functional fillers for plastics

Functional Fillers for PlasticsEdited by M. Xanthos

Functional Fillers for Plastics. Edited by M. XanthosCopyright © 2005 WILEY-VCH Verlag GmbH & Co KGaAISBN 3-527-31054-1

Elias, H.-G.

Macromolecules4 Volumes

2005

ISBN 3-527-31172-6 (Volume 1)

2006

ISBN 3-527-31173-4 (Volume 2)

2007

ISBN 3-527-31174-2 (Volume 3)

2008

ISBN 3-527-31175-0 (Volume 4)

Elias, H.-G.

An Introduction to PlasticsSecond, Completely Revised Edition

2003

ISBN 3-527-29602-6

Meyer, Th., Keurentjes, J. (Eds.)

Handbook of Polymer ReactionEngineering2 Volumes

2005

ISBN 3-527-31014-2

Also of Interest:

Advincula, R. C., Brittain, B., Rühe, J.,Caster, K. (Eds.)

Polymer Brushes2004

ISBN 3-527-31033-9

Urban, D., Takamura, K. (Eds.)

Polymer Dispersions and TheirIndustrial Applications2002

ISBN 3-527-30286-7

Smith, H. M. (Ed.)

High Performance Pigments2002

ISBN 3-527-30204-2

Kemmere, M. F., Meyer, Th., Keurentjes, J.(Eds.)

Supercritical Carbon Dioxidein Polymer Engineering2005

ISBN 3-527-31092-4

Edited by Marino Xanthos

Functional Fillers for Plastics

Editor:

Prof. Dr. Marino XanthosOtto H. York Department of ChemicalEngineering and Polymer ProcessingInstituteNJ Institute of TechnologyGITC Building, Suite 3901Newark, NJ 07102USA

All books published by Wiley-VCH are carefully pro-duced. Nevertheless, editor, authors and publisher donot warrant the information contained in these books,including this book, to be free of errors. Readers areadvised to keep in mind that statements, data,illustrations, procedural details or other items mayinadvertently be inaccurate.

Library of Congress Card No.: Applied for

British Library Cataloguing-in-Publication Data: A catalogue record for this book is available from theBritish Library.

Bibliographic information published by Die Deutsche BibliothekDie Deutsche Bibliothek lists this publication in theDeutsche Nationalbibliografie; detailed bibliographicdata is available in the Internet at <http://dnb.ddb.de>.

© 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

All rights reserved (including those of translation inother languages). No part of this book may be repro-duced in any form – by photoprinting, microfilm, orany other means – nor transmitted or translated intomachine language without written permission fromthe publishers. Registered names, trademarks, etc.used in this book, even when not specifically markedas such, are not to be considered unprotected by law.

Printed in the Federal Republic of Germany

Printed on acid-free paper

Composition TypoDesign Hecker GmbH, LeimenPrinting Strauss GmbH, MörlenbachBookbinding Litges & Dopf Buchbinderei GmbH,Heppenheim

ISBN-13 978-3-527-31054-8ISBN-10 3-527-31054-1

V

Table of Contents

Preface XV

List of Contributors XIX

Part I Polymers and Fillers 1

1 Polymers and Polymer Composites 3

1.1 Thermoplastics and Thermosets 3

1.2 Processing of Thermoplastics and Thermosets 4

1.3 Polymer Composites 6

1.3.1 Types and Components of Polymer Composites 6

1.3.2 Effects of Fillers/Reinforcements – Functions 8

1.3.3 Rules of Mixtures for Composites 11

1.3.4 Functional Fillers 12

References 16

2 Modif ication of Polymer Mechanical and Rheological Properties withFunctional Fillers 17

2.1 Introduction 17

2.2 The Importance of the Interface 18

2.3 Modification of Mechanical Properties 19

2.3.1 General 19

2.3.2 Modulus of Fiber and Lamellar Composites 22

2.3.3 Modulus of Composites Incorporating Particulates 25

2.3.4 Strength of Composites with Fiber and Lamellar Fillers 27

2.3.5 Strength of Composites Incorporating Particulates 30

2.3.6 Toughness Considerations 30

2.3.7 Temperature and Time Effects 31

2.4 Effects of Fillers on Processing Characteristics of Polymers 32

2.4.1 General 32

2.4.2 Melt Rheology of Filled Polymers 32

2.4.3 Processing/Structure/Property Relationships 35


VI Table of Contents

Symbols 37

Subscripts 37

References 38

3 Mixing of Fillers with Plastics 39

3.1 Introduction 39

3.2 Pretreatment of Fillers 42

3.3 Feeding 42

3.4 Melting 46

3.5 Introduction of Solids and Mixing 47

3.6 Venting 52

3.7 Pressure Generation 52

3.8 Process Examples 53

3.9 Further Information 55

References 55

Part II Surface Modif iers and Coupling Agents 57

4 Silane Coupling Agents 59

4.1 Introduction 59

4.2 Production and Structures of Monomeric Silanes 60

4.3 Silane Chemistry 61

4.4 Types of Silanes 62

4.4.1 Waterborne Silane Systems 62

4.4.2 Oligomeric Silanes 62

4.5 Silane Hydrolysis 63

4.6 Reactivity of Silanes Towards the Filler 65

4.7 Combining Silanes and Mineral Fillers 66

4.7.1 Method I 66

4.7.2 Method II 67

4.7.3 Method III 67

4.7.4 Method IV 68

4.8 Insights into the Silylated Filler Surfaces 68

4.8.1 Spectroscopy 68

4.8.2 Pyrolysis-Gas Chromatography 72

4.8.3 Carbon Analysis 72

4.8.4 Colorimetric Tests 73

4.8.5 Acid-Base Titration 73

4.8.6 Empirical Tests for Hydrophobicity 73

4.8.7 Combined Silane/Colorant Surface Modification 74

4.9 Selection of Silanes 74

4.10 Applications of Specific Silanes 76

4.10.1 Vinylsilanes 76

4.10.2 Aminosilanes 78

4.10.3 Methacryloxysilanes 80

Table of Contents VII

4.10.4 Epoxysilanes 81

4.10.5 Sulfur-Containing Silanes 81

4.11 Trends and Developments 82

References 83

5 Titanate Coupling Agents 85

5.1 Introduction 85

5.2 The Six Functions of the Titanate Molecule 88

5.2.1 Effects of Function (1) 88






5.3 Summary and Conclusions 101

References 102

6 Functional Polymers and Other Modif iers 105

6.1 Introduction 105

6.2 General Types of Modifiers and their Principal Effects 106

6.2.1 Non-Coupling Modifiers 106

6.2.2 Coupling Modifiers 106

6.2.3 Reinforcement Promoters 107

6.3 Modifiers by Chemical Type 107

6.3.1 Carboxylic Acids and Related Compounds 107

6.3.2 Alkyl Organophosphates 121

6.3.3 Alkyl Borates 121

6.3.4 Alkyl Sulfonates 121

6.3.5 Functionalized Polymers 121

6.3.6 Organic Amines and their Derivatives 126

References 127

Part III Fillers and their Functions 129

7 Glass Fibers 131

7.1 Background 131

7.2 Production Methods 132

7.3 Structure and Properties 135

7.4 Suppliers 138

7.5 Cost/Availability 138

7.6 Environmental/Toxicity Considerations 140

7.7 Applications 141

References 147

VIII Table of Contents

8 Mica Flakes 149

8.1 Background 149



8.4 Suppliers 156




8.7.1 General 158

8.7.2 Primary Function 158

8.7.3 Other Functions 159

References 161

9 Nanoclays and Their Emerging Markets 163


9.1.1 Clays, Nanoclays, and Nanocomposites 163

9.1.2 Concept and Technology 163


9.2.1 Raw and Intermediate Materials 164

9.2.2 Purification and Surface Treatment 166

9.2.3 Synthetic Clays 167


9.4 Suppliers 170




References 174

10 Carbon Nanotubes/Nanof ibers and Carbon Fibers 175


10.1.1 Types of Carbon Nanotubes/Nanofibers and their Synthesis 175

10.1.2 Types of Carbon Fibers and their Synthesis 178

10.1.3 Chemical Modification/Derivatization Methods 181

10.1.4 Polymer Matrices 182

10.2 Polymer Matrix Composites 183

10.2.1 Fabrication 183

10.2.2 Mechanical and Electrical Property Modification 184




References 192

11 Natural Fibers 195


11.2 Structure and Production Methods 196

Table of Contents IX

11.3 Properties 197

11.3.1 Chemical Components 197

11.3.2 Fiber Dimensions, Density, and Mechanical Performance 198

11.3.3 Moisture and Durability 199

11.4 Suppliers 200



11.7 Applications (Primary and Secondary Functions) 202

11.7.1 Mechanical Property Modification (Primary Function) 203

11.7.2 Environmental Preference and Biodegradability (Secondary Functions) 203

References 205

12 Talc 207



12.3 Suppliers 212




12.6.1 General 215

12.6.2 Applications by Polymer Matrix 216

References 220

13 Kaolin 221



13.2.1 Primary Processing 223

13.2.2 Beneficiation 223

13.2.3 Kaolin Products 226

13.2.4 Calcination 227

13.2.5 Surface Treatment 228

13.3 Properties 228

13.4 Suppliers 230




13.7.1 Primary Function 232

13.7.2 Secondary Functions 236

References 239

14 Wollastonite 241


14.2 Production 241


X Table of Contents

14.4 Suppliers/Cost 244



References 247

15 Wood Flour 249




15.3.1 Wood Anatomy 253

15.3.2 Chemical Components 254

15.3.3 Density 256

15.3.4 Moisture 257

15.3.5 Durability 259

15.3.6 Thermal Properties 260

15.4 Suppliers 262




15.7.1 Thermosets 263

15.7.2 Thermoplastics 264

Authors’ Note 267

References 268

16 Calcium Carbonate 271

16.1 Background 271



16.4 Suppliers 274




16.7.1 General 275

16.7.2 Polyvinyl Chloride 276

16.7.3 Glass Fiber Reinforced Thermosets 277

16.7.4 Polyolefin Moldings 277

16.7.5 Polyolefin Films 280

16.7.6 Polyolefin Microporous Films 280

16.7.7 Bioactive Composites 283

References 284

17 Fire Retardants 285


17.2 Combustion of Polymers and the Combustion Cycle 285

17.3 Fuel 287

Table of Contents XI

17.4 Smoke 287

17.4.1 Toxicity 288

17.4.2 Visibility 288

17.5 Flammability of Polymers 289

17.6 Mechanisms of Fire Retardant Action 290

17.6.1 Condensed Phase 290

17.6.2 Chemical Effects in the Gas Phase 291

17.7 Classification of Fire Retardants 292

17.7.1 Metal Hydroxides 292

17.7.2 Halogenated Fire Retardants 299

17.7.3 Zinc/Boron Systems 303

17.7.4 Melamines 303

17.7.5 Phosphorus-Containing Flame Retardants 303

17.7.6 Low Melting Temperature Glasses 306

17.8 Tools and Testing 307

17.9 Toxicity 308

17.9.1 Metal Hydroxides 308

17.9.2 Antimony Trioxide 309

17.9.3 Brominated Fire Retardants 309

17.9.4 Chlorinated Fire Retardants 310

17.9.5 Boron-Containing Fire Retardants 311

17.9.6 Phosphorus-Containing Fire Retardants 311

17.10 Manufacturers of Fire Retardants 312

17.11 Concluding Remarks 312

Acknowledgements 312

References 314

18 Conductive and Magnetic Fillers 317


18.2 Carbon Black 318

18.2.1 General 318

18.2.2 Varieties 318

18.2.3 Commercial Sources 321

18.2.4 Safety and Toxicity 321

18.2.5 Surface Chemistry and Physics 321

18.3 Phenomena of Conductivity in Carbon Black Filled Polymers 322

18.3.1 General 322

18.3.2 Percolation Theories 323

18.3.3 Effects of Carbon Black Type 324

18.3.4 Effects of Polymer Matrix 324

18.3.5 Other Applications 326

18.4 Distribution and Dispersion of Carbon Black in Polymers 326

18.4.1 Microscopy and Morphology 326

18.4.2 Percolation Networks 327

18.4.3 CB in Multiphase Blends 327

XII Table of Contents

18.4.4 Process Effects on Dispersion 328

18.5 Other Carbon-Based Conductive Fillers 328

18.6 Intrinsically Conductive Polymers (ICPs) 329

18.7 Metal Particle Composites 330

18.8 Magnetic Fillers 333



Appendix: Measurements of Resistivity 334

References 336

19 Surface Property Modif iers 339


19.2 Solid Lubricants/Tribological Additives 340

19.2.1 General 341

19.2.2 Production 342

19.2.3 Structure/Properties 344

19.2.5 Cost/Availability 353

19.2.6 Environmental/Toxicity Considerations 354

19.2.7 Applications 355

19.3 Anti-Blocking Fillers 356

19.3.1 General 356

19.3.2 Silica as an Anti-Blocking Filler 358

19.3.3 Applications of Silicas 362

References 364

20 Processing Aids 367


20.2 Production 368

20.2.1 Magnesium Oxide 368

20.2.2 Fumed Silica 369

20.2.3 Hydrotalcites 369

20.3 Structure/Properties 370




20.4 Suppliers/Manufacturers 373

20.5 Environmental/Toxicological Considerations 375








References 380

Table of Contents XIII

21 Glass and Ceramic Spheres 381


21.2 Production and Properties 381

21.3 Functions 382

21.4 Suppliers 384

References 385

22 Bioactive Fillers 387


22.2 Bone as a Biocomposite 388

22.3 Bioceramics Suitable for Tissue Engineering Applications 389

22.4 Bioceramics as Functional Fillers 390

22.4.1 Hydroxyapatite (HA) 390

22.4.2 Calcium Phosphate Ceramics 393

22.4.3 Calcium Carbonate 394

22.4.4 Bioactive Glasses 395

22.4.5 Apatite–Wollastonite Glass Ceramics (A-W) 396

22.4.6 Other Bioactive Fillers 396

22.5 Modification of Bioceramic Fillers 397

22.6 Fillers Formed In Situ 397


References 399

23 In Situ Generated Fillers: Organic–Inorganic Hybrids 401


23.2 Methodology for the Production of Organic–Inorganic Hybrids 403

23.3 General Properties of Organic–Inorganic Hybrid Materials 404

23.4 Potential Applications of Hybrids in Polymer Composites and Blends 406

23.4.1 Use of Polyimide–Silica Hybrids in Polymer Composites 406

23.4.2 Utilization of Hybrids in Thermosetting Resins 413


References 419

XV

Preface

It is generally accepted that growth in plastics consumption and the development ofnew and specialized applications are related to advances in the field of multicompo-nent, multiphase polymer systems. These include composites, blends and alloys andfoams. Fillers are essential components of multiphase composite structures; theyusually form the minor dispersed phase in a polymeric matrix.

Increased interest in the use of discontinuous fillers as a means to reduce the priceof molding compounds begun about 30 years ago when increasing oil prices madenecessary the replacement of expensive polymers with less costly additives. Whensuch additives had also a beneficial effect on certain mechanical properties (mostlymodulus and strength) they were also known as reinforcing fillers. Since that time,there has been a considerable effort to extend the uses (and functions) of existingfillers by: a) particle size and shape optimization, b) developing value- added materi-als through surface treatments and c) developing efficient methods for their incor-poration in plastics.

The term “filler” is very broad and encompasses a very wide range of materials. Wearbitrarily define in this book as fillers a variety of natural or synthetic solid particu-lates (inorganic, organic) that may be irregular, acicular, fibrous or f lakey and areused in most cases in reasonably large volume loadings in plastics, mostly thermo-plastics. Continuous fibers or ribbons are not included. Elastomers are also not in-cluded in this definition as well as many specialty additives that are used at low con-centrations (e.g. pigments, lubricants, catalysts, etc).

Among the best known handbooks on fillers that appeared in the English languagein the past 25 years are the detailed works edited by Katz and Milewski (1978, 1987)and Zweifel (2001) and the monograph compiled by Wypych (2000). This present vol-ume is not intended to be a handbook listing individual fillers according to theirgeneric chemical structure or name but rather a comprehensive and up-to-date pres-entation, in a unified fashion, of structure /property/ processing relationships inthermoplastic composites containing discontinuous fillers that would help the iden-tification of new markets and applications. For convenience, fillers are grouped ac-cording to their primary functions that include modification of: a) mechanical prop-erties, b) f lame retardancy, c) electrical and magnetic properties, d) surface proper-ties and e) processability. For each filler there is always a series of additional func-tions. Examples include degradability enhancement, bioactivity, radiation absorption,


XVI

damping enhancement, enhancement of dimensional stability, reduced permeabilityand reduced density.

Functional Fillers has been the focus of International Conferences such as those or-ganized annually by Intertech Corp. in N. America and Europe over the past 10 yearsand attended by the editor and several contributing authors of this book and the bian-nual “Eurofillers” Conference. Judging from the interest generated from these con-ferences it became clear that there is a need for a volume that would capture the cur-rent technologies applicable to “commodity” fillers and compare them with new tech-nologies and emerging applications that would ref lect the multifunctional characterof new or modified existing fillers. Examples of advances in the latter category in-clude nanoplatelets of high aspect ratio produced by exfoliation of organoclays,nanoscale metal oxides, carbon nanotubes, ultrafine talc, TiO2 and hydroxyapatiteparticles, ceramers and ormosils, new rheology modifiers and adhesion promoters,and increased usage of natural fibers. This volume is expected to address the needsof engineers, scientists and technologists involved in the industrially important sec-tor of polymers additives and composites.

The book is divided into three main parts:Part I, entitled Polymers and Fillers, contains a general introduction to polymer com-

posites, a review of the parameters affecting mechanical and rheological properties ofpolymers containing functional fillers and an overview of mixing and compoundingequipment along with methods of filler incorporation in molten and liquid polymers.

Part II focuses on the use of Surface Modifiers and Coupling Agents to enhance theperformance of functional fillers and includes sections on silanes, titanates, func-tionalized polymers and miscellaneous low molecular weight reactive additives.

Part III on Fillers and their functions describes in a systematic manner the most im-portant inorganic and organic functional fillers with examples of existing and emerg-ing applications in plastics. Fillers have been grouped into seven families, each fam-ily representing the primary function of the filler. The families and the correspon-ding fillers covered in this book are:

High Aspect Ratio Mechanical Property Modifiers with detailed description of glassfibers, mica f lakes, nanoclays, carbon nanotubes/nanofibers and carbon fibers,and natural fibers, Chapters 7–11.

Low Aspect Ratio Mechanical Property Modifiers with detailed description of talc,kaolin, wollastonite, wood f lour, and calcium carbonate, Chapters 12–16.

Fire Retardants with emphasis on metal hydroxides but also inclusion of antimonyoxide, ammonium polyphosphate, borate salts and low melting temperatureglasses, Chapter 17.

Electrical and Magnetic Property Modifiers with emphasis on carbon black but alsoinclusion of metal particles and various magnetic fillers, Chapter 18.

Surface Property Modifiers with further division into: a) solid lubricants/tribologicaladditives that include molybdenite, graphite, PTFE and boron nitride and b) an-tiblocking fillers such as silica, Chapter 19.

Processing Aids including rheological modifiers such as MgO and fumed silica, andprocess stabilizers such as hydrotalcites, Chapter 20.

Preface

Preface XVII

Under Specialty Fillers (Chapters 21–23) a variety of multifunctional inorganics arediscussed. They include: a) glass and ceramic spheres with primary functionalities asrheology modifiers and enhancers of dimensional stability (solid spheres) or weightreduction (hollow spheres), b) a variety of phosphate, carbonate and silicate calciumsalts and specialty glasses that show bioactivity in tissue engineering applications and,c) in-situ generated fillers such as organic-inorganic hybrids with important functionsas mechanical property or surface modifiers, depending on the system.

For commercially available fillers, contributing authors to chapters of Part III havebeen asked to broadly adhere to a uniform pattern of information that would include:a) production methods of the respective filler, b) its structure and properties, c) a listof major suppliers, information on availability and prices , d) a discussion of envi-ronmental/toxicity issues including applicable exposure limits proposed by regulato-ry agencies, and e) a concluding section on applications that considers both primaryand secondary functions of each filler and presents specific data on properties andinformation on current and emerging markets.

Many authors used government and company websites as sources for updated Ma-terial Safety Data Sheets (MSDS) and information on the threshold limit values (TLV)for the airborne concentration of filler dusts in the workplace. Reliable informationon possible risks to human health or the environment is extremely important to cur-rent and potential users of existing fillers, or new fillers of different origin and dif-ferent particle size/shape characteristics. It should be recognized that health issueshave been responsible in the past for the withdrawal from certain plastics markets ofnatural and synthetic fibrous fillers with unique properties such a as chrysotile as-bestos, microfibers, whiskers and the recently mandated very low content of crys-talline silica in mineral fillers.

In presenting the different topics of this book, efforts were made to produce self-contained chapters in terms of cited references, abbreviations and symbols. Althoughthis may have resulted in duplication of information, it should be useful to readersinterested in only certain chapters of the book. All Tables, Figures and Equations arelabeled in terms of the chapter where they first appear to facilitate cross-referencing.

The authors who contributed to this book all have significant credentials in thefield of fillers and reinforcements for plastics and represent industry, academe, con-sulting and R&D organizations. Their different backgrounds and insight ref lect ontheir chapter contents, style of presentation and the emphasis placed on the infor-mation presented. I would like to thank my colleagues from the Polymer ProcessingInstitute (Drs. Davidson, Patel, Todd), from industry (Drs. Ashton, Dey, Mack andWeissenbach, and Messrs. Duca, Kamena and Monte), from academic institutions(Drs. Flaris, Iqbal, Mascia and Rothon), from US Government laboratories (Drs.Clemons and Caulfield), and our doctoral students at the New Jersey Institute ofTechnology (Ms. Chouzouri and Mr. Goyal) for their hard work and excellent cooper-ation and patience during the long editing process. Special thanks are due to myfriends and coworkers for many years in the field of polymer composites, Dr. LenoMascia of Loughborough University, UK and Dr. Ulku Yilmazer of Middle East Tech-nical University, Turkey for reading Chapters 1 and 2 and offering many useful sug-gestions for their improvement. Finally, many thanks to Dr. Michael Jaffe of NJIT for

XVIII Preface

his input on Chapter 22 and to a great number of colleagues and graduate studentsthat I have been associated with over a 30-year industrial and academic career in poly-mer modification and multicomponent polymer systems. They were instrumental inhelping me summarize the concepts and information presented in this book.

This book was largely completed during a sabbatical leave of the editor/contribut-ing author from the New Jersey Institute of Technology in 2003-2004.

Fort Lee, NJ, USA, November 2004 Marino Xanthos

XIX

Authors

Henry C. AshtonEngineered MaterialsJ. M. Huber Corporation251 Gordon StreetFairmount, GA 30139U.S.A.

Daniel F. Caulf ieldUSDA Forest Service, Forest ProductsLaboratoryOne Gifford Pinchot DriveMadison, WI 53705–2398U.S.A.

Georgia ChouzouriOtto H. York Department of ChemicalEngineeringNew Jersey Institute of TechnologyNewark, NJ 07102U.S.A.

Craig M. ClemonsUSDA Forest Service, Forest ProductsLaboratoryOne Gifford Pinchot DriveMadison, WI 53705–2398U.S.A.

Theodore DavidsonPolymer Processing InstituteSuite 3901, GITC BuildingNew Jersey Institute of TechnologyNewark, NJ 07102U.S.A.

Subir K. DeySonocoMail Code P211, North Second StreetHartsville, SC 29501U.S.A.

Joseph DucaEngelhard Corporation101 Wood AvenueIselin, NJ 08830U.S.A.

Vicki FlarisBronx Community College of the CityUniversity of New YorkUniversity Ave. and W. 181st St.Bronx, NY 10453–3102U.S.A.


List of Contributors

Editor

Marino XanthosOtto H. York Department of Chemical Engineering/Polymer Processing InstituteSuite 3901, GITC BuildingNew Jersey Institute of TechnologyNewark, NJ 07102U.S.A.

XX List of Contributors

Amit GoyalOtto York Department of ChemicalEngineeringNew Jersey Institute of TechnologyNewark, NJ 07102U.S.A.

Zafar IqbalDepartment of Chemistry and EnvironmentalScienceNew Jersey Institute of TechnologyNewark, NJ 07102U.S.A.

Karl KamenaSouthern Clay Products1212, Church StreetGonzales, TX 78629U.S.A.

Helmut MackDegussa AGUntere Kanalstr. 379618 Rheinfelden BadenGermany.

Leno MasciaInstitute of Polymer Technology andMaterials EngineeringLoughborough UniversityLoughborough, LE11 3TUU.K.

Salvatore J. MonteKenrich Petrochemicals, Inc.140 East 22nd StreetBayonne, NJ 07002-0032U.S.A.

Subhash H. PatelPolymer Processing InstituteSuite 3901, GITC BuildingNew Jersey Institute of TechnologyNewark, NJ 07102U.S.A.

Roger N. RothonRothon Consultants/ManchesterMetropolitan University3 Orchard Croft, Guilden SuttonChester, CH3 7SLU.K.

David B. ToddPolymer Processing InstituteSuite 3901, GITC BuildingNew Jersey Institute of TechnologyNewark, NJ 07102U.S.A.

Kerstin WeissenbachDegussa Corp.2 Turner PlacePiscataway, NJ 08855U.S.A.

Marino XanthosOtto H. York Department of ChemicalEngineering/Polymer Processing InstituteSuite 3901, GITC BuildingNew Jersey Institute of TechnologyNewark, NJ 07102U.S.A.

Part IPolymers and Fillers


1Polymers and Polymer Composites

Marino Xanthos

1.1Thermoplastics and Thermosets

Almost 85% of polymers produced worldwide are thermoplastics [1]. They can be di-vided into two broad classes, amorphous and crystalline, depending on the type oftheir characteristic transition temperature. Amorphous thermoplastics are charac-terized by their glass-transition temperature, Tg, a temperature above which the mod-ulus decreases rapidly and the polymer exhibits liquid-like properties; amorphousthermoplastics are normally processed at temperatures well above their Tg. Glass-transition temperatures may be as low as 65 °C for polyvinyl chloride (PVC) and upto as high as 295 °C for polyamideimide (PAI) [1]. Crystalline thermoplastics, or morecorrectly, semicrystalline thermoplastics can have different degrees of crystallinityranging from 20 to 90%; they are normally processed above the melting temperature,Tm, of the crystalline phase and the Tg of the coexisting amorphous phase. Meltingtemperatures can be as high as 365 °C for polyetherketone (PEK), as low as 110 °C forlow density polyethylene (LDPE), and even lower for ethylene–vinyl acetate (EVA)copolymers [1]. Upon cooling, crystallization must occur quickly, preferably within afew seconds. Additional crystallization often takes place after cooling and during thefirst few hours following melt processing.

Over 70% of the total production of thermoplastics is accounted for by the large vol-ume, low cost commodity resins: polyethylenes (PE) of different densities, isotacticpolypropylene (PP), polystyrene (PS), and polyvinyl chloride (PVC). Next in perform-ance and cost are acrylics, acrylonitrile–butadiene–styrene (ABS) terpolymers, andhigh-impact polystyrene (HIPS). Engineering plastics, such as acetals, polyamides,polycarbonate, polyesters, polyphenylene oxide, and blends thereof are increasinglybeing used in high performance applications. Specialty polymers such as liquid-crys-tal polymers, polysulfones, polyimides, polyphenylene sulfide, polyetherketones, andf luoropolymers are well established in advanced technology areas due to their highTg or Tm values (290–350 °C).

Common thermosetting resins are unsaturated polyesters, phenolic resins, aminoresins, urea/formaldehyde resins, polyurethanes, epoxy resins, and silicones. Less


4

common thermosets employed in specialized applications are polybismaleimides,polyimides, and polybenzimidazoles. Thermosetting resins are usually low viscosityliquids or low molecular weight solids that are formulated with suitable additivesknown as cross-linking agents to induce curing, and with fillers or fibrous rein-forcements to enhance properties as well as thermal and dimensional stability. It hasbeen frequently stated that in view of their excessive brittleness many thermosetswould have been nearly useless had they not been combined with fillers and rein-forcing fibers.

1.2Processing of Thermoplastics and Thermosets

The operation by which solid or liquid polymers are converted to finished productsis generally known as polymer processing. Polymer processing consists of severalsteps [2]:

a) Pre-shaping operations involving all or some of the following individual opera-tions:

handling of particulate solids (particle packing, agglomeration, gravitational f low,compaction, and others);

melting or heat softening; pressurization and pumping of the polymer melt; mixing for melt homogenization or dispersion of additives; devolatilization and stripping of residual monomers, solvents, contaminants.

The common goal of the above operations is to deliver thermoplastics or cross-link-able thermosets in a deformable f luid state that will allow them to be shaped by a dieor mold; thereafter, they can be solidified by cooling below Tg or Tm (thermoplastics)or by chemical reaction (thermosets).

b) Shaping operations, during which “structuring” occurs (morphology develop-ment and molecular orientation to modify and improve physical and mechanicalproperties). Principal shaping methods include die-forming, molding, casting, cal-endering, and coating.

c) Post-shaping operations, such as decorating, fastening, bonding, sealing, weld-ing, dyeing, printing, and metallizing.

Following the explosive development of thermoplastics after World War II, manyimprovements and new developments have led to today’s diversity of polymer pro-cessing machines and technologies. Some processes are unique to thermoplastics;some are only applicable to thermosets and cross-linkable thermoplastics, while oth-ers, after certain modifications, can be applied to both thermoplastics and ther-mosets. Table 1-1, adapted from ref. [3], summarizes the principal processing/shap-ing methods. For thermoplastics, extrusion is the most popular, with approximately50% of all commodity thermoplastics being used in extrusion process equipment to

1 Polymers and Polymer Composites

5

produce profiles, pipes and tubing, films, sheets, wires, and cables. Injection mold-ing follows as the next most popular processing method, accounting for about 15%of all commodity thermoplastics processed. Other common methods include blowmolding, rotomolding, thermoforming, and calendering.

Tab. 1-1 Principal processing methods for thermoplastics and thermosets

Thermoplastics Thermosets/Cross-linkable thermoplastics

Extrusion Compression Molding, Transfer Molding, – Pipe, tubing, sheet, cast f ilm, profile Casting– Blown film– Coextrusion, extrusion coating– Wire & cable coating– Foam extrusion

Injection Molding, Resin Injection Molding Injection Molding, Resin Injection Molding (RIM) (RIM)

Foam Molding Polyurethane Foam Molding– Structural– Expandable bead

Thermoforming– Vacuum– Pressure forming

Rotational Molding, Calendering

Open-Mold Reinforced Plastics– Lay-up– Spray-up– Filament Winding

Closed-Mold Reinforced Plastics– Pultrusion– Resin transfer molding (RTM)

The range of processes that may be used for fabricating a plastic product is deter-mined by the scale of production, the cost of the machine and the mold, and the ca-pabilities and limitations of the individual processes. For example, complex and pre-cise shapes can be achieved by injection molding, hollow objects by blow molding orrotational molding, and continuous lengths by extrusion. Processing methods forthermosets, particularly those related to reinforced thermosets involving liquid poly-mers, are often quite different from those employed for thermoplastics.

Increased polymer consumption over the past twenty years has not only stimulat-ed machinery sales, but has also led to a parallel growth in the usage of a large vari-ety of liquid and solid modifiers, including fillers and reinforcements [4]. Significantadvances have been made to accommodate such additives by improving the efficien-cy of polymer mixing/compounding equipment. Thermoplastic resin compounderscombine the polymer(s) with the modifiers in high intensity batch mixers and con-tinuous extruders (mostly twin-screw extruders) and the material is then pumped in-

1.2 Processing of Thermoplastics and Thermosets

6

to a pelletizer to produce the feed for subsequent shaping operations (see Chapter 3).Thermosetting resin suppliers compound heat-sensitive resins with fillers, additives,and/or pigments in a variety of mixers to produce molding compounds in such formsas powders, granules, and pastes to be fed into the molding equipment.

1.3Polymer Composites

Modification of organic polymers through the incorporation of additives yields, withfew exceptions, multiphase systems containing the additive embedded in a continu-ous polymeric matrix. The resulting mixtures are characterized by unique mi-crostructures or macrostructures that are responsible for their properties. The pri-mary reasons for using additives are:

property modification or enhancement; overall cost reduction; improving and controlling of processing characteristics.

Important types of modified polymer systems include polymer composites, poly-mer–polymer blends, and polymeric foams.

1.3.1Types and Components of Polymer Composites

Polymer composites are mixtures of polymers with inorganic or organic additiveshaving certain geometries (fibers, f lakes, spheres, particulates). Thus, they consist oftwo or more components and two or more phases. The additives may be continuous,e.g. long fibers or ribbons; these are embedded in the polymer in regular geometricarrangements that extend throughout the dimensions of the product. Familiar exam-ples are the well-known fiber-based thermoset laminates that are usually classified ashigh performance polymer composites. On the other hand, the additives may be dis-continuous (short), as, for example, short fibers (say <3 cm in length), f lakes,platelets, spheres or irregulars; these are dispersed throughout the continuous ma-trix. Such systems are usually based on a thermoplastic matrix and are classified aslower performance polymer composites compared to their counterparts with contin-uous additives; they form the topic of this book.

Nature uses composites for all her hard materials. These are complex structuresconsisting of continuous or discontinuous fibrous or particulate material embeddedin an organic matrix acting as a glue. Wood is a composite of fibrous cellulose andlignin. Bone is a composite of collagen and other proteins and calcium phosphatesalts. The shells of mollusks (Figure 1-1) are made of layers of hard mineral separat-ed by a protein binder [5]. A similar platy structure providing a tortuous path for va-pors and liquids can be obtained by embedding mica f lakes in a synthetic polymericmatrix (Figure 1-2).


71.3 Polymer Composites

Fig. 1-1 A natural composite: the shell of a mollusk made up of layers of calcium salts separated by protein (reprinted from ref. [5]).

Fig. 1-2 A synthetic composite: SEM photograph of a cross-sectionof a fractured mica thermoset composite showing mica f lakes withthickness ~2.5 µm separated by a much thicker polymer layer(courtesy of the author).

8

Additives for polymer composites have been variously classified as reinforcements,fillers or reinforcing fillers. Reinforcements, being much stiffer and stronger thanthe polymer, usually increase its modulus and strength. Thus, mechanical propertymodification may be considered as their primary function, although their presencemay significantly affect thermal expansion, transparency, thermal stability, etc. Forcomposites containing continuous reinforcements, mostly in thermosetting matri-ces, the long fibers or ribbons, when pre-arranged in certain geometric patterns, maybecome the major component of the composite (they can constitute as much as 70%by volume in oriented composites). For discontinuous composites, the directional re-inforcing agents (short fibers or f lakes) are arranged in the composite in differentorientations and multiple geometric patterns, which are dictated by the selected pro-cessing and shaping methods, most often extrusion or injection molding. In thiscase, the content of the additive does not usually exceed 30–40% by volume. It shouldbe noted, however, that manufacturing methods for continuous oriented fiber ther-moplastic composites are available that are amenable to much higher fiber contents,as used in high performance engineering polymers [6]. In this book, the term rein-forcement will be mostly used for long, continuous fibers or ribbons, whereas theterms filler, performance filler or functional filler will mostly refer to short, discon-tinuous fibers, f lakes, platelets or particulates.

In general, parameters affecting the properties of polymer composites, whethercontinuous or discontinuous, include:

the properties of the additives (inherent properties, size, shape); composition; the interaction of components at the phase boundaries, which is also associated

with the existence of a thick interface, known also as the interphase; this is oftenconsidered as a separate phase, controlling adhesion between the components;

the method of fabrication.

With regard to methods of fabrication, all the processes in Table 1-1 that are appli-cable to unfilled, unmodified thermoplastics can also be used for discontinuous sys-tems (with the exception of expandable bead molding). In addition to thermoform-ing, hot stamping of reinforced thermoplastic sheets containing mostly randomlyoriented continuous or discontinuous fibers is used for the production of large semi-structural parts. Fillers can also be used in the thermoset processes in Table 1-1, of-ten in the presence of the primary continuous fiber reinforcement. The concentra-tion and inherent properties of the additive, as well as its interaction with the matrix,are important parameters controlling the processability of the composite.

1.3.2Effects of Fillers/Reinforcements – Functions

Traditionally, fillers were considered as additives, which, due to their unfavorablegeometrical features, surface area or surface chemical composition, could only mod-erately increase the modulus of the polymer, while strength (tensile, f lexural) re-


9

mained unchanged or even decreased. Their major contribution was in lowering thecost of materials by replacing the more expensive polymer; other possible economicadvantages were faster molding cycles as a result of increased thermal conductivityand fewer rejected parts due to warpage. Depending on the type of filler, other poly-mer properties could be affected; for example, melt viscosity could be significantly in-creased through the incorporation of fibrous materials. On the other hand, moldshrinkage and thermal expansion would be reduced, a common effect of most inor-ganic fillers.

The term reinforcing filler has been coined to describe discontinuous additives,the form, shape, and/or surface chemistry of which have been suitably modified withthe objective of improving the mechanical properties of the polymer, particularlystrength. Inorganic reinforcing fillers are stiffer than the matrix and deform less,causing an overall reduction in the matrix strain, especially in the vicinity of the par-ticle as a result of the particle/matrix interface. As shown in Figure 1-3, the fiber“pinches” the polymer in its vicinity, reducing strain and increasing stiffness [7]. Re-inforcing fillers are characterized by relatively high aspect ratio, α, defined as the ra-tio of length to diameter for a fiber, or the ratio of diameter to thickness for plateletsand f lakes. For spheres, which have minimal reinforcing capacity, the aspect ratio isunity. A useful parameter for characterizing the effectiveness of a filler is the ratio ofits surface area, A, to its volume, V, which needs to be as high as possible for effec-tive reinforcement. Figure 1-4 (from ref. [7]) shows that maximizing A/V and parti-cle–matrix interaction through the interface requires α >> 1 for fibers and 1/α << 1for platelets.

In developing reinforcing fillers, the aims of process or material modifications areto increase the aspect ratio of the particles and to improve their compatibility and in-terfacial adhesion with the chemically dissimilar polymer matrix. Such modificationsmay enhance and optimize not only the primary function of the filler (in this case itsuse as a mechanical property modifier), but may also introduce or enhance addition-al functions. New functions attained by substitution or modification of existing


Fig. 1-3 A cylindrical reinforcing fiber in a polymer matrix: a) in theundeformed state; b) under a tensile load (reprinted with permis-sion of Oxford University Press from ref. [7]).

10

fillers, thus broadening their range of applications, are illustrated by the examples be-low.

As described by Heinold [8], the first generation of fillers soon after the commer-cialization of polypropylene included talc platelets and asbestos fibers for their bene-ficial effects on stiffness and heat resistance. The search for a replacement for as-bestos due to health issues led to calcium carbonate particles and mica f lakes as thesecond-generation fillers. Mica was found to be more effective than talc for increas-ing stiffness and heat resistance, while calcium carbonate proved to be less effectivein increasing stiffness, but increased the impact resistance of PP homopolymers.Surface modification of mica with coupling agents to enhance adhesion and stearatemodification of calcium carbonate to assist dispersion were found to enhance thesefunctions and introduced other benefits such as improved processability, a means ofimparting color, and reduced long term heat ageing. Other fillers imparted entirelydifferent functions. For example, barium sulfate enhances sound absorption, wollas-tonite enhances scratch resistance, solid glass spheres add dimensional stability andincrease hardness, hollow glass spheres lower density, and combinations of glassfibers with particulate fillers provide unique properties that cannot be attained withsingle fillers.

An additional example of a family of fillers imparting distinct new properties isgiven by the pearlescent pigments produced by platelet core-shell technologies [9].These comprise platelets of mica, silica, alumina or glass substrates coated with filmsof oxide nanoparticles, e.g. TiO2, Fe2O3, Fe3O4, Cr2O3 (Figure 1-5). In addition to con-ventional decorative applications, new functional applications such as solar heat re-


Fig. 1-4 Surface area-to-volume ratio, A/V, of a cylindrical particleplotted versus aspect ratio, a = l/d (reprinted with permission ofOxford University Press from ref. [7]).

11

f lection, laser marking of plastics, and electrical conductivity are possible through se-lection of the appropriate substrate/coating combinations.

1.3.3Rules of Mixtures for Composites

Rule of mixtures equations (often modified according to the type, shape, and orien-tation of the reinforcement/filler) are commonly used to describe certain propertiesof composites. For example:

a) Concentrations are usually expressed by volume, as volume fractions of filler, Vf,and matrix Vm, obtained from the volumes, vf and vm, of the individual components:

Vf = vf/(vf + vm) (1-1)Vm = vm/(vf + vm) (1-2)Vf + Vm = 1 (1-3)

b) Volume fractions are also used to predict a theoretical density of the composite,ρ, based on the respective densities of the components and assuming a total absenceof voids:

ρ = Vfρf + (1 – Vf)ρm (1-4)

c) The total cost per unit weight of composite, C, can also be calculated from thevolume fractions and the costs of the individual components and the cost of com-pounding per unit weight of composite, Ci [6].

C = Vf ρf /ρ Cf + (1 – Vf )ρm/ρ Cm + Ci (1-5)

After introducing incorporation costs, the cost of the composite may be higher orlower than that of the unfilled polymer. For low cost commodity plastics such as


Fig. 1-5 SEM photographs of mica f lakes (left) and a cross-sectionof an anatase/mica pigment particle (right) [9].

12

polypropylene, the term filler (implying cost reduction) may be a misnomer sincemanufacturing costs may offset the lower cost of most mineral fillers. For high cost,specialty high temperature thermoplastics, the final cost of, for example, glass fiberreinforced polyetherimide, is usually less than that of the unmodified polymer.

Rule of mixtures equations are also used to describe mechanical, thermal, and oth-er properties, as shown in Chapter 2.

1.3.4Functional Fillers

1.3.4.1 Classif ication and TypesThe term filler is very broad and encompasses a very wide range of materials. In thisbook, we arbitrarily define as fillers a variety of solid particulate materials (inorgan-ic, organic) that may be irregular, acicular, fibrous or plate-like in shape and whichare used in reasonably large volume loadings in plastics. Pigments and elastomericmatrices are not normally included in this definition.

There is significant diversity in the chemical structures, forms, shapes, sizes, andinherent properties of the various inorganic and organic compounds that are used asfillers. They are usually rigid materials, immiscible with the matrix in both themolten and solid states, and, as such, form distinct dispersed morphologies. Theircommon characteristic is that they are used at relatively high concentrations (> 5% byvolume), although some surface modifiers and processing aids are used at lower con-centrations. Fillers may be classified as inorganic or organic substances, and furthersubdivided according to chemical family (Table 1-2) or according to their shape andsize or aspect ratio (Table 1-3). In a recent review [10], Wypych reported more than 70types of particulates or f lakes and more than 15 types of fibers of natural or synthet-ic origin that have been used or evaluated as fillers in thermoplastics and thermosets.The most commonly used particulate fillers are industrial minerals such as talc, cal-cium carbonate, mica, kaolin, wollastonite, feldspar, and barite.

Tab. 1-2 Chemical families of fillers for plastics

Chemical Family Examples

InorganicsOxides Glass (fibers, spheres, hollow spheres, f lakes), MgO, SiO2, Sb2O3,

Al2O3

Hydroxides Al(OH)3, Mg(OH)2

Salts CaCO3, BaSO4, CaSO4, phosphatesSilicates Talc, mica, kaolin, wollastonite, montmorillonite, nanoclays, feldspar,

asbestosMetals Boron, steel

OrganicsCarbon, graphite Carbon fibers, graphite fibers and f lakes, carbon nanotubes, carbon

blackNatural polymers Cellulose fibers, wood f lour and fibers, f lax, cotton, sisal, starchSynthetic polymers Polyamide, polyester, aramid, polyvinyl alcohol fibers


13

Tab. 1-3 Particle morphology of fillers

Shape Aspect ratio Examples

Cube 1 Feldspar, calciteSphere 1 Glass spheresBlock 1–4 Quartz, calcite, silica, baritePlate 4–30 Kaolin, talc, hydrous aluminaFlake 50–200++ Mica, graphite, montmorillonite nanoclaysFiber 20–200++ Wollastonite, glass fibers, carbon nanotubes, wood fibers, asbestos

fibers, carbon fibers

A more convenient scheme, first proposed by Mascia [11] for plastics additives, isto classify fillers according to their specific function, such as their ability to modifymechanical, electrical or thermal properties, f lame retardancy, processing character-istics, solvent permeability, or simply formulation costs. Fillers, however, are multi-functional and may be characterized by a primary function and a plethora of addi-tional functions (see Table 1-4). The scheme adopted in this book involves classifica-tion of fillers according to five primary functions, as follows:

mechanical property modifiers (and further subdivision according to aspect ratio); fire retardants; electrical and magnetic property modifiers; surface property modifiers; processing aids.

Additional functions may include degradability enhancement, barrier characteris-tics, anti-ageing characteristics, bioactivity, radiation absorption, warpage minimiza-tion, etc. Such attributes will be identified in subsequent sections of the book.

Tab. 1-4 Fillers and their functions

Primary Examples of Fillers Additional Examples of FillersFunction Functions

Modification of High aspect ratio: Control of Reduced permeability:mechanical glass fibers, mica, nano- permeability impermeable plate-like fillers: properties clays, carbon nanotubes, mica, talc, nanoclays, glass

carbon /graphite fibers, f lakesaramid / synthetic / Enhanced permeability:natural fibers stress concentrators for in-

Low aspect ratio: ducing porosity: CaCO3, dis-

talc, CaCO3, kaolin, wood persed polymers

f lour, wollastonite, glass spheres

Enhancement Hydrated fillers: Bioactivity Bone regeneration:of f ire Al(OH)3, Mg(OH)2 hydroxyapatite, tricalcium retardancy phosphate, silicate glasses


14

Tab. 1-4 Continued

Primary Examples of Fillers Additional Examples of FillersFunction Functions

Modification Conductive, non-conductive, Degradability Organic fillers:of electrical ferromagnetic: metals, starch, cellulosicand magnetic carbon fibers and nano-properties tubes, carbon black, mica

Modification Antiblock, lubricating: Radiation Metal particles, lead oxide, of surface silica, CaCO3, PTFE, MoS2, absorption leaded glassproperties graphite

Enhancement Thixotropic, anti-sag, Improved Isotropic shrinkage, reduced of processability thickeners, acid scavengers: dimensional warpage: particulate fillers,

colloidal silica, bentonite, stability glass beads, micahydrotalcite

Modification Nucleators, clarifiers, of optical irridescent pigments: fine properties particulates, mica/pigment

hybrids

Control of Flake fillers, glass, BaSO4

damping

1.3.4.2 Applications and TrendsGlobal demand for fillers/reinforcing fillers, including calcium carbonate, alu-minum trihydrate, talc, kaolin, mica, wollastonite, glass fiber, aramid fiber, carbonfiber, and carbon black for the plastics industry is estimated to be about 15 milliontons [12]. Primary end-use markets are building/construction and transportation, fol-lowed by appliances and consumer products; furniture, industrial/machinery, elec-trical/electronics and packaging comprise smaller market segments. Flexural modu-lus and heat resistance are the two critical properties of plastics that are enhanced bythe inclusion of performance minerals. Automotive exterior parts, construction ma-terials, outdoor furniture, and appliance components are examples of applicationsbenefiting from enhanced f lexural modulus. Automotive interior and underhoodparts, electrical connectors, and microwaveable containers are examples of applica-tions requiring high temperature resistance [13].

Recent statistics (2001) suggest a combined demand for performance minerals foruse in plastics for North America and Europe of about 4 million tons per annum, withan average annual growth to 2006 forecasted to be about 4.2% [13]. Data (not includ-ing glass products, natural fibers, or nanofillers, but including TiO2) indicate thehighest demand for ground calcium carbonate (60% of the total), followed by TiO2

(13%), aluminum trihydrate (10%), and talc (10%). Kaolin, mica, wollastonite, andbarites have a much smaller share of the market. When glass and natural fibers areincluded in the statistics, calcium carbonate accounts for 40% of the total market,glass for 30%, and other mineral fillers and natural fibers for 20%. Combinations of


15

fillers are also often used to impart specific combined properties not attainable witha single filler. Among polymers, PVC is still the plastic with the highest filler usage,followed by polyolefins, nylons, and polyesters.

Additional growth in the usage of functional fillers will undoubtedly stem from thecurrent efforts towards: a) identifying new applications for composites containingnanofillers [14,15], b) developing composites containing ultrafine particles (dimen-sions < 3 µm), the latter produced by special grinding methods [16], and c) the in-creased usage of natural fiber (f lax, wood) composites, coupled with the expectedsignificant growth in the use of nanoclay composites in the automotive industry.Some exciting new application areas for composites containing nanoclays, nanosili-cates, carbon nanotubes, ultrafine TiO2, talc, and synthetic hydroxyapatite are:

as structural materials with improved mechanical properties, barrier properties,electrical conductivity, and f lame retardancy;

as high performance materials with improved UV absorption and scratch resist-ance;

as barrier packaging for reduced oxygen degradation; as bioactive materials for tissue engineering applications.

Certain issues need to be addressed prior to further growth in the usage of thesenovel functional fillers. For example, the melt processing of nanocomposites and, toa certain extent, of composites containing natural fibers, still presents compoundingproblems related to feeding, dispersion, aspect ratio retention, and orientation of thereinforcement. Figure 1-6 shows complex agglomerates of commercial montmoril-


Fig. 1-6 SEM photograph of a montmorillonite agglomerate priorto its dispersion into high aspect ratio nanoplatelets; 7000× (courtesyof Dr. S. Kim, Polymer Processing Institute).

16

lonite that need to be exfoliated into nanoparticles by melt compounding. Further-more, concerns have been raised relating to the safety of certain nanomaterials in avariety of products, since their inhalation toxicology has not yet been fully evaluatedand few data exist on dermal or oral exposures [15,17].

References

1 Xanthos, M., Todd, D. B., “Plastics Process-ing”, Kirk–Othmer Encyclopedia of ChemicalTechnology, 4th Ed., John Wiley & Sons,New York, 1996, 19, pp 290–316.

2 Tadmor, Z., Gogos, C. G., Principles of Poly-mer Processing, John Wiley & Sons, Inc.,New York, 1979.

3 Xanthos, M., “Polymer Processing”, Chap-ter 19 of Applied Polymer Chemistry – 21st

Century (Eds.: Carraher, C. E., Craver, C.D.), Elsevier, Oxford, U.K., 2000, pp355–371.

4 Xanthos, M., “The Physical and ChemicalNature of Plastics Additives”, Chapter 14 ofMixing and Compounding of Polymers – The-ory and Practice (Eds.: Manas-Zloczower, I.,Tadmor, Z.), Carl Hanser Verlag, Munich,New York, 1994, pp 471–492.

5 Vogel, S., Cats’ Paws and Catapults, W. W.Norton & Co., New York, 1998, pp 123–124.

6 Raghupathi, N., “Long Fiber ThermoplasticComposites”, Chapter 7 of Composite Mate-rials Technology (Eds.: Mallick, P. K., New-man, S.), Hanser Publishers, Munich, 1990,pp 237–264.

7 McCrum, N. G., Buckley, C. P., Bucknall, C.B., Principles of Polymer Engineering, 2nd Ed.,Oxford University Press, New York, 1997,pp 242–245.

8 Heinold, R., “Broadening Polypropylene Ca-pabilities with Functional Fillers”, Proc.Functional Fillers 95, Intertech Corp., Hous-ton, TX, Dec. 1995.

9 Pfaff, G., “Special Effect Pigments”, Chap-ter 7 of High Performance Pigments (Ed.:Smith, H. M.), Wiley-VCH, Weinheim,2002, pp 77–101.

10 Wypych, G., Handbook of Fillers, ChemTecPubl., Toronto, Ont., Canada, 2000.

11 Mascia, L., The Role of Additives in Plastics,Edward Arnold, London, U.K., 1974.

12 Mahajan, S., Proc. Functional Fillers for Plas-tics 2003, Intertech Corp., Atlanta, GA, Oct.2003.

13 Harris, T., Proc. Functional Fillers for Plastics2003, Intertech Corp., Atlanta, GA, Oct.2003.

14 Harris, P., Industrial Minerals, Oct. 2003,443, 60–63.

15 Cheetham, A. K., Grubstein, P. S. H., Nan-otoday, Elsevier, Dec. 2003, 16–19.

16 Holzinger, T., Hobenberger, W., IndustrialMinerals, Oct. 2003, 443, 85–88.

17 Warheit, D. B., Materials Today, Feb. 2004,32–35.


2Modif ication of Polymer Mechanical and Rheological Propertieswith Functional Fillers

Marino Xanthos

2.1Introduction

Parameters affecting the performance of polymer composites containing functionalfillers are related to:

1. The characteristics of the filler itself, including its geometry (particle shape, par-ticle size and size distribution, aspect ratio), its surface area and porosity, and itsphysical, mechanical, chemical, thermal, optical, electrical, and other properties.Relevant concepts introduced in Chapter 1 are further discussed in this chapterand also in other chapters dealing with specific fillers and surface modifiers.

2. The type and extent of interactions at the phase boundaries, which affect adhesionand stress transfer from the matrix to the filler. Interfacial interactions are also re-lated to surface characteristics of the filler, such as surface tension and surface re-activity. These are parameters that control its wetting and dispersion characteris-tics. The importance of the interface is also emphasized in Chapters 4–6.

3. The method of incorporation of the filler into the polymer melt (discussed inChapter 3) and its distribution in the final product part; processing/structure/property relationships are brief ly discussed in this chapter and are elaborated inother chapters covering specific fillers.

Given the overall importance of mechanical properties, this chapter focuses on pa-rameters controlling such properties as related to the filler, the filler/polymer inter-face, and the method of fabrication. Concepts presented below may also be applica-ble to the modification of other polymer properties (e.g. permeability, thermal ex-pansion) through the addition of functional fillers.


18

2.2The Importance of the Interface

Interactions at phase boundaries affect not only the mechanical behavior, but also therheology and processing characteristics, environmental resistance, sorption and dif-fusion, and many other properties of composites. The strength (tensile, f lexural) of acomposite and its retention at higher temperatures, after prolonged times, and underadverse environmental conditions are particularly affected by interfacial adhesion.The principal sources of information presented in this section are refs. [1–6].

The extent of adhesion at the polymer/filler interface may be related to various pa-rameters associated with adsorption and wetting. Factors related to adsorption of thepolymer onto the filler are types of interfacial forces (primary, secondary bonds), mo-lecular orientation/conformation at the interface, and polymer mobility. Contact an-gle, surface tension, and substrate critical surface tension are among factors relatedto wetting.

For a drop of liquid in equilibrium on a solid surface, Young’s equation relates in-terfacial tensions at the solid/vapor interface, γ1, liquid/vapor interface, γ2, and sol-id/liquid interface, γ12, with the contact angle, θ, which is a measure of the degree ofwetting taking a value of zero for ideal wetting.γ1 = γ12 + γ2 cosθ (2-1)

Critical surface tension, γc, equals the surface tension of a liquid that exhibits zerocontact angle on the solid. Any liquid (melt) with a surface tension less than that ofthe solid’s critical surface tension will wet the surface. Uncoated inorganic fillers mayhave very high surface tension, γ > 200 mJm–2, whereas polymers such as polystyreneand polyethylene have lower surface tension, γ < 50 mJm–2. Thus, polymer melts willspread on the high energy surfaces of fillers, unless the γc value of the filler is reducedby absorbed water layers (γ = 21.8 mJm–2), by contamination with low surface tensionimpurities, or by surface irregularities. This will result in incomplete wetting andvoid formation at the interface.

The need to minimize contact angle in order to maximize the work of adhesion,Wa, is shown by the following Young–Dupré equations:

Wa = γ1 + γ2 – γ12 (2-2)Wa = γ2 (1 + cosθ) (2-3)

The need to minimize unfavorable interfacial interactions by minimizing the in-terfacial tension, γ12, can also be inferred from the following simplified Good–Giri-falco equations:

γ12 = γ1 + γ2 – 2Φ (γ1γ2)0.5 (2-4)

Wa = 2Φ (γ1γ2)0.5 (2-5)

where Φ is an interaction parameter that depends on polarity. Polarity is defined asthe ratio of the polar component of the surface tension to the total surface tension. Φ

2 Modif ication of Polymer Mechanical and Rheological Properties with Functional Fillers

19

is maximal when the polarities are equal (approaching unity) and is minimal (ap-proximately zero) when the polarities are totally mismatched. It follows that finite γ12

and low Wa are the result of disparity between polarities (as, for example, between anon-polar polyalkene and a hydrophilic polar filler surface). Surface modification offillers can reduce γ12, modify the γc of the filler, and reduce polarity differences.

Surface modification of fibrous or non-fibrous fillers through the introduction ofnew functional groups, or the modification of existing ones, may be accomplished byoxidation, thermal treatment, plasma treatment, vapor deposition, ion exchange, orthrough the application of additives that may react or interact with both the filler andthe polymer matrix. Figure 2-1 [7] shows the structure of a hypothetical mineral andthe availability of multiple sites for either direct interactions or reactions with thepolymeric matrix or through additives such as coupling agents. Surface modificationof the mineral for improved adhesion can, in effect, convert an ordinary filler into avalue-added filler with multiple functionalities. Surface modification is further cov-ered in detail in Chapters 4–6 of this book.

2.3Modif ication of Mechanical Properties

2.3.1General

Modification of mechanical properties and, in particular, the enhancement of modu-lus and strength is undoubtedly one of the most compelling reasons for incorporat-ing functional fillers into thermoplastics. Appropriate selection of a filler based on its

2.3 Modif ication of Mechanical Properties

Fig. 2-1 Structure and reactivity of a hypo-thetical silicate mineral. 1: interlamellarspaces; 2: exchangeable cations (acidic poten-tial); 3: variable valence species; 4: reactive

hydroxylic species; 5: Lewis acids; 6: anion-ex-change sites; 7: bridged hydroxyl groups (re-produced from ref. [7]).

20

size and shape, modulus and strength, and density is of paramount importance in or-der to establish its potential reinforcing capacity and to provide guidelines for itsmethod of incorporation into the polymer. For directional fillers with a certain aspectratio (e.g. short fibers and f lakes or platelets) embedded in thermoplastic matrices,the load is transferred from matrix to fibers or f lakes by a shear stress and the endsof the fibers or f lakes do not bear a load. As a result, the properties of the resultingcomposites are inferior to those of equivalent composites containing continuousfibers or ribbons. Although in some cases thermoplastic composites containing con-tinuous fiber or ribbons have been produced, the methods suitable for the produc-tion of short fiber or f lake composites are conveniently those that are normally usedfor the processing of unfilled thermoplastics (e.g., extrusion, injection molding, blowmolding). Rigid filler particles may break during such operations with a concomitantreduction in aspect ratio.

In Table 2-1, average values for modulus and strength of commercially availablecontinuous and discontinuous inorganic fibers/ribbons/f lakes of different densitiesare compared. Data have been obtained from a variety of sources [1,8–15] and oftenmay be subject to variation, considering measurement difficulties, particularly forshort fillers of different origins. In particular, the strength values should be viewedwith caution since they depend on the method of testing and the effects of f laws andedges. Data for typical polymer matrices are also included, as well as for whiskers orsingle-crystal platelets, which are considered to be virtually f law-free and, therefore,have extremely high strengths. Metallic wires have relatively large diameters and aretypically used as continuous reinforcements. The densities of particulate mineralfillers (e.g., calcium carbonate, silica, talc, kaolin, wollastonite, aluminum hydroxide)range from 2.4 to 2.75 g cm–3. The corresponding Young’s modulus values have beenquoted as ranging from 25 to 35 GPa [14,16]. Data are often expressed in terms of spe-cific properties (modulus or strength over density). It is obvious, therefore, that for aspecific application requiring high stiffness and strength combined with lightweight, the choice of filler with the optimal specific properties would be desirable.

In the following sections, theoretical and empirical treatments that have been usedto describe composite modulus and strength are presented for continuous fillers (as-pect ratio approaching infinity), discontinuous directional fillers (finite aspect ratio,>1), and particulates (aspect ratio unity). An attempt is made to demonstrate theprinciples governing the mechanical behavior of polymer composites through mod-el systems rather than real molded parts having variable filler orientation and distri-bution. For ease of analysis, stresses are only applied in tension, since the situationin f lexure or compression becomes significantly more complicated in multiphase,multicomponent systems. The principal sources of information presented in the fol-lowing sections on modulus and strength are refs. [8,10,12,14,15,17–19]. Parameterscontrolling other mechanical properties are also brief ly covered.



Tab. 2-1 Comparison of commercially available high aspect ratiofibers, ribbons, and platelets

Filler Density, Tensile Axial Modulus, GPa Tensile Axial Strength, g cm–3 (average value) MPa (average value)

Inorganic FibersE-glass fibers 2.54 76 1500S-glass fibers 2.49 86 1900asbestos (chrysotile) fibers 2.5 160 2000boron 2.57 400 3600

Organic Fiberscarbon fibers 1.79–1.86 230–340 (7–13)[a] 3200–2500carbon nanotubes 1.2 1000–1700 180000aramid fibers 1.45 124 (5)[a] 2800polyester (terylene) 1.38 1.2 600nylon fibers 1.14 2.9 800UHMWPE (Spectra 900) 0.97 117 2600

Natural Fiberssisal fibers 1.5 16.7 507

(A. Sisalana)jute fibers – 24.1 900

(C. capsularis)f lax fibers 1.52 110 900

(Lin usitatissimum)cotton fibers 1.50 1.1 350wood fibers 0.6 13.5 400

(aver. tropical hardwoods)wood fibers (Kraft) 1.0 72 900

Metallic Wiressteel wire 7.9 210 2390tungsten 19.3 407 2890

Whiskerssilicon nitride 3.2 350–380 5000–7000silicon carbide 3.2 480 20000aluminum oxide 4.0 700–1500 10000–20000

Ribbons, Flakes, Plateletsglass ribbons 2.47–3.84 59–78 up to 21000[b]

SiC platelets 3.2 480 10000AlB2 platelets 2.7 500 6000mica f lakes 2.7–2.9 175 3000[c]

exfoliated silicate 2.8–3.0 170 up to 1000nanoclay platelets

exfoliated graphite platelets 2.0 1000 10000–20000polymers 0.90–1.35 0.2–3.3 8.5–95

(excluding elastomers)

[a] Anisotropic fibers; values in parenthesesrelate to the radial direction.

[b] Extrinsic property depending on manufactur-ing process.

[c] Maximum value for f lakes with perfect edges;in practice, the strength of small f lakes can beas low as 850 MPa.

22

2.3.2Modulus of Fiber and Lamellar Composites

2.3.2.1 Continuous ReinforcementsUniaxially oriented fiber systems with fibers randomly spaced when viewed from anend cross- section are anisotropic materials, as shown in Figure 2-2 [8]. For continu-ous (“long”) fiber composites, in which all fibers are aligned in one direction, sub-jected to a tensile stress applied along the fiber axis (axis 1), the total stress, σcL, equalsthe weighted sum of stresses in the fibers, σf, and matrix, σm:

σcL = Vfσf + (1 – Vf)σm (2-6)

where Vf is the volume fraction of filler.

Table 2-2 summarizes the most commonly used predictive equations for the longi-tudinal tensile modulus, EcL (parallel to the fiber axis) and the transverse tensile mod-ulus, EcT (perpendicular to the fiber axis) of the aligned fiber composite. Such pre-dictions have been verified in a plethora of experimental systems. Derivation of thelongitudinal modulus, Eq. (2-7), assumes an elastic fiber and matrix, equal Poisson’sratios, good adhesion, and isostrain conditions, i.e. strains in the fibers and the ma-trix equal to the strain in the composite. Derivation of the transverse modulus,Eq. (2-8), assumes isostress conditions and is based on a simple model in which allfibers are lumped together in a band normal to the tensile stress applied perpen-dicularly to the fibers (along axes 2 or 3) [19]. Table 2-2 also contains a predictiveequation for the modulus of aligned composites tested at intermediate angles tofibers axes, Eq. (2-9). Modulus values are also given for random planar and 3-Dorientations, Eqs. (2-10) and (2-11), respectively.

With respect to continuous lamellar composites containing ribbons or tapes,isotropy in the plane is essentially provided at large ribbon aspect ratios (width-to-thickness) [12], without any angle dependency. The corresponding Eq. (2-12) is simi-lar to Eq. (2-7) for fibers. To summarize the information in Table 2-2:

The main parameters affecting composite modulus, Ec, are the fiber or ribbonmodulus, the volume fractions, and the angle of application of the stress relative tothe reinforcement axis.

The highest modulus in the fiber composites is obtained in the longitudinal case,at an application angle of 0°. Longitudinal and transverse moduli provide the up-


Fig. 2-2 Definition of axes in aligned fiber com-posites (reproduced with permission of Oxford Uni-versity Press from ref. [8]).

23

per and lower bounds of modulus vs. fiber concentration curves. In the extremecase of fiber misalignment (90°), the fibers play only a minor role in determiningthe overall stiffness of the composite.

Composite modulus decreases rapidly with increasing orientation angle. Experi-ments confirm that even a few degrees of misalignment can significantly reducethe modulus.

Fibers that are randomly oriented in the plane or in space may provide isotropy, butat the expense of the overall composite modulus.

Isotropy in a plane may be achieved with aligned ribbons; the modulus in this caseis approximately equal to that of an oriented continuous fiber composite tested inthe longitudinal direction.

2.3.2.2 Discontinuous ReinforcementsPrediction of the modulus of a short-fiber composite needs to take into account endeffects since isostrain conditions are not satisfied at the fiber ends. Stress builds upalong each fiber from zero at its end to a maximum at its center. As shown in Fig-ure 1-3, at the interface the matrix is severely sheared at the fiber ends.

For aligned short fibers subject to a stress along the fiber axis, the stress borne bythe fibers is no longer σf as defined in Eq. (2-6), but assumes a lower mean value thattakes into account the effect of the fiber ends and the rate of stress build-up along thelength towards the central regions. Eq. (2-7) can be written in a modified form asEq. (2-13) (see Table 2-3), where K is an efficiency parameter that approaches unityfor very long fibers. Values of K calculated by different authors for single fibers andby assuming random overlap are given by Eqs. (2-14) and (2-15). In these equations,the parameter u includes constants such as fiber length and diameter, fiber volumefraction, fiber modulus, and matrix shear modulus. It has been shown [8,14,15] thatfor optimal efficiency in stress transfer to the fibers, the value of u should be as highas possible; this implies that the aspect ratio and the ratio Gm/Ef should also be as


Tab. 2-2 Comparison of modulus equations for continuous fiber and ribbon composites

Tensile Modulus Continuous Fibers Eq. No. Continuous Ribbons (tapes) Eq. No.of High Aspect Ratio

longitudinal (0o) EcL = VfEf + (1 – Vf)Em 2-7 EcL ≈ EcT = VfEf + (1 – Vf)Em 2-12

transverse (90o) 1/EcT = Vf/Ef + [(1 – Vf)/Em] 2-8 EcL ≈ EcT 2-12assume isotropy in plane

intermediate angles (θ ) EcL/Ecθ = cos4θ + sin4θ EcL/EcT + 2-9[a] EcL ≈ EcT 2-12cos2θ sin2θ (EcL/GcLT – 2νcLT) assume isotropy in plane

random orientation Ec = 3/8 EcL + 5/8 EcT 2-10 EcL ≈ EcT 2-12in plane assume isotropy in plane

random 3-D orientation Ec = 1/5 EcL + 4/5 EcT 2-11 – –

[a] GcLT and νcLT are the longitudinal-transverse shear modulus and Poisson’s ratio,respectively, of the composite.

24

high as possible. Eq. (2-13) may be modified as Eqs. (2-16) and (2-17) to account forthe effect of variation in fiber orientation, which is a natural consequence of melt pro-cessing.

An interpolation procedure applied by Halpin and Tsai [17,18] has led to general ex-pressions for the moduli of composites, as given by Eqs. (2-18) and (2-19). Note thatfor ξ = 0, Eq. (2-18) reduces to that for the lower limit, Eq. (2-8), and for ξ = infinity itbecomes equal to the upper limit for continuous composites, Eq. (2-7). By empiricalcurve-fitting, the value of ξ = 2 (l/d) has been shown to predict the tensile modulus ofaligned short-fiber composites in the direction of the fibers, and the value of ξ = 0.5can be used for the transverse modulus.

Eq. (2-13) has also been used to predict the modulus of f lake (platelet) compositescontaining planar oriented reinforcement for uniform arrays of f lakes, Eq. (2-14), andfor random overlap, Eq. (2-15) [10,12,14,19]. Equations for the parameter u are some-what different from those used for fibers, but they still contain the important pa-rameters affecting the modulus of the composite, i.e. aspect ratio, volume fraction,and f lake/matrix modulus ratio. Eq. (2-18) has also been used to predict the modulusof platelet-reinforced plastics [17,20].


Tab. 2-3 Comparison of modulus equations for short fiber, f lake, and particulate composites

Tensile Modulus Short Fibers Eq. No. Flakes Eq. No. Particulates Eq. No.

longitudinal EcL = KVf Ef + 2-13[a] EcL = KVf Ef + 2-13[a] Ec/Em = (1 + ΑΒVf)/ 2-20[d]

(1 – Vf)Em (1 – Vf)Em (1 – ΒχVf)where K ≤ 1 where K ≤ 1

longitudinal Ec/Em = (1 + ξηVf)/ 2-18[b] Ec/Em = (1 + ξηVf)/ 2-18[b] assume 3-D isotropy(1 – ηVf) for ξ = 2(l/d) (1 – ηVf) for ξ = 2(α)

transverse Ec/Em = (1 + ξηVf)/ 2-18[b] EcL ≈ EcT assume 3-D isotropy(1 – ηVf) for ξ = 0.5 assume isotropy

in plane

intermediate Ec = K’KVfEf + 2-16[c] EcL ≈ EcT assume 3-D isotropyangles (θ ) (1– Vf)Em assume isotropy

in plane

[a] For single fibers K = 1 – (tanhu/u) (2-14)or, by assuming random overlap, K = 1 – [ln(u + 1)/u] (2-15)

[b] η = (Ef/Em – 1)/(Ef/Em + ξ) (2-19)where ξ is an adjustable parameter related to aspectratio

[c] K’ = (2-17)

where αi is the fraction of the fibers oriented at an an-gle θ to the direction in which the value of Ec is re-quired and k is the number of intervals of angle defin-ing the orientation distribution; cf. ref. [10], p. 53.

[d] Where A = KE – 1 with KE geometric factor, B = (Ef/Em – 1)/(Ef/Em + Α) (2-21)and χ = 1 + Vf (1/φ2

max – 1/φmax) (2-22)where φmax is the maximum packing fraction of thefiller (0.74 for hexagonal close packing, 0.524 for sim-ple cubic packing, etc.)

αι ι 4cos θi

k

=∑

1

25

In summary, general parameters affecting the modulus of a rigid polymeric mate-rial containing discontinuous directional reinforcements are the moduli of the rein-forcement and the matrix, the ratio Gm/Ef, the volume concentration of the rein-forcement, the reinforcement orientation, and its size (aspect ratio). Values of initialmodulus (tangent) are not usually affected by the extent of interfacial adhesion; how-ever, secant moduli, measured at higher strains, are usually higher in the case of goodadhesion. In Figure 2-3 [18], Ec/Em (the modulus enhancement factor) is plotted as afunction of fiber or f lake aspect ratio; the curves show the effects of aspect ratio andorientation angle. Theoretical plots of this type have been confirmed in many exper-imental systems. Thus, it can be easily inferred that efficient reinforcement occurswhen volume fraction and aspect ratio are fairly large and there is minimum mis-alignment of fibers or f lakes from the axis or plane of application of the stress. Prac-tical limitations are the limited volume fraction that may be attained depending onthe manufacturing method, the reduced aspect ratio during processing, and the com-plex orientation characteristic of injection-molded parts.

2.3.3Modulus of Composites Incorporating Particulates

Parameters affecting the modulus of rigid plastic matrices (Poisson’s ratio < 0.5) con-taining particles with aspect ratio close to unity are primarily the ratio of the moduliof the two phases, and the volume fraction of the filler. Other parameters are the


Fig. 2-3 Modulus enhancement factor for reinforced thermoplasticsas a function of fiber volume fraction, orientation angle, and aspectratio (adapted from ref. [18]).

26

geometry, packing characteristics, and degree of agglomeration of the filler, and itssize and size distribution. The stiffness behavior of particulate composites can berepresented by Eqs. (2-20) to (2-22) included in Table 2-3, which were devised by Kern-er and further modified by Nielsen and co-workers [1,17]. The theories predict thatthe elastic moduli of rigid particulate composites are lower than those of compositeswith directional fillers, but higher than those of composites containing elastomericparticles or voids (foams), as shown in Figure 2-4 [21]. In this figure, the modulus in-crease obtained with rigid particulate fillers with an aspect ratio of approximatelyunity (system 4) is compared with the modulus increases obtained with long fibersin the longitudinal and transverse directions (systems 7 and 2, respectively), and withshort fibers, which are either aligned (system 6) or randomly oriented (system 5).

The theories suggest that the elastic moduli of composites containing particulateswith an aspect ratio of unity should be independent of the dimensions of the fillerand dependent only on the relative moduli of the filler and the matrix, their volumefractions, and geometric factors. Good adhesion between the organic and inorganicphases is assumed, even though they usually have vastly different coefficients of ther-mal expansion. Thus, even in the case of poor adhesion, the theoretical equations arevalid because there may not be any relative motion at the interface as a result of asqueezing force on the filler surface imposed during cooling down from the fabrica-tion temperature. Although the theories predict no dependence of the modulus of the


Fig. 2-4 Relationship between stiffness and filler type and orienta-tion in polymeric materials (reproduced from ref. [21]).

27

composite on the filler dimensions, experiments show that an increase in modulusis obtained when the particle size is decreased and that there is a corresponding de-crease in modulus when the particle size distribution is shifted to higher values.These discrepancies can be attributed to increased particle surface area and surfaceenergy, and increased maximum packing fractions, as discussed in ref. [17].

2.3.4Strength of Composites with Fiber and Lamellar Fillers

2.3.4.1 Continuous ReinforcementsTable 2-4 contains the most important equations proposed to model the tensilestrength of composites containing continuous fibers or ribbons (microtapes) withlarge aspect ratio as a function of degree of interfacial adhesion and reinforcementorientation.

In summary:

Parameters affecting composite strength, σc, are the ultimate strength of the fibersor ribbons, σfu, the ultimate strength of the matrix, σmu, or the stress borne by thematrix when the fibers or ribbons fail, σ’m, volume fractions, the shear strength ofthe matrix or bond strength, τ, and the angle between the direction of stress andthe fiber axis.


Tab. 2-4 Comparison of strength equations for continuous fiber and ribbon composites

Tensile Strength Continuous Fibers Eq. No. Continuous Ribbons Eq. No.(tapes) of high aspect ratio

longitudinal (0°) σcL = σfuVf + σ’m(1 – Vf) 2-23 σcL = σfuVf + σ’m(1 – Vf) 2-23(perfect adhesion)

longitudinal (0°) (no adhesion or very low Vf) σcL = σmu(1 – Vf) 2-24 σcL = σmu(1 – Vf) 2-24

longitudinal (0°) (intermediate adhesion) σcL = KσfuVf + σ’m(1 – Vf) 2-25 σcL = KσfuVf + σ’m(1 – Vf) 2-25

where 0 < K < 1 where 0 < K < 1

transverse (90°) σcT = bond or matrix strength σcL ≈ σcT

assume isotropy in plane for perfect adhesion

intermediate angles (θ ) 1/σ2cθ = cos4θ/σcL + 2-26[a] σcL ≈ σcT

sin4θ/σcT + cos2θ sin2θ assume isotropy in (1/τ2 – 1/σ2

cL ) plane for perfect adhesion

[a] 0–5° longitudinal tensile failure; 5– 45° shear failure; 45–90° transverse tensilefailure.

28

Maximal strength in the fiber composites is obtained in the longitudinal case, i.e.at an angle of 0° in the case of perfect adhesion. Under these conditions, the com-posite efficiently utilizes the strong reinforcement and the overall failure is by fiberor ribbon fracture. The strength decreases rapidly and the mode of failure changeswith increasing angle. At 90°, failure of the fiber composites occurs by failure ei-ther in the matrix or at the interface.

Fibers that are randomly oriented in the plane may provide isotropy, but at the ex-pense of the overall composite strength.

Isotropy in a plane may be achieved with aligned, large aspect ratio ribbons;strength values in this case are approximately equal to those of an oriented contin-uous fiber composite tested in the longitudinal direction.

2.3.4.2 Discontinuous ReinforcementsIn this case, the matrix deforms more than the filler at the fiber ends and shearstresses are set up at the interface. Failure of the composite may occur either by fiberfracture or bond failure (fiber pull-out), depending on the filler aspect ratio. A criti-cal fiber length, Lcr, defines the transition point between the two modes of failure.Similar concepts apply to f lakes (platelets), for which a critical platelet diameter, dcr,defines the transition point. Table 2-5 summarizes common predictive equationsused to describe the tensile strength of composites containing short fibers orplatelets as a function of orientation, aspect ratio, volume fraction, adhesion, andstrength of matrix and fillers. In these equations, the effects of adjacent fibers orf lakes, and of the presence of edges in irregularly shaped f lakes, are ignored. Thecritical aspect ratio, αcr, determining the transition from fiber or platelet fracture tofailure by debonding or shear failure of the matrix at lower stresses, is defined forfibers (Eqs. (2-27) and (2-28)) as:

αcr = σfu/2τ (2-29)

and for platelets (Eqs. (2-32) and (2-33)) as

αcr = σfu/τ (2-34)

where τ is the interface or matrix shear strength and σfu is the fiber or plateletstrength.

In summary:

The fiber or platelet aspect ratio has to be above the critical value for efficient uti-lization of the reinforcing properties of the filler. In addition to aspect ratio, otherparameters affecting composite strength are the ultimate strength of the filler, theultimate strength of the matrix, σmu, or the stress borne by the matrix when thefiller fails, σm, the volume fractions, bond strength or matrix shear strength, andthe angle of application of the stress relative to the fiber axis.

The strengths of composites containing short fiber or platelet-type fillers are low-er than those of their counterparts with continuous reinforcements.


29

Maximal strength for fiber composites is obtained in the longitudinal case, at anangle of 0°, with good adhesion, and at aspect ratios well above the critical value.Composite strength decreases with increasing angle of application of the stress. Inthe case of randomly oriented long fibers (L > Lcr), the (isotropic) compositestrength is much lower than the longitudinal strength of an oriented fiber-filledcomposite.

The strength of a f lake-containing composite is isotropic in the plane of the ori-ented f lakes and much lower perpendicular to the f lake plane axis. The higheststrength for f lake-containing composites is obtained at an angle of applied stressparallel to the f lake surface, with good adhesion, and with aspect ratios well abovethe critical value, when composite failure is by f lake fracture.

Experimental data on the tensile strengths of a variety of polymer/filler combina-tions compiled by Wypych [9] indicate the complexity of the parameters affectingstrength in real systems.


Tab. 2-5 Comparison of strength equations for short fiber, f lake, and particulate composites

Tensile Strength Short Fibers Eq. No. Flakes Eq. No. Particulates Eq. No.

longitudinal; σcL = σfu(1 – Lcr/2L)Vf 2-27[a] σcL = σfu(1 – αcr/α)Vf 2-32[a] σc = σmu(1 – aVfb + 2-35[c]

for fiber L > Lcr + σ’m(1 – Vf) + σ’m(1 – Vf) cVfd) or

or for platelet σc = σmu(1–1.21Vf2/3) 2-36

α > αcr or σc = λσmu – KVf 2-37[d]

longitudinal; σ’cL = τ L/d Vf + 2-28[b] σ’cL = τ α/2 Vf + 2-33[b] 3-D isotropyfor fiber L < Lcr σmu(1 – Vf) σmu(1 – Vf)or for platelet α < αcr

transverse bond or matrix σcL or σ’cL 3-D isotropystrength

random σc = σfuVf/2 (1 – Lcr/L) 2-30 σcL 3-D isotropyorientation in plane; for fiber L > Lcr or for platelet α > αcr

random σc = σfuVf L/4Lcr 2-31 σ’cL 3-D isotropyorientation in plane; if L < Lc

or α < αcr

[a] Failure by reinforcement fracture.[b] Failure by reinforcement debonding (pull out) or shear

failure of the matrix.

[c] a, b, and c are constants depending on particle size andadhesion.

[d] λ = stress concentration factor; K is a constant depend-ing on adhesion.

30

2.3.5Strength of Composites Incorporating Particulates

Table 2-5 contains a general equation, Eq. (2-35), for the effect of particulate fillers onthe tensile strength of a polymer, a common modification of this equation by Nico-lais and Narkis, Eq. (2-36), and an additional equation proposed by Piggott and Leid-ner, Eq. (2-37) [9,17,19]. These equations predict that failure is by either matrix failureor loss of adhesion without utilization of the inherent strength of the particulate. Ex-perimental results for a variety of non-directional filler particles show that in mostcases tensile strength decreases with increasing volume fraction; relatively higher val-ues, however, are obtained with improved adhesion [9].

2.3.6Toughness Considerations

In addition to the effects on modulus and strength, the use of fillers can, in most cas-es, also improve polymer toughness, i.e. its ability to resist crack propagation, nor-mally expressed as the area under the stress/strain curve. Impact strength, a com-mon measure of toughness at high strain rates or under dynamic conditions, is amost important property for real application conditions.

In short-fiber composites, areas around fiber ends, areas of poor adhesion, and re-gions of fiber–fiber contacts may reduce the resistance to crack initiation by acting asstress concentrators. However, fibers may also reduce crack propagation by divertingcracks around the fibers, or by bridging cracks. Materials with high impact strengthspread the absorbed energy throughout as large a volume as possible to preventbrittle failure. Dissipation of energy in short-fiber composites, and hence an increasein impact strength, may be accomplished by: a) mechanical friction, as in the case ofpull-out of the fibers from the matrix, which prevents localization of the stresses, or,b) by controlled debonding of the fibers, which disperses the region of stress con-centration through a larger volume and tends to stop the crack propagation. The en-ergy dissipated in forming a unit amount of new surface, Γ, by pull-out depends onfiber properties such as strength, length, critical length, and volume fraction:

Γ = [σfuLcr2Vf]/12L for L > Lcr (2-38)

Γ = [σfuL2Vf]/12Lcr for L < Lcr (2-39)

The energy dissipated in forming a unit amount of new surface, Γ, by debondingis given by

Γ = [σfu2LdVf]/4Ef (2-41)

where Ld is the debonded fiber length [17].These theoretical equations for maximum energy dissipation have been correlated

with maximum impact strength for impact loads applied parallel to short fibers oflength close to the critical value and in the case of less than perfect adhesion [17].


31

Transverse impact strength is generally lower than longitudinal, since the aforemen-tioned crack propagation toughening mechanisms are inoperative while crack initia-tion factors still exist. Similar equations have been derived for round platelets andf lakes [19]. In practice, the use of f lakes has, in most cases, been shown to adverselyaffect impact strength, possibly because of significant stress concentration effects as-sociated with shape irregularities and sharp corners.

Control of aspect ratio and bond strength appear to be the most important param-eters for impact strength enhancement through the use of fibers, particularly in thecase of brittle matrices. Maximizing impact strength is usually accomplished at theexpense of the longitudinal tensile strength, which is greatest for longer fibers wellbonded to the matrix. The effect of non-directional fillers on toughness is correlatedin a complex manner to the particular testing method and to the particle size, shape,concentration, and rigidity of the filler, the nature of the interface, and the specifictype of polymer matrix. Studies on a variety of rigid particulate fillers ranging in sizefrom 0.8 to 30 µm and added to a rigid matrix showed a peak in falling dart impactstrength at about 2 µm; studies on CaCO3-filled polypropylene confirmed the bene-ficial effect of less than perfect adhesion on impact strength, as was also observed inthe case of fibers [9]. Impact strength is largely determined by de-wetting and forma-tion of narrow zones of highly deformed and voided polymer crazing as a result ofstress application. In brittle rigid polymers, rigid spherical fillers with higher modu-lus than the matrix act as crack initiators, promoting crack propagation and loweringimpact strength. In tough, rigid polymers, toughness may be increased as a resultof enhanced crazing [17]. A compilation of impact strength data for particulate andfibrous fillers [9] in a variety of polymers suggests that for certain systems correla-tions corresponding to reality may be difficult to establish and that one type of testmay contradict the results of another.

2.3.7Temperature and Time Effects

The mechanical behavior of polymer composites is not only defined by short-termproperties such as stiffness, strength, and toughness, but also by long-term proper-ties such as creep, stress relaxation, and fatigue. All such properties, which are af-fected by temperature and type of environment, can be modified through the addi-tion of fillers.

In general, most fillers increase the heat distortion temperature (HDT) of plasticsas a result of increasing modulus and reducing high-temperature creep. Thermo-elastic properties such as coefficient of thermal expansion (CTE) are also affected bythe presence of fillers and have been modeled with a variety of equations derivedfrom the rule of mixtures [8]. For directional fillers, this property is strongly orienta-tion-dependent, and because of the difference between the CTE of the filler and thatof the matrix, internal stresses may lead to undesirable warpage.

In general, fillers also decrease creep and creep rate of polymers, as long as thereis no serious debonding of the particles. At high strains and long times, whendebonding may occur, creep and creep rate may increase dramatically and, accord-


32

ingly, time for rupture may be significantly decreased. Increased adhesion usuallyminimizes these effects.

2.4Effects of Fillers on Processing Characteristics of Polymers

2.4.1General

Unfilled polymers behave as non-Newtonian liquids during melt processing. Meltrheology has been the subject of extensive investigation over the last 20 years and isdocumented in a large number of texts. The significant effect of dispersed particu-lates, fibers or f lakes on polymer melt rheology and melt elasticity are directly relat-ed to processability with respect to both mixing (compounding) and shaping opera-tions. For directional fillers, understanding the f low-induced orientation and thepossibility of particle segregation to the region of highest f luid velocity are of para-mount importance in controlling the microstructure of the final product and itsproperties. The principal sources of the information presented in this section arerefs. [1,2,10,17,21–24].

2.4.2Melt Rheology of Filled Polymers

Filler effects on viscosity and elasticity depend on several parameters, including con-centration, size, shape, and aspect ratio of the filler; interactions with the polymer;shear rate, the presence of agglomerates, fiber/f lake alignment, and surface treat-ment.

Concentration and shear rateIn general, shear and elongational viscosities increase with increasing filler volumefraction. The effect on shear viscosity is more pronounced at low shear rates; “yield”effects due to the formation of structured networks are often encountered at lowshear rates and at high loadings of submicron particles [10,22]. High shear rates tendto orient fibers and f lakes to different degrees depending on their size, rigidity, con-centration, and interactions with the matrix. The increase in viscosity relative to theunfilled matrix becomes less pronounced at higher shear rates, as shown in Figure2-5 [10]. Much larger deviations from Newtonian behavior than for the correspondingpolymer matrix are observed in filled polymer melts. Velocity profiles in circular andslit channels become very f lat, due to a decrease in the power law index, and plug-likef low behavior is observed [1,23]. Increasing the amount of filler, regardless of shape,reduces melt elasticity, as shown by reduced extrudate (die) swell and concomitant ef-fects on normal stress differences [22]. Reduced melt elasticity has significant practi-cal consequences for extrusion and injection-molding.


33

Filler size and shapeParticle size effects may be negligible at high shear rates since for filled systems highshear viscosity is often governed by the matrix characteristics and low shear viscosi-ty is governed by the filler. Low specific area fillers such as large particulate fillersand low aspect ratio fibers or f lakes have less interactions with the polymer and yieldlower viscosities than higher surface area, higher aspect ratio fillers. High shearstresses tend to break not only filler agglomerates but also cause additional reductionof fiber length or f lake diameter, with a concomitant effect on viscosity.

Filler surface treatmentsInterfacial agents that tend to wet or lubricate the filler surface (titanates, stearates,etc.) tend to reduce viscosity. This may result from attenuated interparticle forces anda reduced tendency for f locculation since polymer molecules may slip between treat-ed filler particles encountering less frictional resistance (see Figure 2-6) [10]. Re-duced particle–particle interactions may lead to further f low orientation of fibers and

2.4 Effects of Fillers on Processing Characteristics of Polymers

Fig. 2-5 Shear viscosity of polystyrene/glass fiber suspension ver-sus shear rate at 180 °C (reproduced from ref. [10], Chapter 5).

34

f lakes and to a further decrease in viscosity at high shear rates. However, an increasein viscosity may occur when the surface treatment causes strong adhesion of the fillerto the polymer.

Several equations have been proposed to predict the ratio of the viscosity of thecomposite to that of the unfilled matrix, µc/µm, and to explain viscosity effects as afunction of filler volume fraction, shape factors, aspect ratio, packing characteristics,interaction parameters, and non-Newtonian or yield parameters. Examples are:

1. the Mooney equation [2,23], which is valid over the entire concentration range

ln(µc/µm) = KeVf /[1 – (Vf/φmax)] (2-42)

where φmax is the maximum packing factor, defined as true volume of filler/ap-parent volume occupied by filler, and Ke is a geometric parameter known as theEinstein coefficient (see also Eq. (3-4)), which depends on aspect ratio and degreeof agglomeration and for rods also on degree of orientation, which, in turn, de-pends on shear rate, and

2. the Nielsen equation [17]


Fig. 2-6 Shear viscosity of uncoat-ed and stearate-coated calciumcarbonate filled polystyrene meltversus shear rate (reproduced withpermission from ref. [10],Chapter 5).

35

µc/µm = (1 + ΑΒVf)/(1 – ΒΨVf) (2-43)

where A, B, and Ψ are functions of component properties, packing characteristics,and aspect ratio.

In practical terms, the high viscosities obtained through the incorporation of fibersand f lakes can be reduced either through the use of wetting agents, through filler ori-entation, and/or by reduction of aspect ratio. Reduction of aspect ratio during pro-cessing may be undesirable but is common for fragile fibers and large f lakes (glass,mica); in contrast, organic f lexible fibers may orient and bend without fracture.These effects will lead to an increase in φmax and a decrease in Ke in Eq. (2-42), witha concomitant decrease in the relative viscosity µc/µm [23].

2.4.3Processing/Structure/Property Relationships

Primary filler properties controlling the morphology and properties of plastic prod-ucts are geometry, concentration, density, modulus, strength, and surface chemistry.Additional filler properties related to processing (compounding and shaping) are:

Hardness, which is usually expressed on the Mohs’ scale. Soft fillers are preferredto hard fillers, which tend to cause excessive wear of processing equipment (e.g.,Mohs’ hardness: 1 for talc, 3 for calcite, and 7 for silica).

Thermal properties such as thermal conductivity, which, for most mineral fillers, isabout an order of magnitude higher than that for thermoplastics; specific heat,which is typically about half of that of polymers; and coefficient of thermal expan-sion (CTE) (see Section 2.3.7), which is lower than that for polymers. The net effectis that most fillers (non-fibrous) usually produce a faster rate of cooling in injec-tion molding, lower volume shrinkage, and promote less warpage and shorter cy-cle times. Fibrous fillers, however, may cause differential shrinkage with an in-creased tendency to warp as a result of orientation. Combinations of fibers withf lakes (planar orientation) or spheres (no orientation) tend to minimize warpage.

Thermal stability (up to 300 °C for high temperature thermoplastics), which is re-quired during processing so that there is minimum weight loss or structuralchanges.

Moisture absorption, which needs to be minimized since it may affect the quality ofthe compound or the stability of hydrolytically sensitive matrices such as nylons orpolyesters.

Quantitative predictions of the effects of fillers on the properties of final productsare difficult to make, considering that they also depend on the method of manufac-ture, which controls the dispersion and orientation of the filler and its distribution inthe final part. Short-fiber- and f lake-filled thermoplastics are usually anisotropicproducts with variable aspect ratio distribution and orientation varying across thethickness of a molded part. The situation becomes more complex if one considersanisotropy, not only in the macroscopic composite, but also in the matrix (as a result


36

of molecular orientation) and in the filler itself (e.g., graphite and aramid fibers andmica f lakes have directional properties). Thus, thermoplastic composites are not al-ways amenable to rigorous analytical treatments, in contrast to continuous thermosetcomposites, which usually have controlled macrostructures and reinforcement ori-entation [8,17].

Morphological features resulting from orientation of directional fibers and f lakesin complex f low fields are directly related to the part properties. Attempts have beenmade to model and predict orientation distributions in fiber and f lake composites[10,20]. In injection molding, mold-filling patterns and filler orientation depend,among other factors, on mold geometry and cavity thickness, type and position ofgate, injection speed, and rheology of matrix material [8,10]. Typically, three distinctregions with different fiber orientations can be identified: a) a skin originating fromthe expanding melt front, b) an intermediate layer, in which the fibers are orientedparallel to the f low direction, and c) a core layer, with fiber orientation transverse tothe f low direction. An example of complex fiber orientation in a glass-reinforcedpolypropylene, parallel to the f low direction, is shown in Figure 2-7 [8]. For f lakessuch as mica or talc, f lake orientation during extrusion, injection molding, or blowmolding is also predominantly parallel to the f low direction, with a region of mis-alignment in the core [2,20]. Such morphologies can be modified through the appli-cation of a shear force to the melt as it cools (e.g., SCORIM™), which has a markedeffect on the physical properties [24].

Improper location of the gate or multiple gates may produce a zone in which thetwo melt fronts meet (weldline). For unfilled polymers, this region usually has infe-rior properties compared to other locations within the part; mechanical weakness be-


Fig. 2-7 Section parallel to the f low directionthrough a glass fiber reinforced polypropyleneinjection molding showing longitudinal orien-tation near the mold surface and transverse

orientation in the core region (reproducedwith permission of Oxford University Pressfrom ref. [8]).

37

comes more pronounced in the presence of high aspect ratio fibers and f lakes, whichmay not interpenetrate and instead lie in their most unfavorable orientation for ef-fective reinforcement. A more moderate detrimental effect on properties has beenobserved with low aspect ratio fillers. In general, fibrous fillers cause up to 50% lossin tensile yield strength near weld lines, plate-like fillers up to 30%, and cubic fillersup to 15% [16].

For semicrystalline polymers, fillers may affect crystallinity, size of crystallites, anddirection of crystal growth. The filler surface may provide a large number of nucle-ation sites, although this also depends on surface functional groups and surface treat-ments. In certain polymers, fillers may promote transcrystallinity, which can im-prove adhesion and other properties [10].

Symbols Subscripts

A Nielsen’s equation parameter c compositeB Nielsen’s equation parameter cr criticald fiber or platelet diameter d debondedE tensile modulus f fillerG shear modulus L longitudinalK modulus efficiency parameter m matrixK’ orientation parameter T transverseKe Einstein coefficient u ultimateL lengthu reinforcement parameterV volume fractionWa work of adhesionα aspect ratioγ surface/interfacial tensionΓ fracture surface energyη Halpin–Tsai parameterθ angle (reinforcement orientation)θ contact angleµ viscosityν Poisson’s ratioξ Halpin–Tsai parameterσ stressτ interface or matrix shear strengthφmax maximum packing factorΦ interaction parameterχ packing parameterΨ Nielsen’s equation parameters


38

References

1 Sheldon, R. P., Composite Polymeric Materi-als, Chapters 1, 3, 4, and 5, Applied SciencePublishers Ltd., Barking, Essex, England,1982.

2 Xanthos, M., Chapter 14 in Mixing andCompounding of Polymers – Theory and Prac-tice (Eds.: Manas-Zloczower, I., Tadmor, Z.),Carl Hanser Verlag, Munich, New York,1994.

3 Wu, S., Polymer Interface and Adhesion, Mar-cel Dekker, Inc., New York, 1982.

4 Yosomiya, R., et al., Adhesion and Bondingin Composites, Chapters 1, 2, and 3, MarcelDekker, Inc., New York, 1990.

5 Xanthos, M., Polym. Eng. Sci. 1988, 28,1392.

6 Pukanszky, B., Fekete, E., “Adhesion andSurface Modification”, in Advances in Poly-mer Science, Vol. 139, Springer-Verlag,Berlin, Heidelberg, 1999, 121–153.

7 Solomon, D. H., Hawthorne, D. G., Chem-istry of Pigments and Fillers, p. 6, John Wiley& Sons, New York, 1983.

8 McCrum, N. G., Buckley, C. P., Bucknall,C. B., Principles of Polymer Engineering, 2ndEd., Chapters 6 and 8, Oxford UniversityPress, New York, 1997.


10 Clegg, D. W., Collyer, A. A. (Eds.), Mechani-cal Properties of Thermoplastics, Chapters 1,2, 3, 5, 6, and 9, Elsevier Applied SciencePublishers, Barking, Essex, England, 1986.

11 Callister, W., Jr., Materials Science and Engi-neering – An Introduction, 6th Edition,Chapter 1 and App. B, John Wiley & Sons,Inc., New York, 2003.

