3.2

127.4

131.6

4.2

1.89

3.0

129.4

133.3

3.9

2.56

2.8

128.6

132.4

3.8

2.17

3.0

[From Reference 53 with permission from Wiley Interscience.]

636

Polyolefin Blends

Figure 21.8 Heat flow versus time during isothermal crystallization at different temperatures: (a) PP and (b) 80/10/10/0.4 PP/MAH-g-PP/epoxy/EMI-2,4. (From Reference 53 with permission from Wiley Interscience.)

(57,58). The regime III crystallization was applied, which can be expressed as follows: ð1=nÞ log KðTÞ þ DF=2:3 RT c ¼ A0 ð4b0 ss e T m Þ=ð2:3 kB DH f T c DTÞ

ð21:2Þ

where DT is equal to Tm0 Tc; s and se are the free energies per unit area of the surfaces of the lamellae parallel and perpendicular to the chain direction, respectively, b0 is the distance between two adjacent fold planes, and kB is the Boltzmann constant. The se of pure polypropylene was higher than that of dynamically cured polypropylene/epoxy blends. In all cases, Tm of polypropylene increased as Tc increased, and this was directly related to the size of the polypropylene crystals. The free energy se of polypropylene in the dynamically cured polypropylene/epoxy blends was much lower than that of polypropylene, and it decreased as the epoxy resin content increased, until epoxy resin content in the matrix composition was close to 20 wt%. Above this epoxy resin content, se tend to increase slightly. At low epoxy

Chapter 21

Polyolefin/Epoxy Resin Blends

637

Table 21.6 Kinetic Parameters of PP and Dynamically Cured PP/Epoxy Blends. PP/MAH-g-PP/epoxy/EMI-2,4 100/0/0/0

80/10/10/0.4

70/10/20/0.8

60/10/30/1.2

Tc, C 130 132 134 136 130 132 134 136 130 132 134 136 130 132 134 136

K(T)

n

0.065 0.006 0.002 0.0003 8.65 3.04 0.80 0.18 9.46 4.02 0.84 0.21 6.37 2.75 0.64 0.15

2.15 2.20 2.29 2.35 2.45 2.73 3.02 3.20 2.35 2.63 2.93 2.94 2.52 2.90 3.08 3.16

t1/2, min 5.34 7.86 13.80 30.97 0.39 0.52 0.90 1.40 0.33 0.49 0.83 1.33 0.42 0.67 1.07 1.73

[From Reference 53 with permission from Wiley Interscience.]

resin content, the epoxy particles could act as effective nucleating agents for PP. However, at higher epoxy resin contents (>20 wt%) in the blends, the epoxy particles gave rise to restrictions on PP segment mobility.

21.6

MORPHOLOGY

The morphology of the epoxy/polyolefin blend changed with respect to the modifier used. The morphology of PP/epoxy blends was different from that of functionalized PE/epoxy blends. The scanning electron micrographs of epoxy resin cured with HHPA and epoxy/10 phr AC 5120 blend cured with MNA at 115 C is shown in Fig. 21.9. The distinct fracture lines characteristic of brittle resins was observed on the fracture surface of control (59). In the blend, the bulk of the sample exhibited features similar to those observed for the control and no dispersed particles were found in the bulk of the sample as the polyolefin phase migrated to the top of the sample. The fracture surface of (20 and 30 phr) AC 5120-RVP blends was different from that of unmodified AC 5120 blends with epoxy resin. Although some of the polyethylene migrated to the surface, dispersed polyethylene was seen on the fracture surfaces (Fig. 21.10). The fracture surface of the dispersed particles was different from that of the epoxy phase. The particles fractured in a ductile manner and the adhesion between the dispersed phase and matrix was good. Thus, the dispersed phase could slow down the propagation of the crack by the ductile tearing of the particles. The thickness of

638

Polyolefin Blends

Figure 21.9 SEM micrograph for samples: (a) control (100 Epikote 828/80 HHPA/1 BDMA) and (b) 100 Epikote 828/64 MNA/10 AC 5120/1 BDMA both cured at 115 C. (From Reference 59 with permission from Wiley Interscience.)

the phase-separated layer on the top surface became smaller with the decrease in the content of modified AC 5120. When the functionalized polyethylene was modified with p-t-butyl phenol-glycidyl ether (RVP) to a high degree of conversion (i.e., reacted in the presence of triphenylphosphine), the spherical particles precipitated in a more uniform way and displayed a reduced tendency of migration to the surface.

Figure 21.10 SEM micrograph of (a) 100 Epikote 828/70 HHPA/30 AC 5120-RVP/2 BDMA cured at 115 C, (b) 100 Epikote 828/70 HHPA/30 AC 5120-RVP/3 BDMA cured at 60 C, (c) 100 Epikote 828/70 HHPA/20 AC 5120-RVP/1 BDMA cured at 115 C, and (d) 100 Epikote 828/70 HHPA/20 AC 5120RVP/5 BDMA cured at 60 C. (From Reference 59 with permission from Wiley Interscience.)

Chapter 21

Polyolefin/Epoxy Resin Blends

639

Figure 21.11 SEM micrograph of (a) 100/Epikote 828/78 HHPA/5 AC 5120-RVP/TPP/1 BDMA and (b) 100/Epikote 828/70 HHPA/20 AC 5120-RVP/TPP/1 BDMA, both cured at 115 C (micrographs (a) and (b) have been taken with the same magnification). (From Reference 59 with permission from Wiley Interscience.)

When cured at higher temperatures, independent of their composition, these particles always appeared at the front of brittle fracture planes, suggesting that they acted as crack stoppers (Fig. 21.11a and b). When the resin was cured at lower temperature, higher amount of modified AC 5120 in the resin was required to arrest the propagation of the fracture through each particle (Fig. 21.12a–d). For

Figure 21.12 SEM micrograph for samples (a) and (b) 100 Epikote 828/78 HHPA/5 AC 5120-RVP/ TPP/1 BDMA and (c) and (d) 100 Epikote 828/70 HHPA/20 AC 5120-RVP/TPP/1 BDMA all cured at 60 C. (From Reference 59 with permission from Wiley Interscience.)

640

Polyolefin Blends

the samples cured at 60 C, the particles appeared to fracture in a more ductile way (Fig. 21.12b and d) than those present in systems cured at 115 C (Fig. 21.11a and b), which were more shiny and regular that is, at 60 C, the system being below the melting point of the polyethylene phase retained much of the characteristics of the two individual phases. The particles that precipitated from the high temperature curing schedule, on the contrary, contain epoxy resin and hardener and were likely to be cross-linked. The apparent volume fraction of the particles (VP) calculated from the particle dimension assuming that they have broken exactly in the middle and the theoretical volume fraction (VM) assuming quantitative separation of PE from the liquid resin without occluding any amounts of resin or hardener for selected systems is given in Table 21.7. The volume fraction of the precipitated particles calculated from the scanning electron micrographs was much lower than those calculated theoretically from the total content of the polyethylene in each sample. Although the calculation of theoretical volume fraction values was subjected to some errors, resulting from the assumption that the density of the modifier was the same as that of the modifier, the values showed that only a small portion of PE was precipitated as particles. The most remarkable revelation from these observations was the substantial reduction in the volume fraction of the precipitated phase with increase in the curing temperature. Furthermore, the curing temperature had an appreciable effect on the nature of the precipitated particles. The systems cured at temperatures below the melting point of polyethylene (cured at 60 C) contain particles separated through chemical reactions in addition to those formed through crystallization. When the curing temperature was higher than the melting point of the polyolefin phase (cured at 115 C), the only mechanism for the formation of particles was the spontaneous phase separation determined by solubility considerations. In the first case, the curing process produced particles that are very rich in the crystallizable component, explaining the very ductile surface of the particles. In the second case, the precipitation involved a mixture of modified polyethylene with epoxy resin giving rise to more amorphous particles not too different in composition from the surrounding brittle matrix, but still able to act as crack stoppers. Both systems, however, showed a very good adhesion between the matrix and the particles. The dimensions and the number of the precipitated particles were influenced by both the curing temperature and the amount of modified polyethylene. The dimensions of the particles increased with the Table 21.7

Weight Fractions and Volume Fraction of PE in the Blends.

Epikote 828/HHPA/AC 5120RVP/TPP/BDMA 100/78/5/1 cured at 60 C 100/70/20/1 cured at 90 C 100/70/20/1 cured at 115 C

Wt. fraction, %

VM, %

VP, %

5 17 17

6.6 21.7 21.7

3.7 8.3 4.6

[From Reference 59 with permission from Wiley Interscience.]

Chapter 21

Polyolefin/Epoxy Resin Blends

641

Figure 21.13 SEM micrographs of PP/epoxy blends: (a) 70/30 PP/epoxy, (b) 60/10/30 PP/MAH-gPP/epoxy, (c) 60/10/30/1.2 PP/MAH-g-PP/epoxy/EMI-2,4 (dynamically cured, without etching), and (d) 60/10/30/1.2 PP/MAH-g-PP/epoxy/EMI-2,4 (dynamically cured, with etching). (From Reference 48 with permission from Wiley Interscience.)

content of AC 5120-graft-RVP/TPP and with the use of a lower curing temperature (60 C). Not only the dimensions but also the number of particles decreased when the curing temperature is increased. The micrographs of polypropylene/epoxy blends shown in Fig. 21.13 contain epoxy particles of diameter 3–4 mm dispersed in polypropylene matrix (70/30). When the blend was compatibilized with MAH-g-PP, fine epoxy particles with an average diameter of 1 mm were dispersed in the PP matrix, that is, the maleic anhydride group of the MAH-g-PP reacted with epoxy group to form a copolymer and acted as a compatibilizer (48). The reaction between the two groups was evident from torque measurements during mixing. Figure 21.14 shows the relationship between the torque and time for the PP/ epoxy (70/30) and PP/MAH-g-PP/epoxy (60/10/30) blends at 190 C. The torque at equilibrium of the PP/MAH-g-PP/epoxy (60/10/30) blend was obviously higher than that of the PP/epoxy (70/30) blend due to the reaction between maleic anhydride (MAH) groups of MAH-g-PP and the hydroxyl or epoxy groups of the epoxy resin. Dynamic curing prevented the aggregation of epoxy particles resulting in a finer distribution of domains compared to PP/MAH-g-PP/epoxy blends (Fig. 21.13). The addition of carbon black (CB) to polypropylene/epoxy (70/30) blends significantly changed the usual droplet matrix morphology in the blend (60). The

642

Polyolefin Blends

Figure 21.14 Torque versus time for 70/30 PP/epoxy and 60/10/30 PP/MAH-g-PP/epoxy blends at 190 C. (From Reference 48 with permission from Wiley Interscience.)

morphological change was obvious from the scanning electron micrographs shown in Fig. 21.15. In polypropylene/epoxy blends, polypropylene formed the matrix and epoxy formed spherical particles of 3–20 mm diameter. Less spherical particles and more elongated dispersed phase was seen in the micrographs of carbon black filled polypropylene/epoxy blends (70/30/6). The dichloromethane extracted fracture surfaces are shown in Fig. 21.16. Dichloromethane removed the epoxy phase in PP/epoxy blends but the extraction was not complete in carbon black filled blends due to the strong interaction between epoxy resin and carbon black. Carbon black was preferentially located in the epoxy matrix due to high polarity and lower melt viscosity of epoxy resin at the processing temperature (190 C) (61,62).

Figure 21.15 SEM micrographs of fractured surface of (a) PP/epoxy (70/30) and (b) PP/epoxy/CB (70/30/6). (From Reference 60 with permission from Wiley Interscience.)

Chapter 21

Polyolefin/Epoxy Resin Blends

643

Figure 21.16 SEM micrographs of extracted fracture surface of (a) PP/epoxy (70/30) and (b) PP/ epoxy/CB (70/30/6) 50. (From Reference 60 with permission from Wiley Interscience.)

The formation of elongated dispersed phase was clearer in the optical micrograph shown in Fig. 21.17. The polypropylene continuous phase was transparent and epoxy phase appeared dark in the optical micrograph. However, the cause of elongated structure was not fully understood. The formation of elongated structure could be due to the friction between carbon black coated dispersed particles and the matrix (61), which decreased the interfacial tension between the immiscible polymers (63) or both (64). Based on extraction studies and thermogravimetric analysis, it was concluded that nearly all the carbon black particles were located in the epoxy matrix.

Figure 21.17 Optical micrographs of the PP/epoxy/CB (70/30/6) blend. (From Reference 60 with permission from Wiley Interscience.)

644

Polyolefin Blends

21.7 DYNAMIC MECHANICAL PROPERTIES The dynamic mechanical spectrum of DGEBA epoxy resin with 5 and 20 parts AC 5120-RVP cured at 60 C is shown in Fig. 21.18. A small peak at 90 C in G0 curve was associated with the melting of polyolefin phase. The viscoelastic properties of the blends were dependent on the cure temperature (59). The storage and loss modulus values of the blends cured at 60 C was

Figure 21.18 Dynamic mechanical spectrum of (a) 100 parts of Epikote 828, 78 parts of HHPA, 5 parts of AC 5120-RVP/TPP, 1 part of BDMA and (b) 100 parts of Epikote 828, 80 parts of HHPA, 20 parts of AC 5120-RVP/TPP 1 part of BDMA, all cured at 60 C. (From Reference 59 with permission from Wiley Interscience.)

Chapter 21

Polyolefin/Epoxy Resin Blends

645

Table 21.8 Viscoelastic Properties of Epoxy Resin Modified with Functionalized Polyethylene. G0 , GPa 100 Epoxy/80 HHPA/1 BDMAa 100 Epoxy/78 HHPA/5 AC 5120-RVP/TPP/1 BDMAa 100 Epoxy/80 HHPA/20 AC 5120-RVP/TPP/1 BDMAa CR Epoxy/80 HHPA/1 BDMAb 100 Epoxy/78 HHPA/5 AC 5120-RVP/TPP/1 BDMAb 100 Epoxy/80 HHPA/20 AC 5120-RVP/TPP/1 BDMAb

2.24 2.10 1.68 1.95 1.71 1.86

G00 , MPa

Tg, C

27.0 34.5 27.0 20.0 25.0 41.0

132 129 124 134 129 126

a

Cure temperature ¼ 115 C (postcured). Cure temperature ¼ 60 C (postcured). [From Reference 59 with permission from Wiley Interscience.] b

15–20% lower than that of the blends cured at 115 C (Table 21.8). Overall AC 5120 did not influence the dynamic mechanical properties. The dynamic mechanical spectra of PPC (MW 4900) modified epoxy resin in Fig. 21.19 showed typical features of a phase-separated blend system (47). The relaxations corresponding to the glass transition of polypropylene carbonate and epoxy resin were observed in the spectrum (Table 21.9). The glass-transition peak of epoxy broadened as the polypropylene carbonate content increased. An interesting feature of this system was the decrease in the glass transition of epoxy without any change in the polypropylene carbonate peaks. The shift in the epoxy peak was due to the reaction between polypropylene carbonate and epoxy resin.

Figure 21.19 tan d versus temperature for cured resin with various PPC contents (molecular weight of PPC: 4900). (From Reference 47 with permission from Wiley Interscience.)

646

Polyolefin Blends

Table 21.9 PPC/EP 0/100 20/100 30/100 60/100 100/100

Effect of PPC on the Dynamic Mechanical Properties. Half width, C 23 43 45 58 63

Tg (EP), C 136.9 110.1 92.2 86.3 68.1

Tg (PPC), C — 24.8 22.0 18.9 18.8

[From Reference 47 with permission from Wiley Interscience.]

The storage modulus for the cured resins with different polypropylene carbonate contents (Fig. 21.20) did not change over the temperature range lower than their a-relaxation, compared with the parent epoxy resin. This implied that the epoxy matrix was toughened by the addition of PPC at no expense of its modulus but with a sacrifice to its thermal properties. The molecular weight of polypropylene carbonate

Figure 21.20 Storage modulus versus temperature for cured resin with various PPC contents (molecular weight of PPC: 4900). (From Reference 47 with permission from Wiley Interscience.)

Chapter 21

Polyolefin/Epoxy Resin Blends

647

hardly changed the tan d peak but polypropylene carbonate with molecular weight 2100 showed only a single peak and was transparent indicating a miscible system.