12 Milewski, J. V., Katz, H. S. (Eds.), Handbookof Reinforcement for Plastics, Chapters 4 and5, Van Nostrand Reinhold, New York, 1987.

13 Drzal, L. T., Proc. Functional Fillers for Plas-tics 2003 Conference, Intertech Corp., At-lanta, GA, Oct. 2003, paper 7.

14 Katz, H. S., Milewski, J. V. (Eds.), Handbookof Fillers and Reinforcements for Plastics,Chapters 2, 3, and 20, Van Nostrand Rein-hold, New York, 1978.

15 Kelly, A., Strong Solids, Chapter V and App.A, Clarendon Press, Oxford, England, 1966.

16 Hohenberger, W., Chapter 17 in Plastics Ad-ditives Handbook (Ed.: Zweifel, H.), HanserPublishers, Munich, 2001.

17 Nielsen, L. E., Landel, R. F., MechanicalProperties of Polymers and Composites, Chap-ters 7 & 8, 2nd Ed., Marcel Dekker Inc.,New York, 1994.

18 Mascia, L., Thermoplastics: Materials Engi-neering, Chapter 4, Applied Science Pub-lishers Ltd., Barking, Essex, England, 1982.

19 Piggott, M. R., Load-Bearing Fibre Compos-ites, Chapters 4, 5, 6, & 8, Pergamon Press,Oxford, England, 1980.

20 Xanthos, M., et al., Internat. Polym. Process.1998, 13(1), 58.

21 Osswald, T. A., Menges, G., Materials Sci-ence of Polymers for Engineers, Chapter 8,Hanser Publishers, Munich, Germany,1995.

22 Han, C. D., Multiphase Flow in Polymer Pro-cessing, Chapter 3, Academic Press, NewYork, 1981.

23 Fisa, B., “Injection Molding of Thermoplas-tic Composites” in Composite Materials Tech-nology (Eds.: Mallick, P. K., Newmann, S.),Hanser Publishers, Munich, 1990, 265–320.

24 Hornsby, P. R., “Rheology, Compoundingand Processing of Filled Thermoplastics” inAdvances in Polymer Science, Vol. 139,Springer-Verlag, Berlin, Heidelberg, 1999,155–217.


3Mixing of Fillers with Plastics

David B. Todd

3.1Introduction

Deriving the maximum benefit from the incorporation of fillers into polymers de-pends upon achieving a uniform distribution of well wetted-out individual particlesand/or fibers.

Filler incorporation may be performed in either batch or continuous equipment.For small production requirements or for very viscous products, such as some filledelastomers, double-armed sigma blade mixers (Figure 3-1), Banbury mixers (Fig-


Fig. 3-1 Double-arm sigma blade mixer [17].

40

ure 3-2), or two-roll mills may still be used. Batch mixers allow control of the se-quence of addition of ingredients to obtain the final desired product without the needfor multiple feeders. Disadvantages of batch mixers may be difficulty in emptying,and extra processing being required for the end product shaping.

Bulk molding compounds, wherein chopped fiberglass and/or sisal is incorporat-ed into a plastic (usually thermosetting) matrix along with other fillers, are frequent-ly prepared in double-arm mixers (Figure 3-1). The 180° single curve blade (Fig-ure 3-3B), with relatively large blade-to-housing clearance, was developed with theaim of achieving less fiber degradation and more complete discharge without the ten-dency of the material to wrap around the center wing sections of the sigma blades(Figure 3-3A).

For most applications in the area of filler incorporation into plastics, most proces-sors now use extruders, either single-screw or twin-screw, to achieve the desiredcompounds.

3 Mixing of Fillers with Plastics

Fig. 3-2 Banbury mixer [17].

41

The primary tasks to be accomplished in extrusion compounding generally includethe following steps:

possible pretreatment of the ingredients metering and feeding of the ingredients melting of solid-fed polymers break-up of agglomerates providing uniform distribution of the filler venting developing pressure for discharge.

Further aims are:

avoiding excessive screw and barrel wear minimizing energy consumption.

There may be significant differences between single- and twin-screw extruders inhow they achieve the above functions.

3.1 Introduction

Fig. 3-3 Mixing blades for double-armbatch mixer [17]: A: sigma, B: 180º singlecurve, C: multiwing overlap, D: doubleNaben.

42

3.2Pretreatment of Fillers

Frequently, either the polymer or the filler(s) may contain too much moisture suchthat pre-drying is required, otherwise unremoved moisture may:

interfere with bonding between the filler and polymer contribute to degradation of the polymer lead to undesired bubble formation in the product.

Drying may be accomplished by the passage of low humidity hot air through batchmixers such as ribbon blenders, f low mixers, or high intensity mixers. Polymers suchas nylons and polyesters require drying under vacuum in order to avoid degradationby hydrolysis. Grulke [1] has outlined some of the process engineering considera-tions involved in the slow diffusion process of removing moisture from solids.

3.3Feeding

Generally, single-screw extruders (SSEs), such as that shown in Figure 3-4, are f loodfed; that is, the rate of feed addition is controlled by the extruder screw speed. Dry pre-blending of multiple solid ingredients can be performed in simple ribbon or ploughmixers. With low levels of addition of some non-abrasive fillers, the polymer andfiller can be dry blended and charged to the feed hopper. Feeding large quantities ofpowdered fillers through a downstream feedport in an SSE is difficult, and will like-ly require a crammer feeder and provision for venting the air accompanying low bulkdensity fillers.

The primary advantages of using twin-screw extruders (TSEs) for filler incorpora-tion arise from:

greater volumetric feeding capacity (especially important with low bulk densitypowders)

the possibility of eliminating pre-blending operations f lexibility in porting for downstream feeding and venting independence of feed rate and screw speed greater f lexibility in controlling the mixing action.

Twin-screw extruders (TSEs) are classified as tangential or intermeshing, andcounter-rotating or co-rotating (Figure 3-5). The most prevalent is the co-rotating in-termeshing type (Figure 3-6), which is available from dozens of manufacturers. Mostsuppliers have learned how to solve the problems associated with the incorporationof high loadings of fillers, and some offer special features specifically addressingthese problems.


433.3 Feeding

Fig. 3-4 Typical single-screw extruder [17].

Fig. 3-5 Classification of twin-screw extruders [17].

44

Although the number of screw starts can vary from single to triple, the most com-mon is double, as shown in Figure 3-7. The f light tips of one screw wipe the chan-nels of the opposite screw. A deep intermesh implies a good conveying capacity, buttoo deep an intermesh means a smaller root diameter, and therefore less torque ca-


Fig. 3-6 Co-rotating intermeshing twin-screw extruder [17,19].

Fig. 3-7 Single-, double-, and triple-startco-rotating intermeshing screws [17,19].

45

pability, which can also limit capacity. The optimum operating conditions occur whenall available extruder power is utilized just as the feed intake has reached its limit.

The increasingly faster screw speeds (up to 1200 rpm) and shorter residence times(<20 seconds) now being offered by TSE manufacturers impose a greater precisionrequirement for the feeders. Generally, loss-in-weight feeders are required for the de-sired product uniformity [2], as back-mixing in extruders is so minimal that feed ir-regularities are not dampened out during the brief passage through the extruder.

Metering and feeding of fibers is particularly difficult because of their tendency toclump together. Some combination of a weight belt feeder in turn feeding a side-en-tering crammer feeder is probably the best solution for materials such as fiberglass[3]. With such a feeding system, a large volume of air will also be pumped into the ex-truder, and it is desirable to provide for upstream venting of the starved screw sectioninto which the filler is being fed. Typical barrel sections combining side entry plusventing are shown in Figure 3-8.

Melting of a polymer in the presence of abrasive fillers is undesirable for two rea-sons. The large forces encountered during melting can result in agglomeration (bri-quetting) of powder fillers, and will also cause much more abrasive wear of screwsand barrels. The mechanism for the formation of agglomerates in single-screw ex-truders has been described by Gale [4], and is shown in Figure 3-9.

The best way of overcoming the problems of agglomerates is simply avoiding theirformation, generally by feeding the filler downstream after the polymer is fully melt-ed.

The agglomeration effect accompanying polymer melting in the presence of solidfillers has clearly been shown by Rogers et al. [5], who reported that joint feeding of15 wt. % CaCO3 and 85 wt. % polystyrene pellets led to significant agglomerate for-mation (d > 20 µm) of initially 0.7 µm diameter (d) CaCO3 powder.

3.3 Feeding

Vent

AdditivesPolymer

FlowFig. 3-8 Barrel section for side entryfeeding plus venting [19].

46

The potential for concomitant agglomerate formation with pellet melting is partic-ularly high in kneading block sections of intermeshing TSEs because of the highpressures that can be developed in the compression/expansion cycles of co-rotatingTSEs or during the calendering impact of counter-rotating TSEs. Simultaneous pel-let melting and filler addition can cause excessive screw and barrel wear.

3.4Melting

The generally accepted melting model for SSEs is that described by Tadmor and Go-gos [6], wherein the passive solid bed moves axially downstream being gradually com-pressed in the transition sections such that a melt film forms at the barrel wall by acombination of direct heat transfer through this wall and friction of the unmeltedsolids against it.


Melt pool started

Melt pool growing

Melting completed

Loose granules and filler

Granules deform, fillercompressed into agglomerates

Polymer melted by conductioncontains agglomerates

Filler coats polymer granules

Deformedpolymer granules

Agglomerate formed

Segregated compacted filler on metal surface

Polymer plasticating without filler Polymer plasticating with filler

Fig. 3-9 Mechanism of agglomerate formation [4].

47

Initial melting in TSEs is usually dominated by plastic energy dissipation [7], whichis the unrecovered energy of dynamic compression of the solid pellets in the knead-ing block section of the extruder.

After the initial melting produces a slurry, which predominantly consists of someunmelted solid polymer pellets f loating in a “sea” of the continuous melt phase, com-pletion of the melting process occurs by heat transfer at the melt/polymer interface,with the energy supplied mainly by viscous energy dissipation and to a lesser extentby barrel heat transfer.

In SSEs, it is often prudent to utilize a Maddock (Union Carbide) mixing element(Figure 3-10). This acts as a crude filter, preventing passage of the unmelted f loatersand retaining them until melting reduces their size to less than that of the clearancegaps.

In TSEs, the same filtering effect can be achieved with intermeshing blister rings.

3.5Introduction of Solids and Mixing

After the screw configuration, in which the polymer is melted, a starved section isprovided into which the filler solids can be introduced. Preferably, as previously men-tioned in Section 3.3, provision should also be made for venting the air accompany-ing the powdered solids (as, for example, in Figure 3-8). If oxygen is likely to causedegradation of the base resin, it will be necessary to purge the solids feeder with ni-trogen. The lubrication supplied by the molten polymer and the low pressure re-quirement of the initial incorporation will minimize screw and barrel wear.

The mixing requirement in filler incorporation is usually a combination of disper-sive and distributive actions, as illustrated in Figure 3-11. Dispersive mixing is thebreaking up of agglomerates or the unraveling of fiber bundles. Distributive mixingis achieving equal spatial distribution of the filler throughout the plastic matrix.

Screws by themselves do not provide much mixing. In an SSE, the requirement fordeveloping pressure to form the extrudate leads to some cross-channel mixing.

Shearing forces are required to break-up agglomerates. The level of these shearingforces will depend upon the nature of the filler. For example, intense shearing is usu-ally required when dispersing carbon black. Even more severe shearing is needed todisperse montmorillonite clay into nano-sized platelets.

3.5 Introduction of Solids and Mixing

Fig. 3-10 Maddock (Union Carbide) mixing element.

48

The incorporation of fiberglass is a special case. The glass fibers are supplied inchopped bundles that have been treated with sizings and suitable coupling agents toincrease the adhesion between polymer and glass. These bundles need to be “un-wrapped” and wetted out, ideally with a minimum of fiber breakage. Dispersion ofglass fibers in TSEs is further discussed in Chapter 7.

It is always wise to utilize some sort of mixing enhancer to improve distributivemixing, rather than rely on the mixing normally achieved with screws alone. Figure3-12 illustrates some of the SSE mixing enhancers that have been employed. In gen-eral, the enhancers all improve the distribution of filler through a series of divisionsand recombinations. Simple shear mixing should be interrupted by a combination ofcutting and turning actions. Gale [8] illustrated the homogeneity achieved with ninedifferent mixers, along with the power consumption for each mixing configuration.Gale showed that the Cavity Transfer Mixer was very effective in achieving homo-geneity with relatively low pressure drop and low temperature rise.

As in the case of SSEs, TSE screws themselves do not provide much mixing. Thebulk of the melting and mixing arises from the interplay of kneading paddles, whichare the offset straight elements with the same cross-section as the screws. Kneadingpaddle arrays have some of the same conveying characteristics as screws (Figure3-13), but also provide unique mixing actions not available in single-screw extruders.The dispersion face (Figure 3-14) can provide the high shearing action desired for ag-glomerate break-up. Excellent distribution is achieved by the expansion/compressionaction, as shown in Figure 3-15. Each quarter-turn produces an expansion and com-pression of the process parts of the cross-section, and these squeezing actions causeextensional f low and redistribution of the polymer-filler mixture as it passes downthe extruder. In addition to kneading paddles, toothed gear elements (Figure 3-16)can be effectively used to provide enhanced distributive mixing action.


Fig. 3-11 Dispersive and distributive mixingaspects [17,19].


Fig. 3-12 Mixing enhancers for single-screw extruders [19]:A: pineapple, B: Dulmage, C: Saxton, D: pin, E: cavity transfer.

50 3 Mixing of Fillers with Plastics

S

K-1

K-2

Fig. 3-13 Comparison of kneadingelements and screws [17].

Fig. 3-14 Dispersion face of a kneading paddle [17,19].


Fig. 3-15 Expansion/compression squeezing action ofco-rotating intermeshing bilobal paddles [17,19].

Fig. 3-16 Toothed gear mixing elements [17,19].

52

3.6Venting

Removal of air accompanying low bulk density fillers is essential, as noted in Section3.3. Off-gassing of processing aids associated with some fillers, such as coatings onfiberglass, will also require venting. If a vacuum is required for adequate devolatiliz-ing, the extruder screw(s) must have a melt seal upstream of the vent port zone to pre-vent sucking air or unincorporated powder out of the vent port. The melt seal can beachieved with short sections of reverse screws, reverse kneading blocks, or blisterrings. The screws in the vent port region need to be designed to operate starved,preferably less than half-full, as the mixture may be foaming as it passes the melt seal.

The degree of fill (f) in the screws is approximately equal to the ratio of the net f low(Q) to the drag f low (Qd). The drag f low per revolution is equal to one-half of the vol-ume contained in the open cross-section (a) over the lead length (z):

Qd/N = a · z/2 (3-1)f = Q/Qd (3-2)

For fully intermeshing co-rotating bilobe TSEs, the open cross-section (a) can becalculated from the screw diameter (D) and the channel depth (h) [9]:

a = 3.08 h D (3-3)

3.7Pressure Generation

The pressure required for shaping the product (sheet profile or pellets) depends up-on the f low rate, the aperture geometry, and the viscosity of the filled polymer at theexit shear rate. In general, the viscosity of a polymer-filler mixture (ηc) increases asthe volume fraction concentration (Vf) of the filler is increased. Because of the irreg-ular shapes of most fillers, the viscosity of the mixture will increase more than in thecase of uniformly sized spheres, as predicted by Einstein [10]:

ηc/ηm = 1 + 2.5Vf (3-4)

where ηm is the matrix viscosity.In the case of some platelet-type fillers, such as talc and mica, the platelets can align

with the streamlines at higher shear rates, and the viscosity increase is less than thatpredicted by the above equation (see Chapter 2).


53

3.8Process Examples

Valsamis and Canedo [11] have reported on how the use of different mixing elementsin a Farrel FTX 80 twin-screw extruder affected glass breakage and compound prop-erties in a nylon/fiberglass mixture.

Wall [12] studied a tandem arrangement of melting nylon-6,6 and subsequent ad-dition of 4.5 mm long glass fibers in a short TSE, followed by pressure developmentfor pelletizing in a more slowly rotating SSE. In the ensuing injection-moldingprocess, remelting of the fiber-filled pellets reduced the fiber length from about 500to 300 µm. Grillo et al. [13] studied the effect of TSE mixing elements and die designon glass fiber length retention and ultimate physical properties with glass-filledstyrene/maleic anhydride copolymers.

Using equipment such as that shown in Figure 3-17, Mack [14] has investigated theincorporation of high talc loadings into polypropylene. Although the true talc parti-cle density was around 2.7 g cm–3, the bulk density was only about 0.2 g cm–3, andmay well have been as low as 0.05 g cm–3 when aerated. Thus, a talc stream beingforced into such an extruder may consist of 95% air.

The effect of talc bulk density and talc loading on extruder capacity for a BerstorffZE90-A TSE is shown in Figure 3-18. Increasing the talc ratio made the mixingprocess more difficult and decreased the output rate. Mack also showed that the meltviscosity of the polymer could affect the limits of filler incorporation. For example,the level of incorporation of calcium carbonate (0.34 g cm–3 bulk density) was shownto be higher with less viscous polypropylene (Figure 3-19) because of easier wetting-out of the filler.

3.8 Process Examples

Loss-in-weightfeeder

Agitator

ZSFE

Atmosphericvent

Twin-screwextruder

Fig. 3-17 Horizontal side-feed extruder [14,19].

54

Deep-f lighted TSEs may reach torque-limited capacity before the feeding rate lim-itation. Because the heat capacities of most fillers are only about half that of typicalpolymers, the capacity of such an extruder may actually increase as the filler loadingis increased.

Andersen [15] has provided some guidelines for preparing nano-composite poly-mer compounds, such as clay/polypropylene/maleated polypropylene, in TSEs. Claydelamination was found to be improved by adding all the ingredients in the feed port,as the stresses were greatest as the polypropylene was melting, and there was no ten-dency to form conglomerates as occurs with CaCO3.


Fig. 3-18 Effect of talc bulk density and loading on extruder capacity[14,19].

Fig. 3-19 Effect of polymer viscosity on extruder capacity [14,19].

55

Kapfer and Schneider [16] studied the formation of highly filled compounds inboth a TSE and in a screw kneader, wherein the interrupted screw oscillates as wellas rotates, and the screw channels are wiped by rows of teeth protruding inward fromthe barrel wall. They concluded that the TSE would be preferred when the polymer isnot susceptible to shear degradation, while the reciprocating single screw offers somebenefits when fillers or fibers are highly susceptible to breakage. Deep channels arepreferred with both types of equipment to facilitate easier intake of larger volumes offiller.

3.9Further Information

Virtually all aspects of industrial mixing are described in the recent North AmericanMixing Forum’s Handbook [17]. The SPE Guide on Extrusion Technology and Trou-bleshooting [18] contains chapters particularly relevant to the processing of polymer-ic systems. Compounding operations as perceived from the equipment manufactur-er’s viewpoint are detailed in “Plastics Compounding – Equipment and Processing”[9].

References

1 Grulke, E. A., Polymer Process Engineering,Prentice Hall, Englewood Cliffs, NJ, 1994.

2 Welsch, R., Plast. Additives & Compounding,Sept./Oct. 2003, 40–45.

3 Häuptli, A., Plast. Additives & Compounding,Sept./Oct. 2003, 36–39.

4 Gale, M., Adv. Polym. Technol. 1997, 16,251–262.

5 Rogers, M. J., et al., Proc. 59th SPE ANTEC,2001, 47, 129–133.

6 Tadmor, Z., Gogos, C. G., Principles of Poly-mer Processing, John Wiley & Sons, NewYork, 1979.

7 Qian, B., et al. Adv. Polym. Technol. 2003, 22,69–74.

8 Gale, G. M., Proc. 49th SPE ANTEC, 1991,37, 95–98.

9 Todd, D. B. (Ed.), Plastics Compounding –Equipment and Processing, Hanser Publish-ers, Munich, 1998.

10 Einstein, A., Ann. Phys. 1906, 19, 289 and1911, 34, 591.

11 Valsamis, L. N., Canedo, E. L., Plastics Eng.1997, 53, 37–39.

12 Wall, D., Proc. 45th SPE ANTEC, 1987, 33,778–781.

13 Grillo, J., et al. Proc. 49th SPE ANTEC,1991, 37, 122–127.

14 Mack, M., Plastics Eng. 1997, 53, 33–35.15 Andersen, P. G., “Twin-Screw Guidelines for

Compounding Nanocomposites”, Proc. 60thSPE ANTEC, 2002, 48.

16 Kapfer, K., Schneider, W., Proc. 60th SPEANTEC, 2003, 49, 246–250.

17 Paul, E. L., Antiemo-Obeng, V. A., Kresta, S.M. (Eds.), Handbook of Industrial Mixing –Science and Practice, John Wiley & Sons,Hoboken, NJ, 2004.

18 Vlachopoulos, J., Wagner, J. R. (Eds.), SPEGuide on Extrusion Technology and Trou-bleshooting, Society of Plastics Engineers,Brookfield, CT, 2001.

19 Todd, D. B., Adv. Polym. Technol. 2000, 19,54.

3.9 Further Information

Part IISurface Modif iers and Coupling Agents


4Silane Coupling Agents

Kerstin Weissenbach and Helmut Mack

4.1Introduction

After the paper and coatings markets, the plastics industry is the third largest outletfor white minerals in Europe, as well as in North America. These minerals include anumber of natural products such as talc, kaolin, wollastonite, and ground calciumcarbonate, as well as synthetic products such as precipitated calcium carbonate, alu-minum trihydrate (ATH), magnesium dihydroxide (MDH), synthetic silicas, and sil-icates. More than 37% of all thermoplastics, thermosets, and elastomers worldwideare compounded and reinforced with fillers and fibers. A major need for silanes ascoupling agents arose in the 1940s when glass fibers were first used as reinforce-ments in unsaturated polyester (UP) resins. The first commercial silanes appeared inthe mid-1950s. Since then, they have become the most common and widely used cou-pling agents. Silanes offer tremendous advantages for virtually all market segmentsinvolving polymer/filler interactions and, as a result, are widely used to modify thesurfaces of mineral fillers [1–5].

Surface modification of fillers with silanes may generate the following perform-ance benefits:

improved dimensional stability modified surface characteristics (water repellency or hydrophobicity) improved wet-out between resin and filler decreased water-vapor transmission controlled rheological properties (higher loadings with no viscosity increase) improved filler dispersion (no filler agglomerates) improved mechanical properties and high retention under adverse conditions improved electrical properties.

The world market for silanes used in filler treatment is approximately divided asfollows: Europe 25%, Asia 8%, Americas 66%, rest of the world 1%.


60

Common silanes have the general formulae Y-(CH2)3Si(X)3 (Figure 4-1) andY-(CH2)2Si(CH3)(X)2. The silicon functional group X is a hydrolyzable group chosento react with surface hydroxyl groups of the filler to produce a stable bond, and is usu-ally halogen or alkoxy. The silane coupling agents in commercial use are generallyalkoxy-based and bear one organic group attached to the silicon center, the generalformula being Y-(CH2)3Si(OR)3. The organofunctional group Y is tightly bound to thesilicon via a short carbon chain and links with the polymer. This group has to ensuremaximum compatibility with the resin system. Bonding to the polymer takes place bychemical reactions or physicochemical interactions such as hydrogen bonding, acid-base interaction, interpenetration of the polymer network (entanglement), or electro-static attraction. The group Y may be non-functional or functional (reactive); exam-ples of the latter are vinyl, amino, methacryl, epoxy, mercapto, etc. Most silanes arecolorless or slightly yellowish, low viscosity liquids. Both the non-functional andfunctional organosilanes discussed below are employed in important commercialfiller treatments.

4.2Production and Structures of Monomeric Silanes

Chlorosilanes, as raw materials for organofunctional silanes, are produced technical-ly by reaction of HCl with silicon metal in a fixed-bed reactor or f luidized bed. Prod-ucts derived from this process are tetrachlorosilane, SiCl4, and trichlorosilane,HSiCl3. Tetrachlorosilane is used to manufacture optical fibers, silicic acid esters, andfumed silica. Trichlorosilane is, in addition to other uses, the starting material fororganofunctional and alkyl silanes. Commercially available organofunctional silanesare produced by hydrosilylation reactions of alkenes with hydrogen-containingsilanes. Subsequent esterification affords the standard commercially availableorganofunctional silanes (Scheme 4-1).

4 Silane Coupling Agents

Fig. 4-1 General structure of organosilanes.

Scheme 4-1 Production of organofunctional silanes.

61

4.3 Silane Chemistry

Silanes are adhesion promoters that unite the different phases present in a compos-ite material. These phases are typically organic resins (e.g. ethylene/vinyl acetatecopolymer, EVA) and inorganic fillers (e.g. ATH) or fibrous reinforcements. Silanesform “molecular bridges” to create strong, stable, water- and chemical-resistantbonds between two otherwise weakly bonded surfaces. The properties and effects ofsilanes are determined by their molecular structures. The silicon at the center isbound to the organofunctional group, Y, and the silicon functional alkoxy groups, OR.The organofunctional group, Y, is bound to the polymer by chemical reactions (graft-ing, addition, or substitution) and/or physicochemical interactions (hydrogen bond-ing, acid-base interaction, entanglement, or electrostatic attraction) [6–10]. The sili-con functional groups, OR, usually alkoxy groups, can be hydrolyzed at the first stageof application, liberating the corresponding alcohol. Continuous reaction with wateror moisture results in elimination of all OR groups as alcohol and their replacementby hydroxyl moieties to give silanols (see Scheme 4-2).

In the following text, the term “silanol” is used as a generalized description of themonosilanols, silanediols, and silanetriols formed in the case of trialkoxysilanes.Usually, the hydrolysis of the first OR group is the rate-controlling reaction step. Thefiller surface can react with the silanol intermediates and in subsequent steps theseSi–OH groups react with active OH groups of the inorganic substrate, building upstable covalent (Si–O–substrate) bonds [11].

4.3 Silane Chemistry

Scheme 4-2 Hydrolysis of trialkoxysilanes.

62

4.4Types of Silanes

4.4.1Waterborne Silane Systems

Organofunctional silanes can be applied to fillers in aqueous or aqueous/alcohol so-lutions as well as neat. If the silane is added to an aqueous filler slurry, then it has tobe water-soluble. Some silanes will be rendered water-soluble after hydrolysis and theformation of silanols. These Si–OH functions are very reactive in establishing a co-valent bond between the filler and the silane. In addition to fast hydrolysis, the avail-ability of reactive hydroxyl units on the central silicon atom significantly inf luencesthe reactivity of the silane. In this context, it has long been postulated that monomer-ic silanetriols are exclusively responsible for the activity of the silane and that thesemonomeric units are stable for a period of a few hours to a few days in aqueous so-lutions. However, it has been proven by 29Si NMR spectroscopy that oligomericsilanes are active as well. Finally, the activity of the aqueous silane solution decreasesas a result of cross-linking to give insoluble, polymeric siloxanes (gel structures); thisis evident from a substantial decrease in adhesion. Knowledge of the silanol concen-tration and the degree of oligomerization is, therefore, important when using aque-ous solutions [12].

In terms of devising a stable waterborne silane system, the silanol form of a silaneis desirable since silanols have greater solubility and reactivity than their alkoxysilaneprecursors. A series of waterborne silanes with different functionalities is commer-cially available, and these have high concentrations of active silanol groups and arestable in water for periods of up to a year. These silanes are considered to be free ofvolatile organic compounds (VOC). They are particularly useful in wet processes suchas grinding and milling, where VOCs are undesirable, and the mineral can be treat-ed in situ in the aqueous process slurry.

4.4.2Oligomeric Silanes

Silane hydrolysis and condensation take place on the surface of the mineral filler,thus forming oligomeric silane structures. Oligomeric silanes (e.g. Figure 4-2) arecommercially available. They are low viscosity liquids with high boiling and f lashpoints, releasing a significantly reduced amount of alcohol [13]. Easy handling and


Fig. 4-2 Chemical structure of a vinyl oligomeric silane.

63

safe processing in standard extruders, kneaders, Banbury mixers or co-kneaders isassured as a result of the reduced release of VOCs. Besides low VOC evolution dur-ing application, oligomers may provide additional benefits, such as better wettabilityon the filler and the formation of a more homogeneous, defect-free silane layer onthe filler surface. This can result in smaller amounts of silane being needed toachieve the same final properties. Different types of oligomers are available, rangingfrom homo-oligomers to various types of co-oligomers, the latter combining the ben-efits and properties of both silane monomers.

4.5Silane Hydrolysis

The action of alkoxysilanes starts with hydrolysis. The rate of hydrolysis depends onthe pH as well as on the type of organo- and silicon-functional groups. The siliconfunctional group has a significant inf luence on the hydrolysis rate. The order of re-activity is as follows: propoxy << ethoxy < methoxy. Typically, a large excess of wateris used as reactant; under these conditions, the hydrolysis of alkoxysilanes is found tobe a (pseudo) first-order reaction.

The next important parameter inf luencing the reaction rate is the pH of the silanehydrolysis medium. At high and very low pH values, the rate of hydrolysis is higherthan that at neutral pH, at which silanes are most stable. For example, the rate of re-action of a monomeric trialkoxysilane in acetic acid solution increases by a factor of10 on reducing the pH from 4 to 3. This effect is even more marked on going fromneutral to acidic conditions (pH 3), for which the factor is about 25 to 50 dependingon the method of mixing. As an example, the hydrolysis of 3-methacryloxypropyl-trimethoxysilane proceeds within a few minutes in a low pH environment as a resultof the catalytic action of H+ ions.

The strong dependence of reaction rate on pH is understandable if one takes intoaccount the nucleophilic substitution reaction mechanism for the hydrolysis ofalkoxysilanes in acid media, since H+ ions directly affect the rate-determining step ofthe reaction. The nature of the acid present also affects the hydrolysis behavior. Achange from inorganic (e.g. HCl, H2SO4 or H3PO4) to organic acids (e.g. acetic acid,formic acid, citric acid) has a significant effect on the hydrolysis rate. Experimentalresults on the hydrolysis of trialkoxysilanes, in particular nonpolar, long-chain orbranched alkylsilanes, indicate a type of “micelle formation”, whereby the silane mol-ecules are surrounded by a phase boundary layer akin to the structures found in sur-factant chemistry. The result is that the overall system appears to be homogeneous.In the above model, attack of water molecules at the silicon atom is rendered moredifficult by the formation of a phase boundary layer. Accordingly, an inorganic, fullydissociated acid finds it more difficult to develop its catalytic potential in this “non-polar capsule”, whereas the dual character of acetic acid encourages the transfer ofwater molecules to the reaction center. By comparing hydrolysis rate constants, it canclearly be seen that variation of the structure of the substituents in the alkyltri-alkoxysilane is an important factor [14].

4.5 Silane Hydrolysis

64

The effect of pH on the stability of the formed silanols is different from that on thestability of alkoxysilanes. Silanols are most stable at around pH 3, and their reactivi-ty is higher at a pH lower than 1.5 or higher than 4.5 (Figure 4-3). Silanols condenseto form oligomers and, ultimately, two- and three-dimensional networks.

When considering silane hydrolysis and condensation, different reactivities in dif-ferent pH ranges can be expected. At very low pH, silanes hydrolyze very quickly. Theformed silanols are relatively stable and, over time, form coordinated networks. Atneutral pH, silanes hydrolyze very slowly to silanols, which are unstable and con-dense. Thus, in both cases, there is still a slow reaction in the transition from silanesto Si–O–Si networks. At pH > 8, silanes become highly reactive once more and formsilanols very quickly. These silanols are very unstable and condense very quickly togive uncoordinated Si–O–Si networks. The build-up of Si–O–Si networks cannot becontrolled and the uniform coating of the filler surface becomes more difficult re-sulting in thicker, uncoordinated layers.

The rate of hydrolysis is also inf luenced by the nature of the organic substituent onthe trialkoxysilane. As the polarity of the organic substituent is diminished by in-creasing the length of the non-polar chain, as for example in long-chain alkylsilanes,the hydrolysis rate decreases. This behavior can be explained in terms of the lowersolubility of the nonpolar silanes in the aqueous reaction system and the associatedformation of micellar structures. Incorporating polar moieties (functionalities otherthan alkyl) generally increases susceptibility to hydrolysis. However, it is not possibleto determine whether the increased rate is directly linked to the functionality of thesubstituent or is merely a consequence of better solubility. The pH and/or the use ofa catalyst also have a decisive effect on the hydrolysis behavior of trialkoxysilanes.Without exception, in all the studied functional trialkoxysilanes, complete hydrolysisof the alkoxy substituents to the corresponding silanols takes place within a periodranging from a few minutes to a few hours, depending on the nature of the functionalgroup. Reactivity towards hydrolysis increases with substituent in the following or-der: alkyl < vinyl ≈ methacryloxy < mercapto < epoxy < amino. As the reaction pro-ceeds, not only hydrolysis of the alkoxy groups but also opening of the epoxy ring, as


0 2 4 6 8 100

Reactivity

Silanols Silane

pH

Fig. 4-3 Reactivity of silanes and silanols.

65

in the case of 3-glycidyloxypropyltrimethoxysilane, takes place; this further increasesthe rate of hydrolysis.

4.6Reactivity of Silanes Towards the Filler

Organosilanes rely on the reaction with surface hydroxyl groups to produce a stablecovalent bond and a stable layer on the filler surface [6,7,15]. Thus, they are most ef-fective on fillers with high concentrations of reactive hydroxyls and a sufficientamount of residual surface water. Silica, silicates (including glass), oxides, and hy-droxides are most reactive towards silanes. Silanes are generally not as effective onmaterials such as sulfates and carbonates, although encapsulation with, for example,silica can facilitate stable silane modification even on these surfaces.

As a first step, fixation of the silanol on the filler surface is accomplished throughhydrogen bonding with the surface OH groups. Until the water molecule is elimi-nated and removed from the reaction site, this reaction is thought to be reversible. Aslong as there is only hydrogen bonding, the silane can still migrate on the filler sur-face. The covalent [silane–O–filler] bond eventually fixes the silane on the filler sur-face. In theory, the silane (and in the reactions that follow, the oligomers) form amonolayer on the filler surface (see Scheme 4-3). In reality, the propensity of tri-alkoxysilanes for self-condensation to produce various three-dimensional networksmakes the concept of monolayer coverage based on simple surface reaction of dubi-ous value when considering this type of molecule. The most highly reactive silanolsform oligomers prior to the reaction with the filler surface. This does not normallyaffect the performance of the silane on the surface.

Elucidation of the exact nature of the surface layers and their relationship to thecoating conditions has proven to be difficult. The current understanding is that silanelayers on mineral surfaces are thicker than the postulated theoretical monolayer (seeFigure 4-4). Such layers are very complex and depend on the coating conditions used,the nature of the mineral surface, and the chemistry of the reactive functionalitiespresent. In general, it is known that silane layers, as they are normally formed (and

4.6 Reactivity of Silanes Towards the Filler

Scheme 4-3 Mode of reaction between a silanol and an inorganicsurface.

66

before the incorporation of the treated mineral in a composite), consist of a mixtureof chemisorbed and physisorbed material. The physisorbed silane is readily remov-able by solvent washing, whereas the chemisorbed silane is not extractable.

4.7Combining Silanes and Mineral Fillers

In filler surface pre-treatment procedures, the neat silane (dry Method I) or a silanesolution (slurry Method II) is metered uniformly onto the filler with sufficient mix-ing or tumbling, and this is followed by heat treatment. Other methods include inte-gral silane blending or in situ coating (Method III) and the use of silane concentrates(Method IV).

4.7.1Method I

In the dry procedure, the silane is sprayed onto a well agitated filler. In order to ob-tain maximum efficiency, uniform silane dispersion is essential and this is achievedthrough the high shear rates provided by the mixing equipment. Manufacturers ofprocessing equipment include Loedige, Littleford Day, Hosokawa Alpine, andThyssen Henschel. Most important commercial silane-coating processes are contin-uous and have high throughput rates. Control over the rate of silane addition, thedwell time, and the exact temperature within the system is essential. A certain (ele-vated) temperature is required to promote complete reaction between the silane andthe filler; however, an excessively high temperature may lead to a loss of the silane re-activity. All parameters need adjustment depending on the type of silane employed.

In general, the treated filler is heated after the addition of the silane to remove thereaction by-products, solvents and water, and to completely and permanently bondthe silane to the filler surface. Also important in this step is the control of the by-prod-


Fig. 4-4 Overview of a silanized surface.

67

ucts, such as alcohols. Explosive limits and concentrations of the evolved alcoholshould be considered and special collection systems can be installed to reduce therisk of explosions.

In general, silanes need a certain time to react with the filler surface. Accordingly,dwell times of only 2–3 minutes are common. Since the main covalent reaction iscompleted after about 15–30 minutes, to assure fixation of the silane on the mineralsurface an additional 30 minutes should be allowed at elevated temperatures. Insome cases, a catalyst can help to “activate” a slowly reacting silane or mineral [16].Generally, silane loadings are between 0.7 and 2 wt. % relative to the filler and dependon the filler surface and final application.

4.7.2Method II

The slurry procedure for filler treatment is limited to alkoxysilanes and waterbornesilane systems. Obviously, fillers that react with water cannot be treated according tothis method. The slurry procedure should be considered for commercial treatmentwhen the filler is handled as a slurry during manufacturing. The reaction mediummay be aqueous, a mixture of an alcohol and water, or a variety of polar and non-po-lar solvents. Typically, low concentrations of the silane (up to 5%) are dissolved by hy-drolysis. Alternatively, the silane can be applied as an emulsion. The silane solutionor emulsion can be applied by spraying, dipping, or immersing. Removal of water,solvents, and reaction by-products requires additional steps such as setting, dehydra-tion, and finally drying. Silane loadings are comparable to those achieved by way ofMethod I.

4.7.3Method III

The neat silane can also be added during compounding and would be expected to mi-grate to the filler surface. This in situ treatment procedure provides a means of coat-ing freshly formed filler surfaces, as arise, for example, during silica/rubber com-pounding.

The undiluted silane is added directly to the polymer, either prior to or togetherwith the filler. It is essential that the resin does not react with the silane prematurelyas otherwise the coupling efficiency will be reduced. Typical types of compoundingequipment are internal mixers, kneaders, Banbury mixers, two-roll mills, and ex-truders. The integral blend technique is widely used in resin/filler systems becauseof its great simplicity and possible cost advantages. This is mainly due to the one-stepprocess and the lower raw material costs (untreated mineral plus silane compared topre-treated mineral), despite the fact that more silane is needed to achieve compara-ble performance in the finished composite.

4.7 Combining Silanes and Mineral Fillers

68

4.7.4Method IV

The silane can also be added as a dry concentrate (wax dispersion, dry liquid, or mas-terbatch). Here, the silane is adsorbed at very high levels onto suitable carriers andthen blended with the polymer and filler during compounding. The use of “solid”silanes leads to very effective dispersion even with simple production equipment. Inaddition, an easy and safer handling method is assured. Silane loadings are compa-rable to those achieved by way of the in situ method.

4.8Insights into the Silylated Filler Surfaces

Any silane surface modification needs to prove its value through a practical test.Proper selection of the mineral, silane, and production parameters will lead to opti-mum properties of the composite [17]. Nevertheless, there remains a desire to “see”the silane on the filler surface, to visualize the silane layers, or at least to see the dif-ference between a silane-modified and an unmodified mineral surface [18,19]. Ingeneral, surface analysis depends on the type of organofunctional group on thesilane. To avoid the analysis of physisorbed rather than chemically bound silanes, thetreated mineral can be eluted with an excess of solvent in which the respective silaneis soluble. There exist several analytical methods that can distinguish between chem-ically bonded and physicochemically adsorbed silanes.

4.8.1Spectroscopy

4.8.1.1 FT-IR/Raman SpectroscopySilanes with IR-active organofunctional groups such as methacrylate or vinyl can beeasily detected through their IR absorption bands. The untreated mineral fillershould be used as a control to eliminate the absorption bands inherent to the miner-al. In most cases, this method is semi-quantitative, resulting in the detection of ap-proximate loadings of the silane on the surface [20,21]. Similar results may be ob-tained by means of Raman spectroscopy.

4.8.1.2 MAS-NMR SpectroscopyCharacterization of modified surfaces is also possible by MAS (magic angle spin-ning)-NMR (nuclear magnetic resonance) spectroscopy. Chemically bound andphysicochemically absorbed silane can be distinguished by this method [22]. The 1HNMR spectra of N-(2-aminoethyl)-3-aminopropyltrimethoxysilane (DYNASYLAN®

DAMO) and N-(2-aminoethyl)-3-aminopropyl(methyl)dimethoxysilane (DYNASY-LAN® 1411) are shown in Figures 4-5 and 4-6.


694.8 Insights into the Silylated Filler Surfaces

Fig. 4-5 1H NMR spectrum of N-(2-aminoethyl)-3-aminopropyl-trimethoxysilane (DYNASYLAN® DAMO).

Fig. 4-6 1H NMR spectrum of N-(2-aminoethyl)-3-aminopropyl(methyl)dimethoxysilane (DYNASYLAN® 1411).

70

4.8.1.3 Auger Electron SpectroscopyIn Auger electron spectroscopy (AES), the sample is activated through electrons pro-ducing ionization of atoms of the outer silane layer. During refilling of the resultingvacancy by an outer electron, energy released can be transferred to a third electron,which leaves the solid and can be detected. By this method, two-dimensional silane-coated surfaces can be analyzed [23]. Although it may be assumed that comparableprocesses take place on the surfaces of mineral fillers, AES cannot be used for theiranalysis since mineral fillers are three-dimensional.

Figure 4-7 shows the results of AES surface analysis of a control E-glass sample.The sample was rinsed with ethanol prior to the analysis and, as a result, the ele-mental composition consists only of silicon and oxygen (SiO2) after a few seconds ofsputtering. For AES depth-profiling, the ethanol-rinsed E-glass plate was dipped in a1 wt. % aqueous cationic aminosilane solution for 5 minutes and then dried for 1 hat room temperature.

Figure 4-8 shows two-dimensional visualizations (AES line scans) for a monomer-ic and an oligomeric cationic aminosilane (DYNASYLAN® 1161) on E-glass. The lackof homogeneity of the monomeric cationic aminosilane surface layer is confirmed bya shady SEM (scanning electron microscopy) image (Figure 4-9). In contrast, theoligomeric cationic aminosilane leads to a much more homogeneous surface.

By AES depth-profiling it could be proven that the monomeric cationic aminosi-lane leads to an average silane layer thickness of ca. 10 nm, whereas the oligomericcationic aminosilane forms much thicker silane layers with an average thickness ofca. 200 nm. AES line scans (Figure 4-10) demonstrate the resulting very homoge-neous layer of the oligomeric cationic aminosilane. In contrast, the monomericcationic aminosilane leads to silane pinholes and isolated domains. The data show


Fig. 4-7 AES surface analysis of a control E-glass sample.

71

that oligomeric silanes wet surfaces much more efficiently than monomeric silanesand lead to a more homogeneous silane distribution on the surface.

Other spectroscopic methods are available for characterizing silane-coated sur-faces, including SSIMS (static secondary ion mass spectrometry), AFM (atomic force


Fig. 4-8 AES line scans of monomeric and oligomeric cationicaminosilanes on E-glass.

Secondary electron imagemonomeric cationic aminosilane

50 µm 50 µm

Secondary electron imageoligomeric cationic aminosilane

Fig. 4-9 SEM (scanning electron microscopy) of monomeric andoligomeric aminosilane layers on E-glass.

72

microscopy), ESCA (electron spectroscopy for chemical analysis), and EDX (energy-dispersive X-ray analysis) [24–29].

4.8.2Pyrolysis-Gas Chromatography

Silane loading and the nature of the silane can be quantitatively determined by meansof pyrolysis-gas chromatography (Py-GC). The samples are pyrolyzed and thevolatilized parts of the silane (especially those originating from the organofunctionalgroup) can be subsequently determined by GC [24]. Typical test conditions used to ob-tain the GC trace for the octyltriethoxysilane/TiO2 system shown in Figure 4-11 areas follows: pyrolysis at 700 °C, helium carrier gas, nonpolar column, f lame ionizationdetector [24].

4.8.3Carbon Analysis

Carbon elemental analysis of non-carbon containing minerals such as synthetic TiO2

is a very helpful method to determine the carbon content introduced by a silane sur-face treatment.


18

0

16

14

12

10

8

6

4

2

18

0

16

14

12

10

8

6

4

2

0 100 200 300 400 500 0 100 200 300 400 500

x 104

Monomeric cationic aminosilane Oligomeric cationic aminosilane

Distance [µm]Distance [µm]

Inte

nsity

Inte

nsity

x 104

CC

OO

SiSi

Fig. 4-10 AES line scans for monomeric and oligomeric amino-silane layers on E-glass.

73

4.8.4Colorimetric Tests

Some specific organofunctional silanes can be analyzed by colorimetry with suitablereagents. Examples include the detection of epoxy silanes with bromothymolblue/thiosulfate solution, of methacryloxy and vinylsilanes with permanganate salts,and of amino/diaminosilanes with ninhydrin [29]. The results are semi-quantitativewithin the same system of mineral and silane; in all other cases they are only of aqualitative nature.

4.8.5Acid-Base Titration

A slurry of a treated filler can be analyzed by conventional acid-base titration if thesilane changes the pH of the mineral. For example, application of an aminosilane toan acidic mineral will change its surface pH. Such changes can be monitored by titra-tion in the presence of conventional indicators. The results are semi-quantitativewithin the same system of mineral and silane; in all other cases they are only of aqualitative nature.

4.8.6Empirical Tests for Hydrophobicity

Testing of hydrophobic minerals is comparatively easy, even if this test method is on-ly an empirical one. Placing a mineral that has been treated with, for example, analkyl-, vinyl- or methacryloxysilane into distilled water will immediately show how ef-fective and how uniform the treatment has been through the amount of filler whichf loats on the water. In another empirical method, a drop of water is applied to a pile


Fig. 4-11 Gas chromatogram of a pyrolyzed octyltriethoxysilane/TiO2 system.

74

of the treated mineral. The time required for the water droplet to permeate into themineral provides an indication of the quality of the treatment.

4.8.7Combined Silane/Colorant Surface Modif ication

The incorporation of a suitable colorant into the silane solution may show the distri-bution of the silane on the treated mineral (for example, rhodamine B in isobutyltri-ethoxysilane). It is important that the colorant does not block any of the active siteson the mineral surface and that it spreads as homogeneously on the surface as thesilane itself. The colorant needs to be soluble in the carrier. In this way, the normallyinvisible silane coating may be visualized; a proper treatment will lead to a homoge-neously colored surface.

4.9Selection of Silanes

A list of silanes providing the best property enhancements in a variety of filled poly-mer systems is given in Table 4-1. The list, based on experience, is not exhaustive andvariations are possible. Major suppliers of silanes include Degussa Corp., Dow-Corn-ing, GE-OSI, Shin-Etsu, and Wacker-Chemie GmbH.

Tab. 4-1 Types of polymers and recommended silanes.

Polymer Silane Functionality Examples of Commercial Silanes

acrylic acrylate, methacrylate, 3-(methacryloxypropyl)trimethoxysilanevinyl

butyl rubber diamino N-(2-aminoethyl)-3-(aminopropyl)trimethoxysilaneEVA amino, vinyl 3-(aminopropyl)triethoxysilane,

special aminosilane blends, vinyltriethoxysilane

neoprene mercapto, diamino 3-(mercaptopropyl)trimethoxysilane, N-(2-aminoethyl)-3-(aminopropyl)trimethoxysilane

nitrile rubber mercapto 3-(mercaptopropyl)trimethoxysilanepolyamide amino, 3-(aminopropyl)triethoxysilane,

secondary amino N-(n-butyl)-3-(aminopropyl)trimethoxysilaneunsaturated methacrylate, 3-(methacryloxypropyl)trimethoxysilane, polyester polyether polyether-functional trimethoxysilanepolyester amino, epoxy 3-aminopropyltriethoxysilane, thermoplastic 3-glycidyloxypropyltrimethoxysilanepolyolefin vinyl, alkyl vinyltriethoxysilane

hexadecyltrimethoxysilaneEPR, EPDM vinyl, sulfur, vinyltriethoxysilane, vinyltrimethoxysilane,

mercapto, vinyltris(2-methoxyethoxy)silane, thiocyanato vinyl/alkyl-functional siloxane oligomer

SBR sulfur, mercapto bis(triethoxysilylpropyl)polysulfane, 3-mercaptopropyltrimethoxysilane


75

The following example concerning vulcanized EPDM-filled systems shows the im-portance of correct selection of the silane functionality [4,30]. The surface reactivityof many non-black fillers generally precludes strong bonding with a polymer matrixand, in some cases, leads to poor compatibility and dispersion. These filler surfacescan be made polymer-reactive if careful consideration is given to the choice of theorganofunctionality of the silane. The importance of selecting the right couplingagent with an organofunctionality complementary to the curative/polymer system isillustrated in Table 4-2. The 300% modulus values are compared for two filled EPDMsystems, cured with sulfur and peroxide, respectively. In the formulation, one partper hundred parts of resin (phr) of each silane indicated was introduced during com-pounding of 100 phr of filler in a two-roll mill, and the composite properties were de-termined by standard (DIN 53504) methods.

Tab. 4-2 Comparison of 300% modulus of differently curedEPDM-filled systems containing different silanes.

Silane 300% Modulus, MPa 300% Modulus, MPa(sulfur-cured EPDM, talc-f illed) (peroxide-cured EPDM, clay-f illed)

Control, no silane 490 420Vinyl 430 1110Mercapto 790 1200Amino 790 1440Methacryloxy Not determined 1660

With regard to the sulfur-cured system, the primary amino-functional silane andthe mercapto- functional silane are able to participate in the cure mechanism to a fargreater extent than the vinylsilane, as shown by the higher modulus values. In theperoxide-cured system, all of the silanes promote significant improvements in mod-ulus, but to different degrees depending on their relative reactivity. The methacryloxy-functional silane is considerably more effective than the vinylsilane; this could be pre-dicted from consideration of the relative reactivities of the double-bond moieties (car-bonyl vs. vinyl). The aminosilane also provides a high level of filler–polymer interfa-cial bonding, whereas the mercaptosilane is relatively less effective compared to themethacryloxysilane.

An example of a different reaction of an amino-functional silane with a modifiedpolymer is the reaction of the amino groups with maleated polypropylene to form animide (Scheme 4-4). This reaction is extensively used to couple treated glass fiberswith polyolefins (see Chapter 7).

4.9 Selection of Silanes

76

4.10Applications of Specif ic Silanes

4.10.1Vinylsilanes

Commercial vinylsilanes generally have the vinyl group directly attached to the sili-con atom. The hydrolyzable groups are usually methoxy, ethoxy, or 2-methoxyethoxy.This type of functionality is used in polymers that are cross-linked by a free-radicalprocess (peroxide cure). The vinyl group is, however, not sufficiently reactive for allsystems and the methacryloxy functionality is sometimes preferred, as shown inTable 4-2 [31]. Vinylsilanes, because of their overall cost/performance advantages,have become the industry standard for EPR/EPDM and wire and cable applications.

A particularly important area of application of vinylsilanes is in aluminum trihy-drate (ATH)-filled EVA. The first generation of halogen-free f lame-retardant (HFFR)materials possessed excellent fire and smoke properties, but were mechanically weakand slow to process when compared with the PVC compounds that they were replac-ing. Today, the majority of thermoplastic HFFR cable materials are made ofATH/EVA and this represents a rapidly growing and specialized area of cable pro-duction. Vinylsilane adhesion promoters make possible the high loading levels ofATH required for effective f lame retardancy, the improvement in melt processabilityof the highly filled EVAs, and the enhancement of the mechanical properties of thefinished product. Oligomeric vinylsilanes, in addition, may provide a significantly re-duced quantity of alcohol (and hence lower VOC emissions) upon reaction with mois-ture. Appropriate use of oligomeric vinylsilanes may also reduce compound viscosi-ty and produce a smooth, defect-free surface.


Scheme 4-4 Reaction of an aminosilane with maleated polypropylene.

77

Hydrated fillers such as ATH achieve their f lame retardance by decomposing en-dothermically with the release of water close to the temperature at which the poly-mers themselves decompose (see Chapter 18). They do not have the smoke and cor-rosive gas problems associated with other types of f lame retardants. In order to pro-duce an acceptable HFFR compound at very high ATH loadings of 60 to 65 wt. %, theATH particle size and shape have to be carefully controlled. This is the prime goal ofthe ATH manufacturing process. Experience suggests that large and thick particlesof ATH with a low surface area are required for effective f lame retardation. When us-ing vinylsilanes in HFFR materials, a small amount of peroxide is required to obtaingood coupling.

The tensile properties of various EVA/ATH formulations are summarized inTable 4-3. The basic formulation contained 160 phr of ATH, 1 phr of stabilizer, andvariable amounts of monomeric and oligomeric silanes and peroxide. A co-rotatingtwin-screw extruder was used to produce sheets for the tests. The silane content isbased on the filler; the silane was pre-blended with the EVA. Dicumyl peroxide (DCP)and Irganox 1010 (phenolic stabilizer) were used as peroxide and stabilizer, respec-tively. A control without silane is not included since it led to scorch.

Tab. 4-3 Comparison of EVA/ATH HFFR containing vinylsilanes.

Vinylsilane [phr] Peroxide [phr] Tensile Strength Elongation at [MPa] Break [%]

Oligomeric methoxy- 1.5 no peroxide 12.3 210based vinylsilane

1.6 0.03 16.8 2131.6 0.04 17.4 2031.6 0.05 scorch scorch1.85 0.03 16.9 2131.85 0.05 17.6 2102.1 0.03 16.4 2122.1 0.04 17.4 2042.1 0.05 17.9 199

Monomeric methoxy- 2.5 0.03 16.4 192based vinylsilane

For oligomeric silanes, the absence of peroxide results in poor tensile strength. Inthe presence of peroxide, however, the overall picture changes dramatically. As theATH couples with the EVA, tensile strength increases and water uptake is reducedboth through increasing the cross-link density and by rendering the compound hy-drophobic. Elongation at break is not significantly affected by the presence of perox-ide. Oligomeric vinylsilanes perform better than monomeric vinylsilanes, even at alower concentration of 1.6 phr. At these low silane levels, the vinylsilane/peroxide ra-tio has to be monitored carefully. At a silane concentration of 1.6 phr, the peroxideconcentration should not exceed 0.04 phr because of the risk of scorch. As is clearlydemonstrated in Table 4-3, increased silane levels reduce the risk of scorch signifi-cantly. In general, the properties achieved by the use of oligomeric vinylsilanes are

4.10 Applications of Specif ic Silanes

78

clearly superior to those achieved with commonly employed monomeric vinylsilanes,even at the lowest concentrations.

4.10.2Aminosilanes

4.10.2.1 GeneralAmino-functional silanes are widely used for the surface treatment of fillers such aswollastonite and calcined clay. Commercial silanes of this type are usually based onthe primary γ-aminopropyl functionality. They have a wide versatility, also being usedin epoxies, phenolics, polyamides, thermoplastic polyesters, and elastomers. Unlikefor most other silanes, their aqueous solutions are quite stable as a result of hydro-gen bonding between the silanol groups and the primary amine. An internal five- orsix-membered ring is formed (see Scheme 4-5).

The reactivity of the primary amino group has hampered elucidation of the natureof surface layers resulting from its adsorption on the filler. The amino group itselfmay absorb strongly on a variety of surfaces and has also been shown to be very proneto hydrogencarbonate salt formation with atmospheric carbon dioxide. Modern ana-lytical procedures are needed to elucidate some of the important features of thesecoatings; Ishida [18,19] has given a very good account of the current state of knowl-edge.

4.10.2.2 Calcined Clay Filled PolyamidesImpact strength is very important for mineral-filled polyamides. It is directly relatedto the filler loading level and decreases with increasing loading. In the absence of afine and uniform dispersion of the filler particles, agglomerates acting as stress con-centrators will provide sites at which impact failure originates. Optimizing filler dis-persion would, therefore, minimize the formation of filler agglomerates and yieldmore homogeneous materials with improved impact properties.

The polyamide-6,6 selected in the example below was a general purpose, lubricat-ed material. The filler used was a fine calcined clay that was surface modified with1 wt. % aminosilane by employing a standardized laboratory surface treatmentprocess. The polyamide-6,6 and the silane-treated filler were pre-dried (24 h at 80 °C)prior to compounding. Compounds were prepared at 40 wt. % filler level in a co-ro-tating twin-screw extruder. All injection-molded test samples were conditioned (24 h,


H2N Si(OC2H5)3

+ 3 H2O

– 3C2H5OH

OH

OHSi

OH

HH

N

δ

δ

δ

Scheme 4-5 Autocatalytic hydrolysis of γ-aminopropyltriethoxysilane.

79

23 °C, 50% RH) prior to testing as per the DIN 50014 procedure. Table 4-4 summa-rizes the properties of the nylon-6,6/clay compounds obtained using two differentsilanes. Aminosilane treatment of calcined clay dramatically improves the propertiesof the filled polyamide-6,6. Data are not shown for nylon compounds with 40 wt. %untreated clay since such compositions without silane led to scorch.

Tab. 4-4 Properties of nylon-6,6/clay compounds using twodifferent aminosilanes

Property γ-aminopropyl- N-butyl-γ-aminopropyl-triethoxysilane trimethoxysilane (DYNASYLAN® AMEO) (DYNASYLAN® 1189)

Ultimate tensile strength [MPa][a] 72.0 69.1

Ultimate tensile strength [MPa][b] 36.5 34.7

Flexural modulus [GPa] 3.7 3.5

Charpy impact strength, 32.2 41.7unnotched [kJ m–2]

Izod impact strength, 19.5 30.1unnotched [kJ m–2]

Charpy impact strength, 4.0 4.0notched [kJ m–2]

Izod impact strength, 4.1 4.0notched [kJ m–2]

Melt f low index [g⋅10 min–1]* 43 55

By using (N-n-butyl)-γ-aminopropyltrimethoxysilane, the impact strength of the fi-nal compound can be improved even further. Good calcined clay dispersion in thepolymer phase leads to low compound viscosities (as measured by MFI) and betterprocessability. The use of (N-n-butyl)-γ-aminopropyltrimethoxysilane further reducescompound viscosity and results in higher melt f low rates, improved dispersion, few-er agglomerates, and a smooth, defect-free surface. Treatment with this non-polarsilane also results in reduced moisture uptake by the filler.

4.10.2.3 ATH-Filled EVAIn addition to vinylsilanes, ATH- and MDH-filled polyethylene–ethylene vinylacetate(EVA) copolymers can be effectively coupled with aminosilanes [32]. Generally, fineand uniform ATH dispersion in the polymer leads to HFFR compounds with low vis-cosities (MFI) and high tear strength. The use of 3-aminopropyltriethoxysilane be-came an industrial standard for the production of reliable thermoplastic HFFR com-pounds. New proprietary primary/secondary aminosilane blends have provided fur-ther improvements in the final HFFR compounds [33]. Aminosilanes are also found

* DIN ISO 1133 (Method B), temperature275 °C, preheating time: 1 min, load: 5 kg.

[a] Measurement after 7 days/90 °C immersionin water.

[b] Measurement at 275 °C/5 kg.


80

in thermoplastic, non-peroxide cross-linked EVAs, which is an alternative technologywith widespread use in industry.

The properties of two different EVA/ATH formulations are summarized in Table4-5. The basic formulation contained 160 phr of ATH, 1 phr of stabilizer (Irganox1010), and 1.5 phr of aminosilane based on filler pre-blended with EVA. Compoundswere made in a co-rotating twin-screw extruder and test specimens were producedfrom extruded sheets.

Tab. 4-5 Comparison of EVA/ATH HFFR composites containingaminosilanes.

Silane Ultimate tensile Elongation at Tear strength Water uptake strength break [%] [N mm–1] (14 d @ 70 °C) [N mm–2] [mg cm–3]

Aminopropyl- 16.3 200 10.2 4.02triethoxysilane

Special amino- 16.6 210 11.6 3.75silane blend (DYNASYLAN®

1204, Degussa)

4.10.3Methacryloxysilanes

Methacryloxysilanes provide more reactive forms of unsaturation than vinylsilanesand are used extensively in free-radical curing formulations where the extra reactivi-ty is of benefit. Commercial products usually contain γ-methacryloxypropyl groups.The presence of a carbonyl group in the molecule leads to a tendency for it to orientf lat on the filler surface under certain conditions [34].

Thermosetting filler systems prepared from unsaturated polyesters (UP), methylmethacrylate (MMA), vinyl esters, epoxy (EP), phenolic, and furan resins are used inmany applications. It can be shown that, after pre-treatment of the filler, theorganofunctional silanes cause a marked reduction in viscosity and an enhancementof mechanical properties, such as f lexural and impact strength. This is especially ev-ident after exposure to moisture. When applied onto the filler, the methacryloxysilaneis immediately effective in acrylic casting systems, reducing the viscosity to 24% ofits initial value; in contrast, as an additive the silane brings about a viscosity reduc-tion to only 76% of the initial value even after 24 h storage time.

As an example of the effect of methacryloxysilane on a filled UP system, a formu-lation containing 63% quartz f lour is selected. Pre-treatment of the filler with 1%silane resulted in 38% lower viscosity compared to the silane-free composite. Theorganofunctional silane led to greater retention of the mechanical properties com-pared to the silane-free system, as shown by f lexural strength data after 6 h in boilingwater. Maximum effectiveness was achieved by introducing the silane through pre-treatment of the filler.


81

The effect of the methacryloxy-functional silane in a cristobalite-filled PMMA resincan be visualized by SEM in the case of cryofractured samples that contain 100 phr offiller and either zero or 1 phr of silane, respectively (Figure 4-12). A gap between thepolymer matrix and the filler particle indicating debonding is visible in the absenceof silane. In the presence of silane, no gap between the resin and the cristobalite par-ticles can be detected and the composite breaks in the polymer phase as a result of theimproved adhesion. Similar effects can be observed for highly filled ATH/PMMAsystems. For example, in a system containing 60 wt. % ATH, the viscosity decreasesto one-third of the original value when 0.5% silane is applied to the filler. Flexuralstrength and impact strength are higher by 22% and 35%, respectively, compared tothe values for the silane-free composite.

4.10.4Epoxysilanes

For electronics applications, epoxy resins filled with particulate inorganic fillers pro-vide insulation properties. In addition to decreasing cost, these fillers serve to in-crease hardness, act as a heat sink for the exothermic curing reaction, decreaseshrinkage during curing, and improve other properties, particularly retention of me-chanical and electrical properties after extensive exposure to water. Using 3-glycidyl-oxypropyltrimethoxysilane, the viscosity of an epoxy containing 60 wt. % quartz f lourcoated with 1 wt. % silane drops to 80% of the value of the untreated filler compos-ite. After 6 h in boiling water, f lexural strength and impact strength are higher by210% and 250%, respectively, compared to the values for the silane-free material.

4.10.5Sulfur-Containing Silanes

An important application of these silanes is in tires, where silicas and silanes are usedto reduce rolling resistance. Here, the silane is added during rubber compounding,either neat or in the form of a dry liquid, in order to react with the silanol groups ofthe silica. The rubber-active group of the silane (tetrasulfane, disulfane, thiocyanatoor mercapto group) has a strong tendency to form rubber-to-filler bonds during cur-


Fig. 4-12 Comparison of cristobalite-filled PMMA; left untreated,right treated with 1% methacryloxysilane.

82

ing of the rubber compound [35]. It should be noted that silane pre-treated silica isonly used for technical rubber goods, and not for tires.

4.11Trends and Developments

Future trends for functional, surface-treated minerals in plastics applications are like-ly to include reduced surface moisture for polyamide applications and reduced VOCachieved by the use of oligomeric silanes or aqueous solutions. In particular, lowvolatility silanes are of great interest in the industry as they allow higher treatmenttemperatures and shorter dwell times (increased throughput). Also, the combinationof silane functionalities in a single silane molecule (e.g. combined amino and alkylfunctionalities, alkyl/vinyl, etc.) is of interest for easier handling and better propertiesof the end products. Finally, one challenge remains: an effective and commercially vi-able silane for filled PP is not yet on the horizon.


83

References

1 Ishida, H., Applied Sciences 1993, 230,169–199.

2 Skudelny, D., Kunststoffe 1987, 77(11),1153–1156.

3 Ramney, M. W., et al., Proc. Annual Rein-forced Plastics/Composites Institute, Society ofPlastic Industry (SPI), 1972, 21D, 1–22.

4 Marsden, J. G., Applied Polymer Symposia1970, 14, 107–120.

5 Marsden, J. G., Ziemianski, L. P., Brit.Polym. J. 1979, 11(4), 199–205.

6 Plueddemann, E. P., Macromolecular Mono-graphs 1980, 7, 31–53.

7 Plueddemann, E. P., Addit. Plast. 1978,123–167.

8 Ramney, M. W., et al., Composite Materials1974, 6, 131–172.

9 Atkins, K. E., Gentry, R. R., Gandy, R. C.,Polym. Eng. Sci. 1978, 18(2), 73–77.

10 Plueddemann, E. P., Stark, G. L., ModernPlastics 1977, 54(8), 76–78.

11 Plueddemann, E. P., Leyden, D. E., Collins,W. T. (Eds.), Midland Macromolecular Mono-graphs 1980, 7, 31.

12 Beari, F., et al., J. Organomet. Chem. 2001,625, 208–216.

13 Arkles, B., et al., Modern Plastics 1987, 55(4),138–143.

14 Brand, M., et al., Z. Naturforsch. 1999, 54b,155–164.

15 Rosen, M. R., J. Coatings Technology 1978,50(644), 70–82.

16 Hanisch, H., Steinmetz, J., Plastics Com-pounding 1987, 5, 25–30.

17 Harding, P. H., Berg, J. C., J. Adhesion Sci-ence and Technology 1997, 11(4), 471–493.

18 Ishida, H., Polym. Composites 1984, 5, 101.

19 Ishida, H., Polym. Sci. Technol. 1985, 27, 25.20 Bauer, F., et al., Macromol. Chem. Phys.

2000, 201(18), 2654–2659.21 Feresenbet, E., et al., J. Adhesion 2003, 79,

643–665.22 Görl, U., et al., KGK (Kautschuk Gummi

Kunststoffe) 1999, 52(9), 588–598.23 Hussain, A., Pf lugeil, C., Kleben & Dichten

Adhaesion 1994, 38, 22–25.24 Rotzsche, H., Ditscheid, K.-P., EP 0741293

A2, 1996.25 Trifonova-Van Haeringen, D., et al., Rubber

Chemistry and Technology 1999, 72(5),862–875.

26 Garbassi, F., et al., J. Colloid and InterfaceScience 1987, 117(1), 258–270.

27 Bartella, J., VDI (Verein Deutscher Inge-nieure) Bildungswerk, BW 6998.

28 Albers, P., Lechner, U., Kunststoffe 1991,81(5), 420–423.

29 Hartwig, A., Journal Oberf laechentechnologie1997, 46–51.

30 Sampson, P. N., IEEE Electrical InsulationMagazine 2001, 17, 33–37.

31 Wang, G., et al., J. Appl. Polym. Sci. 2002,85(12), 2485–2490.

32 Schofield, W. C. E., et al., Composite Inter-faces 1998, 5(6), 515–528.

33 Mack, H., 48th International Wire & CableSymposium Proceedings, 401–405, AtlanticCity, NJ, USA, Nov. 15–18, 1999.

34 Hanisch, H., et al., Proc. Plastics Compound-ing, 1987.

35 Hunsche, A., et al., KGK (Kautschuk Gum-mi Kunststoffe) 1997, 50, 881–889.

References

5Titanate Coupling Agents

Salvatore J. Monte

5.1Introduction

Titanate coupling agents impart increased functionality to fillers in plastics. The dif-ferent ways that these additives work in filled polymers can be explained by breakingdown the various mechanisms of the titanate (or zirconate) molecule into six distinctfunctions. Filler pre-treatment and in situ reactive compounding with titanates (andzirconates) to effect coupling, catalysis, and heteroatom functionality in the polymermelt are also discussed.

Esters of titanium or zirconium couple or chemically bridge two dissimilar speciessuch as an inorganic filler/organic particulate/fiber and an organic polymer throughproton coordination. This permits coupling to non-hydroxyl bearing, and thereforenon-silane reactive, inorganic substrates such as CaCO3 and boron nitride as well asorganic substrates such as carbon black and nitramines without the need of water ofcondensation as with silanes. The thermally stable quaternary carbon structure of theneoalkoxy organometallics permits in situ reactions to take place in the thermoplasticmelt. In addition, the coupling of monolayers of a phosphato or a pyrophosphato het-eroatom titanate or zirconate imparts synergistic intumescence to non-halogenatedf lame retardants such as Mg(OH)2 and aluminum trihydrate (ATH); f lame retar-dance function to fillers such as CaCO3; control of the burn rate and burn rate expo-nent of aluminum powder rocket fuels; and extinction of the f lame spread of spallsof polymer-bound nitramines used in propellants and explosives. It is also believedthat the organometallic monolayer covered filler surface becomes a catalysis supportbed for “repolymerization” of the surrounding polymer phase, thus allowing fillersto act as mechanical property improvers. Furthermore, the in situ monomoleculardeposition of titanate on the surface of a particulate, such as a nanofiller, renders theparticulate hydrophobic and organophilic. Under melt compounding shear condi-tions, the titanate assists in the removal of air voids and moisture from the particlesurface, resulting in complete dispersion and formation of a true continuous phase,thus optimizing filler performance.


86

Minor amounts of thermally stable neoalkoxy titanate and zirconate additives mayprovide a means for post-reactor, in situ metallocene-like “repolymerization” catalysisof a filled or unfilled polymer during the plasticization phase. This may result in thecreation of metallocene-like (titanocene or zirconocene) behavior associated with ef-fects such as increased composite strain to failure, resulting in increased impacttoughness, or enhanced polymer foamability. Other effects to be discussed belowwith specific examples are related to enhanced processability, reduced polymer chainscission, shortened polymer recrystallization time, and the compatibilization of dis-similar polymers.

There is a significant body of published information on titanium- and zirconium-based coupling agents. During the period 1974–2002, 1635 patents and 344 technicalpapers appeared. Some detailed historical documentation with more than 500 fig-ures and tables is provided by the author in ref. [1]. References [2–21] are some of thetechnical papers and conference presentations by the author. Table 5-1 provides achemical description of the coupling agents discussed in this chapter, along with analpha-numeric code for the titanates and zirconates invented by the author. The al-pha-numeric code is often used alone in this chapter for the sake of brevity. Table 5-2 indicates the designation of just a few of more than 50 commercial titanates andzirconates available from various vendors, such as Kenrich Petrochemicals, Inc., E. I.du Pont de Nemours and Company, Synetix–Johnson Matthey, Nippon Soda, and Aji-nomoto Fine Techno Co. (Kenrich licensee). Chemical strutures of two common ti-tanales (KR TTS and LICA 38) are shown in the Appendix.

5 Titanate Coupling Agents

Tab. 5-1 Coupling agent chemical description – alpha-numeric code

Code Nomenclature

Silanes[a]

A-187 3-glycidoxypropyl, trimethoxysilaneA-1100 3-aminopropyl, trimethoxysilaneTitanates[b]

KR TTS titanium(IV) 2-propanolato, tris(isooctadecanoato-O)KR 7 titanium(IV) bis(2-methyl-2-propenoato-O), isooctadecanoato-O, 2-propanolatoKR 9S titanium(IV) 2-propanolato, tris(dodecyl)benzenesulfonato-OKR 33CS titanium(IV), tris(2-methyl)-2-propenoato-O, methoxydiglycolylatoKR 38S titanium(IV) 2-propanolato, tris(dioctyl)pyrophosphato-OKR 41B titanium(IV) tetrakis(2-propanolato), adduct with 2 moles (dioctyl)hydrogen

phosphiteKR 46B titanium(IV) tetrakis(octanolato) adduct with 2 moles (ditridecyl)hydrogen

phosphiteKR 55 titanium(IV) tetrakis[bis(2-propenolato methyl)-1-butanolato] adduct with

2 moles (ditridecyl)hydrogen phosphiteKR 112 titanium(IV) oxoethylene-diolato, bis(dioctyl) phosphato-OKR 138S titanium(IV) bis(dioctyl)pyrophosphato-O, oxoethylenediolato (adduct), (dioctyl)

(hydrogen) phosphite-OKR 238S titanium(IV) ethylenediolato, bis(dioctyl)pyrophosphato-OLICA 01 titanium(IV) 2,2-bis(2-propenolatomethyl)butanolato, tris(neodecanoato-O)LICA 09 titanium(IV) 2,2-bis(2-propenolatomethyl)butanolato, tris(dodecyl) benzenesul-

fonato-O

875.1 Introduction

Tab. 5-1 Continued

Code Nomenclature

LICA 12 titanium(IV) 2,2-bis(2-propenolatomethyl)butanolato, tris(dioctyl) phosphato-OLICA 38 titanium(IV) 2,2-bis(2-propenolatomethyl)butanolato, tris(dioctyl)

pyrophosphato-OLICA 38ENP titanium(IV) 2,2-bis(2-propenolatomethyl)butanolato, tris(dioctyl)

pyrophosphato-O/ethoxylated nonyl phenol (1:1)LICA 38J titanium(IV) bis(2-propenolatomethyl)-1-butanolato, bis(dioctyl) pyrophosphato-

O, adduct with 3 moles N,N-dimethylaminoalkyl propenoamideLICA 44 titanium(IV) 2,2-bis(2-propenolatomethyl), tris(2-ethylenediamino)ethylatoLICA 97 titanium(IV) 2,2-bis(2-propenolatomethyl)butanolato, tris(3-amino)phenylatoKS N100 combined mononeoalkoxy titanatesKS N60WE 60% active combined mononeoalkoxy titanates containing water emulsifiers

Zirconates[b]

KZ 55 zirconium(IV) tetrakis[2,2-bis(propenolatomethyl)butanolato], adduct with2 moles bis(tridecyl) hydrogen phosphite

NZ 12 zirconium(IV) 2,2-bis(2-propenolatomethyl)butanolato, tris(dioctyl)phosphato-ONZ 37 zirconium(IV) bis[2,2-bis(2-propenolatomethyl)butanolato],

bis(para-aminobenzoato-O)NZ 38 zirconium(IV) 2,2-bis(2-propenolatomethyl)butanolato,

tris(dioctyl)pyrophosphato-ONZ 39 zirconium(IV) 2,2-bis(2-propenolatomethyl)butanolato, tris(2-propenoato-O)NZ 44 zirconium(IV) 2,2-bis(2-propenolatomethyl), tris(2-ethylenediamino)ethylatoNZ 97 zirconium(IV) 2,2-bis(propenolatomethyl), tris(3-amino)phenylatoKS MZ100 combined trineoalkoxy zirconatesKS MZ60WE 60% active combined trineoalkoxy zirconates containing water emulsifiers

[a] OSi Specialties, GE Silicones.[b] Kenrich Petrochemicals, Inc.

Tab. 5-2 Titanate and zirconate coupling agents form designation[a]

Liquid Form

KR® # = liquid coupling agent, 100% active monoalkoxy, chelate and coordinate titanate.LICA® # = LIquid Coupling Agent, 100% active neoalkoxy titanate.NZ® # = liquid coupling agent, 100% active Neoalkoxy Zirconate.Example: LICA 12 = 100% liquid neoalkoxy titanate.

Powder Masterbatch Form

CAPOW® L® #/Carrier = Coupling Agent POWder.Example: CAPOW L 12/H = 65% LICA 12/35% PPG Hi Sil 233 silica carrier.

Pellet Masterbatch Form

CAPS® L® #/binder = Coupling Agent Pellet System, 20% active neoalkoxy titanate/binder.Example: CAPS L 12/L = 20% LICA 12/10% silica/70% LLDPE binder

Water-Soluble Quat Form Quat Blend (QB) Part RatioDesignation KR 238M :LICA 38J :NZ 38JQB 012 0 : 1 : 2QB 521 5 : 2 : 1

[a] Kenrich Petrochemicals, Inc.

88

5.2The Six Functions of the Titanate Molecule

Organosilanes have long been used to enhance the chemical bonding of a variety ofthermoset resins with siliceous surfaces and more recently of thermoplastics. How-ever, Plueddemann [22] observed that organosilanes are essentially non-functional asbonding agents when employing carbon black, CaCO3, boron nitride, graphite,aramid or other organic-derived fibers.

A discussion of the six functional sites of a titanate (or zirconate) as compared to asilane (see structures in Table 5-1 and Chapter 4) is useful to explain their perform-ance differences and each may be represented as follows:

Titanate Silane(1) (2) (3)(4)(5)(6) (1) (5)

(RO)n-Ti-(-O X R’ Y)4–n (RO)3Si-(-R’ Y)

where:(1) RO = hydrolyzable group/substrate-reactive group with surface hydroxyls or

protons.(2) Ti (Zr), Si = tetravalent titanium, zirconium or silicon. The Ti–O (or Zr–O) bond

is capable of disassociation allowing transesterification, transalkylation, and othercatalyzed reactions such as “repolymerization”, while the Si–C bond is more stableand thus unreactive.

(3) X = binder functional groups such as phosphato, pyrophosphato, sulfonyl, car-boxyl, etc., that may impart intumescence, burn rate control, anti-corrosion, quater-nization sites, dissociation rate/electron-transfer control, etc.

(4) R’ = thermoplastic-specific functional groups such as aliphatic and non-polarisopropyl, butyl, octyl, isostearoyl groups; naphthenic and mildly polar dodecylbenzylgroups; or aromatic benzyl, cumyl or phenyl groups.

(5) Y = thermoset (but also thermoplastic)-specific functional groups such as acry-lyl, methacrylyl, mercapto, amino, etc.

(6) 4–n = mono-, di- or triorganofunctionality. Hybrid titanate (zirconate) couplingagents, such as those containing 1 mole each of a carboxyl [function (3)] and aliphat-ic isostearoyl [function (4)] ligand and 2 moles of carboxyl [function (3)] and acrylyl[function (5)] ligands, are possible.

Therefore, function (1) relates to filler/fiber substrate reaction mechanisms, whilefunctions (2) to (6) are polymer/curative reactive.

5.2.1Effects of Function (1)

The functional site (1) of the titanate molecule is associated with coupling, disper-sion, adhesion, and hydrophobicity effects. These effects are also related to themethod of application of the titanate on the filler surface, as discussed below.


89

5.2.1.1 CouplingIn its simplest terms, the titanate function (1) mechanism may be classed as proton-reactive through solvolysis (monoalkoxy) or coordination (neoalkoxy) without theneed of water of condensation, while the silane function (1) mechanism may beclassed as hydroxyl-reactive through a silanol–siloxane mechanism requiring water ofcondensation. The silane’s silanol–siloxane water of condensation mechanism limitsits reactions to temperatures below 100 °C, thereby reducing the possibility of in situreaction in the thermoplastic or elastomer melt above 100 °C as is possible with ti-tanates. In addition, a variety of particulate fillers such as carbonates, sulfates, ni-trides, nitrates, carbon, boron, and metal powders used in thermoplastics, ther-mosets, and cross-linked elastomers do not have surface silane-reactive hydroxylgroups, while almost all three-dimensional particulates and species have surface pro-tons thereby apparently making titanates more universally reactive.

5.2.1.2 DispersionDispersion of fillers results from the application of electrochemical and mechanicalforces to the interface of the inorganic filler/polymer so as to cause complete deag-glomeration to the attrited or original particle size in an organic phase, completeelimination of air voids and water, and the creation of a true continuous inorgan-ic/organic composition. The coupling of the titanate to the inorganic/organic sub-strate in monolayers allows for elimination of air voids, enhanced hydrophobicity,and a complete continuous phase for stress/strain transfer. Figure 5-1 shows the au-thor’s envisaged “before and after” effect of a titanate monolayer on agglomeratedfillers.

C20 aliphatic mineral oil can be used as a low molecular weight model for poly-olefins. Since it is non-polar and, thus, a poor medium for dispersion of most polarfillers, coupling agent effects can be more easily measured. Figure 5-2 shows the ef-fect of 0.5% isopropyl triisostearoyl titanate (KR TTS; see Table 5-1) on the dispersionof CaCO3 in a non-polar mineral oil. The deagglomeration effect is apparent. Signif-icant viscosity reductions have been observed through the application of the same ti-tanate (at 0.5 to 3 wt. %) on numerous fillers such as 2.5 µm CaCO3 (70 wt. % filler)or clay (30 wt. % filler) in mineral oil, and 40 wt. % TiO2 in dioctyl phthalate plasti-cizer. Figure 5-3 shows the shift in the critical pigment volume concentration point(CPVC) of CaCO3-filled mineral oil using 0.5 wt. % KR TTS. The CPVC is defined asthat point at which addition of more filler to an organic phase will cause incompletewetting due to insufficient organic binder being available to wet the additional inor-ganic filler surface. The shift in the CPVC as a result of “coupling” (function 1) maybe extended from the mineral oil model to filled thermoplastic and thermoset sys-tems allowing higher loading to equivalent oil demand, and improved relative me-chanical properties at any filler loading below the CPVC [1]. Figure 5-4 transitionsfrom the model f luid to polypropylene (PP) polymer as the organic phase and showsthe f lexibility imparted to a sample containing 70 wt. % CaCO3 (3 µm average parti-cle size) in a PP homopolymer using 0.5 wt. % KR TTS with respect to the filler. A PPcomposite containing 70% untreated CaCO3 will normally break in a brittle manner,while the titanate treatment allows 180º bending with no white stress cracking.

5.2 The Six Functions of the Titanate Molecule

90

Voelkel et al. [24] have recently reported on the use of solubility parameters in thecharacterization of silica surface-modified with titanate by inverse gas chromatogra-phy and compared their findings with earlier work on silanes.


Fig. 5-1 Illustration of the dispersion effect of coupling a titanatemonolayer on an agglomerated inorganic (left: no titanate) in anorganic phase, thereby creating a continuous inorganic/organicphase (right: with titanate) by deagglomeration and subsequentelimination of air and water from the interface.

Fig. 5-2 Left: A micrograph of a suspension of a CaCO3 (untreat-ed)/liquid paraffin system. Right: A micrograph of a suspension of aCaCO3 (treated with KR TTS)/liquid paraffin system demonstratingdeagglomeration.

91

5.2.1.3 AdhesionOne of the reasons why the dispersion of inorganics in plastics and the adhesion of aplastic to an inorganic substrate is so difficult is because many thermoplastics andrubbers, such as olefin-based polymers, are non-polar. Titanates and zirconates arewell established adhesion promoters, as discussed in ref. [1]. A recent example of


Fig. 5-3 Pigment volume concentration curves comparing thecritical pigment volume concentration (CPVC) point of untreatedCaCO3-filled mineral oil (left) with 0.5% KR TTS-treated CaCO3-filledmineral oil (right) predicting the ability to fill plastics with higherloadings of filler without detracting from mechanical properties.

Fig. 5-4 A demonstration of the f lexibility of0.5% KR TTS-treated, 70% CaCO3 (2.5 µm)filled PP homopolymer. Untreated 70% CaCO3-filled PP would snap and break at the slightestdeformation.

92

bonding polyolefins to metals appears in ref. [24]. Another example of the adhesionof polyolefins to foil electrodes using KR TTS has been provided by Kataoka [25].

5.2.1.4 HydrophobicityHydrophobicity is a desirable property to impart to fillers, functional particulates,and fibers to provide long-term protection of their composites against corrosion andageing. The application of increasing amounts of titanate on substrates such as Ca-CO3 results in significant changes in hydrophobicity, as shown from contact anglemeasurements with water droplets. In a recent publication by Krysztafkiewicz et al.[26], hydrophobic modification of a silica surface with silanes and coupling agentswas determined on the basis of the heat of immersion and infrared spectroscopy. Thehighest degree of hydrophobicity was observed for silicas modified with 1 wt. % KRTTS (isopropoxy triisostearoyl) and KR 33CS (isopropoxy trimethacrylic) titanates,while it was a little lower after modification with 3 wt. % aminosilane and methacry-loxysilane. The authors also observed that water was necessary for the silane to cou-ple to the silica, while it was not needed in the case of the titanate.

5.2.1.5 Titanate Application ConsiderationsCorrect usage of titanate coupling agents for optimum performance needs to take in-to account the following considerations:

In situ coupling in the melt phase without the need for water of condensation ispossible through the use of monoalkoxy and neoalkoxy titanate or zirconate groupsat temperatures above 200 °C. However, in situ coupling requires careful consider-ation of good compounding principles to avoid localized, inconsistent or incom-plete coupling because of inadequate specific energy input (low shear) caused byreduced polymer viscosity induced by the coupling agent.

Localization and physical absorption of the coupling agent on the filler or fiber thatresults in whole segments of uncoupled particulate surfaces can be largely over-come by using masterbatches of the coupling agent (see Table 5-2).

In order to effect monomolecular level coupling, the titanate or zirconate must besolubilized in the organic (solvent, plasticizer, polymer) phase or finely emulsifiedin water prior to addition of the filler. If the organic phase has a high molecularweight, then sufficient shear and high mixing torque is needed to assure titanatedistribution.

Uniform distribution of the titanate in the dry powder ingredients or accuratelydosing in the melt dictates the matching of the titanate form to that of the polymeror filler by using appropriate liquid, powder or pelletized titanates (see Table 5-2).

High specific energy input during melt compounding for maximum shear/workenergy for dispersion and complete coupling. Titanates reduce the process tem-peratures of most thermoplastics by approximately 10% and by much more for cer-tain thermoplastic polyesters, acting as transesterification catalysts [1,11]. There-fore, when evaluating a titanate in an unfilled or filled thermoplastic, it is impera-tive to compare both compounds processed at the same specific energy input. Theimportance of specific energy input in relation to the dispersion of fillers during


93

single-screw extrusion compounding of particle-filled thermoplastics is discussedin ref. [27].

The ideal amount of coupling agent to use is that amount that will form a mono-layer on the surface of the filler to produce optimal filler dispersion effects, plus theamount that will have an optimal “repolymerization” catalytic effect on the polymer.Again, filler dispersion is defined as the complete deagglomeration of the filler as at-tritted or precipitated so as to allow all moisture and air in the interstices of the ag-glomerates to be replaced with a continuous organic polymer phase. Optimal “re-polymerization” is defined as being the best balance of mechanical properties andsystem rheology. Various experimental methods for measuring viscosity changes in-duced by titanate in mineral oil/plasticizers are useful for determining the requiredamounts based on the organic polymer, on the filler, or the combined organic and in-organic phases. One effective method for dry filler pre-treatment is to apply the neattitanate, either by airless spraying or by adding it dropwise over a period of 1 minute,to a f luidized bed of the filler as created by a Henschel-type mixer operating at lowspeed (1800 rpm). The treated filler thus obtained can then be compared with an insitu treated control to test the effectiveness of the dry treatment method. This is nec-essary because for certain fillers, such as Mg(OH)2, dilution is needed to avoid local-ization and uneven distribution. For example, to prepare linear low density polyeth-ylene (LLDPE) containing 60 wt. % pretreated Mg(OH)2, the required 0.7% LICA 38first has to be diluted with 2.1% of an alkyl phosphate plasticizer.


“Repolymerization” is a concept that has been patented [28] by the author to explainnew and novel rheology and stress/strain effects in thermoplastics and thermosetsobtained with titanate and zirconate that are independent of cross-linking and cura-tive effects. The aromatic (e.g., phenyl, naphthyl, styrenic) or aliphatic (e.g., ethyl,propyl, butyl) backbones that typically make up the thermoplastic macromolecule,liquid chemical compounds, or thermoplastic elastomers are reactive with titanate (orzirconate) [functions (2) to (4)], independent of any curative reaction mechanisms[function (5)]. Thus, the monolayered, organometallic-coupled particulate and/orfiber may be considered as a catalyst support bed for single-site, in situ metallocene-like “repolymerization” [28] of the surrounding polymer.

Currently, published efforts in metallocene (titanocene and zirconocene) chem-istry by major polymer producers appear to be centered on olefin polymers and co-polymers. Metallocene-derived HDPE and engineering plastics seemingly remain afuture goal, while titanate and zirconate esters appear to be efficacious to some de-gree in virtually all polymers synthesized by various routes [1]. Moreover, the ti-tanocene or zirconocene catalysts used in the synthesis of metallocene-derived poly-mers do not remain in the polymer. With “repolymerization”, thermoplastics maynow be regenerated to virgin or recycled more efficiently since the thermally stabletitanate or zirconate ester forms of the relevant organometallics “anneal” or “recon-


94

nect” polymer chain lengths that normally undergo scission during processing andremain in the polymer for subsequent repeat thermal cycles.

Table 5-3, taken from the European Patent application “repolymerization” [28],shows the beneficial effects of titanates on most properties of ABS, PC, PP, and PSthermoplastics. As an additional example, Figure 5-5 shows the “repolymerization”effect of 0.2% LICA 12 on a 50:50 blend of LDPE and PP after six thermal cyclesthrough a twin-screw extruder. The melt index of the control blend without titanateclimbs from 17 to 38, while the value for the blend with titanate is only 24, thus indi-cating a significant decrease in chain scission due to the titanate.

“Repolymerization” appears to affect the isothermal recrystallization time, chainbranching, and the morphology of the polymer chains surrounding the particulate or


25

1 2 3 4 5 6

3540

30

20

151050

Heat Cycles

Mel

t Ind

ex

LDPE/PP – 50/50

Repolymerization

Control

LICA 12

Fig. 5-5 The “repolymerization” effect of0.2% LICA 12 on a 50:50 blend of LDPE andPP after six thermal cycles through a twin-screw extruder. The control melt index increas-

es from 17 to 38. The value for the blend withtitanate increases only to 24, indicating a sig-nificant recombination of cleaved polymerchains.

Table 5-3 “Repolymerization” effect of various neoalkoxy titanatesand zirconates on the mechanical properties of unfilled ABS, PC,PP, and PS thermoplastics.

Resin Titanate/ Tensile Elong. @ Flexural Flexural Notched % H2O, zirconate, % yield, break, % strength, modulus, izod, J m–1 24 h

MPa MPa MPa

ABS Control 48.9 18 82.7 2,826 160 0.30L 44/H, 0.5 64.8 17 289 4,757 256 0.08

PC Control 66.9 65 89.6 2,275 320 0.20NZ 12/H, 0.5 70.3 73 96.5 2,344 421 0.14

PP Control 33.8 120 – 1,447 37.4 –L 12/H, 0.5 38.6 148 – 1,516 74.8 –

PS Control 35.2 10 65.5 2,551 133 –L 12/H, 0.3 40.7 51 68.3 2,551 197 –

95

fiber. Table 5-4 gives examples of filled systems for which easier processing and bet-ter mechanical properties may be attributed to this effect.

Tab. 5-4 Examples of systems with function (2) effects.

System Titanate Coupling agent effects Refs.coupling agent

Talc-filled PP LICA 12 Increase in PP MWD; increased 29dispersion, decreased melt viscosity; increased impact strength, reduced Tg

Nanoclay-filled SBS KR TTS Retention of SBS properties after 30compounding with modified filler; lower temperature of mixing and lower torque

BaSO4-filled peroxide- KR TTS Increased elongation at break; in- 31cured polybutadiene creased tensile strength

Fiber glass filled PPS Lica 09, Increased resin crystallization 1, 32CAPOW NZ 97/H temperature, decreased isothermal

crystallization; increased composite elongation at break, elimination of embrittlement

Carbon black and organic KR 46B Higher solids contents 33pigments in printing inks

Cellulose fibers in mixed CAPS L 12/L Strength increase despite little 34plastics waste evidence of bonding; possible reactive

compatibilization


Pyrophosphato and phosphato titanates (e.g., LICA 38 and LICA 12) may render Ca-CO3-filled LLDPE f lame retardant and provide synergistic effects with conventionalf lame retardants. Figure 5-6 depicts a theoretical monolayer of phosphato titanate[function (3)] coupled to a substrate. Pyrophosphato titanates are efficacious on met-al oxides such as antimony oxide, which is used in halogenated f lame-retardant sys-tems. Because of their non-toxic smoke generation, recent efforts have been directedtowards the use of ATH and Mg(OH)2. ATH must be loaded to a level of 64 wt. % ofthe total compound to generate enough steam to achieve a UL 94V-0 rating, whileMg(OH)2 is usually loaded into thermoplastics in the 40–60 wt. % range. The use oftitanates allows highly filled polyolefins to be processed at ~10% lower temperatures,thus allowing water-releasing f lame retardants to be processed more readily. Exam-ples of the synergisitic benefits of titanate on function (1) coupling, function (2) catal-ysis, and function (3) phosphato and pyrophosphato heteroatom intumescence arepresented in Table 5-5.


96

Tab. 5-5 Examples of systems with function (5) effects.

Filler Polymer Titanate Comments Refs.

ATH, Mg(OH)2 PP Phosphato and High loadings 1, 35pyrophosphate (64 wt. %) to achieve functionality f lame retardancy with-

out loss of mechanical properties

ATH, Mg(OH)2 Miscellaneous Titanates, zirconates Non-toxic f lame 36thermoplastics and retardantelastomers

ATH Dicumyl peroxide Isostearoyl titanate, Increased mechanical 37–39cross-linked LDPE, isopropoxy tris(dioctyl properties through LLDPE pyrophosphoryl) synergism of titanate

titanate and a vinyltriethoxy-silane

ATH Polyolefin mixture Phosphate Flame-proof 40functionality compositions

Mg(OH)2 Polyolefins, EPDM, Isopropyl-triiso- Fire-resistant wire 41,42Polyamide, ABS stearoyl, pyrophos- and cable and

phato moldings

CaCO3 LLDPE Pyrophosphato Self-extinguishing Author’s at 44 wt. % filler data

Highly CaCO3-filled polyolefins treated with phosphate titanate may be convertedinto fire-extinguishing compositions. Data obtained by the author on LLDPE filledwith 44 wt. % CaCO3 and treated with 3 wt. % LICA 38 demonstrated the f lame-re-tardant intumescent effect of the pyrophosphate titanate containing composition as


Fig. 5-6 A graphic representation of the monomolecular depositionof a phosphato titanate [function (3)] to create a phosphatizedflame retardant or anti-corrosive substrate.

97

compared to a titanate-free control. This system has been used as a model for prepar-ing powerful yet safe energetic composites containing high loadings of nitramine ex-plosive (up to 85%) in a cellulose acetate butyrate matrix. In a related patent [43], thephosphate titanate is claimed to permit increased nitramine loadings, improve f lowcharacteristics, and control the spread of burning propellant spalls. Similarly, in sol-id rocket fuel containing aluminum and ammonium perchlorate in a hydroxy-termi-nated polybutadiene/polyurethane binder, the use of pyrophosphato titanate wasshown to control the burning rate, in addition to providing improved dispersion ofthe fillers and improved mechanical properties of the propellant.


CaCO3 and carbon black (CB) are the two largest volume fillers consumed in ther-moplastics and elastomers. Before functions (5) and (6) are discussed, further dis-cussion of the functional effects of titanates on these two fillers and a host of otherinorganic/organic systems in thermoplastics, as noted by the author and other in-vestigators, is instructional.

5.2.4.1 CaCO3-Filled ThermoplasticsThe reactivity of titanates with CaCO3 has been discussed herein and is well-estab-lished [1]. For example, transparent polyolefin films containing titanate and 40 wt. %0.9 µm CaCO3 and many other efficacious CaCO3-filled thermoplastic compositionshave been produced by the author [1] and significant commercial applications exist.Examples of work by others in the area of CaCO3-filled polyolefins are shown inTable 5-6. Specific property data adapted from ref. [46] are shown in Table 5-7. In gen-eral, if overall polyolefin composite strength is desired, and not just filler loading,then 20–30% fine particle CaCO3 loadings using 0.5 wt. % titanate, all compoundedat ~10% lower temperatures than what would be used without titanates, is recom-mended by the author.

Tab. 5-6 Effects of titanate in CaCO3-filled polyolefins.

Polymer Titanate Comments Refs.

PP/HDPE LICA 12 Optimum mechanical properties at 0.7% titanate 44

PP LICA 12 Optimum elongation at break at 20–40 wt. % loading (0.2, 0.3 with impact strengths 35–65% greater than the unfilled and 0.4%) control 45

PP TTS Improved dispersion, higher melt index, higher tensile 46elongation, improved optical properties

PP TTS, Titanates have proved more effective in improving calcium f low properties and impact strength 47stearate


98

Tab. 5-7 Comparison of properties of compression- and injection-molded, untreated, 0.6, and 1.0% KR TTS-treated, unfilled, 40%,and 50% CaCO3-filled PP.