21.8

MECHANICAL PROPERTIES

The mechanical properties of polyolefin modified epoxy resin changed with respect to the type of modifier used. The properties were also dependent on the blend composition and interaction between the components. The mechanical properties of RVP grafted AC 5120 was reported with respect to the change in composition and curing temperature (Table 21.10). The properties of epoxy resin was not seriously affected by the addition of 5% RVP grafted AC 5120. However, the properties decreased as the concentration of modifier increased. But the decrease was due to the lower cross-link density of the cured blend since lesser amount of hardener HHPA (70 parts) was used in the blends instead of 80 parts in the neat resin. The fracture toughness of the blends was influenced by the cure temperature employed for curing the blends. The blends cured at 115 C have 10% more fracture toughness than those cured at 60 C, but the fracture energy remained unaffected by the curing conditions. Fracture toughness increased up to 20% polyolefin content and leveled off above this composition. Fracture energy increased by 20% at low polyolefin concentrations. The charpy impact strength increased by 10–15% by the inclusion of 5–10% modified AC 5120. The lower-than-expected improvement in fracture toughness compared to other thermoplastic toughened epoxy resin was due to the effect of curing agent. The lower cure temperature and anhydride-curing agent resulted in less cross-linked systems compared to amine cured blends. Another factor contributing to the lower fracture toughness was the lower volume fraction of precipitated toughening phase due to the chemical reaction between modified AC Table 21.10 Flexural Properties of Functionalized PE Modified Epoxy Resin. Flexural Stress at Strain at modulus, break, break, GPa MPa % 100 100 100 100 100 100 100 100 a

Epoxy/80 Epoxy/70 Epoxy/78 Epoxy/70 Epoxy/80 Epoxy/70 Epoxy/78 Epoxy/70

HHPA/1 BDMAa HHPA/20 AC 5120-RVP/1 BDMAa HHPA/5 AC 5120-RVP/TPP/1 BDMAa HHPA/20 AC 5120-RVP/TPP/1 BDMAa HHPA/1 BDMAb HHPA/20 AC 5120-RVP/5 BDMAb HHPA/5 AC 5120-RVP/TPP/1 BDMAb HHPA/20 AC 5120-RVP/TPP/1 BDMAb

Cure temperature ¼ 115 C (post cured). Cure temperature ¼ 60 C (post cured). [From Reference 59 with permission from Wiley Interscience.] b

2.37 1.81 2.36 2.17 2.28 1.90 2.21 2.13

99.1 90.1 111.4 86.2 97.0 75.5 96.3 86.6

8.4 6.6 7.7 5.4 7.0 5.4 5.7 5.8

648

Polyolefin Blends

Table 21.11 Mechanical Properties of PPC Modified Epoxy Resin. PPC (phr) Property Tensile strength, MPa Tensile modulus, MPa Elongation, % Fracture toughness, J mm2

0

20

30

60

80

100

51.6 1291 4.3 0.024

67.7 1291 6.7 0.055

62.0 1316 5.8 0.041

48.1 1000 5.6 0.036

8.6 228 15.9 0.022

23.4 558 10.0 0.030

[From Reference 47 with permission from Wiley Interscience.]

5120 and epoxy resin. A third contributing factor was the plasticization effect of solubilized polyolefin and unreacted RVP in the epoxy matrix. The mechanical properties of polypropylene carbonate modified epoxy resin changed with the composition as shown in Table 21.11. 20 phr polypropylene carbonate (MW 4900) was good enough to produce 45% increase in fracture toughness and 33% increase in tensile strength. The same blend showed higher impact strength and shear strength than the neat resin (Fig. 21.21). The dynamically cured PP/MAH-g-PP/epoxy blends have better mechanical properties than PP/epoxy and PP/MAH-g-PP/epoxy blends as shown in Table 21.12. The decrease in impact strength, tensile strength, and elongation at break was due to the poor compatibility between epoxy resin and PP. The addition of MAH-g-PP (10%) as compatibilizer increased the flexural modulus and tensile strength. Dynamic curing of the blend further increased the flexural modulus and impact strength. The mechanical properties of the dynamically cured blends changed further with the variation in epoxy content as shown in the figures (Fig. 21.22a–b).

Figure 21.21 Effect of PPC content on impact and shear strength. (From Reference 47 with permission from Wiley Interscience.)

Chapter 21

Polyolefin/Epoxy Resin Blends

649

Table 21.12 Mechanical Properties of PP and PP/Epoxy Blends.

Composition PP PP/epoxy (70/30) PP/MAH-g-PP/epoxy (60/10/30) PP/MAH-g-PP/epoxy/EMI-2,4 (60/10/30/1.2)

Impact strength, J m1

Tensile strength, MPa

Elongation at break, %

Flexural modulus, MPa

30 18

30.8 25.6

300 12

1250 1577

25

35.2

21

1659

35

35.0

24

2003

[From Reference 48 with permission from Wiley Interscience.]

The impact strength, tensile strength, and flexural modulus increased while elongation at break decreased on increasing the epoxy content in the blends. The impact and tensile strengths leveled off above the 10 phr addition of epoxy resin.

21.9

CONDUCTIVITY STUDIES

Carbon black is the most widely used conducting filler in composite industry. Carbon black filled immiscible blends based on polar/polar (65), polar/nonpolar (63,66), nonpolar/nonpolar thermoplastics (67,68), plastic/rubber and rubber/rubber blends (69,70) have already been reported in the literature. The properties of carbon black filled immiscible PP/epoxy were reported recently by Li et al. (60). The blend system was interesting because one of the components is semicrystalline and the other is an amorphous polar material with different percolation thresholds. The volume resistivity of carbon black filled individual polymers is shown in Fig. 21.23. The volume resistivity decreased with the increase in carbon black content and a marked decrease was observed near the percolation threshold where the transition from insulating to conductive material occurred. The carbon black particles came into contact with each other or close enough to allow electron hop by tunneling, and once percolation threshold was reached additional carbon black loading did not change the volume resistivity much. The percolation threshold for polypropylene and epoxy resin was 8 and 17 phr, respectively. This is due to the difference in the inherent properties of the material. Polypropylene is a highly crystalline polymer in which carbon black can only distribute in the amorphous phase. This selective distribution of carbon black in the amorphous phase was equivalent to increasing the effective concentration of carbon black in the polypropylene matrix. The morphology of PP/CB (100/11) composites where carbon black has reached the percolation threshold is shown in Fig. 21.24a. Carbon black formed chain–like structure within the nonpolar polypropylene matrix in the form of agglomerates rather than as individual particles because of the poor wetting of carbon black. The carbon black distribution in PP was relatively

650

Polyolefin Blends

Figure 21.22 Effect of the epoxy resin content on the mechanical properties of dynamically cured PP/epoxy blends. (From Reference 48 with permission from Wiley Interscience.)

heterogeneous as a result of the selective localization in the amorphous phase. Hence the percolation threshold was low. In contrast to polypropylene, epoxy resin is an amorphous and polar resin containing epoxy, hydroxyl, and ether groups. The high polarity of epoxy resin

Chapter 21

Polyolefin/Epoxy Resin Blends

651

Figure 21.23 Volume resistivity versus CB content of CB filled individual PP and epoxy matrix. (From Reference 60 with permission from Wiley Interscience.)

makes it easily wet the carbon black surface and carbon black particles disperse more uniformly in epoxy resin. As a result, the percolation threshold of epoxy/CB composite was much higher than that of PP/CB composite. The micrographs in Fig. 20.24b (epoxy/CB, 100/20 composites) showed good wetting and uniform dispersion of carbon black in the epoxy matrix. This comparatively uniform carbon black distribution resulted in the high percolation threshold of epoxy/CB composite. The volume resistivity of PP/epoxy blends as a function of carbon black content is shown in Fig. 21.25. The percolation thresholds of carbon black filled 50/50 and 40/60 PP/epoxy blends were found to be about 7.5 and 8 phr, respectively, which were close to the percolation threshold of PP/CB composite. The electrical resistivity and the percolation threshold of carbon black filled immiscible polymer blends depend greatly on the morphology (67,68). The difference in the electrical behavior of various PP/epoxy/CB blends was explained based on the morphological difference. The scanning electron micrographs in Figs. 21.15 and 21.16 showed that carbon black was preferentially localized in the dispersed epoxy phase in the PP/epoxy (70/30) blend. Although the epoxy phase was not continuous, there were narrow gaps between the adjacent carbon black rich epoxy domains. At low carbon black content, carbon black did not percolate in epoxy resulting in very high volume resistivity. When carbon black content was in the vicinity of 6 phr, carbon black reached the percolation level in polypropylene (the percolation threshold of epoxy/CB was about 17 phr). However, conducting paths could be formed via double percolation because of the discontinuity of the epoxy phase, leading to the relatively high volume resistivity when tested with 10 V. In carbon black filled conducting polymers and polymer blends, a higher applied electric field could excite electrons and maintain a higher kinetic and potential energy level, which allowed electrons to proceed in the ‘‘tunneling process’’ or ‘‘jump’’ through the gap (71). This

652

Polyolefin Blends

Figure 21.24 SEM micrographs of freeze-fractured surface of (a) PP/CB (100/11) and (b) epoxy/CB (100/20) composites. (From Reference 60 with permission from Wiley Interscience.)

well explained the significant decrease in volume resistivity when the voltage of 100 V was applied for PP/epoxy (70/30) blends with 6 phr or higher carbon black content. The tunneling effect between the carbon black clusters in adjacent epoxy domains was the dominant conduction mechanism at this high voltage, reducing the volume resistivity of the blends remarkably. The similar electrical behavior of carbon black filled PP/epoxy (60/40) blend was explained using the morphology of the blends. The scanning electron micrographs showed that epoxy phase was dispersed in polypropylene matrix and the addition of carbon black resulted in the preferential localization of carbon black in epoxy phase. As shown in Fig. 21.26 the dispersed epoxy phase was elongated to a great extent. Although the epoxy phase was incompletely continuous, the gaps between some adjacent carbon black rich epoxy

Chapter 21

Polyolefin/Epoxy Resin Blends

653

Figure 21.25 Volume resistivity of PP/epoxy/CB blends at various PP/epoxy ratios. (From Reference 60 with permission from Wiley Interscience.)

domains are very small. In certain regions, the morphology seems to approach the onset of a cocontinuous structure. The volume resistivity of the blends with low carbon black loading remained high because carbon black could not reach the percolation threshold in epoxy resin. Increasing the carbon black content resulted in the gradually slow decrease of volume resistivity due to the increased carbon black concentration in the partially continuous epoxy phase. When the carbon black content was close to or higher than the percolation level of 8 phr in epoxy, the incomplete continuity of epoxy phase made the blends impossible to achieve electrical conduction through double percolation. Thus the volume resistivity tested with 10 V was relatively high. Similarly, when higher voltage (100 V) was applied, the tunneling effect between the adjacent

Figure 21.26 SEM micrographs of freeze-fractured surface of PP/epoxy/CB (60/40/10, etched). (From Reference 60 with permission from Wiley Interscience.)

654

Polyolefin Blends

carbon black clusters dominated the conduction process and thus the volume resistivity was significantly decreased, compared with the results obtained with 10 V applied voltage. 50/50 and 40/60 PP/epoxy blends exhibited cocontinuous morphology. Increasing the carbon black content decreased the volume resistivity, and the blends became conductive through double percolation. The percolation threshold for these systems was 7.5 and 8 phr, respectively. At a given carbon black content, the volume resistivity of carbon black filled PP/epoxy (40/60) blend was higher than that of the carbon black filled 50/50 blend, as shown in Fig. 21.25. This is because the given amount of carbon black is dispersed in more epoxy phase in the PP/epoxy (40/60) blend compared to 50/50 PP/epoxy blend. The processing sequence was another factor, which influenced the volume resistivity of the blends, because it had profound effect on the carbon black distribution and morphology. In addition to simultaneous melt mixing of the blends, epoxy/CB/PP or PP/CB/epoxy blends were prepared by premixing carbon black with epoxy or polypropylene followed by addition of polypropylene or epoxy. The volume resistivity of the various blends is depicted in Table 21.13. In 70/30 PP/epoxy blends containing 6 parts of carbon black, the volume resistivity of simultaneously melt-mixed sample and epoxy/CB/PP composites were similar at 10 and 100 V. The value at 100 V was considerably lower than that at 10 V. In PP/epoxy/CB 40/60/4 samples, epoxy/CB/PP, and simultaneously melt-mixed samples exhibited similar volume resistivity, whereas the volume resistivity of PP/CB/epoxy system was much low. The variation in the volume resistivity was due to the difference in the carbon black distribution and morphology change. The scanning electron micrographs of epoxy/CB/PP (30/6/70) and PP/CB/epoxy (70/6/30) systems are shown in Fig. 21.27. The epoxy phase is dispersed in polypropylene matrix, the particles are elongated and there was no migration of carbon black to PP due to the good affinity between epoxy and carbon black (Fig. 21.27a). Therefore, the blends exhibited similar volume resistivity as that of the simultaneously melt-mixed blends. The micrographs of PP/CB/epoxy system in Fig. 21.27b revealed partial migration Table 21.13 Volume Resistivity of PP/Epoxy/CB (70/30/6) and PP/Epoxy/CB (40/60/4) Blends with Different Processing Sequences. Sample composition PP/epoxy/CB (70/30/6)

Processing sequence Simultaneously melt-blended Epoxy/CB/PP

PP/epoxy/CB (40/60/4)

PP/CB/epoxy Simultaneously melt-blended Epoxy/CB/PP PP/CB/epoxy

[From Reference 60 with permission from Wiley Interscience.]

Volume resistivity, V cm About 1014 (10 V) <108 (100 V) About 1014 (10 V) <108 (100 V) 4.87 109 1.72 1016 7.75 1016 1.66 104

Chapter 21

Polyolefin/Epoxy Resin Blends

655

Figure 21.27 SEM micrographs of freeze-fractured surface of (a) epoxy/CB/PP (30/6/70) and (b) PP/ CB/epoxy (70/6/30) systems. (From Reference 60 with permission from Wiley Interscience.)

of carbon black from PP phase to epoxy phase due to the interactions between them. Carbon black was mainly accumulated on the surface of the epoxy particles and carbon black was also present in the polypropylene phase due to the low melt viscosity of epoxy at the processing temperature. The percolation threshold was achieved in PP/CB system before mixing with epoxy. Hence, the volume resistivity of this system was slightly greater than that of PP/CB system due to the migration of carbon black to the epoxy phase. In epoxy/CB/PP (60/4/40) systems, the percolation threshold was not achieved resulting in high volume resistivity. In PP/CB/epoxy (40/ 4/60) system, the partial migration of carbon black reduced its concentration in polypropylene phase. Extraction of epoxy phase revealed that the continuous phase was polypropylene. Thus, the carbon black particles remaining in polypropylene

656

Polyolefin Blends

Figure 21.28 Volume resistivity versus CB content of the PP/CB/epoxy blend (PP/epoxy ¼ 40/60). (From Reference 60 with permission from Wiley Interscience.)

matrix could maintain the conducting path but with slightly higher volume resistivity. The volume resistivity of PP/CB/epoxy system as a function of carbon black loading is given in Fig. 21.28 showed that the volume resistivity decreased with carbon black loading, and the percolation threshold was reached at about 4 phr carbon black. The percolation threshold was almost half of that for PP/CB composite. Thus, it is possible to reduce the percolation threshold by changing the processing sequence.