Serialno. Sample MFI, Tensile yield Elong. at Brittle Izod impact g/10 min strength, break, % point, °C strength

MPa notched, J/m

Compression-molded1 PP 5.61 34.0 22 32 422 PP:CaCO3 (50:50) 3.87 18.9 14 –7 393 PP:CaCO3 (50:50) 6.00 16.3 43 0 49

+ 0.6% KR TTS4 PP:CaCO3 (60:40) 6.11 20.4 50 22.5 51

+ 0.6% KR TTS5 PP:CaCO3 (60:40) 7.20 22.6 30 21.5 51

+ 1% KR TTSInjection-molded1 PP – 35.9 99 – 502 PP:CaCO3 (50:50) – 20.9 49 – 403 PP:CaCO3 (50:50) – 20.6 65 – 45

+ 0.6% KR TTS4 PP:CaCO3 (60:40) – 23.2 69 – 47

+ 0.6% KR TTS5 PP:CaCO3 (60:40) – 23.8 61 – 52

+1% KR TTS

5.2.4.2 Carbon Black (CB) Filled PolymersCarbon black (CB) is the most extensively used filler in terms of volume in thermo-plastic/thermoset elastomers. One method of quantifying carbon black dispersion isthrough the reduction in resistivity of insulating polymer matrices that occurs byvirtue of the CB’s ability to create a conductive three-dimensional particulate net-work. As the amount of carbon black is increased, the resistivity is decreased. In ad-dition, the efficiency of a fixed amount of carbon black is increased as a result of theincreased dispersion offerred by the use of titanates. As an example, Table 5-8 showsthe effect of increasing amounts of LICA 09 on the resistivity of a CB-filledstyrene/butadiene block copolymer.

Yu et al. [48] investigated carbon black filled polyolefins as positive temperature co-efficient (PTC) materials by studying the effect of coupling agent treatment, compo-sition, and processing conditions. Their data show that an 18 wt. % CB loading inLDPE, treated with a pyrophosphate titanate (KR 38S), creates a composite with a sta-ble PTC intensity, which is much lower than that achieved in the absence of titanate,thus allowing the production of novel “smart” polyolefin composites with more uni-form conductivity control. From PTC intensity vs. CB content plots, this 24-fold in-crease in PTC performance indicates that it would take about 7% additional untreat-ed CB to reach the titanate-treated CB performance level.


99

Tab. 5-8 Resistivity of 3.75% XC-72R conductive black in styrene–butadiene block copolymer/PS (10 mm thick test slab)

Wt. % LICA 09 of 3.75% XC-72R Resistivity

Surface Ω/sq. Volume Ω · cm

Control > 1016 7.8 × 1014

0.67 1.7 × 1012 3.0 × 1012

1.00 2.1 × 108 4.3 × 107

2.00 5.7 × 107 3.7 × 107

It is obvious that any thermally conductive composition can be made more efficientby increasing the loading of the thermal or electron conductive material. Thus, it isnecessary to load the polymer without losing the ability to process the compound, andthen to form a part that has suitable mechanical properties. A shift in CPVC inducedby titanates is usually predictive of an increase in a functional particulate’s perform-ance. Refs. [49,50] provide examples of thermally conductive graphite elastomer(chloroprene and EPDM) compositions incorporating titanates. The effects of pre-treating CB with KR TTS in terms of improving the low-temperature f lexibility ofbutyl rubber and the overall performance of a conductive isobutylene compound aredescribed in refs. [51,52]. It should be noted that KR TTS has been sold commercial-ly and continuously in the U.S.A. since 1974 for the masterbatching of carbon blackfor polyolefin elastomers.

5.2.4.3 Other Functional Inorganics/Organics Used in Thermoplastics and ThermosetsTable 5-9 contains examples of polymer systems incorporating other widely used in-organic fillers that are treated with titanates/zirconates. The principal effects of thecoupling agents in relation to the given application are also shown. Modifications ofTiO2 with pyrophosphato- and phosphito-coordinated titanates to produce a highlyfunctional metal oxide [54] and an acrylic colorant [55] have been described. Titanateshave also been used as surface treatment agents for metal oxides in order to impartenhanced environmental stability [66] and in organic electroluminescent devices [69]for increased brightness.


Organofunctionality can be imparted to any inorganic particulate or substrate usingsuitable titanates or zirconates. For example, a pyrophosphato titanate can be used totreat silica to convert it to an anti-corrosive pigment; an acrylic functional zirconatecan convert TiO2 to a UV-reactive pigment; or a water-insoluble pyrophosphato ti-tanate or zirconate, or blend thereof, can be reacted with a methacrylamide function-al amine to make a water-soluble, anti-corrosive, acrylic functional organometallic ad-ditive for pigments, fillers or surfaces. Such substrates would be compatible with wa-ter-based acrylics of high solids content.


100

As an example of surface functionalization, a recent patent [70] entitled “Surfacefunctionalization of pigments and/or dyes for radiation-curable printing inks and coatingsand their preparation” describes a UV-curable powder coating composition containingTiO2 treated with a zirconate incorporating a radiation-curable functional group [NZ39 (neopentyl (diallyl)oxytriacryl zirconate)]. The cured resin showed good opticalproperties and scratch resistance.


Table 5-9 Effects of titanate/zirconate in miscellaneous functionalinorganics in thermoplastics and thermosets

Filler Coupling agent Polymer Comments Refs.

TiO2 Pyrophosphato Epoxy Improved mechanical properties 53

CaSO4 TTS Acrylic acid- Use in cosmetic pastes 56styrene copolymer

CaCO3, talc, Miscellaneous PP and EPR Good processability at high 57and BaSO4 loadings (70 wt. %)

Fe2O3 KR 38S Styrene/n-butyl Improved performance vs. epoxy 58methacrylate silane in copier toner

Fe3O4 KR TTS Butyl acrylate- Use in magnetic toner 59styrene copolymer

Hydroxyapatite LICA 12, HDPE or Promote adhesion and catalysis 60, 61NZ 12 starch/EVOH and improve mechanical pro-

perties in composites for prosthetic devices

Tungsten Miscellaneous HDPE, poly- Improved conductive positive 62, 63carbide, olefin adhesive temperature coefficient devicesTitanium carbide

Saponite NDZ-201 PP Thermally resistant automotive 64applications

Barium ferrite LICA 38 Thermoplastic Improved electrical and magnetic 65natural rubber properties

Cadmium KR 138S Epoxy Resistance to ozone attack in 1 sulfide photosensitive plates (Table 34)

Clay Miscellaneous Rubber Increased ozone resistance 67

Rare earth-iron- Miscellaneous Miscellaneous Improved coupling efficiency in 68nitrogen bonded magnetspowders

101


In the early stages of the introduction of monoalkoxy titanates, hybrid titanates wereintroduced by the author, in which functions (3), (4), and (5) were intermixed by trans-esterification reactions at the Ti center of the molecules using various organic lig-ands. For example, the seventh in a series of titanates synthesized by the author wasa hybrid titanate (KR 7) consisting of two moles of methacrylic acid and one mole ofisostearic acid (see Table 5-1). The theory, borne out through commercial practice,was that the isostearoyl ligand would stabilize and protect the methacrylic ligandsfrom auto-oxidation.

5.3Summary and Conclusions

There is a wealth of other functional filler work that could be discussed, such as:wood fiber/starch composites [71], the production of dyeable polypropylene throughnanotechnology [72], and a host of fiberglass-, graphite- and aramid-reinforced ther-mosets and thermoplastics. There is also much characterization work to be done tounderstand the function (2) catalysis, “repolymerization”, copolymerization, andcross-linking effects reported in refs. [73–79], which potentially can make any fillerwith such catalytic properties more functional in plastics. The concluding discussionon thermally conductive thermoplastics highly filled with boron nitride and suitablefor electronic packaging refers to a recent patent [80] with potentially significant com-mercial value. Of more than 1837 patents and references in the literature on titanates,this may serve as a representative case history since it covers so many of the issuesraised in this chapter, specifically:

reactivity with a non-silane-reactive inorganic, in this case boron nitride; in situ application of pellet or powder masterbatches of the coupling agent at tem-

peratures in excess of 200 °C for reactive compounding in the melt; dispersion and adhesion effects, or, stated another way, coupling and catalysis ef-

fects, in relation to improved process rheology of filled polymers, increased f low ofthe polymer itself, shift in the CPVC of the filler to polymer ratio, and increasedfiller functionality such as thermal conductivity, enhanced mechanical properties,and enhanced end-product performance;

applicability to a host of polymers, processes, and equipment used in the manu-facture of thermoplastics and thermosets, particularly the ability to withstand theharsh temperature and shear conditions of commonly used high-volume melt pro-cessing such as extrusion, compounding, and injection molding.

The invention [80] relates to a thermally conductive moldable polymer blend com-prising a thermoplastic polymer such as polyethylene terephthalate (PET), polybuty-lene terephthalate (PBT), polyphenylene sulfide (PPS), or polycarbonate (PC) with a

5.3 Summary and Conclusions

102

processing temperature in the range 200–300 °C and having a tensile strength atyield of at least 70 MPa, at least 60 wt. % of a mixture of boron nitride powders hav-ing an average particle size of at least 50 µm, and a coupling agent such as a neoalkoxyor monoalkoxy titanate in a masterbatch form at a level of 0.1–5 wt. %. Such a com-position displays a thermal conductivity of at least 15 W m–1 K–1 and it is capable ofbeing molded using high-speed molding techniques such as injection molding. Thecoupling and/or dispersing agent serves to facilitate better wetting of the boron ni-tride fillers. It also helps to reduce the melt viscosity of the composition and allowsfor higher loading with the fillers. In addition, the coupling and dispersing agent mayalso improve the interfacial adhesion between the polymer and the ceramic fillersand thus provides better physical and mechanical properties.

Appendix

References


1 Monte, S. J., Ken-React® Reference Manual –Titanate, Zirconate and Aluminate CouplingAgents, 3rd Ed., Kenrich Petrochemicals,Inc., Bayonne, NJ, March 1995.

2 Monte, S. J., et al., 33rd Internat. SAMPESymposium, Anaheim, CA, March 1988.

3 Monte, S. J., Sugerman, G., 33rd Internat.SAMPE Symposium, (SAMPE II), Anaheim,CA, March 1988.

4 Monte, S. J., Sugerman, G., Corrosion ’90,Las Vegas, NV, April 1990, Paper No. 432.

5 Monte, S. J., Sugerman, G., Water-Borne &Higher Solids Coatings Symposium, New Or-leans, LA, Feb. 1988.

6 Monte, S. J., Sugerman, G., Corrosion ’88,NACE, St. Louis, MO, March 21–25, 1988.

7 Monte, S. J., Sugerman, G., Western Coat-ings Societies’ 19th Biennial Symposium andShow, Anaheim, CA, USA, March 1989.

8 Monte, S. J., Sugerman, G., 2nd Internat.Conf. on Composite Interfaces (ICCI-II), CaseWestern Reserve University, Cleveland,Ohio, USA, June 1988.

9 Monte, S. J., et al., SPI Urethane Division26th Ann. Techn. Conf., November 1981.

10 Monte, S. J., Sugerman, G., SPI UrethaneDivision 29th Ann. Techn./Marketing Conf.,October 1985.

103References

11 Monte, S. J., Proc. SPE RETEC, WhiteHaven, PA, USA, October 1995.

12 Monte, S. J., Proc. ACS Rubber Div. Conf.,Louisville, KY, USA, Oct. 1996, Paper No.57.

13 Monte, S. J., Rubber Technology International’96, UK & Int’l. Press, a Div. of Auto Inter-mediates Ltd. (1996).

14 Monte, S. J., Polyblends ’97 SPE Div./Sect.Conference, NRCC, Montreal, Canada, Oct.1997.

15 Monte, S. J., Proc. 47th SPE ANTEC, 1989,35, 866–869.

16 Monte, S. J., Plastics Compounding, Novem-ber/December 1989, p. 59–65.

17 Glaysher, W. A., et al., Proc. SPE RETEC,“High Performance Blow Molding” Conf.,Itasca, IL, USA, Oct. 1990, p. 311–335.

18 Monte, S. J., Proc. SPE Recycle RETEC,Charlotte, NC, USA, 30–31 October 1989,Paper 9.8(13).

19 Monte, S. J., Northeast Regional Rubber &Plastics Exposition, September 1994, Mah-wah, NJ, USA.

20 Monte, S. J., Sugerman, G., Compalloy ‘90,Mar. 1990, New Orleans, LA, USA.

21 Monte, S. J., RAPRA Addcon World 2001Conference, October 2001, Berlin, Germany.

22 Plueddemann, E. P., Silane Coupling Agents,Plenum Press, New York, 1982, p. 114.

23 Voelkel, A., Grzeskowiak, T., Chro-matographia 2000, 51(9/10), 606–614.

24 Yamazaki, A., JP 2001288440, KyoritsuChemical Industry Co., Ltd., 2001.

25 Kataoka, M., JP 2002025806 A2 20020125,Tokin Corp., 2002.

26 Krysztafkiewicz, A., et al., J. Mater. Sci.1997, 32, 1333–1339.

27 Wang, Y., Huang, J.-S., J. Appl. Polym. Sci.1996, 60, 1779–1791.

28 Monte, S. J., Sugerman, G., U.S. 4,657,988,Kenrich Petrochemicals, Inc., 1987; EP Ap-pl. 87301634.9-2109, Pub. No. 0 240 137filed 25.02.87, issued 1998.

29 Wah, C. A., et al., Eur. Polym. J. 2000, 36,789–801.

30 Galanti, A., et al., Kautsch. Gummi Kunstst.1999, 52(1), 21–25.

31 Simonutti, F. M., JP 10108925, WilsonSporting Goods Co., 1998.

32 Chen, C.-H., et al., Industrial TechnologyRes. Inst., Hsinchu, Taiwan, U.S. 5,340,861,1994.

33 Fukae, K., Yoshida, I., Advance Color Tech-nology, Inc., U.S. 6,132,922, 2000.

34 Miller, N. A., et al., Polymers & Polymer Com-posites 1998, 6(2), 97–102.

35 Kato, H., et al., Mitsubishi Cable Ind., U.S.4,769,179, 1988.

36 Eichler, H.-J., et al., Alusuisse MartinswerkGmbH, WO 00015710, 1999.

37 Jiang, P., et al., Hecheng Shuzhi Ji Suliao2001, 18(5), 35–38.

38 Wang, G., et al., J. Appl. Polym. Sci. 2002,85(12), 2485–2490.

39 Wang, G. L., et al., Chin. J. Polym. Sci. 2002,20(3), 253–259.

40 Braga, V., et al., Basell Technology Compa-ny BV, Neth., WO 2001048075, EP 1155080,2001.

41 Imahashi, T., Kazuki, K., Kuowa KagakuKogyo K.K., JP 2001312925, 2001.

42 Chiang, W.-Y., Hu, C.-H., Composites, Part A2001, 32A(3-4), 517–524.

43 Monte, S. J., Sugerman, G., Kenrich Petro-chemicals, Inc., U.S. 6,197,135, 2001.

44 Ichazo, M. N., et al., Proc. 57th SPEANTEC, 1999, 45(3), 3900–3902.

45 Doufnoune, R., Haddaoui, M., UniversitéFERHAT-Abbas, Institut de Chimie-Indus-trielle, LPCHP, Sétif, Algérie, Private Com-munication, February 1999.

46 Sharma, Y. N., et al., J. Appl. Polym. Sci.1982, 27, 97–104.

47 Szijártó, K., Kiss, P., Polymer Composites,“Filling of Polymers with the Aid of Cou-pling Agents”, Walter de Gruyter & Co.,Berlin & New York, 1986.

48 Yu, G., et al., J. Appl. Polym. Sci. 1998, 70,559–566.

49 Ogino, M., Bando Chemical Industries,Ltd., JP 03070754, 1991.

50 Hatanaka, T., et al., Uchiyama Manufactur-ing Corp., JP 2002003670, 2002.

51 Kudo, M., et al., Denso Corporation, WO2001092411, 2001.

52 Manabe, T., et al., Kanegafuchi ChemicalIndustry Co., Ltd., JP 2001247732, 2001.

53 Kuwabara, M., Mater. Lett. 1996, 2(6),299–303.

54 Murakata, T., et al., J. Chem. Eng. Jpn. 1998,31(1), 21–28.

55 Koike, Y., Mano, S., JP 10315247, TsuyakkuK. K., 1998.

56 Bodelin-Lecomte, S., Le Gars, G., Fr., EP864322, L’Oreal, 1998.

104 5 Titanate Coupling Agents

57 Ren, Z., et al., Polymers & Polymer Compos-ites 2002, 10(2), 173–181.

58 Young, E. F., O’Keefe, D. J., Xerox Corp.,U.S. Patent 5,489,497, 1996.

59 Kozawa, M., et al., Toda Kogyo Corporation,EP 794154, 1997.

60 Sousa, R. A., et al., Proc. 59th SPE ANTEC,2001, 47, 2550–2554.

61 Vaz, C. M., et al., Biomaterials 2002, 23(2),629–635.

62 Horibe, H., et al., U.S. Appl. 2002137831,09-26-2002.

63 Kataoka, M., NEC Tokin Corp., JP2002226601, 2002.

64 Yang, H., Feijinshukuang Bianjibu Journal2001, 24(2), 24–36.

65 Ahmad, S., et al., Sci. Int. (Lahore) 1998,10(4), 375–377.

66 Uchida, N., et al., Toda Kogyo Corporation,Eur. Pat. Appl., EP 1225600, 2002.

67 Chen, M., et al., J. Appl. Polym. Sci. 2001,82(2), 338–342.

68 Imaoka, N., et al., Asahi Chemical IndustryCo., Ltd., JP 3217057, 2001.

69 Kijima, Y., Sony Corp., U.S. 6312837, 2001.70 Wang, Z., Wu, B., UCB, S.A., Belg., WO

2002048272, 2002.

71 Liu, Z. Q., et al., “Wood Fibre/Starch Com-posites: Effects of Processing and Com-pounding”, Proc. 16th Annual Meeting of thePolym. Processing Soc., Shanghai, China,June 18–21, 2000.

72 Fan, Q., et al., “Dyeable Polypropylene viaNanotechnology”, NTC Project: C01-MD20(formerly C01-D20), National Textile CenterAnn. Report, UMass, Dartmouth, Nov.2002.

73 Kim, C. Y., et al., Korean Institute of Sci-ence and Technology, U.S. 5,237,042, 1993.

74 Kim, C. Y., et al., U.S. 5,714,570, 1998.75 Cho, H. N., et al., U.S. 6,040,417, 2000.76 Lee, S.-S., et al., J. Polym. Sci., Part B: Poly-

mer Physics 2001, 39(21), 2589–2597.77 Kitani, I., et al., Lion Corp., JP 2001232210,

2001.78 Kelley, D. W., U.S. Patent 4,837,272, 1989.79 Schut, J., “Organometallic Esters Enhance

Recycled PET/PC Blends”, Plastics Formu-lating and Compounding, Sept./Oct., 1996.

80 Zhuo, Q., et al., Ferro Corporation, WO00/42098, 2000.

6Functional Polymers and Other Modif iers

Roger N. Rothon

6.1Introduction

Several types of additives are used for modifying the interaction of particulate and fi-brous fillers with polymer matrices. The use of organofunctional silanes and ti-tanates has been described in Chapters 4 and 5. This section covers the other mainapproaches that can be used. For the purposes of the discussion, these additives arereferred to as non-coupling and coupling. Both types have advantages and limita-tions, and the differences between them are described below. Treatment of the non-coupling types is more extensive than in many other works. This is for two reasons.First, they are very important commercially. Secondly, they provide considerable in-sight into basic issues such as the mechanism of reaction/interaction of groups suchas acids and acid anhydrides with filler surfaces; this has not been so well docu-mented for the coupling agent types covered in this section. While division intocoupling and non-coupling types is often based on chemical structures, this can bemisleading. As will be shown later, some modifiers can act as either non-couplingor as coupling types, depending on the formulation. The concept of reinforcementpromoters, first introduced by Ancker and co-workers [1], and discussed later(Section 6.2.3), is a potentially useful alternative way of classifying modifiers in somecases.

The coverage has largely been restricted to additives that can be considered to havesome form of chemical attachment to the filler surface. This eliminates species suchas glycols and some surfactants, which are only physically adsorbed. There is still agray area, however, as some additives can function as polymer modifiers in their ownright, but may also react with a filler surface. Where such filler attachment is thoughtto be likely, and important to the effects observed, then such materials (e.g. bis-maleimides) have been included in this chapter.


106

6.2General Types of Modif iers and their Principal Effects

6.2.1Non-Coupling Modif iers

These additives form a strong, essentially permanent, attachment to the filler surface,but only weakly interact with the polymer phase. They generally contain a filler reac-tive group at the end of a linear hydrocarbon chain, and the main effects usually ob-served are:

reduced filler surface polarity; reduced water adsorption by the filler; improved processing characteristics (faster incorporation, lower viscosity, reduced

energy consumption, etc.; this can lead to less polymer degradation in some com-pounding operations);

reduced tendency of the filler to adsorb important compound additives, such asanti-oxidants and curatives, thus improving their efficiency;

improved filler dispersion; improved surface finish of the final product; low polymer/filler interaction, often resulting in an increase in stress whitening

and a decrease in strength; increased elongation at break and improved impact strength in thermoplastics; reduced water adsorption by the composite, reducing swelling and stabilizing elec-

trical properties; changes in the nucleating effect of filler surfaces in some semi-crystalline thermo-

plastics, notably polypropylene.

6.2.2Coupling Modif iers

These form strong bonds to the filler and the polymer matrix, thus tying, or coupling,the two together. The coupling effects are superimposed on non-coupling ones duesolely to changes in the nature of the filler surface.

The main effects of coupling are on compound properties, notably:

strong filler/polymer adhesion; reduced stress whitening and increased strength; reduced elongation at break; impact strength, which is often improved, but can be unaffected, or even reduced

in some instances; reduced water adsorption and increased property retention under humid condi-

tions.

6 Functional Polymers and Other Modif iers

107

The effects on surface polarity vary considerably, largely due to the nature of thepolymer reactive groups, some of which, such as amino, can be quite polar them-selves. The effects on processing are less clear-cut than with the non-coupling types.They usually depend on structural features other than coupling itself and can also beaffected by at what point in the process the coupling is established (i.e. during com-pounding, or extrusion/injection molding). Strong coupling on its own would be ex-pected to increase melt viscosity and adversely affect processing. While this can be ob-served, it is often masked by other effects. There is also the potential for changes infiller nucleating effects similar to those mentioned for the non-coupling modifiers.

6.2.3Reinforcement Promoters

This is an alternative way of describing the effects of surface modification, especial-ly in thermoplastic systems. It was developed by Ancker and co-workers [1] and refersto modifiers that increase both strength and toughness. While most classical cou-pling agents increase strength, not all increase toughness. Although it is little usedtoday, the Ancker approach is useful, as it is independent of modifier structure ormechanism of action. It will be discussed further in Section 6.3.2.1

6.3Modif iers by Chemical Type

6.3.1Carboxylic Acids and Related Compounds

6.3.1.1 GeneralThe carboxylic acid group is widely used to attach organic species to filler surfaces.Anchoring to the filler is thought to proceed by salt formation with mineral fillers, oresterification with organic fillers such as wood products. The first type of attachmentis effective on fillers with basic and amphoteric surfaces, such as carbonates and hy-droxides, but not so useful on acidic surfaces such as found on silicas and silicates.

6.3.1.2 Saturated Monocarboxylic Acids (Fatty Acids) and their SaltsSurface modification based on the use of fatty acids is the classic non-coupling ap-proach and is widely used commercially. Fatty acids are saturated monocarboxylicacids with the general formula CnH2n+1COOH. The carbon chain can be linear orbranched. Their name derives from the occurrence of some of the higher members,notably stearic acid, in natural fats. They are mainly obtained from natural productsources, with the products containing an even number of carbon atoms being muchmore abundant than those with odd numbers.

Before starting a detailed discussion, some words have to be said about the ap-proach taken here, especially with respect to the fatty acid salts. There is considerablescope for confusion and it has to be admitted that the situation is quite complex and

6.3 Modif iers by Chemical Type

108

by no means clearly resolved at present. The problems stem from the fact that saltssuch as calcium stearate are frequently used as additives in their own right, and caninf luence compound properties without having any filler surface effects. They are al-so often attracted to filler surfaces and may be formed when fatty acids react withfiller surfaces. It is thus almost impossible to separate out the effects of surface andpolymer modification, especially as filler surface treatments based on fatty acids maysplit off salts into the polymer phase, while salts initially in the polymer phase maybecome attached to the filler during processing. For consistency, the approach takenhere is to discuss these additives in terms of filler surface attachment, but it is by nomeans clear that this is necessary for good effects to be obtained with fatty (and oth-er carboxylic) acids and their salts [2].

Methods of ApplicationThe method of application of the fatty acid coatings can have a large inf luence ontheir structure and distribution. There are two methods in commercial use, namelydry and wet coating.

Dry CoatingIn this method, the fatty acid is added to the filler while it is maintained in a dispersedstate, usually by high shear mixing. With the higher MW acids, it is essential for themix to reach the melting point of the fatty acid, if true coating, as opposed to admix-ture, is to occur. Heat is also often necessary to drive off water formed by the reactionand to ensure that all acid is converted to a salt form. In some equipment, the mix-ing procedure will produce sufficient heat, while in others external heating needs tobe applied. In some cases, the fatty acid is dissolved in a small amount of solvent toaid the process. The conditions have to be carefully controlled if a tightly bonded sur-face layer is to be produced and free salt and residual acid are to be minimized. Freesalt formation is favored by high local concentrations of acid, high additive acidity,and high reactivity of the filler surface. Thus, it is most prevalent when trying to coatwith low MW acids and with fillers such as magnesium hydroxide. In some in-stances, a fatty acid salt is used as the coating agent, instead of the free acid, althoughit is by no means clear as to whether this can give rise to strongly bound surfacespecies.

Wet CoatingCommercially, wet coating is usually carried out using an aqueous solution of a suit-able salt of the acid. The sodium salt is most frequently employed, although this leadsto some residual sodium in the product, which can be a problem if water adsorptionand electrical stability are important. Ammonium salts have the advantage of notleaving any cationic residue, the ammonia being driven off either at the coating stageor during drying. The release of ammonia can, however, present handling problems.Again, care has to be taken if free salt formation is to be avoided.


109

Types of Acids and their EffectsThe fatty acids most commonly used for filler treatment are those with linear chainsand with at least 14 carbon atoms, with stearic acid, C17H35COOH, being the mostcommon. For cost purposes, blends rather than single acids are usually employed.These blends can contain significant amounts of unsaturated acids as well as non-acid material and have to be chosen with due care and attention to the requirementsof the final application. Unsaturated components can adversely affect the color of thefinal compound, with yellowing often being observed. Odor can also be an importantconsideration and is again affected by minor components. Significant levels of hy-droperoxides can also be present in fatty acids, or in the coatings derived from them,and can adversely affect compound stability. Organic acids have a strong tendency todimerize, through hydrogen bonding between carboxyl groups, and this can inf lu-ence the adsorption and reaction of fatty acids with filler surfaces.

Properly applied, fatty acids provide the filler with a hydrocarbon-like surface,which is much less polar than the filler itself. As an example, the treatment of a pre-cipitated calcium carbonate with a fatty acid coating was found to reduce the disper-sive component of surface energy from 54 to 23 mJ m–2 [3]. As a result, the filler ismade more compatible with many polymer types, with benefits such as faster incor-poration and mixing, better dispersion, less energy consumption, lower viscosity, andeasier extrusion. The filler generally also has lower adsorbed water content. Many ofthese points are demonstrated by the observed lower overall Brabender torque datafor coated versus uncoated calcium carbonate during mixing with polypropylene.

The effects of fatty acid treatments on composite properties are generally those setout above for non-coupling modifiers. The low degree of interaction between thecoated filler and polymer generally results in voiding at relatively low strains. As a re-sult, tensile strength is usually decreased, while elongation and, sometimes, impactstrength can be increased. Indeed, one of the reasons for the widespread use of fattyacid coated calcium carbonate in homopolymer polypropylene is the high impactstrength that can be achieved. The magnitude of the fatty acid effects can vary con-siderably, and is minimized in the presence of significant amounts of similar surface-active species that are naturally present in some polymers. The most extreme case isprobably found in elastomers such as SBR, where much larger effects are observedin solution-derived polymers than in emulsion ones, as the latter contain high levelsof surfactants [4].

Structure of Surface LayersThe adsorption of fatty acids onto polar surfaces has been widely studied, and usual-ly results in a layer with the carboxyl group at the surface and the hydrocarbon chainsoriented vertically to the surface. If the surface is microscopically smooth enough,and the chains sufficiently long, then the layer can be considered as semi-crystalline.When packed in this way, the area occupied at the surface by one molecule of a satu-rated, linear, fatty acid is about 0.21 nm2. Fatty acids with branched chains, and thosecontaining unsaturation, do not allow such close packing and hence occupy larger ar-eas. Similar layers are believed to form on mineral fillers, but with the acid group con-verted to carboxylate. The conversion of isostearic acid to its carboxylate is illustrated


110

by the FT-IR data presented in Figure 6-1. Monolayer adsorption levels often corre-late well with those predicted based on the known surface area of the filler and thearea occupied by one molecule of surfactant.

Changes in composite properties often peak at about the monolayer level, whencoatings are properly applied. Thus, impact strength in calcium carbonate filled PPhomopolymer appears to be maximal at the monolayer level [5]. Rothon and co-work-ers have provided details of the effect of stearic acid derived coatings on magnesiumhydroxide in an ethylene–vinyl acetate copolymer (18% vinyl acetate) [6]. Elongationat break was found to increase and tensile strength and secant modulus to decreaseuntil the monolayer level was reached. Interestingly, excess coating above a mono-layer was found to increase the ageing rate of the composite (loss of properties atroom or elevated temperature), with a commercial fatty acid blend having a greaterdetrimental effect than a pure acid.

Despite its obvious interest, little has been published regarding the effect of fattyacid chain length on filler performance. As mentioned above, blends approximating


Fig. 6-1 DRIFTS spectra (samples diluted to5 wt. % in finely ground KBr) showing the in-teraction of isostearic acid with aluminum hy-droxide; (a) aluminum hydroxide (Alcan SF11-E) treated with ca. 0.7 monolayers ofisostearic acid, (b) untreated aluminum hy-droxide reference substrate, and (c) substratesubtracted spectrum (i.e., ((a) – (b)) multi-

plied by 5 to enhance peaks) showing: (i) C–Hstretching bands (2960–2850 cm–1), (ii) asym-metric carboxylate carbonyl stretching band(1570 cm–1), (iii) symmetric carboxylate car-bonyl stretching band (1416 cm–1), and (iv)methylenic C–H deformation band (1460cm–1). [Courtesy of Dr. C. M. Liauw, Manches-ter Metropolitan University, Manchester, UK].

111

to stearic acid (C18) are most common and one would intuitively expect that quitelong chains would be needed to increase surface hydrophobicity, considerably longerthan those found in most coupling agents. It is not clear as to whether such longchains are essential, or are merely used due to the relative availability and cost of thestearic acid type compositions. While it has been reported that chain lengths of about14 carbons are necessary in order for crystallization to occur, the data in Figure 6-2from work by DeArmitt and Breese show that good dispersion and reduced viscositycan be achieved with very short chains (2–3 carbons) [7]. Howarth and co-workershave reported chain length effects in polypropylene that seem to vary with the type offiller [8,9]. Thus, they found no difference between C10 and C22 fatty acid treatmenton calcium carbonate, but reported significant differences related to chain length inthe same range when working with magnesium hydroxide. In the latter case, strengthand stiffness were found to increase, but toughness to decrease, with increasingchain length.

The work of Tabtiang and Venables, although using unsaturated acids, appears tobe relevant [10]. Under conditions where coupling did not appear to be a factor, theirresults indicated that a chain length of C10 was sufficient to lead to significant im-provements in ductility of a calcium carbonate/polypropylene composite, as deter-mined by elongation and impact properties. They found no marked additional effectup to chain lengths of C20.

The thermal stability of the coating is little discussed, but is of potential interestwhen the filler is to be processed at high temperatures. Fatty acid surface treatments


IE+3

IE+2

IE+1

IE+0

IE+10 2 4

# Carbons in acid chain

8 10 12

1 Hz

2 Hz

3 Hz

6

Fig. 6-2 The effect of organic acid carbon chain length on theviscosity of calcium carbonate slurries in squalane (a model hydro-carbon f luid) measured at different frequencies. (Reproduced withpermission from ref. [7]).

112

would be expected to start to undergo oxidation below 200 °C. They are more stablein the absence of air, when decomposition probably starts in the range 250–300 °Cand is quite complex, involving decarboxylation reactions. Stability will vary with thepurity of the fatty acid source and the nature of the filler surface, with strongly basicsurfaces tending to promote decarboxylation. It has been reported that magnesiumstearate is less stable than calcium stearate. It has also been reported that coatings de-rived from fatty acids can destabilize some fillers; thus, stearate-coating of magne-sium hydroxide can lower its decomposition temperature [11].

The structure of the surface layers of fatty acid treated fillers can be quite complex.In the ideal case, one would expect a single layer with all of the acid groups in the car-boxylate form and interacting with a basic surface site (i.e., the coating to be in theform of a partial or half-salt, such as MII(OH)COOR). In reality, a variety of otherspecies can be present, including unreacted acid and fatty acid salt not attached to asurface site. In addition, there may be a separate fatty acid/salt phase in the bulk poly-mer, not associated with the filler surface. The relative importance of these will varywith the application. When multi-layers are present, they will be only weakly attachedto the surface layer.

Several studies have been made using organic solutions of fatty acids to treat fillersurfaces. Although this approach is generally not commercially viable, useful infor-mation can be obtained, provided that care is taken in the interpretation. Of interestare the data for the adsorption of stearic and isostearic acids from heptane onto mag-nesium hydroxide reported by Liauw and co-workers [12]. Isotherms indicated amonolayer coverage close to the theoretical one for isostearic acid, but significantlyabove the theoretical coverage for stearic acid. Analysis of the supernatant liquorsshowed that a considerable amount of salt and no free acid was present in the case ofisostearic acid, while the non-adsorbed material from stearic acid was all in the formof unreacted acid. These results were interpreted as showing that both the surface-bonded half-salt Mg(OH)COOR and non-bonded full-salt Mg(RCOO)2 were formed.In the case of isostearic acid, the full-salt is soluble enough to be removed, while it re-mains in the surface layers when stearic acid is used, thus contributing to the appar-ently anomalous monolayer level.

The mechanism of coating from aqueous solution has not been reported in any de-tail. Suess has made some measurements using 14C-labeled acid [13]. As in the caseof adsorption from an organic solvent, a Langmuir-type isotherm showing monolay-er adsorption was observed, albeit at a much lower level. This was postulated to bedue to adsorption as a hydrated complex salt. Presumably, this is converted to the sta-ble coating layer during drying. In many instances, the aqueous phase will containdissolved ions from the manufacturing process and, in such cases, the author has ob-served that the first stage can be precipitation of some form of fatty acid salt, whichsubsequently rearranges to form the surface coating.

The only nonlinear fatty acid to receive any attention has been isostearic acid. Somepatent literature describes the use of such acids, especially for treating ATH [14]. Themolecular footprint area for isostearic acid has been determined to be about0.45 nm2, considerably larger than that for stearic acid. This is because of its branchednature.


113

6.3.1.3 Unsaturated Carboxylic Acids and Related CompoundsThe introduction of a reactive functionality into the polymer, such as unsaturation, of-fers the potential to use organic acids as coupling agents, but mixed results have beenreported. A number of potentially suitable unsaturated products exist, notably male-ic acid and its anhydride, acrylic and methacrylic acids, and unsaturated analogues offatty acids such as oleic acid. The acidities and double-bond reactivities of these com-pounds vary widely depending on their structures, and this probably accounts for themarked differences in their performances. Compounds such as acrylic acid have bothhigh acidity and high double-bond reactivity, due to the proximity of the carbonylgroup to the double bond.

It has been demonstrated that effects consistent with coupling can be achieved withat least some of these additives, although not all structures are effective. It seems thatthe strength of the acid group and the reactivity and accessibility of the unsaturationare important factors.

Although not widely studied, the simplest case is that of peroxide-cured or cross-linked systems, where unsaturation can take part in the curing reaction. Mori has re-ported very good results with some unsaturated carboxylic acids (e.g.H2C=C(CH3)COOC2H4COOH) in a calcium carbonate (filler content not given)filled unsaturated polyester resin [15]. Some of the results are presented in Table 6-1.

Tab. 6-1 Effects of an unsaturated acid coupling agent[a] on themechanical properties of calcium carbonate filled unsaturatedpolyester resin [15].

Filler treatment Tensile Tensile Flexural Flexural Izod impact strength, modulus, strength, modulus, strength, MPa GPa MPa GPa J cm–1

none 18.6 7.9 44.1 7.3 0.8unsaturated acid 36.3 10.3 96.0 9.7 1.6

[a] H2C=C(CH3)COOC2H4COOH.

The data in Table 6-2 show how a commercial form of unsaturated acid improvesprocessability and increases mechanical properties in a peroxide-crosslinked ATH-filled EVA system [16] (Solplus C800 from Lubrizol is an oligomeric material con-taining a carboxylic acid group and an activated double bond).

Most interest in unsaturated additives has been in relation to their use in filledpolyolefins such as polyethylene, polypropylene, and copolymers thereof. In the ab-sence of peroxide-type free radical sources, one needs to rely on what are known asmechano-chemical processes to bring about reaction. Polymer free radicals are gen-erated by shear forces during compounding and can graft onto suitable unsaturation.These shear processes are greatest in the vicinity of the filler particles, which aids theprocess. While mechano-chemical grafting can be effective, it is very dependent onthe processing conditions and can be inhibited by the levels of stabilizer used in manycommercial formulations.


114

Some of the basic effects obtained with ATH-filled EVA in the absence of peroxideare detailed in Tables 6-3 and 6-4 [16].

The effect of anti-oxidant interference and the use of small amounts of peroxide toovercome it are illustrated in Table 6-4.

Tab. 6-2 Effect of an oligomeric unsaturated organic acid[a] in across-linked ATH/EVA formulation[b].

Additive Brabender mixing Tensile strength, MPa Elongation (%)f inal torque, Nm

none 32 8.6 1050.8 % organic acid 28 13.9 180

[a] SOLPLUS C800, Lubrizol Additives.[b] Brabender compounding followed by compression molding,

60 wt. % ATH (Superfine 7, Alcan Chemicals), in EVA (1020 VN5,Elf Atochem).

Tab. 6-3 A comparison of various unsaturated acids and threeorganofunctional silane coupling agents as surface modifiers for60 wt. % ATH-filled EVA[a].

Additive and wt. % level on f iller Final torque, Nm Tensile strength, Elongation at MPa break, %

none 35.2 10.7 59methacrylic acid (0.5) 24.3 9.2 100SOLPLUS C800 (0.8) 23.8 13.7 1794-allyloxybenzoic acid (1.0) 28.3 8.8 49sorbic acid (0.6) 25.8 10.0 48A-172 vinylsilane (0.6) 26.5 10.7 137A-174 methacryloxysilane (0.8) 24.9 10.2 150A-1100 aminosilane (0.8) 27.7 13.4 168

[a] Brabender compounding, 180 °C initial temperature, followed bycompression molding; source of SOLPLUS C800, ATH, and EVA asin Table 6-2; source of silanes: GE Specialties.

As already referred to, the approach of utilizing unsaturated monomers, with orwithout peroxides, has a long history. In one of the earliest efforts, Bixler [17] used acoating based on a doubly unsaturated molecule such as butylene glycol dimethacry-late, an unsaturated anchoring molecule such as maleic acid, and peroxide (in effectforming a polymeric coating with unsaturation and acid groups at the surface). He re-ported good improvements with fillers such as clay and calcium carbonate in poly-ethylene.

Gaylord evaluated various clays in polyethylene using peroxide/maleic anhydridetreatments [18]. He found very variable results, largely inf luenced by the nature of theclay surface, and concluded that some filler surfaces could actually inhibit the free-radical grafting processes.


115

Aishima et al. used a variety of unsaturated acids in pre-treatments for silicatefillers in PE and PP [19] and incorporated a polymerization inhibitor to prevent lossof reactivity during the filler coating process along with a small amount of peroxide.Improvements in both strength and toughness were reported (Table 6-5). They alsofound improvements in dynamic fatigue resistance. Fatigue is a property that is notoften mentioned, but is of considerable practical importance in some applications.

Tab. 6-5 The effect of various unsaturated acids on the propertiesof 50 wt. % nepheline syenite (an anhydrous alkali aluminosilicate)in HDPE [19].

Filler treatment Tensile strength, Elongation at Flexural Notched Izod MPa break, % strength, MPa impact strength,

J cm–1

none 21.6 5 24.5 0.3acrylic acid 35.3 40 45.1 4.5methacrylic acid 30.4 11 34.3 1.5crotonic acid 26.5 9 33.3 1.1sorbic acid 29.4 16 36.3 1.6maleic acid 26.5 8 33.3 0.9itaconic acid 25.5 9 32.3 1.0

In some cases, additives with multiple unsaturation may be employed. These cancross-link and strengthen the polymer, especially in the vicinity of the filler particleswhere most polymer free radicals are generated. If cross-linking is concentratedaround the particles, then processability should not be adversely affected. The con-cept of Ancker et al. of reinforcement promoters for filled thermoplastics, as opposedto coupling agents [1], involves the use of such agents in peroxide- and antioxidant-free systems. They argued that improvement in both strength and toughness was notan inevitable consequence of the simple coupling of the filler to the polymer and that


Tab. 6-4 The effect of anti-oxidant[a] and peroxide on the per-formance of an unsaturated acid[b] in a 60 wt. % ATH-filled EVA[c].

Additive and wt. % level on f iller Final torque, Nm Tensile strength, Elongation at MPa break, %

none 35.2 10.7 59unsaturated acid (0.8) 24.6 13.2 104unsaturated acid (0.8)anti-oxidant (0.3) 24.3 10.6 82unsaturated acid (0.8)anti-oxidant (0.3)tert-butyl perbenzoate (0.01) 30.1 14.0 206

a Irganox 1010, Ciba.b Solplus C800, Lubrizol Additives.c Brabender compounding, followed by compres-

sion molding; initial temperature 180 °C except

when peroxide was present, then 160 °C; sourceof ATH and EVA as in Table 6-2.

116

modifiers that actually achieved this should be called reinforcement promoters. An-cker et al. speculated that such reinforcement promoters function by forming a grad-ed interface between the filler and polymer. A way of quantifying the promoting abil-ity of additives was developed, based on the number and reactivity of the doublebonds present and on the affinity of the additive for the mineral filler surface. Goodcorrelation with experiment was found; some of the results are given in Table 6-6. Itshould be stressed that all of this work is based on peroxide- and anti-oxidant-free sys-tems, and the authors admit that anti-oxidants can have very adverse effects, even onthe best additives. Most of the additives (for example, the triacrylates) identified as re-inforcement promoters do not appear to have any obvious means of chemically at-taching to a filler surface. They may just be strongly physically adsorbed, or there maybe some ester hydrolysis forming an unsaturated acid.

Tab. 6-6 Correlation of properties of filler surface modified ATHin HDPE (50 wt. %) with Ancker’s promotion index [1].

Agent Promotion- Tensile Elongation Notched Izod index strength, at break, % impact strength,

MPa J cm–1

none – 23.5 4.4 0.9trimethylolpropane triacrylate 3.0 38.6 16.3 1.2pentaerythritol triacrylate 4.0 35.0 35.0 3.1diethylene glycol diacrylate –0.1 25.6 12.6 0.8glycerol monoacrylate –2.5 24.9 10.0 0.8

More recently, Tabtiang and Venables [20] studied the reaction of acrylic acid withthe surface of calcium carbonate. The coating was carried out by means of a drymethod and the results further demonstrated the complexities that can be encoun-tered. IR spectrometric analysis of the coated fillers showed that all of the acid wasconverted to the salt form, even when amounts well in excess of that required for amonolayer were used. However, the authors found no evidence of monolayer ad-sorption, with about half of the added material becoming insoluble in xylene, a goodsolvent for calcium acrylate, at all levels of addition. They concluded that, with themethod of coating employed, small droplets react with the filler to produce calciumacrylate, some of which is bound to the filler, or insolubilized in some way, while therest remains as the free salt.

The same workers also examined the effects of the acrylic acid coated filler inpolypropylene. In order to detect grafting of the filler to the polymer they subjectedthe sample to extraction with hot xylene and then to treatment with aqueous acid.This method showed little or no grafting for coatings produced from acrylic acidalone, but extensive grafting was found when peroxide was incorporated into the coat-ing.

The situation with unsaturated long-chain acids such as oleic acid is far from clear.Good results, especially for impact strength, have been reported for such coatings onmagnesium hydroxide in polypropylene [21], although the observed effects are not


117

necessarily those of classical coupling. In support of this, commercial products withsuch coatings appear to be available, although the present author and others havebeen unable to substantiate these claims. According to Ferrigno [22], it is necessaryto stabilize unsaturated acids with anti-oxidants if they are to be successfully em-ployed.

Tabtiang and Venables also examined several long-chain unsaturated acids, withand without peroxide [10]. The position of the double bond, and hence its accessibil-ity and reactivity, varied significantly, thus complicating interpretation. With long-chain acids (> C10), there was little evidence of grafting to polypropylene, even withperoxide present. This may have been due to steric factors, however, as the doublebonds were not at the end of the chain. With 10-undecanoic acid, containing a termi-nal double bond, there was some evidence of grafting, although to a much lesser ex-tent than that observed for acrylic acid. The double bond in acrylic acid is much morereactive, due to its proximity to the carboxyl group. All of the acids studied, exceptacrylic, were found to significantly increase the impact strength, compared to the un-coated filler, when no peroxide was used in the coating. In line with its effect on graft-ing, peroxide had no effect on the longer chain acids, but did markedly reduce the im-pact strength obtained with 10-undecanoic acid.

Acids derived from rosin are sometimes used. These are mostly a mixture of abi-etic and pimaric acids, which are three-ring, cyclic, aliphatic compounds incorporat-ing an acid group and some unsaturation, either in the rings themselves or pendantto them. They have been found by the author to give better results than stearic acidwhen used to treat precipitated calcium carbonates for use in EPDM compounds andthere has reputedly been some commercialization of this application in Japan. Theauthor has, however, not found any benefits from their use with fillers in polypropy-lene.

6.3.1.4 Carboxylic Acid AnhydridesCyclic anhydrides, such as maleic and succinic anhydrides, appear to react readilywith filler surfaces. This reaction has not been studied in detail, but appears to in-volve acid formation through hydrolysis, followed by salt formation. Most interest hasbeen in anhydrides possessing some unsaturation, such as maleic anhydride, whichhave the potential to act as coupling agents. Some of the work with maleic anhydridehas already been mentioned [17,18]. Modeling shows that the area occupied at thefiller surface is about 0.32 nm2 for the anhydride and 0.45 nm2 for the diacid result-ing from hydrolysis [23]. It is not clear as to whether the reaction with the filler sur-face involves one or both of the acid groups.

Although there has been some interest in maleic anhydride itself, most interest hasbeen in alkyl- and especially in polymer-substituted succinic anhydrides, which arediscussed in succeeding sections. Certain long-chain derivatives of succinic anhy-dride have been examined as possible filler surface modifiers, notably dodecyl anddodecenyl succinic anhydrides. The first is a potential non-coupling modifier, whilethe second, having some unsaturation, is a potential coupling agent. Because of therelatively large area occupied at the surface by the diacid, the organic tails cannot packas tightly as in the case of monocarboxylic acids.


118

Although significant effects have been observed in the work carried out to date,both types of modifier have only given results typical of non-coupling additives, suchas stearic acid. There has been little evidence of coupling ability in the case of the un-saturated analogue, even when it was used in conjunction with peroxide [10]. This ispresumably due to steric factors.

Liauw and co-workers [23] found that when dodecenyl succinic anhydride was ad-sorbed onto ATH at 50 °C, most of the interaction was through acid groups formedby hydrolysis and there was little salt formation, with carboxylate salts only appearingat higher temperatures. Monolayer coverage levels were determined by a number ofmethods. Those based on effects such as slurry viscosity were found to correspondreasonably well with that predicted from the footprint area, while that determined byextraction was somewhat higher, presumably due to some multilayer formation.

Tabtiang and Venables studied the reaction of dodecenyl succinic anhydride withcalcium carbonate using a dry blending method [10]. They found that the anhydridewas largely converted to the carboxylate salt, with little or no evidence of any remain-ing anhydride or free acid. The effects in polypropylene were those associated with anon-coupling additive, even when peroxide was present in the coating.

6.3.1.5 BismaleimidesThere has been some interest in products of this type, which were mentioned inpatents as long ago as 1973 [24] and identified by Ancker as having high reinforce-ment indices [1]. They can be thought of as derivatives of maleic anhydride, producedby its reaction with aromatic or aliphatic diamines, and contain two very reactive dou-ble bonds. The product that has attracted most interest has been m-phenylenebis-maleimide (BMI). Commercial use has been limited by toxicity concerns and by col-or formation in some compounds. These products do seem to form strong attach-ments to many filler surfaces, but the mechanism of attachment is not clear; it mayinvolve acid groups resulting from hydrolysis of some of the imide linkages.

BMI has been shown to be an effective cross-linking agent for some elastomers[25]. The cross-linking is believed to proceed through free radicals generated by acharge-transfer mechanism. BMI has also been found to give rise to significant prop-erty improvements in filled polymers such as polyethylene and polypropylene. Graft-ing to the polymer is believed to occur through free radicals produced in a similar wayas in elastomer cross-linking. As with the other unsaturated systems discussed earli-er, it seems that anti-oxidants may interfere with the grafting process and that addi-tion of peroxide can assist performance.

Khunova et al. have demonstrated beneficial effects with many types of mineralfiller in polyolefin compounds and also made a detailed study of the reactions occur-ring [26–29]. They found that BMI had little effect in unfilled compounds, but that itgave significant increases in strength and elongation in filled systems. BMI also ap-peared to enhance the nucleating effect of fillers, but reduced the total crystallinity.Using IR methods, these authors showed that BMI reacted with the polymer via themaleimide alkene, resulting in chain extension and cross-linking reactions. Inpolypropylene, these can partly offset the chain scission that accompanies process-ing, especially at high filler loadings. The effects of BMI are thought to stem pre-


119

dominantly from the regions around the filler particles. As with other unsaturatedsystems, stabilizers are found to inhibit the effects and peroxides to promote them.

Xanthos [30] has reported on the effects of BMI in wood fiber/PP composites. Thetemperatures needed for activation of the BMI system led to degradation of the wood,but pre-grafting of BMI to the polymer, using peroxide, was found to overcome thisproblem and led to a significant increase in tensile strength.

The effect of varying the position of the imide groups in phenylene bismaleimides(PBM) and of using aliphatic bismaleimides has been studied [31]. Using a magne-sium hydroxide filled PP copolymer, the m- and p-isomers (1,3- and 1,4-PBM, re-spectively) were found to have equal effectiveness, while the o-isomer (1,2-PBM) wasconsiderably less effective (see Table 6-7). An aliphatic bismaleimide was also in-cluded in the study and gave broadly comparable results.

Tab. 6-7 The effect of bismaleimide structure on the mechanicalproperties of a 60 wt. % magnesium hydroxide filled polypropyleneblock copolymer[a] [31]

Modif ier and level added Tensile stress Elongation at Impact strength to the formulation at yield (MPa) break (%) (notched Charpy)

(kJ m–2)

unfilled control 27 7 (yield) 18untreated 16.1 0.7 6.41,3-PBM (meta) (5 phr) 26.5 11.1 10.91,2-PBM (ortho) (5 phr) 25.7 3.3 7.01,4-PBM (para) (5 phr) 26.9 9.2 9.0C10BM (6.9 phr) 25.5 16.0 8.6

[a] Mg(OH)2 Magnifin H5 (Martinswerk) in PP RS002P copolymer(Solvay); bismaleimides not pre-coated and used at equimolaramounts.

BMI has quite a high melting point (190 °C), which may restrict its application insome polymers, such as EVA. It also has a tendency for color formation in the pres-ence of many fillers and there is some concern over its possible toxicity. Aliphatic bis-maleimides are of interest as they may be less toxic and less prone to color formation.They also have lower melting points, which may be useful in some polymers, allow-ing their use at lower processing temperatures. Results for some aliphatic bis-maleimides with varying carbon chain length in ATH-filled, thermoplastic EVA, arepresented in Table 6-8.

6.3.1.6 Chlorinated Paraff insThese additives are included here as they are another manifestation of the unsaturat-ed additive approach. This coupling agent technology was developed by researchersat the Ford Motor Company, initially for use with phlogopite mica-filled PP and laterextended to other micas and glass-filled systems [32]. It is quite a complicated tech-nology, but is able to provide the features of coupling at a much lower material cost


120

than with organosilanes. It is based on highly chlorinated paraffin waxes (e.g.ChlorezTM 700, a 70 wt. % chlorinated paraffin with chain lengths in the C20–C30range). These waxes are attracted to the filler surface during melt processing and,given enough residence time, undergo dehydrochlorination, resulting in very reac-tive, conjugated unsaturation that can then mechano-chemically graft withpolypropylene radicals produced as part of the compounding/molding process. Ishi-da and Haung have demonstrated that grafting to PP takes place and that peroxidehelps the process [33]. The mica surface is thought to promote the dehydrochlorina-tion reactions, which seem to proceed more readily with this filler than with glassfiber.

An advantage of the system is that coupling can be delayed until the injection-molding process, which can improve processability and effectiveness. The dehy-drochlorination can be accelerated by the presence of an alkaline species such asmagnesium oxide, which also neutralizes the corrosive hydrogen chloride evolved.Other additives, such as maleic anhydride, peroxides, and organosilanes, can also beincorporated.

While this approach can lead to impressive increases in strength, it has several lim-itations. It requires careful control of the processing conditions, long cycle times tocomplete the surface reaction (which increase processing costs), and can cause cor-rosion problems during processing. Like all mechano-chemical grafting approaches,it is also sensitive to the type and amount of any anti-oxidant used. The acid-func-tionalized polymers described in Section 6.3.5 have ultimately proved to offer a morecost-effective approach.

6.3.1.7 Chrome ComplexesStrange as it may seem now, the first commercial coupling agent was probably VolanA from DuPont, which was available in 1961 [34]. This was produced by condensinga methacrylate compound with chromium(III) chloride and was successfully usedwith glass fibers in unsaturated polyester resins. It is another example of the use ofunsaturated compounds. Today, this technology has been largely displaced by the useof organofunctional silanes, as discussed in Chapter 4.


Tab. 6-8 Effect of γ-aminopropyltriethoxysilane and aliphatic bis-maleimides (DM) with variable chain length on the properties ofATH/EVA-based composites[a] [31]

Modif ier and level added Tensile strength, Elongation at Colorto the formulation (phr) MPa break, (%)

untreated 10.7 59 creamsilane (5 phr) 13.5 144 creamC4DM (5.0 phr) 13.9 114 bright pinkC6DM (5.6 phr) 13.5 126 dark pinkC12DM (7.5 phr) 11.9 163 cream

[a] 60 wt. % ATH (SF7E, Alcan Chemicals) in EVA (1020 VN5, 18% VA,Evatane, AtoFina); DM not pre-coated and used at equimolaramounts.

121

6.3.1.8 Other Carboxylic Acid DerivativesThere has been little success with derivatives other than the unsaturated ones re-ferred to above. The main exception has been in the use of amino acids, which, as dis-cussed later, have proved to be very successful coupling agents for the incorporationof nanolayer silicates into some polyamides. There has also been some interest in hy-droxy acids, such as hydroxy stearic acid and its derivatives. The main current use forthese products seems to be as high efficiency dispersants, especially for inorganicpigments.

6.3.2Alkyl Organophosphates

Acid phosphates provide an alternative anchoring functionality to the carboxyl groupand have been investigated for use on fillers such as calcium carbonate [35–37]. At-tachment to the filler surface is again assumed to be through salt formation. Withalkyl substituents, the properties achieved are generally similar to those achievedwith the less expensive fatty acids. Functional dihydrogen phosphates appear to beable to act as potential coupling agents, especially for calcium carbonate. Most workhas been centered on elastomer systems, where improvements consistent with cou-pling have been reported for a number of treatments. Unsaturation was found to beeffective in peroxide-cured elastomers and a mercapto function to give the best re-sults in a sulfur-cured system. Somewhat surprisingly, little has been published ontheir performance in other polymer types, although some promising results havebeen reported for calcium carbonate in polypropylene and polyethylene [37].

6.3.3Alkyl Borates

The use of these is described in patent literature, but again they seem to offer no ad-vantage over the less expensive fatty acids [38].

6.3.4Alkyl Sulfonates

The use of alkyl sulfonic acids has not received much attention, but DeArmitt andBreeze have reported that they make good dispersants for dolomite in organic f luids[7].

6.3.5Functionalized Polymers

This approach has ultimately proved to be the most successful of the non-silane cou-pling technologies. Here, the potential coupling agent is, in effect, pre-reacted withthe polymer under controlled conditions, introducing filler-reactive groups into thepolymer. The most commonly used groups are carboxylic acids, acid anhydrides, and


122

alkoxysilanes. Much of the basic information discussed above concerning the reac-tion of acid and acid anhydride groups with filler surfaces is relevant to these addi-tives.

There are two main types of functionalized polymers in use; polyolefins andpolybutadienes, and these are discussed separately below.

6.3.5.1 Functionalized Polyolef insThese are the logical conclusion of the work with unsaturated acids and acid anhy-drides discussed above, and are becoming of considerable importance in thermo-plastics applications. Instead of using a one-step process to achieve reaction with thefiller and the polymer, the additive is pre-grafted onto the polymer. While introduc-ing an additional step, this overcomes many of the difficulties mentioned above, es-pecially those relating to the presence of high levels of anti-oxidant during the graft-ing stage. Although initially relatively expensive products, these grafted polymers arenow very competitive in price and their use is rapidly expanding.

Various commercial products are available, notably from Atofina (LotaderTM andOrevacTM), DuPont (FusabondTM), Eastman (EpoleneTM), ExxonMobil (ExxelorTM),and Crompton (PolybondTM). The terminology used in this discussion is as follows:acrylic acid grafted onto PE or PP = AA-g-PE or AA-g-PP; maleic anhydride graftedonto PE or PP = MA-g-PE or MA-g-PP.

These principal commercial products are produced by peroxide grafting of maleicanhydride or acrylic acid onto polyethylene or polypropylene. Unfortunately, thegrafting reactions do not proceed to completion and can leave significant amounts ofunreacted monomer and, possibly, peroxide in the compositions. When applied topolypropylene, they also lead to significant MW reduction due to chain scission. Insome cases, the polyolefin is first thermally degraded to a wax-like material that hasterminal unsaturation, which can then react with maleic anhydride, either thermallythrough an ene reaction or assisted by peroxide [39].

Maleic anhydride does not readily undergo homopolymerization and its graftingleads to the attachment of succinic anhydride groups. When acrylic acid is used, ho-mopolymerization can occur, leading to longer side chains. These two effects are il-lustrated in Figure 6-3.

In some cases, copolymerization is used. Typical of the products made in this wayare terpolymers of ethylene, n-butyl acrylate, and maleic anhydride, such as theLotaderTM products from Atofina, which have proved to be very good coupling agentsin filled EVA copolymers.

Today, most commercial products for filler applications appear to be based onmaleic anhydride, rather than acrylic acid, and what comparative data there are avail-able suggest that the maleated forms are the more effective and can be used at lowerlevels. The acrylic acid forms have higher functionality levels than the maleated ones.

A useful description of the grafting of acrylic acid onto polypropylene can be foundin ref. [40]. Good improvements in both strength and toughness (notched and un-notched Izod impact strength) of a calcium carbonate filled polypropylene were ob-tained, although very high addition levels of the AA-g-PP were required. Improvedbonding of the filler to the polymer was confirmed by electron microscopy.


123

Mai et al. reported on the use of AA-g-PP in ATH-filled PP homopolymer [41] andfound the graft polymer to increase filler-to-polymer wetting and adhesion, as shownby electron microscopy of fracture surfaces. They also found significant increases inf lexural strength, but a decrease in notched impact strength. The loss of impactstrength was most marked at low to moderate filler levels and was least at the 60%level, a figure typical for fire retardant compounds. It was also least at the highest acidgrafting level. The difference in impact results between this work and that describedin ref. [40] may stem from the nature of the filler. The calcium carbonate used wasquite coarse, and probably reduced the impact strength of the polymer (data not giv-en in the paper), while the ATH used was much finer and actually increased the im-pact strength.

Adur [42] has also provided data for the effect of AA-g-PP in PP with a number offillers, including glass fibers and wollastonite. His results showed significant in-creases in most properties (including notched and unnotched Izod impact strengthand heat distortion temperature, HDT) with glass fibers as the filler. With wollas-tonite, there was a significant increase in tensile strength, but no real effect on HDTor impact properties. Again, high levels of additive were found to be needed.

Chun and Woodhams [39] reported significant increases in tensile strength, but lit-tle effect on impact strength, by using a maleated polypropylene wax with mica in aPP homopolymer. Lower levels of additive were needed than in the work with AA-g-PP described above. The authors also found a synergistic effect with some silanetreatments.

Results obtained by the present author and co-workers for a calcium carbonatefilled PP are presented in Table 6-9.

Sigworth has recently described the use of MA-g-PP and MA-g-PE in natural fiberfilled polyolefin compounds [43]. The additives were found to be effective on a widerange of natural products, including wood f lour. The significant improvements re-ported with additive levels of about 5% based on the natural fiber were increased ten-sile, f lexural, and impact strengths; reduced room temperature creep; higher heat


Fig. 6-3 Filler reactive groups introduced into polymer backbonesby maleic anhydride and acrylic acid grafting. Single and multiplegrafted functionalities are respectively shown.

124

distortion temperature; reduced water adsorption; and improved property retentionon moisture ageing.

Functionalized polymers are mainly used as compounding ingredients rather thanbeing precoated onto the filler. Some forms can be converted into aqueous emul-sions, however, and it is believed that these are sometimes used to coat fillers. Theiruse as filler coatings can be useful in reducing dust problems in some cases.

As with the other acid and acid anhydride products already mentioned, bonding toa mineral filler surface is believed to be through salt formation. Bonding to the hostpolymer is believed to be a combination of chain entanglement, chain pinning, andeven co-crystallization. It appears to be important to match the structure of the addi-tive and the host polymer, at least to some degree. Thus, acid-functionalized homo-and co-polymer propylenes are available. It is also necessary to choose the melt f lowrate carefully. High melt f low rates favor processing, but are detrimental to stiffnessand heat distortion temperature. Although the level of acid groups is also clearly im-portant, there is no clear advice on this from the manufacturers. Most graftedpolypropylene products have quite low acid levels, due in part to the MW reductionthat accompanies grafting.

The MW would seem to be important in determining the effectiveness of the in-teraction with the matrix polymer. There have been few useful data published, but Fe-lix and Gatenholm [44] have shown that property improvements in a cellulose-filledpolypropylene become increasingly more marked as the MW is increased up to a lim-it of about 10,000. As will be seen in the next section, a similar MW limit has beenfound for maleated polybutadienes in calcium carbonate filled elastomers.

The current main uses of these functionalized polymers are:

in glass fiber reinforced polyolefins, where the glass fiber is usually pre-coatedwith an amino functional silane that can react with the acid or acid anhydridethrough the formation of an amide linkage (see Chapter 7);

in compounds based on fire-retardant fillers such as aluminum and magnesiumhydroxides (see Chapter 18);

in wood-filled polyolefins, where the acid or acid anhydride is thought to react withcellulosic hydroxyl groups (see Chapter 15).

Maleated polypropylenes also play an important role in the preparation of thermo-plastic aluminosilicate nanocomposites [45]. This could become another importantoutlet, if this technology becomes as widely adopted as some are predicting.


Tab. 6-9 Effect of maleated PP on the properties of calciumcarbonate filled PP.

Additive Secant Tensile Elongation Impact strength modulus, strength, at yield, % (unnotched Charpy), GPa MPa kJ m–2

none 2.4 24.7 1.5 5.6MA-g-PP (7 wt. % on filler) 2.9 33.7 2.3 11.1

125

6.3.5.2 Functionalized PolybutadienesPolybutadienes contain a high level of unsaturation and are thus suitably predisposedfor polymerization and cross-linking reactions. They can also be readily further func-tionalized with the introduction of filler-reactive groups. Two approaches are used toprepare coupling agents from them. Either maleic anhydride can be grafted ontothem, or trialkoxysilane groups can be introduced by hydrosilylation procedures,thereby generating products that can be used on fillers not responsive to acidic addi-tives. Only the maleic anhydride approach is appropriate for this section.

As with polyolefins, reaction with maleic anhydride results in the attachment ofsuccinic anhydride groups, but in this case no peroxide is necessary, the reaction pro-ceeding as an ene addition. High levels of anhydride can be grafted in this way, withthe resulting products being water-soluble (in salt form) and, thus, suitable for fillercoating. They can also be used as compounding additives. Liquid polybutadienes withMW in the region of 10,000–50,000 are most commonly utilized.

These additives are mainly used in elastomer applications, where they are suitablefor both sulfur and peroxide cures. The effects of both acid content and MW havebeen reported in ref. [46]. Property improvements were found to reach a plateau at anumber average MW of about 10,000 and with about 20 wt. % of grafted maleic an-hydride.

While of most commercial interest in elastomers, maleated polybutadienes(MPBD) can also provide good coupling for fillers in other polymers, such as cross-linked ethylene vinyl acetates and thermosets such as poly(methyl methacrylate)(PMMA). Typical results obtained with cross-linked EVA are presented in Table 6-10,where the MPBD additive is seen to give similar effects to those obtained with a con-ventional vinylsilane. The effect on fire retardancy, as measured by the oxygen index,is included to illustrate that modifiers can affect such properties. A decrease in oxy-gen index is often seen when coupling agents are used in cross-linked formulations.Results for a variety of fillers in PMMA can be found in the patent literature [47].

Tab. 6-10 Effect of MPBD[a] in a magnesium hydroxide filled, perox-ide cross-linked EVA composite[b]

Additive Property

Tensile strength, Elongation at Oxygen index, %MPa break, %

none 10 325 42MPBD (2% on filler) 14 210 31vinylsilane (A-172) (2% on filler) 13 190 32

[a] Atlas G-3965, ICI.[b] 125 phr Mg(OH)2 (DP 390, Premier Periclase, Ireland).

Acid-functionalized polybutadienes are most effective with fillers such as calciumcarbonate and aluminum and magnesium hydroxides, but are not so effective withsilicas and other silicates. As with the succinic anhydride derivatives discussed above,it is assumed that the anhydride functional polymers react with the filler surface by


126

ring-opening and salt formation, although there have been few confirmatory studies.Some useful work has been carried out with maleated polybutadiene containingabout 20 wt. % grafted maleic anhydride. FT-IR analysis showed that adsorption on-to magnesium hydroxide from solution at room temperature gave mainly a mixtureof salt and unreacted anhydride peaks, with only a small amount of free acid [48].With ATH, the situation was somewhat different. As in the studies using dodecenylsuccinic anhydride discussed earlier, unreacted anhydride and free acid predominat-ed at room temperature, and temperatures of near 100 °C were necessary in order forsalt formation to be observed. This difference is probably due to the greater reactivi-ty of the magnesium hydroxide surface.

6.3.6Organic Amines and their Derivatives

Alkyl- and arylamines have long had a minor role in treating some filler surfaces, no-tably some clays. They are normally added in a water-soluble ammonium salt form,which can undergo ion exchange with cations such as sodium on the clay surfaces.

Their importance has recently received a significant boost as a result of the grow-ing interest in nanoclays as high value effect fillers (see Chapter 9). This area hasbeen the subject of an excellent review by Alexandre and Dubois [49]. Certain clays,notably montmorillonite, have an aluminosilicate layer structure with exchangeablecations such as sodium trapped between the layers. When these clays are swollen inwater, the inorganic cations can be exchanged for long-chain cations of ammoniumsalts. As a result, the layers can be separated into very thin (about 1 nm) high aspectratio sheets, which can then be dispersed into polymers.

The ammonium salts can be derived from primary, secondary, and tertiary amines.The exact structure and degree of exchange are chosen to suit the application, espe-cially the polarity of the polymer that the layers are to be dispersed in. Quarternaryammonium salts containing a mixture of methyl and long-chain alkyl groups arecommon. In some cases, where some polarity is desired, hydroxy groups can be in-corporated into the structures. A particularly important system is the use of aminoacids, such as 12-aminododecanoic acid. These can provide chains bearing terminalacid groups, with the aid of which the very thin plates can be incorporated into poly-mers such as nylon.

As brief ly mentioned earlier, acid or acid anhydride functional polymers are usedto ensure dispersion and exfoliation of the organoclays in polymers such aspolypropylene. It is popularly believed that this is due to interaction of the functionalpolymer with the onium ion treatment on the clay, but this appears to be a miscon-ception. It seems that the key feature, attracting the functional polymer into the claygalleries, is its interaction with free sites on the clay surface. Thus, complete coatingof the surface is undesirable.


127

References

1 Ancker, F. H., et al., U.S. Patent 4,385,136,1981 (Union Carbide Corp.).

2 Fulmer, M. J., et al., Proc. 58th Society ofPlastics Engineers (SPE) ANTEC, 2000, 46,p. 552.

3 Papirer, E., et al., Eur. Polym. J. 1984, 20(12),1155–1158.

4 Rothon, R. N., Eur. Rubber J. 1984, 166(10),37–42.

5 Taylor, D. A., Paynter, C. D., Proc. Poly-mat’94 Conference, London, U.K., Sept.1994, 628–638.

6 Rothon, R. N., et al., J. Adhesion 2002, 78,603–628.

7 DeArmitt, C., Breese, K. D., Proc. Function-al Fillers for Plastics 2003, Intertech Corp.,Atlanta, GA, October 2003, Paper 16.

8 Haworth, B., Raymond, C. L., Proc. Euro-f illers 97 Conf., Manchester, U.K., Septem-ber 1997, 251–254.

9 Haworth, B., Birchenough, C. L., Proc. Eu-rofillers 95 Conf., Mulhouse, France, Sep-tember 1995, 365–368.

10 Tabtiang, A., Venables, R., Eur. Polym. J.2000, 36, 137–148.

11 Hancock, A., Rothon, R. N., in Particulate-Filled Polymer Composites, 2nd Ed. (Ed.:Rothon, R. N.), RAPRA Technology Ltd.,Shawberry, Shrewsbury, Shropshire, U.K.,2003, 90.

12 Liauw, C. M., et al., J. Adhesion Sci. Technol.2001, 15(8), 889–912.

13 Suess, E., Ph.D. Dissertation, “CalciumCarbonate Interaction with Organic Com-pounds”, Lehigh University, USA, 1968.

14 Bonsignore, P. V., U.S. Patent 4,283,316,1981 (Aluminum Company of America).

15 Mori, A., et al., Japan Patent Appl. No. 62-190857 (filed July 30, 1987), (Nippon SodaCo. Ltd.).

16 Rothon, R. N., et al., Proc. Functional Fillersfor Plastics 2002, Intertech Corp., Toronto,Canada, September 2002, paper 6.

17 Bixler, H. J., Fallick, G. J., U.S. Patent3,471,439, 1969 (Amicon Corp.).

18 Gaylord, N. G., et al., ACS Symposium Series121 (Eds.: Carraher, Jr., C. E., Tsuda, M.),American Chemical Society, Washington,DC, 1979, 469–474.

19 Ishima, I., et al., U.S. Patent 4,242,251, 1980(Asahi, Kasei Kogyo Kabushiki Kaisha).

20 Tabtiang, A., Venables, R., Composite Inter-faces 1999, 6(1), 65–79.

21 Miyata, S., et al., J. Appl. Polym. Sci. 1980,25(3), 415–425.

22 Ferrigno, T. H., U.S. Patent 4,420,341, 1983.23 Liauw, C. M., et al., Plast. Rubber Comp.

Proc. Appl. 1995, 24(2), 211–219.24 Kishikawa, H., et al., Japan Kokai 73 45540,

1973; Chem. Abstr. 1974, 80, 60620p.25 Hill, R. K., Rabinovitz, M., J. Am. Chem.

Soc. 1964, 86, 965.26 Khunova, V., Sain, M. M., Angew. Makromol.

Chem. 1995, 224, 9.27 Khunova, V., Sain, M. M., Angew. Makromol.

Chem. 1995, 224, 11.28 Khunova, V., Liauw, C. M., Chem. Papers

2000, 54(3), 177–182.29 Liauw, C. M., et al., Macromol. Mater. Eng.

2000, 279, 34–41.30 Xanthos, M., Plast. Rubber Proc. Appl. 1983,

3(3), 223–228.31 Liauw, C. M., et al., Proc. Eurofillers 03, Ali-

cante, Spain, September 2003, 145–147.32 Meyer, F. J., Newman, S., Proc. 34th Ann.

Tech. Conf. Reinf. Plast. Compos. Div. SPI1979, Section 14-G.

33 Ishida, H., Haung, Z.-H. Proc. 42nd Societyof Plastics Engineers (SPE) ANTEC 1984, 30,205–208.

34 Yates, P. C., Trebilcock, J. W., Proc. 16thAnn. Tech. Conf. Reinf. Plast. Compos. Div.SPI 1961, Section 8.

35 Nakatsuka, B. T., Kawasaki, H., Itadani, K.,Yamashita, S., J. Colloid Interface Sci. 1981,82(2), 298–306.

36 Nakatsuka, B. T., et al., J. Appl. Polym. Sci.1982, 27, 259–269.

37 Luders, W., et al., U.S. Patent 4,174,340,1979 (Hoechst AG).

38 British Patent 1,555,186, 1977 (Dart Indus-tries).

39 Chun, I., Woodhams, R. T., Proc. 42nd Soci-ety of Plastics Engineers (SPE) ANTEC 1984,30, 132–135.

40 Rahma, F., Fellahi, S., Polymer Composites2000, 21(2), 175–186.

41 Mai, K., et al., J. Appl. Polym. Sci. 2001,80(13), 2617–2623.

42 Adur, A. M., Seminar Papers, Reactive Pro-cessing – Practice and Possibilities, RAPRATechnology Ltd., Shawberry, Shrewsbury,Shropshire, U.K., 1989, paper 5.8.

References

128 6 Functional Polymers and Other Modif iers

43 Sigworth, W., Proc. Functional Fillers forPlastics 2002, Intertech Corp., Toronto,Canada, September 2002, paper 19.

44 Felix, J. M., Gatenholm, P., J. Appl. Polym.Sci. 1991, 50, 699.

45 Manias, E., et al., Chem. Mater. 2001, 13,3516–3523.

46 Rothon, R. N., in Controlled Interphases inComposite Materials, Proc. 3rd Internat.

Conf. on Composite Interfaces (Ed.: Ishida,H.), Elsevier, New York, 1990, 401–406.

47 Rothon, R. N., British Patent, 0,295,005,filed 1988, ICI plc.

48 Rothon, R. N., Proc. High Performance Addi-tives Conf., May 1991, PRI/BPF, London,U.K., Paper 12.

49 Alexandre, M., Dubois, P., Mat. Sci. Eng.2000, 28, 1–63.

Part IIIFillers and their Functions


7Glass Fibers

Subir K. Dey and Marino Xanthos

7.1Background

The term glass covers a wide range of inorganic materials containing more than 50%silica (SiO2) and having random structures. They are often considered as supercooledliquids in a state referred to as the vitreous state. According to the Roman historianPliny, Phoenician sailors made glass when they attempted to cook a meal on a beachover some blocks of natron (a mineral form of sodium carbonate) that they were car-rying as cargo; in doing so, they melted the sand beneath the fire and the mixture lat-er cooled and hardened into glass. The first true glass was probably made in westernAsia, perhaps Mesopotamia, at least 40 centuries ago as a by-product from coppersmelting. The science of glass making was developed over a long period of time fromexperiments with mixtures of silica sand (ground quartz pebbles) and an alkalibinder fused onto the surface. The basic raw materials for the manufacture of ordi-nary soda glass used for windows and bottles are sand, sodium carbonate (soda), andcalcium carbonate (limestone). Typical compositions of these soda-lime-silica glass-es, expressed in % oxides, are about 72–74% SiO2, 14–16% Na2O, 5–10% CaO,2.5–4% MgO, and minor amounts of Al2O3 and K2O [1,2].

It was found that by varying the chemical composition, the mechanical, electrical,chemical, optical, and thermal properties of the glasses, and the ease with which theycould be drawn into fibers, could be modified. Pyrex® glass, which contains about80% SiO2 and a relatively large amount of B2O3 (typically 13%), Na2O (4%), and mi-nor quantities of Al2O3 and K2O, is stronger than soda-lime glass, has better chemi-cal resistance, and a lower coefficient of thermal expansion, but is not so easily drawninto fibers [1]. Various compositions that are more readily drawn into fibers havebeen devised for plastics reinforcement purposes. The most commonly used in gen-eral purpose reinforced plastics is E-glass, a lime-borosilicate glass derived from aPyrex® composition.

The commercialization of glass fibers in the mid-1930s and the development ofpolyester resins during the same period were instrumental in the introduction andestablishment of reinforced plastics/composites (RP/C) based on glass fibers as new


132

construction materials. Significant advances in the processing and applications ofmolded and laminated products were made in the USA during World War II, andthese were followed in the post-war era by the penetration of RP/C into many mar-kets such as automotive, marine, aircraft, appliance, recreational, and corrosion-re-sistant equipment [3]. During the manufacture of glass fibers, after being collimatedinto strands the continuous fibers are further processed into various forms suitablefor use with thermosetting or thermoplastic matrices. Initial applications of glassfibers were as continuous (long) reinforcements in thermosetting resins. Discontin-uous (short) glass fiber reinforced thermoplastics, which constitute the focus of thischapter, were first developed in the late 1940s/early 1950s as glass-reinforced nylonand polystyrene [4], and since then there has been a significant growth in the use ofglass fibers in a variety of commodity and engineering thermoplastics.

7.2Production Methods

Liquid glass is formed by blending the appropriate ingredients for a given composi-tion (e.g. sand, metal salts, boric acid for E-glass) in a high-temperature furnace attemperatures up to 1,600 °C. The liquid is passed through electrically heated plat-inum alloy bushings that may contain up to 4000 holes, through which the glass ismetered into filaments. The diameter of the filaments can, in principle, be variedfrom 2.6 to 27.3 µm and is controlled by composition, viscosity, tip diameter, drawingtemperature, cooling rate, and rate of attenuation. The filaments are drawn togetherinto a strand (closely associated) or “roving” (loosely associated), cooled, and coatedwith proprietary formulations of organic chemicals, the “sizing”, to provide filamentcohesion through film formation, lubrication, protection of the glass from abrasion,and compatibility with the polymer matrix. The collimated filaments from the fiber-forming stage can be wound onto a tube by means of a high-speed winder to give con-tinuous strands that can be characterized according to the diameter of their fila-ments, the linear weight or tex count (g km–1), the direction of twist, and the numberof turns per meter. Strands used in reinforced plastics contain filaments with diam-eters in the 9.1 to 23.5 µm range; those with diameters ranging from 9.1 to 15.9 µmare specifically used for thermoplastics [3]. Figure 7-1 provides a schematic repre-sentation of glass fiber production [5]. The continuous strands may be converted in-to various other forms suitable for open-mold and other thermoset applications, suchas rovings, woven rovings, fabrics, and mats. They may also be cut to specific lengthsto produce chopped strands, or milled to finer sizes for a variety of thermoplastic andthermoset applications. By varying the composition of the initial mixture of raw ma-terials, different types of glasses can be produced.

As discussed in Chapter 2, the surface free energy of glass varies depending onwhether the glass is bare or sized. The high surface energy of bare (pristine) glass fil-aments, which permits their easy wetting, will rapidly decrease due to adsorption ofatmospheric water. In addition, as shown in Figures 7-2 and 7-3, the strength of vir-gin filaments decreases rapidly in air or in water, as well as at elevated temperatures

7 Glass Fibers

133

[6]. Data from glass manufacturers show differences between a standard borosilicateglass (E-glass) and a higher strength, higher modulus glass (R-glass) of different com-position. Sizings applied on the fiber surface immediately after the fibers leave thebushing minimize the loss of strength during subsequent processing. Each sizing isspecially designed for a given molding or compounding process and for a differentmatrix type. In addition to ingredients added to improve the handling characteristicsof the fibers, the use of adhesion promoters such as silane coupling agents con-tributes to enhancing the mechanical properties of composites and particularly theirresistance to ageing (see also Chapter 4). The applied coatings tend to reduce surfacefree energy and, in general, wetting or spreading is not as favorable with these sizingsas it is with truly bare fibers.

The importance of sizing composition in the development of glass fiber gradessuitable for a specific application cannot be underestimated. Sizing composition andits solubility in the polymer are different for room temperature cure of thermosets(HSB, high solubility), high-temperature press moldings (LSB, low solubility), or foruse in reinforced thermoplastics (RTP) [3]. Sizing compositions for chopped strandsor milled fibers that are commonly used with thermoplastics and certain thermosetsinclude both LSB and RTP types. As an example, sizing compositions for glass fibersused for polypropylene (PP) reinforcement are typically based on a hydrolyzedaminoalkyltrialkoxysilane as the coupling agent, and an aqueous, colloidal dispersionof the alkali metal salt of a low MW maleated PP containing an appropriate emulsi-fying agent. The resulting coating provides fiber surfaces with an acceptable degreeof wettability by molten PP and a reasonable compatibility with crystallized PP, thus

7.2 Production Methods

Glass Strand Fabrication

QuarryProducts

Furnace

BushingFilamentForming5 to 24 µm

WindingBasic Strand

Sizing

Fig. 7-1 Flow diagram of glass fiber production. Adapted from ref. [5].

134 7 Glass Fibers

Fig. 7-2 Tensile strength of virgin glass filaments as a functionof temperature. Adapted from ref. [6].

Fig. 7-3 Tensile strength of virgin glass filaments as a function oftime from manufacturing in different environments. Adapted fromref. [6].

135

affording composites with a good overall spectrum of mechanical properties. Thecomplexity of commercial sizing compositions is illustrated with three examples ofwater-based formulations suitable for three different polymers in Table 7-1 [7].

Tab. 7-1 Composition of glass fiber sizings [7].

Components Polyester compatible Polyvinyl acetate Polyurethane compatible (U.S. Pat. 4,752,527) compatible (U.S. Pat. 3,803,069)

(U.S. Pat. 4,027,071)

Solvent water water water

Coupling agent γ-methacryloxypropyl- γ-ethylenediamine γ-aminopropyltriethoxy-trimethoxysilane (propyl)trimethoxysilane silane

or methacrylic acid complex of chro-mium(III) chloride

Film former unsaturated bisphenolic polyvinyl acetate curable blocked poly-glycol-maleic polyester urethane resin emulsion

Anti-static agent cationic organic quaternary ammonium salt

Lubricant polyethyleneimine cationic fatty acid amide polyamide or tetraethylene

pentamine

Strand hardening aqueous methylated agent melamine–formaldehyde

resin

pH control acetic acid acetic acid

Emulsifying condensate of polypro-agent pylene oxide with pro-

pylene glycol

7.3Structure and Properties

The conventional concept of an all-silica vitreous structure is that of silicon–oxygentetrahedral building blocks linked at their corners and randomly organized in a net-work. In glasses, the structure is broken or distorted by the addition of monovalentf luxes and modified by substitution with such ions as aluminum and boron. Thus,the melting point and the viscosity of the vitreous silica is reduced by f luxing withsodium oxide, which is added to the melt as sodium carbonate. Calcium and magne-sium, which are other constituents of glass, are introduced into the structure as net-work modifiers. These modifiers make the structures more complex to hinder crys-tallization of the molten mass during the cooling process. This is why glass is often

7.3 Structure and Properties

136

referred to as a supercooled liquid having no crystallization or melting point and nolatent heat of crystallization or fusion.

Glass fibers are characterized by a relatively high strength and a reasonable cost,but lower modulus as compared to other fibrous reinforcements such as carbon oraramid. The fibers are incombustible, have excellent high-temperature resistanceand chemical resistance, and offer chemical affinity to a variety of resins or couplingagents through their surface silanol groups. In Table 7-2, the properties of glass yarnsare compared with those of aramid and carbon [8].

Tab. 7-2 Comparison of properties of continuous fiberreinforcements for plastics [8].

Property E-glass Carbon[a] Aramid (Kevlar 49)

Tensile strength, MPa 3450 3800–6530 3600–4100Elasticity modulus, GPa 73 230–400 131Elongation to break, % 3–4 1.40 2.5Density, g cm–3 2.58 1.78–1.81 1.44Relative cost 3.7 52–285 44

[a] Values depend on the type of fiber (standard, intermediate or high modulus).

The fibers can be further characterized by their physical and chemical properties,which are governed primarily by the composition of the glass. There are several glassfiber types, with different chemical compositions for different applications. They in-clude:

A-glass; the most common type of glass for use in windows, bottles, etc., but not of-ten used in composites due to its poor moisture resistance.

C-glass; high chemical resistance glass used for applications requiring corrosionresistance.

D-glass; glass with improved dielectric strength and lower density.E-glass; a multi-purpose borosilicate type and the most commonly used glass for

fiber reinforcement.S-glass; a magnesia-alumina-silicate composition with an extra high strength-to-

weight ratio, more expensive than E-glass and used primarily for military and aero-space applications.

The chemical compositions and physical properties of these glasses and theirfibers are shown in Table 7-3.

Other specialty glasses include R-glass, which has higher strength and modulusthan S-glass; AR-glass, which has a high zirconia content and is much more resistantto alkali attack than E-glass; and AF-glass, a general purpose alkali borosilicate withimproved durability compared to container and sheet glasses.

7 Glass Fibers


Tab. 7-3 Chemical compositions and physical properties of variousglass fibers[a].

A-Glass C-Glass D-Glass E-Glass S-Glass

CompositionSiO2 72–73.6 60–65 74 52–56 65Al2O3 0.6–1.0 2–6 0.3 12–16 25B2O3 – 2–7 22 5–10 –K2O 0–0.6 – 1.5 0–2 –Na2O 14–16 8–10 1.0 0–2.0 –MgO 2.5–3.6 1–3 – 0–5 10CaO 5.2–10 14 0.5 16–25 –TiO2 – – – 0–1.5 –Fe2O3 – 0–0.2 traces 0–0.8 –Li2O 0–1.3 – 0.5 – –SO3 0–0.7 0–0.1 – – –F2 – – – 0–1.0 –

PropertyDensity, g cm–3 2.50 2.49–2.53 2.14–2.16 2.52–2.65 2.5Softening point, °C 700 689–750 775 835–860 970Modulus of elasticity, GPa 70–75 69 55 70–75 85Tensile strength, MPa 2450 2750 2500 3400 4600Elongation at break, % 4.3 – 4.5 4.5 –Poisson’s ratio 0.23 – – 0.22 0.23Hardness, Mohs 6.5 – – 6.5 –Refractive index 1.51–1.52 1.54 1.47 1.55–1.57 1.52Dielectric constant at 1 MHz 6.9 6.24 3.56–3.85 5.8–6.7 4.9–5.3Dielectric strength, kV mm–1 – – – 8–12 –Volume resistivity, ohm cm – – – 1013–1014 –Coefficient of linear expansion, 5–8 9.4 3.5 5 5.910– 6 K–1

Thermal conductivity, W m–1 K–1 1.0–1.3Specific heat, J kg–1 K–1 810–1130Refractive index 1.51–1.52 1.54 1.47 1.55–1.57 1.52

[a] Data compiled from various manufacturers’ websites and refs. [2,3,9].

138

7.4Suppliers

Table 7-4 contains information on the major global glass fiber suppliers and theirproducts.

7.5Cost/Availability

Total glass fiber demand in the USA is forecasted to reach 3.5 million tons in 2007,with a significant portion of the envisaged increases being based on opportunities inreinforced plastics [10]. Continuous forms normally used in thermosets are strands(compactly associated bundles of filaments), rovings (loosely associated bundles ofuntwisted filaments or strands) and woven roving fabrics, continuous filament mats(felts of continuous filaments distributed in uniform layers held together by abinder), and chopped strand mats (felts or mats consisting of glass strands choppedto lengths of mostly 50 mm and held together by a binder). Discontinuous forms foruse in thermoplastics processes (extrusion, injection molding) and certain thermosetfabrication methods (injection and compression molding and reinforced-resin injec-tion molding, RIM) include chopped strands and milled fibers. The market price ofglass fibers depends on their type, with A- and E-glasses being the lowest in price. ForE-glass, the most common reinforcement grade, prices are in the range of 1.70–3.00US$/kg, depending on the form and quantities.

Chopped glass strands are produced from continuous rovings and are available atdifferent lengths varying from 3 to 4.5 mm for thermoplastics and from 4 to 13 mmfor thermosets. Nominal aspect ratios with a typical 10 µm diameter fiber (beforecompounding) thus vary from 300 to 1300. Suitably sized grades are available de-pending on the particular thermoplastic (PA, PET, PBT, PP, PC, PPS, PPO, styrenics)or thermoset application (mostly unsaturated polyesters as sheet and bulk moldingcompounds (SMC, BMC)).

Milled fibers are produced from continuous E-glass roving by hammer mills andare used in both thermoplastics and thermosets. Unlike chopped strands, which arereduced in size to specific lengths, the milled glass fiber is reduced in size to an av-erage volume. The lengths of milled fibers (50–350 µm) are substantially smallerthan those of chopped strands. Milled glass is produced in powder and f loccularforms with the powdery form generally having shorter fiber lengths and higher bulkdensity than the f loccular form. Various grades available for different applications(thermoplastics, thermosets, PTFE) differ in bulk density and type of sizing. More in-formation on the types and characteristics of available glass fiber products can befound via the suppliers’ websites and in refs. [3,9,11]. Silver-coated conductive glassfibers with powder resistivities ranging from 1.6 to 3.5 mΩ cm and mean lengthsranging from 175 to 50 µm are also available for applications requiring electromag-netic interference (EMI) shielding properties [12].

7 Glass Fibers

1397.5 Cost/Availability

Tab. 7-4 List of major glass fiber suppliers

Name, website Location, product

Aiken, SC. E-glass, Advantex,S-2 glassyarns and single end high strength glassrovings.

Turkey. E-glass reinforcements

Japan. E-glass yarn and direct sized rov-ing, chopped strand, chopped strand mat,milled glass.

Canada. Boron- and f luoride-free E-CRglass fiber rovings.

Germany. Glass fiber strands, mats, androvings.

China. E- and C-glass fiber fabric, yarnsand roving

E-glass assembled and direct rovings,yarns, and chopped strands

Korea. E-glass direct and assembled rov-ings, milled glass, woven roving, and veil

Germany. Chopped strands and microglass fiber.

China. Specialty fibers and fabrics

Japan. E-glass and AR-glass yarns, mats

Japan. Producer of fine E-glass yarns,high strength glass, and E-glass rovings,chopped strand mat, filament mat

USA. Multi-national manufacturer of ad-vanced glass material systems

USA. Glass fiber for non-woven and tex-tile applications

France. Glass fabrics and tapes

US subsidiary of CSG. Manufacturer ofE-glass in Texas and Mexico. Supplier ofAR-glass, R-glass, TwinTex

Hong Kong, China. Manufacturer of AR-glass and E-glass in China. Roving,chopped strand, yarn, tape, mesh, fabric,cloth, scrim, gun roving

Czech Republic. Glass fiber filamentsand chopped strands. Part of the Saint-Gobain Group

Advanced Glassfiber Yarns LLC –http://www.agy.com

Camely AF – http://www.camelyaf.com.tr

Central Glass Co., Ltd. – http://www.cgco.co.jp

Fiberex, Ltd – http://www.fiberex.com/

Glasseiden GmbH – http://www.glasseide-oschatz.de/

JINWU Glass Fibre – http://www.jwfg.com

Johns Manville – http://www.jm.com

KCC Glass Fiber –http://www.kccworld.co.kr/eng/product/

Lauscha Fiber International GmbH – http://www.lfifiber.com/

Nanjing Fiberglass Research Institute –http://www.fiberglasschina.com/english/

Nippon Electric Glass Co., Ltd. –http://www.neg.co.jp/eng/index.html

Nitto Boseki Co., Ltd. – http://www.nittobo.co.jp/

Owens Corning – http://www.owenscorning.com/

PPG Fiber Glass – http://www.ppg.com/

Saint-Gobain Group – http://www.saint-gobain.fr/

Saint-Gobain Vetrotex – http://www.sgva.com/index.html

Texas Fiberglass Group – http://www.fiberglass.to

Vertex AS – http://www.vertex.cz/

140

7.6Environmental/Toxicity Considerations

Commercial glass fibers consist of the basic glass component plus organic surfacesizings at a level of <5% of the overall weight. Short-term exposure to glass fibers maycause irritation of the skin and possibly irritation of the eyes and upper respiratorytract (nose and throat). Fiberglass is a non-burning material, although the organicbinders may burn. It is generally considered to be an inert solid waste not requiringhazardous disposal procedures.

Health effects after long-term exposures to glass fibers have been the topic of nu-merous studies. The potential for inhaled glass fibers to cause any health hazard de-pend on their “respirability”, i.e. their potential to enter the lower respiratory tract.Only fibers of less than approximately 3.5 µm in diameter are considered respirableand hence potentially carcinogenic. Other criteria for respirable fibers [13] are alength/diameter ratio greater than 3, and a length larger than 5 µm.

According to a study conducted at the National Cancer Institute in 1970, glassfibers less than 3 µm in diameter and greater than 20 µm in length were found to bepotent carcinogens in the pleura of rats. In another more recent laboratory study [14],animals exposed to very high concentrations of respirable microfibers with a meandiameter of 0.5 µm on a long-term basis developed lung tumors, fibrosis, andmesotheliomas. Since then, studies have continued to appear, showing that fibers ofthis size not only cause cancer in laboratory animals, but also cause changes in theactivity and chemical composition of cells, leading to changes in the genetic structurein the cellular immune system. Although these cell changes may be more common(and possibly more important) than cancer, it is the cancer-causing potential of glassfibers that has attracted most attention.

According to various reports, the concentrations of glass fibers to which the work-ers in fiber glass manufacturing plants are exposed are far lower than the concentra-tions to which asbestos workers were exposed. However, statistically significant in-creased levels of lung cancer among workers handling glass fibers have been report-ed [15–18] in several industry-sponsored epidemiological studies conducted in thelate 1980s in the USA, Canada, and Europe. More recently, it has been shown that an8-hour exposure to 0.043 glass fibers per cubic centimeter of air is sufficient to causelung cancer in one-in-every-thousand exposed workers during a 45-year working life-time [19]. Other major epidemiological studies in the USA, Europe, and Canada in-volving 21500 workers in fiber glass manufacturing showed no increased incidenceof lung cancer or non-malignant respiratory disease [13]. Other epidemiological stud-ies published in 1997 and listed in ref. [13] did not provide evidence of increased in-cidence of cancer in populations working in the plants for a long time. For more cur-rent information, please refer to the appropriate MSDS from glass fiber suppliers.

OSHA considers glass fibers as a “nuisance” dust with a permissible occupationalexposure limit (8 hour time-weighted average, TWA) of 15 mg m–3 as total dust and5 mg m–3 as respirable dust. ACGIH’s TWA limit for inhalable glass fiber dust is5 mg m–3 and 1 fiber cm–3 for the respirable fraction. The International Agency forResearch on Cancer (IARC), the U.S. National Toxicology Program (NTP), and OSHA

7 Glass Fibers

141

do not list continuous-filament glass fibers as a carcinogen. The diameters of mostfillers used to reinforce fibers are greater than 6 µm and therefore they do not con-form to the “respirable“ criteria, although chopped, crushed, or severely mechanical-ly processed fibers may contain a very small amount of respirable fibers. Repeated orprolonged exposure to respirable fibers may cause fibrosis, lung cancer, or mesothe-lioma. However, measured airborne concentrations of these respirable fibers inenvironments in which fiberglass has been extensively processed have been shownto be extremely low and well below the threshold limiting value (see, for example,ref. [20]).

7.7Applications

As a result of their inherent physical and mechanical properties and availability inhigh aspect ratios, the primary function of short glass fibers is enhancement of themechanical properties of a variety of thermoplastics and thermosets. The character-istics of glass fiber thermoplastic composites, such as high strength-to-weight ratio,good dimensional stability, good environmental resistance, good electrical insulationproperties, ease of fabrication, and relatively low cost, make them particularly suit-able in a variety of applications such as automotive, appliances, business equipment,electronics, sports, and recreational.

Typical glass fiber concentrations in commodity thermoplastics, engineering ther-moplastics, and thermosets may range from 10 to 50 wt. %. The effects on the me-chanical properties of a specific polymer, as discussed in Chapters 2 and 3, largely de-pend on the type of polymer, the filler concentration, the retention of fiber aspect ra-tio in the final molded product, dispersion and fiber orientation, and the type of sur-face treatment (coupling agent, sizing). Figures 7-4 to 7-8 contain representative dataon the mechanical properties of miscellaneous glass fiber reinforced thermoplastics.The figures have been constructed from replotted data generated by materials sup-pliers and included in ref. [21]. In all cases, the glass used was chopped strands ofE-glass, and was assumed to bear appropriate surface treatments.

Figures 7-4 and 7-5 show the effects of increasing glass concentration on the ten-sile moduli and Charpy impact strengths of several injection-molded engineeringthermoplastics; data on polypropylene, a commodity resin of much lower moduluscommonly reinforced with glass fibers, are also included for comparison purposes.

Figures 7-6 to 7-8 contain data on the f lexural properties and Izod impact strengthsof six engineering thermoplastics. In general, the addition of glass increases the mod-uli and strengths compared to the unfilled resin. Impact strengths are also seen to in-crease, except in the case of the inherently tough PC and PPO resins. It is of interestto note that for certain ductile resins such as polyphenylene ether sulfone (PPSU) andPC-ABS and PC-PBT blends, it has recently been shown that surface coatings withlow adhesion to the matrix can give improved impact strengths at the expense of f lex-ural strength, but without a modulus compromise [22]. For comparison purposes, thedata in Figures 7-6 to 7-8 include figures for combinations of glass fibers with mica

7.7 Applications

142 7 Glass Fibers

Fig. 7-5 Charpy impact strength as a function of glass fiber contentfor several engineering thermoplastics. PP data are included forcomparison.

Fig. 7-4 Tensile modulus as a function of glass fiber content for severalengineering thermoplastics. PP data are included for comparison.

or mineral fillers that are commonly used to control warpage and improve dimen-sional stability. Additional extensive information on the properties of injection-mold-ed glass-reinforced thermoplastics can be accessed via the websites of glass suppliers,resin producers, and compounders, and found in ref. [3].

Fig. 7-6 Comparison of f lexural modulus of glass fiberreinforced engineering thermoplastics. Data for the un-filled resins are included for comparison. (To convertfrom psi to Pa, multiply by 6895).

1437.7 Applications

Fig. 7-7 Comparison of f lexural strength of glass fiberreinforced engineering thermoplastics. Data for the un-filled resins are included for comparison. (To convertfrom psi to Pa, multiply by 6895).

144

Adhesion in polyolefin–glass fiber composites can be increased by suitable reactivemodification of the non-polar matrix. The effect of glass fibers on the properties of apredominantly polyethylene matrix, before and after grafting with maleic anhydride,is shown in Table 7-5 [23]. Maleation of the matrix further enhances its tensile prop-erties, impact strength, and heat distortion temperature (HDT) by virtue of improvedadhesion as a result of reaction of the aminosilane groups present on the glass sur-face with the pendant polymer anhydride groups (see also Chapters 4 and 6). Fig-ures 7-9 and 7-10 show a comparison of the fracture surfaces of unmodified and

7 Glass Fibers

Fig. 7-8 Comparison of impactstrength of glass fiber reinforcedengineering thermoplastics. Data forthe unfilled resins are included forcomparison.

145

maleated composites; the microphotographs clearly show pull-out regions and barefiber surfaces in the case of the unmodified non-polar matrix.

Tab. 7-5 Properties of an injection-molded glass fiber[a] reinforcedpolyolefin[b] before and after maleation of the matrix [23].

Property Unf illed Unmodif ied Maleated unmodif ied matrix matrix matrix + 20% glass + 20% glass

Tensile yield strength, MPa 24.8 – –Tensile break strength, MPa 10.8 36.5 42.1Tensile yield elongation, % 9.2 – –Tensile break elongation, % 23 3.5 4.8Flexural modulus, MPa 1030 2620 2340Flexural strength, MPa 30.3 49.4 51.3Izod impact strength notched, J m–1 48.0 53.4 80.1Izod impact strength unnotched, J m–1 1014 208 272HDT at 1.82 MPa, °C 47 79 84

As an example of the importance of aspect ratio, Table 7-6 contains data on injec-tion-molded 30 wt. % glass fiber–polyamide composites containing fibers of differ-ent diameters [24]. Assuming that the average fiber length is approximately the samein all molded samples, it is clearly evident that increasing fiber diameter (decreasingaspect ratio) results in inferior mechanical properties (with the exception of notched

[a] Fiber glass OCF457AA, Owens Corning.[b] Matrix: polyolefin based on a model recyclable

stream containing at least 80% HDPE, modi-

fied with peroxide/maleic anhydride in a twin-screw extruder.

7.7 Applications

Fig. 7-9 SEM fracture surface showing poor adhesion between theunmodified matrix and glass fibers [23].

146

impact strength). As a result of compounding and molding processes, the fiberlength was reduced from an initial 3–5 mm to less than 0.45 mm. Similar trends werereported at higher glass fiber concentrations (50 wt. % and 63 wt. %). The observedaspect ratio effects are more pronounced in long-fiber reinforced thermoplastics(LFRTP), for which pellets with long initial fiber lengths are produced by pultrusion-type operations rather than extrusion compounding. Longer initial fiber lengths areexpected to yield longer final lengths, and hence higher aspect ratios, in injection-molded composites.

Tab. 7-6 Effect of glass fiber diameter on properties of nylon-6,6containing 30% E-glass fibers [24].

Glass f iber diameter, µmProperty 10 11 14 17

Tensile strength, MPa 184 183 174 164Tensile modulus, GPa 9.73 9.76 9.55 9.50Tensile strain, % 2.83 2.80 2.70 2.49Flexural strength, MPa 287 285 270 249Flexural modulus, GPa 9.20 9.21 9.29 9.01Unnotched impact strength, J m–1 979 894 775 568Notched impact strength, J m–1 130 130 135 139

7 Glass Fibers

Fig. 7-10 SEM fracture surface showing improved adhesion be-tween the maleated matrix and glass fibers [23].

147

References

1 Piggott, M. R., Load-Bearing Fibre Compos-ites, Chapter 3, Pergamon Press, Oxford,England, 1980.

2 Callister, W. D., Jr., Materials Science and En-gineering, An Introduction, 6th Ed., Chapter13, John Wiley & Sons, Hoboken, NJ, 2003.

3 Milewski, J. V., Katz, H. S. (Eds.), Handbookof Reinforcement for Plastics, Chapter 14, VanNostrand Reinhold, New York, 1987.

4 Sheldon, R. P., Composite Polymeric Materi-als, Chapter 1, Applied Science Publishers,Ltd., Barking, Essex, England, 1982.

5 Saint-Gobain Vetrotex technical informa-tion; http://www.vetrotexna.com/busi-ness_info/gstrand.html#11

6 Saint-Gobain Vetrotex technical informa-tion;http://www.vetrotextextiles.com/pdf/RGlass%20DS2000.pdf

7 Michigan State University, Intelligent Sys-tems Laboratory;http://islnotes.cps.msu.edu/trp/rtm/siz_basc.html

8 See ref. [2], Chapter 16, and Appendix B.9 Hohenberger, W., Chapter 17 in Zweifel,

H., (Ed.), Plastics Additives Handbook,Hanser Publishers, Munich, 2001.

10 Freedonia Group, “Reinforced Plastics to2007” market report, April, 2003; summaryaccessed via http://freedonia.ecnext.com

11 Wypych, G., Handbook of Fillers, ChemTecPubl., Toronto, Ont., Canada, 2000, pp187–188.

12 Potters Industries, Inc., Technical Informa-tion on Conduct-O-Fil® Silver-Coated GlassFibers, Valley Forge, PA.

13 European Glass Fibre Producer Associa-tion, “Continuous Filament Glass Fibre andHuman Health”, APFE Publication, March2002, Brussels.

14 Cullen, R. T., et al., Inhalation Toxicology2000, 12, 959–977.

15 Shannon, H. S., et al., Annals OccupationalHygiene 1987, 31(4B), 657–662.

16 Goldsmith, J. R., Am. J. Ind. Med. 1986, 10,543–552.

17 Marsh, G. M., et al. J. Occupational Medicine1990, 32, 594–604.

18 Boffetta, P., et al., Scand. J. Work, Environ-ment, and Health 1992, 18, 279–286.

19 Infante, P. F., et al., Am. J. Ind. Med. 1994,26, 559–584.

20 PPG Industries, Inc., Material Safety DataSheet for “Fiber Glass Continuous Fila-ment”, revised Feb. 2004.

21 CAMPUS® Computer-Aided Material Pre-selection by Uniform Standard;http://www.campusplastics.com/.

22 Galluci, R. F., Proc. 62nd SPE ANTEC,2004, 50, 2718.

23 Xanthos, M., et al., Polym. Compos., 1995,16(3), 204.

24 BASF Corp., “An Advanced High Modulus(HMG) Short Glass Fiber Reinforced Ny-lon-6: Part I”, technical article available athttp://www.basf.com/PLASTICSWEB/dis-playanyfile?id=0901a5e180086583.

7.7 Applications

8Mica Flakes

Marino Xanthos

8.1Background

Mica is a term for a group of more than 35 phyllosilicate minerals with a layered tex-ture and perfect basal cleavage. This perfect cleavage, due to weak bonding betweenthe layers, results in splitting or delamination of the mica layers into thin sheets. Mi-cas compose roughly 4% of the Earth’s crustal minerals and are common in all threemajor rock varieties, i.e. igneous, sedimentary and metamorphic [1]. Micas as a groupare variable in chemical composition and in physical and optical properties. They arebasically complex potassium aluminosilicates with some aluminum atoms replacedby magnesium and iron, and may contain minor amounts of a variety of other ele-ments. Muscovite and phlogopite, the most important commercial types, haveunique characteristics such as chemical inertness, superior electrical and thermal in-sulating properties, high thermal stability, and excellent mechanical properties.

Micas are used in sheet and ground forms. High quality sheet mica is used princi-pally in the electronic and electrical industries. Built-up mica produced by mecha-nized or hand setting of overlapping splittings and alternate layers of binders andsplittings, and reconstituted mica (mica paper) are primarily used as electrical insu-lation materials. Commercial micas are divided into “wet ground” and “dry ground”depending on the method of production. In addition to its more recent widespreaduse as a functional filler for plastics, dry-ground mica has several other applications.It is used in tape-joint cement compounds for gypsum dry wall, in the paint industryas a pigment extender, in the well drilling industry as an additive to drilling muds, inthe rubber industry as a mold-release compound, and in the production of rolledroofing and asphalt shingles. Wet-ground mica, which retains the brilliance of itscleavage faces, is mostly used in pearlescent paints and in the cosmetics industry[1,2].

Since the early days of the development of phenolic molding compounds for elec-trical applications, mica has been extensively used as a filler of choice. It was not un-til the late 1960s/early 1970s that the realization of the importance of the aspect ratioand interfacial adhesion in platelet/f lake-containing polymers initiated R & D efforts


150

at Canadian Universities and industrial laboratories that led to the commercializationof grades suitable as reinforcements for plastics. Early work was carried out at theUniversity of Toronto, Toronto, Ontario [3–6], Fiberglass Canada, Sarnia, Ontario [7],and Marietta Resources International, Boucherville, Québec; in 1975, the latteropened a plant that now has a capacity of 30,000 tons per annum, producing high as-pect ratio (surface-treated and untreated) phlogopite grades, in addition to tradition-al mica products [8]. During that same period, fundamental work on the mechanicsof f lake-reinforcement resulted in the development of predictive equations for mod-ulus, strength, and toughness of f lake-reinforced composites (see Chapter 2), which,in most cases, were confirmed by experiments. Research efforts from that periodwere reviewed by Woodhams and Xanthos in 1978 [9].

In the mid-1970s/early 1980s, the potential of high aspect ratio (HAR) mica as a re-inforcement in a variety of thermoplastics and thermosets, and the parameters af-fecting its performance, were described in a series of technical bulletins from Mari-etta Resources International and presentations at Society of Plastics Engineers andSociety of Plastics Industry conferences (see, for example, ref. [10]). This early work,complemented by publications from Canadian and United States University re-searchers and Canadian Government laboratories, provided significant informationon the understanding of:

the coupling differences between muscovite and phlogopite, particularly inpolypropylene;

the rheological characteristics of f lake-containing melts; the importance of f low-induced f lake orientation in relation to properties and

weld-line strength; the differences between f lake reinforcement and reinforcement with fibers and ir-

regular fillers.

During that period, it also became clear that, as for fragile glass fibers, retention off lake aspect ratio and f lake orientation (prerequisites for effective reinforcement)strongly depend on the type of polymer (thermoplastics vs. thermosets) and the pro-cessing/shaping method. It also became obvious that retention of high aspect ratiofor large diameter f lakes would only be favored by low-shear processing methods ap-plicable to thermosets; these would ensure planar orientation of the f lakes withouthaving to consider the complications arising from the f low of filled thermoplasticmelts in circular channels and irregularly shaped dies. It is unfortunate that, even atthe present time, high aspect ratio, smaller diameter, thin f lakes that would be lesssusceptible to mechanical breakdown are not generally available. The status of the de-velopments in mica-reinforced plastics during the aforementioned period was re-viewed by Hawley in 1987 [8]. It should be noted that the principles of f lake rein-forcement developed on mica more than 20 years ago are directly applicable to themuch higher aspect ratio montmorillonite-based organoclay reinforcements intro-duced in the last decade as plastics reinforcements (see also Chapter 9).

The multiple functions of mica have been outlined in Chapter 1 of this book, alongwith an example of its role in the search for multifunctional fillers for polypropylene

8 Mica Flakes

151

compounds for automotive applications. Mica-reinforced thermoplastics, such aspolypropylene, polyethylene, nylon and polyesters, are now established in a variety ofautomotive applications and consumer products where mica supplements or re-places glass fibers and other mineral fillers. The wider use of mica in many applica-tions has been limited by low impact strength and low weld-line strength in certainplastics. These issues are the focus of continuing R & D efforts by materials suppli-ers and compounders/molders.

In addition to its primary function as a high aspect ratio mechanical property en-hancer, mica is also used as a modifier of electrical properties and as an importantcomponent of sound-deadening formulations; it is also used for reducing permeabil-ity, improving dimensional stability, and as a modifier of optical properties. The mul-tiple functions of mica are compared with those of other fillers in Table 1-4.


Mica occurs worldwide, with large deposits in the Unites States, Canada, France, Ko-rea, Malaysia, Mexico, Russia, Madagascar, and India, with smaller deposits in someEuropean countries. Muscovite mica, the most common form, is found in acidic ig-neous rocks such as granites and also forms very large “books” in pegmatites. Inmetamorphic rocks, muscovite occurs in lower grades of purity. Phlogopite, the sec-ond most common form of mica, is found in ultrabasic igneous rocks, which ref lectsits high magnesium content. Biotite, the lesser of the micas in terms of commercialimportance, contains more iron than magnesium, is brown to black in color, and isfound in granites and intermediate igneous rocks [1].

Mining mica is typically accomplished through quarrying, although it is occasion-ally feasible to use underground mining methods. In producing large mica sheets forelectronic and high-temperature applications, the fragility of mica limits the feasiblemethods that can be used for its extraction. Mica blocks are split into thin slices andindividually assessed, trimmed, and sorted by size, color, and quality.

Mica is relatively easy to cleave while in coarse f lake form, but as grinding proceedsbreakage perpendicular to the cleavage plane supersedes delamination. Thus, theproduction of high aspect ratio, well-delaminated, small-sized f lakes is a challenge.The production of ground mica generally involves steps that are determined by thepurity of the host ore. Preliminary size reduction and purification from other min-erals may involve various f lotation steps, magnetic separation, f lotation cells, and hy-drocyclones or air-table separation. The actual dry grinding may involve impactprocessors such as rotor mills, high-speed hammer and cage mills, and pin mills. Inthe SuzoriteTM process [8], using ore of high purity, i.e. with up to 90% phlogopite,deposit, impurities such as feldspar and pyroxene are removed after primary milling.Further milling takes place in closed systems containing air classifiers that removewell-delaminated f lakes and return thicker f lakes for reprocessing. Finer f lakes canbe made by feeding these materials into f luid energy mills. A variety of other meth-ods have been proposed for separating high aspect ratio f lakes [4,8].


152

Wet grinding may also be employed, although at higher cost and with lower capac-ity than in the case of dry grinding. Wet grinding is still performed batchwise in chas-er mills, where rollers revolve on the surface of the mica with a smearing action forseveral hours. The heavier thick f lakes settle out, whereas the well-delaminated micaoverf lows into settling tanks. Other wet-grinding methods (log mills, vibro-energymilling, high-pressure water jets, and ultrasonics) and techniques for separating highaspect ratio f lakes (e.g., sedimentation, water elutriation) are reviewed in refs. [8,11].


Commercial micas are only available in the muscovite and phlogopite forms. Theircompositions are subject to variation due to isomorphous substitutions. The basicmica structure is a sandwich, where the outer layers are silica tetrahedra in whichsome of the silicon atoms have been substituted by aluminum and the middle layerconsists of aluminum, magnesium, iron, and f luorine plus hydroxyl groupsarranged in an octahedral fashion. Muscovite, KAl2(AlSi3O10)(OH)2, has a gibbsite,Al(OH)3, structure sandwiched between the silica sheets. It is water-white with apinkish or greenish hue (Figure 8-1). Commercial phlogopite mica, K(Mg, Fe)3(Al-Si3O10)(OH, F)2, is dark brown to black in color depending on the iron content. Themiddle layer of the sandwich is brucite, Mg(OH)2, with iron substitution in bothFe(II) and Fe(III) forms. These three-layer units are about 100 nm thick, and areloosely held together by potassium ions in 12-fold coordination with the oxygenatoms, so that the interlayer forces are rather weak and permit cleavage. In Figure 8-2, the structure of mica is compared with those of the other phyllosilicate fillers kaoli-nite and montmorillonite.

Chemical analyses of typical commercial muscovites and phlogopites are given inTable 8-1. Table 8-2 summarizes typical physical and mechanical properties. As ex-pected, data vary depending on the mineral source, sample type and size, and methodof property measurement. Certain micas expand upon heating to temperatures above600 °C. Muscovites generally lose their combined water at lower temperatures. Char-acteristic properties of both muscovite and phlogopite are high modulus (more thandouble that of glass fibers; 172 vs. 70 GPa) and high strength, low coefficient of ther-mal expansion, high thermal conductivity and temperature resistance, good dielectricproperties, low hardness, low coefficient of friction, and good chemical resistance.

A key characteristic of mica that is not apparent from Table 8-2 is its planar isotropydue to its plate-like nature. The listed properties are similar in both the “x” and “y”directions in the plane (but not always in the “z” direction), giving rise to the isotrop-ic properties characteristic of oriented mica composites (see Chapter 2). Thus, unlikefibers, mica reinforces equally in the two directions in the plane.

The morphological features of ground micas vary depending on the method usedfor their delamination. Dry-ground mica f lakes are not of even thickness but arestepped due to uneven delamination, which also results in feathered edges. Figure 8-3 shows sharp edges, a broad size distribution, and incomplete delamination for a

8 Mica Flakes

153

commercial dry-ground mica; irregular surfaces are also shown in Figure 1-2 ofChapter 1 for ultrasonically delaminated f lakes. In contrast, the wet-grinding processpolishes the f lakes to an even thickness and rounds off the edges (see Figure 1-5),usually producing a high aspect ratio product. Thus, wet-ground mica has more lu-bricity and has a higher “sparkle” than the dry-ground product.

Natural micas have a very low ion-exchange capacity as compared to the bentoniteclays described in Chapter 9 and those sites that exchange ions are all on the outersurfaces. However, the reactive surface groups of micas, some appearing on the facesafter delamination, are amenable to treatment with a variety of additives that may im-prove dispersion/adhesion in a variety of polar or nonpolar polymeric matrices. De-tails of additives and methods used for surface treatments are included in refs. [8,18].For polar polymers (nylons, thermoplastic polyesters, and polyurethanes), aminosi-lanes and aminostyrylsilanes have been found to be efficient. With nonpolar poly-mers such as polyolefins, appropriate coupling is needed to increase strength values


Fig. 8-1 Diagrammatic sketch of the structure of muscovite [12].

154

above those of talc or calcium carbonate filled systems and closer to those of glassfiber filled systems. Azidofunctional and aminostyrylsilanes have been found to beparticularly effective for PP homopolymers. Maleic anhydride or acrylic acid func-tionalized polyolefins, often in combination with aminosilanes, are lower cost, effec-tive coupling agents [19], and are discussed in more detail in Chapter 6 of this book.The effect of acid number of anhydride-based coupling agents on mica-filled PP hasbeen reviewed in ref. [20]. Low MW chlorinated polyalkenes are also a low-cost alter-native to silanes and are discussed in Chapter 6 of this book and in ref. [8]. Other ex-amples of less conventional, but highly effective surface treatments, also discussed inChapter 6, include bismaleimides, which were shown to increase the tensile strengthof PP/40% mica injection-molded samples by 25% [21]. The response of micas to avariety of adhesion promoters, including silanes, titanates, functionalized polyolefinsand others has been reviewed in refs. [18,22].

8 Mica Flakes

Fig. 8-2 Comparison of the structure of mica with those of kaoliniteand montmorillonite. Adapted from ref. [13].


Tab. 8-1 Chemical analysis of commercial phlogopite andmuscovite micas [8,14,15].

Phlogopite[a] Muscovite[b] Muscovite[c]

SiO2 40.7 47.9 45.6Al2O3 15.8 33.1 33.1MgO 20.6 0.69 0.38FeO 7.83 – –Fe2O3 1.21 2.04 2.48K2O 10.0 9.8 9.9Na2O 0.1 0.8 0.6BaO 0.5 – –CaO tr. 0.5 0.2TiO2 0.1 0.6 tr.Cr2O3 – – –MnO tr. tr. tr.F 2.16 – –P tr. 0.03 tr.S tr. 0.01 tr.H2O combined 1.0 4.3 2.7H2O free 0.01 0.1 0.25

[a] Suzorite Mica Products.[b] The English Mica Co.[c] Inderchand Rajgarchia & Sons (P) Ltd.

Tab. 8-2 Properties of muscovite and phlogopite micas [8,14–17].

Property Muscovite Phlogopite

color white, off-white, amber, yellow, ruby, green light brown

crystal structure monoclinic monoclinichardness, Mohs 3–4 2.5–3.0density, g cm–3 2.7–3.2 2.75–2.9pH, aqueous slurry 6.5–8.5 7.5–8.5water solubility trace tracerefractive index 1.55–1.61 1.54–1.69tensile modulus, GPa 172 172tensile strength, MPa 255–296[a] 255–296[a]

3100[b]

690–900[c] 690–900[c]

linear coefficient of thermal expansion, per °C(perpend. to cleavage) 15–25 × 10– 6 –(parallel to cleavage) 8–9 × 10– 6 13–15 × 10– 6

chemical resistance very good gooddielectric constant at 104 Hz 2.0–2.6 5.0–6.0max. temp. with little or no decomposition, °C 500–530 850–1000thermal conductivity, W m–1 K–1 2.5 × 10–5 2.5 × 10–5

[a] Measured on sheets with stressed edges [8].[b] Measured on sheets with edges unstressed [8].

[c] Calculated for the effective strength of highaspect ratio f lakes in plastics [8].

156

8.4Suppliers

As of June 2004, 27 suppliers were identified worldwide [23], producing different mi-ca products. Major United States producers include Azco Mining, Inc., EngelhardCorp., Franklin Industrial Minerals, Minerals Technologies, Inc., Pacer Corp., Whit-taker Clark & Daniels, Inc., and Zemex Corp. Zemex Corp. operates the Suzorite phl-ogopite plant in Boucherville, Québec, that processes ore from a deposit located inSuzor Township, Québec, and which is described as the largest known phlogopitebody in the world [1]. European producers include Acim Jouanin S.A. (France), Co-gebi (Belgium), Kemira Pigments Oy (Finland), Microfine Minerals Ltd. (UK), Sig-mund Lindner GmbH (Germany), WTL International Ltd. (UK), and Ziegler & Co.,GmbH (Germany). Several Indian producers are mostly specializing in mica sheetsrather than ground products.

According to the U.S. Geological Survey “Minerals 2002” [2], ten plants produceddry-ground mica in the states of NC, GA, AZ, SC, NM, and SD, while four plants pro-duced wet-ground mica in the United States to a total of 98,000 tonnes. Imports fromCanada amounted to about 15,000 tonnes. In 2002, consumption of dry-ground micain plastics applications in the USA accounted for 5% of the total market of about100,000 tonnes [2]. U.S. production increased to about 107,000 tonnes in 2003, most-ly as a result of increased production in GA, NM, and SD. This is about one-third ofthe worldwide mine production, the latter estimated to be about 300,000 tonnes in2003. Countries with the highest levels of mine production, and correspondinglylarge estimated reserves, are the United States and Russia, followed by the Republicof Korea, Canada, Brazil, and India, among others [24].

8 Mica Flakes

Fig. 8-3 SEM microphotograph of commercial dry-ground mica.Courtesy of Dr. S. Kim, Polymer Processing Institute, Newark, NJ.

157


In 2003, average prices for U.S. mica ranged from $200/tonne for dry ground to$1000/tonne for wet ground [24]. Micronized mica was quoted at $535–930/tonne[25]. Prices for surface-treated grades vary and are usually more than twofold higherthan those for their untreated counterparts; however, the profit margin for suchgrades decreases as more compounders are using non-silane reactive compounds orfunctionalized polymers as adhesion promoters with untreated grades. Grades suit-able for plastics compounding can be phlogopite or muscovite, dry ground or wetground, untreated or surface-treated, and used with polar or nonpolar polymers.Typical sizes for grades suitable for plastics applications are –40+100 mesh,–60+120 mesh, –100 mesh, and –325 mesh (below 45 µm). Flake thickness dependson the degree of delamination. Average aspect ratios, although not always specified,may range from 10:1 up to 100:1 depending on the aspect ratio definition (usuallyequivalent diameter over thickness) and the method of measurement (see, for exam-ple, refs. [3,4,26]). Grades for paints and pearlescent pigments are usually 85–95%–325 mesh wet ground or micronized.


Finely divided mica is generally considered as a nuisance dust with an applicable OS-HA PEL of 3 mg m–3 and a respirable ACGIH TLV of 3 mg m–3 (TWA, 8 hour peri-od). This is valid for mica containing less than 1% crystalline silica [27,28]. Mica itselfis not listed as a carcinogen by OSHA, NTP, or IARC. However, crystalline silica thatmay be present as a contaminant is classified as carcinogenic to humans with anACGIH TLV (respirable) of 0.05 mg m–3 and an OSHA PEL (respirable) of0.1 mg m–3. Crystalline silica levels in Suzorite phlogopite mica (CAS No.: 12001-26-2) may vary in the range 0.1 to 1% [27]. Mica with less than 1% silica is considered anuncontrolled product according to the Canadian WHMIS. Mica is unreactive andnonf lammable, and is not classified as hazardous waste. It meets FDA criteria cov-ering its safe use in articles intended for food contact use and is listed in the U.S.Code of Federal Regulations, Title 21, parts 175 and 177, under “indirect food addi-tives” [27].

8.6 Environmental/Toxicity Considerations

158

8.7Applications

8.7.1General

Processing methods for mica-filled thermoplastics include extrusion, injection-molding, thermoforming, structural foam molding, blow-molding, and rotomolding.Mica may be incorporated into thermoplastics by twin-screw extruder melt com-pounding or, with certain polymer types, particularly in powder form, by direct in-jection-molding of dry blends. In general, the free-f lowing mica f lakes disperse inmolten resins more easily than fillers containing aggregates that need to be brokenor glass fibers that need to be separated into filaments. The relatively low Mohs hard-ness of 2.5–3.0 results in lower abrasion and wear to metallic equipment than withglass fibers. Polypropylene/mica compounds have found significant uses in the au-tomotive industry for such products as under-the-hood components, trim, dash-board, and grille opening panels. Uses in other thermoplastics include HDPE pack-aging films, rotomolded HDPE and LLDPE tanks, polyolefin structural foam for au-tomotive parts, speaker cabinets, and thermoplastic polyesters for automotive appli-cations such as distributor systems.

In thermosets, traditional processing methods were casting (as for epoxies) andcompression/transfer molding (as for phenolic compounds) or compression mold-ing/lamination as in the case of impregnated mica paper. In new applications, micais being incorporated into certain hybrid glass fiber/unsaturated polyester compos-ites for marine, automotive, and household applications produced by spray-up, lay-up, and combinations thereof. It is also finding use in automotive applications inRTM and reinforced RIM polyurethanes, where it is added to the polyol componentreplacing part of the milled glass. Applications of micas in thermoplastics and ther-mosets have been reviewed in ref. [8].

8.7.2Primary Function

The primary function of mica that has led to significant applications in automotiveand other industries is modification and improvement of mechanical properties.General effects are significant increases in modulus, which in most cases is inde-pendent of the degree of interfacial adhesion but is still dependent on orientation;usually an increase in tensile and f lexural strength, the effects being strongly de-pendent on the degree of adhesion and extent of orientation. Elongation usually de-creases. Often, mica’s outstanding performance in improving stiffness is unsatisfac-torily offset by reduced impact strength, the latter depending on the type of test(notched vs. unnotched, falling dart), the type of polymer, f lake orientation, size, load-ing level, and interfacial adhesion. Thermomechanical properties such as heat dis-tortion temperature and creep resistance generally improve upon the addition of mi-ca, the effects being strongly dependent on adhesion [8].

8 Mica Flakes

159

As discussed in Chapter 2, for most processing methods for thermoplastics, thef low-induced orientation of f lakes is predominantly parallel to the f low direction,with a region of misalignment in the core. In injection-molding, such morphologiescan be modified through the application of shear to the melt as it cools (e.g.SCORIM™), which has a marked effect on orientation and physical properties. Inmica-reinforced thermosets, processes such as compression molding and laminationensure mostly planar orientation (Figure 1-2 in Chapter 1) and isotropic properties asdiscussed in Chapter 2. The issue of reduced weld-line strength due to the resultingunfavorable f lake orientation when two f low fronts meet may be mitigated by ap-propriate selection of the process conditions (injection speed, melt temperature), mi-ca concentration, and mica size [29,30].

Comparisons of the effects of mica, glass fibers, talc, and calcium carbonate on theproperties of a polyolefin matrix (unmodified and chemically modified by maleation)are shown in Tables 16-1 and 16-2 of this book [31]. Even after normalizing for equalloadings, the effects on mechanical properties generally follow the sequence glassfibers > mica f lakes > talc > calcium carbonate. Table 8-3 shows the effect of pre-treat-ment with a sulfonylazidosilane (now discontinued) on the properties of injection-molded 40%-filled mica compounds and a comparison with commercially available30% glass fiber compounds. The possibility of approaching the mechanical proper-ties of glass fiber compounds through selection of the appropriate surface treatmentis clear.

Tab. 8-3 Comparison of properties of untreated and azidosilane-treated 40% mica/PP injection-molding composites with those ofcommercial 30% glass fiber/PP and unfilled PP [8,10,18].

Property Unf illed 40% 40% 30% PP Untreated Azidosilane- Glass

Mica treated Mica f ibers

Tensile strength, MPa 32.9 28.7 43.4 44.1Flexural strength, MPa 31.5 45.5 66.5 70.7Flexural modulus, GPa 1.26 6.51 7.70 6.51Izod impact strength notched, J m–1 24.0 32.0 34.7 74.8Izod impact strength unnotched, J m–1 No break 203 235 502HDT at 1.85 MPa, °C 56 89 108 125

8.7.3Other Functions

The mechanical properties of ternary PET composites containing 40–45% total glassfibers and mica are shown in Figures 7-6 to 7-8 of this book and are compared withthose of composites containing 30% glass. Flexural modulus is higher for theglass/mica composites than for the all-glass composites; however, the addition of mi-ca decreases f lexural strength and impact strength. In such composites, mica acts asa multifunctional filler, controlling warpage and improving dimensional stability. Mi-

8.7 Applications

160

ca reduces thermal expansion and shrinkage, the effects becoming more pronouncedwith increasing filler concentration and increased interfacial interactions. Such ef-fects have been discussed earlier in relation to injection-molded PP/mica containingPP-g-MA and more recently in relation to rotomolded MDPE/mica and LLDPE/micacontaining PE-g-MA [32,33]. Warpage resulting from residual stresses and differen-tial shrinkage can be high in fiber-reinforced thermoplastics, but is reduced by theplanar orientation of mica f lakes. Combinations of glass (10–15%) and mica(20–30%) are often used as a best compromise between shrinkage and warpage andmechanical properties. For example, a commercial mica/glass fiber PET compoundhas a notched Izod impact strength of 74.7 J m–1 and a relative warpage measured onan annealed disc of 5; the corresponding values for an all-glass fiber compound are128 J m–1 and 125, respectively [8].

Mica f lakes embedded in a polymer and properly oriented in a plane can provide atortuous path to vapors and liquids, similarly to the natural composites shown in Fig-ure 1-1. Barrier properties can be imparted in blow-molded containers, packagingfilms, and corrosion-resistant coatings not only by mica but also by other imperme-able lamellar fillers, including glass f lakes, talc, and nanoclays. In blown LDPE film,the addition of 10% mica was found to reduce the oxygen permeability from 4.16 to3.03 Barrer [34]. Assuming an impermeable, fully oriented lamellar filler, Eq. (8-1)[35] may be used to predict the composite permeability, Pc, perpendicular to the fillerplane as a function of the matrix permeability, Pm, filler volume fraction, Vf , matrixvolume fraction, Vm, and filler aspect ratio α:

Pc/Pm = Vm/tf (8-1)

where the factor

tf = 1 + (α/2) Vf (8-2)

represents tortuosity (actual path of solvent or gas over film thickness).In addition to particle aspect ratio and concentration, permeability is affected by

several other factors, including polymer crystallinity, particle orientation, and adhe-sion. Modified equations to accommodate misalignment effects and the presence ofthe (assumed) impermeable-to-oxygen crystalline phase of HDPE in mica/HDPEfilms have been proposed [34].

Mica can also modify optical properties in semicrystalline polymers by acting as anucleating agent and as a substrate for oxide deposition in pearlescent pigments pro-duced by platelet core-shell technologies as discussed in Chapter 1. Figure 1-5 in thisbook shows a cross-section of an anatase/mica pigment particle produced by thistechnology.

An additional function of mica may be manifested in damping applications. Fillersmay introduce a broadening in the damping (tan δ = G’’/G’) transition region of apolymer, shifting it to longer times or higher temperature. The broadening of thetransition region may be useful in vibration-damping and sound-deadening materi-als. Flake-filled elastomers and plastics often have high mechanical damping and for

8 Mica Flakes

161

this reason vibration-damping materials may contain f lakes that facilitate the con-version of the energy of vibration into heat rather than emitting it to the air. Part ofthis damping may result from one layer of a f lake such as mica or graphite slidingover another layer when the material is deformed [36]. For a specific polymer system,it has been shown that mica f lakes are more effective than talc, CaCO3, or TiO2, withdamping increasing with filler concentration [37].

References

References

1 Natural Resources Canada, Minerals andMining, Statistics On-Line, “Mica”, ac-cessed at http://mmsd1.mms.nrca.gc.ca

2 Hedrick, J. B., “Mica”, Minerals Yearbook2002, United States Geological Survey, ac-cessed at www.mineralsusgs.gov

3 Lusis, J., et al., Polym. Eng. Sci. 1973, 13(2),139.

4 Kauffman, S. H., et al., Powder Technol.1974, 9, 125.

5 Woodhams, R. T., US Patent 3,799,799,1974.

6 Woodhams, R. T., Xanthos, M., US Patent4,112,036, 1978.

7 Maine, F. W., Shepherd, P. D., Composites1974, 193.

8 Hawley, G. C., “Flakes”, Chapter 4 in Hand-book of Reinforcement for Plastics (Eds.: Katz,H. S., Milewski, J. V.), Van Nostrand Rein-hold, New York, 1987.

9 Woodhams, R. T., Xanthos, M., Chapter 20in Handbook of Fillers and Reinforcements forPlastics (Eds.: Katz, H. S., Milewski, J. V.),Van Nostrand Reinhold Co., New York,1978.

10 Xanthos, M., et al., Proc. 35th SPE ANTEC,1977, 23, 352.

11 Hawley, G. C., “Mica”, in Pigment Hand-book: Vol. 1: Properties and Economics (Ed.:Lewis, P. A.), John Wiley & Sons, New York,1988, pp. 227–256.

12 Grimm, R. E., Applied Clay Mineralogy, Mc-Graw-Hill, New York, 1962, p. 23.

13 Kingery, W. D., Introduction to Ceramics,John Wiley & Sons, New York, 1967, pp.130–131.

14 ZEMEX Industrial Minerals, Boucherville,Québec, Canada, Technical information on“Suzorite mica – products and applica-tions”.

15 Inderchand Rajgarchia & Sons (P) Ltd.,“Physical properties of mica”, accessed athttp://www.icrmica.com

16 Engelhard Corp., Hartwell, GA, USA, tech-nical information on mica.


18 Hawley, G. C., Proc. Coupling Agents andSurface Modifiers ’99, Intertech Corp., At-lanta, GA, September 22–24, 1999.

19 Xanthos, M., Polym. Eng. Sci. 1988, 28,1392.

20 Olsen, D. J., Hyche, K., Proc. 47th SPE AN-TEC, 1989, 35, 1375.

21 Xanthos, M., Plast. Rubber Proc. & Appl.1983, 3(3), 223.

22 Canova, L. A., Proc. 58th SPE ANTEC, 2000,46, 2211.

23 Global Industry Analysts, Inc., “Mica”, Mar-ket research report, 06/2004; Summary ac-cessed at http://www.the-infoshop.com

24 Hedrick, J. B., “Mica (natural), Scrap andFlake”, Mineral Commodity Summaries,U.S. Geological Survey, January 2004;accessed at www.mineralsusgs.gov

25 Industrial Minerals Prices, Ind. Minerals,Feb. 2004, pp 72–73.

26 Canova, L. A., Proc. 45th Ann. Conf., Com-pos. Inst., SPI, Feb. 12–15, 1990, 17-F, 1–5.

27 Suzorite mica products, Inc., Boucherville,Québec, Canada, MSDS Suzorite Mica,September 2003.

28 NIOSH, The Registry of Toxic Effects ofChemical Substances, “Silicate, mica” ac-cessed at http://www.cdc.gov/niosh/rtecs

29 Ferro, J. P., Proc. Functional Fillers for Plas-tics 1995, Intertech Corp., Houston, TX,USA, December 4–6, 2003.

162 8 Mica Flakes

30 Dharia, A., Rud, J. O., Proc. 60th SPE AN-TEC, 2002, 48, 2093.

31 Xanthos, M., et al., Polym. Compos. 1995,16(3), 204.

32 Robert, A., et al., Proc. 58th SPE ANTEC,2000, 46, 1399.

33 Robert, A., Crawford, R. J., Proc. 57th SPEANTEC, 1999, 45, 1478.

34 Xanthos, M., et al., Internat. Polym. Process.1998, 13(1), 58.

35 Nielsen, L. E., J. Macromol. Sci., 1967, A1,926.

36 Nielsen, L. E., Landel, R. F., MechanicalProperties of Polymers and Composites, Chap-ter 8, 2nd Ed., Marcel Dekker, Inc., NewYork, 1994.

37 Wypych, G., Handbook of Fillers, ChemTecPubl., Toronto, Ontario, Canada, 2000, pp.112–115 and 807–808.

9Nanoclays and Their Emerging Markets

Karl Kamena

9.1Introduction

9.1.1Clays, Nanoclays, and Nanocomposites

“Nanoclay” is the term generally used when referring to a clay mineral with a phyl-losilicate or sheet structure with a thickness of the order of 1 nm and surfaces of per-haps 50–150 nm in one dimension. The mineral base can be natural or synthetic andis hydrophilic. The clay surfaces can be modified with specific chemistries to renderthem organophilic and therefore compatible with organic polymers. Surface areas ofnanoclays are very large, about 750 m2 g–1. When small quantities are added to a hostpolymer, the resulting product is called a nanocomposite.

Nanoclays and nanocomposites have generated a tremendous amount of researchinterest and curiosity, and it is estimated that hundreds of millions of dollars havebeen invested globally in order to investigate relevant technologies and products.Commercialization has not been rapid to date, but realistically the understanding anddevelopment of this concept to and through product stages is detailed and time-con-suming. It is likely that nanoclays and nanocomposites will continue to satisfy nicheapplications and markets, and will begin to grow in substantial volume incrementsas producers and users grow more comfortable with this developing technology.

9.1.2Concept and Technology

The nanocomposite concept appears to have its origin in pioneering research con-ducted in Japan by Unitika Ltd. in the 1970s [1] and separately by Toyota Central Re-search and Development Laboratories in the late 1980s [2]. The theory was that if nan-oclays could be fully dispersed or exfoliated to high aspect ratio platelets in polymersat relatively low levels (2–5 wt. %), a number of mechanical and barrier propertieswould be enhanced. The original work at both Unitika and Toyota CRDL was based


164

on an in situ process for the preparation of nylon-6 nanocomposites. According to thismethod, a nanoclay is introduced into the caprolactam monomer stage of the process,and the caprolactam is intercalated into the clay galleries. Under appropriate reactorconditions, the caprolactam polymerizes and the platelets are further expanded andbecome exfoliated to become an integral part of the bulk polymer [3,4]. Toyota re-ported that NCH (nanocomposite nylon-6/clay hybrid) materials provided signifi-cant improvements in mechanical, thermal, and gas barrier properties at 2–5 wt. %loadings of montmorillonite. Toyota CRDL has also prepared nylon-6/clay nanocom-posites (NCC) by melt-compounding techniques. Typical mechanical properties arelisted in Table 9-1 [2].

Tab. 9-1 Tensile properties and impact strengths of nylon-6nanoclay materials [2].

Specimen Type Tensile Strength Tensile Modulus Charpy Impact (Montmorillonite, wt. %) (MPa) (GPa) Strength (kJ m–2)

NCH-5 (4.2) 107 2.1 2.8NCC-5 (5.0) 61 1.0 2.2Nylon-6 (0) 69 1.1 2.3

Other methods to prepare nanocomposites include a solvent-assisted process,whereby a co-solvent is employed to help carry the monomer into the galleries and issubsequently removed from the polymer system, and direct polymer melt intercala-tion methods, which involve the direct addition of nanoclays to a polymer melt undershear conditions at elevated temperatures, allowing their direct exfoliation into thepolymer [5].

Following on from the Unitika and Toyota CRDL work, there has been a largeamount of investigation in many industrial and academic environments, and muchof the effort has been targeted at achieving exfoliation in a technologically and eco-nomically feasible manner [6]. During the period from November 2000 to September2002, there were 11 international conferences devoted specifically to developments innanocomposites, with over 350 papers presented by academic, government, and in-dustry researchers [7].


9.2.1Raw and Intermediate Materials

Many nanoclays are based on the smectite clay known as montmorillonite, ahydrated sodium calcium aluminum magnesium silicate hydroxide, (Na,Ca)(Al,Mg)6(Si4O10)3(OH)6·nH2O. Montmorillonite is found throughout the world in smallquantities in its natural geological state. In large deposits, where the mineral is found

9 Nanoclays and Their Emerging Markets

165

in greater than 50% concentrations admixed with a variety of other minerals, it isknown as bentonite. Commercially attractive deposits of bentonite are located inmany geographical areas, ranging from the United States (particularly in Wyoming)to Western Europe, the Middle East, China, etc.

Natural montmorillonite clays are most commonly formed by the in situ modifica-tion of volcanic ash resulting from volcanic eruptions in the Pacific and WesternUnited States during the Cretaceous period (85–125 million years ago). Opinions dif-fer concerning the process and time of modification of the ash to clay. Certainly thechange began with contact with water. The instability of the ash made for ease of dis-solution and reaction with the available marine chemistry. Probably the single mostimportant factor in the formation of the clay was the availability of sufficient magne-sium in the marine sediment environment. Ensuing chemical and structuralchanges took place throughout the deposits’ entire geological history. It is estimatedthat the resulting deposits in Wyoming, for example, alone contain over 1 billion tonsof available clay.

Geological maps are available from areas around the globe where clay deposits havebeen detected. Such maps and associated area photographs and topographical mapsassist in guiding the decision for exploration drilling. Such activity is termed “borehole” drilling and its first step is to drill bore hole samples on centers covering areasof 50–300 foot (15–90 meters). The use of global positioning satellite (GPS) tech-niques ensures accurate surveying to within a few centimeters.

Generally speaking, in characterizing a deposit the prospector/producer will lookfor the following information:

purity, crystallography, chemistry, particle size, morphology, charge, dispersion characteristics, response to processing parameters.

If the decision is to mine, the results from the bore hole tests guide the miners andheavy equipment operators in removing the clay from the earth and depositing it instockpiles. Depending on the ash fall volume, the depth of a deposit can be from afew centimeters to several meters and the length up to hundreds of meters. The borehole data give the profile of the deposit, and the mining techniques follow accord-ingly.

After the overburden has been removed, layers of clay are formed into disks and al-lowed to sun-dry before removal. The clay is removed from the pit in layers, and theensuing stockpile is constructed layer by layer. This construction is done in an exact-ing manner to maximize crude clay homogenization.


166

9.2.2Purif ication and Surface Treatment

The diagram in Figure 9-1 summarizes the process for separating the montmoril-lonite clay from other non-clay minerals, such as quartz, gravel, and limestone, fol-lowed by surface treatment. Copious quantities of water are used to ensure that themontmorillonite clay is in an exfoliated condition so that larger particles can be re-moved through various separation techniques. Strategic sampling points existthroughout the process. Statistical Process Control tools assist the production opera-tors in maintaining the process within natural variation limits. Important control fac-tors in the process are as follows:

solids/water ratio, counter ion optimization, purity, pre-organic reaction particle size, organic/inorganic ratio, post-organic reaction dispersive characteristics, post-milling solids/moisture ratio, post-milling dispersive characteristics, post-milling particle size, packaging aesthetics.

The organic modifier used in the surface treatment is generally a quaternary am-monium compound, although other onium ions, i.e. phosphonium, can be consid-


Fig. 9-1 Flow chart of clay processing.

167

ered. The reaction that occurs is an ion-exchange reaction wherein the positivelycharged quaternary salt replaces the sodium cations on the clay surface. During thereaction, as the clay is being converted to an organoclay, it changes from hydrophilicin nature to oleophilic.

9.2.3Synthetic Clays

Synthetic clays may be prepared using a variety of chemical sources providing thenecessary elements, namely silicon, oxygen, aluminum, and magnesium, amongothers. Synthetic clays are the subject of current research, but there is little publicknowledge concerning the various technologies being investigated. Natural clayswould appear to have an inherent raw material cost advantage, but the ability to con-trol purity, charge density, and particle size is an appealing objective.

One synthetic clay that has been on the market for several years is a syntheticmica prepared from a natural raw material, talc, which is treated in a high tem-perature electric furnace with alkali silicof luoride. The chemical structure isNaMg2.5SiO4O10(FαOH1–α)2 (0.8 ≤ α ≤ 1.0). The producer of this material claims low-er impurity levels and higher aspect ratios than with natural montmorillonitespecies.


Silicon and oxygen are common to all clay minerals, and the combination with otherelements such as aluminum, magnesium, iron, sodium, calcium, and potassium,and the numerous ways in which the elements can be linked together make for alarge number of configurations. An important distinction in clay mineral propertiesis the capacity of certain clays to change volume by absorbing water molecules fromother polar ions into their structures. This is called the swelling property. Clays aredivided into swelling and non-swelling type materials, and swelling types are calledsmectites. Of the many smectite varieties, montmorillonite appears to be most suit-able as the basis for a nanoclay.

Silica is the dominant constituent of montmorillonite clays, with alumina being es-sential as well. Clays have a sheet structure consisting of two types of layers, the sili-ca tetrahedral and alumina octahedral layers. The silica tetrahedral layer consists ofSiO4 groups linked together to form a hexagonal network of repeating units of com-position Si4O10. The alumina layer consists of two sheets of close-packed oxygens orhydroxyls, between which octahedrally coordinated aluminum atoms are embeddedin such a position that they are equidistant from six oxygens or hydroxyls. The twotetrahedral layers sandwich the octahedral, sharing their apex oxygens with the latter.These three layers form one clay sheet. Figure 9-2 shows the structure of a dioctahe-dral smectite consisting of two-dimensional arrays of silicon–oxygen tetrahedra andtwo-dimensional arrays of aluminum- or magnesium–oxygen–hydroxyl octahedra.


168

If the octahedral positions were occupied by alumina, the structure would not cor-respond to montmorillonite but to that of the inert mineral pyrophyllite. So, ex-tremely important to the structure of clays is the phenomenon of isomorphous sub-stitution. Replacement of trivalent aluminum by divalent magnesium or iron(II) re-sults in a negative crystal charge. The excess negative charge is compensated on theclays’ surface by cations that are too large to be accommodated in the interior of thecrystal. Further, in low pH environments, the edges of the clay crystal are positivelycharged and compensated by anions. The structure of montmorillonite is comparedwith the structures of kaolinite and muscovite mica in Figure 8-2 of Chapter 8.

In order to convert montmorillonite clay into a nanoclay compatible with organicpolymers, an ion-exchange process is performed to treat the clay surfaces. Generally,an organic cation, such as that from a quaternary ammonium chloride, is used tochange the hydrophilic/hydrophobic characteristics of the clay (Figure 9-3).

Typical characteristics of montmorillonite clays are as follows:

shape: platelet; size: 1 nm thick, 75–150 nm across; charge: unit cell 0.5–0.75 charge 92 meq/100 g clay; surface area: >750 m2 g–1; specific gravity 2.5 (lower for alkyl quaternary ammonium bentonites); modulus: ~170 GPa; particle: robust under shear, not abrasive (Mohs hardness 1–2).

Figure 9-4 shows nanoplatelets and their precursor, bentonite rock.


Fig. 9-2 Structure of smectite clay.


R1

R3 R4

R2

N

+ Cl–

N = NitrogenR1 through R4 = combination of aliphaticchains, methyl and/or benzylgroups

Quaternary Ammonium Compound

Fig. 9-3 Structure of quaternary ammonium compounds.

Fig. 9-4 Morphologies of bentonite rock and exfoliatedmontmorillonite.

170

9.4Suppliers

The following are among the major nanoclay suppliers worldwide:

Southern Clay Products, Inc., Gonzales, TX, USA Nanocor (Division of AMCOL Int’l.), Arlington Heights, IL, USA Süd-Chemie AG, Moosburg, Germany Laviosa Chimica Mineraria S.p.A., Livorno, Italy Kunimine Industries, Tokyo, Japan Elementis Specialties, Inc., Heightstown, NJ, USA Pai Kong Nano Technology Co., Ltd., Taoyuan Hsien, Taiwan CO-OP Chemical, Ltd., Tokyo, Japan

Typical products are untreated and surface-treated grades. Surface-treated gradesdiffer in their degree of hydrophobicity and the type of cation introduced through ionexchange.


Nanoclays are relatively new commercial products and, as such, the cost/price struc-ture has yet to be established. Although a variety of products appear to be availablefrom the suppliers listed, the quantities being sold are small and ref lect specialty anddevelopmental pricing policies. As the market grows and matures, it is expected thatprices for materials will be in the range US $2–4 per pound (US $4.5–9 per kg).


The health and environmental issues for nanoclays specifically are minimal andmanageable. Sister organoclay products have been used for many years in a host ofindustrial and consumer products. The perception that nanoclays are somehow dif-ferent because of the prefix “nano” may be the problem. Nanoclays only become“nano” when they are placed in a host polymer matrix, whereupon they cannot beseparated or distinguished from the bulk polymer and other constituents. Crystallinesilica is a naturally occurring component that may be present in commercial alkylquaternary ammonium bentonite (CAS No. 68953-58-2) at concentrations <0.5% [8].Crystalline silica dust (see also Chapter 19) when inhaled presents a health hazard tohumans and is regulated to very low permissible exposure limits.


171

9.7Applications

Nanoclays, in addition to their primary function as high aspect ratio reinforcements,also have important additional functions such as thermal and barrier properties andsynergistic f lame retardancy. Some of the factors responsible for good performancein nanocomposites are:

intercalation (surfactant & polymer); interfacial adhesion or wetting; exfoliation (dispersion and delamination).

Under appropriate conditions, the gallery spaces can be filled with monomer,oligomer, or polymer. This increases the distance between platelets, swelling the clay.Clay platelets swollen with polymer are said to be intercalated. If the clay swells somuch that it is no longer organized into stacks, it is said to be exfoliated, as shown inFigure 9-5.

Figure 9-6 is a schematic of the various dispersion mechanisms operative in pro-ducing nanoplatelets of very high aspect ratio. The nominal size of a dry nanoclayparticle is about 8 µm. Comprising the particle are approximately 1 million clayplatelets consisting of bundles of platelets called tactoids. Through a combination ofchemistry and processing/shear techniques, the particle is separated into tactoidsand the platelets are peeled from the tactoid to become fully dispersed or exfoliated.

The primary appeal of a clay/polymer nanocomposite is that much smaller quan-tities of the nanoclay can be used to enhance polymer performance without detract-ing from other key characteristics. A comparison of properties achieved with talc andwith nanoclay in TPO (thermoplastic elastomer) is shown in Figure 9-7. Indeed, oneof the major challenges has been to develop fully exfoliated products to obtain themaximum benefit of nanoclays. During the dispersion process, particles are shearedinto tactoids and platelets peel from the tactoids to become fully dispersed or exfoli-

9.7 Applications

Fig. 9-5 Comparison of intercalated and exfoliated clays.

172

ated in the host matrix (see Figure 9-8). During compounding, important process pa-rameters are clay feed position, type of twin-screw extruder, and screw design/speed.There are numerous publications discussing the effects of process conditions on de-gree of exfoliation [9,10].


8 µm Particle~1 million Platelets

Chemistry Chemistry/Processing Processing

Dispersion

Tactoids/Intercalants

PartialDispersion

Dispersion




Fig. 9-6 Schematic of the different clay dispersion mechanisms.

250

200

150

100

50

0

Res

ura

l Mo

du

lus,

kp

al

ReactorGrade TPO

Mineral Percent

Four times lessNano than talc

0 5 10 15 20 25

CA53 TalcCA53 15APB3200

Fig. 9-7 Flexural modulus vs. filler concentration in a thermoplasticelastomer, TPO; comparison of talc (CA53) with nanoclay (CA53-15A) in the presence of maleated PP (PB3200).

173

A variety of potential host systems, including polyamides, polyolefins, PVC, TPU,PLA, EVA, ionomers, rubber, recycled streams, and polymer blends have been evalu-ated with regard to the incorporation of nanoclays. Although exfoliation has beenachieved in many polymers, it has not led to significantly improved mechanical prop-erties other than modulus. The high degree of interest in the nanocomposite concepthas not yet resulted in a plethora of commercial products. However, products areemerging with increasing frequency as producers, processors, and users gain moreexperience with the products and envisage potential commercial applications. In thecase of the nanocomposites used in the 2004 Chevrolet Impala side moldings, Gen-eral Motors reports a weight saving of 7% as well as a better overall surface quality be-cause the filler is so fine it does not disrupt the surface of the part. The fine fillersare also said to improve mar resistance. Usually, when a stone hits a rocker panel, thewhite color that appears is due to the eye being able to see the filler. In the case ofnanocomposites, this effect should be lessened due to the inherently smaller size ofthe filler. As with any new technology and product, challenges have had to be over-come. For example, the tooling required some design changes, shrinkage rates weredifferent, and color recipes needed to be changed. Manufacturing the nanocompos-ite was also a challenge: introducing a nanoclay at low levels into a TPO requires gooddistribution and dispersion of the dry product into the polymer melt so that the nan-oclay can be substantially exfoliated.

The following are additional examples of the multifunctional character of nan-oclays:

ThermalA nano-nylon-6 has been commercialized by Unitika for an engine cover which re-quired substantially higher heat distortion temperatures than achieved with nylon-6.At a 4 wt. % loading of synthetic mica, the DTUL (at 1.8 MPa) was increased from70 °C for neat nylon-6 to 152 °C. Also, f lexural strength increased from 108 to158 MPa and f lexural modulus from 3.0 to 4.5 GPa [11].

9.7 Applications

Fig. 9-8 Monitoring the dispersion of nanoclays by transmission electron microscopy.

174

BarrierSeveral companies offer commercial nano-nylon products with improved barrierproperties and maintaining clarity. Targeted packaging applications include multi-layer PET bottles for high oxygen barrier demands for beer bottles, and f lexiblemultilayer films for meats and cheeses. The permeability of nylon barrier resins isgenerally reduced by a factor of two to four with less than 5% nanoclay addition [12].Nylon nanocomposites are also being considered for automotive applications such asfuel tanks and lines [13].

Synergistic Flame RetardancyNanocomposites have been demonstrated to reduce f lammability, particularlythrough lowering peak heat release. In combination with conventional f lame retar-dants such as magnesium hydroxide or aluminum trihydrate, several polyolefin-based wire and cable products have been developed which incorporate 5% nanoclayto reduce the amount of conventional FR agents required and to improve physicalproperties [14,15].

References

1 Fujiwara, S., Sakamoto, T., Japanese PatentNo. JPA51-109998, 1976.

2 Kato, M., Usuki, A., Chapter 5 in Polymer-Clay Nanocomposites (Eds.: Pinnavaia, T. J.,Beal, G. W.), John Wiley & Sons, New York,2000.

3 Usuki, A., et al., J. Mater. Res. 1993, 8,1179–1184.

4 Usuki, A., et al., J. Mater. Res. 1993, 8,1174–1178.

5 Vaia, R., Chapter 12 in Polymer-ClayNanocomposites (Eds.: Pinnavaia, T. J., Beal,G. W.), John Wiley & Sons, New York, 2000.

6 Kamena, K., “Nanocomposites: The Path toCommercialization”, Conference Proceed-ings, Principia Partners, Baltimore, June4–5, 2001.

7 Kamena, K., Functional Fillers for Plastics2002 Conference, Intertech Corp., Toronto,Canada, Sept. 18–20, 2002.

8 Southern Clay Products, Inc., MaterialsSafety Data Sheet, “Cloisite 20A”, RevisedSept. 2003.

9 Pinnavaia, T. J., Beal, G. W. (Eds.), Polymer-Clay Nanocomposites, multiple chapters,John Wiley & Sons, New York, 2000.

10 Utracki, L. A., Cole, K. C., Proc. 2nd Inter-nat. Symposium on PolymericNanocomposites, Oct. 2003, National Re-search Council Canada; Polym. Eng. Sci.2004, 44, 6.

11 Yasue, K., et al., Chapter 6 in Polymer-ClayNanocomposites (Eds.: Pinnavaia, T. J., Beal,G. W.), John Wiley & Sons, New York, 2000.

12 Defendini, B., “High barrier polyamide-6nanocomposite and oxygen scavenger”,Proc. Nanocomposites 2002 Conference, Euro-pean Plastics News, Amsterdam, TheNetherlands, Jan. 28–29, 2002.

13 Nakamura, K. “Examining progress towardsdeveloping polyamide nanocomposites forautomotive applications”, Proc. Nanocom-posites 2004 Conference, European PlasticsNews, Brussels, Belgium, March 17–18,2004.

14 Gilman, J., Kashiwagi, T., Chapter 10 inPolymer-Clay Nanocomposites (Eds.: Pinnava-ia, T. J., Beal, G. W.), John Wiley & Sons,New York, 2000.

15 Beyer, G., “Nanocomposites – a new con-cept for f lame-retardant polymers”, PolymerNews 2001, 26, 370–378.


10Carbon Nanotubes/Nanof ibers and Carbon Fibers

Zafar Iqbal and Amit Goyal

10.1Introduction

Carbon nanotube and nanofiber reinforced polymer nanocomposites and micron-sized carbon fiber-based polymer composites look set to have a significant impact onemerging advanced products ranging from aerospace, automotive and PEM (proton-exchange membrane) fuel cell parts, to surgical implants and to components fornanoelectronics. The area of micron-scale carbon fiber filled composites, unlike thatof the emerging field of carbon nanotube and nanofiber-based nanocomposites, isrelatively mature. Although both areas are discussed in this chapter, we focus moreon the rapid advances being made in the field of carbon nanotube-based nanocom-posites, discuss some new developments in conventional micron-scale and sub-micron scale carbon fiber composites, and point out possible synergies.

10.1.1Types of Carbon Nanotubes/Nanof ibers and their Synthesis

There has been intense interest in carbon nanotubes (CNTs) since their discovery byIijima in 1991 [1], in large part because they possess unique structural and electron-ic properties. Single-wall carbon nanotubes (SWNTs) are the fundamental form ofcarbon nanotubes, with unique electronic properties that emerge due to their one-di-mensionality; an SWNT is a single hexagonal layer of carbon atoms (a graphenesheet) that has been rolled up to form a seamless cylinder. Three types of SWNTs withdiffering chirality can be formed, as depicted in Figure 10-1. The one-dimensionalunit cell shown has a circumference given by the chiral vector C = na + mb, where nand m are integers equivalent to the roll up vectors and a and b are unit vectors of thehexagonal lattice. A multiple-wall carbon nanotube (MWNT) is a stack of graphenesheets rolled up into concentric cylinders. This stacking results in a loss of some ofthe unique one-dimensional properties present in the single (SWNT) and double-wall (DWNT) tube structures. The walls of each MWNT layer are parallel to the cen-tral axis. A stacked cone or herringbone arrangement can also be formed by catalytic


176

chemical vapor deposition (CVD), which can grow with a hollow, tubular center.These structures have relatively large diameters (typically ≥ 50 nm) compared withthe near molecular-scale dimensions of SWNTs and the low nanoscale dimensions ofmost MWNTs, and are therefore referred to as carbon nanofibers (CNFs).

MWNTs were first synthesized using a non-catalytic carbon arc discharge methodby Iijima [1]. SWNTs were initially synthesized in 1–2% yields in soot generated in anarc struck between graphite electrodes containing a few percent Fe, Co or Ni byBethune et al. [2] and by Iijima and Ichihashi [3]. Smalley and co-workers [4] thenscaled-up SWNT synthesis using a dual laser ablation technique with transition met-al particles incorporated in the graphite target. This method could produce SWNTsin yields of up to 70%. The tubes are formed catalytically in the extremely high tem-perature of the ablation plume with a narrow distribution of diameters around 1.3nm and, due to van der Waals forces, generally assemble into bundles or ropes of par-allel SWNTs. Soon afterwards, Journet et al. [5] showed that yields of SWNT bundlesof about 50%, similar in size to those produced by laser ablation, can be obtained us-ing the arc-discharge method when catalyst particles of rare earth metals such as yt-trium are incorporated together with transition metals in the graphite rods. CVDmethods involving the decomposition of hydrocarbon precursor gases, typically eth-ylene and acetylene, in the presence of transition metal (iron, cobalt or nickel) cata-lysts on a support material such as alumina or silica, have been used to make CNFs

10 Carbon Nanotubes/Nanof ibers and Carbon Fibers

Figure 10-1 Armchair (n,m = 5,5) (top), zig-zag (9,0) (middle), and chiral (10,5) (bottom)single-wall nanotubes. All armchair tubes aremetallic, whereas only one-third of the chiral

tubes has metal character. (n,m) are the roll-up vectors and are proportional to the tube di-ameter. The dangling bonds at the tube endsare saturated by hemispherical fullerene caps.

177

[6,7] and MWNTs [8] at temperatures in the 550 to 1000 °C range. MWNTs grown bythe CVD technique, however, have high defect densities in their structures. Arc-grown MWNTs, on the other hand, are largely defect-free because growth occurs atplasma-generated temperatures in excess of 2000 °C. Recently, plasma-enhancedCVD (PE-CVD) growth of MWNTs has emerged as a technique for the growth of ver-tically aligned MWNTs and CNFs [9,10].

Since the late 1990s, largely defect-free SWNTs have been grown at near or above90% purity by CVD techniques involving the catalytic decomposition of methane attemperatures near 1000 °C [11,12], by the catalytic disproportionation of carbonmonoxide (CO) under high pressures (the so-called high-pressure carbon monoxideor HIPCO process) and at temperatures above 1000 °C [13], as well as at 1 atmosphereand temperatures below 1000 °C [14–17], using catalysts supported on silica andMgO. Cheng et al. [18] produced SWNTs at 1200 °C with undetermined purity levelsby heating a f low of benzene together with ferrocene and thiophene precursors toform f loating catalytic particles, whereas Maruyama et al. [19] generated SWNTs attemperatures down to 550 °C using ethanol under low-pressure conditions. Becauseof the etching effect of OH radicals produced on decomposition of the alcohol, non-SWNT phases, such as amorphous and non-tubular nanocarbons and MWNTs, werenot formed. Maruyama’s group has also been able to grow vertically aligned SWNTson catalyst-coated quartz substrates using ethanol as precursor in a thermal CVDprocess [20]. Low-pressure conditions using either ethylene or propylene as the car-bon source were employed by Sharma and Iqbal [21] to grow and observe in real-timeboth SWNTs and MWNTs in situ in an environmental transmission electron micro-scope.

As-synthesized SWNTs are typically bundled and comprise a range of tube diame-ters and chiralities. A method to grow single diameter, individual SWNTs is to formthem inside zeolites with selected pore sizes. Catalyst-free SWNTs with a diameter of0.42 nm, corresponding to that of the smallest fullerene, C20, were grown by Wang etal. [22] by this method. Another catalyst-free method [23], which provides thin bun-dles of SWNTs with a narrow diameter distribution in the 1.2 to 1.6 nm range, in-volves horizontal templated growth of the tubes on the Si face of hexagonal siliconcarbide (6H-SiC), although growth occurs only at temperatures above 1500 °C. Morerecently, PE-CVD with methane as the carbon source has been used for the first timeto grow SWNTs in the 550 to 900 °C temperature range. In the first reported study,SWNTs were grown bridging the pores of a zeolite positioned on an Ni plate [24]. Inthe second study, largely semiconducting SWNTs were grown on ferritin (a precursorfor nanoscale Fe catalyst particles) on silica [25], and in the third study SWNTs wereformed on sol-gel produced bimetallic Co-Mo catalysts on MgO [26]. It remains to beseen as to whether more controlled alignment of SWNTs can be obtained by the PE-CVD technique.

10.1 Introduction

178

10.1.2Types of Carbon Fibers and their Synthesis

Micron-sized carbon fibers presently used contain at least 90% carbon and are most-ly produced by heat treatment or controlled pyrolysis of different precursor fibers.Fibers may also be vapor grown in the absence of fibrous precursors. The most preva-lent precursors are polyacrylonitrile (PAN), cellulose fibers (such as viscose, rayon,and cotton), petroleum or coal tar pitch, and certain phenolic fibers. Pitch is a tar-likemixture of hundreds of branched organic compounds with differing molecularweights formed by heating petroleum or coal. The so-called mesophase of pitch is ina liquid-crystalline state. The structures of PAN, cellulose, and a phenolic resin aredepicted in Figure 10-2.

Micron-sized carbon fibers can be classified in terms of the precursor fiber mate-rials as PAN-based, mesophase or isotropic pitch-based, rayon-based, and phenolic-based. The synthesis process involves a heat treatment of the precursor fibers to re-move oxygen, nitrogen, and hydrogen to form the carbon fibers. It is well establishedin the literature that the mechanical properties of the carbon fibers are improved byincreasing the crystallinity and orientation, and by reducing defects in the fiber. Thebest way to achieve this is to start with a highly oriented precursor and then maintainthe initial high orientation during the process of stabilization and carbonizationthrough tension.


Figure 10-2 Molecular structures of polymeric precursors for mi-cron-sized carbon fibers.

179

10.1.2.1 PAN-Based Carbon FibersThere are three successive stages in the conversion of a PAN precursor into high-per-formance carbon fibers:

1. Oxidative stabilization: The PAN precursor is first stretched and simultaneouslyoxidized in the temperature range 200–300 °C. This treatment converts thermo-plastic PAN to a non-plastic cyclic or ladder compound.

2. Carbonization: After oxidation, the fibers are carbonized at about 1000 °C withouttension in an inert atmosphere (normally nitrogen) for a few hours. During thisprocess, the non-carbon elements are removed as volatiles to give carbon fibers ina yield corresponding to about 50% of the mass of the original PAN.

3. Graphitization: Depending on the type of fiber required, the fibers are treated attemperatures in the range 1500–3000 °C; this step improves the ordering and ori-entation of the crystallites in the direction of the fiber axis.

10.1.2.2 Carbon Fibers from PitchCarbon fiber fabrication from pitch generally involves the following four steps:

1. Pitch preparation: Essentially an adjustment in the molecular weight, viscosity,and crystallite orientation for spinning and further heating.

2. Spinning and drawing: In this stage, the pitch is converted into filaments, withsome alignment in the crystallites to achieve directional characteristics.

3. Stabilization: In this step, cross-linking is introduced to maintain the filamentshape during pyrolysis. The stabilization temperature is between 250 and 400 °C.

4. Carbonization: The carbonization temperature is typically in the range 1000–1500 °C.

Carbon fibers made by the spinning of molten pitches are of interest because of thecarbon yield approaching 99% and the relatively low cost of the starting materials.The formation of melt-blown pitch webs is followed by stabilization in air and car-bonization in nitrogen. Processes have been developed for use with isotropic pitchesand with anisotropic mesophase pitches. The mesophase pitch-based and melt-blowndiscontinuous carbon fibers have a structure comprised of a large number of smalldomains, each domain having an average equivalent diameter ranging from 0.03 mmto 1 mm, and a nearly unidirectional orientation of folded carbon layers assembled toform a mosaic structure on the cross-section of the carbon fibers. The folded carbonlayers of each domain are oriented at an angle to the direction of the folded carbonlayers of the neighboring domains on the boundary.

Carbon f ibers from isotropic pitchIsotropic pitch or a pitch-like material, such as molten polyvinyl chloride, is meltspun at high strain rates to align the molecules parallel to the fiber axis. The ther-moplastic fiber is then rapidly cooled and carefully oxidized at a low temperature(<100 °C). The oxidation process is rather slow, so as to ensure stabilization of thefiber by cross-linking to make it infusible. However, upon carbonization, relaxation

10.1 Introduction

180

of the molecules takes place, producing fibers with no significant preferred orienta-tion. This process is not industrially attractive due to the lengthy oxidation step, andbecause only low-quality carbon fibers with no graphitization are produced. Thesefibers are used as fillers in various plastics to form thermal insulation materials.

Carbon f ibers from anisotropic mesophase pitchHigh molecular weight aromatic pitches that are mainly anisotropic in nature are re-ferred to as mesophase pitches. The pitch precursor is thermally treated above 350 °Cto convert it to mesophase pitch, which contains both isotropic and anisotropic phas-es. Due to shear stresses occurring during spinning, the mesophase molecules ori-ent parallel to the fiber axis. After spinning, the isotropic part of the pitch is made in-fusible by cross-linking in air at a temperature below its softening point. The fiber isthen carbonized at temperatures up to 1000 °C. The main advantage of this processis that no tension is required during stabilization or graphitization, unlike in the caseof rayon or PAN precursors.

10.1.2.3 Carbon Fibers from RayonThe conversion of rayon fibers into carbon fibers is a three-stage process:

1. Stabilization: This is essentially an oxidative process that involves different steps.In the first step, between 25 and 150 °C, there is physical desorption of water. Thenext step is a dehydration of the cellulose unit between 150 and 240 °C. Finally,thermal cleavage of the cyclosidic linkage and scission of ether bonds and someC–C bonds occurs via free radical reactions (240–400 °C), which is followed byaromatization.

2. Carbonization: Heat treatment between 400 and 700 °C converts the carbonaceousresidue into graphite-like layers.

3. Graphitization: Graphitization is carried out under strain at 700–2700 °C to obtainhigh modulus fibers through a longitudinal orientation of the planes.

10.1.2.4 Carbon Fibers from Phenolic ResinsMicron-sized carbon fibers can be synthesized from phenolic resin fibers such asKynol [27]. The carbon fibers prepared in this way are typically obtained in an acti-vated form, which produces well developed mesopores for use in applications as highsurface area adsorbents.

10.1.2.5 Vapor-Grown Carbon FibersVapor-grown carbon fibers (VGCFs) comprise a large family of filamentous nanocar-bons. They can be distinguished in terms of the arrangement of the graphene layersin their molecular-scale structures: they can be “plate-like”, with near-parallelgraphene layers that are approximately perpendicular to the fiber axis, or they canhave the “fish-bone” microstructure with stacked cones of graphene planes. Sub-mi-cron (50 to 200 nm diameter) VGCFs of the “fish-bone” structure approach the di-mensions of MWNTs and are referred to as CNFs (see above) in this chapter. VGCFsand CNFs are generally grown by depositing carbon by the high temperature (typi-


181

cally in the 900 to 1200 °C range) decomposition of a hydrocarbon (usually methane)catalyzed by finely divided transition metal catalyst particles. Depending on the cata-lyst, different growth forms are found: one-directional growth (the fiber grows withthe catalyst at the tip “tip-mode” or at the rear “rear-mode”), bidirectional growth (si-multaneous growth in two opposite directions with the catalyst particle in the mid-dle), multi-directional growth (more than two fibers grow out of one catalyst particle:“octopus fiber”), as well as branched growth (a larger catalyst particle explodes dur-ing the growth resulting in branched growth of a number of smaller fibers).

10.1.3Chemical Modif ication/Derivatization Methods

The development of carbon nanotube-based nanocomposites was initially impededby the inability to uniformly disperse the nanotubes in the polymer matrix due to alack of compatibility between the chemical structures of the two components. Com-patibility has now been achieved in many cases by chemical modification or derivati-zation of the nanotubes. Some degree of derivatization is achieved following nan-otube synthesis by the adsorption of electron-withdrawing oxygen on the tube wallsand by the formation of acidic COOH groups as a result of acid purification proce-dures to remove the catalyst and support as well as amorphous/microcrystalline car-bon produced as an impurity. Derivatization may also improve the solubility of thenanotubes in certain organic solvents and in water, and also permit covalent interac-tion between the nanotube and polymer, which leads to better adhesion at the nan-otube/polymer interface, resulting in the formation of nanocomposites with excep-tionally high mechanical strength.

Two approaches have been utilized to achieve derivatization. The first has involvedchemical modification of the nanotube surface, while the second has involved chem-ical interaction with various defects on the graphitic walls of the tubes and at the tubeends. The surface modifications reported in the literature for nanotubes have beensomewhat similar to those achieved on the C60 fullerene, although closer examinationreveals notable differences in reaction type and in the location and symmetry of thechemistry involved. On the other hand, defect site functionalization involves chem-istry that is not applicable to the fullerenes because they are free from similar defects.

A large amount of literature exists on the chemical modification of carbon nan-otubes, but detailed understanding is still lacking because of the paucity of theoreti-cal calculations and simulations. Several research groups have reported the success-ful functionalization of both SWNTs and MWNTs [28–33]. These modifications haveinvolved the direct attachment of functional groups such as f luorine or hydrogen tothe graphitic walls, reactions with nitrenes and carbenes, or the use of carboxylic acidgroups bonded to the nanotube walls produced on oxidation of shortened and un-bundled tubes. Chen et al. [28] first reported the use of acid groups for attaching longalkyl chains to SWNTs via amide linkages. There is now ample evidence that nan-otube-bound carboxylic acid groups are the sites at which a variety of functionalgroups for the solubilization of both shortened and full-length carbon nanotubes areattached. For example, it has been shown that esterification of the carboxylic groups

10.1 Introduction

182

can be used to functionalize and solubilize nanotubes of any length [34–36]. Mono-,di-, and trinitroanilines have recently been attached to SWNTs via carboxylic groupsthrough reaction with thionyl chloride [37]. Multiple sulfonate groups, OSO3H, havebeen chemically grafted onto MWNTs [38], and these have been compounded withemeraldine base polyaniline to form composites with enhanced electrical conductiv-ity and thermal properties due to concomitant doping of the polymer by the sul-fonated nanotubes in the course of composite fabrication. Solubilization in water hasbeen achieved by wrapping with polymers such as polyvinyl pyrrolidone (PVP) [39] orpolyethylene imine (PEI) [40,41], and by reaction with glucosamine [42]. SWNTs havealso been effectively dispersed/solubilized in water by their sonication in the pres-ence of the single-stranded version of the central polymeric molecule in biology, DNA[43], or in the presence of enzyme molecules suitable for use in biosensing and bio-fuel cells [44]. In the case of DNA, molecular modeling suggests that single-strandedDNA binds to SWNTs through π-stacking interactions that result in helical wrappingof the nanotube sidewalls [43].

For carbon nanofibers and conventional micro-fibers, the key to the formation ofhigh strength polymer composites is the adhesion of the fibers to the polymeric ma-trices. The adhesion forces are still not fully understood, primarily because the sur-faces of the carbon fibers are complex with respect to their structure and chemistry.The forces result from different interactions across the interface, which include dis-persive interactions of the van der Waals type involving London forces, non-disper-sive interactions involving acid-base processes, and covalent chemical bonds. Typicalsurface treatment involves oxidation in air or ozone to form oxygen-containing func-tional groups. Alternative approaches involve the use of plasma-induced surfacemodification [45] or electrochemical anodization in an acidic electrolyte such as phos-phoric acid [46]. The surface groups produced consist of basic pyrone-like structures,neutral quinines, and acidic carboxylic groups. The strength of the compositesformed has been correlated with the surface roughness discerned by means of de-tailed scanning electron and tunneling microscopies [47].

10.1.4Polymer Matrices

As discussed in Section 10.1.3, PVP and DNA have been used to wrap and water-sol-ubilize SWNTs. For specific actuator, electrical, and electro-optic applications,SWNTs have been wrapped with piezoelectric polyvinylidene f luoride/trif luoroeth-ylene copolymer [48] or with conjugated polymers [49,50]. A conjugated polymer thathas been used to form a composite with MWNTs and an electron-transport layer inlight-emitting diodes is poly(m-phenylene-vinylene-co-2,5-dioctyloxy-p-phenylene-vinylene) (PmPV) [51]. Wrapping coupled with electron-doping has been achievedwith polyethylene imine (PEI) to form p-n junction devices [40,41].

Thermosetting epoxy resins are widely used in the fabrication of carbon fiber-based composites for aerospace applications. High-temperature amorphous thermo-plastics with high impact strength, which include polycarbonate, polysulfones, poly-ether imide, polyether sulfones, and partially crystalline polyether ether ketone, are


183

alternative polymers bearing functional groups that can undergo selective interac-tions with the functional groups formed on the carbon fiber surface. For the fabrica-tion of electrically conductive bipolar plates for proton-exchange membrane fuelcells, chemically passive polymers such as polypropylene (PP) are preferred [52],whereas poly(acrylonitrile–butadiene–styrene) (ABS), polystyrene (PS), and high-im-pact polystyrene (HIPS) are used in the fabrication of composites for applicationswhere high impact strength is required.

10.2Polymer Matrix Composites

10.2.1Fabrication

In contrast to short carbon fiber reinforced thermoplastics, which are processed byconventional melt processing techniques, the limited availability, high cost, and thedifficulties encountered in achieving a high degree of dispersion continue to presentchallenges in the manufacture of carbon nanotube composites. Currently, most car-bon nanotube-reinforced composites are prepared in the laboratory using the so-called solution-evaporation method [53–56]. The solvent and curing agent may varywith different polymer matrices. The general procedure involves dissolving the poly-mer to form a first solution, dispersing/dissolving SWNTs or MWNTs to form a sec-ond solution, mixing the two solutions with the aid of ultrasonication, and finallycasting films or solid parts from the mixed solution and subjecting them to a curingprocess.

In order to achieve more uniform nanotube dispersion in composites, Haggen-mueller et al. [57] developed an alternative melt mixing method consisting of a com-bined solution-evaporation technique to prepare a thin SWNT-polymer film followedby repeated compression molding of the latter. The resulting product was reported toyield compositionally uniform films. Using a small batch mixer, adequately dis-persed nanotube composites based on PP, ABS, PS, and HIPS have been prepared[58].

Another technique, known as the dry powder mixing method, has been employedby Cooper et al. to produce nanotube-reinforced poly(methyl methacrylate) (PMMA)composites [59]. Like most of the currently used fabrication methods for nanotube-based polymer composites, this technique is a combination of several protocols in-cluding solution-evaporation, sonication, kneading, and extrusion. More specifically,these workers used ultrasonic techniques to blend carbon nanotubes with PMMAparticles, and the blend was later extruded to orient the nanotubes. Yang et al. [60]prepared small-scale batches of ABS nanocomposites without the use of solvents orultrasonic techniques with good dispersion of the nanotubes.

Another method used, known as extrusion free-form fabrication (EFF), belongs toa family of manufacturing processes in which different parts are built in layers. It isa solid freeform fabrication (SFF) technique, whereby the feed is in the form of a sol-

10.2 Polymer Matrix Composites

184

id. A 3-D computer model is generated and transferred to a computer supported bythe SFF software. The model is sliced into layers and the geometrical information isfed for each layer of the part, which is then built layer by layer. Carbon nanotubes andcarbon fibers are well-suited to this technique because they do not clog the nozzles.In EFF, the solid feed material is placed in a heated head and is forced through a noz-zle by a piston into a specified shape. Once a layer is complete, the support base islowered in the z-direction. The EFF process helps in tailoring the alignment of fibersin composites since the extrusion path can be changed in different parts. A study ofSWNTs and VGCFs mixed with ABS polymer by means of Banbury mixing and EFFwas conducted by Shofner et al. [61]. A high degree of dispersion of the nanotubesand fibers was achieved without porosity. For both VGCFs and SWNTs, sizable ten-sile strength and modulus improvements were observed.

10.2.2Mechanical and Electrical Property Modif ication

Carbon fibers have been used in both thermosetting and thermoplastic polymer com-posites for a long time, imparting higher modulus and strength and lighter weightthan glass fibers (see the comparison of properties in Table 2-1), electrical and ther-mal conductivity, chemical resistance, and reduced wear. Specific examples of theireffects on thermoplastics and thermosets may be found in the handbooks and gen-eral references listed in Chapters 1 and 2. However, with the discovery of near mo-lecular scale carbon nanotubes and advances in understanding their mechanical andelectrical properties over the past decade, new nanocomposites based on these novelmaterials are now possible. Experimental estimates of SWNT strength are in therange of 13–52 GPa and tensile modulus is of the order of 1 TPa [62–64], values muchhigher than those for carbon fibers (see also Table 2-1). Electrical resistivity and ther-mal conductivity measurements along the length of a bundle of SWNTs indicate val-ues of approximately 10– 4 Ω cm and 200 W m–1 K–1, respectively [65]. Two main is-sues to be addressed for the effective use of nanotubes and the translation of their ex-traordinary mechanical properties to a composite are alignment and uniform dis-persion in the polymer matrix. This is because SWNTs form in bundles and tend toagglomerate through weak van der Waals forces. A great deal of work has been doneaimed at overcoming this and several surfactant-based and organic solutions havebeen identified that are able to disperse and chemically functionalize nanotubes[66–71]. Another issue being addressed is that of interfacial bonding between thetubes and the polymer matrix, which affects the efficiency of load transfer across thenanotube–polymer interface.

Several studies on the characterization and fabrication of carbon nanotube–poly-mer nanocomposites have highlighted the important roles of the parameters dis-cussed in Chapter 2 (orientation, dispersion, and interfacial adhesion) in determin-ing the properties of the composites. Jia et al. [72] used an in situ process for the fab-rication of a PMMA/MWNT composite. An initiator was used to open up the π bondsof the MWNTs in order to increase the linkage with the PMMA. The formation ofC–C bonds results in a strong interface between the nanotubes and the PMMA. For


185

samples mixed with carbon nanotubes, smaller amounts of initiator are required andimproved mechanical properties are obtained. In another example, simple sonicationof SWNTs in solvents such as DMF was found not to yield good dispersion accordingto the method of Haggenmueller et al. [57] introduced above; therefore, repetitivefilm forming with sonication and drying followed by mixing at higher temperaturesand pressure was required to obtain a uniform dispersion (Figure 10-3). The meltcould be formed as a film or spun into fibers. With the introduction of SWNTs, thedraw ratio is reduced and a roughened surface results for the fiber, as viewed underan optical microscope. Mechanical and electrical properties are improved with in-creasing SWNT concentration. Melt processing, therefore, appears to be a very effec-tive method for realizing targeted mechanical and electrical properties in the bulkcomposites.


Figure 10-3 Optical micrographs of anSWNT–PMMA nanocomposite containing1 wt. % purified soot: (a) After only sonicationand drying. The as-cast film was repeatedly

subjected to hot pressing (180 °C, 13.5 kN,3 min) and is shown here after (b) 1 cycle, (c) 5 cycles, (d) 20 cycles. (Reproduced withpermission from Elsevier [57]).

186

Nanotubes were found to be oriented in the extrusion f low direction, increasingthe impact strength of the PMMA nanocomposites formed by the method of Cooperet al. [59]. Jin et al. [69] proposed a method of casting a suspension of carbon nan-otubes in a solution of the thermoplastic polymer polyhydroxy amino ether (PHAE)in chloroform. In this study, it was found that the resulting nanocomposites could bestretched up to five times their original length without breaking under varying me-chanical loads at a temperature 90–100 °C. The tubes were aligned inside the poly-


Figure 10-4 (a) TEM image of an internal frac-ture surface of a composite sample about90 nm in thickness after being microtomedparallel to the stretching direction. The nan-otubes are aligned parallel to the stretching di-rection and fiber pull-out is observed. In someareas, nanotubes bridge the microvoids (ormicrocracks) in the matrix and presumably en-

hance the strength of the composite. (b)Cross-sectional view of the same compositemicrotomed perpendicular to the stretchingdirection. Cross-sections of the nanotubesand nanoparticles are observed. (Reproducedwith permission from American Institute ofPhysics [69]).

187

mer matrix, as indicated by X-ray diffraction and transmission electron microscopystudies (Figure 10-4). Highly aligned SWNTs in polystyrene and polyethylene havebeen obtained by Haggenmueller et al. [73] using a twin-screw extruder. Compositefibers obtained with 20% nanotube loadings showed a 450% increase in elastic mod-ulus relative to polyethylene fibers.

Carbon nanotube/polystyrene nanoporous membranes with aligned MWNTs tra-versing the membrane thickness have recently been fabricated by Hinds et al. [74].The structures formed are depicted in Figure 10-5. These nanoporous membraneshave the ability to gate molecular transport through the cores of the nanotubes, of-fering potential applications in chemical separations and sensing.


Figure 10-5 A. As-grown alignedMWNTs produced by a Fe-cat-alyzed chemical vapor depositionprocess. B. Schematic of a targetmembrane structure. C. Scanningelectron micrograph of MWNT-polystyrene composite membrane.Scale bar represents 2.5 microns.(Reproduced with permissionfrom Science AAAS [74]).

188

Carbon nanotubes have been introduced into conducting polymers, such aspoly(m-phenylene-vinylene-co-2,5-dioctyloxy-p-phenylene-vinylene) (PmPV), as anelectron-transport layer in organic light-emitting diodes. Their introduction led to asignificant increase in efficiency and an increase in electrical conductivity of four or-ders of magnitude [49]. The normalized photoluminescence intensity and electricalconductivity as a function of MWNT loading for these composites are shown in Fig-ure 10-6. Helical wrapping of the conducting polymer around the nanotubes hasbeen modeled by Lordi and Yao [75]; such a model is depicted in Figure 10-7.


Tables 10-1 and 10-2 list U.S. and international companies that supply nanotubes, re-lated nanocarbon materials, and carbon fibers. The prices at the time of writing arealso given where available. Note that the cost of pure SWNTs still remains very high.


Fullerene soot with a high SWNT content has been tested to assess its biochemicalactivity [76]. The dermatological trial results did not show any signs of health hazardsrelated to skin irritation or allergic risks. To determine whether carbon nanotubesand in particular SWNTs present any significant health hazards, Huczko et al. [77]performed tests routinely used in the patho-physiological testing of asbestos-induced


Figure 10-6 Normalized photoluminescence (PL) intensity and conduc-tivity for conducting polymer-nanotube composite films as a functionof nanotube to polymer mass ratio. (Reproduced with permission fromAmerican Institute of Physics [49]).

189

diseases. No abnormalities of pulmonary function or measureable inf lammationwere detected in guinea pigs. As also discussed in Chapter 1, care should be exercisedwith the use of nanofillers since their toxicology has not yet been fully explored.

Carbon micro-fibers easily form dust during handling, which gets dispersed in theatmosphere. The fibers also tend to stick to human skin or mucous membranes caus-ing pain and itching. Protective wear for the skin, eyes, and throat therefore needs tobe worn to prevent these hazards. Local air exhausts and ventilators can help in re-moving the dust. Protective cream or gloves need to be used during handling of thefibers. Since the fibers are electrically conductive, care should be taken around ex-posed electrical circuits and outlets. Some general purpose grades of carbon fibermay ignite at temperatures below 150 °C in the presence of air or fuel. If heated toabove 400 °C in the presence of air or fuel, the fibers burn slowly but stop burning assoon as the fuel source is removed.

Carbon fiber waste should be treated as industrial rather than household waste.Local governments may have their own local codes for disposing of carbon fiberwastes. On the positive side, carbon is thought to have good compatibility with hu-man tissue. Carbon fibers and fiber composites have therefore been used extensive-ly as components for artificial body parts and devices.


Figure 10-7 (a) Model of cis-poly(phenylacetylene) wrapped perfectlyaround a (10,10) SWNT. (b) Model of trans-poly(phenylacetylene),which has a slightly smaller diameter, distorts around the SWNT,and wraps more tightly. (Reproduced with permission from Journalof Materials Research [75]).

190

10.5Applications

Carbon fiber composites are widely used in the aerospace industry, and with the de-creasing price of the fibers they are increasingly being used in the automobile, ma-rine, sports, and construction industries. In aerospace, epoxy/carbon composites areused in the Space Shuttle payload door, its manipulator arm, and its booster tail andfins. There is extensive use of carbon fiber/epoxy composites in helicopter structuresas well as in commercial aircraft. Carbon fiber composites have started to be used inautomobiles, mainly for saving weight. Here, carbon nanotubes, in particular cost-ef-fective MWNTs and CNFs, are starting to be used in car bumpers and gasoline tanks.In the future, with decreasing prices, nanotubes could be used in these compositesto obtain much higher strength at much lower loading levels.

In addition to their primary function as mechanical property modifiers, the highelectrical conductivity of carbon fibers provides carbon-based composites with staticdissipation and radio frequency shielding characteristics. This opens up a wholerange of applications, and with carbon nanotubes this can be achieved at extremelylow loading levels. One application with future potential is the use of carbon nan-otubes to fabricate bipolar interconnecting f low-field plates for fuel cells [52]. Awhole range of futuristic applications in nanoelectronics is also emerging withSWNTs and MWNTs. The high thermal conductivity of carbon fibers and the dia-mond-like thermal conductivity of SWNTs make their composites highly attractivematerials for heat sinks in electronics. The low density of the composites comparedto copper makes them even more attractive for aerospace electronics.


Tab. 10-1 List of carbon nanotube and related material suppliers.

Company Products Price[a] Location

Applied Nano- Nanotube materials N/A Chapel Hill, NC, USAtechnologies Inc. (MWNTs/SWNTs)Applied Science Inc. MWNTs and CNFs CNFs: $145/kg Cedarville, OH, USA

(nanofibers)Bucky USA Fullerenes, MWNTs, MWNT: $100–150/g Houston, TX, USA

and SWNTs SWNT: $250–750/gCarbolex SWNTs $100–60/g Lexington, KY, USACarbon Nano- HIPCO SWNTs $500/g Houston, TX, USAtechnologies Inc.Carbon Solutions SWNTs $50–400/g Riverside, CA, USACatalytic Materials Ltd. Suppliers of nanofibers $60–40/g Holliston, MA, USA

and MWNTsFullerene Int. Corp. Fullerenes and nanotubes N/A Tucson, AZ, USA(FIC) supplierGuangzhou Single- and multiwall MWNT: $10–20/g, Guangdong,

carbon nanotubes SWNTs: $100–150/g P. R. China

[a] Prices, where available, are given at the time of writing.Price ranges ref lect purity and quantities.


Tab. 10-1 Continued

Company Products Price[a] Location

Hyperion Catalysis MWNTs N/A Cambridge, MA, USAInternationalIljin Nanotech MWNTs & SWNTs N/A Kangseo Ku, Seoul,

S. KoreaMER Corporation Various (from fullerenes N/A Tucson, AZ, USA

to SWNTs)Metrotube SWNTs N/A Tokyo, JapanMitsui XNRI MWNTs N/A Tokyo, JapanNano-C Fullerenes N/A Westwood, MA, USANanoCarbLab SWNTs N/A Moscow, RussiaNanocs International MWNTs N/A New York, USANanocyl MWNTs and SWNTs N/A Namur, BelgiumNanoledge MWNTs/SWNTs SWNT: 65 €/g Montpellier, FranceNanolab MWNTs MWNT: $125–165/g Newton, MA, USANanoMaterials Inorganic nanotubes N/A Longmont, CO, USA

and nanospheresNanomirae Nanofibers – herringbone N/A Guro-gu, Seoul,

and spiral MWNTs S. KoreaRosseter Holdings Ltd. MWNT, SWNTs, and MWNT: $20/g Limassol, Cyprus

nanohornsSeldon Laboratories MWNTs N/A Windsor, VT, USASouthWest Nano SWNTs SWNT: $500/g Norman, OK, USATechnologies, Inc.Sun Nanotech MWNTs N/A Jiangxi, P. R. ChinaTsinghua-Nafine MWNTs N/A Beijing, P. R. ChinaNano-Powder

[a] Prices, where available, are given at the time of writing.Price ranges ref lect purity and quantities.

Tab. 10-2 List of carbon fiber suppliers.

Company Products Trademark Location

Amoco Fabrics and Fibers pitch-type fibers THORNEL USAAsahi Kasei Corporation HI CARBOLON JapanToho Rayon Co. PAN-type fibers BESFIGHT JapanToray PAN-type fibers TORACA JapanMitsubishi Rayon pitch-type fibers Dialead JapanNGF pitch-type fibers N/ABASF PAN-type fibers CELION USAHercules PAN-type fibers MAGNAMITE USASGL Carbon SIGARTEX Germany/USAZoltek PANEX USARK Carbon Fibers Ltd. CURLON UKCourtauld Ltd. COURTELLE UKAshland CARBOFLEX USA

192

References


1 Iijima, S., Nature 1991, 354, 56–58.2 Bethune, D. S., et al., Nature 1993, 363,

603–605.3 Iijima, S., Ichihashi, T., Nature 1993, 363,

605–607.4 Thess, A., et al., Science 1996, 273, 483–487.5 Journet, C., et al., Nature 1997, 388,

756–758.6 Tibbetts, G. G., J. Cryst. Growth 1984, 66,

632–638.7 Tennent, H. G., 1987, USA Patent

4,663,230.8 Amelinckx, S., et al., Science 1994, 265,

635–639.9 Ren, Z. F., et al., Science 1998, 282,

1105–1107.10 Meyyappan, M., et al., Plasma Sources Sci.

Technol. 2003, 12, 205–216.11 Kong, J., et al., Nature 1998, 395, 878–879.12 Kong, J., et al., Chem. Phys. Lett. 1998, 292,

567–574.13 Nikolaev, P., et al., Chem. Phys. Lett. 1999,

313, 91–97.14 Kitiyanan, B., et al., Chem. Phys. Lett. 2000,

317, 497–503.15 Lan, A., et al., Appl. Phys. Lett. 2002, 81,

433–435.16 Lan, A., et al., Chem. Phys. Lett. 2003, 379,

395–400.17 Goyal, A., Wang, Y., Sharma, R., Iqbal, Z.,

2004, unpublished data.18 Cheng, H., et al., Appl. Phys. Lett. 1998, 72,

3282–3284.19 Maruyama, S., et al., Chem. Phys. Lett. 2002,

360, 229–234.20 Murakami, Y., et al., Chem. Phys. Lett. 2004,

385, 298–303.21 Sharma, R., Iqbal, Z., Appl. Phys. Lett. 2004,

84, 990–992.22 Wang, N., et al., Chem. Phys. Lett. 2001, 339,

47–52.23 Derycke, V., et al., Nano Lett. 2002, 2,

1043–1046.24 Koto, T., et al., Chem. Phys. Lett. 2003, 381,

422–426.25 Li, Y., et al., Nano Lett. 2004, 4, 317–321.26 Maschmann, M., Goyal, A., Iqbal, Z., Fish-

er, T. S., Gat, R., 2004, Nano Lett., submit-ted.

27 Lin, R. Y., Economy, J., Appl. Polym. Symp.1973, 21, 143–152.

28 Chen, J., et al., Science 1998, 282, 95–98.29 Michelson, E. T., et al., Chem. Phys. Lett.

1998, 296, 188–194.30 Bahr, J. L., et al., J. Am. Chem. Soc. 2001,

123, 6536–6542.31 Pekker, S., et al., J. Phys. Chem. 2001, B105,

7938–7943.32 Lin, Y., et al., J. Phys. Chem. 2003, 107,

914–919.33 Sun, Y. P., et al., Acc. Chem. Res. 2002, 35,

1096–1104.34 Sun, Y. P., et al., Chem. Mater. 2001, 13,

2864–2869.35 Riggs, J. E., et al., J. Phys. Chem. 2000,

B104, 7071–7076.36 Fu, K., et al., Nano Lett. 2001, 1, 439–441.37 Wang, Y., Malhotra, S., Iqbal, Z., Proc. MRS

2004, Vol. 800, in press.38 Dai, L., Mau, A. W. H., Adv. Mater. 2001, 13,

899–913.39 O’Connell, M. J., et al., Chem. Phys. Lett.