21.10 CONCLUSION Polyolefin modification of epoxy resin received great attention during the past one decade. Attention has been given to improve the miscibility of polyethylene with epoxy resin by modification of the components. The presence of reactive groups in the blend component was the decisive factor in determining the miscibility. The miscibility of polyethylene in epoxy resin increased with the increase in the functionality of the polymer. The modification of epoxy resin reduced the miscibility because of the reduction in functionality. Polyethylene modified epoxy resin were heterogeneous when the curing agent hexahydrophthalic anhydride was present in the blends. Kinetic studies of polypropylene carbonate modified epoxy resin cured with methyltetrahydrophthalic anhydride revealed that the predominant reaction during curing was the reaction between epoxy resin and methyltetrahydrophthalic anhydride and a small amount of reaction is possible between polypropylenecarbonate and methyltetrahydrophthalic anhydride. The rate of crystallization in PP/epoxy and dynamically cured PP/epoxy blends were higher than that of pure PP. The epoxy particles acted as nucleating sites in both cases and the smaller particles in the

Chapter 21

Polyolefin/Epoxy Resin Blends

657

dynamically cured blends further increased the number of nucleating sites and rate of crystallization. Although the rate of crystallization increased, the percentage crystallinity was less than that of pure PP. The analysis of crystallization kinetics using Avrami’s equation revealed that the polypropylene spherulites grow in three-dimensional spherical forms. The size of the polypropylene spherulites decreased with the increase in the amount of epoxy in the blends. The morphology of functionalized polyethylene modified epoxy resin was dependent on the functionality of polyethylene. The oxidized homopolymer with an acid functionality of 2.0 modified with p-t-butyl phenol-glycidyl ether system showed dispersed polyethylene morphology and the precipitated particles were fractured in a ductile manner. The precipitation at curing temperatures (60 C) lower than the melting point of modified polyethylene was mainly due to crystallization and at higher curing temperatures (115 C), above the melting temperature; the precipitation was due to the curing of epoxy resin. The volume fraction of precipitated polyethylene was less than that of the theoretical volume fraction because of the reaction between the blend components. In polypropylene/epoxy blends, the blend composition was so adjusted that epoxy formed the dispersed phase. The size of the epoxy domains increased with the increase in epoxy content in the blends. On compatibilization with maleic anhydride graft polypropylene, the domain size decreased further. The dynamically cured polypropylene/epoxy blends had finer domains than the compatibilized blends. The addition of carbon black to the blends significantly altered the morphology of the blends. The fracture surfaces contained more elongated epoxy particles than the spherical ones. The carbon black migrated to epoxy phase due to the high polarity and low melt viscosity of epoxy resin. The dynamic mechanical properties of epoxy resin were not much affected by the addition of modified polyethylene to epoxy resin. The polypropylenecarbonate modified epoxy resin showed typical characteristics of a phase-separated blend. The glass-transition temperature of epoxy phase decreased with the addition of more and more polypropylenecarbonate. The mechanical properties were dependent on the type of polyolefin used. A higher cure temperature favored more increase in fracture toughness in modified polyethylene toughened epoxy resin. A 20 phr addition of polypropylenecarbonate to epoxy resin resulted in 45% increase in fracture toughness. The compatibilized polypropylene/epoxy blends exhibited higher tensile, flexural, and impact strengths. Conductivity studies of carbon black filled polypropylene/epoxy blends are very new. The volume resistivity of the blends was dependent on morphology and processing sequence. It was shown that by judicious selection of mixing sequence the percolation threshold can be manipulated as per the requirements. From the above discussion, it is clear that the investigation on polyolefin epoxy blends are mainly focused on the miscibility and crystallization studies. Polyolefin was solubilized in epoxy by functionalizing the polyolefin. However, attempts are yet to be made to have the complete miscibility of polypropylene in the epoxy matrix. An understanding of the morphology development during curing is essential for explaining the phase-separation mechanism and ultimate morphology of the blends. The improvement in the mechanical properties of polyolefin/epoxy blends is less than

658

Polyolefin Blends

that of other thermoplastic modified epoxy resins. A detailed and systematic study for improving the mechanical properties of the blends is required for using these blend systems for various applications.

NOMENCLATURE s se kB DW AC 1450 AC 5120 AC 540 AC 6702 AC 80 ATBN b0 BDMA CA CA CB CTBN CTPB DGEBA DMP-302 DSC EHTPB EMI-242 FTIR HHPA HTBN HTPB MAH MNA MTHPA MTHPA PCL-TESi PE PEEK PP

Free energy per unit area of the surfaces of the lamellae parallel to the chain direction Free energy per unit area of the surfaces of the lamellae perpendicular to the chain direction Boltzmann constant Crystallite size distribution Ethylene–acrylic acid–vinyl acetate terpolymer Ethylene–acrylic acid copolymers with functionality approximately 2.0 Ethylene–acrylic acid copolymers with functionality approximately 1.2 Oxidized homopolymer with an acid functionality of 0.26 Ethylene–acrylic acid–vinyl acetate–vinyl alcohol terpolymer Amine terminated butadiene acrylonitrile rubber Distance between two adjacent fold planes N,N-Dimethylbezylamine 1,4,5,6,7,7-Hexachloro-5-norborene-2,3-dicarboxylic anhydride Chlorendic anhydride Carbon black Carboxyl terminated butadiene acrylonitrile rubber Carboxyl terminated polybutadiene Diglycidyl ether of bisphenol-A 2,4,6-Tris(dimethylaminomethyl)phenol Differential scanning calorimetry Epoxidized hydroxyl terminated polybutadiene Ethylene-4-methane imidazole Fourier transform infrared spectroscopy Hexahydrophthalic anhydride Hydroxyl terminated butadiene acrylonitrile random copolymer Hydroxyl terminated polybutadiene Maleic anhydride Methyl nadic anhydride Methyltetrahydrophthalic anhydride Methyltetrahydrophthalic anhydride Triethoxysilyl terminated polycaprolactone elastomer Polyethylene Polyether–ether ketone Polypropylene

Chapter 21

PPC RIPS RVP Si t1/2 Tc Tm Tp UCST VM VP X(t)

Polyolefin/Epoxy Resin Blends

659

Polypropylene carbonate Reaction-induced phase separation p-t-Butyl phenol-glycidyl ether Rate of nucleation Halftime for polypropylene crystallization Crystallization temperature Melting temperature Crystallization peak temperature Upper critical solution temperature Theoretical volume fraction Volume fraction of the particles Relative crystallinity at different crystallization times

REFERENCES 1. B. Ellis, Chemistry and Technology of Epoxy Resins, Blackie Academic and Professional, London, 1993. 2. L. T. Manzione, J. K. Gilham, and C. A. McPherson, J. Appl. Polym. Sci. 26, 889 (1981). 3. V. Nigam, D. K. Setua, and G. N. Mathur, Rub. Chem. Technol. 73, 830 (2001). 4. M. L. Arias, P. M. Frontini, and R. J. J. Williams, Polymer 44, 1537 (2003). 5. P. S. Achary, P. B. Latha, and R. Ramaswamy, J. Appl. Polym. Sci. 41, 151 (1990). 6. B. U. Kang, J. Y. Jho, J. Kim, S. S. Lee, M. Park, S. Lim, and C. R. Choe, J. Appl. Polym. Sci. 79, 38 (2001). 7. A. Gu and G. Liang, J. Appl. Polym. Sci. 89, 3594 (2003). 8. V. D. Ramos, H. M. da Costa, V. L. P. Soares, and R. S. V. Nascimento, Polym. Test. 24, 387 (2005). 9. C. W. Wise, W. D. Cook, and A. A. Goodwin, Polymer 41, 4625 (2000). 10. J. Zhang, H. Zhang, and Y. Yang, J. Appl. Polym. Sci. 72, 59 (1999). 11. P. B. Latha, K. Adhinarayanan, and R. Ramaswamy, Int. J. Adhes. Adhes. 14, 57 (1994). 12. V. Nigam, D. K. Setua, and G. N. Mathur, J. Appl. Polym. Sci. 87, 861 (2003). 13. S. T. Lin, and S. K. Huang, J. Polym. Sci. Part A: Polym. Chem. 34, 1907 (1996). 14. S. T. Lin and S. K. Huang, Eur. Polym. J. 33, 365 (1997). 15. P. Liu, L. He, H. Ding, J. Liu, and X. Yi, J. Appl. Polym. Sci. 97, 611 (2005). 16. I. Blanco, G. Cicala, C. L. Faro, and A. Recca, J. Appl. Polym. Sci. 89, 268 (2003). 17. R. J. Varley, J. H. Hodgkin, and G. P. Simon, Polymer 42, 3847 (2001). 18. K. Mimura, H. Ito, and H. Fujioka, Polymer 41, 4451 (2000). 19. I. Martinez, M. D. Martin, A. Eceiza, P. Oyanguren, and I. Mondragon, Polymer 41, 1027 (2000). 20. I. Blanco, G. Cicala, O. Motta, and A. Recca, J. Appl. Polym. Sci. 94, 361 (2004). 21. B. Francis, V. L. Rao, S. Jose, K. V. S. N. Raju, R. Ramaswamy, and S. Thomas, Polym. Eng. Sci. 45, 1645 (2005). 22. B. Francis, S. Thomas, J. Jose, R. Ramaswamy, and V. L. Rao, Polymer 46, 12372 (2005). 23. M. Naffakh, M. Dumon, J. Dupuy, and J. F. Gerard, J. Appl. Polym. Sci. 96, 660 (2005). 24. M. Wang. Y. Yu, X. Wu, and S. Li, Polymer 45, 1253 (2004). 25. S. Shin and J. Jang, J. Appl. Polym. Sci. 65, 2237 (1997).