2001, 342, 265–271.40 Shim, M., et al., J. Am. Chem. Soc. 2001,

123, 11512–11513.41 Zhang, Y., Grebel, H., Iqbal, Z., 2004, un-

published data.42 Balzano, L., et al., Chem. Mater. 2002, 14,

1853–1858.43 Zheng, M., et al., Nature Materials 2003, 2,

338–342.44 Wang, Y., Malhotra, S., Iqbal, Z., 2004, un-

published data.45 Mittal, K. L., Proc. of 2nd International Sym-

posium on Polymer Surface Modification,Newark, New Jersey, May 24–26, 1999.

46 Park, S. J., et al., J. Chem. Eng. Japan 2001,34, 396–400.

47 Krekel, G., et al., J. Mater. Sci. 1994, 29,2968–2980.

48 El-Hami, K., Matsushige, K., Chem. Phys.Lett. 2003, 368, 168–171.

49 Fournet, P., et al., J. Appl. Phys. 2001, 90,969–975.

50 Coleman, J. N., et al., Phys. Rev. B 1998, 58,R7492–R7495.

51 Czerw, R., et al., Proc. SPIE 2001, 4590,153–161.

52 Iqbal, Z., et al., 2003, US Patent 6,572,997.53 Mitchell, C. A., et al., Macromolecules 2002,

35, 8825–8830.


54 Wood, J. R., et al., Composites: Part A 2001,32, 391–399.

55 Qian, D., Dickey, E. C., J. Microscopy 2001,204, 39–45.

56 Sandler, J., et al., Polymer 1999, 40,5967–5997.

57 Haggenmueller, R., et al., Chem. Phys. Lett.2000, 330, 219–225.

58 Qian, D., et al., Appl. Phys. Lett. 2000, 76,2868–2870.

59 Cooper, C. A., et al., Composites Science andTechnology 2002, 62, 1105–1112.

60 Yang, S., et al., Polym. Degrad. Stab. 2004,83, 383–388.

61 Shofner, M. L., et al., Composites: Part A2003, 34, 1207–1217.

62 Yu, M. F., et al., Phys. Rev. Lett. 2000, 84,5552–5555.

63 Walters, D. A., et al., Appl. Phys. Lett. 1999,74, 3803–3805.

64 Gere, J. M., Timoshenko, S. P., Mechanics ofMaterials, PWS Publishing Co., Boston,1990.

65 Hone, J., et al., Appl. Phys. Lett. 2000, 77,666–668.

66 Dufresne, A., et al., J. Mater. Sci. 2002, 37,3915–3923.

67 Barraza, H. J., et al., Nano Lett. 2002, 2,797–802.

68 Safadi, B., et al., J. Appl. Polym. Sci. 2002,84, 2660–2669.

69 Jin, L., et al., Appl. Phys. Lett. 1998, 73,1197–1199.

70 Thostenson, E. T., Chou, T. W., J. Phys. D:Appl. Phys. 2002, 35, L77–L80.

71 Thostenson, E. T., Chou, T. W., J. Phys. D:Appl. Phys. 2003, 36, 573–582.

72 Jia, Z., et al., Materials Science and Engineer-ing 1999, A271, 395–400.

73. Haggenmueller, R. et al. , J. Nanosc. Nan-otech. 2003, 3, 105–110.

74 Hinds, B. J., et al., Science 2004, 303, 62–65.75 Lordi, V., Yao, N., J. Mater. Res. 2000, 15,

2770–2779.76 Huczko, A., Lange, H., Fullerene Sci. Tech.

2001, 9, 247–250.77 Huczko, A., et al., Fullerene Sci. Tech. 2001,

9, 251–254.

11Natural Fibers

Craig M. Clemons and Daniel F. Caulf ield

11.1Introduction

The term “natural f ibers” covers a broad range of vegetable, animal, and mineralfibers. However, in the composites industry, it usually refers to wood fiber and agro-based bast, leaf, seed, and stem fibers. These fibers often contribute greatly to thestructural performance of the plant and, when used in plastic composites, can pro-vide significant reinforcement. Below is a brief introduction to some of the naturalfibers used in plastics. More detailed information can be found elsewhere [1–4].

Although natural fibers have been used in composites for many years, interest inthese fibers has waned with the development of synthetic fibers such as glass andcarbon fibers. However, recently there has been a resurgence of interest.

One of the largest areas of recent growth in natural fiber plastic composites is theautomotive industry, particularly in Europe, where natural fibers are advantageouslyused as a result of their low density and increasing environmental pressures. Most ofthe composites currently made with natural fibers are press-molded, although a widerange of processes have been investigated [1,5]. Table 11-1 shows the consumption ofnatural fibers by the European automotive industry and projections of future totalconsumption [1]. Flax is the most widely used natural fiber in the European automo-tive industry, comprising 71% of the natural fibers consumed in 2000. Most of this isshort-fiber f lax obtained as a by-product of the textile industry [5]. Natural fibers aretypically combined with polypropylene, polyester, or polyurethane to produce suchcomponents as door and trunk liners, parcel shelves, seat backs, interior sunroofshields, and headrests [6].

Increased technical innovations, identification of new applications, continuing po-litical and environmental pressures, and government investment in new methods forfiber harvesting and processing are leading to projections of continued growth in theuse of natural fibers in composites, with expectation of reaching 100,000 tonnes perannum by 2010 [1,6].


196

Tab. 11-1 Consumption of natural fibers tonnes in the Europeanautomotive industry. (Reproduced with permission from ref. [1]).

Fiber type 1996 1999 2000 2005 2010

Flax 2,100 15,900 20,000 –[a] –Hemp[b] 0 1,700 3,500 – –Jute 1,100 2,100 1,700 – –Sisal 1,100 500 100 – –Kenaf 0 1,100 2,000 – –Coir 0 0 1,000 – –Total 4,300 21,300 28,300 50,000–70,000 >100,000

[a] Not estimated.[b] Industrial hemp is listed as a controlled substance in the U.S. and

cannot be used in commercial production of composites.

11.2Structure and Production Methods

The major steps in producing natural fibers for use in plastics include harvesting ofthe fiber-bearing plants, extraction of the fibers, and further processing of the rawfiber to meet required purity and performance aspects for use in plastic composites.

Methods exist for the harvesting of most natural fibers since they are used in themanufacture of products other than composites. For example, fibers derived fromwood are used in the paper and forest products industries, f lax fiber is used to makelinen and cigarette papers, and jute fiber is used in making rope and burlap [4]. Sincemany natural fibers are an annual crop, issues such as storage and variability in thegrowing season need to be considered. Europe is making large investments in newharvesting and fiber separation technologies for natural fibers such as f lax [7].

Fiber extraction procedures depend on the type of plant and portion thereof fromwhich the fibers are derived (e.g., bast, leaves, wood), as well as the required fiber per-formance and economics. Fiber bearing plants have very different anatomies (e.g.,tree versus dicotyledonous plants) and often fibers are derived from agriculturalresidues or by-products from industry [7]. Consequently, the processing needs candiffer greatly.

Wood is primarily composed of hollow, elongated, spindle-shaped cells (called tra-cheids or fibers) that are arranged parallel to each other along the trunk of the tree[8]. These fibers are firmly cemented together and form the structural component ofwood tissue. Fibers are extracted from wood by mechanical or chemical means dur-ing the pulping process.

Bast fibers such as f lax or kenaf have considerably different structures comparedto wood and, consequently, are processed quite differently. They are found in the in-ner bark of the stems of dicotyledons, and typically account for less than 30% of thestem [7]. Inside the inner bark is a woody core (called the “shive”) with much short-er fibers [4]. Fiber strands are removed from the bast. These fiber strands are sever-al meters long and are actually fiber bundles of overlapping single ultimate fibers.

11 Natural Fibers

197

Bast fibers are processed by various means that may include retting, breaking,scutching, hackling, and combing [4,7]. The exact process depends, to a large extent,on the type of plant and fiber source. For example, f lax fiber can be obtained fromdifferent f lax plants or from by-products of linen or f lax seed production [4]. Usefulnatural fibers have also been derived from other parts of plants, including leaves(e.g., sisal), seeds (e.g., coir), or grass stems [3]. The production of these fibers variesgreatly depending on the fiber type.

Much of the natural fiber used in composites is currently made into fiber mats,which are often needled, thermally fixed with small amounts of polymer fibers, orotherwise modified to improve handling, and then press molded [3]. However, thisadditional step comes with increased cost. The use of short fibers in more conven-tional processes such as injection molding is projected to increase in the future [9].When using natural fibers, practical processing issues such as feeding and meteringof the low bulk-density fibers, as well as fiber bridging, must be addressed. Someprogress has been reported on methods for pelletizing fibers to facilitate their intro-duction into polymer compounding equipment [10].

11.3Properties

11.3.1Chemical Components

The structure and chemical make-up of natural fibers varies greatly and depends onthe source and many processing variables. However, some generalizations are possi-ble. Natural fibers are complex, three-dimensional, polymer composites made upprimarily of cellulose, hemicellulose, pectins, and lignin [11]. These hydroxyl-con-taining polymers are distributed throughout the fiber wall. The major chemical com-ponents of selected natural fibers are listed in Table 11-2, reproduced with permissionfrom [3].

Tab. 11-2 Chemical compositions (%) of selected natural fibers [3].

Species Cellulose Lignin Pectin

Flax 65–85 1–4 5–12Kenaf 45–57 8–13 3–5Sisal 50–64 – –Jute 45–63 12–25 4–10Hardwood 40–50 20–30 0–1Softwood 40–45 36–34 0–1

Of the three major components, cellulose shows the least variation in chemicalstructure and can be considered the major framework component of the fiber. It is ahighly crystalline, linear polymer of anhydroglucose molecules with a degree of poly-

11.3 Properties

198

merization (n) of around 10,000. It is the main component providing the strength,stiffness, and structural stability. Hemicelluloses are branched polymers containingfive- and six-carbon sugars of varied chemical structure, the molecular weights ofwhich are well below those of cellulose but which still contribute as a structural com-ponent of wood [12]. Portions of the hemicelluloses are polymers of five-carbon sug-ars and are called pentosans [1].

Lignin is an amorphous, cross-linked polymer network consisting of an irregulararray of variously bonded hydroxy- and methoxy-substituted phenylpropane units[12]. The chemical structure varies depending on its source. Lignin is less polar thancellulose and acts as a chemical adhesive within and between fibers.

Pectins are complex polysaccharides, the main chains of which consist of a modi-fied polymer of glucuronic acid and residues of rhamnose [3]. Their side chains arerich in rhamnose, galactose, and arabinose sugars. The chains are often cross-linkedby calcium ions, improving structural integrity in pectin-rich areas [3]. Pectins areimportant in non-wood fibers, especially bast fibers. The lignin, hemicelluloses, andpectins collectively function as matrix and adhesive, helping to hold together the cel-lulosic framework structure of the natural composite fiber.

Natural fibers also contain lesser amounts of additional extraneous components,including low molecular weight organic components (extractives) and inorganic mat-ter (ash). Though often small in quantity, extractives can have large inf luences onproperties such as color, odor, and decay resistance [12]. The high ash content of somenatural materials, such as rice hulls, causes some concern about their abrasive na-ture.

11.3.2Fiber Dimensions, Density, and Mechanical Performance

Due to different species, a natural variability within species, and differences in cli-mates and growing seasons, natural fiber dimensions as well as physical and me-chanical performance can be highly variable. Methods of producing fibers with morereproducible properties are the goal of a major research effort [13].

Most natural fibers have a maximum density of about 1.5 g cm–3. Though somenatural fibers, such as wood, are hollow and have low densities in their native state,they are often densified during processing. Nevertheless, even the maximum densi-ty of these fibers is considerably less than that of inorganic fibers such as glass fibers.As such, their low density makes them attractive as reinforcements in applicationswhere weight is a consideration.

Table 11-3 summarizes the dimensions and Table 11-4 the mechanical propertiesof selected natural fibers. Though variable, high aspect ratios are found, especially forf lax and hemp. The mechanical performance of the fibers is good, but not as good asthat of synthetic fibers such as glass. However, their densities are considerably lower.The balance of significant reinforcing potential at low cost and low density is part ofthe reason why they are attractive to industries such as the automotive industry.

11 Natural Fibers

199

Tab. 11-3 Dimensions of selected natural fibers.

Length (mm) Width (µm)Fiber type

Avg. Range Avg. RangeRef.

Flax 33 9–70 19 5–38 [22]Hemp[a] 25 5–55 25 10–51 [22]Kenaf 5 2–6 21 14–33 [22]Sisal 3 1–8 20 8–41 [22]Jute 2 2–5 20 10–25 [22]Hardwood 1 – – 15–45 [8]Softwood – 3–8 – 15–45 [8]

[a] Industrial hemp is listed as a controlled substance in the U.S. andcannot be used in commercial production of composites.

Tab. 11-4 Mechanical properties of selected organic and inorganic fibers.

Fiber/f iber bundles Density Stiffness Strength Strain Ref.(g cm–3) (GPa) (MPa) (%)

Glass 2.49 70 2700 – [23]Kevlar 1.44 124 2800 2.5 [24]Nylon-6 1.14 1.8–2.3 503–690 17–45 [24]Polypropylene 0.91 1.6–2.4 170–325 80–100 [24]Polyester (staple) 1.38 1.5–2.1 270–730 12–55 [24]Flax[a] 1.4–1.5 50–70 500–900 1.5–4.0 [3]Hemp[a,b] 1.48 30–60 300–800 2–4 [3]Jute[a] 1.3–1.5 20–55 200–500 2–3 [3]Softwood 1.4 10–50 100–170 – [3]Hardwood 1.4 10–70 90–180 – [3]

[a] Fiber bundles[b] Industrial hemp is listed as a controlled substance in the U.S. and

cannot be used in commercial production of composites.

11.3.3Moisture and Durability

The major chemical constituents of natural fibers contain hydroxyl and other oxygen-containing groups that attract moisture through hydrogen bonding [15]. The mois-ture content of these fibers can vary greatly depending on the fiber type. The pro-cessing of the fiber can also have a large effect on moisture sorption. Table 11-5 showsthe wide range of moisture contents for different natural fibers at several relative hu-midities.

This hygroscopicity can create challenges both in composite fabrication and in theperformance of the end product. If natural fibers are used, a process that is insensi-tive to moisture must be used or the fibers must be dried before or during process-ing. Natural fibers absorb less moisture in the final composites since they are at leastpartially encapsulated by the polymer matrix. However, even small quantities of ab-

11.3 Properties

200

sorbed moisture can affect performance. Moisture can plasticize the fiber, alteringthe composite’s performance. Additionally, volume changes in the fiber associatedwith moisture sorption can reduce fiber–matrix adhesion and damage the matrix[14]. Methods of reducing moisture sorption include adequately dispersing and en-capsulating the fibers in the matrix during compounding, limiting fiber content, im-proving fiber–matrix bonding, chemically modifying the fiber, or simply protectingthe composite from moisture exposure.

Natural fibers undergo photochemical degradation when exposed to UV radiation[15]. They are degraded biologically because organisms recognize the chemical con-stituents in the cell wall and can hydrolyze them into digestible units using specificenzyme systems [15]. Though the degradability of natural fibers can be a disadvan-tage in durable applications where composites are exposed to harsh environments, itcan also be an advantage when degradability is desired.

Due to their low thermal stability, natural fibers are generally processed with plas-tics for which high temperatures are not required (less than about 200 ºC). Abovesuch temperatures, many of the polymeric constituents in natural fibers begin to de-compose. Since cellulose is more thermally stable than other chemical constituents,highly pulped fibers that are nearly all cellulose have been used to extend this pro-cessing window [16].

The release of volatile gases can, before, during, and after processing, lead to odorissues in applications where the composite is in an enclosed environment, such as inmany automotive applications, and especially when moisture is present [17].

11.4Suppliers

Natural fibers are used to manufacture a variety of products – linen, geotextiles, pack-aging, and specialty papers, for example. Natural fibers can be obtained from grow-ers, distributors, importers, and as by-products from other manufacturing process-es. Additionally, some companies sell semi-finished products made from naturalfibers (e.g., non-woven mats) that can be further processed into composites. Due to

11 Natural Fibers

Tab. 11-5 Equilibrium moisture content at 27 °C of selected natural fibers [22].

Equilibrium moisture content (%)Fiber

30% relative humidity 65% relative humidity 90% relative humidity

Bamboo 4.5 8.9 14.7Bagasse 4.4 8.8 15.8Jute 4.6 9.9 16.3Aspen 4.9 11.1 21.5Southern pine 5.8 12.0 21.7Water hyacinth 6.2 16.7 36.2Pennywort 6.6 18.3 56.8

201

the huge variety and diverse nature of natural fibers, there are currently few good re-sources that list a large number of manufacturers. However, with the growing use ofnatural fibers in plastics, they are beginning to be listed in plastics industry re-sources. For example, the following is a list of some of the major suppliers of naturalfibers to the North American composite industry from one industry resource [18].

Danforth Technologies (Point Pleasant, New Jersey, USA) Flaxcraft (Cresskill, New Jersey, USA) JRS Rettenmaier (Schoolcraft, Michigan, USA) Kenaf Industries of South Texas (Lasara, Texas, USA) Rayonier (Jesup, Georgia, USA) Creafill Fibers (Chestertown, MD, USA) Rice Hull Specialty Products (Stuttgart, Arkansas, USA)

Some of the major European fiber suppliers listed by another, online database [19]are:

AGRO-Dienst GmbH, Germany Badische Naturfaseraufbereitung GmbH, Germany Holstein Flachs GmbH, Mielsdorf, Germany Procotex SA Corporation, Belgium SANECO, France


Cost and availability of various natural fibers depend greatly on locale, region, andimport markets. For example, jute is commonly grown in India and Bangladesh, f laxis prevalent in Europe, and many non-wood natural fibers have to be imported intothe United States. Although non-wood agricultural fibers and agricultural fiberwastes are abundant worldwide, their source can be diffuse and infrastructure fortheir collection, purification, and delivery is sometimes limiting. Although there isincreasing interest in commercial uses of industrial hemp worldwide, it is listed as acontrolled substance in the U.S. and cannot be used in the commercial production ofcomposites.

The cost of natural fibers greatly depends on such factors as fiber type, location, de-gree of refinement, other markets for natural fibers, etc. However, some pricing fornatural fibers of interest in North America is given in Table 11-6 [25].


202


The environmental benefits of wood fibers and other natural fibers have been an im-portant inf luence on their use, particularly in Europe. Natural fibers are derivedfrom a renewable resource, do not have a large energy requirement to process, andare biodegradable [20].

Generally speaking, natural fibers are not particularly hazardous. However, natu-ral fibers have low thermal stability relative to other reinforcing fibers and can de-grade, release volatile components, and burn. Some basic precautions include avoid-ing high processing temperatures, using well-ventilated equipment, eliminating ig-nition sources, and using good dust protection, prevention, and control measures.Due to the wide variety of fibers classified as natural fibers, it is difficult to make spe-cific comments. For information on environmental and health risks, users shouldconsult their suppliers.

11.7Applications (Primary and Secondary Functions)

Recently, there has been a resurgence of interest in the use of natural fibers as rein-forcements in plastics due to their good mechanical performance and perceived en-vironmental advantages. Considerable funds have been expended on research aimedat introducing f lax, hemp, kenaf, and other natural fibers, especially in the automo-tive industry, with the greatest success being achieved in the use of mat technologiesin panel applications. However, issues such as: 1) lack of established delivery chan-nels, 2) processing complications due to the low density of the fibers (feeding, me-tering, and bridging), and 3) performance issues such as odor control, are limitingthe wider use of natural fibers as reinforcements in thermoplastics. These are activeareas of research and development.

11 Natural Fibers

Tab. 11-6 Pricing of selected natural fibers (Reprinted with permission from ref. [25]).

Fiber type Price (FOB, US$/kg)

Flax: core (6.4 mm and less) 0.44Flax: long (>6.4 mm) >0.73Kenaf 0.68Sisal (~3.2 mm) 1.03Hemp:[a] core (6.4 mm and less) 0.55Hemp:[a] long (>6.4 mm) >0.88

[a] Industrial hemp is listed as a controlled substance in the U.S. andcannot be used in commercial production of composites.

203

11.7.1Mechanical Property Modif ication (Primary Function)

The mechanical performance of natural fiber reinforced plastics varies greatly de-pending on the type of natural fibers, fiber treatments, the type of plastic, additives,and processing methods. Natural fibers are added to plastics to improve mechanicalperformance such as stiffness and strength without increasing the density or cost toomuch. The generally low impact performance of natural fiber composites tends tolimit their use and addressing this issue is an active area of research [1].

Natural fibers are hydrophilic and do not tend to be easily wetted or to bond wellwith many matrix materials, particularly the commodity thermoplastics. Couplingagents, such as maleated polyolefins, silanes, and isocyanates, are often necessary foradequate performance. A wide variety of coupling agents and fiber surface modifi-cations and treatments has been investigated for use in natural fiber plastic compos-ites and these are reviewed elsewhere [21].

Table 11-7 shows the mechanical performances of polypropylene composites madewith several different natural fibers. Not surprisingly, the fibers (i.e., pulp fibers, ke-naf fiber) are more effective reinforcements than the particulates (i.e., wood f lour;low aspect ratio wood fiber bundles). Wood fibers are an order of magnitude strongerthan the wood from which they derive [11] and the higher aspect ratio improves stresstransfer efficiency, particularly when a coupling agent is used. The addition of a cou-pling agent, namely maleated polypropylene, was found to improve performance, es-pecially f lexural and tensile strengths and unnotched impact energy. Fiber prepara-tion methods proved to have a large effect on reinforcing ability. The high perform-ance dissolving pulp fibers have nearly all of their non-cellulose components re-moved and are more effective than the lower cost, thermomechanical pulp fibers.

11.7.2Environmental Preference and Biodegradability (Secondary Function)

The environmental advantages of natural fibers are an important inf luence, particu-larly in Europe. Natural fibers are derived from renewable resources and often fromindustrial by-products. Wood and other natural fibers are derived from renewable re-sources, do not have a large energy requirement to process, and are biodegradable [6].They are lighter than inorganic reinforcements, which can lead to benefits such as fu-el savings when their composites are used in transportation applications. Naturalfibers can be used to reinforce biodegradable polymers since natural fibers them-selves are biodegradable.

11.7 Applications (Primary and Secondary Functions)

204

Authors’ Note

The Forest Products Laboratory is maintained in cooperation with the University ofWisconsin. This article was written and prepared by U.S. Government employees onofficial time, and it is therefore in the public domain and not subject to copyright.The use of trade or firm names in this publication is for reader information and doesnot imply endorsement by the U.S. Department of Agriculture of any product orservice.

11 Natural Fibers

Tab. 11-7 Effect of selected natural fibers on the mechanical performance of severalpolypropylenes (all composites contain 40 wt. % fiber).

Izod Impact[b] Flexural Properties[c] Tensile Properties[d]

Filler Type Coupling Notched Un- Max. Modulus Max. Modulus Elongation Ref.Agent notched strength strength at break

(J m–1) (J m–1) (MPa) (GPa) (MPa) (GPa) (%)

PP-1[a]

None No 20.9 656 38.3 1.19 28.5 1.53 5.9 [26]Wood f lour[e] No 22.2 73 44.2 3.03 25.4 3.87 1.9 [26]Wood f lour Yes[g] 21.2 78 53.1 3.08 32.3 4.10 1.9 [26]Thermomechanical No 22.2 90 48.9 3.10 29.7 3.68 2.1 [26]pulp (softwood)[f ]

Thermomechanical Yes[g] 21.3 150 76.5 3.50 50.2 3.89 3.2 [26]pulp (softwood)

PP-2[h]

None No 24 – 41 1.4 33 1.7 >>10 [28]Dissolving pulp Yes[g] – – 82.7 3.43 60.4 4.67 4.5 [27](softwood)[i]

Kenaf[j] Yes[k] 28 160 82 5.9 56 6 1.9 [28]

[a] Fortilene 3907, polypropylene homopolymer, meltf low index = 36.5 g/10 min., Solvay Polymers, DeerPark, TX, USA.

[b] ASTM D-256 [29].[c] ASTM D-790 [30].[d] ASTM D-638 [31].[e] Grade 4020 (American Wood Fibers, Schofield, WI,

USA).[f ] Laboratory produced from a mixture of pines.

[g] Maleated polypropylene (MP880, Aristech, Pittsburgh,PA, USA).

[h] Fortilene 1602, polypropylene homopolymer, meltf low index = 12 g/10 min., Deer Park, TX, USA.

[i] High purity cellulose pulp (Ultranier-J, Rayonier Inc.,Jessup, GA, USA).

[j] Kenaf strands obtained from AgFibers, Inc., Bakers-field, CA, USA.

[k] Maleated polypropylene, G-3002, Eastman ChemicalCompany, Longview, TX, USA.

205

References

References

1 Bledski, A. K., et al., “Natural and Wood Fi-bre Reinforcement in Polymers”, Rapra Re-view Reports, Rapra Technology Ltd., Shaw-berry, Shrewsbury, Shropshire, UnitedKingdom, 2002, 13(8), 144 pages.

2 Rials, T. G., Wolcott, M. P., “Physical andMechanical Properties of Agro-BasedFibers”, Chapter 4 of Paper and Compositesfrom Agro-Based Resources (Eds.: Rowell, R.M., Young, R. A., Rowell, J. K.), CRC Press,Inc., Boca Raton, FL, 1997, 63–81.

3 Lilholt, H., Lawther, J. M., “Natural OrganicFibers”, chapter 1.10 of Comprehensive Com-posite Materials, Vol. 1: Fiber Reinforcementsand General Theory of Composites (Ed.:Chou, T.-W.), Elsevier, New York, 2000,303–325.

4 McGovern, J. N., et al. “Other Fibers”,Chapter 9 of Pulp and Paper Manufacture,Volume 3: Secondary Fibers and Non-WoodPulping, 3rd ed. (Eds.: Hamilton, F.,Leopold, B., Kocurek, M. J.), TAPPI, At-lanta, GA, 1987, 110–121.

5 Plackett, D., “The Natural Fiber-PolymerComposite Industry in Europe – Technolo-gy and Markets”, Proc. Progress on Woodfi-bre-Plastic Composites Conference 2002, Mate-rials and Manufacturing Ontario and Uni-versity of Toronto, Toronto, Ontario, 2002,10 pages.

6 Suddell, B. C., Evans, W. J. “The IncreasingUse and Application of Natural Fibre Com-posite Materials within the Automotive In-dustry”, Proc. 7th International Conferenceon Woodfiber-Plastic Composites, ForestProducts Society, Madison, WI, 2003.

7 Young, R. A., “Processing of Agro-BasedResources into Pulp and Paper”, Chapter 6of Paper and Composites from Agro-Based Re-sources (Eds.: Rowell, R. M., Young, R. A.,Rowell, J. K.), CRC Press, Inc., Boca Raton,FL, 1997, 135–245.

8 Miller, R. B., “Structure of Wood”, Chapter2 in the Wood Handbook: Wood as an Engi-neering Material, General Technical Report,FPL-GTR-113, USDA Forest Service, ForestProducts Laboratory, Madison, WI, USA,1999, 463 pages.

9 Kaup, M., et al., “Evaluation of a MarketSurvey 2002: The Use of Natural Fibre inthe German and Austrian Automotive In-

dustries: Status 2002, Analysis and Trends”,presented at EcoComp 2003 Conference,September 1–2, 2003, London, U.K.

10 Jacobson, R., et al., “Low Temperature Pro-cessing of Ultra-Pure Cellulose Fibers intoNylon-6 and Other Thermoplastics”, Proc.6th International Conference on Woodfiber-Plastic Composites, Forest Products Society,Madison, WI, 2001, 127–133.

11 Rowell, R. M., “Opportunities for Lignocel-lulosic Materials and Composites”, Chapter2 of Emerging Technologies for Materials andChemicals from Biomass (Eds.: Rowell, R.M., Schultz, T. P., Narayan, R.), AmericanChemical Society, Washington, DC, 1992.

12 Pettersen, R. C., “The Chemical Composi-tion of Wood”, Chapter 2 of The Chemistryof Solid Wood (Ed.: Rowell, R. M.), Ameri-can Chemical Society, Washington, DC,1984, pp 76–81.

13 Kenny, J. M., “Natural Fiber Composites inthe European Automotive Industry”, Proc.6th International Conference on Woodfiber-Plastic Composites, Forest Products Society,Madison, WI, 2001, pp 9–12.

14 Peyer, S., Wolcott, M., “Engineered WoodComposites for Naval Waterfront Facilities”,2000 Yearly Report to Office of Naval Re-search, Wood Materials and EngineeringLaboratory, Washington State University,Pullman, WA, 2000, 14 pages.

15 Rowell, R. M., “Penetration and Reactivityof Cell Wall Components”, Chapter 4 of TheChemistry of Solid Wood (Ed.: Rowell, R. M.),American Chemical Society, Washington,DC, 1984, p 176.

16 Sears, K. D., et al., “Reinforcement of Engi-neering Thermoplastics with High PurityCellulose Fibers”, Proc. 6th InternationalConference on Woodfiber-Plastic Composites,Forest Products Society, Madison, WI, 2001,pp 27–34.

17 Bledzki, A. K., et al., “Odor Measurement ofNatural Fiber Filled Composites used forAutomotive Parts”, Proc. 9th Annual GlobalPlastics Environmental Conference: PlasticsImpact on the Environment, Detroit, MI, Feb-ruary, Society of Plastics Engineers, 2003.

18 Anonymous, “Special Report: Who’s Whoin WPC”, in Natural and Wood Fiber Com-posites, Principia Partners, 2003, 2(8), p 4.

206 11 Natural Fibers

19 “N-FibreBase”, www.n-fibrebase.net, M-Base Engineering+Software GmbH,Aachen, Germany, 2002.

20 Suddell, B. C., Evans, W. J., “The IncreasingUse and Application of Natural Fibre Com-posite Materials within the Automotive In-dustry”, Proc. 7th International Conferenceon Woodfiber-Plastic Composites, ForestProducts Society, Madison, WI, 2003.

21 Lu, J. Z., et al., “Chemical Coupling inWood Fiber and Polymer Composites: A Re-view of Coupling Agents and Treatments”,Wood and Fiber Science 2000, 32(1), 88–104.

22 Rowell, R. M., et al., “Characterization andFactors Affecting Fiber Properties”, Proc.Natural Polymers and Agrofibers Based Com-posites: Preparation, Properties, and Applica-tions, Embrapa InstrumentaçãoAgropecuária, São Carlos, Brazil, 2000,115–134.

23 Chamis, C. C., “Laminated and ReinforcedMetals”, in Encyclopedia of Composite Materi-als and Components (Ed.: M. Grayson), JohnWiley and Sons, New York, NY, 1983, p 613.

24 Billmeyer, F. W., Jr., Textbook of Polymer Sci-ence, 3rd ed., John Wiley and Sons, NewYork, NY, 1984, 502–503.

25 Anonymous, “Wood Flour Prices Rise, Nat-ural Fiber Prices Holding Steady”, Naturaland Wood Fiber Composites, Principia Part-ners, 2003, 2(7), 4–5.

26 Stark, N. M., Forest Products Journal 1999,49(6), 39–46.

27 Clemons, C. M., unpublished data.28 Sanadi, A. R., et al., Ind. Eng. Chem. Res.

1995, 34(5), 1889–1896.29 Test Method D256-02e1: “Standard Test

Methods for Determining the Izod Pendu-lum Impact Resistance of Plastics,” AnnualBook of ASTM Standards, Vol. 08.01, ASTMInternational, West Conshohocken, PA,2002.

30 Test method D790-03: “Standard Test Meth-ods for Flexural Properties of Unreinforcedand Reinforced Plastics and Electrical Insu-lating Materials”, see ref. [29].

31 Test Method D638-02a: “Standard TestMethod for Tensile Properties of Plastics”,see ref. [29].

12Talc

Vicki Flaris


Talc is a natural mineral found worldwide and is the major constituent of rocksknown as soapstone or steatite [1,2]. Montana ores have 85–95% talc, while New York,Vermont, and Canada ores contain 35–60% talc, with the remainder of the ore beingmagnesium carbonate. The purest of talcs are found in Montana, while the whitestcome from California; however, the latter are more abrasive because of asbestos-re-lated hard contaminants. Vermont talcs contain higher percentages of magnesiumand iron. Talc can be gray, green, blue, pink or even black [1,3].

Talc is mined in open-pit (for the majority of talc deposits) or underground opera-tions [3,5]. There are seven to eight steps to producing talc. The first step is overbur-den removal. This involves the removal of waste rock covering the talc vein by giantshovels (which can shift up to 1,500 tons of rock an hour) [4]. Secondly, the exposedtalc is extracted using shovels, and different ore types are sorted; this step is knownas the talc extraction step. Thirdly, the crude ore is crushed with rollers or jaw crush-ers to a size of 10–15 cm and sorted according to content and brightness by tech-niques such as hand sorting or state-of-the-art laser image analysis technology.

The type of further processing is dependent on the purity of the ore (dry vs. wetprocessing). Pure Montana talc can be dry processed. The fourth and fifth steps inmanufacturing talc involve grinding and classification. It is in these areas that mostadvances have been made in the last 20 years as consumers have realized the greatereffectiveness of fine particles (large surface area) over coarser ones [6]. Only for moredemanding applications such as the pharmaceutical industry is further purificationnecessary.

The size of the crushed ore can be further reduced using roller mills or cone crush-ers. The requirements of the talc–plastic composite determine the fineness of the talcneeded. A standard grinder roller produces 50 µm particles; finer grades are in therange 10–40 µm, and the finest grades are in the range 3–10 µm [2]. Roller-milledproducts are used in low impact strength polymer filled parts, such as fans in auto-motive under-the-hood applications. Fine micronizing (1–12 µm) is carried out in


208

hammer mills, tube mills, pebble mills or f luid-energy mills. Ball- or rod-millingwith steel media can discolor the talc, so ceramic grinding media are used instead.Acceptable grinding rates can be achieved [5]. Grind is measured in the plant by topsize, loose bulk density, and/or median particle size. Most suppliers report the lastmeasurement. The median particle size can be measured using laser light scatteringtechniques or from Stokes law settling rates [3]. It has only been possible to obtainsub-micron products since the advent of improved classification methods [6]. Classi-fication is usually carried out by air methods. To obtain the desired grades, theprocess conditions and type of equipment are critical.

High purity talc (97–98%) is obtained by means of wet methods. Manufacturersuse techniques such as froth f lotation, sedimentation, spray-drying, magnetic sepa-ration, centrifugation, and hydrocycloning. After the application of techniques suchas f lotation, the material is filtered, dried, and milled by jet mill micronization or inan impact mill. Bleaching agents are used when brightness is a major concern.

The sixth and seventh steps involve treatment of certain talcs. Some talc grades aresurface-treated with silanes for the rubber industry. Others are treated with glycolstearate to improve dispersability and processing. Amine-coated talcs are used for fer-tilizers, and cationic talcs for pitch control in paper-making. The surface treatment al-so helps with compatibilization reactions of certain components of polymer blends[1,4].

The last step in the production of talc involves delivery of the powdered mineral inbags, semi-bulk bags or bulk. Talc can also be delivered in pellet form or as a slurry.


Pure talc is a hydrated magnesium silicate with the chemical formulaMg3Si4O10(OH)2. The central brucite plane is chemically bonded by bridging oxygenatoms to two tetrahedral silica planes (see Figure 12-1). Talc, unless heated to above800 °C, has a plate-like structure (see Figure 12-2). The planar surfaces of the indi-vidual platelets are held together by weak van der Waals forces, which means that talccan be delaminated at low shearing forces. This makes the mineral easily dispersableand accounts for its slippery feel [1,5]. In contrast, the layers in mica are held togeth-er by ionic forces, while in kaolin hydrogen-bonding forces hold aluminosilicate lay-ers together. The size of an individual talc platelet can vary from 1 to >100 µm de-pending on the deposit. The platelet size determines the talc’s lamellarity. Highlylamellar talc has large individual platelets, whereas compact (microcrystalline) talchas smaller platelets [4]. Talc is accompanied by the mineral chlorite, in which Mg2+

ions have been substituted by Al3+ or Fe3+. Other mineral contaminants are magne-sium carbonate, mica, quartz, sericite, and often tremolite, a type of amphibole as-bestos [7]. The composition of talc is dependent on the source and the presence oftremolite. The U.S. Montana talcs are considered to be asbestos- and tremolite-free.California plate-like talcs have minor amounts (<3%) of tremolite. Hard talcs contain5–25% tremolite. Industrial talcs mined in New York contain 25–50% tremolite and

12 Talc


Fig. 12-1 Molecular structure of talc. Magnesium has octahedralcoordination (courtesy of Luzenac Inc.).

Fig. 12-2 Micrograph of Ultratalc 609 (courtesy of SpecialtyMinerals, USA).

210

other asbestiform minerals [1]. Certain talc compounds have odor issues dependenton the source. The odor is believed to result from interaction of the talc with stabiliz-ers added to the polymer-based formulation.

The theoretical chemical composition of pure talc by weight is 19.2% magnesium,29.6% silicon, 50.7% oxygen, and 0.5% hydrogen. In terms of metal oxides it is 31.7%MgO and 63.5% SiO2, with the remaining 4.8% being H2O. Other elements found inimpure talcs in variable amounts are Ca, Al, and Fe. Trace elements include Pb, As,Zn, Ba, and Sb [1,2].

Talc-filled composites have low gas permeability and high resistivity because of theplate-like nature of the impermeable talc particles and the resulting tortuous, com-plicated diffusion path. Talc is also unique in its ability to easily delaminate and canbe used as a lubricant. Talc is the softest mineral on the Mohs hardness scale (equiv-alent to 1 and used as a standard). Commercial grades are, however, usually some-what harder as a result of impurities [5]. In general, as the mineral is soft it is also lessabrasive. This is advantageous as there is reduced wear on processing equipment(such as extruders). The surface of talc is hydrophobic, which has been explained interms of the high ionic character of the central magnesium plane, uniform polarityand symmetry of the structure, and neutrality of the layers. The hydrophobic natureof talc allows it to be more compatible with polymer resins. Hydrophobicity can befurther increased through coating with zinc stearate. The physical and chemicalproperties of talc are summarized in Table 12-1 [1,4,5].

Tab. 12-1 Characteristics of talc [1,2,24].

Property Data

Crystal structure monoclinicTypical chemical composition, wt. %

MgO 24.33–31.9SiO2 46.4–63.5CaO 0.4–13Al2O3 0.3–0.8Fe2O3 0.1–1.8

Platelet aspect ratio 5–20Density, kg m–3 2.58–2.83 × 103

pH 9.3–9.6Oil absorption (ASTM D281) 20–57%Refractive index 1.54–1.59Mohs hardness 1–1.5Brightness 78–93Thermal conductivity, W K–1 m–1 0.02Specific heat, J kg–1 K–1 8.7 × 102

Coefficient of thermal expansion, K–1 8 × 10– 6

Particle size and shape are very critical with regard to the final mechanical proper-ties of the composite. For a medium porosity particle material, the specific surfacearea is in the range 3–20 m2 g–1. Typical particle size distributions of coarse and fine

12 Talc

211

products are shown in Figure 12-3. Coarser products have an average size of about10–20 µm and a top size of up to 75 µm. For the coarser grades, particle size infor-mation is indicated by top-particle-size sieve data. Particle size information for thefiner grades is obtained by either sieve or sedigraphic light-scattering water tech-niques. Fine talcs may be as low as 1 µm in size with an upper size limit of 12 µm.Particle thickness varies in the range 0.2–6 µm. Fine talcs (for example, through therecent introduction of nanotalcs from Nanova LLC) are very important in plastics ap-plications [6].

Talc is inert to most chemical reagents. It contains about 5 wt. % water, which ischemically bound to the magnesium oxide or the brucite layer. The associated wateris lost between 380 °C and 500 °C. Above 800 °C, talc progressively loses hydroxylgroups, and above 1050 °C it recrystallizes to form enstatite (anhydrous magnesiumsilicate) through an endothermic reaction with the liberation of water. The estimatedmelting point of talc is 1500 °C [1,3,5].

The surface chemistry of talc is not well understood. Reactive groups that form up-on talc’s fractured surface are: a) weakly acidic terminal hydroxyl groups (HO–Si →H+ + [O–Si]–); b) sites of proton release through polarization of water molecules(Mg2+ + H2O → 2H+ + MgO); c) Lewis acid sites (metal ions react with paired elec-trons on water molecules); d) octahedral iron ox/redox sites (FeII/FeIII for any Fe pres-


Fig. 12-3 Typical particle size distribution of talc products(reproduced from ref. [5]).

212

ent in talc ore); e) strongly acidic Brønsted sites (Mg2+ + HO–Si- → H+[Si–O–Mg]+);f ) weakly basic sites (Mg(OH)2) [3]. Interaction with polymeric matrices may pre-sumably be achieved through some of the above functionalities. Luzenac® R7 talc isa result of improvements in a proprietary surface modification technology. R7 talc in-teracts synergistically with the rubber domains of a thermoplastic olefin elastomer(TPO) matrix to improve f lexural modulus and maintain impact strength. It also im-proves UV stability, color retention, and thermal stability (see Table 12-2). R talc is ef-fective with ethylene-propylene-diene monomer (EPDM) and ethylene-propylenerubbers (EPR), but not as effective with metallocene elastomers containing TPOs. Re-cently, Clark [8] presented evidence of improvements in scratch and mar resistancewhen talc is grafted to TPO. The need for higher performance materials will be thedriving force behind the development of new surface modification technologies [3].

Tab. 12-2 Thermal stability of homopolymer PP (a comparisonbetween unmodified and surface-modified talc of the same particlesize) [8].

wt. % Talc0 15 27 40

Days to FailureLuzenac® R7 Talc 132 118 105 90Unmodified Talc 132 95 80 18

12.3Suppliers

The largest producer of talc worldwide and in the USA is the Luzenac Group, with to-tal annual production increasing from 1.26 million tons in 2000 to 1.33 million tonsin 2002 following the acquisition of the Three Springs mine in Australia [9]. Majortalc suppliers and grades recommended for use in plastics are listed in Table 12-3.


As of the end of 2003, most suppliers quoted typical grade prices in the range of US$0.15/kg to $0.66/kg, with some premium grades costing up to $1.48/kg. The nano-talcs are quoted at $5.50/kg to $12/kg. Availability of any of the talcs does not seemto be an issue. Growth in talc consumption in plastics is forecast at about 4% per an-num to 2006, in line with increased use in polypropylene. In 2001, demand for talc inplastics in North America was around 180,000 tons per annum. Six countries – Chi-na, USA, India, Finland, France, and Brazil – accounted for 80% of the total worldoutput in 2002; China accounted for 40% and the USA for 13%. On the supply side,

12 Talc


Tab. 12-3 Major talc suppliers.

Supplier Grades

ATC/Possehl USA BHS-325, 400, 600, 800, 1000, 1250, 1500, 2000, (distributors of Asada Japan) 5000One Maynard Drive, Park Ridge, NJ 07656, USATel.: (201) 307 1500Fax: (201) 307 0540Website: www.possehl.com

Canada Talc Ltd. Cantal® 45–80, 45–85, 45–90, 290, 490, 690, (subsidiary of Dynatec Mineral Products) Talcor® 5600Suite 715, 734 – 7th Avenue S.W., Calgary, AB, T2P 3P8, CanadaTel.: (403) 261 3999Fax: (403) 264 2959Website: www.dynatec.ca

IMI Fabi USA (produces talc at Natural Benwood 2202, 2203, 2204, 2207, 2210, 2213Bridge, NY, Benwood, WV sites) Talc HTPultra 5, 5c, 10, 10cBenwood Plant, Second & Marshall Street, Talc HTP 05, 05c, 1c, 2, 3, 4Benwood, WV 26031, USATel.: (304) 233 0050Fax: (304) 232 0793Website: www.hitalc.com; www.imifabi.com

Luzenac America Inc. Cimpact® 610, 699, 710, 710R, 710HS345 Inverness Drive South, Suite 310, Jetfil® 7C, 350, 575C, 625C, 700CCentennial, CO 80112, USA Nicron® 674Tel.: (303) 643 0400, 1-800-3250299 Mistron® ZSC, AB, NT, 400CFax: (303) 6430446 Artic Mist®

Website: www.luzenac.com/us.html Stellar® 410, 420, 510Vertal® 97, UA40Luzenac® 8230Luzenac® R7

Nanova LLC (subsidiary of Nanomat Inc.) NanoTalcTM TP1001061 Main St. Bldg#1, North Huntingdon, PA 15642, USATel.: (724) 978 2190Fax: (724) 861 6119Website: http://www.nanomat.com/nanova/left.htm

Non-Metals Inc.- Affiliate of CNMIEC, China PP500, 12501870 W. Prince Rd., Suite 67, Tucson, AZ 85705, USATel.: (520) 690 0966, 1-800-320 0966Fax: (520) 690 0396Website: www.nonmetals.com

214

world production of talc fell by around 10% from 1996 to 2002, due to lower outputfrom China and the USA [9,10].


Safety issues regarding talc have been controversial. The debate revolves around as-bestos-type impurities, particularly tremolite. Most commercial talc grades containno asbestos, and those that do contain only trace amounts. The Occupational Safetyand Health Administration (OSHA) and Mine Safety and Health Administration(MSHA) have an 8-hour time-weighted exposure limit for non-asbestos containingtalc dust of 3 × 10– 6 kg m–3. For the personal product industries, more rigorous con-trols are instituted and set by the American Industrial Hygiene Association (AIHA)and the Cosmetic Toiletry and Fragrance Association (CTFA). Talc is approved by theU.S. Food and Drug Administration (FDA) for use in polymeric compounds in con-

12 Talc

Tab. 12-3 Continued

Supplier Grades

Specialty Minerals Inc. (SMI acquired Polar MICROTALC® MP 12–50, 30–36, 44–26Minerals, Pfizer Microtalc, ULTRATALC® 609and Barretts Minerals) ABT® 250035 Highland Ave., Bethlehem, PA 18017, USA CLEARBLOC® 80Tel.: 1-800-801 1031 MICROTUFF® AG 609, 101, 121, 262, 445Fax: (610) 882 8726 POLYBLOC® antiblock 9102, 9103, 9103S, 9107, Website: www.mineralstech.com 9110, 9310, 9405, 9410

OPTIBLOC® 10, 25FLEX TALC™ 610, 815, 1222ULTRATALC® 609 & MICROTUFF® AG 609 – ultrafine grades less than 1 µm

R. T. Vanderbilt Company Inc. Nytal® 100, 100R, 200, 3X, 300, 300H, 3300, 30 Winfield St., P.O. Box 5150, Norwalk, 400, 7700CT 06856-5150, USA Vantalc® 6H, 6H-II, F2003, F2504, PC, RTel.: (203) 853 1400Fax: (203) 853 1452Website: www.rtvanderbilt.com

Zemex Industrial Minerals Pioneer 2606, 4317, 4319, 4404(owns Suzorite Mineral Products (SMP), Van Horn, TX)1040 Crown Pointe Parkway, Suite 270, Atlanta, GA 30338, USATel.: (770) 392 8660Fax: (770) 392 8670Website: www.z-i-m.com

215

tact with food (Listing – Title 21, Code of Federal Regulations Part 178.3297, “Col-orants for Plastics”). There are minimal concerns with skin contact, other than dry-ness with continuous exposure. Eye contact causes a mild mechanical irritation,while ingestion is of no concern. Approved dust masks should be worn when theAmerican Conference of Governmental Industrial Hygienists (ACGIH) thresholdlimit value (TLV) of 2 × 10– 6 kg m–3 is exceeded, according to the National Institute ofOccupational Safety and Health (NIOSH) and now also OSHA. Studies by Porro etal., Siegal et al., McLaughton et al., and Kleinfeld et al. [11] have shown that long-terminhalation can lead to mild scarring of the lungs (pneumonoconiosis symptoms:wheezing, chronic cough, shortness of breath). Pneumonoconiosis was found to becaused by fibrous varieties of talc and the particle length rather than the compositionof the talc seemed to be of importance. Other adverse health effects are associatedwith covering of the lungs (pleural thickening) [3,5,12].

12.6Applications

12.6.1General

Talc is an important reinforcing filler for plastics, in particular polypropylene (PP).The major benefits of incorporating talc into plastics are summarized in Table 12-4according to its primary and secondary functions, with examples obtained from refs.[1,2,3,8,13,17,21]. Primary reasons for using talc include improvements in mechani-cal properties such as heat def lection temperature (HDT), rigidity, creep resistance,and sometimes impact resistance, as well as lower shrinkage. Additional secondarybenefits, because of the f laky nature of talc, include improvements in dimensionalstability (as it orients along f low lines during molding) and lower permeability; oth-er benefits are reduced coefficient of thermal expansion (CLTE), increased bright-ness, shortening of the injection molding cycle due to nucleation, and its use as ananti-block additive [7,8,10,13,14]. Adverse effects include reduction in toughness andelongation at break in certain polymers, reduced weld line strength, and for certainpolymer/stabilizer package combinations a reduction in long-term thermal ageingand UV resistance. In the rubber industry, talc is used to increase stiffness andprocessability.

Talc-filled masterbatches are available at up to 75 wt. % loadings. Talcs densifiedthrough Zero Force compaction/densification technology (Luzenac) have shown bet-ter mechanical properties and higher throughput rates during processing than regu-lar talcs subjected to the compressive forces of pelletization [3]. Dark talc is usedwhere color is not of prime importance, such as in certain exterior and interior au-tomotive applications, and when the reinforcing properties are not so critical. Whitetalcs are used in applications involving washers and dryers and in garden furniture.Composites of lower f lexural modulus are based on talcs containing more granularminerals such as quartz or magnesite. Composites of higher f lexural strength and

12.6 Applications

216

modulus are based on talc grades that are more homogeneous in terms of particlesize distribution and that contain a higher percentage of coarser particles [7].

12.6.2Applications by Polymer Matrix

12.6.2.1 Polyolef insDue to its plate-like nature, talc is used in linear low density polyethylene (LLDPE) asan anti-blocking agent, preventing two or more contacting film layers from stickingtogether (see Chapter 20), and also as a nucleating agent [7]. Recently, talc has found

12 Talc

Tab. 12-4 Primary and secondary benefits of talc in plastics.

Function Examples

Primary FunctionMajor improvements in HDT – In homopolymer PP, increase by 60 ºC at

40% loading– In PP copolymer, increase by 75 ºC at 40%

loading– In commingled post-consumer stream (80%

PE, remainder PET, PS, PP, PVC) at 40%loading, a 6 ºC increase; in maleated streamwith 40% loading, a 13 °C increase

Major improvements in modulus – In homopolymer PP, double stiffness at 40%loading

– In PP copolymer, fourfold increase at 40%loading

– In commingled stream at 40% loading, a130% increase

Minor improvements in impact resistance – In PS with an elastomer and talc (see for certain polymers Figure 12-6)

– Only maleated commingled stream with 40%loading showed an increase in Izod impactstrength (22%)

Increase in tensile strength – 2–10% concentration of talc in HDPE increas-es tensile strength by 15–80%

– In commingled post-consumer stream at 40%loading, an 8% increase, whereas in maleatedstream with 40% loading, a 19% increase.

Secondary FunctionsCLTE/mold shrinkage decrease – At 30% loading, mold shrinkage decreased by

57% in homopolymer PP and by 39% in PPcopolymer

Lower permeability – With 2–10% talc loading in HDPE a decreasein permeability of 15–55%

Nucleating capacity – Talc-filled PS in PS foamEnhance moisture barrier – 20% loading in PP increased barrier by 50%Anti-tack agent – Rubber blends

217

some new applications, such as reducing the necessary dosage of f luorocarbon-elas-tomer polymer processing aids (PPA) and enhancing the moisture barrier in pack-aging films [15–19]. Because of its non-polar nature, a major application of talc inhigh density polyethylene (HDPE) (where a 2–10 wt. % concentration leads to a15–80% increase in tensile strength) is in wires and cables. Grades coated with zincstearate are used in cross-linked low density polyethylene (XLPE) wire coatings asf lame retardancy aids. These talc grades increase char build-up and also act asthixotropic agents and reduce dripping.

Talc is mostly used with PP, with typical loadings in the range 10–40 wt. %. It pro-vides benefits such as stiffness, dimensional stability, enhanced thermoforming,opacity, whiteness, and high temperature heat resistance for automotive (fan shroudsand blades) and appliance (washer tubs, pump housings, spin baskets) applications[20]. Small amounts of stabilizers are additionally needed for more demanding auto-motive applications, such as under-the-hood, bonnet, dashboard, bumper, and inte-rior and exterior trim. Other applications include fascias and kickplates. The purityof talc affects its efficiency in improving thermal properties, since even low levels ofmetal ions can catalyze polymer degradation. Weaknesses of talc in PP applicationsare low scratch resistance and low impact strength. Table 12-5 summarizes the effectsof mineral loading on the properties of PP homopolymer and copolymer [5,21].

Increasing talc concentration has a direct effect on stiffness, while particle size hasa less pronounced effect. Macrocrystalline talc (talc with large aspect ratio, length tothickness) will impart greater stiffness than microcrystalline talc. The addition of talcgenerally reduces notched impact strength, the effect being largely dependent on par-ticle size. For example, in a 30% loaded copolymer blend, fine talc can promote duc-tile fracture whereas coarse talc can promote brittle fracture. The fineness of the talcis even more critical in relation to low temperature impact strength. Figure 12-4 sum-marizes the wide range of stiffness/toughness properties that can be achieved withdifferent talc grades in polypropylene resins [22].

Among other properties, talc is beneficial in reducing mold shrinkage. Clark andSteen [3] have shown that at a 30 wt. % loading, mold shrinkage is reduced by 57% inhomopolymers and by 39% in impact copolymers. In a 30% talc-filled PP compound,the filler decreases the coefficient of thermal expansion by half in the temperaturerange 50–150 ºC; thermal expansion does not seem to be affected by particle size. At

12.6 Applications

Tab. 12-5 Typical properties of talc-filled homopolymer andcopolymer PP [5,21].

Property Homopolymer CopolymerUnf illed 20 wt. % 40 wt. % Unf illed 20 wt. % 40 wt. %

Density, 103 kg m–3 0.903 1.05 1.22 0.899 1.04 1.22Flexural modulus, MPa 1655 2482 3275 756 2206 2896Yield tensile strength, MPa 35.5 34.1 31.4 27.6 27.9 25.8Rockwell R Hardness 99 98 95 82 87 85Heat def lection temp., °C (455 kPa) 97 123 131 85 117 127Notched Izod impact strength, 45.7 32.0 20.9 133.5 53.4 32.0J m–1 (22 °C)

218

a 15 wt. % loading, talc increases the PP crystallization temperature from 115 to123 ºC. Talc can also be used to impart barrier properties in PP films. At a 20 wt. %loading, the moisture vapor transmission rate (MVTR) in a 0.2–0.25 mm film is re-duced by 50%. The thermal stability is affected by the crystallinity of the talc (high pu-rity talc is more thermally stable), its specific surface area, and its heavy metal con-tent. Heavy metal cations linked to the silica lattice slightly affect the thermal stabili-ty, while carbonate anions can decrease the thermal performance of talc-filled PPcompounds [3,21].

Surface treatment improves properties by increasing dispersion of the particlesand providing adhesive sites [23]. It has been shown that a maleic anhydride basedcoupling agent can increase tensile and f lexural properties [13]. For 20 and 40% talc-filled PP, levels of 20 wt. % of an acrylic acid modified polypropylene yield optimalphysical properties, such as HDT and tensile and f lexural properties, but have no ef-fect on impact strength [23]. The bonding mechanism is believed to involve: a) bond-ing between magnesium ions of the talc surface and the acrylic acid of the modifiedPP; a magnesium salt is formed and complexed by oligomeric acrylate ions and heldbetween platelets of the poly (Si4O11

6–) ions, and b) co-crystallization of the chains ofthe acid-modified PP with the PP matrix. Particle-size recommendations for talc-filled polyolefins are summarized in Figure 12-5 [3].

12 Talc

Fig. 12-4 Polypropylene-talc systems covering a wide range of stiff-ness/toughness properties (reproduced with permission fromIndustrial Minerals Information [22]).

219

12.6.2.2 Polyvinyl ChlorideA major application of talc in polyvinyl chloride (PVC) is in f looring, where talc load-ings can be as high as 50%. The talc has to be ultra-dry for this application to preventstreaking (moisture vaporizing during processing). Another minor application ofsurface-treated talcs is in the wire and cable industry, where requirements of tensilestrength and dielectric properties must be met [5].

12.6.2.3 Styrenics and Thermoplastic ElastomersIn polystyrene (PS), talc is used in combination with an elastomer to overcome re-duction in impact properties (note the improvement in Table 12-6 from ref. [5]). An-other commercial application is in PS foam, where it is used as a cell nucleatingagent. In rubber blends, talc is used as an anti-tack agent to prevent newly formedgoods from sticking together. In some specialty elastomers it reduces air and f luidpermeability [5].

Other market applications of talc involving polymers outside the filled thermo-plastics industry include [3,5]:

interior and exterior architectural coatings, where it is used to control gloss andsheen, improve opacity, tint strength, and weatherability, and enhance viscosity andsag resistance;

12.6 Applications

0 10 20 30 40 50 60

Top Size, µm

12

11

10

9

8

7

6

5

4

3

2

1

0

Med

ian

Part

icle

Siz

e, µ

m Commodity-Appliance

Automotive-Under-Hood

Good Performance Automotive

TPO and Nucleation

Fig. 12-5 Range of particle sizes for talc-filled polyolefin applica-tions (reproduced from ref. [3]).

220

industrial coatings to provide f latting and sandability, and to enhance package sta-bility and water and chemical resistance;

paints for hiding power, matting effect, and satin finish; roofing [25].

References

1 Wypych, G., Fillers, ChemTec Publ., Toron-to, Ont., Canada, 1993, pp 43–46.

2 Wypych, G., Handbook of Fillers, ChemTecPubl., Toronto, Ont., Canada, 2000, pp150–153, 663–667.

3 Clark, R. J., Steen, W. P., Chapter 8 inHandbook of Polypropylene and Polypropy-lene Composites (Ed.: Karian, H. G.),Marcel Dekker Inc., New York, 2003, pp281–309.

4 www.luzenac.com5 Sekutowski, D., Section IV in Plastics Addi-

tives and Modifiers Handbook (Ed.: Eden-baum, J.), Chapman & Hall, London, U.K.,1996, pp 531–538.

6 Harris, P., Industrial Minerals, Oct. 2003,443, 60–63.

7 Cordera, M., Proc. Functional Fillers andFibers for Plastics 98, Intertech Corp. 5thInternational Conference, Beijing, P.R. Chi-na, June 1998.

8 Clark, R., Proc. Functional Fillers for Plas-tics 2003, Intertech Corp., Atlanta, GA, Oct.2003.

9 www.roskill.com/reports/talc10 Harris, T., Proc. Functional Fillers for Plas-

tics 2003, Intertech Corp. Atlanta, GA, Oct.2003.

11 www.cdc.gov/niosh/pe188/14807-96.html.12 www.rtvanderbilt.com.13 Xanthos, M., et al., Polym. Compos. 1995,

16(3), 204.

14 Gill, T. S., Xanthos, M. J. Vinyl. Addit. Tech-nol. 1996, 2(3), 248.

15 Fazzari, A. M., Modern Plastics Encyclope-dia, McGraw-Hill, New York, 1997, 74(13),C3.

16 Graff, G., Modern Plastics 1998, 75(5),32–33.

17 Deutsch, D. R., Radosta, J. A., Proc. Poly-olefins XI, The SPE International Confer-ence on Polyolefins, Houston, TX, 1999, pp657–677.

18 Chapman, G. R., et al., The SPE Interna-tional Conference on Additives for Poly-olefins, 1998, Houston, TX, pp 149–171.

19 Amos, S., Deutsch, D. R., Proc. TAPPI Poly-mers, Laminations and Coatings Conf.,TAPPI Press, Atlanta, GA, 1999, pp829–847.

20 Posch, W., Proc. Functional Fillers for Plas-tics 2003, Intertech Corp., Atlanta, GA, Oct.2003.

21 J. A. Radosta, Proc. Functional Fillers forPlastics 95, Intertech Corp., Houston, TX,1995.

22 Holzinger, T., Hobenberger, W., IndustrialMinerals, Oct. 2003, 443, 85–88.

23 Adur, A. M., Flynn, S. R., Proc. 45th SPEANTEC, 1987, 33, 508.


25 www.mineralstech.com

12 Talc

Tab. 12-6 Properties of elastomer-modified, talc-containingpolystyrene [5].

High Impact 40% Surface-Treated 30% Surface-Treated Polystyrene Talc, 10% TPE, Talc, 10% TPE,

50% Crystal PS 60% Crystal PS

Falling-weight impact strength, J 4.3 3.6 3.6Notched Izod impact strength, J m–1 69.4 74.8 101.5Flexural modulus, GPa 2.38 4.2 3.71Tensile yield strength, MPa 37.1 35.0 37.1

13Kaolin

Joseph Duca

13.1Introduction

The term kaolin encompasses a group of minerals, the dominant one being kaolin-ite. In industry, the term kaolin is used to refer mainly to the mineral kaolinite, andthis is the implied meaning throughout this chapter.

The lesser minerals of the kaolin group comprise hydrated aluminosilicates suchas dickite, nacrite, and halloysite [1]. Structurally, kaolinite consists of an aluminaoctahedral sheet bound on one side to a silica tetrahedral sheet, stacked alternately.The two sheets of kaolinite form a tight fit, with the oxygen atoms forming the linkbetween the two layers (Figure 13-1) [2]. The theoretical composition for theAl2O3·2SiO2·2H2O mineral is 46.3% SiO2, 39.8% Al2O3, 13.9% H2O [3].

Kaolin is considered to be a phyllosilicate mineral. Phyllosilicates are characterizedby an indefinitely extended sheet of rings, in which three of the tetrahedral oxygensare shared, while every fourth oxygen is apical and points upwards. These phyllosili-cates also often have a hydroxyl group centered between the apical oxygens. This oc-curs through bonding of the silica sheet to a continuous sheet of octahedra, with eachoctahedron tilted onto one of its triangular sides. In kaolin, these octahedra containthe trivalent aluminum cation. To balance the charge, only two of every three alu-minum octahedral positions are occupied by aluminum cations to form a gibbsitestructure. Hence, a layer of silica rings is joined to a layer of alumina octahedrathrough shared oxygens resulting in a plate-like morphology. The individual kaolinparticle has an oxygen surface on one side and a hydroxyl surface on the other side,as shown in Figure 13-1 [2]. This means that such layers can stack through hydrogenbonding to the lamella above and below. Consequently, kaolin is often shown to be inwhat are called “booklets”, which are stacks of plates, one on top of the other, and con-nected through hydrogen bonding (Figure 13-2). Interestingly, talc, on the otherhand, has a characteristic soft, slippery feel, due to easy sliding and delamination ofits platelets, which are held together through relatively weak van der Waals forces (seealso Chapter 12) [1]. In contrast, kaolin interlaminar bonding is due to stronger hy-drogen bonding involving an oxygen face from the silica layer bonded to a hydroxyl


222

face from the gibbsite layer. Unlike talc, kaolin booklets can only be delaminated bysignificant grinding [1].

Kaolin is an extremely versatile white functional filler, with applications in papercoating, paper filling, paints, plastics, rubber, and inks. The plastics and adhesives in-dustries consume some 65,000 tons of kaolin in the U.S. per year. This can be ex-pected to increase, especially as the price of resins increases. Among other marketsare pharmaceuticals, cracking catalysts, and ceramics.


Kaolins are classified as either primary or secondary. Primary kaolins are formed bymodifications of crystalline rocks such as granite. The source of this kaolin is foundwhere it is formed. Conversely, secondary kaolin deposits are sedimentary and areformed by erosion of primary deposits. The secondary deposits contain much morekaolinite (about 85–95%) than the primary deposits, which contain only 15–30%. Thebalance of the ore consists of quartz, muscovite, and feldspar in the primary depositsand quartz, muscovite, smectite, anatase, pyrite, and graphite in the secondary de-posits. Kaolin, also known by the common term clay, is usually open-pit mined in theUSA from vast deposits in Georgia, South Carolina, and Texas. The ore is notprocessed in one specific way. There are also distinct methods for ore beneficiation,each method adding value to the mineral.

13 Kaolin

Fig. 13-1 Structure of kaolin.

223

13.2.1Primary Processing

Commercially available kaolin grades can be produced by air-f loating or wet-pro-cessing. Air-f loated kaolin is the least processed and therefore the least expensive.Here, the mineral is crushed, dried, and pulverized. Thereafter, it is f loated in an airstream and classified using an air classifier. The finer particles are separated fromthe coarse ones and from the non-kaolin particles, which are referred to as grit.

The much more sophisticated wet process delivers products that are higher in pu-rity, having consistent high quality and with a broader range of particle sizes and dis-tributions. The wet process starts near the mine site with the formation of a clay-wa-ter suspension in a step referred to as “blunging”. In this step, kaolin slurry is pro-duced to contain 40–50% solids. This slurry is then pumped to the process plant fordegritting, manipulation of particle size distribution and/or morphology, leaching ofcolor bodies, and other value-adding process steps. Lastly, apron-drying and pulver-ization or spray-drying into bead form takes place. A schematic of the wet processcontaining the various beneficiation steps discussed below is shown in Figure 13-3.The final product is packaged in 25 kg up to one ton bags. Twenty-ton containers areoften employed for overseas shipments. Additionally, the final product might arriveas a high solid content slurry, which is transported to the customer in rail cars.

13.2.2Benef iciation

Particle size and shape may be manipulated by centrifugation using a combinationof solid-bowl and disk-nozzle centrifuges to classify the particles according to size.


Fig. 13-2 “Booklets” of kaolin.

224 13 Kaolin

Fig. 13-3 Overview of the wet kaolinproduction process.

225

The interplay between the two types of centrifuges allows for a wide array of particlesize distributions. The efficiency of separation of coarse and fine particles dependson many process variables, including the initial particle size distribution, the viscos-ity of the feed stream, the retention time in the centrifuge, and the mechanical con-figuration of the centrifuge itself.

An additional manipulation is delamination, which changes both the apparent par-ticle size distribution and the morphology of the particles. Coarse kaolin slurry ispumped into an agitation tank containing plastic or glass-bead media. The objectiveis to cleave the stacks of kaolin without breaking the platelets and causing excessivefines. Delamination results in the formation of higher aspect ratio individualplatelets by breaking the weak hydrogen bonds that hold the stacked platelets togeth-er as booklets (see Figure 13-4). The delaminated platelets originating from Georgia,U.S.A., average about 0.15 µm in thickness and about 0.6 µm in diameter. Kaolin sup-pliers such as Engelhard Corporation, Thiele, J. M. Huber, and Imerys all mine outof Georgia, U.S.A. Higher aspect ratios are possible depending on the kaolin feed in-to the delaminating process. As discussed in Chapter 2, high surface area and highaspect ratio are essential for efficient stress transfer in polymer composites, althoughthey may result in higher viscosity. The particle size distribution also has an impacton melt viscosity, with a narrower distribution corresponding to a higher viscosity.This effect is more significant at higher kaolin loadings.

The brightness and color of the kaolin can be improved through various methods.High-intensity magnetic separation (HIMS) is a continuous, semi-batch process thatuses cryogenic, superconducting magnets cooled by liquid nitrogen and helium.Kaolin slurry is passed through a steel-wool matrix within a magnetic coil. The ma-trix retains paramagnetic impurities such as iron oxide and anatase. HIMS increas-es the GE brightness of the mineral by 5 points, the latter being defined as the direc-tional ref lectance at 457 nm [4]. Periodically, the HIMS is shut down and the steel


Fig. 13-4 Scanning electron micrograph showing individual kaolinplatelets.

226

wool matrix is f lushed with water. The discharge is often a deep-red slurry of mag-netic rejects consisting mainly of dark reddish-purple hematite and yellowish iron-enriched anatase. Another option for improved brightness is froth f lotation, whichremoves (although not entirely) anatase from the kaolin slurry. Flotation decreasesthe content of anatase color bodies from about 2% down to 0.7%. During the process,the iron-stained anatase-laden froth rises to the top of the vessel and is skimmed offand discarded. Another step is reductive bleaching, which converts the iron’s valencefrom +3 to +2, thereby dispelling the color and enhancing brightness by about 5points [5].

By lowering the pH of the kaolin slurry to 3 with sulfuric acid, some of the iron issolubilized. The addition of a strong reducing agent, such as sodium hydrosulfite, re-duces the iron and keeps it in a soluble Fe(II) state, such that it can be removed byrinsing during filtration [6]. Oxidative bleaching is used to render organic color bod-ies colorless since less than 1% organic carbon content can discolor the kaolin to anunacceptable degree [7]. Two oxidative bleaching agents are often used: sodiumhypochlorite, which is added to the slurry in aqueous form, and ozone, which is bub-bled through a gas-liquid contact tower. The organic matter is oxidized to colorlesscarbon dioxide gas. Another way to improve brightness is selective f locculation,which involves the f locculation of discrete TiO2 particles, followed by quiescent set-tling of a portion of the anatase at the bottom of a collection vessel and its subsequentremoval. Kaolin can also be f locculated with a high molecular weight anionic poly-mer, leaving the titaniferrous contaminants in the suspension and thereby discarded.

13.2.3Kaolin Products

Kaolin products are shipped in both wet and dry forms. For the plastics industry, thefiller obviously needs to be in dry form. Filtering takes place on vacuum filters to cre-ate a cake consisting of about 60% solids. The cake is rinsed with water to removesoluble salts and then dried in rotary dryers. It is then pulverized. The final productsobtained in this way are referred to as acid clays since their final pH is 3.5 to 5. Onthe other hand, the filter cake can be redispersed by adding anionic dispersants andadjusting the pH to neutral. At this point, the dispersed slurry may be spray-dried toa moisture level of typically less than 4%, resulting in a pre-dispersed product withhigh bulk density and good bulk f low properties. The product is in its final form atthis stage, although it may be pulverized prior to packaging depending on the end-use application. It can be packaged in different bag sizes or into bulk hopper cars. Forthe production of slurries with high solids content, the dispersed filter product maybe passed through a vacuum evaporator, increasing the solids content from 60% to70%. Another method commonly utilized to achieve 70% solids is to blend the spray-dried product with the filter product in a highly agitated vessel.

13 Kaolin

227

13.2.4Calcination

Calcination is a process utilized to produce value-added kaolin products. Commonly,there are two families of calcined kaolin. One is “metakaolin”, which is produced byheating spray-dried, pulverized kaolin to approximately 550–600 °C. At this temper-ature, the water of crystallization is released. With this 14% loss in weight and theconcomitant changes in the crystalline structure, the mineral becomes highly reac-tive and rich in soluble alumina. In PVC insulation, such as low-voltage wire appli-cations, the incorporation of metakaolin significantly increases volume resistivity asa result of its high dielectric capacity and good thermal insulating properties. The sec-ond family of products produced by calcination is referred to as “fully calcined”.These are formed as a result of increasing the calcination temperature to approxi-mately 1000 °C, at which an exothermic reaction occurs. At this higher temperature,the kaolin is significantly improved in terms of its whiteness and brightness, pro-vides better light scatter, and, hence, can extend TiO2 pigments with concomitant sig-nificant cost savings. Calcined products are pulverized for control of residual free mi-ca or abrasive silica, which would otherwise shorten the life of processing machineparts. Calcination of either type (meta or full) changes the kaolin structure from crys-talline to amorphous. Sintering or fusing of individual clay particles takes place cre-ating a porous material (Figure 13-5), which enhances light scatter. The calcined ma-terials, due to their higher porosity, are capable of considerably higher oil absorptionand their refractive index increases. The mineral becomes somewhat harder, in addi-tion to other key properties that are altered as a result of calcination.


Fig. 13-5 Scanning electron micrographs of calcined kaolin.

228

13.2.5Surface Treatment

Surface treatment is another value-adding step that can improve the performance ofkaolin. Since the filler is naturally very hydrophilic due to its hydroxyl groups, a treat-ment can be applied to render its surface hydrophobic or organophilic. These surface-modified kaolins are especially useful in the plastics and rubber industries, wherethey improve adhesion and dispersion and hence act more effectively as functionalfillers. Silanes, titanates and fatty acids, as discussed in Chapters 4–6, respectively,may be used to modify the surface characteristics of either hydrous or calcinedkaolins, promoting deagglomeration, often lower viscosities, and improved mechan-ical and electrical properties.

13.3Properties

Typical chemical compositions of hydrous and calcined kaolins are shown inTable 13-1 [8]. The major effects of the removal of water of hydration are an increasein refractive index, a moderate increase in the otherwise low Mohs hardness, a de-crease or increase in specific gravity depending on the extent of calcination, and a de-crease in dielectric constant (Table 13-2) [8]. The calcined materials, due to theirgreater void structure, are capable of considerably higher oil absorption, as indicatedin Table 13-3, which also includes comparisons of additional properties of variouskaolin grades. Tables 13-4 and 13-5 [9] summarize some of the features of hydrousand calcined kaolins and the corresponding benefits observed in various plastic andrubber formulations.

Tab. 13-1 Typical chemical compositions (%) of water-washedhydrous and calcined kaolin[a] [8].

Metal Oxide Hydrous grade Calcined grade

Al2O3 38.8 44.7SiO2 45.2 52.5Na2O 0.05–0.3 0.2TiO2 0.6–1.7 0.6–1.8CaO 0.02 0.1Fe2O3 0.3–0.9 0.3–1.0MgO 0.03 0.3K2O 0.05–0.2 0.2Loss on ignition 13.6 – 14.2 < 0.5

[a] Volatile-free basis.

13 Kaolin

229

Tab. 13-2 Typical physical properties of water-washed hydrous andcalcined/silane-treated kaolin grades [8].

Property Hydrous grade Calcined and/or silane-treated grades

Melt temperature (°C) ~1800 ~1800Specific heat capacity (J K–1 g–1) 0.84–0.92 0.84–0.92Free moisture, % < 1.0 < 0.5Mohs hardness 2–2.5 2.5–3Refractive index 1.56 1.62Dielectric constant 2.6 1.3Specific gravity 2.58 2.5–2.63

Tab. 13-3 Miscellaneous properties of specific kaolin grades.

Property Air-f loated Water-washed Water-washed Calcined and hydrous delaminated silane surface

hydrous treated

Residue >44 µm 0.3–1.5 0.01 0.01 0.015

G.E. brightness 70–81 83–92 87–92 90–96

pH 4.5–6.5 3.5–8.0 6.0–8.0 5.0–6.0

Median particle size 0.3–1.3 0.2–4.5 0.4–1.0 0.8–2.0(ESD[a], µm)

Oil absorption rubout 28–36 31–46 38–46 50–95(g/100 g)

Surface area B.E.T. (m2 g–1) 10–22 12–22 11–15 7–12

Specific gravity 2.58 2.58 2.58 2.63

Cost ($/kg) 0.11–0.22 0.26–0.55 0.26–0.48 0.35–1.21

[a] Equivalent spherical diameter.

13.3 Properties

230

Tab. 13-4 Features and benefits of hydrous kaolin [9].

Features Benef its

Inert Low soluble saltsCorrosion and chemical resistance

Soft plate-like particles Non-abrasive; low wear on equipmentBarrier properties

Particle size distribution Wide range – very fine to coarse; narrow distributionsBrightness Good color in non-black compoundsLow residual contaminants Good dispersion and low abrasion

High purityDelamination Higher aspect ratio; improved modulus and barrier properties

Tab. 13-5 Features and benefits of calcined kaolin [9].

Features Benef its

Inert Low soluble saltsElectrical, corrosion, and chemical resistance

Structured and porous particles Abrasion resistanceHigher oil absorptionTiO2 extension

Surface treatment Improved dispersion and adhesionHigher modulusImproved tear and impact strengthHydrophobic

Brightness Good color in non-black compoundsLow residual contaminants Good dispersion and low abrasion

High purityLow free moisture Compatibility with resins; little effects on cure rate

13.4Suppliers

Major suppliers of kaolin with their corresponding capacities in millions of tons perannum (mtpa) are Imerys with >5.0 mtpa, Huber Engineered Materials with ~2.0 mt-pa, Engelhard with ~2.4 mtpa, and CADAM/PPSA (suppliers to the paper market on-ly) with ~1.4 mtpa (2 mtpa by 2006/2007) [10].

Imerys leads the European market for polymer end use. The aforementioned fourcompanies supply the majority of the world’s water-washed kaolin and are major ex-porters of the mineral. Other key producers are Thiele with 1.25 mtpa and Quarzw-erke owned by Amberger Kaolinwerke (AKW) with 0.9 mtpa. Other U.S. producersof air-f loated grades are KT Clay (owned by Imerys) with >500,000 tpa and Uniminwith >500,000 tpa. It should be noted that European kaolin comes mostly from pri-mary deposits and does not lend itself to the air-f lotation process [10].

13 Kaolin

231


Kaolin is readily available in Middle Georgia, U.S.A., where the major U.S. suppliersmine, process and package, and export to Europe, Asia, and throughout the world.Other sources of kaolin are likewise mined and sold outside the U.S.A. The cost ofthe mineral f luctuates, especially with energy prices having a direct impact on theproduction of calcined kaolin. Generally, the 2004 truckload price structure, FOB(free on board), is:

air-f loated grades: 0.07–0.22 $/kgwater-washed hydrous grades: 0.33–0.55 $/kgcalcined grades: 0.51–1.10 $/kgsurface-treated grades: 0.99–1.52 $/kg

U.S. products are available in 50 lb (22.5 kg) bags, semi-bulk bags, and mixed withcertain products in bulk.

In North America, about 11,000 tons of hydrous kaolin were sold for $2.1 millionto the plastics industry in 1999; the forecast for 2004 is a 2.5% growth in tons per an-num to 12,000 tons [11]. Of the calcined kaolin consumed, use in plastics accountedfor 10% of the total volume, behind that used in paper and paint. In North America,about 58,000 tons were sold for $23.8 million in 1998; the forecast for 2004 is a 5%growth in tons per annum to 74,000 tons. For air-f loated kaolin in North America,about 22,000 tons were sold for $1.4 million to the plastics market in 1999, with prac-tically no growth forecasted for 2004 [11].


With the exception of possible dust issues, kaolin is a non-toxic mineral. A dust maskapproved by NIOSH is highly recommended. The Food and Drug Administrationgives kaolin a ‘generally regarded as safe’ status (GRAS). Kaolin is used in many con-sumer products in the cosmetic and pharmaceutical industries. It is employed inhousehold products, such as sunscreen lotions, toothpastes, and even anti-diarrheatreatments. The total crystalline silica content in china clay is less than 0.1%. Kaolinis considered a naturally occurring chemical substance, as per TSCA, 40 CFR710.4(b). In the event of slurry shipment, low levels (ppm) of biocide will have beenadded for long-term protection against microbial degradation.


232

13.7Applications

13.7.1Primary Function

Kaolin, being plate-like filler with a relatively low aspect ratio [6,7,12,13], may impartcertain mechanical property improvements to thermoplastics. Like other plate-likematerials, kaolin can improve dimensional stability (warp resistance), isotropy, andsurface smoothness, a common problem with high aspect ratio fibrous reinforce-ments. Calcined kaolin, particularly after surface treatment with aminosilanes, pro-vides an array of benefits in nylon matrices. Surface-treated calcined kaolin results inmaximum dispersion and has the greatest effect on mechanical properties. Hydrouskaolin, due to its water of hydration, would not be a candidate for use in hydrolytical-ly unstable polymers. The particle size of the mineral, and the method of applicationand concentration of the silane are important. Finer grade kaolin will enhance themechanical properties of a composite more so than coarse grades, assuming com-plete dispersion of the particles in the matrix. Data support the fact that silane pre-treatment delivers superior performance compared to in situ addition of the couplingagent (see also Chapter 4). Figures 13-6 and 13-7 [14] show graphical comparisons ofresults obtained with the two methods in the case of injection-molded nylon-6,6 con-taining 40% calcined kaolin with a mean particle size of 1.4 µm. Both methods, thatis, in situ treatment and pre-treatment, lead to performance improvements; however,the pre-treatment method provides significantly better results. This is especially pro-nounced for unnotched Izod and Gardner impact strengths, which are sensitive tothe extent of dispersion in the polymer matrix. Moreover, it is demonstrated that thelevel of treatment should be optimized based on the cost of the silane and the level ofperformance sought.

Figure 13-8 shows the morphology of an aminosilane-treated 0.8 µm calcinedkaolin in nylon-6,6.

As with nylons, kaolin can also improve the properties of polyolefins. A study [9]comparing silane-treated kaolin with Montana talc and ground calcium carbonate atdifferent loadings in HDPE showed that kaolin gave higher impact and tensilestrengths (Table 13-6). The talc, having the highest aspect ratio of the minerals eval-uated, gave the best modulus improvement.

In another study, a highly impact-modified copolymer PP resin was extruded andthen injection molded. Surface-treated kaolin doubled the f lexural modulus over theunfilled resin at a loading of 30 wt. %, with a small effect on impact strength. Again,the higher tensile and f lexural moduli of the talc composites are, as expected, due tothe higher aspect ratio of the mineral (Table 13-7) [9].

13 Kaolin

23313.7 Applications

Fig. 13-6 Effect of silane addition method on properties of 40 wt. %kaolin-filled nylon-6,6.

234 13 Kaolin

Fig. 13-7 Effect of silane addition level on properties of 40 wt. %kaolin-filled nylon-6,6.

Fig. 13-8 Scanning electron micrograph of a polished cross-sectionof 40 wt. % aminosilane-treated 0.8 µm calcined kaolin in nylon-6,6.


Tab. 13-6 Comparison of mechanical properties of mineral-filled HDPE [9].

Mineral Tensile Tensile Tensile Flexural Gardner Notched strength, elongation, modulus, modulus, impact, J izod impact MPa % MPa MPa strength,

(1% secant) J m–1

Unfilled 21.1 17.0 454.7 813.0 > 35.2 45.6

Kaolin[a]

20% loading 28.3 15.0 668.3 1151 > 35.2 70.540% loading 32.6 10.7 916.4 1778 > 35.2 44.5

Kaolin[b]

20% loading 29.9 14.5 682.1 1240 > 35.2 97.040% loading 33.3 10.3 1033 2101 > 35.2 66.2

Calcium carbonate, 3.0 µm20% loading 24.1 13.6 620.1 1068 31.7 31.3

Montana talc, 1.5 µm20% loading 23.8 11.7 723.4 1433 8.36 23.840% loading 27.6 7.7 1089 2680 1.87 26.0

[a] TRANSLINK® 445 (Engelhard).[b] TRANSLINK® 555 (Engelhard).

Tab. 13-7 Mechanical properties of mineral-filled polypropylene copolymer [9].

Mineral type Mineral Tensile yield Tensile Tensile Flexural Notched loading, wt. % strength, elongation, % modulus, modulus, GPa izod impact

MPa MPa strength, (1% Secant) J m–1

Unfilled 0 20.9 37.2 337.6 730.3 779Treated kaolin 20 20.7 23.2 509.9 1102 699Treated kaolin 30 21.2 17.5 565.0 1378 657Talc 20 21.1 21.8 565.0 1350 355Talc 30 21.4 14.7 682.1 1784 207

236

13.7.2Secondary Functions

13.7.2.1 Improvement of Electrical PropertiesKaolin also serves as a functional filler in wire and cable insulation. As in other ap-plications, surface-treated kaolin offers the best improvement in performance, pro-viding a high degree of hydrophobicity and suppressing polar species at the mineralsurfaces. Moreover, surface-treated grades allow for better compatibility with the poly-mer and minimize the penetration of water through the insulation. Polymers oftenemployed in this market are PVC, LDPE, LLDPE, X-PE, PP, and TPE, among others.

In general, volume resistivity, dielectric strength, and dielectric loss are all im-proved by the addition of kaolin. Finer particle size provides lower dielectric loss,minimizing heat build-up to the point of insulation breakdown. Vinyl functionaltreated calcined kaolin outperforms untreated kaolin in terms of tensile elongationand electrical property retention in cross-linked polyethylene after high temperatureageing in air and water (Table 13-8) [15]. All calcined grades give excellent initial pow-er factor as a result of the low dielectric loss of the amorphous kaolin structure. How-ever, after ageing, the surface-treated grade performed better as a result of the hy-drophobicity imparted by the vinyl functional silane (see also Chapter 4).

Table 13-8 Evaluation of kaolin in cross-linked polyethylene typicallyused in medium voltage cable [15].

Type Loading (phr) Tensile Elongation % Retention Power factor % strength, MPa at break, % of elongation (after 168 h in

(after 168 h @ water at 75 °C)120 °C in air)

Unfilled 0 14.5 307 103 0.05

Satintone 5® 50 16.4 131 45 1.94calcined kaolin (0.8 µm)

Translink 77® 50 18.8 123 94 0.12vinylsilane-treated (0.8 µm)

Note: Satintone 5® and Translink 77® are Engelhard Corp. kaolin grades.

Metakaolin is an effective functional filler in plasticized PVC wire insulation atloadings of about 10 phr. It better protects the insulation from cracking or “treeing”,a term that describes the physical breakdown of the cable polymer matrix due tomoisture or other external inf luences. Volume resistivity is greatly improved throughthe incorporation of metakaolin as compared to the incorporation of other minerals(see Table 13-9) [15], although the color is not as white as with higher temperature ful-ly calcined kaolin. In addition, metakaolin shows high acid solubility and thereforeacts as an acid scavenger to assist in stabilizing the PVC against degradation. The

13 Kaolin

237

kaolin types used in wire and cable generally have low soluble salt and free moisturecontents.

Tab. 13-9 Effect of different fillers at 10 phr on volume resistivity oflow voltage PVC wire [15].

Filler Volume Resistivity, Ohm cm

Ground calcium carbonate 7 × 1012

Hydrous kaolin 30 × 1012

Hydrous kaolin, silane-treated 50 × 1012

Calcined kaolin 50 × 1012

Calcined kaolin, silane-treated 100 × 1012

Metakaolin 300 × 1012

13.7.2.2 TiO2 ExtensionTiO2 is an excellent pigment for color and opacity in plastics. TiO2 extension is com-monly achieved with kaolin, thereby lowering cost yet still maintaining properties. Acase study was conducted with a typical 50 wt. % TiO2 LLDPE composite. In injection-molded parts, replacement of 10% of the TiO2 with calcined kaolin allowed retentionof both color and opacity. At 20% replacement of TiO2, only a slight drop in opacityoccurred and color intensity was completely maintained (Table 13-10) [16].

Hydrous kaolin may also be used to replace or extend TiO2, but not always with thesame efficiency as calcined grades. In blown LLDPE film (Table 13-11) [17], hydrouskaolin can replace up to 10% TiO2 with no loss in color or opacity. At 20% TiO2 re-

13.7 Applications

Tab. 13-10 Color of injection-molded LLDPE containing TiO2/kaolincombinations [16].

Solids % Color “L” Color “a” Color “b” Opacityvalue value value

TiO2 % replacement

Control – 0.95 96.72 –0.51 1.64 83.3

Ultrex 96 TM 5 0.97 96.77 –0.56 1.67 82.110 0.98 96.38 –0.60 1.58 82.015 0.99 96.51 –0.58 1.70 80.720 1.03 96.35 –0.64 1.45 80.1

ASP-170® 5 0.95 96.38 –0.58 1.62 79.410 1.10 96.13 –0.46 1.95 79.115 0.94 95.79 –0.48 1.80 79.220 0.98 95.54 –0.43 2.24 78.1

Materials: LLDPE Dowlex 2553 (Dow), TiO2 Tri-Pure® R101 (DuPont), Ultrex 96 TM engineeredcalcined kaolin (Engelhard Corp.), ASP-170® spe-cialty hydrous kaolin (Engelhard Corp.).Conditions: Letdown ratio, 50:1; Battenfeld press,

85 ton; barrel temperature, 230–250 °C. Testing:Color data generated from 7.5 cm diameter injec-tion-molded discs and opacity measured on nom-inal 0.575 mm thick press-outs from the let-downs.

238

placement, just a slight loss in opacity may occur. Another application where kaolinhas delivered cost savings is in white PVC siding. Here again, TiO2 extension can berealized up to a level of 20% [17].

Tab. 13-11 Color of blown LLDPE films containing TiO2/kaolin com-binations [17].

TiO2 % Color Color Color Opacityreplacement “L” value “a” value “b” value

Letdown 25:1Control – 92.6 –1.0 0.1 35.7

ASP-170® 5 92.4 –1.0 0.3 38.610 92.4 –1.0 0.3 38.520 91.9 –1.1 0.3 34.3

Letdown 100:1Control – 91.3 –1.1 –0.1 18.5

ASP-170® 5 91.3 –1.2 –0.1 19.810 91.1 –1.2 –0.1 19.320 91.1 –1.1 –0.1 17.9

An additional evaluation [18] of the extension of rutile TiO2 by different fillers wasconducted by assessing blister resistance in two types of liquid, thermosetting poly-ester resins [(namely, ortho- and ortho-neopentyl glycol (NPG)]. The performance ofthe filler/pigment portion of the cured polyester gel coat was compared after expo-sure to water at 65 °C for 1150 hours. Untreated kaolin provided better blister resist-ance than untreated talc or calcium carbonate. Vinylsilane surface-treated kaolin gavethe best blister resistance, reducing the size and extent of blisters as a result of im-proved compatibility and the formation of a hydrolytically stable, strong mineral/resin interface.

Tab. 13-12 Comparison of filler performance in polyester gel coat after ageing [18].

Sample TiO2 % Mineral % Blister rating Blister rating ortho ortho-NPG

Rutile TiO2 control 25 0 2.5 1.1Untreated calcined kaolin 10 15 1.0 0.7Vinylsilane-treated kaolin 10 15 0.8 0.0Aminosilane-treated kaolin 10 15 0.9 0.5Untreated talc 10 15 3.0 2.5Wet-ground untreated 10 15 2.8 2.2calcium carbonate

Blister rating: 0 = none; 1 = slight; 2 = moderate; 3 = severe.

Materials: LLDPE Dowlex 2553 (Dow); TiO2 Tri-Pure® R101 (DuPont); Kaolin ASP-170® specialtyhydrous kaolin (Engelhard Corp.)Conditions: Blow-up ratio: 3:1, die diameter:25 cm spiral, barrel temperature: 190–245 °C.

Two letdown ratios during production of blownfilm, 25:1 and 100:1.Testing: Opacity was measured on nominal0.025 mm thick films.

13 Kaolin

239

References

1 Ciullo, P. A., Industrial Minerals and TheirUses – A Handbook and Formulary, NoyesPublications, Westwood, NJ, 1996.

2 Washabaugh, F., “Kaolins for Rubber Appli-cations”, Engelhard Corp., Iselin, NJ, 1995.

3 Murray, H., Applied Clay Science 1991, 5(3),379–385.

4 Popson, S., “Measurement and Control ofthe Optical Properties of Paper”, TechnicalPaper, Technidyne Corp., New Albany, IN,1989.

5 Finch, E., Industrial Minerals, March 2002,414, 64–66.

6 Prasad, M., Reid, K., Technical Report“Kaolin: Processing, Properties and Appli-cations”, University of Minnesota, MN,1990.

7 Kogel, J., et al., “The Georgia Kaolin – Geol-ogy and Utilization”, SME: Society for Min-ing, Metallurgy, and Exploration, Inc., AnnArbor, MI, 2002, p. 41; www.smenet.org

8 Engelhard Corporation, “Performance Min-eral Reinforcements for Plastics and Rub-ber”, Technical Brochure, Engelhard Corp.,Iselin, NJ, 2002; www.engelhard.com

9 Fajardo, W., “Kaolin in Plastic and RubberCompounds”, Technical Report, EngelhardCorp., Iselin, NJ, 1993.

10 Moore, P., Industrial Minerals, August 2003,431, 24–35.

11 Kline & Company, Inc., “Extender andFiller Minerals North America, Kaolin Re-port”, Kline & Company, Inc., Little Falls,NJ, 2000–2002.

12 Bundy, W., Kaolin Genesis and Utilization,The Clay Minerals Society, Boulder, CO,1993, p. 56–57.

13 Pickering, Jr., S., Murray, H., “Kaolin” inIndustrial Mineral and Rock, 6th Ed., SME:Society for Mining, Metallurgy, and Explo-ration, Inc., Ann Arbor, MI, 1994;www.smenet.org

14 Carr, J., “Kaolin Reinforcements: An AddedDimension”, Technical Report, EngelhardCorp., Iselin, NJ, 1990.

15 Khokhani, A., “Kaolins in Wire & Cable”,Technical Report, Engelhard Corp., Iselin,NJ, 1994.

16 Fajardo, W., “ASP®170 & Ultrex™ 96 Spe-cialty Pigments”, Technical Bulletin,TI2303, Engelhard Corp., Iselin, NJ, 1997.

17 Sherman, L., “Stretch TiO2”, Plastics Tech-nology Online, July 1999; http://www.plastic-stechnology.com/articles/199907fa1.html

18 Washabaugh, F., “The Effect of Gel Coat Ex-tenders on the Performance of PolyesterLaminates”, Technical Bulletin, TI2180, En-gelhard Corp., Iselin, NJ, 1990.

13.7 Applications

14Wollastonite

Marino Xanthos

14.1Introduction

Wollastonite is a naturally occurring acicular (needle-like) silicate mineral, named in1822 after the English chemist and mineralogist W. H. Wollaston. Chemically, it is acalcium metasilicate with the formula CaSiO3. When pure, wollastonite is white; im-purities from substituted elements or accompanying minerals may change the colorto cream, pink or even red. Wollastonite forms through metamorphism (heat andpressure) of limestones. In one method of formation, silica and calcite in silica-bear-ing limestones react to form wollastonite and carbon dioxide. Wollastonite can alsoform by the passage of siliceous hydrothermal solutions through limestone zones.During the formation of the wollastonite structure, a series of accessory mineralsmay be introduced; these need to be removed during beneficiation by wet processingand/or high intensity magnetic separation to produce commercial, high puritygrades. Wollastonite became widely known as an important industrial mineral in thelate 1950s, rapidly attaining applications in plastics, paints, friction products, ceram-ics, and metallurgy; very recently, it has also been used as an active component in bio-medical composites (see Chapter 22).

14.2Production

The ore is first crushed and milled sufficiently to segregate the accompanying min-eral impurities. Weakly magnetic garnet and diopside can be removed by magneticseparators, whereas calcite may be removed by f lotation. The extent to which the aci-cular shape (aspect ratio) of the mineral is preserved during milling of the benefici-ated products dictates its uses. Powder grades having a low aspect ratio (3:1 to 5:1) areproduced in impact mills to 325 mesh and 400 mesh, and 8 µm up to 25 µm top sizefor micronized grades. Higher aspect ratio (10:1 to 20:1) grades, particularly useful asfunctional fillers for plastics, may range from about 200 µm to 20 µm in average nee-


242

dle lengths. They are carefully attrition milled with removal of fines by air separation.Grades surface-treated with coupling agents are also produced through a series ofblending/coating and drying steps. Some further information on the production ofcommercial grades is provided by the two major U.S. producers in their publications[1,2] and also in publications available from the U.S. Geological Survey website [3].


The theoretical composition of calcium silicate is 48.3% CaO and 51.7% SiO2. Basedon unit cell parameters, the specific gravity is calculated to be 2.96. In commercialgrades, deviations from the theoretical specific gravity are due to various impurityions that substitute for calcium in the crystal lattice or impurity minerals such as cal-cite, garnet, and diopside. Typical chemical compositions of wollastonite grades forplastics obtained from various suppliers are listed in Table 14-1.

Tab. 14-1 Typical chemical analyses of wollastonite filler grades.

Composition Wt. %[a] Wt. %[b] Wt. %[c]

CaO 47 44.0 47SiO2 49.5 50.0 50.0MgO 0.2 1.5 0.3Al2O3 0.6 1.8 0.3Fe2O3 0.43 0.3 1.0TiO2 traces not reported 0.05MnO 0.29 <0.1 0.1Na2O 0.02 0.2 not reportedK2O 0.11 not reported 0.1

The structure of wollastonite is characterized by chains made up of silica tetrahe-dra connected side-by-side through octahedrally coordinated calcium. This structureleads to the growth of acicular crystals, which preserve their shape upon cleavage(Figure 14-1). The principal characteristics of grades suitable for plastics applicationsare: white color, low water absorption, good thermal stability, low thermal expansioncoefficient, and low loss on ignition; other characteristics are relatively high Mohshardness (about 4.5), high alkalinity, and dissociation by mineral acids. Reportedmodulus values are about 30,000 MPa, similar to those of other silicates [4]. The phys-ical and chemical properties of wollastonite are summarized in Table 14-2 (from refs.[4,5] and suppliers’ literature).

The incorporation of wollastonite into thermoplastics results in enhanced dimen-sional stability, an increase in f lexural modulus, and a higher heat def lection tem-perature (HDT). It is used as an economic alternative/supplement to glass fibers. It

[a] Wolkem India Ltd. (KEMOLIT® grade).[b] R. T. Vanderbilt Co., Inc. (VANSIL® grade).

[c] NYCO Minerals, Inc. (NYAD® grade).

14 Wollastonite

243

is particularly amenable to surface treatment with silanes, titanates, polymeric esters,and other additives to promote lower compound viscosity, easier dispersion, and im-proved physical properties. Abrasive wear related to the relatively high hardness ofthe mineral (Mohs 4.5 vs. about 1 for talc) is a concern for process equipment. On theother hand, the high hardness of the mineral leads to superior scratch and mar re-sistance in certain formulations.


Fig. 14-1 Microphotograph of acicular wollastonite (VANSIL® WG,T. Vanderbilt Co., Inc.), 170×.

Tab. 14-2 Typical properties of wollastonite.

Property

Color whiteCrystal system triclinicSpecific gravity 2.8–2.9Coefficient of thermal expansion, K–1 6.5 × 10– 6

Specific heat, J kg–1 K–1 1003Melting point, °C 1540Transition temperature, °C (to pseudowollastonite) 1200Hardness (Mohs) 4.5–5Refractive index 1.63–1.67pH (10 wt. % slurry) 9.0–11Loss on ignition, % (950 °C) 0.1–6Thermal conductivity, W m–1 K–1 2.5Dielectric constant, 104 Hz 6

244

14.4Suppliers/Cost

India, China, USA, Mexico, and Finland are the major wollastonite producing coun-tries with worldwide production in 2002 estimated to be between 550,000 tons and600,000 tons [6]. An estimated 20% of the world production is used in plastics appli-cations [4], with surface-treated grades being available from most suppliers.

Wolkem India Ltd. claims to be the largest world producer, with a production of160,000 tons in 2002–03 [7]. U.S. producers include NYCO Minerals Inc. and R. T.Vanderbilt Co. Inc., both mining wollastonite from deposits in New York State withbeneficiation plants located near the mines. Estimated U.S. production in 2002 wasabout 120,000 tons, with plastics accounting for about 37% of U.S. sales [8]. In 2002,production in Finland and Mexico was estimated to be about 20,000 tons and40,000 tons, respectively, while production in China was estimated to be about300,000 tons [6]. Quarzwerke GmbH, Germany, is a supplier of a variety of silane-treated grades.

Prices for U.S. produced acicular wollastonite (200 mesh to 400 mesh), ex-works,ranged from US$190 to US$258 per ton, whereas the price for high aspect ratio(15:1–20:1) was about $320 per ton. Chinese wollastonite FOB prices were quoted asUS$80 to US$110 depending on mesh size [9]. Prices for specialty surface-treatedgrades can be as high as $1,700 per ton [6].


Current recommended exposure limits (REL) set by the U.S. National Institute of Oc-cupational Safety and Health (NIOSH) and permissible exposure limits (PEL) set bythe Occupational Safety and Health Administration (OSHA) are a TWA (time-weighted average) of 10 mg m–3 (total particulates) and 5 mg m–3 (respirable particu-lates). Such values, which are similar to those for other particulates such as kaolin,require appropriate precautions for exposure to dust. Because it is an acicular min-eral, wollastonite has come under the close scrutiny of various federal agencies as apossible health hazard. With respect to carcinogenic effects, in a NIOSH medical sur-vey “no definite association of wollastonite exposure and excess morbidity could bedemonstrated”. From an update of this study, it was concluded that prolonged expo-sure to excessive wollastonite dust may affect pulmonary functions, although exces-sive exposure to any dust may aggravate pre-existing respiratory conditions [10]. Ini-tial exposure to the high aspect ratio grades may produce minor skin irritation. In areview of in vitro, in vivo, and epidemiological studies, it was concluded that there isno evidence to suggest that wollastonite presents a health hazard [11].

14 Wollastonite

245


Wollastonite imparts in thermoplastic matrices the usual attributes of stiff, reinforc-ing mineral fillers, i.e. higher modulus and higher heat def lection temperature, butthe magnitudes of the enhancements are smaller than with higher aspect ratio fillerssuch as glass fibers or mica. Increase in modulus is usually accompanied by lowerelongation at break. Impact strength is usually lower, although fine particle sizegrades and partial substitution by glass fibers may enhance the notched impactstrength of certain resins. Above a certain filler concentration, f lexural and tensilestrength values are usually increased compared to the unfilled matrix, the effect de-pending on the polymer type, the aspect ratio of the filler, and the type of surfacetreatment. Parts made with wollastonite have low coefficients of thermal expansion,low mold shrinkage, and good dimensional stability. In addition to its primary func-tion as a mechanical property modifier, wollastonite also imparts good scratch andmar resistance. Fine surface-treated grades impart a good surface finish. Issues thatneed to be addressed with wollastonite are warpage with thermoplastic polyesters, theabrasive nature of certain grades promoting equipment wear, and low weld linestrength.

Common thermoplastics reinforced with wollastonite and their applications [7] are:

polyamide-6 and -6,6 (automotive parts, electric motors, gears, power tool hous-ings);

polypropylene, PP (furniture, battery cases, fan blades, automotive under thehood);

polybutylene terephthalate, PBT (electrical and electronic components).

Other thermoplastics with which wollastonite shows reinforcing effects are poly-carbonate, polystyrene, and vinyl plastisols. In polyurethanes and LDPE, it also im-proves electrical insulation properties.

In polyamides, high aspect ratio grades usually modified with aminosilanes areused at concentrations up to 40 wt. %. For compositions in which wollastonite is usedas partial replacement for glass fibers, the wollastonite/glass fibers combinationusually results in higher tensile strength and modulus, as well as in higher notchedimpact strength and HDT than the all-wollastonite composition at the same loading.Representative data are shown in Table 14-3 [1]. The importance of aspect ratio andtype of surface treatment has recently been confirmed in a study of a series ofpolyamide-6,6 composites containing 40 wt. % fillers from different suppliers [12].

The growth in the use of wollastonite in PP compounds has been accompanied bythe development of appropriate coupling agents for the incompatible non-polar poly-mer/polar filler interface. A particularly effective coupling agent was found to be asulfonyl azide silane produced by Hercules, Inc. in the late 1970s/early 1980s, butwhich is now discontinued. Filler treatment with 1% of this particular silane wasshown to increase the room temperature tensile strength of a 40 wt. % wollas-tonite/PP compound by 48%, its 100 °C tensile modulus by 44%, and its HDT by 12%


246

[13]. A variety of other silanes, titanates, and maleated polypropylenes have been usedas coupling agents, the latter in combination with aminosilanes [14,15]. For automo-tive applications, wollastonite competes with glass, mica, and talc. In a comparativestudy on surface-modified wollastonite, mica, and milled glass fiber injection-mold-ed PP compounds containing 20–25 wt. % fillers, glass and mica gave the highestmodulus, strength, and HDT values, while wollastonite provided the highest un-notched impact strength [1]. In pigmented mineral-filled PP compounds marketed aspotential alternatives to ABS and PC/ABS blends for automotive interior compo-nents, wollastonite was shown to increase scratch hardness and to reduce scratchdepth and scratch visibility [16]. In contrast, talc was found to have poorerscratch/mar resistance. In injection-molded parts, these effects were shown to de-pend on the injection-molding conditions, particle size, and filler coating [17].

The use of amino-functionalized silanes has been shown to be particularly effectivein polycarbonate composites containing 50 wt. % wollastonite. In the presence of thetreated filler, f lexural and tensile strength values exceeded those of the unfilled resinunder both dry and wet conditions. As expected, modulus values were less affectedby the surface treatment [18].

Common wollastonite-filled thermosets and their applications are:

Polyurethane (PUR) automotive parts produced by RRIM (reinforced-resin injec-tion molding), in which wollastonite is used as a replacement for milled glassfibers.

Phenolic resins, improving machinability and electrical properties with applica-tions in appliances, brakes, furniture insulators, foundry moldings.

Unsaturated polyesters in automotive SMC, BMC, replacing glass fibers by up to30%.

Epoxy-based molding and casting compositions (with loadings up to 50% for re-duced shrinkage and improved dimensional stability and electrical properties), andanti-corrosive metal primers as an extender.

14 Wollastonite

Tab. 14-3 Properties of wollastonite and wollastonite/glass fiber re-inforced polyamide-6,6.

Property Unf illed 40 wt. % 25 wt. % Polyamide-6,6 Wollastonite[a] Wollastonite[a]/

15 wt. % Glass Fibers

Tensile yield strength, MPa 84 99.4 126Elongation at break, % 60 3 3Flexural modulus, GPa 2.9 7.35 9.8Izod impact strength notched, J m–1 53.4 32.04 53.4HDT, °C, at 185 kPa 90.5 229 254Specific gravity 1.14 1.50 1.50

[a] G Wollastocup® 1100 (NYCO Minerals Inc.)

247

References

1 Copeland, J. R., Chapter 8 in Handbook ofReinforcement for Plastics (Eds.: Milewski, J.V., Katz, H. S.), Van Nostrand Reinhold,New York, 1987.

2 Ciullo, P. A., Robinson, S., “Wollastonite –A Versatile Functional Filler”, Paints andCoatings Industry Magazine, Nov. 2002; ac-cessed March 21, 2004, via http://www. rt-vanderbilt.com.

3 http://www.Minerals.USGS.gov.4 Hohenberger, W., Chapter 17 in Plastics Ad-

ditives Handbook (Ed.: Zweifel, H.), HanserPublishers, Munich, 2001.

5 Wypych, G., Handbook of Fillers, pp.167–169, ChemTec Publ., Toronto, Ont.,Canada, 2000.

6 Virta, R. L., “Wollastonite” in U.S. Geologi-cal Survey Minerals Yearbook 2002, Vol. IMetals and Minerals; accessed March 21,2004, via http://www.Minerals.USGS.gov.

7 Mahajan, S., Proc. Functional Fillers for Plas-tics 2003, Intertech Corp., Atlanta, GA, Oct.2003.

8 Hawley, G. C., Mining Eng. 2002, 55(6), 55.

9 Anonymous, “Prices”, Industrial Minerals2004, 438, p. 81.

10 Material Safety Data Sheet (draft), SectionXI, Product Vansil® EW-10, R. T. VanderbiltCo., Inc.; accessed on-line March 25, 2004,via http://www.rtvanderbilt.com.

11 Emerson, R. J., “Understanding the HealthEffects of Wollastonite”, accessed March 21,2004, via www.nycominerals.com.

12 Hawley, G. C., Jaworski, B., Proc. 56th SPEANTEC, 1998, 44, 2847.

13 Copeland R. J., Rush, O. W., ModernPlastics, March 1979, p. 68

14 Roberts, D. H., Proc. 56th SPE ANTEC,1998, 44, 1427.

15 Anonymous, “Wollastonite in Polypropy-lene”; Accessed March 24, 2004, viahttp://www. rtvanderbilt.com.

16 Chu, J., et al., Polym. Eng. Sci. 2000, 40(4),944.

17 Wong, T. L., et al., Proc. 57th SPE ANTEC,1999, 45, 3162.

18 Anonymous, Plastics Additives & Compound-ing 2003, 5(3), 40–45.

References

15Wood Flour

Craig M. Clemons and Daniel F. Caulf ield

15.1Introduction

The term “wood f lour” is somewhat ambiguous. Reineke [1] states that the termwood f lour “is applied somewhat loosely to wood reduced to finely divided particlesapproximating those of cereal f lours in size, appearance, and texture”. Though itsdefinition is imprecise, the term wood f lour is in common use. Practically speaking,wood f lour usually refers to wood particles that are small enough to pass through ascreen with 850-micron openings (20 US standard mesh).

Wood f lour has been produced commercially since 1906 [2] and has been used inmany and varied products including soil amendments, extenders for glues, and ab-sorbents for explosives. One of its earliest uses in plastics was in a phenol–formalde-hyde and wood f lour composite called Bakelite. Its first commercial product was re-portedly a gearshift knob for Rolls Royce in 1916 [3]. Though once quite prevalent asa filler for thermosets, its use has diminished.

In contrast to its use in thermosets, large-scale use of wood f lour in thermoplasticshas only occurred within the last few decades. Recent growth has been great; use ofwood–plastic composites has grown from less than 50,000 tonnes in 1995 to nearly600,000 tonnes in 2002 (Figure 15-1) [4]. Most of this has been due to the rapid growthin the manufacture of exterior building products such as railings, window and doorprofiles, and especially decking (Figure 15-2).

Due to its low thermal stability, wood f lour is usually used as a filler only in plas-tics that are processed at temperatures lower than about 200 °C. The great majorityof wood–plastic composites use polyethylene as the matrix (Figure 15-3). This is due,in part, to the fact that many of the early wood–plastic composites were developed asan outlet for recycled film. Some manufacturers also use combinations of thermo-plastics and thermosets as the matrix material.


250 15 Wood Flour

Fig. 15-1 Market demand for wood and natural fiber plastic com-posites in North America and Europe. (Reprinted with permissionfrom ref. [4]).

Fig. 15-2 Current applications and market size of plastics withwood flour or natural fibers. (Reprinted with permission from ref. [4]).

251


Wood f lour is derived from various scrap wood from wood processors. High qualitywood f lour must be of a specific species or species group and must be free from bark,dirt, and other foreign matter. Many different species of tree are offered as wood f lourand are often based on the regional availability of clean raw materials from wood-pro-cessing industries. The most commonly used wood f lours for plastic composites inthe United States are made from pine, oak, and maple. Many reasons are given forspecies selection, including slight color differences, regional availability, and famil-iarity. Some species, such as red oak, can contain low MW phenolic compounds,which may cause stains if the composite is repeatedly wetted [5].

Though there is no standard method of producing wood f lour, some generalitiescan be discussed. The main steps in wood f lour production are size reduction andsize classification. If larger raw materials are used, their initial size may be reducedusing equipment such as a hammer mill, hog, or chipper [2]. Once coarsely ground,the wood is pulverized by grinding between disks as in attrition mills, beating withimpactors or hammers as in hammer mills, or crushing between rollers as in rollermills [2]. Other mills can also be used but are less common.

Pulverizing results in a mixture of particles that contains fiber bundles and fiberfragments. These particles typically have aspect ratios (i.e., length-to-diameter ratios)of only 1–5 (Figure 15-4). These low aspect ratios allow wood f lour to be more easilymetered and fed than individual wood fibers, which tend to bridge. However, the lowaspect ratio limits their reinforcing ability [6].

Once pulverized, the wood can be classified using vibrating, rotating, or oscillatingscreens. Air classifying is also used, especially with very finely ground wood f lours[1]. Wood f lour particle size is often described by the mesh of the wire cloth sievesused to classify the particles. Table 15-1 lists the US standard mesh sizes and theirequivalent particle diameters. However, different standards may be used interna-


Fig. 15-3 Plastics used in wood–plastic composites. (Reprintedwith permission from ref. [4].)

252

tionally [7]. Most commercially manufactured wood f lours used as fillers in thermo-plastics are less than 425 µm (40 US standard mesh). Very fine wood f lours can costmore and increase melt viscosity more than coarser wood f lours, but compositesmade with them typically have more uniform appearance and a smoother finish. Ifground too finely, fiber bundles become wood dust, fragments that no longer re-semble fibers or fiber bundles.

Tab. 15-1 Conversion between US standard mesh and particlediameter.

Mesh U.S. [41] Particle Diameter (µm)

20 85025 71030 60035 50040 42545 35550 30060 25070 21280 180

100 150120 125140 106170 90200 75230 63270 53325 45400 38

15 Wood Flour

Fig. 15-4 Scanning electron micrograph of pine wood f lour.

253

Wood f lour is commonly packaged in: 1) multi-walled paper bags (approximately23 kg or 50 lbs), 2) bulk bags (typically 1.5 cubic meter or 55 cubic feet), or 3) bulktrailers [8]. Wood f lour is typically supplied to the customer with moisture contentsin the range 4–8% and must be dried before use in thermoplastics. Some wood f lourmanufacturers offer standard grades, while others prefer to customize for individualbuyers and applications. Specifications depend on the application, but include sizedistribution, moisture content, species, color, and cost.


15.3.1Wood Anatomy

As with most natural materials, the anatomy of wood is complex. Wood is porous, fi-brous, and anisotropic. Wood is often subdivided into two broad classes, namely soft-woods and hardwoods, which are classified by botanical and anatomical featuresrather than actual wood hardness. Figures 15-5 and 15-6 are schematics of a softwoodand a hardwood, respectively, showing the typical anatomies of each wood type. Soft-woods (or Gymnosperms) include pines, firs, cedars, and spruces among others; hard-woods (or Angiosperms) include species such as the oaks, maples, and ashes.

Wood is primarily composed of hollow, elongated, spindle-shaped cells (called tra-cheids or fibers) that are arranged parallel to each other along the trunk of the tree[9]. The lumen (hollow center of the fibers) can be completely or partially filled withdeposits, such as resins or gums, or growths from neighboring cells called tyloses [9].These fibers are firmly cemented together and form the structural component of


Fig. 15-5 Schematic of a softwood.

254

wood tissue. The length of wood fibers is highly variable, but averages about 1 mm(1/25 in.) for hardwoods and 3 to 8 mm (1/8 to 1/3 in.) for softwoods [9]. Fiber diam-eters are typically 15–45 µm. When wood is reduced to wood f lour, the resulting par-ticles are actually bundles of wood fibers rather than individual fibers and can con-tain lesser amounts of other features such as ray cells and vessel elements. Furtherinformation on wood anatomy can be found in refs. [10,11].

15.3.2Chemical Components

Wood itself is a complex, three-dimensional, polymer composite made up primarilyof cellulose, hemicellulose, and lignin [12]. These three hydroxyl-containing polymersare distributed throughout the cell wall. The chemical compositions of selectedwoods are shown in Table 15-2.

Of the three major components, cellulose shows the least variation in chemicalstructure. It is a highly crystalline, linear polymer of anhydroglucose units with a de-gree of polymerization (n) of around 10,000 (Figure 15-7). It is the main componentproviding the wood’s strength and structural stability. Cellulose is typically 60–90%crystalline by weight, and its crystal structure is a mixture of monoclinic and triclin-ic unit cells [13,14]. Hemicelluloses are branched polymers composed of various five-and six-carbon sugars, the molecular weights of which are well below those of cellu-lose but which still contribute as a structural component of wood [15].

Lignin is an amorphous, cross-linked polymer network consisting of an irregulararray of variously bonded hydroxy- and methoxy-substituted phenylpropane units[15]. The chemical structure varies depending on its source. Figure 15-8 depicts partof a softwood lignin structure, illustrating a variety of possible structural compo-

15 Wood Flour

Fig. 15-6 Schematic of a hardwood.

255

nents. Lignin is less polar than cellulose and acts as a chemical adhesive within andbetween the cellulose fibers.

Additional organic components, called extractives, make up about 3–10% of the drywood grown in temperate climates, but significantly higher quantities are found inwood grown in tropical climates [15]. Extractives include substances such as fats, wax-es, resins, proteins, gums, terpenes, and simple sugars, among others. Many of theseextractives function in tree metabolism and act as energy reserves or defend againstmicrobial attack [15]. Though often small in quantity, extractives can have large in-f luences on properties such as color, odor, and decay resistance [15]. Small quantities(typically 1%) of inorganic matter, termed ash, are also present in wood grown in tem-perate regions.

Cellulose forms crystalline microfibrils held together by hydrogen bonds, whichare in turn cemented to lignin in the wood fiber cell wall. The microfibrils are alignedin the fiber direction in most of the cell wall, winding in a helix along the fiber axis.The angle between the microfibril and fiber axes is called the microfibril helix angle.The microfibril helix angle is typically 5–20° for most of the cell wall [16] and variesdepending upon many factors including species and stresses on the wood duringgrowth.


Tab. 15-2 Approximate chemical compositions (%) of selectedwoods [15].

Species Cellulose[a] Hemicellulose[b] Lignin[c] Extractives[d] Ash

Ponderosa Pine 41 27 26 5 0.5Loblolly Pine 45 23 27 4 0.2Incense Cedar 37 19 34 3 0.3Red Maple 47 30 21 2 0.4White Oak 47 20 27 3 0.4Southern Red Oak 42 27 25 4 0.4

[a] Alpha cellulose content as determined byASTM D 1103 [41].

[b] Approximate hemicellulose content deter-mined by subtracting the alpha cellulose con-tent from the holocellulose content valuesfrom ref. [15].

[c] Klason lignin content as determined by ASTMD 1106 [42].

[d] Solubility in 1:2 (v/v) ethanol/benzene accord-ing to ASTM D 1107 [43].

Fig. 15-7 Chemical structure of cellulose [15].

256

15.3.3Density

The bulk density of wood f lour depends on factors such as moisture content, parti-cle size, and species, but typically is about 190–220 kg m–3 (12–14 lbs ft–3) [8]. Becauseof its low bulk density, special equipment, such as a crammer, is sometimes used toaid feeding of wood f lour.

As a filler, wood f lour is unusual in that it is compressible. Though the density ofthe wood cell wall is about 1.44 to 1.50 g cm–3 [17], the porous anatomy of solid woodresults in overall densities of about 0.32 to 0.72 g cm–3 (20 to 45 lb ft–3) when dry [18].However, the high pressures encountered during the processing of plastics can causethe hollow fibers of the wood f lour to collapse or fill them with low molecular weightadditives or polymers. The degree of collapse or filling will depend on such variablesas particle size, the processing method, and additive viscosity, but wood densities in

15 Wood Flour

Fig. 15-8 A partial softwood lignin structure [15].

257

composites approaching the wood cell wall density can be attained in high-pressureprocesses such as injection molding. Consequently, adding wood fibers to commod-ity plastics such as polypropylene, polyethylene, and polystyrene increases their den-sities.

Even these higher densities are considerably lower than those of inorganic fillersand reinforcements. This density advantage is important in applications whereweight is important, such as in automotive components. Recently, chemical foamingagents and microcellular foaming technology have been investigated with a view toreducing the density of wood–plastic composites [19–21].

15.3.4Moisture

The major chemical constituents of the cell wall bear hydroxyl and other oxygen-con-taining groups that attract moisture through hydrogen bonding [26]. This hygroscop-icity can cause problems both in composite fabrication and in the performance of theend product.

Moisture sorption in wood is complex and the final equilibrium moisture contentis affected by temperature and humidity. The equilibrium moisture content can alsovary by up to 3–4% (although usually less) depending on whether it is approachedfrom a higher or lower humidity (i.e., wood exhibits a moisture sorption hysteresis).Table 15-3 shows approximate equilibrium moisture contents for wood at differenttemperatures and humidities at a midpoint between the hysteresis curves.

Tab. 15-3 Equilibrium moisture content for wood at differenttemperatures and humidities [18].

Temperature Moisture content (%) at various relative humidities(°C) (°F) 10% 20% 30% 40% 50% 60% 70% 80% 90%

–1.1 30 2.6 4.6 6.3 7.9 9.5 11.3 13.5 16.5 21.04.4 40 2.6 4.6 6.3 7.9 9.5 11.3 13.5 16.5 21.010 50 2.6 4.6 6.3 7.9 9.5 11.2 13.4 16.4 20.915.6 60 2.5 4.6 6.2 7.8 9.4 11.1 13.3 16.2 20.721.1 70 2.5 4.5 6.2 7.7 9.2 11.0 13.1 16.0 20.526.7 80 2.4 4.4 6.1 7.6 9.1 10.8 12.9 15.7 20.232.2 90 2.3 4.3 5.9 7.4 8.9 10.5 12.6 15.4 19.837.8 100 2.3 4.2 5.8 7.2 8.7 10.3 12.3 15.1 19.5

Wood f lour usually contains at least 4% moisture when delivered, which must beremoved before or during processing with thermoplastics. Though moisture couldpotentially be used as a foaming agent to reduce density, this approach is difficult tocontrol and is not common industrial practice. Commercially, moisture is removedfrom the wood f lour: 1) before processing using a dryer, 2) by using the first part ofan extruder as a dryer in some in-line process, or 3) during a separate compoundingstep (or in the first extruder in a tandem process).


258

Once dried, wood f lour can still absorb moisture quickly. Depending on the ambi-ent conditions, wood f lour can absorb several weight percent of moisture withinhours (Figure 15-9). Even compounded material often needs to be dried prior to fur-ther processing, especially if high weight percentages of wood f lour are used. Fig-ure 15-10 shows moisture sorption curves for compounded pellets of polypropylenecontaining 40% wood f lour at different humidities.

The hygroscopicity of wood f lour can also affect the end composite. Absorbedmoisture interferes with and reduces hydrogen bonding between cell wall polymersand alters the mechanical performance of the product [22]. A moisture content of upto about 30% can be adsorbed by the cell wall, with a corresponding reversible in-crease in apparent wood volume. The wood volume V1, at a moisture content M, hasbeen roughly approximated by [16]:

V1 = V0 (1 + 0.84 M ρ) (15-1)

where V0 is the dry volume and ρ is the specific gravity of the wood when dry.Volume changes due to moisture sorption, especially repeated moisture cycling,

can lead to interfacial damage and matrix cracking [23]. This damage, and the result-ing irreversible mechanical property reductions after the composites have been ex-posed to a humid environment or to liquid water, have been discussed in a numberof papers [23–25]. Water uptake depends on many variables, including wood f lourcontent, wood f lour particle size, matrix type, processing method, and additives suchas coupling agents. Many manufacturers of wood–plastic composites used in exteri-or applications limit wood f lour content to 50–60% by weight and rely on the partialencapsulation of the wood by the polymer matrix to prevent significant moisturesorption and the consequent negative effects.

15 Wood Flour

Fig. 15-9 Moisture sorption of wood f lour at several relativehumidities and 26 °C.

259

15.3.5Durability

Wood will last for decades in exterior environments, especially if it is stained, paint-ed or otherwise protected. However, wood–plastic composites are not commonly pro-tected. In fact, a common selling point for wood–plastic composites is that they arelow maintenance materials and do not require painting or staining for use in outdoorapplications.

The surface of wood undergoes photochemical degradation when exposed to UVradiation. This degradation takes place primarily in the lignin component and resultsin a characteristic color change [26]. Hence, wood–plastic composites containing nopigments usually fade to a light gray when exposed to sunlight. Photostabilizers orpigments are commonly added to wood–plastic composites to help reduce this colorfade when they are used in exterior environments.

Mold can form on surfaces of wood–plastic composites. Mold growth has been at-tributed to various effects, among them moisture sorption by the wood f lour, build-up of organic matter on the composite surface, and the lubricants used in processingthe composites. The relative contributions of these factors to mold growth are uncer-tain. Although mold does not reduce the structural performance of the composite, itis an aesthetic issue.

Wood is degraded biologically because organisms recognize the celluloses andhemicelluloses in the cell wall and can hydrolyze them into digestible units using spe-cific enzyme systems [26]. If the moisture content of the wood f lour in the compos-


Fig. 15-10 Moisture sorption of compounded pellets of polypropy-lene containing 40% wood f lour at several relative humidities and26 °C [49].

260

ite exceeds the fiber saturation point (approximately 30% moisture), decay fungi canbegin to attack the wood component leading to weight loss and significant reductionin mechanical performance. Figure 15-11 shows the weight loss due to exposure of awood–plastic composite to the decay fungi Gleophylum Trabeum in a laboratory soilblock test. Decay does not commence until a moisture threshold of about 15% isreached. Since HDPE does not absorb moisture, the average moisture content of thewood f lour in the composite would be expected to be roughly twice that shown. Thissuggests that when the moisture content of the wood f lour reaches about 30%, ap-proximately the fiber saturation point, significant decay begins. Additives such aszinc borate are sometimes added to wood–plastic composites to improve fungal re-sistance.

15.3.6Thermal Properties

Figure 15-12 shows a thermogravimetric analysis of wood f lour and its constituents.The onset of degradation differs for the major components of wood, with cellulose be-ing the most thermally stable. Due to its low thermal stability, wood f lour is usuallyused as a filler only in plastics that are processed at low temperatures, lower thanabout 200 °C. Above these temperatures, the cell wall polymers begin to decompose.High purity cellulose pulps, from which nearly all of the less thermally stable ligninand hemicelluloses have been removed, have recently been investigated for use inplastic matrices such as nylon that are processed at higher temperatures than mostcommodity thermoplastics [27].

15 Wood Flour

Fig. 15-11 Weight loss due to fungal attack (Gloeophyllum trabeum)as a result of moisture sorption. Extruded composites of high densi-ty polyethylene containing 50% wood f lour [50].

261

Because of its practical performance, the thermal properties of wood have been ex-tensively investigated [17]. Understandably, this work has generally been performedon uncompressed wood at moisture contents typical of those found in service. Infor-mation on dry, compressed wood as might be found in a wood–thermoplastic com-posite is lacking. Additionally, thermal properties vary depending on the chemistryand structure of the wood. Factors such as extractive content, grain direction, and fib-ril angle are important. Though precise numbers are not known, some approxima-tions may be made for a broad discussion on the topic.

The thermal expansion of wood is less than that of the commodity plastics com-monly used as matrices. Thermal expansion coefficients for wood are directional andare roughly given by [17]:

α = A ρ × 10– 6 (15-2)

where α is the coefficient of thermal expansion (in K–1), ρ is the specific gravity(oven-dried basis), and A is roughly 50–80 perpendicular to the fiber direction andabout 5–10 times less in the fiber direction. This roughly yields an average of about70 × 10– 6 K–1 if we assume a density of 1.5. This is about half that of polypropylene(150 × 10– 6 K–1) and 3.5 times less than that of low density polyethylene


Fig. 15-12 Thermogravimetric analysis of wood and wood compo-nents. (Reprinted with permission from ref. [51]).

262

(250 × 10– 6 K–1), both of which are commonly used as matrix materials for wood–plas-tic composites [28].

The specific heat of dry wood does not show a strong dependence on specific grav-ity and is roughly 0.324 cal g–1 K–1 or 1,360 J kg–1 K–1 [17]. This is about half the spe-cific heat of common polyolefins such as polypropylene and polyethylene, for whichthe values are approximately 2–3,000 J kg–1 K–1.

The thermal conductivity, k, of dry wood has been reported to increase approxi-mately linearly with specific gravity, ρ, according to [17]:

k = 0.200 ρ + 0.024 (15-3)

where k is in units of W m–1 K–1. Assuming the specific gravity of compressedwood f lour in a composite to be 1.5, the thermal conductivity is calculated as0.32 W m–1 K–1. This is the same order of magnitude as the values reported forpolypropylene and polyethylene (0.17–0.51 W m–1 K–1).

Thermal diffusivity is a measure of the rate at which a material changes tempera-ture when the temperature of its surroundings changes. The thermal diffusivity, h, isthe ratio of thermal conductivity k to the product of specific heat, c, and density, D[17]:

h = k / c D (15-4)

Calculations for wood f lour yield a value of 0.16 × 10– 6 m2 s–1, compared to0.11–0.17 × 10– 6 m2 s–1 for polypropylene and polyethylene.

15.4Suppliers

There is a wide range of wood f lour suppliers and they cater to a number of differentindustries. These are both large companies that have broad distribution networks, aswell as small, single source suppliers catering to single customers. Because of the var-ied and disperse nature of these suppliers, there are currently few good resources thatlist wood f lour manufacturers.

Wood–plastic composite manufacturers obtain wood f lour either: 1) directly fromforest products companies such as lumber mills and furniture, millwork, or windowand door manufacturers that produce it as a by-product, or 2) commercially fromcompanies that specialize in wood f lour production.

With a growing number of wood f lour suppliers targeting the wood–plastic com-posites industry, they are beginning to be listed in plastics industry resources. Thefollowing is a list of major suppliers of wood f lour to the U.S. wood–plastic compos-ite industry from one industry resource [29].

American Wood Fibers (Schofield, Wisconsin) Composition Materials (Fairfield, Connecticut)

15 Wood Flour

263

Lang Fiber (Marshfield, Wisconsin) Marth Manufacturing (Marathon, Wisconsin) P. J. Murphy Forest Products (Montville, New Jersey)


As with most materials, wood f lour costs are variable and depend on such factors asvolume, availability, particle size, and shipping distance. However, wood f lour is typ-ically about $0.11–0.22 kg–1 ($0.05–0.10 lb–1) in the United States. Narrow particlesize distributions and fine sizes tend to increase cost. Because there are many smallmanufacturers and the volume is small relative to other wood products (solid wood,wood composites, and paper), information on wood f lour availability is scarce.


The environmental benefits of wood and other natural fibers have been an importantinf luence on their use, particularly in Europe. Wood f lour is derived from a renew-able resource, does not have a large energy requirement to process, and is biodegrad-able [30].

Wood is a commonly used material and most people are very comfortable with itsuse. Most of the risk in using wood f lour lies in the facts that: 1) it has low thermalstability and can degrade and burn resulting in a fire and explosion hazard that isgreater than that with solid wood, and 2) inhalation of finely ground wood f lour canlead to respiratory difficulties. Some basic precautions include avoiding high pro-cessing temperatures, using well-ventilated equipment, eliminating ignition sources,and using good dust protection, prevention, and control measures. For detailed in-formation on environmental and health risks, manufacturers should consult theirsuppliers and material safety data sheets. Regulatory bodies such as the Occupation-al Safety and Health Association (OSHA) also have good information on health andsafety aspects of wood dust (see, e.g., www.osha.gov/SLTC/wooddust/index.html).


15.7.1Thermosets

In thermosetting adhesives, wood f lour has been used for several functions. As an ex-tender, it is added to reduce cost while retaining bulk for uniform spreading. Unfor-tunately, wood f lour extenders generally also reduce the durability of a given resin.


264

As a filler, it is added to thermoset adhesives to control penetration when bondingwood and to improve the characteristics of the hardened film [31].

High weight percentages of wood f lour have been used with thermosets such asphenolic or urea/formaldehyde resins to produce molded products. The wood f louris added to improve toughness and reduce shrinkage on curing. Wood filler may beadded to the thermoset resin at elevated temperatures to form a moldable paste. Thispaste is then transferred to a mold and cured under heat and pressure [32]. Alterna-tively, a mixture of powdered thermoset resin and wood f lour is poured directly intoa mold and pressed under heat and pressure. Although this type of composite wasvery prevalent throughout much of the 20th century, often under trade names such as“Bakelite”, its use has diminished considerably. However, a variety of products, suchas some salad bowls, trays and cutting boards, are still manufactured from wood ther-moset composites.

15.7.2Thermoplastics

15.7.2.1 Mechanical Property Modif ication (Primary Function)Wood f lour is often added to thermoplastics as a low cost filler to alter mechanicalperformance, especially the stiffness of low melt temperature, commodity thermo-plastics such as polypropylene and polyethylene, without increasing their density ex-cessively. Wood is much stiffer than the commodity thermoplastics usually used asmatrices. Additionally, wood and pulp fibers can nucleate crystal growth inpolyalkenes, resulting in a transcrystalline layer that can inf luence mechanical be-havior [33,34].

The wood f lour stiffens these plastics but also embrittles them, reducing proper-ties such as elongation and unnotched impact energy. Tensile and f lexural strengthsare at best maintained, but more often decreased in the absence of a coupling agent.Many different coupling agents have been investigated for use in wood–plastic com-posites and these are reviewed elsewhere [35,36]. When a coupling agent is desired,maleated polyalkenes are most often used commercially. However, even when a cou-pling agent is used, improvements in strength are limited by the low aspect ratio ofthe wood f lour (Figure 15-13). It is often unclear as to how much of the strength in-crease is due to better wetting and dispersion and how much is due to the increasedbonding. Small amounts of thermosets have also been added to wood–plastic formu-lations to improve mechanical performance [37].

Table 15-4 summarizes some typical mechanical property changes when woodf lour from various species of tree is added to an injection-molding grade ofpolypropylene [44]. Both tensile and f lexural moduli increase with the addition ofwood f lour. Composites with hardwoods (maple and oak) yield the highest f lexuralmoduli at 60% wood f lour content; i.e., approximately four times that of unfilledpolypropylene. The heat def lection temperature is also approximately doubled byadding 60% wood f lour. The increase in modulus with the addition of wood f lourcomes at the expense of elongation, a drastic reduction in unnotched impactstrength, and a general decrease in tensile strength.

15 Wood Flour

265

The effects of wood f lour particle size on the mechanical performance of wood-filled polypropylene are summarized in Table 15-5. The largest particle size range(0.425–0.600 µm) yields the lowest performance, except in the case of notched impactstrength. For the three smaller particle size ranges, properties generally decrease asparticle size is reduced, except for the unnotched impact strength. Similar trendswere found when broader particle size ranges more typical of commercially availableblends were investigated [38].

15.7.2.2 To Impart Wood-Like Properties (Primary Function)Wood f lour is also often added to make plastics perform more like wood. Customersand builders have a certain familiarity with wood in applications such as decking andrailings (the largest wood–plastic composite market) and often desire an alternativethat may have similar attributes. As a result, many manufacturers add 50–60% woodf lour to plastics to impart some wood-like qualities.

Adding wood f lour to thermoplastics results in a composite with a wood color, al-though color-fade remains an issue in exterior applications. If desired, extrudedwood–plastic composites can usually be painted or stained. Though not nearly as stiffas solid wood, these composites are stiffer than unfilled plastics. They do not usual-ly require special fasteners or design changes such as shorter spans in applicationssuch as deck boards. Adding wood f lour also improves dimensional stability with re-spect to temperature changes. Although more expensive than wood, many con-sumers have been willing to pay for the lower maintenance required whenwood–plastic composites are used.

However, wood–plastic composites are heavier than wood and do not have as goodmechanical performance, often having only one-third to a half of the modulus of elas-ticity of wood [39]. The swelling of wood with moisture can create irreversible dam-


Fig. 15-13 Comparison between wood f lour and wood fiber as areinforcement in polypropylene. Injection-molded composites;3% maleated polypropylene added as a coupling agent. (Derivedfrom ref. [56]).

266

age, and the exact durability of wood–plastic composites in exterior applications is un-known. Most deck board manufacturers currently offer a ten-year warranty. As a re-sult of these limitations, much research and development effort is underway to try toincrease structural performance, improve durability, and decrease weight without toogreat an increase in cost.

15 Wood Flour

Tab. 15-4 Effect of adding wood f lour on the mechanical performance of injection-molded polypropylene[a] [44].

Izod Impact strength[b] Flexural properties[c] Tensile properties[d]

Filler Notched Unnotched Max. Modulus Maximum Modulus Elongation Heat Content f lexural of elasticity tensile of elasticity at max. def lection

strength strength strength temp.[e]

(wt. %) (J m–1) (J m–1) (MPa) (GPa) (MPa) (GPa) (%) (%)

No filler0 15.0 600 34.7 1.03 28.5 1.31 10.4 55Ponderosa pine20 15.4 128 41.6 1.89 26.5 1.99 5.7 6930 19.0 95 43.1 2.58 24.6 3.24 3.1 7640 20.8 76 44.2 3.22 25.5 3.71 2.3 8550 20.5 58 41.8 3.66 23.0 4.25 1.7 8960 21.1 41 38.8 4.04 20.1 4.56 1.4 91Loblolly pine20 12.4 120 40.6 1.71 24.9 2.14 4.8 6430 12.7 75 41.2 2.28 23.7 2.53 3.5 7240 13.7 49 39.3 2.84 21.4 3.28 1.8 7850 13.8 42 37.1 3.40 19.7 3.97 1.3 7960 10.4 31 34.1 3.81 17.8 4.32 0.9 77Maple20 12.8 113 46.2 2.16 27.9 2.87 4.1 6930 15.0 87 46.5 2.47 27.1 3.33 3.3 8840 16.5 63 45.4 3.23 25.6 4.72 2.0 10450 17.7 49 42.1 4.16 24.0 5.20 1.4 11160 17.9 44 38.0 4.35 19.9 4.77 1.1 110Oak20 14.6 87 44.1 1.83 27.2 2.48 4.7 7430 17.5 63 45.9 2.87 25.7 3.81 2.4 9840 18.6 68 44.8 3.39 25.2 4.19 2.1 10050 20.9 46 42.8 3.99 23.4 4.79 1.5 11260 18.8 33 38.1 4.60 19.8 5.05 1.2 114Avg. COV12 6 2 4 2 8 10 2

[a] Fortilene 3907, polypropylene homopolymer, SolvayPolymers, Deer Park, TX, USA.

[b] ASTM D-256 [45].[c] ASTM D-790 [46].

[d] ASTM D-638 [47].[e] ASTM D-648 [48].[f ] Wood f lour was commercial grade from American

Wood Fibers, Schofield, WI, USA.

267

15.7.2.3 Environmental Preference and Biodegradability (Secondary Function)The environmental advantages of wood and other natural fibers have been an im-portant inf luence, particularly in Europe. Wood f lour is manufactured from indus-trial by-products, mitigating a disposal issue. Wood f lour and other natural fibers arederived from a renewable resource, do not have a large energy requirement toprocess, and are biodegradable [40]. These features are advantageous in applicationswhere environmental benefits and impact are important. Though not specificallyadded to plastics to impart biodegradability, wood f lour can be used as a filler inbiodegradable polymers where its biodegradability is an attribute rather than thedetriment it is sometimes considered to be in more durable composites.

Authors’ Note

The Forest Products Laboratory is maintained in cooperation with the University ofWisconsin. This article was written and prepared by U.S. Government employees onofficial time, and it is therefore in the public domain and not subject to copyright.The use of trade or firm names in this publication is for reader information and doesnot imply endorsement by the U.S. Department of Agriculture of any product or serv-ice.


Table 15-5 Mechanical properties of composites made frompolypropylene[a] filled with 40 wt. % wood f lour [37].

Izod Impact strength[b] Flexural properties[c] Tensile properties[d]

Particle Notched Un- Max. Modulus Maximum Modulus Elongation size notched f lexural of tensile of at max. range strength elasticity strength elasticity strength

(µm) (J m–1) (J m–1) (MPa) (GPa) (MPa) (GPa) (%)

No filler– 15.0 600 34.7 1.03 28.5 1.31 10.4Composites with 40% wood flour[e]

425–600 22 54 38.7 2.69 21.8 3.20 2.3180–250 20 79 42.6 3.15 25.5 3.61 2.3106–150 19 84 42.9 3.00 24.9 3.47 2.253–75 16 91 41.4 2.89 24.3 3.46 2.1Avg. COV 5 10 1 2 1 7 6

[a] Fortilene 3907, polypropylene homopolymer,Solvay Polymers, Deer Park, TX, USA.

[b] ASTM D-256 [45].[c] ASTM D-790 [46].

[d] ASTM D-638 [47].[e] Specially screened wood f lour from American

Wood Fibers, Schofield, WI, USA.

268

References

15 Wood Flour

1 Reineke, L. H., Wood Flour, U.S. Forest Ser-vice Research Note FPL-0113, USDA ForestService, Forest Products Laboratory, Madi-son, WI, USA, 1966, 13 pages.

2 Panshin, A. J., et al., Chapter 13: “WoodFlour”, in Forest Products: Their Sources, Pro-duction, and Utilization, McGraw-Hill, NewYork, 1962, p. 265–271.

3 Gordon, J. E., The New Science of Strong Ma-terials (or why you don’t fall through thef loor), 2nd ed., Princeton University Press,Princeton, NJ, 1988, p. 179.

4 Morton, J., Rossi, L., “Current and Emerg-ing Applications for Natural and WoodFiber Composites”, Proc. 7th InternationalConference on Woodfiber-Plastic Composites,Madison, WI, Forest Products Society, 2003.

5 Rowe, J. W., Conner, A. H., “Extractives inEastern Hardwoods – A Review”, GeneralTechnical Report FPL 18, USDA Forest Ser-vice, Forest Products Laboratory, Madison,WI, 1979, p. 17.

6 Bigg, D. M., et al., J. Thermoplastic Compos-ite Materials 1988, 1, 146–161.

7 “International Sieve Chart”,www.reade.com/Sieve/international_sieve.html ©1997 Reade Advanced Materials.

8 “Wood Flour: Discover the Possibilities”,product literature, P. J. Murphy ForestProducts Corp., Montville, NJ.

9 Miller, R. B., “Structure of Wood”, Chapter2 in the Wood Handbook: Wood as an Engi-neering Material, General Technical Report,FPL-GTR-113, USDA Forest Service, ForestProducts Laboratory, Madison, WI, USA,1999, 463 pages.

10 Panshin, A. J., de Zeeuw, C., Textbook ofWood Technology, 4th ed., McGraw-Hill,New York, 1980.

11 Hoadley, R. B., Identifying Wood, AccurateResults with Simple Tools, Taunton Press,Newtown, CT, 1990.

12 Rowell, R. M., “Chemical Modification ofWood”, Forest Products Abstracts 1983, 6(12),363–382.

13 Imai, T., Sugiyama, J., Macromolecules 1998,31, 6275–6279.

14 Wada, M., et al., “The Monoclinic Phase isDominant in Wood Cellulose”, MokuzaiGakkaishi 1994, 40(1), 50–56.

15 Pettersen, R. C., “The Chemical Composi-tion of Wood”, Chapter 2 of The Chemistryof Solid Wood (Ed.: Rowell, R. M.), Ameri-can Chemical Society, Washington DC,1984, pp 76–81.

16 Parham, R. A., Gray, R. L., “Formation andStructure of Wood”, in The Chemistry of Sol-id Wood (Ed.: Rowell, R. M.), AmericanChemical Society, Washington DC, 1984,pp. 1–56.

17 Kellogg, R. M., “Physical Properties ofWood”, in Wood: Its Structure and Properties(Ed.: Wangaard, F. F.), Pennsylvania StateUniversity, College Park, PA, 1981, pp191–223.

18 Simpson, W., TenWolde, A., “Physical Prop-erties and Moisture Relations of Wood”,Chapter 3 in The Wood Handbook: Wood asan Engineering Material, General TechnicalReport, FPL-GTR-113, USDA Forest Ser-vice, Forest Products Laboratory, Madison,WI, USA, 1999, p. 2-1 to 2-4.

19 Karayan, V., “The Effect of Chemical Foam-ing Agents on the Processing and Proper-ties of Polypropylene-Wood Flour Compos-ites”, Proc. 6th International Conference onWoodfiber-Plastic Composites, Forest Prod-ucts Society, Madison, WI, 2001, pp249–252.

20 Turng, L.-S., et al., “Applications of Nano-Composites and Woodfiber Plastics for Mi-crocellular Injection Molding”, Proc. 7th In-ternational Conference on Woodfiber-PlasticComposites, Forest Products Society, Madi-son, WI, 2003.

21 Matuana-Malanda, L., et al., J. Cellular Plas-tics 1996, 32(5), 449–469.

22 Winandy, J. E., Rowell, R. M., “The Chem-istry of Wood Strength”, Chapter 5 of TheChemistry of Solid Wood (Ed.: Rowell, R. M.),American Chemical Society, Washington,DC, 1984, p 218.

23 Peyer, S., Wolcott, M., “Engineered WoodComposites for Naval Waterfront Facilities”,2000 Yearly Report to Office of Naval Re-search, Wood Materials and EngineeringLaboratory, Washington State University,Pullman, WA, 14 pages.

24 Stark, N. S., J. Thermoplastic Composite Ma-terials 2001, 14(5), 421–432.

269References

25 Balatinecz, J. J., Park, B.-D., J. ThermoplasticComposite Materials 1997, 10, 476–487.

26 Rowell, R. M., “Penetration and Reactivityof Cell Wall Components”, Chapter 4 of TheChemistry of Solid Wood (Ed.: Rowell, R. M.),American Chemical Society, Washington,DC, 1984, p. 176.

27 Sears, K. D., et al., “Reinforcement of Engi-neering Thermoplastics with High PurityCellulose Fibers”, Proc. 6th InternationalConference on Woodfiber-Plastic Composites,Forest Products Society, Madison, WI, 2001,pp. 27–34.

28 Osswald, T. A., Menges, G., Materials Sci-ence of Polymers for Engineers, Carl HanserVerlag, New York City, NY, 1996, pp458–459.

29 Anonymous, “Special Report: Who’s Whoin WPC”, in Natural and Wood Fiber Com-posites, Principia Partners, 2003, 2(8), p. 4.

30 Suddell, B. C., Evans, W. J., “The IncreasingUse and Application of Natural Fibre Com-posite Materials within the Automotive In-dustry”, Proc. 7th International Conferenceon Woodfiber-Plastic Composites, Madison,WI, 2003.

31 Marra, A. A., Technology of Wood Bonding:Principles in Practice, Van Nostrand Rein-hold, New York, NY, 1992, p. 438.

32 Dahl, W. S., Woodf lour, The Mercury Press,The Parade, Northampton, England, 1948,119 pages.

33 Quillin, D. T., et al., J. Applied Polymer Sci-ence 1993, 50, 1187–1194.

34 Lee, S. Y., et al., Inf luence of Surface Char-acteristics of TMP Fibers to the Growth ofTCL on the Linear Fibers Surface (I): As-pect in the Surface Roughness”, Proc. 6thInternational Conference on Woodfiber-PlasticComposites, Forest Products Society, Madi-son, WI, 2001, pp 107–118.

35 Lu, J. Z., et al., “Chemical Coupling inWood Fiber and Polymer Composites: A Re-view of Coupling Agents and Treatments”,Wood and Fiber Science 2000, 32(1), 88–104.

36 Bledski, A. K., et al., Polymer and PlasticsTechnology and Engineering 1998, 37(4),451–468.

37 Wolcott, M. P., Adcock, T., “New Advancesin Wood Fiber-Polymer Formulations”,Proc. Wood-Plastic Conference, Baltimore,MD, Plastics Technology Magazine andPolymer Process Communications, 2000,pp. 107–114.

38 Stark, N. M., Berger, M. J. “Effect of ParticleSize on Properties of Wood Flour Rein-forced Polypropylene Composites”, Proc.4th International Conference on Woodfiber-Plastic Composites, Forest Products Society,Madison, WI, 1997, pp 134–143.

39 English, B. W., Falk, R. H., “Factors that Af-fect the Application of Woodfiber-PlasticComposites”, Proc. Woodfiber-Plastic Com-posites: Virgin and Recycled Wood Fiber andPolymers for Composites Conf., The ForestProducts Society, Madison, WI, 1995, pp189–194.

40 Test Method ASTM E11-01: “Standard Spec-ification for Wire Cloth and Sieves for Test-ing Purposes”, Annual Book of ASTM Stan-dards, Vol. 14.02, ASTM International, WestConshohocken, PA, 2001.

41 Test Method ASTM D1103-60 (1977):“Method of Test for Alpha-Cellulose inWood”, Annual Book of ASTM Standards,Vol. 04.10, ASTM International, West Con-shohocken, PA, 2001.

42 Test Method ASTM D1106-96 (2001): “Stan-dard Test Method for Acid-Insoluble Ligninin Wood”, see ref. [41].

43 Test Method ASTM D1107-96: “StandardTest Method for Ethanol-Toluene Solubilityof Wood”, see ref. [41].

44 Berger, M. J., Stark, N. M., “Investigationsof Species Effects in an Injection-Molding-Grade, Wood-Filled Polypropylene”, Proc.4th International Conference on Woodfiber-Plastic Composites, Forest Products Society,Madison, WI, 1997, pp 19–25.

45 Test Method D256-02e1: “Standard TestMethods for Determining the Izod Pendu-lum Impact Resistance of Plastics,” AnnualBook of ASTM Standards, Vol. 08.01, ASTMInternational, West Conshohocken, PA,2002.

46 Test Method D790-03: “Standard Test Meth-ods for Flexural Properties of Unreinforcedand Reinforced Plastics and Electrical Insu-lating Materials”, see ref. [45].

47 Test Method D638-02a: “Standard TestMethod for Tensile Properties of Plastics”,see ref. [45].

48 Test Method D648-01: “Standard TestMethod for Def lection Temperature of Plas-tics Under Flexural Load in the EdgewisePosition”, see ref. [46].

49 English, B., et al., “Waste-Wood DerivedFillers for Plastics”, General Technical Re-

270 15 Wood Flour

port FPL-GTR-91, Madison, WI, U.S. De-partment of Agriculture, Forest Service,Forest Products Laboratory, 1996, pp282–291.

50 Clemons, C. M., Ibach, R. E., “The Effectsof Processing Method and Moisture Historyon the Laboratory Fungal Resistance ofWood-HDPE Composites”, Forest ProductsJournal, in press.

51 Nakagawa, S., Shafizadeh, F., “ThermalProperties” in Handbook of Physical and Me-chanical Testing of Paper and Paperboard(Ed.: Mark, R. E.), Marcel Dekker, NewYork, 1984.

52 Stark, N. M., Forest Products Journal 1999,49(6), 39–45.

16Calcium Carbonate

Marino Xanthos

16.1Background

Calcium carbonate is the most common deposit formed in sedimentary rocks. Nat-ural CaCO3 used as a filler in plastics is produced from chalk, limestone or marblefound in the upper layers of the Earth’s crust to a depth of about 15 km. Chalk is asoft textured, microcrystalline sedimentary rock formed from marine microfossils.Limestone is also of biological origin but it is harder and denser than chalk, havingbeen compacted by various geological processes. Marble is even harder, having beingsubjected to metamorphosis under high pressures and temperatures, which resultedin recrystallization with the separation of impurities in the form of veins. The originsof calcium carbonate in deposits and in living organisms are discussed in ref. [1].

The sedimentary rocks consisting mainly of calcite crystals are processed by stan-dard mining procedures and then subjected to grinding and classification. In addi-tion to natural ground calcium carbonate (GCC), there also exists a chemically pro-duced form known as precipitated calcium carbonate (PCC), which may be finer andof higher purity, but also more expensive than the natural one.

Calcium carbonate is an abundant, largely inert, low cost, white filler with cubic,block-shaped or irregular particles of very low aspect ratio (see Table 1-3 in Chapter 1of this book). Its use yields a cost reduction in a variety of thermoplastics and ther-mosets, and it can have moderate effects on mechanical properties. In terms of itsprimary function as a mechanical property improver, its advantages in thermoplas-tics are slightly increased modulus and often an increase in impact strength. Thesebenefits are accompanied by shrinkage reduction and improved surface finish. Interms of secondary functions, calcium carbonate may act as a surface property mod-ifier, as a processing aid, and as a stress concentrator by introducing porosity instretched films. These functions may be enhanced or modified by the application ofsuitable surface treatment agents, such as stearates or titanates.


272


More than 90% of the CaCO3 used in plastics is processed by conventional grindingmethods. The following processes are used to produce GCC [1,2]:

1. Dry grinding followed by air separation. The product is a relatively coarse powderwith the finest ground material having a median diameter of ∼12 µm and a broadsize distribution. It is suitable for inexpensive dark f loor tiles and vinyl foam car-pet backing.

2. Wet grinding, whereby the coarse particles are removed by centrifuging.3. Beneficiation and grinding. Through crushing, disintegrating, removal of iron

and silica by magnetic and/or f lotation means, and finely grinding, powders hav-ing a median particle size between about 1 and 10 µm are produced. These prod-ucts are the most commonly used in plastics.

Grinding costs increase as degree of fineness increases, until wet milling becomesmore economical than dry grinding in spite of the additional drying costs [3]. If thewet grinding process is used, the material may often be delivered to the customer ina slurry form. The typical particle morphology of the ground materials is rhombohe-dral.

Precipitated calcium carbonate (PCC) is also known as synthetic CaCO3 since sev-eral chemical operations may be involved in its manufacture. In earlier syntheticprocesses, it was obtained as a by-product during the manufacture of Na2CO3 by theammonia process, or during the manufacture of NaOH by the soda-lime process [2].These methods have been replaced by a direct synthetic process that involves calci-nation of CaCO3 at 900 °C to produce quick lime, CaO, conversion into slaked lime,Ca(OH)2, by mixing with water, and reaction with CO2, the latter recovered from thecalcination process. Following carbonation, the suspension is filtered, and the col-lected solid is dried and deagglomerated in grinders. Depending on the process con-ditions, various morphologies (spherical, discrete or clustered acicular, prismatic,rhombohedral, scalenohedral, orthorhombic) and crystal forms (calcite, aragonite)are possible. A schematic of the process may be found in ref. [1] and photographs ofthe obtained crystals accessed via the supplier’s website [4].

PCC products are of high purity with a very fine, regular particle size, a narrow par-ticle size distribution, and high surface area. An advantage of the process is that thereis no co-product to separate, obviating the need for any additional steps. However, itis a higher cost process and more energy intensive than that involved in GCC pro-duction. The median particle size is 0.7–2.0 µm, with the primary particle size assmall as 20–70 nm. The high specific surface area may adversely affect rheology athigh loadings through excessive adsorption of stabilizers/plasticizers in PVC, andcan cause excessive viscosity increase in unsaturated polyesters.

Most of the CaCO3 produced for use in plastics is surface-treated, usually with fat-ty acids, to produce a hydrophobic filler with improved rheological properties and, insome cases, improved mechanical properties. Surface coating with 0.5–1.5% stearic

16 Calcium Carbonate

273

acid may be accomplished by dry methods in high intensity mixers above the melt-ing temperature of the stearic acid. The challenge is to convert as much of the coat-ing to a bound surface layer as possible, leaving a minimum of unbound salt and re-maining free acid (see Chapter 6 and ref. [3]). Titanates and zirconates are also used(see Chapter 5) and, less frequently, other additives such as carboxylated polybutadi-ene [5].


Calcium carbonate occurs in different crystalline forms. The most widespread is cal-cite, which has either a trigonal-rhombohedral or a trigonal-scalenohedral crystal lat-tice. The fundamental properties are shown in Table 16-1. Another form is the or-thorhombic aragonite, which is less stable and can be converted to calcite by heating.Vaterite, a third form, is unstable and over time will transform into the other twoforms. Aragonite has a higher density (2.8–2.9 g cm–3), a higher single refractive in-dex (1.7), and a somewhat higher Mohs hardness (3.5–4) than calcite. Its other prop-erties are very similar. Both minerals are white and their refractive indices are nothigh enough to interfere with effective coloration.

Tab. 16-1 Properties of ground calcite [1,2,6].

Property Value

Loss on ignition, 950 °C 43.5Density, g cm–3 2.7Hardness, Mohs 3Water solubility, g/100 mL 0.0013Acid solubility highYoung’s modulus, MPa 35000Thermal conductivity, W m–1 K–1 2.5Thermal expansion coefficient, K–1 10–5

Volume resistivity, Ohm cm 1010

Dielectric constant at 104 Hz 8–8.5Refractive indices 1.48, 1.65pH in 5% water slurry 9.0–9.5

Commercial GCC grades, wet or dry ground, contain 94–99% CaCO3, MgCO3 asthe major impurity, and minor quantities of alumina, iron oxide, silica or manganeseoxide. PCC may contain 98–99% CaCO3, or even higher concentrations for pharma-ceutical grades.


274

16.4Suppliers

Omya and Imerys are among the largest suppliers mining worldwide. Other Euro-pean suppliers include Provençale S.A. (France), Reverté S.A. (Spain), Alpha Calcit(Germany), Mineraria Sacilese (Italy), Dankalk (Denmark), Solvay (Germany), andSchaefer Kalk (Germany). Major U.S. suppliers in addition to Omya and Imerys in-clude J. M. Huber (GA), Specialty Minerals, Inc. (PA), Columbia River Carbonates (WA),Global Stone PenRoc, Inc. (PA), and Franklin Industrial Minerals (TN). All these sup-pliers list their products on their respective websites, and some of them also providePCC grades. There are numerous other suppliers in the Asia-Pacific region and inother European countries.

GCC grades available for use in plastics are classified according to fineness andtype of surface treatment. Ultrafine grades have mean particle size 1–2 µm. Finegrades are of size 3–7 µm, and surface-treated grades have mean particle size (basedon untreated material) in the range 1–3 µm [7]. Grades of different particle size arerecommended depending on the polymer, the intended application, and the fabrica-tion method. PCC grades can be submicron in mean particle size.


Prices vary depending on degree of fineness and surface treatment, with coated finegrades being the most expensive [8]. For GCC, early 2004 FOB prices in US$ per tonare $110–160 for 5–7 µm grades and $140–290 for 2–0.5 µm grades. For PCC, the cor-responding FOB (free on board) prices in US$ per ton are $250–270 for fine(0.4–1 µm) grades and $375–750 for ultrafine (0.02–0.36 µm) surface-treated grades.

In a recent marketing report [9], it is estimated that the world production of highbrightness, finely ground calcium carbonate amounts to 20–30 million tons per an-num, with the most likely consumption figure being around 24–25 million tons. Thepaper industry is the most important consumer of GCC. The plastics and paint in-dustries consume a combined 10–12 million tons of high quality GCC, with PCCcomprising a very small percentage of the overall market. World production is almostequally split between Europe/N.America and the Asia-Pacific region, where produc-tion is dominated by China and Japan [10]. Two large companies with worldwide op-erations, Omya and Imerys, are the major producers, accounting for a significant pro-portion of the global supply. A number of other medium-sized suppliers have opera-tions in more than one country. Over the years, there has been a considerable con-solidation of both producers and consumers of GCC and this trend is expected tocontinue.


275


Calcium carbonate is practically harmless with no known occupational diseases as-sociated with its handling. It is non-toxic and non-hazardous, and most grades meetUnited States FDA codes. The United States OSHA regulates the substance underthe generic total particulate limit of 15 mg m–3. Specific exposure limits for the vari-ous forms of natural CaCO3 dust containing less than 1% quartz according to the var-ious U.S. regulatory agencies are as follows:

OSHA PEL: 8-hour TWA 15 mg m–3 (total), TWA 5 mg m–3 (respiratory). NIOSH REL: TWA 10 mg m–3 (total), TWA 5 mg m–3 (respiratory). ACGIH: 8-hour TLV, TWA of 10 mg m–3 dust.

Further details on environmental/toxicity considerations can be found in suppli-ers’ material safety data sheets (MSDS).

16.7Applications

16.7.1General

The primary function of calcium carbonate as a filler is to lower costs, while havingmoderate effects on mechanical properties. However, depending on the particularpolymer system, it may also be considered as a multifunctional filler with a variety ofspecific effects on rheology, processing, and morphology. In any case, the introduc-tion of surface-treated calcium carbonate and of ultrafine grades has undoubtedly ledto the development of new applications. Examples of the importance of surface treat-ments on the rheological, mechanical, and other properties of CaCO3-filled plasticsare presented in Chapters 5 and 6 of this book.

The majority of calcium carbonate applications (about 80%) are in polyvinyl chlo-ride, PVC (f lexible, plastisols, rigid), and thermosets, primarily fiber glass reinforcedunsaturated polyesters, UP. The remaining 20% are in polyolefins (primarily injec-tion-molded parts and films), EPDM, polyurethane, polyamide, rubber, ABS, etc. Asone of the oldest commodity fillers, a significant amount of technical informationhas been published on its uses in plastics over the years. A summary of traditionaland novel applications in specific polymers, which also demonstrates its multiplefunctions, is presented below.

16.7 Applications

276

16.7.2Polyvinyl Chloride

In filled PVC, 80% of the mineral market is accounted for by CaCO3 (ground 70%,precipitated 10%). Important characteristics of GCC are its ready availability in dif-ferent particle sizes and its low cost. It imparts increased stiffness, often improvesimpact strength in rigid formulations, and results in low shrinkage and improved di-mensional stability.

Filler levels in f lexible PVC are usually in the range of 20 to 60 phr (parts per hun-dred of resin) of 3 µm filler in such applications as upholstery coverings, hose extru-sion compounds, etc. [2]. In general, finer grades have a lesser detrimental effect onphysical properties. There are special requirements for the use of CaCO3 (up to15 phr) in vinyl electrical wire insulation compounds, where ionic impurities need tobe carefully controlled in order to meet applicable volume resistivity standards. InPVC plastisols and organosols, loading is usually 20 to 100 phr and a wide range ofparticle sizes is used, from coarse grades in carpet backing to ultrafine precipitatedgrades and coated grades that control rheological properties. Loadings up to 400 phrof fine GCC grades may be used in PVC f loor tiles, where the installation require-ment of easy break of a notched tile is achieved by the high filler loading.

Up to 5 phr of 2–3 µm surface-treated or untreated CaCO3 is used in pipes carry-ing potable water, meeting applicable industry and government standards. Surface-treated grades improve rheology and smoothness of extrusion, whereas uncoatedgrades serve as anti-plate-out agents. Up to 40 phr is used in other pipe, conduit, fit-ting, and duct applications. In rigid PVC extrusion, injection molding, and sidingcompounds, the incorporation of 3 µm CaCO3 at levels ranging from 10 to 40 phr usu-ally results in decreased impact and tensile strengths. However, the use of ultrafine(about 1.1 µm) surface-treated filler may increase Izod or Gardner impact strengthswith minimal effects on tensile strength [11]. Relevant technical information onCaCO3-filled PVC may be found in trade literature from PVC resin manufacturers,CaCO3 suppliers, and stabilizer manufacturers.

PCC is characterized by fine particle sizes and a narrow size distribution. In highvalue rigid PVC applications such as sidings, vertical blinds, or pipes, it imparts spe-cific properties such as improved mar resistance, increased surface gloss, reducedf lex whitening, and reduced plate-out. These benefits are in addition to increasedimpact strength and higher modulus. The improved impact strength of rigid PVC in-corporating coated ultrafine PCC at concentrations of 5–10 phr is maintained even attemperatures below 0 °C; this may allow a reduction in the concentration of expen-sive elastomeric additives required to maintain ductility [12].

PCC can be considered as a processing aid in rigid PVC. Its fine size is compatiblewith the PVC primary particles, improving dispersion of the components of the for-mulation. This may result in shorter fusion/gelation times. More complete gelationmay provide a matrix that has fewer defect sites than with coarser CaCO3. Further ad-vantages claimed by material suppliers [12] are elimination of plate-out; eliminationof surface defects, thereby improving surface finish; and increased output. PCC may


277

also act as an acid acceptor for secondary stabilization of PVC, neutralizing chlorideions, and as a rheology modifier, promoting uniform cell structure in PVC foams.

16.7.3Glass Fiber Reinforced Thermosets

Significant uses of GCC in unsaturated polyesters are in bulk molding compounds(BMC) and sheet molding compounds (SMC). Typical particle sizes range from 3 to10 µm at concentrations from 150 to 220 phr (parts per hundred of resin) dependingon the filler size. In addition to cost reduction, the filler improves the surface finishby reducing shrinkage and modifies rheology to prevent segregation of the glassfibers. Key requirements for these grades are low cost, low moisture content, lowresin demand, and absence of contaminants that interfere with the thickening reac-tion in SMC [2,7,13]. Coarser grades are used in so-called ‘hand-lay-up’ and ‘spray-up’applications, and even coarser grades (up to 60 µm) with low resin demand are usedfor cultured marble made by casting at loadings up to 300 phr. Table 6-1 in Chapter 6of this book contains an example of improved adhesion in filled UP through the useof unsaturated acids.

16.7.4Polyolef in Moldings

When compared to other fillers of different shape and aspect ratio (glass fibers, mi-ca f lakes, talc), calcium carbonate in polyolefins usually provides higher ductility andunnotched impact strength, albeit at the expense of modulus, tensile strength, andHDT. These effects are shown in Table 16-2 for injection-molded composites con-taining a predominantly HDPE matrix that was designed to simulate a recyclablecomposition [14]. Note that the lower aspect ratio CaCO3 and talc are used at 40%whereas the higher aspect ratio glass and mica are used at lower concentrations(20–25%).

Effects of surface treatment in CaCO3-filled polyolefins are presented in Chapters 5and 6. With regard to titanate treatment, data on the improvement of dispersion, rhe-ology, processability, and certain mechanical properties are provided in Tables 5-6 to5-8. The effect of deagglomeration afforded by titanates is shown in Figure 5-2. Fig-ure 5-3 shows the effects of titanate treatment in allowing higher filler loadings inthermoplastics through reducing viscosity. Figure 5-4 demonstrates the f lexibilityimparted to a 70 wt. % surface-treated CaCO3-filled PP homopolymer by the presenceof 0.5% of a titanate coupling agent.

Fatty acid (stearic, isostearic) and other polar coupling agent effects on the rheolo-gy and properties of CaCO3-filled polyolefins are discussed in detail in Chapter 6. Thebeneficial effects of maleated PP in filled PP compounds are shown in Table 6-9.Modification of the polyolefin matrix of Table 16-2 by maleic anhydride graftingthrough reactive extrusion increases the tensile, f lexural, and unnotched impactstrengths of the untreated CaCO3 compositions without having any significant effecton modulus, HDT, or elongation at break. Increased tensile or f lexural strength is ac-

16.7 Applications

278

companied by the disappearance of a yield point and a decrease in unnotched impactstrength (Table 16-3) [14]. Similar effects are observed when comparing the unmodi-fied with the modified matrix in composites containing surface-treated CaCO3. Com-parison of the fracture surfaces shown in Figures 16-1 and 16-2 is indicative of im-proved coating of the particles with the maleated matrix, which presumably resultsin improved adhesion. An interesting example of the multifunctional characteristicsof CaCO3 is included in Table 5-5 (Chapter 5). In LLDPE, a 44% loading of a fillertreated with a pyrophosphato titanate can lead to self-extinguishing characteristics.

Tab. 16-3 Properties of injection-molded maleated polyethylenecontaining different fillers [14][a].

Property 20% 25% CaCO3 CaCO3 talcglass mica 40% 40% 40%

uncoated coated

Tensile yield strength, MPa – – – – 29.4Tensile break strength, MPa 42.1 26.9 33.1 34.0 29.2Tensile yield elongation, % – – – – 3.0Tensile break elongation, % 4.8 3.7 8.2 7.8 3.4Flexural modulus, MPa 2340 2620 1850 2020 2290Flexural strength, MPa 51.3 40.7 38.2 39.5 37.4Izod impact strength notched, J m–1 80.1 58.7 80.1 58.7 58.7Izod impact strength unnotched, J m–1 272 208 422 315 251HDT at 1.82 MPa, °C 84 65 49 52 60

[a] Note: Fiber glass OCF457AA, Owens Corning;Suzorite mica 200H-K; CaCO3 Atomite/Mi-crowhite 25, ECC Intern; CaCO3 Kotamite,coated ECC Intern.; Talc, Microtalc MP1250,

Pfizer; Matrix, Polyolefin based on at least80% HDPE, modified with peroxide/maleicanhydride in a twin-screw extruder.


Tab. 16-2 Properties of injection-molded polyethyleneincorporating different fillers [14][a].

Property Unf illed 20% 25% 40% 40% 40% control glass mica CaCO3 CaCO3 talc

uncoated coated

Tensile yield strength, MPa 24.8 – – 23.4 20.8 26.9Tensile break strength, MPa 10.8 36.5 25.0 21.8 18.3 26.7Tensile yield elongation, % 9.2 – – 5.5 5.3 3.0Tensile break elongation, % 23 3.5 2.8 8.7 13 3.3Flexural modulus, MPa 1030 2620 2480 1800 1590 2380Flexural strength, MPa 30.3 49.4 36.5 33.9 30.6 35.0Izod impact strength notched, J m–1 48.0 53.4 58.7 48.0 53.4 37.4Izod impact strength unnotched, J m–1 1014 208 139 534 406 245HDT at 1.82 MPa, °C 47 79 65 49 50 53

[a] Note: Fiber glass OCF457AA chopped strands46 mm, Owens Corning Fiberglas Corp.; Su-zorite mica 200H-K, Suzorite mica Products;CaCO3 Atomite/Microwhite 25, uncoated ECC

Intern; CaCO3 Kotamite, coated ECC Intern.;Talc, Microtalc MP1250, Pfizer; Matrix, Poly-olefin based on at least 80% HDPE.

279

Significant applications, particularly in automotive (automotive interior, exterior,under the hood), household items, and furniture, exist for molded filled PP. Particlesize usually ranges from 3 to 10 µm at 20–40% filler content. Primary effects with re-spect to the unfilled resin are increased stiffness and heat distortion temperature, re-tention or improvement of impact strength, and improved dimensional stability.Shrinkage decreases with increasing concentration. CaCO3 also provides moreisotropic shrinkage and lower warpage than other filler shapes and may also act as acrystal nucleator in PP. A stearate-coated PCC ultrafine (mean particle size40–70 nm) grade has been shown to promote the formation of greater amounts of the

16.7 Applications

Fig. 16-1 SEM fracture surface of an injection-molded specimenshowing poor adhesion between the unmodified polyethylene ma-trix and untreated calcium carbonate [14].

Fig. 16-2 SEM fracture surface of an injection-molded specimenshowing improved adhesion between the maleated polyethylene ma-trix and untreated calcium carbonate [14].

280

β-crystalline phase, which has been reported to be tougher than the usually prevail-ing PP α-phase [15]. In PP sheet thermoforming, 30% filler incorporation increasesstiffness and can yield approximately the same levels of performance as PS or PVCwith the same shrinkage characteristics. In Figure 16-3, CaCO3-filled PP is comparedin terms of stiffness with several unfilled polymers and PP containing other fillers.In thermoforming, the presence of the filler can also increase line speeds due tomore rapid cooling by virtue of the increased thermal conductivity. PP molding com-pounds containing 10–40% CaCO3 are available from a plethora of compounders.

16.7.5Polyolef in Films

Calcium carbonate is used in polyolefin film applications where high clarity is not re-quired. It improves stiffness, dart impact resistance, and tear strength in LLDPE andHMW HDPE blown films and biaxially oriented PP (BOPP) films. This may allowdown gauging and lower material costs. In addition, it improves processing charac-teristics by increasing thermal conductivity, thus enabling the part to be cooled moreswiftly; it also reduces specific heat, allowing more rapid temperature increases andoverall increased line speeds and productivity [16,17]. CaCO3 levels of 20–35 wt. %may increase output by as much as 20–40%. Typical grades for LLDPE films are wet-ground, surface-treated materials with a median diameter of 1.4 µm added as a mas-terbatch.

An additional function of CaCO3 in polyolefin and other films (polyester and cel-lulose acetate) is its ability to increase surface roughness at low filler content and tominimize “blocking”, i.e., the tendency of adjacent film surfaces to stick together un-der pressure (see also Chapter 19). Examples of the use of CaCO3 as an anti-block inLLDPE can be found in refs. [1,6]. Related to its surface roughening capacity is theability of CaCO3 to improve printability in extrusion coating, allowing better f low ofthe printing ink on the film surface.

16.7.6Polyolef in Microporous Films

Microporous films have quickly developed, with a great majority of the product linesin infant and adult health care, construction, protective apparel, and medical sectorsconsuming large quantities of GCC [18]. Such films are manufactured from highlyfilled, non-porous precursor polyolefin films containing up to 60% GCC by stretch-ing on- or off-line [19]. Porosities down to 1 µm can be induced by interfacial debond-ing of the CaCO3 particles, which act as stress concentrators in the stretching step.Figure 16-4 shows the surface characteristics of an experimental LLDPE/60% CaCO3

stretched film with an average pore size of <2 µm [20]. The films are termed ‘breath-able” since, after stretching, an interconnected pathway is created for the transport ofwater vapor but not of liquid water.



Fig.

16-

3C

ompa

riso

n of

fle

xura

l mod

ulus

of C

aCO

3-fi

lled

PP w

ithun

fille

d an

d un

fille

d po

lym

ers;

to c

onve

rt p

si in

to P

a m

ultip

ly b

y68

95. (

Rep

lott

ed d

ata

from

CA

MPU

S da

taba

se c

ourt

esy

of D

r. S.

K.

Dey

, Son

oco

Cor

p.)

282

Manufacturing involves the following three steps:

extrusion compounding, usually in a TSE; film extrusion by casting or blowing; rapid stretching to thin films of thickness ≤ 25 µm, either uniaxially or biaxially, to

produce uniform microvoids; combination with a non-woven material to provide a cloth-like product with good

strength characteristics.

In order to provide maximum dispersion and deagglomeration at these high load-ings, extrusion compounding usually involves addition of the filler downstream afterthe polymer has fully melted. Appropriate selection of the screw configuration andthe removal of trapped air are important, as discussed in Chapter 3.

Important characteristics of the stretched, porous films are microstructure, me-chanical properties, porosity, and water vapor transmission rate, WVTR. Porosity andaverage pore size have been shown to correlate fairly well with WVTR and can be usedto optimize GCC grade characteristics (Figure 16-5). Important characteristics of theselected GCC grade are mean particle size (1–2 µm), particle size distribution, sur-face area, and stearic acid coating levels. Figure 16-6 shows comparative data of mod-ulus in the machine (MD) and transverse (TD) directions, along with WVTR data ofsix microporous LLDPE films containing different grades of GCC, all at 60 wt. %[20,21]. As expected, an increase in porosity, as manifested by increased WVTR, re-sults in decreased modulus. These effects depend significantly on surface area, par-ticle size, and particle size distribution, and less on stearic acid level in the studied1–2% range.


Fig. 16-4 Surface scanning electron micrograph of microporousstretched LLDPE/60% CaCO3 film [20].

283

16.7.7Bioactive Composites

Among the various fillers evaluated for bone regeneration, a pure form of vaterite hasbeen shown to promote bioactivity in degradable polymers (polylactic acid) throughapatite formation. Figure 22-4 in Chapter 22 shows the formation of an apatite struc-ture through contact with simulated body f luids. Bioactive fillers may be a very prom-ising area for expanded use of highly purified forms of calcium carbonate.

16.7 Applications

Pore Dia. Sitze @ Pop.Max., microns

Ave

.WV

TR

, gh–1

m–2

Fig. 16-5 WVTR versus pore diameter for microporous stretchedLLDPE/60% CaCO3 films containing different filler grades [20].

WVTR [gh–1m–2] vs. modulus [MPa]

Fig. 16-6 Average WVTR versus tensile modulus of microporousstretched LLDPE/60% CaCO3 films containing different filler grades[20].

284

References

1 Wypych, G., Handbook of Fillers, ChemTecPubl., Toronto, Ont., Canada, 2000, pp.48–58 and p. 800.

2 Katz, H. S., Milewski, J. V. (Eds.), Handbookof Fillers and Reinforcements for Plastics,Chapter 5, Van Nostrand Reinhold Co., NewYork, 1978.

3 Rothon, R. N., “Mineral Fillers in Thermo-plastics: Filler Manufacture and Characteri-zation” in Advances in Polymer Science, Vol.139, Springer-Verlag, Berlin, Heidelberg,1999, 67–107.

4 Specialty Minerals, Inc. (SMI), at www.min-eralstech.com

5 Solvay Chemicals U.S., “Coated Precipitat-ed Calcium Carbonate” MSDS No Win-nofil-1003, Revised 10-2003.

6 Zweifel, H., (Ed.), Plastics Additives Hand-book, Chapters 7 and 17, Hanser Publish-ers, Munich, 2001.

7 Lamond, T. G., Proc. Functional Fillers ‘95,Intertech Corp., Houston, TX, Dec. 1995.

8 Industrial Minerals Prices, Ind. Minerals,Feb. 2004, pp 72–73.

9 Materials Markets Consulting, “CalciumCarbonate Supply, 2001” at www.mineral-net.co.uk/consult_GCC.html

10 Wilson, I., Ind. Minerals, Febr. 2004,pp 40–45.

11 Crowe, G., Kummer, P. E., Plastics Com-pounding, Sept./Oct. 1978, 14–23.

12 Solvay Precipitated Calcium Carbonate,technical information at www.solvaypcc-com/market/application

13 Champine, N., Proc. Functional Fillers ‘95,Intertech Corp., Houston, TX, Dec. 1995.

14 Xanthos, M., et al., Polym. Compos. 1995,16(3), 204.

15 Kotek, J., et al., Eur. Polym. J. 2004, 40, 679.16 Hancock, M., Proc. Functional Fillers &

Fibers for Plastics ‘98, Intertech Corp., Bei-jing, P.R. China, June 1998.

17 Guy, A. R., Proc. Functional Fillers for Plas-tics 2003, Intertech Corp., Atlanta, GA, Oct.2003.

18 Moreiras, G., “GCC in Microporous Films”,Proc. 3rd Minerals in Compounding Conf.,Cologne, Germany, April 2001; Ind. Miner-als, July 2001; also available atwww.omya.com

19 Clemensen, P. D., Proc. Functional Fillers forPlastics 2001, Intertech Corp., San Antonio,TX, Sept. 2001.

20 Zhang, Q., et al., Proc. 60th SPE ANTEC,2002, 48, 2840.

21 Xanthos, M., Wu, J., Proc. 62nd SPE AN-TEC, 2004, 50, 2641.


17Fire Retardants

Henry C. Ashton

17.1Introduction

The chemistry and dynamics of combustion have been well described in many pub-lications, but for the purposes of this chapter a short overview is helpful. Combustionmay be defined as the rapid uncontrolled reaction of oxygen with a substrate. A fea-ture of this process is that it is generally exothermic and as a consequence of the heatof combustion the reaction of the fuel with oxygen is accelerated resulting in a self-propagating reaction. While the reaction can be viewed as an extreme case of oxida-tion, the process is more difficult to control, and as with thermal oxidative degrada-tion of polymers, once the process is started it is extremely difficult to control or stop.Thus, any fire-retardant strategy should focus on the prevention of initiation of acombustion reaction or intervene rapidly if combustion has been initiated.

This chapter is focussed on the combustion of polymeric materials and the strate-gies used to counteract fire. In many applications, such as electrical, electronic, trans-port, building, etc., the use of polymers is restricted by their f lammability despite theother advantages that their use brings.

17.2Combustion of Polymers and the Combustion Cycle

The increasing diffusion of synthetic polymers into the marketplace has greatly in-creased the “fire risk” and the “fire hazard”, that is, respectively, the probability of fireoccurrence and its consequence either on humans or on structures.

To enable effective use of polymers, f lame-retardant materials have to be added topolymer-based formulations. The role of these additives is to:

slow down polymer combustion and degradation (fire extinction), reduce smoke emission, avoid dripping of hot or burning material.


286

A key safety element of a fire situation is to allow people the maximum amount oftime possible to escape a fire. Fire safety regulations are developed with this factor inmind. The stringency of the regulations often depends on the time needed to safelyescape the environment of a fire.

As shown in Figure 17-1, the three required elements to sustain combustion are asource of fuel, heat, and oxygen. Removal of any one of these elements from the cy-cle can prevent combustion. It takes heat to initiate combustion (∆H 1), and oncecombustion begins heat is liberated (∆H 2). This will correspond to the total heat re-leased, as measured by cone calorimetry (see Section 17.8).

Several techniques are available whereby this combustion cycle can be interrupted.Among the materials that can be used as fire-retarding agents, some function as ab-sorbers of heat, thus lowering the ambient temperature below the critical tempera-ture for ignition. Others function by providing a vapor barrier that dilutes the con-centration of the incoming oxygen required to feed the fire and thus slow down thekinetics of combustion. Another approach offering a barrier to combustion is to usematerials that will cross-link or intumesce, providing a networked structure that pre-vents the ingress of oxygen to the site of combustion.

Dilution of the available fuel (polymer system) can be accomplished by the use ofmaterials such as inorganic fillers that allow for the use of diminished amounts ofthe more combustible organic polymers. Another class of fire retardants comprisesspecies that can interact with oxygen to form molecular species that in effect dilutethe free radical character of oxygen and render it less reactive to the fuel source.

Physical characteristics of materials as well as chemical composition inf luencecombustion behavior. Factors such as surface area, availability of air (oxygen) to im-pinge on the site of ignition, heat transfer to and from the f lame, inter alia, can in-f luence the rate and extent of combustion.

17 Fire Retardants

Fig. 17-1 The combustion cycle.

287

17.3Fuel

There is a wide and structurally complex variety of polymers. However, all of themwill burn under appropriate conditions, reacting with oxygen from the air, releasingheat and generating combustion products. The oxidation reaction in a f lame takesplace in the gas phase. Hence, for liquids or solids to burn there must be conversionto gaseous species. In the case of liquids, this can be accomplished by evaporation orby boiling from the liquid surface. For most solids, decomposition by heating (pyrol-ysis) is necessary to afford products that have a molecular weight that is sufficientlylow for volatilization. This requires that the surface temperature of the solid be suffi-ciently high to generate species with energies higher than those required for evapo-ration alone. Typically, the surface temperature of burning polymeric materials hasto be close to the temperatures at which carbon–carbon and carbon–hydrogen bondscission will take place, i.e. at approximately 400 °C [1,2].

17.4Smoke

Perhaps the best working definition of smoke is given by the National Fire ProtectionAssociation, in NFPA 92B: “The airborne solid and liquid particulates and gases evolvedwhen a material undergoes pyrolysis or combustion together with the quantity of air that isentrained or otherwise mixed into the mass”.

Ideally, the combustion of a linear hydrocarbon can be represented by the equation:

CnH2n+2 + [(3n + 1)/2]O2 → nCO2 + (n + 1)H2O

Under these conditions, the only products expected are carbon dioxide and water.However, in most cases, combustion involves the generation of more than thesespecies as products. In cases of incomplete combustion the formation of carbonmonoxide often occurs with the concomitant formation of lower molecular weight oroligomeric species.

In general, the volatile components make a complex mixture, the composition ofwhich ranges from simple molecules such as hydrogen and low molecular weight hy-drocarbons (e.g. ethylene) to higher molecular weight species that can only volatilizeat higher temperatures.

Smoke particles are mainly of two kinds:

1. Carbonaceous solid particles that produce black smoke, often called soot.2. Liquid droplets that form as some gas molecules cool and condense producing

light colored smoke.

The droplets described above may also contain some finely divided solid materialssuch as minerals originally present in the combusting material; gases can also be en-

17.4 Smoke

288

trained. The relative amounts of solids, liquids, and gases depend on the ambientconditions, particularly the temperature. There are two characteristics of smoke thatdeserve mention, namely toxicity and visibility.

17.4.1Toxicity

Every year approximately 48,000 people die from the toxic effects of smoke in theUnited States [3]. Carbon monoxide is recognized as being the major intoxicant inmany cases. Carbon monoxide converts blood hemoglobin to carboxyhemoglobinand in so doing prevents the formation of the oxygen complex oxyhemoglobin that isrequired for cellular metabolism. Exposure to carbon monoxide at a level of 2% in aircan incapacitate a person in two minutes, and unless there is immediate medical at-tention it is likely that death will follow in a matter of minutes [4]. Some attempts havebeen made to quantitatively predict the toxic effects of fire, but overall accurate pre-dictions are difficult [5–7].

Another intoxicant is hydrogen cyanide, which is found in the combustion prod-ucts of nitrogen-containing polymers, including wool, acrylonitrile, polyurethanes,polyamides, and urea–formaldehyde resins. This is estimated to be 10 to 40 timesmore toxic than carbon monoxide. Hydrogen cyanide functions by rendering cells in-capable of accepting oxygen from oxyhemoglobin. Levels below 1 ppm in humanblood can lead to toxicity.

Carbon dioxide is not highly toxic but exposure to it can render the effects of othertoxic agents more severe as it causes increased breathing rates as a result of depletionof the available oxygen in the air. In extreme cases of exposure it can cause suffoca-tion. Other toxic agents that deserve mention are hydrogen chloride and acrolein. Hy-drogen chloride is a severe irritant and causes lung damage. Acrolein, which is acombustion product of cotton, wood, and paper, is also a severe irritant and pro-longed exposure (20 minutes or more) can be lethal. A full treatment of the toxic ef-fects of products of combustion has been given by Purser [8].

17.4.2Visibility

The most common optical method for measuring the density of smoke particles is tomake use of the Beer–Lambert Law, which is described by the following equations:

I/I0 = exp(–αlc) and A = log (I/I0) (17-1)

where the absorption coefficient α = 2.303ε, A = absorbance, ε = molar absorptivi-ty, I0 = intensity of the incident light, I = intensity of light after passing through thematerial, l = pathlength of light, c = concentration of absorbing species. The extinc-tion coefficient, defined as the fraction of light lost to scattering and absorption per

17 Fire Retardants

289

unit distance in a participating medium, is a measure of the opacity of the smoke andcan be calculated using the above relationships.

Optical methods are favored for measuring smoke density because they provide in-formation directly relevant to visibility through smoke. Other methods involve weigh-ing particulates and calculating the weight loss on combustion relative to the weightloss of the combustible. This method correlates with smoke measurements by opti-cal methods [9–11].

17.5Flammability of Polymers

The f lammability of polymers can be determined by their reactivity with oxygen. Thelimiting oxygen index (LOI), defined as the minimum amount of oxygen required tosustain combustion under specified conditions, is a quantitative measure of the ten-dency of materials to burn (Table 17-1). In a f lame-retarded system, study of the LOIcan provide information on the effectiveness of f lame-retardant materials.

Tab. 17-1 Limiting oxygen index (LOI) values for selected polymers [12].

Material LOI Material LOI

Polyoxymethylene 15.7 Neoprene 26.3–40Polyurethane foam (f lexible) 16.5 Nomex 26.7–28.5Natural rubber foam 17.2 Modacrylic fibers 26.8Polymethylmethacrylate 17.3 Leather 34.8Polyethylene 17.4 Phenol–formaldehyde resin 35Polypropylene 17.4 PVC 37–42Polyacrylonitrile 18 Polyvinylidene chloride 60ABS 18.3–18.8 Polytetraf luoroethylene 95Cellulose 19 Polystyrene 18–19Nylon 20.1–26 Styrene–acrylonitrile 19Wood (birch) 20.5 Polyethylene terephthalate 20–23Polycarbonate 22.5–28 Epoxy 21–25Wood (red oak) 23 Polyvinylidene f luoride 44Wool 23.8 Fluorinated ethylene propylene copolymer >95

Depending on the polymer and the applicable fire retardance standards, f lame re-tardants are chosen to interfere with one or more stages of the combustion process:heating, decomposition, ignition, f lame spread, smoke density. Fire retardants haveto inhibit or even suppress the combustion process.

Fire retardants can be divided into halogen-containing and non-halogen-contain-ing. This is a useful categorization as it leads into an explanation of their differentmodes of actions. Chief among the non-halogen fire retardants are aluminum trihy-drate (ATH) and magnesium hydroxide (MOH), as well as materials such as

17.5 Flammability of Polymers

290

melamine and derivatives, ammonium polyphosphates, antimony oxide, and boron-containing materials such as zinc borate.

17.6Mechanisms of Fire Retardant Action

Flame retardants can act chemically and/or physically in the condensed phase and/orin the gas phase. In reality, combustion is a complex process occurring through si-multaneous multiple paths that involve competing chemical reactions. Heat pro-duces f lammable gases from pyrolysis of the polymer and if the required ratio be-tween these gases and oxygen is attained ignition and combustion of the polymer willtake place.

17.6.1Condensed Phase

In the condensed phase, three types of processes can take place:

1. Breakdown of the polymer, which can be accelerated by f lame retardants, leads toits pronounced f low which decreases the impact of the f lame.

2. Flame retardants can leave a layer of carbon (charring) on the polymer’s surface.This occurs, for example, through the dehydrating action of the f lame retardantgenerating double bonds in the polymer. These processes form a carbonaceouslayer as a result of cyclization and cross-linking (see Figure 17-2).

3. Heat absorption through materials such as ATH that have a very high heat capac-ity.

Flame-retarding polymers by intumescence is essentially a special case of a con-densed phase activity without apparent involvement of radical trap mechanisms inthe gaseous phase. Intumescence involves an increase in volume of the burning sub-strate as a result of network or char formation. This char serves as a barrier to theingress of oxygen to the fuel and also as a medium through which heat can be dissi-pated (Figure 17-2).

In intumescence, the amount of fuel produced is also greatly diminished and charrather than combustible gases is formed. The char constitutes a two-way barrier, bothfor hindering the passage of combustible gases and molten polymer to the f lame aswell as for shielding the polymer from the heat of the f lame. Many intumescent sys-tems have been developed in the past 25 years. These can be generally formulated toconsist of three basic ingredients:

a “catalyst” (acid source) promoting charring, a charring agent, a blowing agent (spumific compound).

17 Fire Retardants

291

The catalyst or acid source can consist of ammonium phosphate or polyphosphatesalts, phosphoric acid derived amides, or alkyl or haloalkyl phosphates. Charringagents are based on molecular structures that can form cross-linked networks suchas pentaerythritol, sorbitol, melamine, and phenol–formaldehyde resins. Other poly-meric systems capable of intumescence are some polyamides and polyurethanes.Blowing agents help to form a porous structure in the char and can facilitate its for-mation. Common blowing agents are based on urea and urea–formaldehyde resins,melamines, and polyamides that can liberate moisture.

17.6.2Chemical Effects in the Gas Phase

In the gas phase, the process of combustion is slowed by reactive species that inter-fere chemically with the propagation process of the fire. The f lame retardants them-selves or species derived from them interfere with the free radical mechanism of thecombustion process. This slows or stops the exothermic processes that occur in thegas phase and results in a cooling of the system and a reduction in the supply of f lam-mable gases. Hydrogen halides, HX (X = Br or Cl), produced by the reaction of halo-genated organic compounds, R–X, with a polymer, P–H, can react with the excited-state HO· and H· radicals to produce the less reactive halogen free radicals X· lead-ing to an overall decrease in the kinetics of the combustion as shown below:

R –X + P–H → HX + RPHX + H·→ H2 + X·HX + OH· → H2O + X·

Table 17-2 shows the components of polymer combustion. In the presence of halo-genated fire retardants, the active radical species such as OH·, O·, and H· can bequenched in the gas phase to form species such as H2O, H2, and HX, which are rel-atively less reactive in the combustion cycle.

A schematic summary of the mechanism of action of f lame retardants is given inFigure 17-3. The relationship between condensed-phase and gas-phase processes is

17.6 Mechanisms of Fire Retardant Action

Fig. 17-2 Char and intumescence formation.

292

indicated by referring to specific f lame retardants. The inf luence of both chemicaland physical processes should also be noted.

Tab. 17-2 Components of polymer combustion.

Gas-phase components of polymer comustion

Ever-increasing concentrations of the The excited-state free radicals have been excited-state reactive OH*, O*, and H* removed and rendered unavailable to the com-free radical species bustion cycleCO2 OH* O2 O* CO OH H* CO2 CO HOH HX HX HHPolymer without RX flame retardant Polymer with flame retardant RX

17.7Classif ication of Fire Retardants

17.7.1Metal Hydroxides

Metal hydroxides, particularly aluminum trihydrate and magnesium hydroxide, con-tribute to several fire-retardant (FR) actions. They first decompose endothermicallyand release water. The endothermic decomposition serves to remove heat from thesurroundings of the f lame and thus cool the f lame. This is often referred to as the“heat sink” phenomenon. Pyrolysis decreases in the condensed phase as a result ofthis action. The release of water dilutes the amount of oxygen capable of ingress tothe f lame and avoids the critical fuel/oxygen ratio (physical action in the gas phase).Both mechanisms combat ignition. In some fire tests used for electric/electronic ca-

17 Fire Retardants

Fig. 17-3 Chemical mechanism of f lame retardance.

293

ble applications, the ignition has to be significantly delayed. Thus, metal hydroxidesare suitable for these applications. Moreover, after the degradation, a ceramic-basedprotective layer is created, which improves insulation (physical action in the con-densed phase) and gives rise to a smoke suppressant effect (chemical action in thecondensed phase). The ceramic-based protective layer ensures efficient protection ofthe polymer during combustion, leading to a severe decrease in the heat released.

17.7.1.1 Aluminum TrihydrateAluminum trihydrate is well known as the largest volume f lame retardant used in theworld, with a consumption of around 200,000 tons per annum, representing 43% ofall f lame-retardant chemicals in volume (about 29% in value). Grades available in-clude general purpose (particle sizes of 5 to 80 µm), high loading (5 to 55 µm), su-perfine (controlled surface areas from 4–11 m2 g–1), ultrafine (15–35 m2 g–1), and lowelectrolyte level. Also known as ATH or “hydrated alumina”, it is technically alu-minum hydroxide, with the chemical formula Al(OH)3. The term “hydrate” becamepart of the common name because chemically combined water is released during itsdecomposition in fire.

ATH came into wide use as a f lame retardant in the 1960s, primarily as a result ofdemands of consumer driven safety legislation for carpet backing and fiberglass-re-inforced polyester products. For these applications, ATH imparts f lame retardanceand smoke suppression. End use markets for ATH include transportation, construc-tion, cast polymers, electrical/electronic, wire and cable, leisure, and appliances. Spe-cific polymer applications range from thermosets such as “solid surface”, sheet-molding compounds (SMC), and bulk-molding compounds (BMC), to wire and cableand other thermoplastic applications, for which it is compounded with polyvinyl chlo-ride (PVC), polyolefins, or ethylene propylene diene (EPDM) rubber.

Thermodynamic properties and f lame retardanceAt room temperature, ATH is very stable but when the temperature exceeds 205 °C,as could happen in a fire situation, it begins to undergo an endothermic decomposi-tion with a heat of reaction of –298 kJ mol–1. This decomposition, as shown below, iskinetically slow between 205 °C and 220 °C, but above 220 °C (see Figure 17-4 com-paring ATH and MOH) it becomes very rapid:

2 Al(OH)3 → Al2O3 + 3H2O

Water that is released during this decomposition dilutes the gases that feed com-bustion. This dilution slows down the rate of polymer combustion by formation of avapor barrier that prevents oxygen from reaching the f lame. The combustion of poly-mers is also retarded since ATH acts as a heat sink (heat capacity of Al2O3·3H2O =186.1 J K–1 mol–1) and absorbs a portion of the heat of combustion. It is also believedthat aluminum oxide (Al2O3) formed during this decomposition aids in the forma-tion of an insulating barrier on the surface of the burning polymer, which acts to in-sulate the polymer from fire.