660

Polyolefin Blends

26. A. Bonnet, B. Lesteriez, J. P. Pascault, and H. Sautereau, J. Polym. Sci. Part B: Polym. Phys. 39, 363 (2001). 27. S. M. Shin, D. K. Shin, and D. C. Lee, J. Appl. Polym. Sci. 78, 2464 (2000). 28. T. Iijima, S. Mimura, M. Fujimaki, T. Taguchi, W. Fukuda, and M. Tomoi, J. Appl. Polym. Sci. 61, 163 (1996). 29. T. Iijima, K. Fujimoto, and M. Tomoi, J. Appl. Polym. Sci. 84, 388 (2002). 30. Y. Cao, Y. Shao, J. Sun, and S. J. Lin, J. Appl. Polym. Sci. 90, 3384 (2003). 31. H. S. Kim and P. Ma, J. Appl. Polym. Sci. 69, 405 (1998). 32. F. Fenouillot, C. Hedreul, J. Forsythe, and J. P. Pascault, J. Appl. Polym. Sci. 87, 1995 (2003). 33. L. Hu, H. Lu, and S. Zheng, J. Polym. Sci. Part B: Polym. Phys. 42, 2567 (2004). 34. H. Lu and S. Zheng, J. Polym. Sci. Part B: Polym. Phys. 43, 359 (2005). 35. Q. Guo, C, Harrats, G. Groeninckx, and M. H. J. Koch, Polymer 42, 4127 (2001). 36. G. Kortaberria, P. Arruti, N. Gabilondo, and I. Mondragon, Eur. Polym. J. 40, 129 (2004). 37. J. C. Cabanelas, B. Serrano, and J. Baselga, Macromolecules 38, 961 (2005). 38. M. S. Li, C. C. M. Ma, M. L. Lin, M. S. Lu, J. L. Chen, and F. C. Chang, Polymer 38, 845 (1997). 39. C. C. Su, E. M. Woo, C. Y. Chen, and R. R. Ru, Polymer 38, 2047 (1997). 40. C. C. Su and E. M. Woo, Macromolecules 28, 6779 (1995). 41. S. Zheng, H. Zheng, and Q. Guo, J. Polym. Sci. Part B: Polym. Phys. 41, 1085 (2003). 42. S. Zheng, Q. Guo, and C. M. Chan, J. Polym. Sci. Part B: Polym. Phys. 41, 1099 (2003). 43. G. V. Poel, S. Goossens, B. Goderis, and G. Groeninckx, Polymer 46, 10758 (2005). 44. Y. Ni, and S. Zheng, Polymer 46, 5828 (2005). 45. A. Saxena, B. Francis, V. L. Rao, and K. N. Ninan, J. Appl. Polym. Sci. 100, 3536 (2006). 46. M. Frigione, D. Acierno, and L. Mascia, J. Appl. Polym. Sci. 73, 1457 (1999). 47. Y. Huang, J. Wang, B. Liao, M. Chen, and G. Cong, J. Appl. Polym. Sci. 64, 2457 (1997). 48. X. Jiang, H. Huang, Y. Zhang, and Y. Zhang, J. Appl. Polym. Sci. 92, 1437 (2004). 49. G. S. Bennet, R. J. Farris, and S. A. Thompson, Polymer 32, 1633 (1991). 50. B. Francis, S. Thomas, G. V. Asari, R. Ramaswamy, S. Jose, and V. L. Rao, J. Polym. Sci. B: Polym. Phys. 44, 541 (2006). 51. B. Francis, V. L. Rao, S. Jose, B. K. Catherine, R. Ramaswamy, and S. Thomas, J. Mater. Sci. 41, 5467 (2006). 52. Z. Zhong, S. Zheng, J. Huang, X. Cheng, Q. Guo, and J. Wei, Polymer 39, 1075 (1998). 53. X. Jiang, Y. Zhang, and Y. Zhang, J. Polym. Sci. B: Polym. Phys. 42, 1181 (2004). 54. M. Avrami, J. Chem. Phys. 7, 1103 (1939). 55. M. Avrami, J. Chem. Phys. 9, 177 (1941). 56. L. Mandelkern, Crystallization of Polymers McGraw-Hill, New York, 1964. 57. J. D. Hoffman, Soc. Plast. Eng. Trans. 4, 315 (1964). 58. J. D. Hoffman and R. L. Miller, Polymer 38, 3151 (1997). 59. M. Frigione, D. Acierno, and L. Mascia, Adv. Polym. Technol. 18, 237 (1999). 60. Y. Li, S. Wang, Y. Zhang, and Y. Zhang, J. Appl. Polym. Sci. 99, 461 (2006). 61. O. Breuer, R. Tchoudakov, M. Narkis, and A. Siegmann, J. Appl. Polym. Sci. 64, 1097 (1997). 62. O. Breuer, R. Tchoudakov, M. Narkis, and A. Siegmann, J. Appl. Polym. Sci. 73, 1655 (1999). 63. K. Cheah, M. Forsyth, and G. P. Simon, J. Polym. Sci. B: Polym. Phys. 38, 3106 (2000). 64. E. Segal, R. Tchoudakov, M. Narkis, and A. Siegmann, J. Polym. Sci. B: Polym. Phys. 41, 1428 (2003). 65. K. Cheah, M. Forsyth, and G. P. Simon, Synth. Met. 102, 1232 (1999). 66. R. Tchoudakov, O. Breuer, and M. Narkis, Polym. Eng. Sci. 36, 1336 (1996).

Chapter 21

Polyolefin/Epoxy Resin Blends

661

67. F. Gubbels, R. Jerome, Ph. Teyssie, E. Vanlathem, R. Deltour, A. Calderone, V. Parente, and J. L. Bredas, Macromolecules 27, 1972 (1994). 68. M. Sumita, K. Sakata, S. Asai, K. Miyasaka, and H. Nakagawa, Polym. Bull. 25, 265 (1991). 69. B. G. Soares, F. Gubbels, R. Jerome, Ph. Teyssie, E. Vanlathem, and R. Deltour, Polym. Bull. 35, 223 (1995). 70. N. C. Das, T. K. Chaki, D. Khastgir, and A. Chakraborty, Adv. Polym. Technol. 20, 226 (2001). 71. J. C. Huang, Adv. Polym. Technol. 21, 299 (2002).

Index

Adhesion, 200, 201, 222 AFM, 218 Arrhenius equation, 79 Bimodal distributions branch, 63, 64, 71 lamellar thickness, 63, 71 molar mass, 65, 68, 76 Bis-alkylphenoldisulphide (BAPD), 469 Bis(diisopropyl)-thiophosphoryl disulfide (DIPDIS), 445, 446, 469 Blown film extrusion, 65 Branch density, 72 Branch distribution, 62–64, 68, 69, 72 Brittle-ductile transition temperature (TBT), 413 Bromobenzene, 466 Carbon black, 444, 462 Chain slip, 77 Chlorinated penetrant, 466 Chlorinated rubber, 450, 452, 454, 461, 462 Chlorobenzene, 466 Chlorohydrocarbon, 466 Chlorosulfonated polyethylene (SPE), 450, 454, 460 N-(4-Chlorophenyl) acrylamide, 452 N-Chlorothioamides, 445 Cocrystallization, 69 Coefficient of linear thermal expansion (CLTE), 203, 204 Compatibility, 224, 225, 247, 264, 442–443, 449–453, 455, 460, 461, 469 Compatibilization, 509, 519, 523 Compatibilizer, 504, 505, 507–513, 517–519, 522, 523 Compression, 466 Continuous phase, 72, 286–290

COP, 59, 503–505, 507–509, 517–519, 522, 523 Copolymer, 28, 270, 273, 274, 277, 282, 284, 289, 293, 296, 298–300 Covulcanization, 448, 452 Crack, 454, 463, 465, 467 Creep resistance, 68 Cross-linking, 270, 273, 278, 279, 284, 290, 291, 293, 295, 298–301 Cross-linking agent, 421, 431 Crystal, 228, 231, 233, 235, 237, 239, 242, 246, 250, 252, 253, 255, 260–263 Crystallization cocrystallization, 69 kinetics, 511, 515–519, 523 lamellar thickness, 69, 71, 72 rate, 510, 515, 518, 519, 522 slow crystallization, 69 Cyclopolyolefin, 28–29 Damping factor curve, 78 Deformation, 225, 242, 244–246, 263, 264, 293 Differential scanning calorimetry (DSC), 69–71, 73, 74, 201, 413 Diffusivity, 466 Dimethyl itaconate (DMI), 414 2,5-Dimethyl 2,5-bis(tert-butyl peroxy) hexane, 414 Dipole–dipole, 442 Dispersed phase, 455 Dithiocarbamate, 445 DMA, 414, 416 DMI-modified PP, 414 Drawability, 225, 264 Dynamic elastic modulus (E0 ), 456 Dynamic mechanical properties, 456

Polyolefin Blends, Edited by Domasius Nwabunma and Thein Kyu Copyright # 2008 John Wiley & Sons, Inc.

663

664

Index

Dynamic mechanical spectroscopy (DMS), 78, 79 Dynamic mechanical thermal analysis (DMTA), 431, 432 Dynamic radiation curing, 412, 438 Dynamic shear moduli (G0 ), 229, 236 Dynamic vulcanization (DV), 419, 421, 425, 435 Elastic properties, 528, 534, 539, 540, 543, 547, 550 Enthalpy, 442 Entropy, 442 EPDM-g-MAH, 450 Ethylene–1-octene copolymer, 214, 220 Ethylene–a-olefin copolymers (ECP), 203, 204 Ethylene–butene copolymer (EBR), 214, 226, 228, 229, 244, 247 Ethylene–hexene copolymer, 214 Ethylene–hexene-1 copolymer (EHR), 230, 231, 233, 236, 238, 244–247, 250, 251, 253, 256–258, 260, 262, 264 Ethylene–propylene copolymer (EPM), 236–241, 247–250, 264, 412 Ethylene propylene diene monomer (EPDM), 7–11, 15–17, 198, 206, 207–217, 222, 224, 242, 412–438, 441, 469, 530–537, 542, 546–550 Ethylene–propylene rubber (EPR), 5, 6, 8–11, 14–17, 198–200, 202, 203, 224, 225, 234, 242, 247 Ethylene vinyl acetate copolymer (EVA), 8, 11, 15, 17 Extender oil, 207 Fatigue resistance, 420, 442 Finite element modeling (FEM), 208, 210, 430 Flory–Huggins theory, 75 Flow direction (FD), 204 Folded-chain lamellae, 219 Fracture, 415, 419, 420, 424–428, 432 Fringe-micelle, 219 FT-IR spectroscopy, 416 Functionalization, 269, 270, 275, 277, 284, 286, 290, 292, 295, 297, 299, 301, 530, 531, 534, 536, 538, 539, 543, 547, 551

Gel permeation chromatography (GPC), 201 Gibb’s free energy, 442 Glass transition temperature (Tg), 229, 231, 234, 236, 239, 241–245, 249–251, 253, 254, 265, 432, 456, 457, 460 Grafting, 270–301 Grafting efficiency, 272–274, 277, 279, 284, 286, 287, 294, 297, 299 Halobutyl rubber, 445 Heat capacity (Cp), 419 Heterogeneous slip, 78 High heat deflection temperature (HDT), 436 High impact polypropylene (hiPP), 198 Homogeneous slip, 78 Hydrogen-bonding, 442 Hydroxyl functional group, 462 Immiscible, 225, 231, 233, 238, 241–244, 246–250, 261, 264 Impact copolymer PP (ICP), 198–202, 221 Impact PP copolymer (IPC), 198, 199 Impact strength, 411–413, 420, 421, 430–432, 438, 527, 528, 530, 543, 244, 547–551 Initiators, 273, 278, 279, 283, 296, 300 In situ grafting and dynamic vulcanization (ISGV), 432 In situ polymerization, 411 Interactions, 283, 289–292 Interfacial adhesion force, 254 Interlaminar deformation, 77 Intermolecular chain transfer, 64 Interphase interaction, 528, 530, 531, 533–535, 537, 543, 544, 547, 550, 551 Intramolecular branch distribution, 63, 72 Intramolecular chain transfer, 64 Ion-dipole, 442 Isopropyl alcohol, 447 Itaconate, 414 Izod impact strength, 412, 413, 420–432 Lamella, lamellar, lamellae, 226, 231, 233, 237, 238, 241, 244, 246, 248–250, 252, 253, 255, 256, 263 Lamellar fragmentation, 77

Index Lamellar rotation, 77 LCP blends, 501, 509–511, 519, 523 concentration, 502, 515 copolymer, 508, 510, 517, 519, 523 domains, 508, 509, 515, 520 phase, 502, 504, 510, 515, 523 Light scattering, 200, 412 Liquid-liquid phase separation, 68, 72, 74 Loss modulus (E00 ), 226, 228, 234, 235, 240 Lower critical temperature (LCST), 207 Macroradicals, 272–275, 277, 279, 282–284, 287, 290, 294, 297, 299 MAH, 432, 434 MAH-g-EPDM, 452, 469 Maleic anhydride (MAH), 212, 213, 274–276 Maleic anhydride grafted PP, 416 Matrix phase, 455 MBTS, 452, 462 Mechanical properties, 198, 199, 203, 211, 221, 411, 512, 414, 417, 421, 502, 506, 507, 523 Melt extrusion, 411 Melt flow index (MFI), 287–290, 294, 297, 298, 301 Melt mixing, 412 Melt strength, 62, 64–68, 76 Mesophase, 531, 533, 535 Metallocene catalyst, 61, 66, 68, 73, 199, 201, 203, 214, 221 Miscibility, 76, 77, 224, 225, 231, 235, 238, 442 Molar mass distribution, 61, 62, 65–68 Molecular weight distribution (MWD), 200, 201, 207 Monomethyl itaconate (MMI), 414 Morphology, 77–79, 198, 199, 207, 412, 416–424, 431, 435, 437, 438 Na-neutralized ionomer, 425 Natural rubber (NR)/EPDM, 441, 442, 445–469 Nitrobenzene, 466 Non-Newtonian liquid, 416 Notch radius (R), 413 Nucleation, 510, 516, 517–519, 523

665

Olefins, 269, 274, 275, 277, 286, 293, 298 Optical microscopy, 74, 75, 413 Organic peroxide, 421 Oriented crystal fraction (Ioc), 215 Ozone, 442, 444, 463–465 PA/PO blends, 528–531, 533–535, 543 PE-g-SBH, 502–505, 508, 509, 517, 519, 521, 522 Permeability, 466 Permittivity, 462, 463 Peroxide, 198, 272, 275, 279, 282, 286, 288, 294–296 Phase diagram, 63, 72, 74, 75, 207 Phase dissolution, 448 Phase separation, 226, 231, 250, 262, 278 Phenolic resin, 200, 421, 428, 435 Photocrosslinking, 431 Photoinitiation, 431 Polar monomers, 270, 274, 293, 300 Polyamide, 7, 8, 9 Polybutadiene rubber (BR), 450, 454 Polybutene (PB), 5, 6, 9, 17, 225 Polyethylene high density, 5, 17, 225 linear low density, 5, 8, 9, 225 low density, 5, 9, 17 polyethylene/ethylene–propylene rubber, 473 polyethylene–polyamide, 39, 42, 513, 519, 522 polyethylene–polypropylene, 74, 270, 292, 294 polyethylene–polystyrene, 44–48 Polyethylene naphthalate (PEN), 7, 17 Polyethylene terephthalate (PET), 7, 17 Polyheptylene-1, 28 Polyhexylene-1, 28 1,4-Polyisoprene, cis-, trans, 29 Polymethyl methacrylate (PMMA), 7, 17 Polymethylpentene (PMP), 5, 17 Polyoctanomer (TOR), 445 trans-Polyoctene rubber, 450 Polyoctylene-1, 28 Polyolefin–polyamide blends, 43, 48 Polyolefin–polystyrene blends, 43–45 Polypentene-1, 28 Polyphenylene ether (PPE), 7, 29