17.7 Classif ication of Fire Retardants

294

Ground versus precipitated ATHATH can be produced in a variety of particle sizes and particle size distributions thatare controlled by grinding or precipitation processes. Both precipitated and groundATH products can be used as long as processing temperatures stay below the onsetof decomposition temperature of 200 °C. Mechanically ground ATH has a wide par-ticle size distribution and a better packing fraction than precipitated ATH and thiscontributes to less dusting, faster incorporation into the polymer matrix, and lowercompound viscosity (by as much as 50%) compared to precipitated grades. Table 17-3 shows a chemical analysis of wet ground and precipitated ATH.

Tab. 17-3 Chemical analysis of wet-ground and precipitated ATH.

Parameter Precipitated ATH Wet-ground ATH

Al(OH)3, % 99.2 99.2SiO2, % 0.05 0.005Fe2O3, % 0.035 0.007Na2O, % soluble 0.01 0.005

The most significant difference between the two grades is in their morphology, asillustrated in Figure 17-5. Precipitated ATH, which is allowed more time to crystal-lize, displays a hexagonal platelet morphology and a more homogeneous particle sizedistribution. This can be a benefit in some electrical insulation applications, wheremineral homogeneity can inf luence dielectric properties. However, this homogene-ity may be offset by difficulty in processing compared with ground ATH; in the latter,

17 Fire Retardants

Fig. 17-4 Comparison of the thermal degradation properties ofmagnesium hydroxide and aluminum trihydrate.

295

packing fractions are controlled allowing for ease of processing and, thus, an im-proved distribution of the mineral in the polymer matrix.

The relative ease of processing of ground ATH can allow higher levels to be incor-porated into a polymer composite and this in turn enables FR standards requiringhigh loading levels to be more easily attained. The relative ease of incorporation andprocessing of the ground ATH compared to the precipitated product is manifested bylower viscosities: for example, in a 55% ATH-loaded EPDM compound Mooney vis-cosities were measured as 28 and 52, respectively, for the wet-ground and precipitat-ed grades.

A comparison of the combustion properties of EPDM composites containing wet-ground and precipitated ATH (Table 17-4) shows significant differences. It is likelythat the lower average heat release rate (HRR) and average mass loss rate for the wet-ground ATH composite may derive from a more homogeneous distribution of theATH in the polymer. This is a good example of the inf luence of the physical proper-ties of a fire-retarding material on the performance in combustion testing.

Tab. 17-4 Cone calorimetric data comparison of wet-ground andprecipitated ATH in EPDM.

Cone calorimetry data @ 35 kW m–2 Precipitated ATH Wet-ground ATH

Peak HRR, kW m–2 164 111Peak HHR, s 489 272Average HHR, kW m–2 123 84Total heat, mJ m–2 80 85Average effective heat of combustion, MJ kg–1 29 34Average mass loss rate, g s–1 m–2 4.4 2.9


Fig. 17-5 Scanning electron micrographs of ground (20,000×) andprecipitated ATH (30,000×).

296

ApplicationsIn reinforced thermoset applications, the principal properties that ATH imparts aref lame retardance, smoke suppression, thermal conductivity, optical translucency,and chemical stability. Aside from the benefits imparted in the above properties, theuse of ATH as a resin extender also affords lower cost formulations with an accept-able balance of performance properties.

Finely divided ATH or MOH produced in precipitation processes or finely groundand surface-treated ATH grades can be used in wire and cable applications, most no-tably in PVC and EPDM. The FR tests that are relevant to this application areUL94/horizontal burn, VW-1, UL-1685, UL-1666, and UL-910. ATH has the followingfire-retarding effects:

cooling of the polymer and reduction of pyrolysis products; insulation of the substrate through a combination of aluminum oxide (derived

from ATH when it loses water) with the char formed; dilution of combustion gases by release of water vapor.

It is furthermore:

unleachable owing to its insolubility in water; non-toxic; non-corrosive.

So-called “Solid Surface” is a solid, non-porous surfacing composite material de-rived from ATH and a polyester or acrylic resin that is cured either thermally or atroom temperature. The color and pattern run throughout its thickness with a soft,deep translucence and a natural feeling of warmth. “Solid Surface” resists heat,stains, mildew, and impact. Incidental damage to the surface can be easily repaired tomaintain the appearance. It can be easily cut, routed, and shaped so that fabricatorscan create distinctive designs and customized patterns. As a result of the high load-ing levels of ATH, which often exceed 60 wt. %, the material easily resists fire even ifit is contacted with red-hot metal. For this reason the material finds use in kitchensand homes as counter tops and bathroom fixtures.

Surface-Treated ATHTo enhance the overall performance of ATH-filled compounds, a wide selection ofchemical modifications of the ATH surface can be carried out with surfactants,stearates, and organofunctional silanes. As described in Chapters 4, 5, and 6, thesechemical treatments can aid processing and improve mechanical properties, chemi-cal resistance, electrical performance, and f lame retardance. Although many of thesurface treatments used in ATH or MOH are organic materials that contain com-bustible functional groups, the effect of these treatments on FR properties is usuallynegligible. In fact, as a consequence of the increased loading levels and improved dis-persion that these treatments afford as compared to the non-treated counterparts, theFR performance of composites is often enhanced by using treated ATH or MOH.

17 Fire Retardants

297

Table 17-5 shows the various surface treatments that are suitable for a variety ofpolymers and the associated impact on end-use properties. Surface treatments can beclassified under three broad headings:

1. encapsulating & non-reactive,2. reactive with the ATH but non-coupling to the polymer matrix,3. reactive and coupling [13].

Tab. 17-5 Chemical surface treatments applied to ATH for specificpolymers.

Surface modif ication Polymer type Properties affected

AX 5102, a blend of anionic Epoxy, phenolic, polyolefins, Lower viscosity; improved dis-and non-ionic surfactants urethane persion; higher loading

potential; improved surface profile

ST, proprietary surfactant Latex, polyester, polyolefins, Lower viscosity; improved sus-urethane pension; better air release

properties; faster dispersion; improved wet-out

SL, isostearic acid EVA, polyolefins Improved processing and dis-persion; higher loading potential

Hyf lex silane, Y-5889 OSi Epoxy polyester, urethane Lower viscosity; higher loading product, ethoxylated potential; faster dispersion; propylsilane improved physical properties;

high polymer compatibility

SP, RC-1 (vinylsilane + Acrylic, EPDM, EVA, neo- Improved mechanical proper-methylsilane oligomer mix) prene, polyester, polyolefins, ties; better processing; in-

PVC, SBR creased water resistance

Isobutyl(trimethoxy)silane Epoxy, polyester, urethane Lower viscosity; higher loading potential; faster dispersion; enhanced physical properties; high polymer compatibility

SA, AMEO 3-amino- EVA, nitrile, phenolic, PVC, Improved mechanical proper-(propyl)triethoxysilane urethane ties and processing

SM, A-189, 3-mercapto- EPDM, neoprene, nitrile Greater abrasion resistance; (propyl)trimethoxysilane higher mechanical properties;

improved processing

SE, A-187, 3-glycidoxy- Epoxy, phenolic Improved processing; improved (propyl)trimethoxysilane mechanical properties; im-

proved water resistance

SH, A-174, methacryloxy- Acrylic, EPDM, EVA, Improved mechanical proper-(propyl)trimethoxysilane polyester, polyolefins, SBR ties and processing; improved

water resistance


298

It is clearly evident that the end-use properties are significantly inf luenced by thechoice of ATH and the chemical surface treatment. The ATH products of the futurewill include new combinations of morphology and surface treatments as a way to op-timize the desired combination of end-use properties. Table 17-6 shows a survey ofsurface treatment options for ATH in wire and cable applications using a variety ofpolymers where f lame retardance and smoke suppression are required.

Tab. 17-6 ATH surface treatments used in a variety of polymers.

Polymer Stearic acid (SL) Aminosilane (SA) Vinylsilane (SP) Proprietary silane

XL-EVA + + + +TP-EVA + + +XL-polyolefin + + + +TP-polyolefin + + + +NylonPVC + +

SummaryATH is used in a variety of polymers in different areas of application. The productsof the future will be tailored to conform to specific end-use requirements of the cus-tomer. This will be done by simultaneous control of morphology and optimization ofsurface treatments as well as optimization of the end-use formulations. Users of ATHwill continue to seek lower processing costs through ease of processing and polymercompatibility, which can be achieved and optimized by the methods outlined above[14–16]. Table 17-7 summarizes the various f lammability and electrical property stan-dards where ATH plays a major role.

Tab. 17-7 Standards met with ATH use.

Standard Description Target

UL94 vertical V0ASTM E84 (UL 723) tunnel test < 450 smoke, < 25 f lame spreadASTM D495 arc resistance > 200 secondsASTM E662 NBS smoke chamber < 200 @ 4 minutesASTM E1354 cone calorimeterASTM E648 radiant panel critical heat f lux of f loor coverings

using a radiant panelASTM D635 horizontal burnASTM D2863 limiting oxygen indexASTM D149 dielectric strengthASTM D150 dielectric constantASTM D257 volume resistivityASTM E906 OSU testANSI Z124 torch testMVSS-302 auto interior f lammabilityUBC 17-5 corner room burn testCAL 133 contact furniture f lammability

17 Fire Retardants

299

17.7.1.2 Magnesium HydroxideMagnesium hydroxide functions in a manner similar to ATH. It liberates water ofcrystallization in a fire situation and in so doing it forms a barrier to the ingress ofoxygen required for combustion. The material is also capable of absorbing heat, al-though the heat capacity value of 77.0 J K–1 mol–1 (crystalline) is lower than that forATH.

Magnesium hydroxide is produced by extraction from ores, such as magnesite,dolomite, or serpentinite, and from brine/seawater. The material used as a f lame re-tardant is generally of high purity (>98.5%). It is most often obtained from seawa-ter/brine, although other ore-derived products can also be very pure. Three majorprocesses relating to magnesium hydroxide are isolation from seawater/brine; theAman process, which involves thermal decomposition of brine to yield a high puritymagnesia that is converted to magnesium hydroxide; and the Magnifin® process, inwhich magnesium oxide is isolated from serpentinite ore and converted to magne-sium hydroxide.

Magnesium hydroxide is a white powder. Median particle sizes range from 0.5 to5 µm and specific surface areas typically range from 7 to 15 m2 g–1, depending on sizeand morphology. It is used at high loading levels, usually in the range 40–65 wt. %.A key difference compared to ATH is the higher decomposition temperature of ca.320 °C (Figure 17-4). This affords an advantage over ATH in that it may be used incommodity and engineering thermoplastics as well as thermoset resin systemsprocessed at temperatures above 200 °C. From a chemical standpoint, magnesiumhydroxide is a much more alkaline material than ATH. Since slurries in water (~5%w/w) can yield pH values of 10.5 or more, this property may need to be considered insome formulations.

Magnesium hydroxide finds use in applications such as wire and cable, appliancehousings, construction laminates, roofing, piping, and electrical components. Dif-ferent grades are developed for different applications. The use of surface-treatedgrades is increasing as market demands move towards increased product and per-formance differentiation [17,18].

17.7.2Halogenated Fire Retardants

17.7.2.1 Mechanism of ActionThe chemical mode of action of halogenated fire retardants has been described pre-viously in Section 17.6.2. The mode of action depends on the reaction of halogen-based free radicals with excited-state fire-propagating components in the gas phaseresulting in a lower system temperature and a decreased reaction rate for the fire-sus-taining chemical processes.

Brominated and chlorinated organic compounds are generally used as fire retar-dants. Iodine-containing materials are avoided as they tend to be less stable than theirchlorinated or brominated counterparts. The choice of retardant depends on severalfactors, such as the polymer type, the behavior of the halogenated fire retardant un-der the processing conditions (stability, melting, distribution, etc.), and/or its effect


300

on the properties and long-term stability of the resulting material. It is advantageousto use an additive that introduces a halogen species into the f lame in the same tem-perature range as that in which the polymer degrades into volatile combustible prod-ucts. Then, the fuel and inhibitor would both reach the gas phase according to the“right place at the right time” principle.

The most effective fire-retardant (FR) polymeric materials are halogen-based poly-mers (e.g., PVC, chlorinated PVC, polyvinylidene f luoride (PVDF)) and additives(e.g., chlorinated paraffins (CP), tetrabromobisphenol A (TBBA)). However, the im-provement of fire performance depends on the type of fire test, i.e. the application.

Halogenated organic fire retardants may be classified as additives rather than asfunctional fillers. Many are liquids, others are high molecular weight polymers, andothers exhibit a certain degree of miscibility with the polymer matrix; thus, they donot fall within this book’s scope of solid (mostly inorganic) fillers forming a distinctdispersed phase. They are, however, brief ly covered since they are commonly usedsynergistically in combination with more traditional inorganic fillers such as anti-mony trioxide and molybdenum oxide, or occasionally with aluminum trihydrate.

17.7.2.2 Synergy with Antimony TrioxideAntimony trioxide (Sb2O3) is often used as a synergist in combination with halo-genated fire retardants. For efficient action, trapping of excited-state free radicalsneeds to occur in the gas phase in the f lame and hence the radical trapping agentneeds to reach the f lame in the gas phase. The addition of antimony trioxide allowsthe formation of volatile antimony species (halides or oxyhalide), which interrupt thecombustion process by inhibiting H* radicals through the series of reactions shownin Figure 17-6. This mechanism is an illustration of the synergy between halogenat-ed f lame retardants and the antimony trioxide.

Loading levels of antimony trioxide can often be quite high and in some cases thishas led to toxicity concerns.

17.7.2.3 Bromine-Containing Fire RetardantsBromine-containing fire retardants may be used in many polymeric applications aseffective gas-phase f lame retardants. They may be broadly classified into aliphaticand aromatic materials. In general, the aliphatic materials are more effective f lameretardants by virtue of their lower decomposition temperatures than the aromaticbromine-containing FRs, which have a high degree of thermal stability. In some sys-tems, the decreased chemical lability of the aromatic systems can allow for action overa longer time cycle. This can be advantageous in the case of longer burning fires or

17 Fire Retardants

Fig. 17-6 Mechanism of action of anti-mony trioxide as a f lame retardant [11].

301

in cases where a fire may have taken hold and the more kinetically labile retardantshave become exhausted [19].

The mechanism of action of halogenated f lame retardants has been outlined inTable 17-2 and Figure 17-3. These materials function mainly in the gas phase by theliberation of volatile bromine compounds, such as hydrogen bromide, HBr, which in-hibit the reactions that propagate the f lame. Hydrogen bromide reacts readily withexcited-state radicals of oxygen, hydroxyl, and hydrogen to generate molecular speciesthat have little or no tendency to propagate the f lame. One important reaction in thisprocess is that of HBr with HO• to form H2OBr•. This prevents the reaction of HO•

with CO to form CO2, which is a highly exothermic process and assists the propaga-tion of combustion.

A wide range of brominated f lame retardants is available and the choice of mate-rial is often made with consideration of the polymer processing operations and con-ditions. Ideally, the f lame-retardant material should remain stable during the normalprocessing operations of the polymer. Thermogravimetric weight loss studies are of-ten used to determine the stability of many polymer additives, including f lame retar-dants, over the projected range of processing conditions. In addition to low MWcompounds (see the examples in Table 17-8), higher MW materials such as bromi-nated polystyrene (MW 200,000) are used when non-blooming characteristics are de-sirable. Brominated epoxy fire retardants are also available as oligomers with molec-ular weights in the range of 1600–3600 g mol–1 or as more defined polymers with mo-lecular weights in the range of 10,000–60,0000 g mol–1. These materials can be ter-minated with either epoxy groups or with brominated aromatic rings. Theepoxy-terminated materials are suitable for high-end engineering polymer applica-tions providing high thermal and UV stability, do not bloom, and are easily processed.The brominated phenyl group terminated polymers are used in HIPS and ABS,where they provide good processability with low migration; high thermal and UV sta-bility; good HDT and impact strength, and low metal adhesion. The choice of bromi-nated epoxy depends upon the molecular weight, which correlates well with the on-set of softening. Generally, the lower the molecular weight the greater the ease of in-corporation into polymers. However, this benefit can be offset by greater mobility ofthe additive.

17.7.2.4 Chlorine-Containing Flame RetardantsThe use of chlorinated additives as f lame retardants is often in conjunction with oth-er materials, such as phosphorus- or antimony-containing additives. The choice of achlorinated additive is dependent on the stability of the material, its volatility, its chlo-rine content, and the kinetics of the reaction in the gas phase during combustion.The use of such materials, however, is decreasing owing to the toxicity of polychlori-nated biphenyls (PCBs) and products with related molecular structures.

Examples of chlorinated additives include chlorinated paraffins derived from thechlorination of petroleum distillates and represented by the general formulaCnH2n+2–mClm. They typically contain chlorine in the range of 30 to 70% by weightand range in length from 10 to 30 carbon units and they may be liquid or solid. Thesematerials function by releasing hydrogen chloride, which reacts in the gas phase to


302

decrease the concentration of the reactive hydroxyl, oxygen, or hydrogen radicalspecies that propagate the f lame. The main application of chlorinated resins is asplasticizers for f lexible PVC in combination with dioctyl phthalate or diisononyl ph-thalate, where they serve to improve f lame-retardant properties in applications suchas f looring and cables. Solid grades with high chlorine content are used in thermo-plastics such as low density polyethylene (LDPE) in cable jacketing in combinationwith antimony trioxide.

Other examples of chlorinated additives include chlorinated alkyl phosphates, withmain applications in rigid and f lexible polyurethane foams, and chlorinated cy-cloaliphatics such as dodecachlorodimethanodibenzocyclooctane. The latter is usedwith various synergists, such as antimony trioxide and zinc borate, in numerous poly-mers including polyamide, polyolefins, and polypropylene.

17 Fire Retardants

Tab. 17-8 Examples of aliphatic and aromatic brominated f lame retardants.

Material CAS no. MW Br Melting Advantages Applicationcontent point, °C(%)

AliphaticTrisbromoneopentyl 19186-97-1 1018 67 230 UV, heat PP, ABSphosphate stability

Trisbromoneopentyl 36483-57-5 325 73 65 Low leaching, Poly-alcohol heat stability urethanes

Dibromoneopentyl 3296-90-0 262 60 110 High FR perf. UPEsglycol transparency

Hexabromocyclo- 25637-99-4 642 73 180 High purity and EPS, XPSdodecane efficiency

AromaticTris(bromophenyl)- 25713-60-4 1067 67 230 Non-blooming PE, ABS triazine epoxy

Tetrabromobisphenol 21850-44-2 943 68 115 High efficiency PP, ABS, PSA bis(2,3-dibromo-propyl ether)

Octabromodiphenyl 32536-52-0 801 78 120–185 High surface ABS, oxide quality styrenic

copolymers

Decabromodiphenyl 1163-19-5 959 83 305 Excellent General oxide thermal stability purpose

Tetrabromobis- 79-94-7 544 58 181 High reactivity PC, epoxy, phenol A and efficiency phenolic,

ABS

Brominated triethy- 155613-93-7 857 73 240–255 Good f low and PA, phenyl indane impact styrenics

Pentabromotoluene 87-83-2 487 82 299 PE, PP, rubber

303

17.7.3Zinc/Boron Systems

Zinc borate is a boron-based fire retardant available as a fine powder with a chemicalcomposition of (ZnO)x(B2O3)y(H2O)z. The most commonly used grades have thestructure 2ZnO·3B2O3·zH2O.

Zinc borate can be used as a fire retardant in PVC, polyolefins, elastomers,polyamides, and epoxy resins. In halogen-containing systems, it is used in conjunc-tion with antimony trioxide, while in halogen-free systems it is normally used in con-junction with other FRs, such as aluminum trihydrate, magnesium hydroxide, or redphosphorus. In a small number of specific applications, zinc borate can be usedalone.

17.7.4Melamines

Melamine (2,4,6-triamino-1,3,5-triazine), or melamine-derived f lame retardants,represent a small but fast-growing segment in the f lame-retardant market. In thisfamily, three chemical groups can be defined: pure melamine; melamine derivatives,i.e. salts with organic or inorganic acids such as boric acid, cyanuric acid, phosphor-ic acid or pyro/polyphosphoric acid; and melamine homologues containing multi-ring structures. Melamine-based f lame retardants show excellent f lame-retardantproperties and versatility in use because of their ability to employ various modes off lame-retardant action. Melamines are only brief ly mentioned in this chapter sincethey may be considered as additives rather than as true functional fillers.

17.7.5Phosphorus-Containing Flame Retardants

The element phosphorus (red phosphorus) is known to be an effective inhibitor ofcombustion. Phosphorus is found in both organic and inorganic materials that havef lame-retardant properties. Phosphate esters are widely used in PVC or in polymersthat have high hydroxyl group content. Phosphorus-containing polyols are also usedin the production of polyurethane foams.

17.7.5.1 Mechanism of ActionThe mechanism of action of phosphorus depends to a large extent on its chemical en-vironment. Under conditions when no heteroelements are present other than oxy-gen, phosphoric or phosphonic acids are formed. These accelerate the loss of volatilegroups from the burning polymer chain and provide a less combustible vapor barri-er.

The mode of action in the gas phase involves the formation of reactive phosphorus-based species such as PO•, P•, and HPO that are capable of removing the free radi-cals that can drive combustion. The reaction scheme outlined below for a triaryl phos-


304

phate is indicative of the underlying chemical mechanism of phosphorus-basedf lame retardance:

R3P=O → PO• , P•

H• + PO• + M → HPO + MHO• + PO• → HPO + 1/2 O2

HPO + H• → H2 + PO•

17.7.5.2 Ammonium PolyphosphateThis material is an inorganic salt of polyphosphoric acid and ammonia with the

general structure [NH4PO3]n. There are two main families of ammonium polyphos-phate as shown below:

APP I NH4+OP(=O)O–[NH4PO3]n–OP(=O)O+NH4

APP II NH4+OP(=O)O–[NH4PO3]m–P(=O)O–[PO3NH4]nOP(=O)O+NH4

Crystal phase I APP (APP I) is characterized by a variable linear chain length,showing a lower decomposition temperature (approximately 150 °C) and a higher wa-ter solubility than crystal phase II (APP II) (see Tables 17-9 and 17-10). APP II has across-linked and branched structure. The molecular weight is much higher than thatof APP I with n > 1000.

Tab. 17-9 Solubility and properties of some commercial gradesof APP I.

Monomer units pH Solubility in water Median particle in chain (n) (g/100 mL) size (µm)

20 5–6 5 1240 5–6 4 1460 5–6 3 1680 5–6 2 20

Tab. 17-10 Solubility and properties of some commercial gradesof APP II.

Material type pH P2O5 (%) Solubility in water Median particle (g/100 mL) size (µm)

Material 1 5–6 72 0.5 18Material 2 5–6 72 0.5 15Material 3 5–6 72 0.5 7Material 4 5–6 72 0.5 4

APP is a stable, non-volatile material. It slowly hydrolyzes in contact with water toproduce monoammonium phosphate (orthophosphate). This process is autocatalyt-ic and is accelerated by higher temperatures and prolonged exposure to water. The

17 Fire Retardants

305

temperature of decomposition of APP is related to the length of the polymer chain.Long-chain APP starts to decompose at temperatures above 300 °C to give polyphos-phoric acid and ammonia. Short-chain APP, defined as that having a chain length be-low 100 monomer units, begins to decompose at temperatures above 150 °C.

Ammonium polyphosphate (APP) and APP-based systems are very efficient halo-gen-free f lame retardants mainly used in polyolefins (PE, PP), epoxy resin thermosetsystems, polyurethanes, unsaturated polyester phenolic resins, and others. APP is anon-toxic, environmentally friendly material and it does not generate additionalquantities of smoke due to intumescence. Compared to other halogen-free systems,APP can be used at lower loadings. In thermoplastic formulations, it exhibits goodprocessability, good retention of mechanical properties, and good electrical proper-ties. In thermosets, it can be used in combination with ATH to obtain a significantreduction of the total amount of FR filler. Such combinations are used in construc-tion and electrical applications. The processability can be improved when the totalfiller level is reduced. Table 17-11 summarizes some benefits of APP in several poly-mer systems.

Table 17-11 Benefits of APP in several polymer systems.

Applications Benef its

Thermoplastics Inorganic polymerPolyethylene High FR efficiencyPolypropylene Halogen-free

Thermosets Environmentally friendlyPolyurethane Excellent processabilityEpoxy resin Good mechanical propertiesUnsaturated polyester Good electrical propertiesPhenolic resin Low smoke generated

Ammonium polyphosphate acts as a f lame retardant by intumescence as shown inFigure 17-7.

When plastic or other materials that contain APP are exposed to an accidental fireor heat, the f lame retardant starts to decompose, commonly into polymeric phos-phoric acid and ammonia. The polyphosphoric acid reacts with hydroxyl or othergroups of a synergist to form an unstable phosphate ester. Dehydration of the phos-phate ester then ensues. Carbon foam is built up on the surface against the heatsource (charring). The carbon barrier acts as an insulation layer, preventing furtherdecomposition of the material.

(NH4PO3)n + heat (> 250 °C) → (HPO3)n + nNH3

(HPO3)n + polymer → carbon char + H3PO4

–H2O


306

Co-synergists may also be used. Addition of synergetic products such as pentaery-thritol derivatives, carbohydrates, or spumific agents will significantly improve thef lame-retardant performance of APP.

17.7.6Low Melting Temperature Glasses

Low melting temperature phases (glasses) have long been considered as f lame-re-tardant polymer additives. Low melting temperature glasses can improve the thermalstability and f lame retardance of polymers by:

providing a thermal barrier for both polymer and char that may form as a com-bustion product;

providing a barrier to retard oxidation of the thermally degraded polymer and com-bustion of the char residue;

providing a “glue” to maintain the structural integrity of the char; providing a coating to cover over or fill-in voids in the char, thus providing a more

continuous external surface with a lower surface area; creating potentially useful components of intumescent polymer additive systems.

A low-melting temperature glass f lame-retardant system for PVC consisting of Zn-SO4/K2SO4/Na2SO4 may be used. This system is an excellent char former and asmoke suppressant. Low-melting glasses have also been tested with transition metalssuch as Ni, Mn, Co, and V and with main group metals such as Al, Ca, and Ce. It hasbeen shown that silica gel in combination with potassium carbonate is an effectivefire retardant (at a mass fraction of only 10% total additive) for a wide variety of com-mon polymers such as polypropylene, nylon, polymethyl methacrylate, polyvinylal-cohol, cellulose, and to a lesser extent for polystyrene and styrene–acrylonitrilecopolymers. The mechanism of action for these additives seems to involve the for-mation of a potassium silicate glass during combustion.

17 Fire Retardants

Fig. 17-7 Char formation in a polymer containing ammoniumpolyphosphate as a f lame retardant.

307

The action of ammonium pentaborate (APB) as a glass former in polyamide-6 hasbeen studied. The degradation of APB into boric acid and boron oxide provides a lowmelting glass at temperatures of interest, and then the char formed by the degrada-tion of polyamide-6 is protected by the glass, which induces the fire-retardant effect.

Zinc phosphate glasses, with glass transition temperatures in the range 280 °C to370 °C, have recently been developed. These may be compounded with engineeringthermoplastics such as polyarylether ketones, aromatic liquid polyesters, polyarylsul-fones, perf luoroalkoxy resins or polyetherimides to afford a significant increase inLOI relative to the unfilled polymer. The incorporation of low-temperature glasses asconventional fillers in commodity polymers can, however, produce measurablef lame-retardance effects as well. Nevertheless, the loading must be high enough(>60 wt. % in polycarbonate) to significantly increase the LOI.

17.8Tools and Testing

While a complete survey of the testing techniques for f lame retardants is beyond thescope of this chapter, testing methods such as cone calorimetry, the requirements ofthe UL 94 testing protocols, and radiant heat panels deserve mention here.

The cone calorimeter (Figure 17-8) is perhaps the most valuable research instru-ment in the field of fire testing. It functions by burning sample materials cut to spe-cific geometry and measuring the evolved heat using the technique of oxygen deple-tion calorimetry. The basis of this technique is that the heat released during the com-bustion of materials is directly proportional to the quantity of oxygen consumed inthe combustion process. The apparatus owes its name to the shape of the truncatedconical heater that is used to irradiate the test specimen with f luxes of up to100 kW m–2. Heat release is a key measurement in the progress of a fire [20–22]. TheFTT cone calorimeter has been produced to meet all existing standards (includingISO 5660, ASTM E 1354, ASTM E 1474, ASTM E 1740, ASTM F 1550, ASTM D 6113,NFPA 264, CAN ULC 135, and BS 476 Part 15).

The UL 94 standard provides a preliminary indication of a polymer’s suitability foruse as part of a device or appliance with respect to its f lammability. It is not intend-ed to ref lect the hazards of a material under actual fire conditions. UL 94 incorpo-rates the following tests: 94HB, 94V, 94VTM, 94-5V, 94HBF, 94HF, and radiant pan-el (ASTM D 162). The 94HB test describes the horizontal burn method. Methods 94Vand 94VTM are used for vertical burn, a more stringent test than 94HB. The 94-5Vtest is for enclosures, that is, for products that are not easily moved or are attached toa conduit system. The 94HBF and HF tests are used for non-structural foam materi-als, i.e. acoustical foam. The radiant panel test is an ASTM (E162) test to determinethe f lame spread of a material that may be exposed to fire.

Flame spread, critical heat f lux, and smoke production rate of f loor coverings andcoatings are measured using a radiant heat source according to DIN 4102-14 and ENISO 9239 for classification into DIN 4102 and Euroclasses.

17.8 Tools and Testing

308

17.9Toxicity

In considering the toxicity of f lame retardants one has to consider the toxicity of ma-terials liberated during combustion from originally non-toxic materials as well as thetoxicity of the FR materials alone. The most toxic materials typically liberated fromthe combustion of polymers include carbon monoxide and hydrogen cyanide (fromnitrogen-containing polymers). When considering the production of toxic chemicalsthat can arise from the use of f lame retardants such as halogenated f lame retardants,the negative effects often have to be considered as offsetting, by orders of magnitude,the potential toxicity from gas evolution that would happen in the case of a non-f lame-retarded polymer. This is not to say that the toxicity effects, e.g. of halogenatedfire retardants, should be ignored and no efforts should be undertaken to minimizethem, but the danger has to be set in the context of their use.

17.9.1Metal Hydroxides

Aluminum trihydrate is not considered a toxic material and is classified as GRAS(generally recognized as safe) in paper and is specifically listed for use in polymers.It is not approved for use in direct food contact but is approved for use in indirect food

17 Fire Retardants

Fig. 17-8 Schematic of a cone calorimeter.

309

additives such as polymers, cellophane, and rubber articles intended for repeateduse, paper and paperboard components, and also in defoaming agents used in themanufacture of paper and paperboard. The human therapeutic category is as anantacid and an anti-hyperphosphatemic [23]. From a hygiene perspective ATH is anuisance dust and is mildly irritating.

Magnesium hydroxide meets the specifications of the Food Chemicals Codex [24].It is used in food with no limitation other than current good manufacturing practice.The affirmation of this ingredient is as GRAS as a direct human food ingredient. Itis permitted for indirect use in polymers. It is a non-toxic material and the humantherapeutic category is as an antacid and a cathartic [25]. The material can be irritat-ing owing to its potential to develop alkalinity on contact with water.

17.9.2Antimony Trioxide

Antimony trioxide is a material that is suspected of being a human carcinogen buthas not been proven to be so. Epidemiological studies suggest that occupational ex-posure to antimony trioxide could be associated with an increased occurrence of lungcancer [26]. Concerns about human carcinogenicity have arisen because the materialhas been shown to be carcinogenic in animal studies [27,28]. Antimony trioxideproved to be carcinogenic in female, but not male, rats after inhalation exposure, pro-ducing lung tumors. The studies with rats tend to indicate that there is an increasedrisk of cancer developing from long-term chronic exposure than from acute or occa-sional exposure [29–33]. Antimony trioxide has been reported to induce DNA dam-age in bacteria. In summary, antimony trioxide is often treated as though it were car-cinogenic in humans, a position that is supported in practice by many toxicologists.Regardless of the defined regulatory status, appropriate environmental and engi-neering controls are recommended in handling this material. Antimony trioxidemost likely exerts its effect as a contact irritant, since patch tests for allergy usuallyproduce no reaction. Further information on its effects on humans can be found inrefs. [34,35].

17.9.3Brominated Fire Retardants

At the current time there are studies underway to determine the toxicological char-acteristics and effects on hygiene of a number of halogen-containing FRs. Work hasbeen published in Sweden suggesting the possibility of environmental accumulationresulting in the presence of these materials in the food chain from fish to humans(breast milk) [36]. The use of some brominated FRs such as polybrominatedbiphenyls and brominated diphenyl ether has been either phased out or limited inEurope. There is also increasing pressure to phase out higher polybrominateddiphenyl ethers, which, although not believed to bioaccumulate or be toxic in them-selves, are suspected to degrade to the more problematic lower brominated diphenylethers. Without prejudice to the outcome of these or other studies, it is fair to say that

17.9 Toxicity

310

concern over the toxicity of halogenated FRs has led many consumers to consider al-ternative FRs that do not contain halogens. Public concern is driven in part by anxi-ety over the health effects of dioxins and dioxin-like materials.

Regarding brominated f lame retardants, Cullis [37] stated that unless suitable met-al oxides or metal carbonates are also present, virtually all of the available bromine iseventually converted to gaseous hydrogen bromide. This is corrosive and a powerfulsensory irritant. In a fire situation, however, it is always carbon monoxide (CO) or hy-drogen cyanide (HCN), rather than an irritant, that causes rapid incapacitation. Ow-ing to its high reactivity, hydrogen bromide is unlikely to reach dangerously high con-centration levels [37].

The toxicity profile of brominated f lame retardants, in contrast to those of otherf lame retardants, is relatively well understood. Studies have been carried out by theWorld Health Organization (WHO), the Organization for Economic Cooperation andDevelopment (OECD), and the European Union (EU). The World Health Organiza-tion concluded in the early 1990s [38] that any risk presented by the major brominat-ed f lame retardants, such as decabromodiphenyl oxide and tetrabromobisphenol A,was minimal and manageable. Toxicology testing is ongoing on several brominatedf lame retardants of the diphenyl ether family and hexabromocyclododecane in ac-cordance with industry’s commitments under the OECD Voluntary Industry Com-mitment and EU risk assessment programs [39,40].

17.9.4Chlorinated Fire Retardants

A major concern with respect to the use of chlorinated FRs is their potential to formknown hazardous materials, particularly dioxins or dioxin-like materials. In analogyto the situation with the brominated analogues, the most likely product of combus-tion of chlorinated FRs is HCl. However, the potential for formation of other poten-tially hazardous materials is real and has to be considered carefully [41].

The number of mixed-halogen congeners based on dibenzodioxins or dibenzofu-rans is in excess of 4000. Interestingly, at the present time, relatively few of these sub-stances are regulated under U.S. or European laws [42]. Among all these substances,only the 2,3,7,8-substituted chlorinated, brominated and mixed-halogenated diben-zodioxins and dibenzofurans are of concern in terms of toxicity, because dioxin toxi-city is a receptor-mediated event and only these congeners have the appropriate con-formation to interact with the biological receptors. They also accumulate in the fat ofanimals and humans [43].

Some countries (e.g., Germany) have set regulations for the maximum content ofsome 2,3,7,8-substituted polychlorinated dibenzo-para-dioxins and dibenzofurans inproducts. The availability of relevant data on f lame retardants in the open literatureis limited, especially for some existing chemicals produced before regulations forcommercialization were strengthened in several countries. The International Pro-gram for Chemical Safety (IPCS) has issued evaluations for some f lame retardantsand is preparing evaluations for others.

17 Fire Retardants

311

17.9.5Boron-Containing Fire Retardants

There is little information on the health effects of long-term exposure to boron. Mostof the studies have been on short-term exposures. The Department of Health and Hu-man Services, the International Agency for Research on Cancer, and the U.S. Envi-ronmental Protection Agency (EPA) have not classified boron with regard to its hu-man carcinogenicity. In one animal study, no evidence of cancer was found after life-time exposure to boric acid in food. No human studies are available [44].

Zinc borate is also commonly used as a fire retardant. The toxicity of zinc is low[45] and data indicate that relatively large amounts of zinc may pass for years throughthe kidneys and gastrointestinal tract in humans without causing any detectable clin-ical damage [46]. Some zinc salts are irritants and corrosive on skin contact and wheningested they can act as emetics [47].

17.9.6Phosphorus-Containing Fire Retardants

Organic phosphate esters are commonly used as fire retardants in many applications,including housings for personal computers and video display monitors. While therehave been no definitive findings to date that would directly negate the use of thesematerials, the European Committee for Standardization has established a workinggroup to prepare a standard addressing the risk associated with the presence of or-ganic chemical compounds in toys [48]. Their priority list of chemicals includes morethan 50 f lame retardants, most of them chlorine-, bromine- or phosphorus-based.The Electronics Industry Alliance (EIA) is currently working on a “Material Declara-tion Guide” [49]. All listed “materials of interest” must be disclosed, if they are in-corporated into parts at levels greater than 0.1%. FRs are listed in this material cate-gory, including organophosphorus FRs [36].

There is an evolving concern about the use of some particular organophosphates,such as tricresyl phosphate, tri-n-butyl phosphate, and triphenyl phosphate[36,50–52]. It has been stated [51] that animal studies have shown that these phos-phate esters can cause allergies, learning problems, and deterioration in sperm pro-duction, and can inf luence the white and red blood cell counts of humans. In a re-cent publication [52], levels of f lame retardants in indoor air at an electronics scraprecycling plant and other work environments were considered; 15 brominated FRsand 9 organophosphorus compounds were analyzed. Although the measured con-centrations found in all these studies were very low and orders of magnitudes belowany occupational limit values (and so the levels of exposure provided very significantmargins of safety), health related concerns were still raised by the researchers.

There are few demonstrated toxicological concerns associated with the handlingand use of ammonium polyphosphate (APP). Thus, APP does not require hazardwarning labels and has been shown to cause acute oral toxicity in rats only at a levelof 5000 mg/kg. APP is capable of causing mild skin irritation. As in the case of most

17.9 Toxicity

312

phosphorus-containing materials, APP should be prevented from entering surfacewaters as it may contribute to eutrophication of stagnant water.

17.10Manufacturers of Fire Retardants

Table 17-12 is a list of the major global producers of fire retardants. Statistics on theproduction of fire retardants may be obtained from the Flame Retardant ChemicalsAssociation (FRCA) in the US (www.fireretardants.org) and from the EuropeanFlame Retardants Association (EFRA), which may be contacted through www.cefic-efra.com.

17.11Concluding Remarks

Fire retardants have proven benefits. Their use has saved an inestimable number ofpeople from death and injury. Their use has enabled the proliferation of plastics totheir present extent in society. Nevertheless, there are valid concerns about the safetyof individual fire retardants that need to be addressed on an individual basis. Thereis also a need to collect data to fill the gaps that exist for some existing fire-retardantmaterials. Only through the acquisition of valid data in a thorough scientific mannercan many of the current and emerging concerns about fire-retardant materials beconclusively addressed and the benefits of their fire retardance be realized to thegreatest possible extent [53–55].

Acknowledgements

The author thanks Huber Engineered Materials for permission to publish this chap-ter. The following colleagues are acknowledged: David Temples of J. M. Huber’s Poly-mer Technology Center, Fairmount, GA, for generating data for Figure 17-4; Dr.Thomas Lynch of J. M. Huber’s Polymer Technology Center, Fairmount, GA, for in-formation and data on the topic of chemical surface treatment of minerals.

17 Fire Retardants

31317.11 Concluding Remarks

Tab. 17-12 Major global producers of fire retardants.

Aluminum trihydrate ALCAN Alcoa Martinswerk J. M. Huber Corp. Custom Grinders R. J. Marshall Shandong Corp. Franklin Minerals Aluchem Nabaltech GmbH Albemarle Inc.

Magnesium hydroxide Kisuma Incemin Dead Sea Bromine Company Penoles J. M. Huber Corp. Martinswerk

Antimony trioxide Great Lakes Chemical Corp. Albemarle Corp. Campine Sica

Halogenated FRs Great Lakes Chemical Corp. Albemarle Corp. Clariant Inc. Akzo Nobel Eurobrom Dover Corp. Ineos Chlor Caffaro

Melamines DSM Melamine Buddenheim Iberica Akzo Nobel Argolinz

Phosphorus-containing FRs Clariant Buddenheim Iberica Great Lakes Chemical Corp.

Boron compounds US Borax Inc. Joseph Storey William Blythe Rio Tinto Borax Europe Ltd. Buddenheim Iberica

314

References

17 Fire Retardants

1 Kasahiwagi, T., Proc. 25th Internat. Sympo-sium on Combustion, 1994, The CombustionInstitute, Pittsburgh, PA, pp 1423–1437.

2 Beyler, C. M., Hirschler, M. M., “ThermalDecomposition of Polymers” in SFPEHandbook of Fire Protection Engineering, 1sted. (Eds.: Di Nenno, P. J., et al.), NationalFire Protection Assoc., Quincy, MA, 1988,pp 1–119.

3 Fire and Smoke, Understanding the Hazards,National Research Council, National Acade-my of Sciences, Washington DC, 1986.

4 Hartzell, G. E., Fire and Materials 1988, 13,53–60.

5 Gottuk, D. T., et al., Fire Safety Journal 1995,24, 315–331.

6 Pitts, W. M., Progr. Energy and CombustionScience 1995, 21, 197–237.

7 Pitts, W. M., Proc. 5th Internat. Symp., Inter-nat. Assoc. for Fire Safety Science, 1997, In-terscience Communications Ltd., Mel-bourne, Australia, pp 535–546.

8 Purser, D. A.,“Toxicity Effects of Combus-tion Products”, in SFPE Handbook of FireProtection Engineering, 2nd ed. (Eds.: DiNenno, P. J., et al.), National Fire ProtectionAssoc., Quincy, MA, 1995, pp 85−146.

9 Tewarson, A., “Improved Fire and SmokeResistant Materials” in ref. [2].

10 Robertson, A. F., Fire Technology, 1975, 11,80.

11 Fenimore, C. P., Jones, G. W., Combust.Flame 1966, 10, 295.

12 Cote, A. E., Fire Protection Handbook, 18thEd., National Fire Protection Association,Quincy, MA, 1997, Table A-5.

13 Ashton, H. C., et al., Proc. SPE RETEC Poly-olef ins 2003, Houston, TX, Feb. 24–26,2003, pp 395–407.

14 Ashton, H. C., Proc. Functional Fillers forPlastics 2001, Intertech Corp., San Antonio,TX, September 11–13, 2001.

15 Green, D. W., Dallavia, Jr., A. J., Paper pre-sented at the 46th Ann. Conf. of the Compos-ites Institute, The Society of the Plastics In-dustry, Inc., Feb. 18–21, 1991.

16 Green, D., Dallavia, Jr., A. J., Proc. 46th SPEANTEC, 1989, 34, 619.

17 Robertson, A. F., ASTM Spec. Tech. Publ.No. 344, ASTM International, West Con-shohocken, PA, 1975, p. 33.

18 Gerards, T., Proc. FILPLAS ‘92, May 19–20,1992, Manchester, UK.

19 Bar Yaakov, Y. et al., Flame Retardants 2000,87–97.

20 Babrauskas, V., “The Cone Calorimeter”(Section 3/Chapter 3), pp. 3-37 to 3-52 inref. [8].

21 Test Method ASTM E 1354 “Standard TestMethod for Heat and Visible Smoke Re-lease Rates for Materials and Products Us-ing an Oxygen Consumption Calorimeter”,Annual Book of ASTM Standards, Vol. 04.07,ASTM International, West Conshohocken,PA, 2001.

22 Kaplan, H. L., et al., Combustion Toxicology,Principles and Test Methods, Technomic Pub-lishing Company, Inc., Lancaster, PA, 1983,pp 7–49.

23 The Merck Index, An Encyclopedia of Chemi-cals, Drugs and Biologicals, 11th ed., Merckand Co., Rahway, NJ, 1989, p. 57.

24 Food Chemicals Codex, 4th edition, Nation-al Academy Press, March 1, 1996.

25 See ref. [23], p. 892.26 Website: http://toxnet.nlm.nih.gov/cgi-

bin/sis/search/f?./temp/~AAAXEa4wa27 Groth, D. H., et al., J. Toxicol. Environ.

Health 1986, 18, 607–626.28 International Agency for Research on Can-

cer, IARC, Monographs on the Evaluation ofthe Carcinogenic Risk of Chemicals to Man,1972 – Present (Multivolume work), WorldHealth Organization, Geneva, 1989, p. V47296.

29 Patty, F. (Ed.), Industrial Hygiene and Toxi-cology: Volume II: Toxicology, 2nd ed., Inter-science Publishers, New York, 1963, p. 995.

30 Clayton, G. D., Clayton, F. E. (eds.), Patty’sIndustrial Hygiene and Toxicology: Volume2A, 2B, 2C: Toxicology, 3rd ed., John Wiley& Sons, New York, 1981–1982, p. 1510.

31 ACGIH, Documentation of the ThresholdLimit Values and Biological Exposure Indices,4th ed., American Conference of Govern-mental Industrial Hygienists, Inc., Cincin-nati, Ohio, 1980, p. 20.

32 See ref. [28], p. 47 302.33 ACGIH, Threshold Limit Values for Chemical

Substances and Physical Agents and BiologicalExposure Indices for 2002, A2; Suspected hu-man carcinogen. /Antimony trioxide pro-

315References

duction/ TLVs & BEIs: 15, American Con-ference of Governmental Industrial Hy-gienists, Cincinnati, OH, 2002.

34 Rom, W. N. (ed.), Environmental and Occu-pational Medicine, 2nd ed., Little, Brownand Company, Boston, MA, 1992, p. 815.

35 Prager, J. C., Environmental ContaminantReference Databook, Vol. 2, Van NostrandReinhold, New York, 1996, p. 90.

36 Carlson, H., et al., Environ. Sci. Technol.2000, 34, 3885–3889.

37 Cullis, C. F., Proc. Internat. Conf. on FireSafety, 1987, 12, 307–323.

38 World Health Organization, “BrominatedDiphenyl Ethers, IPCS EnvironmentalHealth Criteria 162”, Geneva, 1994.

39 EEC, “Council Regulation EEC 793/93 of 23March, 1993, on the Evaluation and Controlof the Risks of Existing Substances”, EECOfficial Journal, L 84, 5 April, 1993, pp.1–75.

40 Hardy, M. L., “Regulatory Status and Envi-ronmental Properties of Brominated FlameRetardants Undergoing Risk Assessment inthe EU: DBDPO, OBDPO, PEBDPO, andHBCD”, 6th European Meeting on Fire Retar-dancy of Polymeric Materials, Lille, France,September 24–26, 1997.

41 Pomerantz, A., et al., Environ. Health Per-spectives 1978, 24, 133–146.

42 Kurz, R., Paper presented at the BromineScience and Environmental Forum, JapanSeminar, Tokyo, November 6, 1998.

43 Brominated Flame Retardants/CEM Work-ing Group, “Polybrominated Dibenzodiox-ins and Dibenzofurans (PBDDs/PBDFs)from Flame Retardants ContainingBromine: Assessment of Risk and Pro-posed Measures”, Report No. III/4299/89

in German to the Conference of Environ-mental Ministers, Bonn, Germany, Septem-ber 1989.

44 Agency for Toxic Substances and DiseaseRegistry (ATSDR), Toxicological Profile ForBoron, U.S. Department of Health and Hu-man Services, Public Health Service, At-lanta, GA, 1992.

45 ACGIH, Documentation of the ThresholdLimit Values and Biological Exposure Indices,5th ed., American Conference of Govern-mental Industrial Hygienists, Cincinnati,OH, 1986, p. 645.

46 Browning, E., Toxicity of Industrial Metals,2nd ed., Appleton–Century–Crofts, NewYork, 1969, p. 353.

47 See ref. [30], p. 2039.48 C&EN, “Safety of Toys: Report on the

Progress of the Mandated Work CoveringOrganic Chemical Compounds in Toys”,2001, 79(52), 19.

49 Electronic Industry Alliance, EIA MaterialDeclaration Guide, 2000, p. 25.

50 Carlson, H., et al., Environ. Sci. Technol.1997, 31, 2931–2936.

51 Svenska Dagbladet, “Nytt plastämne kannge allergi” (New plastic material causes al-lergies), November 11, 1999.

52 European Flame Retardants Association,EFRA, “Position Paper on Triphenyl Phos-phate (TPP)”, 2000; http://efra.cefic.org

53 Lloyd’s List, 2 Aug. 1995, p. 3., Informa,London, UK.

54 Hall, J. R., The Total Cost of Fire in the Unit-ed States Through 1994, National Fire Pro-tection Agency, Quincy, MA, 1997.

55 Tsuda, Y., J. Med. Soc. Toho (Japan) 1996,43(3), 188–192.

18Conductive and Magnetic Fillers

Theodore Davidson

18.1Introduction

Polymers are often favored materials for the manufacture of devices and compo-nents. They combine ready formability and molding characteristics with good insu-lating and dielectric properties. These attributes are sufficient to make plastics andelastomers functional for many purposes. In some instances, however, it would bedesirable to have polymers that are able to dissipate static charges or shield their con-tents from electromagnetic fields. For other applications, the control of static charge(as in xerography) or the control of conductivity as a function of current density (self-limiting switches and heaters) is attractive. All of these functions have been achievedby the addition of conductive fillers to polymers. In this chapter, some of these ap-plications and the roles played by functional fillers in controlling polymer conductiv-ity are described. In a separate discussion it is shown how the properties of perma-nent magnetization may be conferred upon polymers through the assembly of com-posite materials consisting of a magnetic solid dispersed in a polymer matrix orbinder. These range from the familiar refrigerator magnets and magnetic recordingtape to data disks that can store information at very high densities.

It is widely observed that when carbon black (CB) is added to an insulating polymerthere is a threshold concentration Vc at which the composite changes from insulatingto conductive. Conductivity (and resistivity) change by several orders of magnitudefor a small increment in carbon black concentration. This threshold, which typicallyoccurs in the range 3–15 wt % CB, has been the subject of many investigations. Un-fortunately, in these systems there are not only many independent variables, such asthe surface chemistry and dispersibility of various blacks but, to complicate thesestudies, there are also significant effects of polymer interfacial chemistry, variousmodes of compounding, and segregation to phase boundaries. An attempt is madehere to illustrate several of these effects with examples from the literature. The se-lected examples are not all-inclusive and do not attempt to address the very large vol-ume of patent literature in this field. Previously, the field has been surveyed by Sichel[1] and Rupprecht [2] and reviewed by Huang [3].


318

18.2Carbon Black

18.2.1General

Carbon black (often referred to as “black” or “CB” in this chapter) consists of ag-glomerates of small assemblies of carbon particles called aggregates, which areformed in the processes by which solid carbon condenses from the vapor. An ag-glomerate consists of a number of aggregates held together physically as opposed tothe continuous pseudo-graphitic structure of the aggregates. Figure 18-1 shows aschematic of CB formation. Blacks are variously described as “high structure” or “lowstructure”, which correlates with their spatial extent, the former having larger di-mensions than the latter. Their structure can be further characterized by the adsorp-tion of liquid dibutyl phthalate (DBP or DBPA), with those blacks that take up moreDBP being termed “high structure”. They typically have a more reticulated form,whereas the low structure blacks are more compact [4].

18.2.2Varieties

Carbon blacks are made by the combustion of natural gas, or acetylene, or various hy-drocarbon oil feedstocks under reducing conditions. In former times, “channelblacks” were made by thermal decomposition of natural gas in an open system withthe black being collected on lengths of channel iron in the shape of long inverted veessupported over a line of gas f lames.

Increased efficiency and reduced loss to the surroundings is achieved in closed-fur-nace systems, yielding “furnace black” or “lampblack”. Thermal decomposition ofnatural gas produces “thermal blacks”, with large primary particles (≤ 500 nm) and

18 Conductive and Magnetic Fillers

Fig. 18-1 Schematic diagram of carbon black formation. (Afterref. [5], Cabot Corporation).

319

surface areas of 6–8 m2 g–1. Details of these various types are given by Kühner andVoll [6] and by Medalia [7]. Currently manufactured grades cover a range from ap-proximately 9 to 300 m2 g–1 in external surface area, 30 to 180 cm3 per 100 g in DBPAabsorption, and 0.5 to 15% “volatiles” [7]. There is a continuous exothermic processfed by acetylene that yields “acetylene blacks”, which are more crystalline and have ahigh degree of structure. These are difficult to densify but are useful for their en-hanced electrical conductivity and anti-static properties when compounded into rub-ber and plastics.

Tab. 18-1 Representative grades of carbon black used in plastics.(adapted from ref. [7]).

Type Surface area by nitrogen adsorption, m2 g–1 DBPA, cm3/100 g Volatiles, %

high color 240 50 2.0medium color 200 117 1.5medium color 210 74 1.5regular color 140 114 1.5regular color 84 102 1.0low color 42 120 1.0aftertreated 138 55 5.0conductive 254 178 1.5conductive 67 260 0.3conductive 950 360 1.0

When dry, CB forms aggregates (Figure 18-2). The proportions of these groupingshave been measured for 19 different blacks and may be related to their “structures”as classified by DBPA uptake (Table 18-2) [8].

Some 90% of CB production goes to the rubber industry. Of the non-rubber con-sumption, 36% is for additives to plastics, while 30% goes into printing inks [6]. Theprocess technology for making each type of carbon black is a large subject in itself.Surveys can be found in refs. [6,9]. The mechanism of formation, involving nucle-ation, aggregation, and agglomeration of the aggregate particles, has been dealt withby Bansal and Donnet [10].

18.2 Carbon Black

320 18 Conductive and Magnetic Fillers

Fig. 18-2 Rendering of shape categories for carbon black aggregates.Compare with data in Table 18-2. (After ref. [8], Marcel Dekker, Inc.).

Tab. 18-2 Weight percent of aggregates in four shape categories forvarious CB grades in the dry state. (After ref. [8])[a].

Carbon black DBPA cm3/100 g Weight % of each shape category1 2 3 4

CD-2005 174 0.1 4.8 17.8 77.3N358 155 0.1 7.9 34.9 57.1HV-3396 138 0.2 4.2 33.4 62.2N121 134 0.4 8.8 28.7 62.1N650 129 0.2 9.2 47.0 43.6N234 124 0.3 9.0 32.5 58.3N299 124 0.4 10.0 33.2 56.4N351 120 0.1 9.2 46.9 43.8N550 120 0.6 13.8 45.3 40.3N339 118 0.2 9.5 36.6 53.7N110 115 0.3 8.7 31.1 59.9N220 115 0.6 11.9 34.0 53.5N330 100 0.2 10.2 44.1 45.5N660 91 0.4 15.4 52.5 31.7N630 78 0.4 21.4 49.0 29.2N774 77 1.3 20.8 46.3 31.6N326 72 1.6 23.4 35.2 39.8N762 67 2.5 22.4 47.7 27.4N990 35 44.9 34.8 14.4 5.9

[a] Note: N designates furnace blacks with “normal” curing in rubbercompounds. The first digit after N relates to particle size or surfacearea. The second and third numbers are arbitrary grade designators.

321

18.2.3Commercial Sources

There are numerous manufacturers of basic carbon black. A partial list would includeAkzo Nobel, Cabot Corp., Columbian Chemicals, Chevron–Phillips, and DegussaCorp. Most manufacturers describe the various grades that they produce on theirwebsites.

As with many fillers, colorants, and additives, CB is frequently compounded witha thermoplastic to form a concentrate or masterbatch, which is then “let down” anddispersed in a chosen polymer to the final desired concentration. Masterbatches areavailable from a variety of compounders including Americhem, Inc., Ampacet Corp.,Cabot Corp., Colloids Ltd. (UK), Hubron Manufacturing Div., Ltd. (UK), PolychemUSA, A. Schulman, Singapore Polymer Corp. (PTE.) Ltd., and others.

18.2.4Safety and Toxicity

Carbon black has been assigned an Occupational Exposure Limit (OEL) of 3.5 mg m–3

according to the American Conference of Governmental Industrial Hygienists(ACGIH) regarding its threshold limit value (TLV). The same value has been assignedby the National Institute for Occupational Safety and Health (NIOSH) regarding rec-ommended exposure limits (REL), and the Occupational Safety and Health Admin-istration (OSHA) regarding permissible exposure limits (PEL). A detailed discussionof CB safety, health, and environmental considerations can be found on the websiteof the Cabot Corporation [11].

18.2.5Surface Chemistry and Physics

In general, the quenching step of the manufacturing process presents an opportuni-ty for oxidation of the fine CB powder. Surface groups can be further enhanced in-tentionally by oxidation with acid or ozone. Surface groups formed in this way are pri-marily carboxylates and phenolics. An untreated black typically has pH > 6 and avolatile content of 0.5–1.5%. Oxidized blacks have pH < 6 and a volatile content of3–10%.

Pantea and co-workers [12] examined ten different CBs having a range of surfaceareas from 150 to 1635 m2 g–1 and DBP absorptions, as a measure of structure, rang-ing from 115 to 400 cm3 per 100 g. In compacted powder specimens of these highlyconductive CBs, no correlation was found between measured conductivity and sur-face concentrations of oxygen or sulfur. The measured ranges were from “not de-tectable” (n.d.) to 0.8 atom % oxygen and from n.d. to 0.3 atom % sulfur, as deter-mined by X-ray photoelectron spectroscopy (XPS). However, upon examining the C1s

spectra at high resolution, particularly with regard to the C1s peak at 284.5 eV associ-ated with graphitic character, it was found that the full-width at half-maximum(FWHM) of this peak correlated well with the observed conductivities. Static second-

18.2 Carbon Black

322

ary-ion mass spectra (SIMS) and low-pressure nitrogen adsorption isotherms con-firm the effect that graphitic surface structure has in inf luencing “dry” CB conduc-tivity [12].

18.3Phenomena of Conductivity in Carbon Black Filled Polymers

18.3.1General

When the concentration of CB in a polymer is progressively increased, there is a crit-ical volume fraction, Vc, at which the electrical conductivity increases by several or-ders of magnitude [13,14]. This critical volume fraction differs for various polymersand for different blacks (see Figure 18-3 and Table 18-3). As expected, it is also de-pendent upon mixing and on the resulting mesoscopic distribution of CB in the in-sulating matrix polymer. It has been assumed that the carbon particles need notmake continuous physical contact, but rather that within some small distance, say2 nm, electrons can tunnel through the insulating polymer and provide pathways forconduction. These observations have led to explanations of conductivity based on per-colation theory, which is formulated to account for the movement of charge in a dis-ordered medium. Of course, there are some agglomerates of CB in closer contact, andwithin aggregates a different “graphitic” mechanism of conduction may be expected[8,12].


Fig. 18-3 Log resistivity vs. concentration curves for various carbonblacks in HDPE. (After ref. [18], Marcel Dekker, Inc.).

323

Tab. 18-3 Threshold concentrations for XE-2, a highly conductivecarbon black, in various polymers. (After ref. [18])

Polymer Threshold concentration, wt. %

High density polyethylene (MI = 20) 4.0High density polyethylene (MI = 2) 2.6Cross-linked HDPE 1.8Low density polyethylene (LDPE) 1.6Polypropylene (PP) 1.6Poly(ethylene-co-14%vinylacetate), EVA 1.2Poly(ethylene-co-28%vinylacetate), EVA 3.2High-impact polystyrene, HIPS 2.8Transparent styrene-butadiene copolymer, SB 2.0Polycarbonate, PC 4.4Polyphenylene oxide, PPO 1.6Polyphenylene sulfide, PPS 1.2Polyacetal, POM 1.0Polyamide-6,6, PA-66 1.4Oil-extended thermoplastic elastomer, TPE 2.4

18.3.2Percolation Theories

Systems involving granular metal particles and polymers have been modeled interms of mixtures of random voids and conductive particles. There is also an invert-ed random void model [15]. For CB in polymers, percolation is the process by whichcharge is transported through a system of interpenetrating phases. In these systems,the first phase consists of conductive particle agglomerates and the second is an in-sulating polymer. Since charge transport is at least in part by tunneling through thininsulating regions, the combined theory is referred to as the tunneling–percolationmodel (TPM) [16].

As Balberg notes in a recent review: “The electrical data were explained for manyyears within the framework of inter-particle tunneling conduction and/or the frame-work of classical percolation theory. However, these two basic ingredients for the un-derstanding of the system are not compatible with each other conceptually, and theirsimple combination does not provide an explanation for the diversity of experimen-tal results” [17]. He proposes a model to explain the apparent dependence of percola-tion threshold critical resistivity exponent on the “structure” of various carbon blackcomposites. This model is testable against predictions of electrical noise spectra forvarious formulations of CB in polymers and gives a satisfactory fit [16].

18.3 Phenomena of Conductivity in Carbon Black Filled Polymers

324

18.3.3Effects of Carbon Black Type

The conductivity of a polymer–CB composite depends upon such factors as:

carbon black loading, the physical and chemical properties of the chosen carbon black, the chemistry of the polymer and its morphology in the solid state, mixing and finishing processes used to create the composite.

For various CBs dispersed in the same matrix, e.g. high density polyethylene(HDPE), the critical concentration Vc ranges from 8 to 62 wt. % (see Figure 18-3).Printex XE-2 shows the lowest critical volume fraction, while MT-LS shows the high-est. When the XE-2 blend has reached its asymptotic minimum resistivity (at 14%CB), blends of acetylene black, N-550, and MT-LS at this loading in HDPE are still in-sulating. Comparable results for various CBs in polypropylene have also been re-ported [19].

Figure 18-4 shows that the profile of resistivity vs. concentration differs markedlyfor the four blacks studied in styrene-butadiene rubber (SBR). A possible explanationwas thought to lie in differences in the surface chemistry of the blacks, particularlyin the amount of oxygen-containing groups present on their surfaces, as suggestedby Sichel and co-workers [14]. Foster [5] has reported resistivity measurements on anamine-cured epoxy matrix and on a thermosetting acrylic polymer with various CBs.The data show that “post-treated” grades (bearing more surface functionality) have ahigher surface resistivity than untreated furnace blacks. Also, low surface area, lowstructure CBs dispersed in these polymers show consistently higher resistivity thanhigh structure, high surface area blacks. On the other hand, recent XPS and SIMSanalyses by Pantea et al. [12] suggest that “…the concentration of non-carbon ele-ments on the carbon black surface is not a determining factor for the electrical con-ductivity.” According to these authors, the amount of graphitic surface characterseems to be determinative. However, the ultra high vacuum environment of XPS andSIMS may promote desorption of oxygenated species and other organic compounds.It would be of interest to measure blacks of higher surface functionality by these tech-niques.

18.3.4Effects of Polymer Matrix

When comparing the same CB in different matrices, the threshold concentrations ofXE-2 (a high structure CB) range from 1.0 wt. % in polyacetal to 4.4 wt. % in polycar-bonate (see Table 18-3). There are explicable variations in Vc among the two types ofpolyethylene of different MI cited. It is notable, however, that all the values of Vc arewithin a fairly limited range of 1–5 wt. %. This suggests dominance of the CB prop-erties over those of the matrix [18].


325

Differences may be observed in both the surface and bulk conductivities for a giv-en CB in various polymers as, for example, was reported by Narkis et al. [19] in com-paring PP, PS, HIPS, PE, and Noryl®. There is even a subtle effect of composition onresistivity in the case of PP homopolymer vs. PP copolymer because the latter con-tains a rubbery component (Table 18-4). The CB apparently segregates into the rub-bery phase and a higher loading of CB is needed to reach levels of conductivity com-parable to that seen with the homopolymer [19]. Effects of titanate surface treatmentson the conductivity of CB compounds are described in Chapter 5.

Tab. 18-4 Inf luence of type of polypropylene (homopolymer vs.copolymer) and conductive filler on surface resistivity and impactstrength of filled composites. (After ref. [19])[a]

Homopolymer Copolymer

carbon black, wt. % 1.2 1.2 3.7 4.6 4.8 5.1glass fibers, wt. % 15.4 15.4 10 15.4 10 10surface resistivity, Ω/sq 106 >1012 108 106 105 104

notched Izod impact strength, J m–1 57 117 140 115 140 135

[a] Note: CB is Ketjenblack EC 600 JD (Akzo, The Netherlands).The glass fibers are 10 µm in diameter and 3 mm long(Vetrotex, Owens-Corning).

18.3 Phenomena of Conductivity in Carbon Black Filled Polymers

Fig. 18-4 Log resistivities vs. concentrations for various carbon blacksdispersed in SBR compounds. (After ref. [18], Marcel Dekker, Inc.).

326

18.3.5Other Applications

A variety of electronic devices based upon dispersed particles in a polymer matrixhave been fabricated. Composites of CB in certain thermoplastics exhibit the proper-ty of a step change from low to high resistance when the device temperature exceedsa certain value. This positive temperature coefficient of resistance (PTCR) behaviorarises from Joulean heating, which causes thermal expansion and, in some cases,melting of the polymer matrix, thereby disrupting the continuity of the CB filler path-ways. When the device cools, continuity is restored [20]. This effect has been used tocreate temperature-limited heating tapes and self-resetting fuses. The microphysicsof this switching has recently been analyzed [21].

A device consisting of 15 vol. % nickel powder and 40 vol. % SiC powder dispersedin a silicone resin binder has been fabricated. Thermal curing was performed withthe aid of a peroxide initiator. The resulting device was found to switch from a highresistance state (>106 Ω) to a conductive state (1 to 10 Ω) when subjected to electro-static discharge circuit transients. This response effectively shunts transients toground in times of <25 ns [22].

One of the major applications of CB-loaded polymers is in materials capable of dis-sipating an electrostatic charge. These are referred to as ESD compounds. Accordingto the Electronics Industry Association (EIA), “conductive” materials have surface re-sistivity of < 105 Ω/sq, dissipative materials 105 to 1012 Ω/sq, and insulators >1012 Ω/sq. In practice, electrostatic dissipative materials have preferred resistivity inthe range of 106 to 109 Ω/sq [19].

As an example, a commercial ESD compound uses as little as 1 to 2 wt. % of highstructure Ketjenblack EC 600 (Akzo, Netherlands) to achieve stable resistivities of 106

to 109 Ω/sq in combination with glass fibers at loadings of 10 to 25 wt. % to create aninorganic interface at which the CB presumably concentrates [19]. The glass fibers are~3 mm long and 10 µm in diameter, giving an aspect ratio of 300. The resulting com-pound is ideal for electrostatic dissipation requirements and uses a much lower load-ing of CB than would be required without the glass fiber. Without glass fibers to pro-mote CB clustering, a loading of 25 vol. % would be typical to achieve resistivity of theorder of 108 Ω/sq.

18.4Distribution and Dispersion of Carbon Black in Polymers

18.4.1Microscopy and Morphology [8]

Carbon black was a much studied substance during the evolution of the transmissionelectron microscope. It was a tour-de-force to resolve the graphitic planes in carbonblack [23]. With the scanning electron microscope (SEM), more fully dimensionalviews became available. In recent years, with the advent of the atomic force micro-


327

scope (AFM), there have been attempts to correlate fine-scale microstructure withelectrical conductivity.

18.4.2Percolation Networks

It has been possible to directly image the percolation network at the surface of a CB−polymer composite. An early report was that of Viswanathan and Heaney [24] on CBin HDPE, in which it was shown that there are three regions of conductivity as a func-tion of the length, L, used as a metric for the image analysis. Below L = 0.6 µm, thefractal dimension, D, of the CB aggregates is 1.9 ± 0.1. Between 0.8 and 2 µm, the da-ta exhibit D = 2.6 ± 0.1, while above 3 µm, D = 3, corresponding to homogeneous be-havior. Theory predicts D = 2.53. “It is not obvious that the carbon black/polymer sys-tem should be explainable in terms of standard percolation theory, or that it shouldbe in the same universality class as three-dimensional lattice percolation problems”[24]. Subsequent experiments of this kind were carried out by Carmona [25,26].

It has been correctly realized that most microscopies, including AFM and electri-cal probes, reveal only a two-dimensional section of a three-dimensional conductingnetwork. Gubbels [27,28] applied methods of image analysis to blends consisting oftwo polymers plus CB. The frequently observed segregation of CB to the phaseboundaries or interphase was confirmed.

18.4.3CB in Multiphase Blends

It is known that in some polymers, addition of CB to levels above Vc causes a decreasein mechanical properties. This is a large effect for some polymers, but is minor forothers [3]. Tensile elongation varies with CB loading for polycarbonate (PC) andpolypropylene (PP), but is little affected in ethylene vinyl acetate (EVA) and an ethyl-ene–octene copolymer (Engage®) [3].

Both for reasons of economy and to minimize unwanted effects on mechanical be-havior, it is desirable to use the minimum concentration of CB to achieve the requiredelectrical properties. By employing polymer blends, it is possible to create mor-phologies in which the CB additive concentrates in one phase or, better, at the inter-phase. There, the CB aggregates approach each other closely and the percolationthreshold is low. Examples of systems [27–29] where such phase segregation can oc-cur are:

polyolefins with other crystallizable or non-crystallizable polymers [30–33], phase-separating block copolymers such as SBS, SEBS, etc. [34], other incompatible polymer blends [35,36].

In a non-crystallizable polymer such as atactic polystyrene, Vc is close to 8 wt. %.In semicrystalline PE it is 5 wt. %, presumably due to segregation of the CB to thenon-crystalline phase or the phase boundaries in the PE. This segregation is further

18.4 Distribution and Dispersion of Carbon Black in Polymers

328

enhanced in PE/PS blends when the composition allows for continuity of the PEphase and “double percolation” of the phases and conductive regions [27,37]. This hasbeen observed at a 45:55 (w/w) ratio of PE/PS in a melt-blended composition withmore than 0.4 wt. % (0.2 vol. %) carbon black [27]. The effect seems to depend uponthe relative interfacial tensions of the polymers and the CB in a manner consistentwith the independent observations of Miyasaka et al. [38].

18.4.4Process Effects on Dispersion

All of the above situations reiterate the fact that observed conductivities are highly de-pendent on the morphology or microstructure of the composite and on the mixingprocess by which it is obtained. Regarding mixing, see Chapter 3 of this book as wellas refs. [39–42]. Manas-Zloczower and co-workers considered that CB agglomerateswould break up into particles roughly half the size of the original under the action ofmixing stresses [43]. Another viewpoint is that agglomerates “erode” as small piecesbreak off from their surface [44]. Subsequent studies in the laboratory of Manas-Zloc-zower using a transparent cone-and-plate viscometer with f luids and polymer meltscovering a wide viscosity range showed that both mechanisms are operative. The ear-ly stages are predominately agglomerate break-up and subsequent erosion [45].

It is well known that in a process such as injection molding there are segregationeffects that inherently lead to skin-core differences in CB distribution and to morecomplex spatial distributions. Even with the simplest f lows and part shapes there canbe measurable variations of resistivity between various locations in a single part [46].

A different approach to controlling the dispersion of CB is based on a collective mo-tion of f luid elements called chaotic advection. The theory behind this mode of mix-ing has been set out by Aref [47,48]. Danescu and Zumbrunnen [49,50] have builtequipment incorporating an eccentric cavity formed between two offset cylinders. Insuch a device, the high structure CB Printex XE-2 (Degussa Corp.), pre-compoundedinto atactic polystyrene, was let down to loadings ranging from 0.4 to 2.5 wt. %. Adrop of two orders of magnitude in resistivity was observed in the concentrationrange from 0.8 to 1.0 wt. %. The conductivity threshold value, 0.8 wt. % CB, provedto be 70% lower than the percolation threshold of this system compounded by con-ventional means. Further studies have been published, including an apparatus forcontinuously extruding CB−polymer films with chaotic advection [51].

18.5Other Carbon-Based Conductive Fillers

For carbon nanotubes, which are discussed in detail in Chapter 10, conductivity isachieved at lower loadings (by weight), but these materials are difficult to disperse inmolten polymers. Methods of surface functionalization and lower-cost manufactur-ing must be developed before carbon nanotubes can be expected to find wider use asconductive fillers [52,53]. As an alternative to nanotubes, Fukushima and Drzal [54]


329

have observed conductivity thresholds of less than 3 vol. % in composites containingacid-etched or otherwise functionalized exfoliated graphite. These composites showretained or improved mechanical properties compared to other carbon-filled poly-mers.

For carbon fibers (see also Chapter 10) increasing the aspect ratio of the fillershould, predictably, lower the resistivity compared to an equiaxed filler. It is interest-ing to observe that for one such carbon fiber compounded into five thermoplastics,the resistivity vs. loading curves are identical; that is, the critical volume concentra-tion of fiber is ≈7 vol. %. This is the concentration at which the volume resistivitydrops from 1012 to 101 Ω cm [55]. Inevitably, for all thermoplastic matrices, attritionof the carbon fiber (as measured by a diminished L/D) increases the resistivity of thecomposite (Table 18-5). King and co-workers [56] have shown that still lower resistiv-ities are produced when carbon fiber is combined with other fillers such as graphitepowder or CB. It seems that the effect from increased internal interfacial area is dom-inant. Narkis et al. [19,57] showed that even glass fibers create interfacial regions inwhich added CB at low concentrations can significantly increase conductivity.

Tab. 18-5 Effects of compounding technique and screw type on thefiber length and resistivity of compounds filled with carbon fibers.(After ref. [55]).

Matrix Extruder Concen- Resistivity, Fiber length, Fiber aspect Polymer compounder tration, Ω cm mm ratio form

vol. %

ABS single-screw 0.20 0.60 0.44 55 pelletsABS twin-screw 0.20 1.70 0.24 30 pelletsPPO, Noryl® single-screw 0.25 0.630 0.18 22 pelletsPPO, Noryl® twin-screw 0.25 3.30 0.11 14 pelletsPolyamide-6,6 single-screw 0.19 0.56 0.23 29 pelletsPolyamide-6,6 twin-screw 0.19 1.04 0.13 16 pelletsPPS, Ryton® single-screw 0.21 3.40 0.15 19 powderPPS, Ryton® twin-screw 0.21 1.25 0.22 27 powder

18.6Intrinsically Conductive Polymers (ICPs)

Soon after the discovery of intrinsic conductivity in polymers, experiments were per-formed aimed at dispersing these organic materials in host polymers with a view toimproving processability and increasing stability. This quest is ongoing, althoughsome ICP-polymer blends are already being sold commercially e.g. by RTP Compa-ny, Winona, MN, and by Eeonyx Inc. of Pinole, CA. For example, polyaniline (PANI)shows improved processability and stability when conjugated to a protonic acid suchas poly(styrene sulfonic acid) or to the small molecule dodecylbenzenesulfonic acid[58,59]. In this state, it is more compatible with certain polymers. Such blends, as in-terpenetrating polymer networks, show Vc < 1% [60]. ICPs exhibit a volume resistiv-

18.4 Distribution and Dispersion of Carbon Black in Polymers

330

ity of 105 Ω cm compared to 103 to 109 Ω cm for a CB dispersion. This comparison isfor each additive in a matrix of polypropylene [61]. In PANI/poly(styrene sulfonicacid) dispersions, the particle sizes are <1 µm and thin coatings with resistivities of1–10 Ω cm can be cast as transparent films [60].

18.7Metal Particle Composites

Most polymers can be filled with metal particles to render them conductive. Howev-er, consideration must be given to specific aspects of the possible combinations [62].For example, ABS and PC are widely used for instrument, TV, and similar housings,but copper is known to promote degradation of PC so an alternative filler is oftenpreferable [63].

Polymers that undergo phase separation upon solidification may contribute to seg-regation of conductive filler to the interphase or to non-crystalline regions. This canproduce favorably high conductivities with less metal because of the local concentra-tion of metal particles. This is illustrated in Figure 18-5 [64]. Note the two cases: whenpolymer and metal particles are of comparable size (panels A and B) and when themetal particles are very much smaller (panels C and D; Figure 18-5); note also the dif-ferent microstructures for V < Vc and for V > Vc.

This effect of metal segregation is borne out by measurements of resistivity vs.vol. % metal filler, as shown in Figure 18-6. A threshold for conductivity in metal-filled polymers was first observed by Gurland [65] in a composite of silver spheres in


Fig. 18-5 Schematic illustrations of “random” (A and B) vs. “segre-gated” (C and D) particle distributions. (After ref. [63], MarcelDekker, Inc.).

331

phenol–formaldehyde resin (Bakelite®). The threshold or critical volume concentra-tion, Vc observed was 0.38.

Vc depends on particle shape but, in general, for equiaxed particles Vc is 0.4 forbroad and 0.2 for narrow particle size distributions (PSD), respectively [66]. General-ly, when equiaxed metal particles are used, a narrow PSD creates conductive path-ways at lower loadings than a broad PSD [66]. Among metal fibers, aluminum fiberswith aspect ratio 12.5:1 showed a threshold at ~12 vol. % in either a thermosettingpolyester or polypropylene. As the aspect ratio of metal fibers is increased, the thresh-old loading for conductivity decreases. For example, the incorporation of stainlesssteel fibers 6 mm long and 8 µm in diameter into ABS led to a resistivity of 0.70Ω cmat 1 vol. % loading [66,67]. Metal fibers may be coated or sized to increase adhesionto the polymer matrix.

The acoustic shielding effectiveness of various fillers vs. frequency is shown in Fig-ure 18-7 [66]. Adopting the guideline that 30 to 40 dB of attenuation will meet the ma-jority of requirements, all materials in Figure 18-7 meet this criterion up to at least50 MHz. Polymers filled with aluminum f lakes or metal fibers are effective in termsof cost and performance and may offer significant economic advantages over the la-borious methods used for painting or coating parts made of unfilled polymers. Insome instances, metal-filled polymers may be pigmented to give colored products.This may be used as a design advantage since carbon black fillers give only blackparts.

18.7 Metal Particle Composites

Fig. 18-6 Inf luence of specimen microstruc-ture and filler content on electrical resistivityfor a random vs. a segregated distribution ofmetal particles. The “random distribution”

() is 50% silver in phenol–formaldehyderesin while the segregated case () is 7% sil-ver in PVC. (After ref. [63], Marcel Dekker, Inc.).

332

Static charge decay times for carbon- or metal-filled and metal-painted polycar-bonate as compared to neat polycarbonate are shown in Table 18-6. All of thefilled/painted polycarbonates performed similarly: a charge of ±5 kV decayed in lessthan 0.06 s compared to more than 100 s for neat PC [63]. Metal-filled composites mayshow ohmic behavior in the interior but appear non-ohmic overall due to surface de-pletion of the conductive species [68].

Highly conducting composites can be produced with low concentrations of metalfillers provided that their aspect ratio is high. Fiber L/D is a significant considerationin processing such materials [66]. A summary of the effects of processing on carbonfibers in various polymer melts is presented in Table 18-5 [55]. Metal fibers are bet-ter able to withstand processing than metallized glass or carbon fibers. Stainless steelfibers at a loading of 1 vol. % produce a composite with an RF signal attenuation ofup to 50 dB. Such compounds can be successfully processed in screw extruders andmay, if desired, be colored [69].


Fig. 18-7 Shielding effectiveness of differentmaterials as a function of radio frequency: A:nickel-coated glass fiber composite; B: copperplate; C: carbon black composite; D: steel; E:

20% aluminum fiber in polyester; F: 30% alu-minum flake in polyester; G: unfilled polymer.(After ref. [63], Marcel Dekker, Inc.).

333

18.8Magnetic Fillers

In polymers filled with magnetic materials, the magnetic moment is simply propor-tional to the volume loading of magnetic particles. As an example, consider the caseof magnetite (Fe3O4), a ferrite, in polypropylene or polyamides. In practice, to makeresin-bonded magnets, relatively high filler levels are used, for example 60–80 wt. %or 25–45 vol. % [70]. A significant drop in electrical resistivity of this composite isnoted at 44 vol. % magnetite. In terms of process conditions, the maximum practicalloading of magnetite was found to be 80 wt. % when operating at 90% maximumtorque and 300 rpm using a downstream side feeder on a twin-screw extruder withL/D = 39. In these experiments, the production rate of the magnetic compound was85 kg h–1 [70].

Other magnetic particles compounded into plastics include barium ferrite, alnico,samarium cobalt (SmCo), and rare earth iron borides. An overview of magnetics ter-minology and the properties of some representative permanent magnets, i.e. ferrite,alnico, SmCo, and NdFeB, has been given by Trout [71]. Magnetic powders are gen-erally pre-compounded with the matrix polymer and then injection molded to givenear net-shape parts or stock shapes. The matrix can provide an added benefit by de-creasing corrosion effects on the magnetic fillers. A recommended general treatmentof magnetic materials and their physics may be found in the book by Cullity [72].Neodymium/iron/boron powder is a preferred material for the production of bond-ed magnets. Manufacturers make these composites by various processes: injectionmolding of ferrites such as barium ferrite, sintering of SmCo, and processing of Nd-FeB by sintering, hot pressing, and molding.

18.8 Magnetic Fillers

Tab. 18-6 Static charge decay times of selected polycarbonate-based plastics. (After ref. [63])

Material Applied voltage (%) Decay time (s)At +5 kV At –5 kV

Polycarbonate 50 > 100 >10010 > 100 > 1000 > 100 > 100

Polycarbonate coated 50 0.02 0.02with nickel paint 10 0.03 0.04

0 0.05 0.05Polycarbonate filled 50 0.02 0.02with 30% metal 10 0.04 0.03

0 0.06 0.05Polycarbonate filled 50 0.02 0.02with 40% carbon 10 0.04 0.03

0 0.05 0.05

334

In the manufacture of magnetic tape and f loppy disks, the magnetic filler is dis-persed in a concentrated polymer solution, which is coated onto a substrate and ori-ented by an externally applied magnetic field before the composite coating “sets” bydrying or cross-linking. Orientation is very important for recording media in order toachieve high magnetic remanence, which aids in avoiding inadvertent demagnetiza-tion. Some of the technology for preparing magnetic recording media has beendescribed [73] but much is withheld as trade secrets.


Real composite materials are often far from homogeneous in their properties or mi-crostructure. In some instances, an immediate technical problem may be dealt withby adjusting one parameter or another, but no fundamental discriminating meas-urements can be made on such systems. Therefore, it is to be hoped that future stud-ies will describe the polymers and fillers in considerable detail, that the mixing andprocessing variables will be annotated, and that the microstructures of the resultingcomposites will be characterized. Only then will meaningful measurements andcomparisons be possible among these various filled plastics, which have such inter-esting electrical and magnetic properties.

Acknowledgements

The author appreciates receiving bibliographical assistance from Mr. Bruce Slutskyof the Van Houten Library at NJIT and Ms. K. Fitzgerald.

Appendix: Measurements of Resistivity

One distinguishes “bulk” or volume resistivity (Ω cm), in which the dimensions ofthe conductor are considered, from surface resistivity (Ω/square). Definitions andmethods can be found in ASTM D4496-87 and BS 2044 for volume or “bulk” resis-tivity and in IEC 167 and AFNOR C26-215 for surface resistivity. Figure 18-8 showsmeasurement set-ups and illustrates the concepts involved [74]. Table 18-7 classifiesplastic compounds according to their resistivity level.


33518.9 Concluding Remarks

Fig. 18-8 Set-ups for measuring surface and bulk resistivity. (After ref. [74]).

Tab. 18-7 Resistivity classification of conductive thermoplastics(after refs. [3,19,69]).

Bulk resistivity, Ω cm Surface resistivity, Ω/sq

Undoped 1014 to 1016 1012 to 1014

Antistatic 109 to 1014

Dissipative 105 to 109 106 to 109

Conductive 100 to 105

EMI Shielding < 1 < 106

336

References


1 Sichel, E. K. (Ed.), Carbon Black-PolymerComposites: The Physics of Electrically Con-ducting Composites, Marcel Dekker, NewYork, 1982.

2 Rupprecht, L. (Ed.), Conductive Polymers andPlastics in Industrial Applications, PlasticsDesign Library, Norwich, NY, 1999.

3 Huang, J.-C., Adv. Polym. Technol. 2002,21(4), 299–313.

4 Donnet, J.-B., Bansal, R. C., Wang, M.-J.,(Eds.), Carbon Black: Science and Technology,2nd Ed., Marcel Dekker, Inc., New York,1993.

5 Foster, J. K., “Effects of Carbon Black Prop-erties on Conductive Coatings”, presentedat the 2nd International Exhibition of PaintIndustry Suppliers, São Paulo, Brazil, 1991.On the Web at http://www.cabotcorp.com/

6 Kühner, G., Voll, M., Manufacture of CarbonBlack, Chapter 1 in ref. [4] op. cit.

7 Medalia, A. I., Chapter 15 in Mixing andCompounding of Polymers: Theory and Prac-tice (Eds.: Manas-Zloczower, I., Tadmor, Z.),Hanser Publishers, Munich and Cincinnati,1994.

8 Hess, W. M, Herd, C. R., Microstructure,Morphology, and General Physical Properties,Chapter 3 in ref. [4] op. cit.

9 Dannenberg, E. M., Carbon Black, inKirk–Othmer Encyclopedia of ChemicalTechnology, 3rd Edition, John Wiley, NewYork, 1978, vol. 4, pp 631–666.

10 Bansal, R. C., Donnet, J.-B., Mechanism ofCarbon Black Formation, Chapter 2 inref. [4] op. cit.

11 Cabot Corporation, “Carbon Black User’sGuide: Safety, Health, and EnvironmentalInformation,” on the websitehttp://www.cabot-corp.com/

12 Pantea, D., et al., Appl. Surface Sci. 2003,217, 181–193.

13 Janzen, J., J. Appl. Phys. 1975, 46(2),966–969.

14 Sichel, E., et al., Phys. Rev. 1978, B18(10),5712–5716.

15 Balberg, I., Phys. Rev. 1998, B57, 13351.16 Rubin, Z., et al., Phys. Rev. 1999, B59(19),

12196–12199.17 Balberg, I., Carbon 2002, 40, 139–143.18 Probst, N., “Conducting Carbon Black”,

Chapter 8 in ref. [4] op. cit.

19 Narkis, M., et al., Novel Electrically Conduc-tive Injection-Moldable Thermoplastic Com-posities for ESD Applications, in ref. [2] op.cit.

20 Doljack, F. A., IEEE Trans. Components, Hy-brids, and Mfg. Tech. 1981, Vol. CHMT-4, 4,372–378.

21 Azulay, D., et al., Phys. Rev. Lett. 2003, 90,236601.

22 Rector, L., Hyatt, H., Proc. 56th SPEANTEC, 1998, 1381–1385.

23 Heidenreich, R. D., et al., J. Appl. Cryst.1968, 1, 1.

24 Viswanathan, R., Heaney, M. B., Phys. Rev.Lett. 1995, 75(24), 4433–4436 and errata inPhys. Rev. Lett. 1996, 76(19), 3661.

25 Carmona, F., Ravier, J., Carbon 2002, 40,151–156.

26 Carmona, F., Ravier, J., Physica B 2003, 338,247–251.

27 Gubbels, F., Macromolecules 1994, 27,1972–1974.

28 Gubbels, F., et al., Macromolecules 1995, 28,1559–1566.

29 Knackstedt, M. A., Roberts, A. P., Macro-molecules 1996, 29, 1369–1371.

30 Tchoudakov, R., et al., Polym. Eng. Sci. 1996,36, 1336.

31 Thongruang, W., et al., Polymer 2002, 43,2279–2286.

32 Thongruang, W., et al., Polymer 2002, 43,3717–3725.

33 Thongruang, W., et al., J. Polym. Sci.: PartB: Polymer Physics 2002, 40, 1013.

34 Tchoudakov, R., et al. pp. 51–56, in ref. [2]op. cit.

35 Zhang, M. Q., et al., Macromolecules 1998,31, 6724–6726.

36 Feng, J., et al., Polym. Eng. Sci. 2003, 43,1058–1063.

37 Levon, K., et al., Macromolecules 1993, 26,4061–4063.

38 Miyasaka, K., et al., J. Mater. Sci. 1982, 17,1610–1616.

39 Todd, D. B. (Ed.), Plastics Compounding:Equipment and Processing, Hanser Publish-ers, Munich, 1998.

40 Hornsby, P. R., Advances in Polymer Science1999, 139, 155–217.

41 Manas-Zloczower, I., Tadmor, Z. (Eds.),Mixing and Compounding of Polymers: Theo-

337References

ry and Practice, Hanser Publishers, Munichand Cincinnati, 1994.

42 White, J. L., et al. (Eds.), Polymer Mixing:Technology and Engineering, Hanser Publish-ers, Munich and Cincinnati, 2001.

43 Manas-Zloczower, I., et al., Rubber Chem.Technol. 1982, 55, 1250.

44 Shiga, S., Furuta, M., Rubber Chem. Technol.1985, 58, 1.

45 Hong, C. M., et al., Proc. 61st SPE ANTEC,2003, 1625–1629.

46 Jana, S. C., Polym. Eng. Sci. 2003, 43, 570.47 Aref, H., J. Fluid Mech. 1984, 143, 1.48 Aref, H., Physics of Fluids 2002, 14, 1315.49 Danescu, R. I., Zumbrunnen, D. A., pp.

77–83 in ref. [2] op. cit.50 Danescu, R. I., Zumbrunnen, D. A., pp.

85–91 in ref. [2] op. cit.51 Chougule, V. A., Zumbrunnen, D. A., Proc.

61st SPE ANTEC, 2003, 1299–1303.52 Potschke, P., et al., Polymer 2003, 44, 8061.53 Sandler, J. K. W., et al. Polymer 2003, 44,

5893.54 Fukushima, H., Drzal, L. T., Proc. 61st SPE

ANTEC, 2003, 2230–2234.55 Bigg, D. M., Adv. Polymer Technol. 1984,

4(3/4), 255.56 Heiser, J. A., et al., Adv. Polym. Technol.

2004, 23(2), 135–146.57 Narkis, M., et al., Proc. 56th SPE ANTEC,

1998, 1375–1380.58 Haba, Y., et al., Synthetic Metals 2000, 110,

189–193.59 Segal, E., et al., J. Polym. Sci.: Part B: Polym.

Physics 2001, 39, 611–621.60 Heeger, A. J., Synthetic Metals 2002, 125,

23–42.61 Dahman, S., “All Polymeric Compounds:

Conductive and Dissipative Polymers in ESD

Control Materials,” Proc. of the EOS/ESDSymposium, Las Vegas, 2003.

62 Bhattacharya, S. K. (Ed.), Metal-Filled Poly-mers: Properties and Applications, MarcelDekker, Inc., New York, 1986.

63 Kusy, R. P., Applications, Chapter 1 inref. [62], op. cit.

64 Kusy, R. P., J. Appl. Phys. 1977, 48, 5301.65 Gurland, J., Trans. Metal. Soc AIME, 1966,

236, 642.66 Bigg, D. M., Electrical Properties of Metal-

Filled Polymer Composites, Chapter 3 inref. [62], op. cit.

67 Bigg, D. M., Stutz, D. E., Polymer Composites1983, 4, 40.

68 Reboul, J. P., Nonmetallic Fillers, Chapter 6in ref. [62], op. cit.

69 Murphy, J., Plastics Additives &Compounding, 2000, 2(9), 22–27.

70 Duifhuis, P. L., Janssen, J. M. H., PlasticsAdditives and Compounding 2001, 3(11),14–17.

71 Trout, S. R., Understanding Permanent Mag-net Materials: An Attempt at Universal Mag-netic Literacy, Paper presented at “CoilWinding 2000” Conf. Available on the In-ternet athttp://spontaneousmaterials.com/Papers/CoilWinding2000.pdf

72 Cullity, B. D., Introduction to Magnetic Mate-rials, Addison-Wesley, Reading, MA, 1972.

73 Mee, C., Daniel, E., Magnetic Recording, Mc-Graw-Hill, New York, 1988.

74 Keithley Instruments Co., Electrometer Mea-surements, 2nd Edition, Cleveland, OH,1977.

Author’s note. Mention of a supplier, product, orprocess in this chapter does not constitute an en-dorsement or recommendation.

19Surface Property Modif iers

Subhash H. Patel

19.1Introduction

A variety of organic/inorganic additives, in either solid or liquid form, are incorpo-rated into plastic articles to achieve desired surface property modifications. Depend-ing upon the specific function they perform in modifying the surface property, theymay be categorized as follows:

1. Lubricants Prevent sticking to processing equipment2. Anti-blocking and slip agents Prevent sheet and film sticking3. Anti-fogging agents Disperse moisture droplets on films4. Coupling agents Enhance interfacial bonding between the

filler and the polymer matrix5. Anti-static agents Prevent static charge build-up on surfaces6. Wetting agents Stabilize filler dispersions

In the paints and coatings industry, surface modifiers are used to modify the ap-pearance or to improve the performance characteristics of a cured film. Typical per-formance features include anti-blocking, slip, abrasion resistance, matting, andscratch/mar resistance [1].

It would be beyond the scope of this chapter to discuss all the organic/inorganic ad-ditives used by the plastics industry for surface property modification. Thus, the dis-cussion is limited to two types of additives meeting the definition of functional fillersas used in this book. These are: a) inorganic or organic lubricants/tribological modi-fiers, and b) anti-blocking inorganic fillers. The emphasis of the chapter is placed onthe first type of fillers. Fillers of the second type, the anti-blocking function of whichis not covered in other chapters of the book, are brief ly presented. Fillers improvingscratch/mar resistance as a secondary functionality are included in other chapters ofthe book.


340

19.2Solid Lubricants/Tribological Additives

Different types of solid materials most widely used at the present time include [2,3]:

1. Specific fillers (layer-lattice solids): molybdenite or molysulfide (MoS2) andgraphite.

2. Polymers: PTFE (polytetraf luoroethylene), polychlorof luoroethylene, silicones.3. Less frequently used materials: ceramics, e.g., BN (boron nitride), aramid or car-

bon fibers; miscellaneous, e.g. calcium f luoride, cerium f luoride, tungsten disul-fide, mica, borax, silver sulfate, cadmium iodide, lead iodide, and talc.

Among the above-listed materials, MoS2 and graphite are the predominant solid lu-bricants. In dry powder form, these are effective lubricants due to their lamellarstructures. The lamellae orient parallel to the surface in the direction of motion asshown in Figure 19-1 [3].

Such lamellar structures are even able to prevent contact between highly loadedstationary surfaces. In the direction of motion, the lamellae easily shear over eachother resulting in low friction. While larger particles perform best on relatively roughsurfaces at lower speeds, finer particles perform best on relatively smooth surfacesand at higher speeds. A comparison of various solid lubricants with respect to theircoefficients of friction is shown in Figure 19-2 [4].

19 Surface Property Modif iers

Particle orientation after initial sliding

Fig. 19-1 Schematic of solid lubrication mechanism.Reproduced with permission from ref. [3].

Fig. 19-2 Comparison of coefficient of friction of various solidlubricant powders. Reproduced with permission from ref. [4].

341

Solid/dry lubricants are useful under conditions when conventional liquid lubri-cants are inadequate [3], namely high temperatures, conditions of reciprocating mo-tion, and extreme contact pressures.

19.2.1General

19.2.1.1 MolybdeniteMolybdenite or molybdenum disulfide (MoS2) or Molysulfide is a mineral [5] foundin granites, syenites, gneisses, and crystalline limestones. A NIOSH (National Insti-tute for Occupational Safety and Health) website [6] has listed a total of 31 synonymsfor molybdenite. The use of molybdenite as a lubricant was apparently recorded inthe early 17th century by John Andrew Cramer [7]. Chrysler, an automotive manu-facturer, was the first to widely use MoS2 grease in the 1960s.

19.2.1.2 GraphiteGraphite is one of the oldest and most widely used (due to its lower price comparedto MoS2 or BN) solid lubricants. It is obtained both as a soft mineral and as a man-made (synthetic) product. Graphite powders are used as solid lubricants in threeways: 1) in dry films, 2) as an additive in liquids (oils) or semi-solids (greases), and 3)as a component of self-lubricating (internally lubricated) composites. A NIOSH web-site [8] has listed a total of 55 synonyms for graphite.

19.2.1.3 Polytetraf luoroethylene (PTFE)PTFE is a perf luorinated, straight chain, high molecular weight synthetic polymer[9]. In contrast to most inorganic functional fillers, PTFE is an organic filler having aunique combination of high thermal and chemical resistance together with the low-est friction coefficient of any known internal lubricant, high purity, and dielectricproperties. The features and benefits of PTFE include excellent slip, anti-blocking,improved stability against polishing, and improved abrasion, scratch, mar, and scuffresistance [1]. There are a variety of synonyms and trade names for PTFE, withTef lonTM being the most well known.

19.2.1.4 Boron Nitride (BN)BN is a synthetic, high temperature, white solid lubricant, used for parts that mustbe highly resistant to wear. Although it was discovered in the early 19th century, it wasnot developed as a commercial material until the latter half of the 20th century [10].BN is often referred to as “white graphite” because it is a lubricious material with thesame plate-like hexagonal crystal structure as black graphite [11]. In the same waythat carbon exists as graphite and diamond, BN can be synthesized in hexagonal (softlike graphite) and cubic (hard like diamond) crystal forms. A NIOSH website [12] haslisted a total of 24 synonyms for BN.