666

Index

Polyphenylene ether–polystyrene blends, 29 Polyphynele oxide (PPO), 7, 17 Polyphenylene sulphide (PPS), 7, 17 Polypropylene, (PP), 5, 6, 9, 16, 18, 27, 28, 36, 38–40, 42, 46–48, 224, 235, 411–438 atactic polypropylene, 256, 265 isotactic polypropylene, 6, 224, 265 polypropylene–polyamide, 49 polypropylene–polystyrene, 47 syndiotactic polypropylene, 235, 265 Poly(propylene-co-ethylene), 214 Polystyrene (PS), 8, 9, 29 Polyvinyl chloride (PVC), 7, 9, 450 Position annihilation lifetime spectroscopy (PALS), 215 Postreactor blend, 201, 221 Power law equation (Ostwald–de Waele equation), 458 PP/EPDM, 418–421 PP-g-MAH, 434 PP-g-MAH/Zn-SEPDM, 416 PP-grafted maleic anhydride (PP-g-MAH), 212, 213 PP(PE)/EOC, 299 PP(PE)/styrene, 300, 309 Pressure–volume–temperature (PVT) relationship, 75 Pro-Am, Pro-hd, 467 Propylene-dominant component, 214 Propylene–ethylene block copolymer, 214 Propylene–ethylene propylene copolymers blends, 38 Propylene–hexene copolymers blends, 38 Propylene–poly(butene-1) blends, 47–49 Proton NMR spectroscopy, 445 Reactive extrusion, 270, 289, 294 Rheology, 76, 412, 431, 438 Rheometric mechanical spectrometry (RMS), 432 Ruthenium tetraoxide (RuO4), 199, 226, 227, 232, 234 SAXS, 73 Scanning electron microscopy (SEM), 358, 415, 503, 546 Self-nucleation and annealing (SSA), 72

Semiflexible LCP, 502, 508, 519, 523 Shear storage moduli (G0 ), 229, 265 Shear thinning, 64–66, 68, 76 Sheet extrusion, 65 Silanol (Si-OH), 447 Silica, 448, 456, 462, 464 Size exclusion chromatography (SEC), 201 Small angle light scattering (SALS), 245 Small angle neutron scattering (SANS), 37, 68, 86, 203 Solubility parameter, 283, 295, 300, 301 Solution blending, 411 Solvent extraction, 68 Sorptivity, 466 Spherulite, 231, 238, 239, 244, 248, 250, 255 Stearic acid, 444 Stereoblock PP, 214 Stirling’s approximation, 30 Strain hardening, 78 Strain-induced crystallization, 218 Stress–strain curve, 77, 243, 262, 263, 455, 573 Stress whitening, 243, 244, 245, 247 Styrene–butadiene rubber (SBR), 198 Sulfonated EPDM ionomer, 414 Surface tension, 34–35, 44, 45 Surface-modified carbon nanofiber (MCNF), 220 Surfactant, 44, 45 Swelling, 539, 540, 541, 542 Talc, 200 Temperature gradient extraction fractionation (TGEF), 200 Temperature rising elution fractionation (TREF), 202 Tensile strength, 68, 211, 212, 220, 397, 412, 423, 428 Thermal properties, 412 Thermodynamics, 30, 32, 33 Thermoplastic elastomers (TPEs), 8, 18, 419 Thermoplastic interpenetrating polymer network (IPN), 425 Thermoplastic polyolefin (TPO), 416, 442 Thermoplastic polyolefin elastomer (TPO), 198, 199

Index Thermoplastic vulcanizate (TPV), 8, 18, 198, 207–214, 221, 412, 419–425, 430, 436–438 Time–temperature master curve, 79 Transmission electron microscope (TEM), 37–43, 47, 49, 74, 199, 201–210, 215, 432 UHMWPE, 5, 9, 10, 18 ULDPE, 5, 18 Ultrablend, 455, 456, 460 Ultrasonic curing, 412, 438 Ultrasonic irradiation, 418 Unoriented crystal fraction (Iuc), 215

667

Upper critical solution temperature (UCST), 33, 37, 63, 72, 74, 75 Van der Waals, 443 WAXS, 73, 208, 215–217, 220 William–Landel–Ferry (WLF) equation, 79 Young’s modulus, 412, 453 Ziegler–Natta, 60, 61, 63, 64, 68, 69, 77 Zn-neutralized ionomer, 425 ZnO, 444

Polyolefin Foams

Polyolefin Composites

Blends Dyeing

Blends Dyeing

Biomedical Applications Polymer Blends

Blends of Natural Rubber

Advances in Polyolefin Nanocomposites

Polymer Blends and Alloys

Phase Behavior of Polymer Blends

Synthetic fibres: Nylon, polyester, acrylic, polyolefin

Polyolefin Characterization: The First International Conference on Polyolefin Characterization (Macromolecular Symposia)

Polymer Blends Handbook 2 vols set

Synthetic Fibres : Nylon, polyester, acrylic, polyolefin

Progress in Olefin Polymerization Catalysts and Polyolefin Materials: Proceedings of the First Asian Polyolefin Workshop, Nara, Japan, December

Thermodynamics of Polymer Blends (Polymer Thermodynamics Library , Vol 1)

Polyolefin Fibres: Industrial and Medical Applications (Woodhead Publishing in Textiles)

Mixing Essential Oils for Magic: Aromatic Alchemy for Personal Blends

Biodegradable Polymer Blends and Composites from Renewable Resources

Polymer Blends Handbook, Volume 1 and Volume 2

Biomedical Applications Polymer Blends (Advances in Polymer Science)

Thermal Oxidation of Polymer Blends: The Role of Structure

Handbook of thermoplastic polyesters: homopolymers, copolymers, blends, and composites

Nano- and Micro-Mechanics of Polymer Blends and Composites

IRJET- Exergetic Performance of Zeotropic Blends in Domestic Refrigerator and its Alternative Refrigerants

Designing with Blends: Conceptual Foundations of Human-Computer Interaction and Software Engineering

Polymer Blends and Composites in Multiphase Systems (Advances in Chemistry 206)

Polymers, Polymer Blends, Polymer Composites And Filled Polymers: Synthesis, Properties, and Applications

Dairy Blends Market Development Analysis, Share and Recent Trends By 2026

IRJET- Performance Evaluation and Emission Testing of Simarouba Seeds Oil Biodiesel Blends in CI Engine

IRJET- An Investigation on the Performance of Diesel Engine using Karanja Oil Diesel Blends

Polyolefin Blends

Polyolefin Foams

Polyolefin Composites

Blends Dyeing

Blends Dyeing

Biomedical Applications Polymer Blends

Blends of Natural Rubber

Advances in Polyolefin Nanocomposites

Polymer Blends and Alloys

Phase Behavior of Polymer Blends

Synthetic fibres: Nylon, polyester, acrylic, polyolefin

Polyolefin Characterization: The First International Conference on Polyolefin Characterization (Macromolecular Symposia)

Polymer Blends Handbook 2 vols set

Synthetic Fibres : Nylon, polyester, acrylic, polyolefin

Progress in Olefin Polymerization Catalysts and Polyolefin Materials: Proceedings of the First Asian Polyolefin Workshop, Nara, Japan, December

Thermodynamics of Polymer Blends (Polymer Thermodynamics Library , Vol 1)

Polyolefin Fibres: Industrial and Medical Applications (Woodhead Publishing in Textiles)

Mixing Essential Oils for Magic: Aromatic Alchemy for Personal Blends

Biodegradable Polymer Blends and Composites from Renewable Resources

Polymer Blends Handbook, Volume 1 and Volume 2

Biomedical Applications Polymer Blends (Advances in Polymer Science)

Thermal Oxidation of Polymer Blends: The Role of Structure

Handbook of thermoplastic polyesters: homopolymers, copolymers, blends, and composites

Nano- and Micro-Mechanics of Polymer Blends and Composites

IRJET- Exergetic Performance of Zeotropic Blends in Domestic Refrigerator and its Alternative Refrigerants

Designing with Blends: Conceptual Foundations of Human-Computer Interaction and Software Engineering

Polymer Blends and Composites in Multiphase Systems (Advances in Chemistry 206)

Polymers, Polymer Blends, Polymer Composites And Filled Polymers: Synthesis, Properties, and Applications

Dairy Blends Market Development Analysis, Share and Recent Trends By 2026

IRJET- Performance Evaluation and Emission Testing of Simarouba Seeds Oil Biodiesel Blends in CI Engine

IRJET- An Investigation on the Performance of Diesel Engine using Karanja Oil Diesel Blends

Recommend Documents