19.2 Solid Lubricants/Tribological Additives

342

19.2.2Production

19.2.2.1 MolybdeniteBy far the majority of the world production of MoS2 comes from the USA, Chile,Canada, and China. Smaller quantities are mined in the rest of the world, includingMexico, Australia, South Korea, Namibia, and several European countries.

There are three types of mines/ore bodies from which molybdenite can be recov-ered [13]:

1. Primary mines, from which solely molybdenite is recovered.2. By-product mines, from which molybdenite is recovered during copper recovery.3. Co-product mines, from which both molybdenite and copper-bearing minerals are

recovered.

The raw ore is pulverized using a series of crushers and rotating ball and/or rodmills to fine particles. This liberates the molybdenite from its host rock. The productis then beneficiated by f lotation separation, subsequent regrinding, and ref lotationto increase the molybdenite content of the new concentrate stream by steadily re-moving unwanted material. The final concentrate may contain 70−90% molybdenite.An acidic leach may be employed to dissolve copper and lead impurities, if required.A schematic of the production of molybdenum compounds, including MoS2, can befound in ref. [13].

19.2.2.2 GraphiteGraphite is considered an “archaic” industrial mineral since it has been mined for itsuseful properties (lubrication, pigmentation, writing, etc.) for thousands of years [14].There are two types of graphite used in industry: natural and synthetic.

Natural graphite is obtained in three commercially used varieties: f lake, crystallinevein, and amorphous graphite [14−17]. Most f lake graphite is formed in a metamorphicgeological environment by the heat- and pressure-induced transformation of dis-persed organic material. Flake graphite is removed from its enclosing “ore” rock bycrushing the rock and separating the graphite f lakes by froth f lotation. “Run ofmine” graphite is available in 80−98% carbon purity ranges. However, most proces-sors are also capable of supplying 99% carbon f lake graphite through various post-f lotation purification methods. The impurities in f lake graphite are virtually identi-cal in composition to the enclosing rock.

Crystalline vein graphite is unique, as it is believed to be naturally occurring py-rolytic (deposited from a f luid phase) graphite. Vein graphite gets its name from thefact that it is found in veins and fissures in the enclosing “ore” rock. This variety isformed from the direct deposition of solid, graphitic carbon from subterranean, high-temperature pegmatitic f luids. It typically shows needle-like macro-morphology andf lake-like micro-morphology. Due to the natural f luid-to-solid deposition process,vein graphite deposits are typically more than 90% pure, with some actually reaching99.5% graphitic carbon in the “as-found” state. Vein graphite is mined using con-


343

ventional shaft or surface methods. Although several small deposits of this type ofgraphite are known to exist worldwide, Sri Lanka is the only area presently producingcommercially viable quantities of this unique mineral.

Most commercial grade amorphous graphite is formed from the contact or regionalmetamorphism of anthracite coal. It is considered a seam mineral and is extracted us-ing conventional, coal-type mining techniques. Synthetic graphite, also known as “ar-tificial graphite,” is a man-made product. Synthetic graphite is manufactured by heat-treating amorphous carbons, i.e. calcined petroleum coke, pitch coke, etc., in a re-ducing atmosphere at temperatures above 2500 °C. At high temperatures, the “pre-graphitic” structures present in these “graphitizable carbons” become aligned inthree dimensions. The result is the transformation of a two-dimensionally orderedamorphous carbon into a three-dimensionally ordered crystalline carbon. Feedstocksfor synthetic graphite production are chosen from product streams that have a highconcentration of polynuclear aromatic structures that can coalesce to graphene layersunder the inf luence of heat.

19.2.2.3 Polytetraf luoroethylene (PTFE)Commercially, PTFE is produced from the monomer tetraf luoroethylene by two dif-ferent polymerization techniques, viz., suspension and emulsion polymerization.These processes give two vastly different physical forms of chemically identical PTFE.While suspension polymerization produces granular PTFE resin, emulsion poly-merization produces an aqueous PTFE dispersion and PTFE fine powders (after co-agulating the dispersion).

Fine powder PTFE resins, which are relevant to this chapter, are made by variationsof the emulsion polymerization technique. It is extremely important that the disper-sion retains its integrity during polymerization, but is subsequently sufficiently un-stable to be able to coagulate into a fine powder [18]. Emulsion polymerization latexhas an average particle size of about 150−300 nm. However, by using low reactionconversion it is possible to obtain 100 nm particles [19]. In addition, through the tech-nology of perf luorinated microemulsion polymerization [20,21], it is possible to ob-tain particles in the size range 10−100 nm. Suspension polymer is obtained as reac-tor beads with dimensions of a few millimeters, which, after post-treatment andmilling, may be ground to micron particle size [22].

19.2.2.4 Boron Nitride (BN)BN has at least four crystal structures, viz., hexagonal (h-BN), cubic or zinc blende orsphalerite (c-BN), wurtzite (w-BN), and rhombohedral (r-BN). Of these, the first twoare commercially important, and only h-BN is used as a solid lubricant. Three majorroutes in current use for the production of h-BN are:

B2O3 + 2NH3 → 2BN + 3H2O (T = 900 °C)B2O3 + CO(NH2)2 → 2BN + CO2 + 2H2O (T > 1000 °C)B2O3 + 3CaB6 + 10N2 → 20BN + 3CaO (T > 1500 °C)


344

These processes yield refractory grades with 92−95% BN and 5−7% B2O3. The B2O3

is removed by evaporation in a second step by reheating to >1500 °C to obtain ceramicgrade with >98.5% BN. The c-BN is usually prepared from the hexagonal form at highpressures (4−6 GPa or 40–60 kbar) and temperatures (1400–1700 °C) in the presenceof lithium or magnesium nitride catalysts.

19.2.3Structure/Properties

19.2.3.1 MolybdeniteMolybdenite, or “Moly Ore” as it is sometimes called, is a very soft, very high luster,metallic mineral, which could be easily confused with graphite [23]. Whereas graphitehas a darker black-silver color and a black-gray or brown-gray streak, molybdenite hasa bluish-silver color and streak. Color pictures of molybdenite from various parts ofthe world can be seen at various websites [23−32].

Figure 19-3 [3] shows a comparison of the crystal structures of MoS2 and graphite.Molybdenite’s hexagonal crystal structure is composed of molybdenum ions sand-wiched between layers of sulfur ions. The sulfur layers are strongly bonded to themolybdenum, but are not strongly bonded to other sulfur layers, rendering it soft-ness, easy shear and perfect cleavage [23]. Thus, the principle of action of molybden-ite as a dry lubricant is based on the formation of bonds between the metal and sul-fur. These bonds slip under shear forces, but are continuously reformed, therebyholding the lubricating film on the surface [33].

Salient physical and chemical properties of molybdenite with regard to its use as afiller are summarized in Table 19-1. Its lubrication performance often exceeds that ofgraphite, and it is also effective in vacuo or under an inert atmosphere up to 1200 °C,temperatures at which graphite cannot be used. The temperature limit of 400 °C for


MoS2 structure Graphite structureS

S

S

S

Mo

Mo

Fig. 19-3 Comparison of crystal structures of MoS2 and graphite.Reproduced with permission from ref. [3].

345

the use of molybdenite in air is imposed by oxidation. It begins to sublime at 450 °C.It is insoluble in water, dilute acids, and most organic solvents. When used as a lu-bricant, the particle size should be matched to the surface roughness of the plastic ormetal substrate. Large particles may result in excessive wear by abrasion caused byimpurities, while small particles may promote accelerated oxidation [3].

Tab. 19-1 Comparison of physical/chemical properties of MoS2,graphite, PTFE, and BN [3,7,9,11,23,26,27,31,32,34,35,38,41,42,50,91–98].

Property Molybdenite Graphite PTFE Boron nitride

Chemical formula MoS2 C -(-CF2-CF2-)- BN

CAS # 1317-33-5 7782-42-5 9002-84-0 10043-11-5

Molecular weight 160.07 12.01 Up to 107 24.82

Color Lead gray/bluish Black silver White-to-trans- Whitegray lucent

Luster Metallic Metallic to dull – –

Streak Green to bluish- Black gray to – –gray brownish gray

Magnetism Nonmagnetic Strongly Nonmagnetic Nonmagneticdiamagnetic

Crystal system Hexagonal Hexagonal Various forms, Cubic (abrasive)e.g. trigonal, Hexagonal hexagonal (lubricant)

Cleavage Perfect basal, Perfect in one – –easy to remove directionsmall f lakes

Crystal features Crystals are Thin f lakes are From triclinic to Layered or zinc f lexible, but not f lexible but disordered blendeelastic, greasy feel inelastic hexagonal

at 19 °C

Water absorption, – 0.5–3.0 Nil (<0.01%) 0.0–1.0% @ r.t.

Density (g cm–3) 4.7–5.0 1.4–2.4 2.14–2.20 2.27 (hexagonal)3.48 (cubic)

Mohs hardness 1–1.5 1–2 50–60 (Shore-D) 2.0

Melting point, °C 1185, sublimes Withstands 327, at >400 2700–3000 at 450 temps. up to appreciable (melting)

2820 decomposition 3000 (sublimation)2700 (dissociation in vacuum)

Onset of oxidation 360 450 >400 >2000in air, °C


346

Tab. 19-1 Continued.

Property Molybdenite Graphite PTFE Boron nitride

Maximum 420 550 260 1200 (oxidizing operating temp., atm.)°C (without 3000 (inert atm.)oxygen ingress)

Minimum –180 –20 –200 –operating temp., °C

Products of MoO2, MoO3 CO, CO2 Mainly monomer, At >2200 °C, boron oxidation, trif luoroacetate, oxides, nitrogen, decomposition hexaf luoropro- f luoride fumes.

pene, mono- and With strong oxi-dif luoroacetic dizer, may produce acid, etc. ammonia

Solubility Soluble in hot Soluble in No solvent at Degrades in hot H2SO4, aqua molten iron room temperature conc. alkali. Not regia, HNO3. wetted by molten Insoluble in water, metalsdilute acids, and most solvents

Resistance to Good Very good Excellent Goodchemicals

Resistance to Poor Good Good Goodcorrosion

Thermal con- 0.13–0.19 At 273 K, 160 0.20 20 (at r.t.)ductivity, (natural), 80 W m–1 K–1 (parallel to c-axis),

250 (perpen-dicular to c-axis)

Coeff. of ther- 10.7 Overall: 7.8 160 (from 25 to 0.46 (perpen-mal exp. × 10– 6, (293 K); 8.9 100 °C) dicular)K–1 (500 K) 0.6 (parallel)

Electrical resist- – 1.2 × 10– 6 – 1011 (at r.t.)ivity, ohm m (natural)

Friction 0.03–0.06 0.08–0.10 0.04 (dynamic)coefficient 0.09 (static) 0.12

19.2.3.2 GraphiteGraphite and diamond, polymorphs of carbon, share the same chemistry, but havevery different structures (cubic for diamond vs. hexagonal for graphite) and proper-ties. While diamond is the hardest mineral known to man, graphite is one of the soft-est. Diamond is the ultimate abrasive and an excellent electrical insulator, whereasgraphite is a very good lubricant and a good conductor of electricity [34]. Table 19-1contains representative properties of graphite of relevance to its use as a filler.


347

Due to its impervious laminar structure, f lake graphite is an effective coating ad-ditive and a barrier filler in plastics. When properly dispersed, overlapping graphitelamellae form a tough, impervious coating, which is not only lubricious but also in-ert and both electrically and thermally conductive. Also, f lake graphite is non-photo-reactive and is not bleached or affected by ultraviolet light. Figure 19-4 shows a f lakegraphite pinacoid surface [15]. The morphology of f lake graphite is consistently lam-inar regardless of particle size.

Also known as “expandable graphite”, intumescent f lake graphite is a synthesizedintercalation compound of graphite that expands or exfoliates when heated. This ma-terial is manufactured by treating f lake graphite with various types of intercalationreagents, which migrate between the graphene layers. If exposed to a rapid increasein temperature, these intercalation compounds decompose into gaseous products,which results in high inter-graphene layer pressure that pushes apart the graphitebasal planes. The result is an increase in the volume of the graphite of up to 300times, a lowering of bulk density, and an approximately tenfold increase in surfacearea. Intumescent f lake graphite may be used as a fire-suppressant additive. Its fire-suppressant function may be affected by mechanisms associated with: a) the forma-tion of a char layer (see also Chapter 17), b) the endothermic exfoliation, which effec-tively removes heat from the source, and c) out-gassing from decomposing intercala-tion reagents that displace oxygen in an advancing f lame front.

In polymer applications, crystalline vein graphite, in addition to lubricity, may offersuperior performance since it has slightly higher thermal and electrical conductivity,which result from its high degree of crystalline perfection and good oxidation resist-ance. Amorphous graphite is the least “graphitic” of the natural graphites. However,the term “amorphous” is a misnomer since this material is truly crystalline. Thisgraphite variety is “massive” with a microcrystalline structure (anhedral), as opposed


Fig. 19-4 Scanning electron micrograph (SEM) of f lake graphitepinacoid structure. Reproduced with permission from ref. [15].

348

to f lake and vein, both of which have relatively large, visible crystals (euhedral). Amor-phous graphite is typically lower in purity than other natural graphites due to the in-timate contact between the graphite “microcrystals” and the mineral ash with whichit is associated. Amorphous graphite tends to be much less ref lective in both large andsmall particle sizes. Therefore, it has a darker color, bordering on black, while othernatural graphites have a color closer to “silver-gray.”

The morphology of synthetic graphite is generally a function of particle size. Forparticles larger than about 20 µm, the macroscopic morphology is very similar to thatof the coke feed used to manufacture the graphite. However, as size is reduced belowabout 20 µm, the basic f laky structure common to all graphites becomes apparent inthe primary particle.

19.2.3.3 Polytetraf luoroethylene (PTFE)PTFE is generally considered as a thermoplastic polymer, retaining a very high vis-cosity at 327 °C. It can be employed at any temperature from −200 °C to +260 °C [9].The following properties and those shown in Table 19-1 are important if PTFE is tobe used as a functional filler.

Thermal PropertiesPTFE is one of the most thermally stable plastic materials. At 260 °C, it shows mini-mal decomposition and retains most of its properties; appreciable decomposition be-gins only at above 400 °C. The arrangement of the PTFE molecules (crystalline struc-ture) varies with temperature. There are various transition points, with the most im-portant ones being that at 19 °C, corresponding to crystal disordering relaxation, andthat at 327 °C, which corresponds to the disappearance of the crystalline structure;PTFE then assumes an amorphous aspect conserving its own geometric form. Thelinear thermal expansion coefficient varies with temperature. By contrast, thermalconductivity of PTFE does not vary with temperature and is relatively low, so that thematerial can be considered to be a good insulating material.

Surface PropertiesThe molecular configuration of PTFE imparts its surfaces with a high degree of an-ti-adhesiveness, and for the same reason these surfaces are hardly wettable. PTFEpossesses the lowest friction coefficients of all solid materials, between 0.05 and 0.09.The static and dynamic friction coefficients are almost equal, so that there is noseizure or stick-slip action. Wear depends upon the condition and type of the othersliding surface and obviously depends upon the speed and loads.

Mechanical PropertiesPTFE maintains its tensile, compressive, and impact properties over a broad tem-perature range. Hence, it can be used continuously at temperatures up to 260 °C,while still retaining a certain compressive plasticity at temperatures near absolute ze-ro. PTFE is quite f lexible and does not break when subjected to stresses of 0.7 MPaaccording to ASTM D-790. Its f lexural modulus is about 350 to 650 MPa at room tem-perature, about 2000 MPa at -80 °C, and about 45 MPa at 260 °C. The Shore D hard-


349

ness, measured according to ASTM D-2240, has values between D50 and D60. PTFEexhibits “plastic memory”, i.e. if subjected to tensile or compression stresses belowthe yield point, part of the resulting deformations remain after discontinuance of thestresses. If the piece is reheated, the induced strains tend to release themselves with-in the piece, which resumes its original form.

Environmental ResistancePTFE is practically inert towards known elements and compounds. It is attacked on-ly by the alkaline metals in their elemental state, by chlorine trif luoride, and by ele-mental f luorine at high temperatures and pressures. PTFE is insoluble in almost allsolvents at temperatures up to about 300 °C. Fluorinated hydrocarbons cause a cer-tain swelling, which is, however, reversible; some highly f luorinated oils, at temper-atures over 300 °C, exercise a certain dissolving effect. Resistance to high-energy ra-diation is rather poor.

Electrical PropertiesThe dielectric strength of PTFE varies with the thickness of the sample and decreas-es with increasing frequency. It remains practically constant up to 300 °C and doesnot vary even after prolonged exposure to high temperatures (6 months at 300 °C).PTFE has very low dielectric constant and dissipation factor values, which remain un-changed up to 300 °C in a frequency field of up to 10 GHz, even after a prolongedthermal treatment.

19.2.3.4 Boron Nitride (BN)Hexagonal BN powder exhibits the same characteristics of solid lubricants as seen forgraphite and molybdenum disulfide. These include crystalline structure, low shearstrength, adherence of the solid lubricant film, low abrasivity, and thermochemicalstability [3]. In many instances, (h)BN exceeds the performance levels of these con-ventional solid lubricants, particularly in its adherence and thermochemical stability.Figure 19-5 [35] shows a typical scanning electron micrograph of a commercial BNpowder.

Table 19-2 shows the effect of temperature of synthesis of BN on various proper-ties, viz., surface area, crystallinity, coefficient of friction, and oxygen content [35]. Itmay be noted that with increase in synthesis temperature, the coefficient of frictiondecreases. Typical properties of (h)BN relevant to its use as a filler are summarizedin Table 19-1.

Figure 19-6 [35,36] compares the coefficient of friction at various temperatures forgraphite, MoS2, talc, and (h)BN. In contrast to (h)BN, graphite and MoS2 undergo ma-jor increases in coefficient of friction (lose their lubricity) between 400 °C and 500 °C;talc shows increases in coefficient of friction at much lower temperatures. The abili-ty to retain lubricity at elevated temperatures is an important characteristic of (h)BN.Inorganic ceramic materials such as (h)BN have inherent advantages over polymerssuch as PTFE and other low-melting-point materials. BN has an oxidation thresholdof approximately 850 °C, and the rate of reaction is negligible even up to 1000 °C.


350

Tab. 19-2 Effect of synthesis temperature on properties of (h)BN [35].

Temperature of 800 1400 1900 2000Synthesis, °C

Surface area, m2 g–1 50–100 20–50 10–20 <10Crystallinity Turbostatic Quasi turbostatic Meso graphitic GraphiticCoefficient of friction 0.6 0.4 0.2–0.3 0.15Oxygen content, % >5 1.5–5 0.5–1.5 <0.5


Fig. 19-5 Scanning electron micrograph (SEM) of BN powder (GEAdvanced Ceramics-Grade: AC 6004). Reproduced with permissionfrom ref. [35].

1.2

1.0

0.8

0.6

0.4

0.2

0 100 200 300 400 500 600 700

Centigrade

Fric

tion

coef

ficie

nt µ

Fig. 19-6 Changes in coefficient of friction as a function of temper-ature for different solid lubricants. Reproduced with permissionfrom ref. [35].

351

19.2.4 Suppliers/ManufacturersThe website of the International Molybdenum Association (IMOA) [37] includes a listof the IMOA member companies that offer “unroasted” Mo concentrates containingmolybdenite. Table 19-3 lists suppliers/manufacturers of MoS2. Table 19-4 lists sup-pliers/manufacturers of different types and grades of graphite. Tables 19-5 and 19-6list major suppliers/manufacturers of BN and PTFE powders, respectively.

Tab. 19-3 Major suppliers/manufacturers of MoS2.

Supplier/Manufacturer Grades

AML Industries, Inc. Amlube 510 (technical grade)Warren, OH, USA Amlube 511 (fine technical grade)Fax: 1-330-399-5005

Climax Molybdenum Company TechnicalUSA, Germany, United Kingdom, Tokyo, Technical fineThe Netherlands Super fine (suspension)[email protected]

EM Corporation E-4: purified MoS2 powderWest Lafayette, IN, USA Parma-Silk: mixtures of MoS2 and graphiteTel.: 1-800-428-7802

Jinduicheng Molybdenum Mining Corp., Reagent gradeGermany, Japan, P.R. China, USA Super fine gradeE-mail: [email protected] Grade 1E-mail: [email protected]: [email protected]

Langeloth Metallurgical Company, LLC Large particle grade 40.0 µm Langeloth, PA, USA Technical grade 10.0 µm E-mail: [email protected] Technical fine grade 3.0 µm

Super fine grade 1.6 µm

Thompson Creek Metals Company High performance molybdenum (HPM), Englewood, CO, USA MoS2 of exceptionally high quality available in Fax: 303-761-7420 different grades

Tab. 19-4 Major suppliers/manufacturers of graphite.

Producer / Supplier Grades/Types

AML Industries, Inc. Natural graphite powder: AmLube 611Warren, OH, USA High purity graphite powder: AmLube 610, 613Fax: (330) 399 5005

Applied Carbon Technology Natural graphite: Grades A–HSomerville, NJ, USA Synthetic graphite: L101(very pure); J101 (high ash)Phone: (908) 707-0807 Amorphous graphite: P100 & P103

Asbury Graphite Mills, Inc. Flake graphite, amorphous graphite, vein graphiteAsbury, NJ, USA synthetic graphite, intumescent f lake graphite www.asbury.com (expandable graphite)


352

Tab. 19-4 Continued

Producer / Supplier Grades/Types

Superior Graphite Co. ThermoPURE Products:Chicago, IL, USA Purified crystalline f lake and crystalline vein graphite: Fax: +1 312 559 9064 carbon (LOI): 99.7–99.9%Fax: +1 800 542 0200 Purified synthetic graphite: carbon (LOI): 99.7–99.9%

Superior Graphite Europe Ltd. Signature Products:Sundsvall, Sweden Natural crystalline vein and f lake graphite: carbon Fax: +46 60 13 41 28 (LOI): 80–99%

Amorphous graphite: carbon (LOI): 60–90%(microcrystalline); synthetic graphite: carbon (LOI): 98.0–99.7%

Timcal Ltd., Synthetic graphite (Timrex): grades KS, T, SFG, Bodio, Switzerland MX/MB, HSAG, KB/KL, SLPwww.timcal.com Natural graphite (Timrex): grades BNB, GA/GB

Tab. 19-5 Major suppliers/manufacturers of BN.

Supplier/Manufacturer Grades/Types

GE Advanced Ceramics BN grades:(formerly Advanced Ceramics Corp.): HCP, HCPH, HCPL, AC6004, NX, HCV, AC6003, Headquarters: Cleveland, OH, USA AC6097, AC6069, AC6028, AC6103, AC6091, HCR48, www.advceramics.com HCJ48, HCM

Industrial Supply, Inc. (h)BN grades: PG (premium), SG (standard), USA CG (custom)Fax: 1-970-461-8454

National Nitride Technologies Co., Ltd. BN powder grades: N, NA, NW, S, SW, SAPomona, CA, USAwww.nntbn.com

Saint-Gobain Advanced Ceramics Combat BN grades: MCFP, SHP705, PSHP605, (formerly Carborundum Co.) PHPP325; CarboTherm™ BN powders; Amherst, NY, USA CarboGlide™ BN powderswww.bn.saint-gobain.com

San Jose Delta Associates, Inc.Santa Clara, CA, USAFax: 1-408-727-6019

Wacker-Chemie GmbH BN powdersMünchen, Germanywww.wacker.com


353

Tab. 19-6 Major suppliers/manufacturers of PTFE.

Producer/Supplier Grade

Asahi Glass Co., Ltd. Fluon™ PTFE powdersTokyo, Japan G100 series: Fine particle powderswww.agc.co.jp/english G200 series: Pre-sintered extrusion powders

G300 series: Agglomerated (free f lowing)Lubricants: L150J, L169J, L170J, L172J, L173J

Daikin America, Inc. DAIKIN-POLYFLON™ PTFE fine powdersOrangeburg, NY, USA grades: F-104U, F-107, F-201, F-201L, F-205, F-301, Phone: 1-800-365-9570 F-303, F-303HG

Daikin Industries, Ltd.Osaka, JapanFax: 81-6-6373-4390

Daikin Chemical Europe GmbHDüsseldorf, [email protected]

DuPont Corp. Zonyl: MP 1000, 1300, 1400, 1500, 1600, 1200, etc.Wilmington, DE, USAPhone: 1-800-441-7515

Dyneon LLC, Dyneon GmbH & Co. KG Dyneon PTFE fine powders:USA Fax: +1 800 635 8061 TF 2021, 2025, 2029, 2071, 2072, TFX 2035, TFM 2001, Germany Fax: +49 6107 772 517 etc.

Dyneon PTFE micropowders: TF 9201, 9205, 9207, J14, J24, etc.

Lubrizol Micronized powder grades:World Headquarters, Wickliffe, Pinnacle 9000, 9001, 9002, 9007, 9008, 9500OH, USAwww.lubrizol.com

Shamrock Technologies, Inc. Micronized powders:Dayton, NJ, USA Fluoro M290, E, HP, FG, Raven 5372www.shamrocktechnologies.com

Solvay Solexis, Inc., USA Micronized powder grades: Polymist F-5, 5A, 5A EX, Italy, Japan, France, UK, Brazil, 510, XPH-284China. www.solvaysolexis.com Algof lon

19.2.5Cost/Availability

Prices of MoS2 depend on degree of fineness and order size, with prices in the rangeof $20/kg for small quantities up to 10 kg. The costs of various grades of PTFE pow-ders for lots of 2000−3000 kg are in the $5.50−$6.50/kg range depending upon thecomposition/property of the product. The costs of BN powders for lots of 100 kg areabout $70−80/kg for refractory grades and $100−120/kg for ceramic grades, with lit-tle or no increase expected in the immediate future. High purity grades can be pur-chased at $200−400/kg, depending on quality and size of the order.


354

With regard to graphite, prices in the $230−750/ton range are reported, dependingon type and grade. Flake graphite is available in sizes ranging from 0.5 mm f lakes to3 µm powder. The morphology of f lake graphite is consistently laminar regardless ofparticle size. Intumescent graphite is available in purities ranging from 80−99% car-bon. Both coarse and fine grades are available. The degree of intumescence, alsoknown as “expandability”, generally ranges from 80 times to 300 times volume in-crease. Products can be specified as low (acidic), neutral, and high (alkaline) pH to al-low compatibility with a variety of aqueous and non-aqueous coating/paint systems[14,15,17].

Crystalline vein graphite commercial grades are available with purities in the range85−99% carbon. Sized materials from 2.5 cm to 3 µm are available. Most of the cur-rent supply of amorphous graphite available in the USA is imported from Mexico andChina. Amorphous graphite is typically lower in purity than other natural graphites.Commercial grades of amorphous graphite are available with purities in the range 75−85% and in sizes ranging from 10 cm lumps to 3 µm powder. This graphite variety istypically lower in cost than other types, but is still lubricious, conductive, and chem-ically stable.

Synthetic graphite is typically available in purities above 99%. High purity is therule rather than the exception with this material because the feedstocks used to makeit are typically petroleum-based materials that are inherently low in mineral impuri-ties, and the manufacturing method tends to expel impurities, which are vaporizedat the high process temperatures. This variety of graphite is available in sizes from1.2 cm down to 3 µm [14,15,17].

19.2.6Environmental/Toxicity Considerations

19.2.6.1 MolybdeniteAccording to the IMOA [6,38], MoS2 has been shown not to be harmful to rats by in-halation or by ingestion (MoS2 in 1% w/v aqueous methylcellulose at a dose level of2000 mg/kg bodyweight).

The 4 h LC50 (lethal concentration for 50% kill) was determined to be more than2.82 mg L−1 in air. The inhalation hazard associated with acute exposure to MoS2 isthus low. MoS2 has been shown not to be harmful to rats, guinea pigs, and rabbits incontact with skin (acute lethal dermal dose >2000 mg/kg bodyweight), not to causesensitization by skin contact, and not to be irritating to the skin or eyes. Up to 1998,there had been no reports of dermatitis in exposed workers. The following Occupa-tional Exposure Limits are applicable: OSHA (Occupational Safety and Health Ad-ministration) TWA (time-weighted average; total dust), 10 mg m–3 ; DFG (DeutscheForschungsgemeinschaft) MAK (maximum concentration in the workplace) TWA(total dust), 15 mg m–3 ; ACGIH (American Conference of Governmental IndustrialHygienists, Inc.) TLV (threshold limit values) TWA, 10 mg m–3 .


355

19.2.6.2 GraphiteThe following exposure limits have been reported [39]: TLV, 2.0 mg m–3 as respirabledust (ACGIH 1996−1997); OSHA PEL (permissible exposure limits), TWA: 15 mppcf(million particles per cubic foot); NIOSH REL (recommended exposure limits),TWA: 2.5 mg m–3 (respiratory); NIOSH IDLH (immediately dangerous to life orhealth concentration): 1250 mg m–3 .

With regard to the effects of long-term or repeated exposure to dust, lungs may beaffected resulting in graphite pneumoconiosis. OSHA’s proposed 8-hour TWA PELfor synthetic graphite was 10 mg m–3 (total particulate), and this limit is establishedby the final rule; the 5 mg m–3 limit for the respirable fraction is retained. TheACGIH also has a TLV-TWA limit of 10 mg m–3 for graphite as total dust [40].

19.2.6.3 Polytetraf luoroethylenePTFE is not classified as dangerous according to European Commission directives. Itshould be handled in accordance with good industrial hygiene and safety practice.This material has not been tested for environmental effects, and according to the In-ternational Agency for Research on Cancer (IARC), it belongs to Group 3, unclassifi-able as to carcinogenicity in humans. For PTFE thermal decomposition products, airconcentration should be controlled since they are likely to be toxic monomers, mono-or dif luoroacetic acid, trif luoroacetate, hexaf luoropropene, etc. [41].

19.2.6.4 Boron NitrideBN is a non-f lammable, non-reactive solid material. It is supplied in the form of anodorless white powder. It is considered a nuisance dust [42]. Exposure to BN has notbeen shown to result in direct poisoning. Under normal operating conditions or ther-mal decomposition, BN has not been shown to liberate free boron. According to MS-DS-102 [43], BN powders may contain about 0.1–8.0% boron oxide (CAS #1303-86-2),which may cause irritation, alter kidney function, and produce changes in the bloodas a result of occupational exposure. Boron oxide may cause reversible effects that aregenerally not life-threatening. The exposure limits for boron oxide are as follows:ACGIH, 10 mg m–3 TWA; OSHA, total dust 10 mg m–3 TWA; NIOSH, 10 mg m–3

TWA.

19.2.7Applications

Some typical applications of MoS2, graphite, PTFE, and BN in various plastics are list-ed in Table 19-7. Specific recent examples include:

A process for preparing high-strength UHMWHDPE (ultra high molecular weighthigh density polyethylene) composite plastics, useful for manufacturing mechani-cal drive and rotation parts, involves the incorporation of 5−15% MoS2 dry powder[44].

The outer edges of blades of the rotor used in air motor are made of polyamide oracetal copolymer containing MoS2 as a dry lubricant [45].


356

A helical blade used in f luid compressors made of plastics such as PTFE, PFA (per-f luoroalkoxy resin), PEEK (polyether ether ketone), PES (polyether sulfone), PEI(polyether imide), PAI (polyamide imide), PPS (polyphenylene sulfide) or LCP (liq-uid-crystal polymer) contains solid lubricants, including MoS2, graphite, BN, etc.[46].

The screens for waste water filtration are made of fiber-reinforced plastic contain-ing solid lubricants such as MoS2 [47].

With the addition of 30% graphite, the friction coefficient of polyamide-6 was re-duced by 30% with only a small increase in wear, while for polystyrene, both thefriction coefficient and wear were decreased. A further increase in the graphiteconcentration increased both the wear and the friction coefficient [33]. Yan et al.[48] reported similar observations for a PTFE/graphite system and rationalized theincreased wear rate in terms of increased porosity with an increase in graphite con-centration.

Combinations of MoS2 and graphite have generally been found to exhibit a syner-gistic effect in extreme pressure and anti-wear characteristics, with the level of syn-ergism depending on the ratio of the two components [7].

For PES (polyether sulfone) lubricated with 20% PTFE, the dynamic coefficient offriction decreases from 0.37 to 0.11 and the wear factor drops from 1500 to 3 [49].

Optimum PTFE loadings of 15% in amorphous and elastomeric base resins and20% in crystalline base resins provide the lowest wear rates. Higher PTFE loadingshave minimal effects in terms of further reduction in wear rate, although the coef-ficient of friction will continue to decrease [49]. The effects of PTFE on the wearcharacteristics of various engineering resins are strongly dependent on the type ofresin [49].

In addition to its primary function as a lubricating filler and wear-resistant barrierfor applications [35,50] such as sliding or rotating motion (Table 19-7), secondaryfunctions of the non-abrasive hexagonal BN powder are: a) enhancement of ther-mal conductivity [51] (see also Chapter 5) due to excellent particle-to-particle con-tact that provides a superior thermal path, b) rheology modification during pro-cessing of PE films, c) enhancement of dielectric strength, d) high-temperaturecomposite stability due to its chemical inertness and low coefficient of thermal ex-pansion, e) improvement of crystal nucleation efficiency during processing, and f )enhanced mold-release characteristics.

19.3Anti-Blocking Fillers

19.3.1 GeneralFor plastic sheets/films, “blocking” is generally defined as a condition that occurswhen two or more sheets/films stick together when stacked on top of one another.For coatings, blocking is a measure of the ability of a given coating to resist adhesionto itself (or another freshly coated surface) or adhesion to another substrate [17].Thus, an anti-blocking agent can be defined as an additive that prevents the undesir-able sticking together or adhesion of coated surfaces under moderate pressure, or un-


357

der specified conditions of temperature, pressure, and humidity. The antiblockingagents function by producing invisible, micro-surface imperfections on the film/coating surface, which entrap air and voids on a microscopic scale, thereby reducingthe adhesion/sticking of the film layers to each other [52]. In other words, anti-block-ing additives in particulate form simply act as “spacer bars” between the two film lay-ers/surfaces [53], lowering the interlayer blocking (adhesive) force. Two parameters,namely the particle size and the number of anti-block particles on the film surface,dominate the anti-blocking effect, besides other factors such as type of film material.It has been shown that the higher the surface roughness (i.e., the more particles thatthere are on the surface), the better the anti-blocking performance. In general, smallparticle size anti-blocks are preferred for thin films (20–30 µm), while coarse parti-cles are employed for thick films (>30 µm). Agglomerates of anti-blocks reduce theanti-blocking performance [53]. Selection of the appropriate anti-block depends onthe polymer type and the quality requirements of the final film product.

Several inorganic fillers/organic additives, such as silica, talc, kaolin, CaCO3, tita-nia, zeolites, cross-linked acrylic copolymers, spherical silicon beads, etc., are em-ployed in the plastics/coatings industry to attain the desired blocking performance.Some of these fillers are discussed elsewhere in this book in terms of their primaryfunction; only amorphous silica forms (natural and synthetic) used for anti-blockingwill therefore be discussed in this chapter.

Silicas, which are in competition with carbon blacks as functional fillers for plas-tics and rubbers, have one significant advantage: their white color [54]. The most im-portant role of silicas is as elastomer reinforcements, inducing an increase in me-chanical properties. Other functions, in addition to their use as anti-blocks for PE, PP,

19.3 Anti-Blocking Fillers

Tab. 19-7 Various applications of MoS2, graphite, PTFE, and BNin different polymers.

Polymer(s) MoS2 Graphite PTFE BN Refs.

Polyolefins X X X X 67, 76, 89Acrylonitrile rubber X 68Polyurethane X 69PTFE copolymer X X 72Epoxy X X 73, 81Acetal, HDPE X X 69, 71, 75EPR, NBR, rubber products X X X 70, 77, 90Nylon 12 X 78PTFE, PA, PAI, PI, PPS X X X X 74, 79, 80, 87Thermoplastics X 82PPQ (poly(phenylquinoxaline)) X 83Thermoset (Kerimid) X 84PC, polyester, phenolic X 85PE, FEP X X 86PCO (polycyclooctene) X 88

358

and other films, are: a) to promote adhesion of rubber to brass-coated wires and tex-tiles, b) to enhance the thermal and electrical properties of plastics, c) in accumula-tor separators, and d) as rubber chemical carriers.

Since fumed silica enhances several different properties in formulations, such asthickening liquids or improving f lowability of powders, it is used in a variety of ap-plications [55], including plastics, rubbers, coatings, cosmetics, adhesives, sealants,inks, and toners.

19.3.2Silica as an Anti-Blocking Filler

19.3.2.1 ProductionThe term silica is used for the compound silicon dioxide, SiO2, which has severalcrystalline forms as well as amorphous forms, which may be hydrated or hydroxylat-ed [56]. It is the chemical inertness and durability of silica that has made it very pop-ular in many applications [33].

Natural silicas can be divided into crystalline and amorphous. Crystalline varietiesinclude sands, ground silica (silica f lour), and a form of quartz – Tripoli. The amor-phous types used as anti-blocks include diatomaceous earth (DE) or diatomite. DE isa chalky sedimentary rock composed of the skeletons of single-cell aquatic organ-isms, the diatomites, grown in a wide variety of shapes and varying in size from10 µm to 2 mm. Figure 19-7 shows an example of a DE deposit. The skeletons arecomposed of opal-like, amorphous silica (SiO2(H2O)x) having a wide range of porous,fine structures. The natural grades are uncalcinated powders classified according toparticle size distribution. During the calcination process, the moisture (~40% in DE)is also removed due to the high process temperature, which may also cause sinteringof DE particles to clusters.

Amorphous synthetic silicas, used as anti-blocks, are produced by two differentprocesses: pyrogenic or thermal (generally referred to as fumed silica grades) and thewet process (known as precipitated or particulated silica). The ingredients used in the


Fig. 19-7: An example of a DE deposit with a single principal di-atom present, ×1000. Courtesy of World Minerals, Inc.

359

manufacture of fumed silica are chlorosilanes, hydrogen, and oxygen, and theprocess involves the vapor-phase hydrolysis of silicon tetrachloride in a hydrogen/oxygen f lame [55]. The reactions are:

2H2 + O2 → 2H2OSiCl4 + 2H2O → SiO2↓ + 4HCl↑(Overall reaction):SiCl4 + 2H2 + O2 → SiO2↓ + 4HCl↑

Silica can be precipitated from a sodium silicate solution by acidifying (with sulfu-ric or hydrochloric acid) to a pH less than 10 or 11, usually using a lower concentra-tion than in the silica gel preparation method described below. Typical reactions are:

3SiO2 + Na2CO3 → 3SiO2·Na2O + CO2

(SiO2·Na2O) aq. + 2H+ + SO42– → SiO2 + Na2SO4 + H2O

The product is separated by filtration, washed, dried, and milled. The final productproperties, such as porosity, specific surface area, size and shape of particles and ag-glomerates, density, hardness, etc., are dependent on process variables such as reac-tant concentration, rates of addition, temperature, and fraction of theoretical silicateconcentration in the reaction [33].

Synthetic silica gel, SiO2(H2O)x, is a solid, amorphous form of hydrated silicondioxide distinguished by its microporosity and hydroxylated surface [53]. It is pro-duced according to the following reaction [33]:

Na2O(SiO2)x + H2SO4 → xSiO2 + Na2SO4 + H2O

The product, containing about 75% water, is dried in a rotary kiln, washed with hotalkaline water (which reinforces the matrix, decreases shrinkage, and produces larg-er pores), and milled to produce xerogels. Replacing the water with methanol beforedrying or supercritical drying produces aerogels with up to 94% air space.

In a recent U.S. patent [57], a new process for producing ultra-fine silica particleshas been described. The process involves directing a plasma jet onto a silicon-con-taining compound so as to form silica vapor, and condensing this vapor on a collec-tion surface. The silica particles have high porosity and surface area and may be usedas catalyst supports or as anti-blocking or anti-slipping agents in plastic films.

19.3.2.2 Structure/PropertiesSynthetic silica gel is a major anti-blocking filler usually used for high quality filmapplications. The structure of silica gel is an interconnected random array of poly-merized spheroidal silicate particles with 2–10 nm diameter and 300–1000 m2 g−1

surface area [53]. Its refractive index of 1.46 is very close to those of PE (~1.50) and PP(~1.49), which helps in retaining the high transparency and clarity of the films. Be-cause of the highly porous structure, synthetic silica gel provides many particles perunit weight, directly enhancing the anti-blocking performance vs. a non-porous ma-


360

terial. Its high purity, >99.5 wt. %, and low level of deleterious impurities furtherminimizes the likelihood of any major quality deterioration of polymer films [53].Amorphous diatomite, DE, used in commercial applications, contains 86−94 wt. %SiO2, with a skeletal structure that contains 80–90% voids. Various properties of di-atomite, fumed silica, and precipitated silica are compared in Table 19-8 [53,56]. Fur-ther information on fumed silica may be found in Chapter 20.

Table 19-8 Comparison of properties of fumed silica vs. diatomiteand precipitated silica [53,56,59].

Property Diatomite, DE Fumed Silica Precipitated Silica

CAS # 61790-53-2 112945-52-5 112926-00-869012-64-2 7699-41-4

SiO2, % 85.5–92.0 96.0–99.9 97.5–99.4CaO, % 0.3–0.6 trace 0.5Na2O, % 0.5–3.6 trace 0–1.5Loss on ignition, % 0.1–0.5 1.0–2.5 3–18Decomposition temp., °C 2000 2000 2000Thermal conductivity, W m–1 K–1 0.015 0.015 0.015Thermal exp. coeff., K–1 0.5 × 10– 6 0.5 × 10– 6 0.5 × 10– 6

Specific heat, J kg–1 K–1 794 794 794Surface area, m2 g–1 0.7–3.5 15–400 45–700pH, aq. suspension 9–10 3.5–8 4–9Water solubility, g/100 mL none 0.015 0.015Acid solubility none, except HF none, except HF none, except HFTrue density, g cm–3 2.0–2.5 2.16 2.0–2.1Hardness, Mohs n.a. 6.5 1.0Refractive index 1.42–1.48 1.45 1.45Volume resistivity, Ω cm n.a. 1013 1011–1014

Surface resistivity, Ω cm 5 × 109 5 × 109 5 × 109

Dielectric constant, 104 Hz 1.9–2.8 1.9–2.8 1.9–2.8

19.3.2.3 Suppliers/ManufacturersThe United States is the world’s largest producer and consumer of diatomite. Othermajor producers include Russia, France, Germany, Mexico, Spain, Italy, Brazil, andPeru.

Table 19-9 lists various suppliers/manufacturers of various types/grades of silicaanti-blocks.

19.3.2.4 Environmental/Toxicity ConsiderationsIn evaluating the effects of exposure to silica, it is important to differentiate betweenamorphous and crystalline forms. According to NIOSH [58], at least 1.7 million U.S.workers are exposed to respirable crystalline silica in a variety of industries and oc-cupations. Silicosis, an irreversible but preventable disease, is the illness most close-ly associated with occupational exposure to the material, which is also known as sili-ca dust. Occupational exposures to respirable crystalline silica are associated with the



Tab. 19-9 Major suppliers/manufacturers of anti-block silicas.


Cabot Corporation Cab-O-Sil (hydrophilic fumed silica): L-90, LM-130, World Headquarters: Boston, MA, USA LM-150, M-5, MS-55, H-5, HS-5, EH-5www.cabot-corp.com Cab-O-Sil (hydrophobic fumed silica): TS-500, TS-530,

TS-610, TS-720Cab-O-Sil (densified silica grades): LM-150D, M-7D, M-75D

Degussa AG Sipernet 310, 320, etc.Frankfurt, Germany Sident 8, 9, 10, [email protected]

Degussa Corp. Ultrasil 360, 880, 7005, etc.Parsippany, NJ, USA Aerosil 200, 300, OX 50, etc.Fax: 1-973-541-8501

Fuji Silysia Chemical, Ltd. Silysia (micronized, synthetic, amorphous, pure, Res. Triangle Park, NC, USA colloidal silica)Fax: 919-544-5090 Type – G (SiO2 powder)www.fuji-silysia.com

Grefco Minerals, Inc. Dicalite (diatomaceous earth (DE) fillers)Bala Cynwyd, PA, USAwww.grefco.com

Harwick Standard Distribution Corp. Silica S (fumed silica)Akron, OH, USAwww.harwickstandard.com

INEOS Silicas Americas Gasil (synthetic silicas)Joliet, IL, USAwww.ineossilicas.com

Minerals Technologies Optibloc – 25, 10, etc. (clarity anti-block)Specialty Minerals Sylobloc 45 (synthetic silica)Dillon, MT, USA ABT - 2500www.mineralstech.com

OCI International, Inc. Micloid (micronized silica)Englewood Cliffs, NJ, USA,Fax: 1-201-569-3005

PPG Industries, Inc. Lo-Vel 27, 275, 28, 29, 66Pittsburgh, PA, USA, www.ppg.com Hi-Sil T-600, T-700 (precipitated silica)

Sibelco Cristobalite Minbloc®M , M4000-M6000Dessel, Belgium, www.sibelco.be

Unimin Corp. Minblock HC (sodium aluminum silicate, 60.3% SiO2)Tamms, IL, USA HC 500, 1400, 2000, 2100Houston, TX, [email protected]

W.R. Grace & Co. Ludox (colloidal silica for anti-blocking)Grace Davison PERKASIL KS-series (precipitated silica)www.gracedavison.com PERKASIL SM-series

362

development of silicosis, lung cancer, pulmonary tuberculosis, and diseases of therespiratory tract. Crystalline silica has been classified by the IARC as carcinogenic tohumans when inhaled.

In general, fumed or pyrogenic silica, which is X-ray amorphous, is non-toxic, doesnot cause silicosis, and is safe to work with. However, excessive inhalation should beavoided by appropriate ventilation or the wearing of protective masks [59].

OSHA’s current limit for amorphous silica is 20 mppcf, which is equivalent to6 mg m–3 TWA (ACGIH 1984), measured as total dust. The ACGIH has establisheda limit for this dust (measured as total dust containing < 1% quartz) of 10 mg m–3 (8-hour TLV-TWA). In contrast, TLVs for quartz, cristobalite, and tridymite dusts are on-ly 0.05 mg m–3 . Diatomaceous earth composed of the skeletons of prehistoric or-ganisms known as diatoms is largely non-crystalline; the presence of varyingamounts of crystalline quartz has led, in the opinion of the ACGIH (1986/Ex. 1-3, p.520), to conf licting results in studies of the pulmonary effects of exposure to this col-orless to gray, odorless powder. Acute toxicity LD50/LC50 values for amorphous sili-ca relevant to classification are: oral > 5000 mg/kg rat, dermal > 5000 mg/kg rabbit,inhalation > 0.139 mg/L/4 h rat. In general, according to FDA regulations, amor-phous silica is generally recognized as safe (GRAS).

19.3.3Applications of Silicas

Some recent applications of silica anti-blocks for plastic films are given below. In gen-eral, a typical concentration of 2500–4000 ppm of natural silica is employed forLDPE, while 1000–2000 ppm of synthetic silica is employed for PP, LDPE, LLDPE,and PET.


Tab. 19-9 Continued


Grace Silica GmbH ELFADENTDüren, Germany SYLOWHITE

W. R. Grace Specialty Chemicals DURAFILL(Malaysia)Grace Brasil Ltda.

Wacker-Chemie GmbH HDK S13, V15, N20, T30, T40, V15P, N20P München, Germany (hydrophilic fumed silicas)www.wacker.com HDK H15, H18, H20, H30, H15P (hydrophobic)

World Minerals Inc. ActivBlock (for polyolefins)Santa Barbara, CA, USA Diatomite (natural silica) products:www.worldminerals.com/activblock.asm Celite, Kenite, Diactiv, Primisil, Diafil

World Minerals Europe S.A.Fax: 33 1 41 91 5732

363

Davidson and Swartz [52] have studied the anti-blocking performance of varioussynthetic silicas, talc, and diatomaceous earth fillers for high clarity polyethylenefilm applications and reported anti-blocking performance vs. haze, yellowness index(YI), coefficient of friction (COF), and gloss for LDPE, LLDPE, and mLDPE. The an-tagonism between five commercial grades of erucamide and seven commercialgrades of inorganic anti-block agents used in LDPE film formulations has been in-vestigated by Peloso et al. [60]. Use of silica from rice husk ash (RHA) as an anti-block-ing agent in LDPE film has been studied by Chuayjuljit et al. [61]. It was found thatLDPE film with 2000−3000 ppm RHA silica showed similar properties to commer-cial LDPE films filled with 500−1000 ppm silica in terms of blocking behavior, me-chanical strength, and film clarity.

In a review article, Shiro [62] has compiled the applications of anti-blocking fillers,such as silica, kaolin, talc, etc., in plastics, transparent films, magnetic recording ma-terials, electrical insulating films, etc. Water-soluble polyester films and coatings formagnetic recording materials with improved anti-blocking property, adhesion, andsolvent resistance are prepared using silica as an anti-blocking filler [63]. Similarly,pressure-sensitive adhesives, with good anti-blocking property, heat resistance, andabrasion resistance containing silica gel have been developed [64]. Thermoplasticfilms for use in stretch/cling wrapping are reported to contain a silica-based anti-blocking polyolefin layer on the opposite side from the cling layer [65]. Coating com-positions for imparting a suede finish to leather substitutes contain polyurethanes,polybutylene, and silica [66].

Synthetic silicas are used preferentially as anti-blocks in PP and PVC films com-pared to diatomite and other mineral fillers. For thicker films, coarser silicas with anaverage particle size of about 11 µm are used. The advantages of using synthetic sili-cas are their high efficiency and adsorption capacity (due to high porosity) for PVCplasticizers. As a secondary function, anti-blocking silicas not only create a micro-rough surface, but also improve printability of PVC films. Anti-blocks are also usedin the skin layer of multilayer films containing polymers such as PA, EVOH, ABS,etc., and in various applications of polyester films, such as magnetic tapes, videotapes, packaging films, capacitors, etc. For magnetic films, very fine silica (<1−2 µm)or submicron-sized calcium carbonate or China clay is used, besides other mineralssuch as barites and aluminum oxides [53].


364

References


1 Rohr, E., Paint & Coatings Industry 2003,19(10), 110; accessed at www.pcimag.com

2 SpecialChem S.A., “Additives for tribologi-cal polymers used in bearing applications”,July 15, 2003, article accessed at www.spe-cialchem4polymers.com/resources/articles

3 Tribology-abc, “Solid lubricants/dry lubri-cation” at www.tribology-abc.com/abc/solidlub.htm

4 GE Advanced Ceramics, “Lubrication” atwww.advceramics.com/geac/applications/lubrication/

5 “Molybdenite” at www.mindat.org/min-2746.html

6 National Institute for Occupational Safetyand Health (US NIOSH), “Molybdenumsulfide” atwww.cdc.gov/niosh/rtecs/qa47aba8.html

7 Klofer, A., “A study of synergistic effects ofsolid lubricants found in lubrication pack-ages of constant velocity joints”, Thesis, Im-perial College of Science Technology andMedicine, London, March 14, 1997, ac-cessed atwww.mitglied.lycos.de/akloefer/Technik2/Technik2.html

8 National Institute for Occupational Safetyand Health (US NIOSH), “Graphite” atwww.cdc.gov/niosh/rtecs/md9364d0.html

9 Gapi Group “Virgin PTFE” at www.gapi-group.com/ptfe2.HTM

10 Ceram Research, “Boron Nitride” atwww.azom.com/details.asp?ArticleID=78

11 Accuratus, “Boron Nitride” at www.accura-tus.com/boron.html

12 National Institute for Occupational Safetyand Health (US NIOSH), “Boron Nitride”at www.cdc.gov/niosh/rtecs/ed7704c0.html

13 International Molybdenum Association, UK(IMOA) at www.imoa.org.uk/

14 Asbury Carbons, “Graphite” at www.as-bury.com

15 Tamashausky, A. V., “Graphite – A multi-functional additive for paint and coatings”,Paint & Coatings Industry 2003, 19,(10), 64;accessed at www.pcimag.com

16 Graphtek LLC , “Graphite” at www.graph-tekllc.com/graphite.htm

17 Koleske, J. V., et al., “Two Thousand Four –Additives Guide”, Paint & Coatings Industry2004, 20(4), 32.

18 Gangal, S. V., “Fluorine Compounds, Or-ganic”, in Kirk–Othmer Encyclopedia ofChemical Technology, 3rd Ed., Vol. 11, JohnWiley & Sons, New York , 1996, p. 4.

19 Bladel, H., Felix, B., Hintzer, K., Lohr, G.,Mitterberger, W. D., U.S. Patent 5,576,381,Hoechst Aktiengesellschaft, Germany, 1994.

20 Giannetti, E., Visca, M., US Patent4,864,006, Ausimont S.p.A., Italy, 1987.

21 Visca, M., Chittofrati, A., US Patent4,990,283, Ausimont, S.p.A., Italy, 1988.

22 Marchese, E., Kapeliouchko, V., AusimontS.p.A, Italy, “High solids-content PTFE na-noemulsions”, accessed at www.fon-dazione-elba.org/73.htm

23 “The Mineral Molybdenite” at http://miner-al.galleries.com/minerals/sulfides/molyb-den/molybden.htm

24 “Molybdenite Gallery” atwww.mindat.org/gallery.php?min=2746

25 “Sulphides” atwww.geocities.com/ijkuk/ik_sulphide.htm

26 “Molybdenite” at www.theimage.com/min-eral/molybdenite/index.htm

27 “Molybdenite” at www.webmineral.com/da-ta/Molybdenite.shtml

28 “Molybdenite” atwww.gwydir.demon.co.uk/jo/minerals/molybdenite.htm

29 “Systematic Mineralogy – Sulfides” athttp://geology.csupomona.edu/drjessey/class/gsc215/minnotes17.htm

30 “Molybdenite”, atwww.chem4kids.com/misc/photos/molyb-denite1.jpg

31 “Molybdenite” atwww.mineralgallery.co.za/molybdenite.htm

32 “Mineral description: Molybdenite” atwww.geology.neab.net/minerals/molyb-den.htm


34 “Graphite” athttp://mineral.galleries.com/minerals/ele-ments/graphite/graphite.htm

35 Lelonis, D. A., et al., “ Boron Nitride Pow-der – A High Performance Alternative forSolid Lubrication”; Article accessed atwww.advceramics.com

36 Deacon, R., Proc. Royal Soc. 1957, 243A,464.

365References

37 International Molybdenum Association, UK(IMOA), “Concentrates – Unroasted” atwww.imoa.info/Default.asp?Page=197

38 International Molybdenum Association, UK(IMOA) “Molybdenum disulfide” atwww.imoa.org.uk/Default.asp?Page=155

39 National Institute for Occupational Safetyand Health (US NIOSH), “Graphite (Natur-al)” atwww.cdc.gov/niosh/ipcsneng/neng0893.html

40 National Institute for Occupational Safetyand Health (US NIOSH), “Graphite, Syn-thetic” at www.cdc.gov/niosh/pel88/syn-graph.html

41 Ellis, D. A., et al., Nature 2001, 412, 321.42 Electronic Space Products International,

MSDS of BN at www.espi-metals.com/ms-ds’s/Boron%20Nitride.htm

43 Saint-Gobain Advanced Ceramics, MSDS102 for Combat BN powders, Revision Date:02/09/2004, Amherst, NY.

44 Yichuan, J., CN 1338484, Faming ZhuanliShenqing Gongkai Shoumingshu, China,2002; AN: 2003: 204208

45 Cooper Power Tools GmbH & Co., Ger-many, DE 20021980, 2002; AN: 2002:446050.

46 Fujiwara, T., et al., US 6354825, KabushikiKaisha Toshiba, Japan, 2002.

47 Hiroshi, K., Koji, S., JP 2004042010, Ya-mayichi Technos K.K., Japan, 2004; AN:2004: 117492.

48 Yan, F., et al., J. Appl. Polym. Sci. 1996,61(7), 1231–1236.

49 LNP Engineering Plastics, “A Guide toPlastic Gearing” atwww.lnp.com/LNP/Products/Available-Brochures/WearResistant.html

50 Saint-Gobain Advanced Ceramics Corp.,“Boron Nitride” at www.bn.saint-gobain.com

51 Zhuo, Q., et al., WO 00/42098, Ferro Cor-poration, 2000.

52 Davidson, D. L., Swartz, D., Proc. 62nd SPEANTEC, 2004, 50, 1111.

53 Zweifel, H., (Ed.), Plastics Additives Hand-book, Chapters 7 and 17, Hanser Publish-ers, Munich, 2001.

54 Biron, M., “Silicas as Polymer Additives” ar-ticle accessed at www.specialchem4poly-mers.com/resources/articles/article.aspx?id=1295

55 Cabot Corp., “Untreated Fumed Silica:General Application Guide”, Technical Bul-letin at www.cabot-corp.com

56 Willey, J. D., “Amorphous Silica”, inKirk–Othmer Concise Encycl. Chem. Technol.3rd Ed., John Wiley & Sons, Inc., 1985,p. 1054.

57 Debras, G., US 6495114, Fina Research,S.A., Feluy (BE), 2002.

58 National Institute for Occupational Safetyand Health (US NIOSH), “Silica” atwww.cdc.gov/niosh/topics/silica/default.html

59 Katz, H. S., Milewski, J. V. (Eds.), Handbookof Fillers and Reinforcements for Plastics,Chapter 8, Van Nostrand Reinhold Co., NewYork, 1978.

60 Peloso, C. W., et al., Polymer Degrad. andStab. 1998, 62(2), 285–290.

61 Chuayjuljit, S., et al., J. Appl. Polym. Sci.2003, 88(3), 848–852.

62 Shiro, M., Purasuchikkusu 1995, 46(10), 38;AN: 1995: 882737.

63 Juzo, S., et al., JP 06056979, Toray Indus-tries Japan, 1994; AN: 1994: 458262.

64 Tsutomu, S., JP 01168779, Toppan MooreCo. Ltd., Japan, 1989; AN: 1990: 8634.

65 Masten, P., EP 317166, Exxon Chemicals,1989; AN: 1989: 555491.

66 Takamitsu, D., JP 61285268, Honny Chemi-cals, Japan, 1986; AN: 1987: 41861.

67 Allod GmbH & Co. K.-G., Germany, DE20118862, 2002; AN: 2002: 240365.

68 Kazuhiko, K., Kimihiro, N., JP 09040807,Mitsubishi Cable Ind., Japan, 1997; AN:1997: 264610.

69 Kazumasa, H., Hiroshi, Y., JP 08252874,Mitsubishi Belting Ltd., Japan, 1996; AN:1996: 733587.

70 Ludomir, S., et al. PL 168619, Politechnika#odzka, Poland, 1996; AN: 1996: 485617.

71 Daisuke, S., Takatoshi, A., Tyoichi, S., US5,508,581, Nikon Corp., Japan, 1996; AN:1996: 298576.

72 Antonio, C., Ewald, M., EP 658611, Rings-dorff Sinter GmbH, Germany, 1995; AN:1995: 890103.

73 Jaroslav, S., Ladislav, M., CS 276646, TosHulin, Czechoslovakia, 1992; AN: 1994:558821.

74 Heinz, K. W., Erich, H., DE 4200385, Glyco-Metall-Werke Glyco & Co. KG, Germany1993; AN: 1994: 109032.

366 19 Surface Property Modif iers

75 Sunao, I., et al., JP 05247351, MatsushitaElectric Co., Japan, 1993; AN: 1994: 78661.

76 Yongwei, H., CN 1064906, Nanning Auto-mobile Flexible Axel Factory, China, 1992;AN: 1993: 519870.

77 Ludomir, S., et al., PL 144270, Politechnika#odzka, Poland, 1988; AN: 1990: 120439.

78 Itaru, O., Masaaki, S., JP 62264601, SeikoEpson Corp., Japan 1987; AN: 1989: 106850.

79 Bharat, B., ASTM Special Tech. Publ. 1987,947, 289–309.

80 Vinogradova, O. V., et al., Trenie i Iznos1996, 17(4), 544–549; AN: 1997: 615201.

81 Ciora, P., et al., RO 89228, Intreprindereade Utilaj Chimic, Romania, 1986; AN: 1987:555861.

82 Bezard, D., Oesterreichische Kunststoff-Zeitschrift 1988, 19, 1-2, 18, 20-1, 24; AN:1988: 438801.

83 Korshak, V. V., et al., Trenie i Iznos 1986, 7,1, 16-20; AN: 1986: 208216.

84 Toray Industries, Inc., Japan, JP 60020958,1985; AN: 1985: 204960.

85 Avaliani, D. I., Arveladze, I. S., Soob-shcheniya Akademii Nauk Gruzinskoi SSR1979, 96(1), 149–152; AN: 1980: 77260.

86 Hatzikiriakos, S., Rathod, N., Korea-Aus-tralia Rheo. J. 2003, 15(4), 173.

87 Behncke, H. and Kaehne, H. H., DE1937390, 1971; AN: 1971: 112839.

88 Liu, C., Mather, P. T., Proc. 62nd SPE AN-TEC, 2004, 50, 3080.

89 Muliawan, E., Hatzikiriakos, S. G., Proc.62nd SPE ANTEC, 2004, 50, 256.

90 Haberstroh, E., et al., Proc. 62nd SPE AN-TEC, 2004, 50, 355.

91 Rodriguez, F., Principles of Polymer Systems,3rd Ed., Hemisphere Publ. Corp., 1989.

92 Asahi Glass Company, “Fluon PTFE” atwww.agc.co.jp/english/chemicals/jushi/ptfe/PTFE3.htm

93 Jaszczak, J. A., “The Graphite Page” atwww.phy.mtu.edu/faculty/info/jaszczak/graphite.html

94 Gapi Group, “Virgin PTFE datasheet” atwww.gapigroup.com/virgin2.HTM

95 Ferro-Ceramic Grinding, Inc., “CeramicProperties” at www.ferroceramic.com

96 “Boron Nitride” at www.a-m.de/englisch/lexikon/bornitrid.htm

97 “BN – Boron Nitride” atwww.ioffe.rssi.ru/SVA/NSM/Semicond/BN/basic.html

98 Handbook of Chemistry and Physics, 67thEd., CRC Press, Boca Raton, 1986–87, p. B-109.

20Processing Aids

Subhash H. Patel

20.1Introduction

It is widely accepted by the plastics community that “processing aids” is a term givento additives that are employed to overcome processing problems of various plastics,predominantly polyvinyl chloride (PVC). Typically, processing aids are combinationsof high molecular weight polymers/copolymers (such as polymethyl methacrylate(PMMA), styrene–acrylonitrile (SAN) copolymer), oligomers, or other resins, which,when added to PVC, enhance fusion and improve melt properties. Thus, processingaids, when added at low concentrations, contribute to processing improvements,higher productivity, and better product quality.

Addition of small amounts of a processing aid to a polymer may lead to a majorchange in the rheology/viscosity and/or morphology of the material, thereby im-proving processability. Rheology modifiers are compounds that alter the deformationand f low characteristics of matter under the inf luence of stress. They are used ex-tensively in paints, coatings, plastisols, and liquid thermosetting agents prior tocross-linking, to modify viscosity/shear rate characteristics or to impart thixotropy. Athixotropic coating exhibits a time-dependent decrease in viscosity with increasedshear rate up to a limiting value (due to a loss of structure in the coating system). Thisloss is usually temporary and the system will regain its original state given enoughtime. The generated curve is commonly known as a “thixotropic” or “hysteresis” loop[1].

Rheological additives can be roughly divided according to their chemical nature in-to inorganic and organic thickeners, with a subsequent distinction being made be-tween thickeners for solvent-borne systems and thickeners for water-borne systems[2]. A variety of inorganic fillers and organic compounds may be used in cross-link-able liquid systems. Inorganic thickeners/viscosity modifiers are unmodified or or-ganically modified fillers such as bentonites, hectorites, silicas, and organoclays. Al-kaline earth oxides or hydroxides are a special class of reactive fillers acting as thick-eners/viscosity modifiers used in sheet- and bulk-molding compounds (SMC/BMC)based on unsaturated polyesters (UP).


368

The majority of the processing aids employed by the thermoplastics industry areorganic compounds, and these usually lower melt viscosity. In other words, process-ing aids are also rheology modifiers. However, viscosity stabilizers that act by mini-mizing polymer degradation may also be considered as processing aids in thermallysensitive systems such as PVC. However, in view of the scope and title of this book,one can expand the definition of processing aids to inorganic fillers that act as rheol-ogy modifiers or rheology stabilizers for a variety of thermosets, coatings/paints,thermoplastics, and elastomers. Three representative fillers, the primary function ofwhich is to assist processing, are discussed in this chapter:

Magnesium oxide (MgO or magnesia) as a thickening agent/viscosity modifier inSMC/BMC.

Fumed silica as a rheology modifier in liquid cross-linkable coatings. Hydrotalcites as process stabilizers (acid scavengers) in PVC and as acid neutraliz-

ers in polyethylene (PE) and polypropylene (PP).

20.2Production

20.2.1Magnesium Oxide

The majority of magnesium oxide produced today is obtained from the processing ofnaturally occurring magnesium salts present in magnesite ore, magnesium chloride-rich brine, and seawater [3–5]. Worldwide, about two-thirds of the magnesium oxide(some 11 million tonnes per annum) produced is extracted from seawater (co-pro-duced with salt production) and the rest comes from the natural mineral depositsmagnesite, dolomite, and salt domes [5]. Large mineral deposits of magnesite are lo-cated in Australia, Brazil, Canada, China, Europe, the Russian Federation, Turkey,and the U.S. When heated from 700 °C to 1000 °C, magnesium carbonate thermallydecomposes to produce magnesium oxide and carbon dioxide:

MgCO3 → MgO + CO2

Other sources of MgO are underground deposits of brine, which are essentiallysaturated salt solutions containing magnesium chloride and calcium chloride. De-posits of brine located approximately 750 m below ground are used in the Martin Ma-rietta’s process [3]. The process involves the conversion of MgCl2 into Mg(OH)2

through a series of steps involving the reaction of brine with a calcined dolomiticlimestone (CaMg(CO3)2), followed by filtration and calcination of the obtainedMg(OH)2 to produce magnesium oxide:

2Mg(OH)2 → 2MgO + 2H2O

20 Processing Aids

369

Several types of kilns can be used in the calcination step. Calcination not only con-verts magnesium hydroxide to magnesium oxide, but is also the most important stepfor determining the activity of the product as related to its end use.

The process of extraction of magnesia from seawater may involve reaction of limeslurry with MgCl2 and MgSO4 dissolved in seawater to form Mg(OH)2, which is fur-ther calcined as in the Premier Periclase process [4], or thermal decomposition of aconcentrated MgCl2 brine in a special reactor as in the Dead Sea Periclase process [6].

Magnesium oxide may be further refined and purified by rehydrating it to magne-sium hydroxide slurry:

2MgO + 2H2O → 2Mg(OH)2

and addition of compounds to remove contaminants that may have remained fromthe original reaction. The hydroxide is then converted back into the oxide by calcina-tion [3].

20.2.2Fumed Silica

Fumed silica, or fumed silicon dioxide, is produced by the vapor-phase hydrolysis ofsilicon tetrachloride in an H2/O2 f lame. The reactions are shown in Chapter 19. Hy-drophilic fumed silica bearing hydroxyl groups on its surface is produced by thisprocess. Hydrophobic fumed silica is made by processing fumed hydrophilic silicathrough in-line hydrophobic treatments, such as with silanes, siloxanes, silazanes,etc. [1]. Examples of different types of hydrophobic fumed silica coatings includeDMDS (dimethyldichlorosilane), TMOS (trimethoxyoctylsilane), and HMDS (hexa-methyldisilazane).

20.2.3Hydrotalcites

Hydrotalcite is a natural mineral with a white color and pearl-like luster. It ismined in small quantities in Norway and the Ural area of Russia. Natural hydrotal-cite is a hydrated magnesium-, aluminum-, and carbonate-containing mineral with alayered structure that is represented alternatively as 6MgO·Al2O3·CO2·12H2O orMg6Al2(OH)16CO3·4H2O.

Natural hydrotalcite deposits are generally found intermeshed with spinel and oth-er materials due to the existence of non-equilibrium conditions during formation ofthe deposits. Other minerals, such as penninite and muscovite, as well as heavy met-als, are also found in natural hydrotalcite deposits. There are as yet no known tech-niques for separating these materials and purifying the natural hydrotalcite.

Kyowa Chemicals Company, Japan, was the first in the world to succeed in the in-dustrial synthesis of hydrotalcite in 1966 [7]. Synthetically produced hydrotalcite canbe made to have the same composition as natural hydrotalcite, or, because of f lexi-bility in the synthesis, it can be made to have a different composition by replacing the

20.2 Production

370

carbonate anion with other anions, such as phosphate ion [8]. In general, the hydro-talcite compositions are prepared from aqueous solutions of soluble magnesium andaluminum salts, which are mixed in a molar ratio ranging from about 2.5:1 to 4:1, to-gether with a basic solution containing at least a twofold excess of carbonate and asufficient amount of a base to maintain the pH of the reaction mixture in the rangeof 8.5 to 9.5. The product particle size is determined by controlling process parame-ters, including reaction temperature, reaction stirring speed, reaction pH, reactantaddition rates, and reactant concentrations [8]. Synthetic hydrotalcites can be given aplate-like, a needle-like or a spheroidal morphology through control of the synthesisparameters. The empirical formulae for some synthetic hydrotalcites are as follows[9,10]:

DHT-4A® (Kyowa Chemical Ind., Japan)Mg4.5Al2(OH)13(CO3)·3.5H2O

L-55RII® (General Chemical Corp., USA)Mg4.35Al2(OH)11.36(CO3)1.67·xH2O

Baeropol MC 6280® (Baerlocher GmbH, Germany)Mg4Al2(OH)12 (CO3)·2.85H2O

Hysafe 539 (J. M. Huber, USA)Mg4.5Al2(OH)13(CO3)·3.5H2O

20.3Structure/Properties


Magnesium oxide is a white powder broadly similar to calcium oxide and is rarelyfound in Nature as such but more commonly as the carbonate form, including theless common mineral complex with calcium carbonate (carnallite) [5].

Three basic types or grades of “burned” magnesium oxide can be obtained from thecalcination of magnesium hydroxide, with the differences between each grade beingrelated to the degree of reactivity remaining after exposure to a range of extremelyhigh temperatures.

Temperatures used when calcining to produce refractory grade magnesia are in therange 1500–2000 °C and the magnesium oxide is referred to as “dead-burned” sincemost, if not all, of the reactivity has been eliminated. A second type of MgO producedby calcining at temperatures in the range 1000–1500 °C is termed “hard-burned”. Thethird grade of MgO is produced by calcining at temperatures in the range700–1000 °C and is termed “light-burned” or “caustic” magnesia. This material has awide reactivity range and is used in plastics, rubber, paper, and other industrial ap-plications. A typical scanning electron micrograph of “light burned” MgO is shownin Figure 20-1 [3].

Some typical properties of MgO are given in Table 20-1 [11,12].

20 Processing Aids

371

Tab. 20-1 Typical properties of MgO [11,12].

Property Value

Density, g cm–3 3.6 (varies)Melting point, °C 2800Boiling point, °C 3600Refractive index 1.7085Thermal conductivity, W m–1 K–1 at 273 K 42Thermal expansion, K–1 at 273 K 10.8 × 10– 6

Dielectric constant at 1 MHz 9.65Young’s modulus, GPa 249Solubility in water, % 0.00062

20.3.2Fumed Silica

Fumed silica (silicon dioxide) is generally regarded as a unique material because ofits unusual particle characteristics, enormous surface area, and high purity. It is afine, white, and extremely f luffy powder. When added to liquids or polymers withwhich its refractive index (1.46) is a close match, it appears colorless or clear. Unlikecrystalline silicas, amorphous fumed silica is safe to handle, thus eliminating the se-rious health problems associated with crystalline silica dust. Its high surface area,ranging from 130−380 m2 g–1, affects dispersibility, rheology control, thixotropic be-havior, and reinforcement efficiency [12]. The higher the surface area, the more therheological control and thixotropic behavior increases, and the greater the potentialfor reinforcement. However, dispersion becomes more difficult. Various properties

20.3 Structure/Properties

Fig. 20-1 A typical SEM of “light burned”MgO submicron particles with surface area0.1–1.0 m2 g–1. Reproduced with permissionfrom ref. [3].

372

of fumed silica are compared with those of precipitated silica in Table 19-8, Chap-ter 19.

20.3.3Hydrotalcites

Natural hydrotalcites, with the structure Mg6Al2(OH)16CO3·4H2O, crystallize in atrigonal-hexagonal scalenohedral system, are characterized by a lamellar structure,have a density of 2.06 g cm–3, and a Mohs hardness of 2 [13,14]. Figure 20-2 shows athree-dimensional, double-layered structure of a typical hydrotalcite [7] consisting ofmagnesium and aluminum hydroxide octahedrons interconnected through theiredges. Additional interstitial anions between the layers compensate the charge of thecrystal and determine the size of the interlayer distance (basal spacing). While hy-drotalcites are accessible through the corresponding metal salts [15], the metal alco-holate route may provide certain advantages over other synthetic routes, such as vari-ation of the Al/Mg ratio over a wide range, high purity, and controlled anion content[16]. Compared to aluminum trihydrate (pH 8−9), hydrotalcites are even more alka-line by nature. Basicity is adjustable by increasing the Mg/Al ratio and/or incorpo-rating anions other than hydroxyl.

Synthetic hydrotalcites employed in plastics applications are finely divided, free-f lowing, odorless, amorphous powders. For polyolefin applications, hydrotalciteswith an average particle size of 0.5 µm and a specific surface area of less than20 m2 g–1 (BET method) are recommended. Their density varies around 2.1 g cm–1

depending on the composition. Sodium, zinc, and calcium stearate coated hydrotal-cites are available, which show enhanced compatibility with polyalkenes [9]. Fig-ure 20-3 is a scanning electron micrograph of a typical submicron-sized synthetic hy-drotalcite [17].

20 Processing Aids

Fig. 20-2 Schematic of the double-layered metal hydroxidestructure of a typical hydrotalcite. Reproduced from ref. [7].

373

A useful and unique feature of hydrotalcite lies in its special acid-adsorbing mech-anism and its inherent anion-exchanging property. For example, in the case of hy-drochloric acid, the CO3

2– in the hydrotalcite structure is easily replaced by Cl–, andthese chloride ions are incorporated into the crystal structure to produce chlorinatedcompounds such as Mg4.5Al2(OH)13Cl2·3.5H2O, which is insoluble in both water andoil. Also, the Cl– is not desorbed from the crystal structure up to a temperature of ap-proximately 450 °C [7,9]. These reactions allow hydrotalcite to be used as an HClscavenger in PVC stabilization and as a neutralizer of acidic catalyst remnants inpolyalkenes.

20.4Suppliers/Manufacturers

Major suppliers/manufacturers of MgO are given in Table 20-2. Major suppliers/manufacturers of hydrotalcites are given in Table 20-3. Major suppliers/manufactur-ers of fumed silica are covered along with suppliers of anti-block silicas in Table 19-9,Chapter 19.

20.4 Suppliers/Manufacturers

Fig. 20-3 Scanning electron micrograph of synthetic hydrotalcitesubmicron particles [17].

374

Tab. 20-2 Suppliers/manufacturers of magnesium oxide.


AluChem, Inc. Deadburned MgO grades:Reading, OH, USA ADM-98H, ADM-99Pwww.aluchem.com

Baymag Baymag 30, 40, 96Calgary, Alberta, Canada (Natural MgO)www.baymag.com

Dead Sea Periclase RA-150 Active MgOMishor Rotem D.N. Arava, Israel, 86805www.periclase.com

Martin Marietta Magnesia Specialties, LLC MagChem HSA-10, HAS-30Baltimore, MD 21220, USA MagChem 40, 50, 60www.magspecialties.com MagChem 50M

(micronized, high & moderate activity grades)

Premier Chemicals, LLC MAGOX – Super PremiumMiddleburgh Heights, OH 44130, USA MAGOX – Premiumwww.premierchemicals.com

Scora S. A. SCORA MgO (for plastics, rubbers and Caffiers, France elastomers, paints and inks)www.scora.com

Tab. 20-3 Suppliers/manufacturers of hydrotalcites.


Baerlocher GmbH Baeropol MC 6280Unterschleissheim, Germanywww.baerlocher.com

DOOBON, Inc. CLC-120Chungcheongbuk-do, South Koreawww.doobon.co.kr

General Chemical Corporation L-55R IIParsippany, NJ 07054, USAwww.genchemcorp.com

J. M. Huber Corporation Hysafe 310, 510, 539Atlanta, GA 30339, USAwww.hubermaterials.com

Kyowa Chemical Industry Co., Ltd. DHT-4AKagawa, Japan Alcamizerwww.kyowa-chem.co.jp/english

Sasol North America, Inc. PURAL MG 61 HTHoustn, TX 77079, USA PURAL MG 30, MG 50, MG 70www.sasolnorthamerica.com

Sud-Chemie, Inc. SORBACID 911Louisville, KY 40232, USA SYNTAL

HYCITE 713 (distributed by Ciba Specialty Chemical)

20 Processing Aids

375

20.5Environmental/Toxicological Considerations


A NIOSH (National Institute for Occupational Safety and Health) website [18] has list-ed 24 synonyms for MgO. According to the material safety data sheet for a grade rec-ommended for use in plastics (MagChem HAS-10) [3], MgO does not fit the defini-tion of a hazardous material and no LD50 or LC50 data are available for MgO dust.MgO is stable at ambient temperatures and pressures. Exposure to water may causethis product to slowly hydrate, through an exothermic reaction. MgO products maypresent a nuisance dust hazard, but are not f lammable or combustible. MgO is notconsidered a teratogen, mutagen, sensitizer, toxin or carcinogen by the NTP (Nation-al Toxicology Program), the IARC (International Agency for Research on Cancer), orOSHA. However, if magnesium oxide is heated in a reducing atmosphere to the pointof volatilization (i.e., >1700 °C), magnesium oxide fumes may be generated. Exposurelimits for MgO fumes are reported as follows: ACGIH: 10 mg m–3 TWA; OSHA: finalPELs (permissible exposure limits), total particulate 15 mg m–3 TWA.

20.5.2Fumed Silica

It is important to differentiate between amorphous and crystalline forms of silicawhen evaluating its effects upon exposure. In general, fumed or pyrogenic silica,which is X-ray amorphous, is non-toxic, does not cause silicosis, and is safe to workwith. However, excessive inhalation should be avoided by proper ventilation or thewearing of protective masks [12]. According to MSDS for hydrophilic fumed silica(CAS #112945-52-5), no acute toxic effects are expected upon eye or skin contact or byinhalation [19]. This material does not contain any reportable carcinogenic ingredi-ents or any reproductive toxins at OSHA or WHMIS (Workplace Hazardous Materi-als Information System) reportable levels. It is not mutagenic in various in vitro andin vivo test systems. A long-term exposure exceeding the TLV can lead to damagingeffect as a result of mechanical overloading of the respiratory tract. Animal testshave shown no indication of carcinogenic or reproduction effects. The ACGIHrecommends a TLV-TWA of 10 mg m–3 measured as total dust containing less than1% quartz. Acute toxicity LD50/LC50 values relevant to classification are: oral> 5000 mg/kg rat, dermal > 5000 mg/kg rabbit, inhalation > 0.139 mg/L/4 h rat.

20.5.3Hydrotalcites

Hydrotalcites are, in general, environmentally safe, non-toxic, non-corrosive, andnon-volatile. According to MSDS for Hysafe 539 (J. M. Huber), hydrotalcites with thegeneral name magnesium aluminum hydroxycarbonate (CAS #11097-59-9), OSHA –

20.5 Environmental/Toxicological Considerations

376

PEL values are 15 mg m–3 dust and 5 mg m–3 respirable dust and the ACGIH – TLVis 10 mg m–3 total dust [20]. The product itself is non-combustible. Acute toxicity val-ues of LD50 oral (rat) >10000 mg/kg have been reported. Dust inhalation may causerespiratory tract irritation.

20.6Applications


In plastics/rubber adhesives manufacture/compounding operations, magnesium ox-ide has different functions depending upon its type/grade and the specific applica-tion. For example, viscosity development in unsaturated polyester SMC may be con-trolled using magnesium oxide. In the production of polychloroprene rubber, mag-nesium oxide regulates the rate of vulcanization and prevents the development of“scorch”. In polychloroprene−phenol resin adhesives, magnesia acts as an acid ac-ceptor and is an integral part of the bonding system. In chlorosulfonated polyethyl-ene it acts as a cross-linker [21]. In all these and other acid-acceptor systems (e.g.,chlorinated polyethylene), fine particle size (usually less than 325 mesh) and closelycontrolled surface area (as high as 160−170 m2 g–1) or reactivity are vital [3,6,21].

The primary role of magnesium oxide in unsaturated polyesters (UP) is as a reac-tive filler/thickening agent in SMC/BMC manufacturing. Incorporation of up to 5%MgO in the formulation allows the production of a tack-free sheet within a few daysat room temperature. The sheets can then be easily handled prior to the high-tem-perature molding step. The increased viscosity assists dispersion uniformity and im-proves the f low characteristics of the compound prior to the onset of the cross-link-ing reaction.

Two mechanisms have been proposed in the literature to account for the interac-tion of polyester and magnesium oxide in the reactive monomer (e.g. styrene) medi-um. One is a chain-extension mechanism and the second is the formation of a coor-dinate complex, also known as the two-stage thickening mechanism [22]. The com-mon starting point for these two mechanisms is the formation of basic and neutralsalts with the polyester carboxylic acid (-COOH) end groups according to the follow-ing reactions:

—-COOH + MgO → —-COOMgOH—-COOH + HOMgOOC—- → —-COOMgOOC—- + H2O~~UP~COO-+Mg+-OOC~UP~COO-+Mg+-OOC~UP~~~

In the chain-extension mechanism, it is postulated that dicarboxylic acid groups onthe UP chains react with MgO to produce a very high molecular weight (MW) species(through condensation polymerization), and thus give rise to a large increase in vis-cosity. However, this theory only applies to those UP molecules terminated by car-

20 Processing Aids

377

boxylic groups with the structure HOOC~~~COOH. For other possible polyesterstructures, such as HOOC~~~OH and HO~~~OH, the MW of the polyester will on-ly increase two-fold or not at all if the chain-extension mechanism is followed. Ac-cordingly, this mechanism cannot explain the large increase of viscosity in the thick-ening of UPs with structures having terminal OH functional groups, as reported inother publications [23].

In the two-stage mechanism, it is postulated that a high MW salt is formed initial-ly, and then a complex is formed between the salt and carbonyl groups of the esterlinkages, as shown below. The second stage of this theory is considered to be re-sponsible for the large increase in viscosity. Several publications support this mech-anism, for example refs. [23,24].

In recent work [25,26], attempts have been made to produce similar magnesiumsalts by melt processing mixtures of an unsaturated polyester oligomer (in the ab-sence of a reactive monomer) with various amounts of MgO. As an example, Fig-ure 20-4 show the effects of increasing MgO concentration on the reaction betweenUP and MgO in a batch mixer at 220 °C. As the torque data indicate, the reaction rate

~~~~UP-C-O-UP~~~

~~~UP~COOMgOOC~UP~COOMgOOC~UP~~~

O

O

~~~UP-C-O-UP~COOMgOOC~UP~COOMgOOC~UP~

20.6 Applications

Time (s)

Torq

ue (

Nm

)

Fig. 20-4 Changes in torque of an unsaturated polyester oligomerupon addition of various amounts of MgO in a batch mixer at220 °C [25,26].

378

increases significantly with increasing MgO concentration. Thus, depending on con-centrations and processing conditions, different products with different rheologicalcharacteristics and different glass transition temperatures are possible. Such prod-ucts may be considered as thermoplastic ionomers containing unsaturation thatmight be useful for further reactions.

20.6.2Fumed Silica

Both hydrophilic and hydrophobic silicas are used in solvent-borne coatings to im-prove rheological properties, and as f low control agents and anti-settling additives forpigments [1]. Fumed silica is a weak acid, bearing hydroxyl groups on its surface. Thethickening mechanism of liquid coating systems can be explained in terms of hydro-gen-bond formation between neighboring aggregates of silica, leading to the forma-tion of a regular network. Some of these hydrogen bonds may be broken under shearforces, resulting in reduced viscosity. Usually, fumed silica contains 0.5−2.5% mois-ture, which not only aids the thickening process, but also facilitates curing of somepolyurethane pre-polymer systems [21]. Fumed silica is a thixotropic additive, which,when dispersed into epoxy or polyester resins, increases viscosity, imparts thixotrop-ic behavior, and adds anti-sag and anti-settling characteristics during the potlife of theresins.

Besides its principal use in coating/paint systems, fumed silica may also be usedin thermoplastics. The melt viscosity of poly(ethylene 2,6-naphthalate) (PEN) was re-ported to be decreased following the incorporation of small amounts of fumed silicananoparticles. Additional effects on mechanical properties and crystallization behav-ior obtained by using untreated and surface-treated silica are reported in ref. [27].

In addition to their function as rheology modifiers, silica particles, colloidal orfumed, and clays are among the most widely studied inorganic fillers for improvingthe scratch/abrasion resistance of transparent coatings. These fillers are attractivefrom the standpoint that they do not adversely impact the transparency of coatingsdue to the fact that the refractive indices of the particles closely match those of mostresin-based coatings. The drawback of silica-based fillers is that high concentrationsof the particles are generally required to show a significant improvement in thescratch/abrasion resistance of a coating, and these high loadings can lead to variousother formulation problems associated with viscosity, thixotropy, and film formation[28,29].

20.6.3Hydrotalcites

Synthetic hydrotalcites are basically used as process stabilizers for PVC, acting asscavengers for the HCl evolved during processing and hence minimizing degrada-tion. They are mainly used as co-stabilizers with metal soap stabilizers to improve col-or and heat stability [9] at concentrations from 0.5 to 3.5 phr. Hydrotalcites are alsoused as acid neutralizers in polyolefins, particularly those containing residual acidic

20 Processing Aids

379

products from Ziegler−Natta catalysts used for polymerization. In the latter case, theyare part of the normal stabilizer packages at typical concentrations in PP, LLDPE, andHDPE not exceeding 500 ppm [9]. They function as processing aids, minimizing cor-rosion of metallic surfaces during processing and also lead to less discoloration of themolded products. Studies on model hydrotalcite/phenolic anti-oxidant compositions,in the absence of polyolefins containing residual acidity, confirmed that color(“pink”) development at processing temperatures is due to enhanced hydrotalcite/an-ti-oxidant interactions, and occurs irrespective of the hydrotalcite used [30].

The hydrotalcites may also exhibit a series of additional functions. For example,due to their ability to release water and carbon dioxide at relatively low temperatures,they have been extensively evaluated as f lame retardants. Recent data obtained witha mass-loss calorimeter indicated that an EVA polymer filled with 50 wt. % hydrotal-cite had the slowest heat release rate and the lowest evolved gas temperature as com-pared with aluminum hydroxide or magnesium hydroxide. XRD data, combined withthermal analysis results, indicated that the layered structure of hydrotalcite may playa role in the degradation mechanism. The improved fire resistance of EVA filled withhydrotalcite also results from its intumescing behavior [31].

A hydrotalcite compound containing specific interlayer ions showed excellent abil-ity to absorb infrared rays and excellent light transmission when used in an agricul-tural film [32].

Hydrotalcite has also been shown to be a suitable replacement for lead-containingstabilizers in heat- and water-resistant chlorosulfonated polyethylene formulations[33]. Finally, recent research on bioactive composites (see also Chapter 22) suggeststhat hydrotalcite promotes both thermal and hydrolytic degradation of poly(L-lacticacid) and may also promote bioactivity. In addition to their reinforcing characteris-tics, the inherent acid-neutralizing capacity of hydrotalcites provides pH control dur-ing polymer biodegradation, which is usually accompanied by the formation of acidiclow molecular weight fragments. This could be of considerable importance in in vivotissue engineering applications [17], where rigorous pH stability is required.

20.6 Applications

380

References

1 Koleske, J. V., et al., 2004 Additives Guide,Paint & Coatings Industry Magazine, 2004,20(4), 32; also available at www.pcimag.com

2 Manshausen, P., “Role and Function ofRheological Additives in Modern Emulsionand Industrial Coatings”, available atwww.pcimag.com/cda/articleinformation

3 Martin Marietta Magnesia Specialties LLCtechnical information at www.magspecial-ties.com/students.htm and www.magspe-cialties.com/MSDS

4 Premier Periclase, Ireland, at www.premier-periclase.ie/ and at www.psi-net.org/chem-istry/s2/magnesiumoxide.pdf

5 Chemicals Australia Consultants atwww.chemlink.com.au/mag&oxide.htm

6 Dead Sea Periclase atwww.periclase.com/products/rubber04.htm

7 Kyowa Chemical Industry Co. Ltd., techni-cal information at www.kyowa-chem.co.jp/english/

8 Cox, S. D., Wise, K. J., Minerals Technolo-gies, USA, US 5,364,828 (1994).

9 Zweifel, H., (ed.), Plastics Additives Hand-book, Chapter 4, Hanser Publishers, Mu-nich, 2001.

10 Ashton, H., Proc. Functional Fillers for Plas-tics 2003, Intertech Corp. Atlanta, GA, Oct.2003.

11 Crystran Co., “Magnesiun Oxide – DataSheet” at www.crystran.co.uk/mgodata.htm

12 Katz, H. S., Milewski, J. V. (Eds)., Handbookof Fillers and Reinforcements for Plastics,Chapters 8 and 10, Van Nostrand ReinholdCo., NY (1978).

13 Mindat., “Hydrotalcite” atwww.mindat.org/min-1987.html

14 Mineralogy Database “Hydrotalcite – min-eral data” athttp://webmineral.com/data/Hydrotalcite.shtml

15 Miyata, H., et al., Clay and Minerals 1977,25, 14–18.

16 Sasol North America, Product Information,at www.sasoltechdata.com/

17 Chouzouri, G., et al., Proc. 62nd SPE AN-TEC 2004, 50, 3366.

18 NIOSH (National Institute for OccupationalSafety and Health), “Magnesium Oxide” atwww.cdc.gov/niosh/rtecs/om3abf10.html

19 Cabot Corp., MSDS for CAB-O-SIL® un-treated fumed silica, revised Aug. 2000.

20 J. M. Huber Corp., MSDS for Hysafe® 539,revised Dec. 17, 2003.

21 Wypych, G., Handbook of Fillers, ChemTecPubl., Toronto, Ont., Canada, 2000, pp.132–137, 651–652.

22 Judas, D., et al., J. Polym. Sci.: Polym. Chem.Ed. 1984, 22, 3309.

23 Vancsó-Szmercsányi, I., Szilágyi, Á, J.Polym. Sci.: Polym. Chem. Ed. 1974, 12, 2155.

24 Rao, K. B., Gandhi, K. S., J. Polym. Sci.:Polym. Chem. Ed. 1985, 23, 2135.

25 Wan, C., “Reactive Modification of Poly-esters and Their Blends”, Ph.D. Thesis,New Jersey Institute of Technology, Newark,NJ, 2004.

26 Wan, C., Xanthos, M., Proc. 61st SPE AN-TEC, 2003, 49, 1503.

27 Seong, H. K., et al., Proc. 62nd SPE ANTEC,2004, 50, 1357.

28 “Nanoparticle Composites for Coating Ap-plications”, Paint & Coatings Industry maga-zine, May 2004, accessed atwww.pcimag.com/cda/articleinformation/

29 Degussa Corp., Technical Library Docu-ment GP-89, accessed at www.epoxyprod-ucts.com/silica.html

30 Patel, S. H., et al., J. Vinyl & Additive Tech.1995, 1(3), 201.

31 Camino, G., Polym. Degrad. and Stability2001, 74(3), 457.

32 Takahashi, H., Okada, A., Kyowa ChemicalIndustry Co. Ltd., US 6,418,661 (2002).

33 Fuller, R. E., Macturk, K. S., KGK-Kautschukund Gummi Kunststoffe 2000, 53(9), 506.

20 Processing Aids

21Glass and Ceramic Spheres

Marino Xanthos

21.1Introduction

Glass and ceramic spheres are used extensively in thermoplastics and thermosets.They may be solid or hollow, with densities varying from 2.5 g cm–3 to 0.1 g cm–3. Thespheres aspect ratio of unity may provide only moderate positive effects for their useas mechanical property modifiers (see Chapter 2). More important functions for bothsolid and hollow spheres are associated with their spherical form; they include en-hanced processability and dimensional stability. Overall reduced composite density isan additional, and most important, function of hollow spheres.

21.2Production and Properties

Solid glass spheres are produced by firing crushed glass with subsequent collectionand cooling of the spheroid product, or by melting the formulated glass batch andsubsequent break-up of the free-falling molten stream to form small droplets [1]. Thespheres are mostly based on the A-glass composition (see Chapter 7), although E-glass spheres are also available. A-glass is recommended for all polymers except foralkali-sensitive resins such as polycarbonate, acetal, and PTFE, for which the use ofE-glass is recommended [2]. Commonly used mean particle sizes may range from200 to 35 µm. Essentially all glass beads are coated with appropriate coupling agentsdepending on the intended use. Metal-coated, e.g. silver-coated, spheres are also avail-able for shielding applications. They are not considered as hazardous materials andnuisance dust OSHA exposure limits are applicable.

Hollow glass spheres are produced by heating crushed glass containing a blowingagent. During the process of liquefaction the surface tension causes the particles toassume a spherical shape. The gas formed from the blowing agent expands to formhollow spheres. Other methods involve passing spray-dried sodium borosilicate so-lutions through a f lame or spray-drying modified sodium silicate solutions [1,3]. The


382

mean particle size may range from 20 to 200 µm depending on the manufacturer, andas in the case of the solid beads they are also coated with a variety of coupling agents.The crush strength of hollow spheres is determined by the thickness of the walls and,as expected, the higher the sphere density the higher the crush strength. Available ef-fective densities are much lower than those of all polymers, ranging from 0.15 to0.8 g cm–3; these values correspond to crush strength values ranging from 2 to>150 MPa. Low density spheres are shear-sensitive and broken fragments can causeexcessive wear in processing equipment. This largely restricted their first uses to liq-uid or low-pressure processes, such as casting or in open-mold reinforced thermosets(see Chapter 1), and selected thermoplastic operations such as plastisol processing.The development of higher strength spheres led to uses in higher shear, thermosetclosed-mold applications, and eventually in extrusion and injection molding of ther-moplastics. Denser hollow spheres with density 0.6 g cm–3 are reported to have suf-ficient crush strength to withstand injection molding [4]. Some physical properties ofsolid and hollow glass spheres are compared in Table 21-1 [3,5]. It should be notedthat the chemical properties of both solid and hollow glass spheres are basically thoseof their precursor silicate glasses.

Tab. 21-1 Comparison of A-glass based solid and hollow glass spheres

Property Solid Hollow

softening temp., °C 700 700density, g cm–3 2.3–2.5 0.1–1.1hardness (Mohs) 5.5–6 5modulus, GPa 60–70 200thermal conductivity, W m–1 K–1 0.7 0.0084thermal expansion coefficient, K–1 8.6 × 10– 6 8.8 × 10– 6

dielectric constant, 104 Hz 5 1.5

Ceramic hollow spheres are aluminosilicates produced from a variety of mineralsor reclaimed from f ly ash waste. Ceramic spheres have higher densities than glassbeads, but are less expensive, more rigid, and mechanically more resistant, appar-ently due to their thicker walls. Their true densities may vary from 0.3 to 0.8 g cm–3

with mean particle sizes ranging from 30 to 125 µm. Figures 21-1 and 21-2 show com-mercial ceramic microspheres with a broad particle size distribution; wall thicknessis estimated to be about 1 µm. The properties of the available ceramic spheres aredescribed in detail in refs. [1,3].

21.3Functions

The spherical form of these fillers gives rise to distinct functions and properties com-pared with other directional fillers. These include:

21 Glass and Ceramic Spheres

383

high packing fractions, little effect on viscosity, and high attainable loading levels(up to 50 vol. %);

better f low characteristics than high aspect ratio fillers; more uniform stress distribution around the spherical inclusions and better di-

mensional stability; no orientation effects and enhanced isotropy; reduced uniform and predictable shrinkage and less warpage in injection-molded

parts;

21.3 Functions

Fig. 21-1 Photomicrograph of commercial ceramic microsphereswith a top size of about 50 µm and a broad particle size distribution;300× (courtesy of Dr. S. Kim, Polymer Processing Institute, Newark, NJ).

Fig. 21-2 Photomicrograph of a broken ceramic microsphere with awall thickness of about 1 µm; 7430× (courtesy of Dr. S. Kim, Poly-mer Processing Institute, Newark, NJ).

384

smoother surfaces than with directional fillers.

Effects on mechanical properties depend on the particle size, volume fraction, andsurface treatment (see Chapter 2). With solid spheres, mechanical properties that arepositively affected are usually modulus, compressive strength, and, in certain cases,tensile strength. Properties that are adversely affected are ductility, manifested aselongation at break, and often impact strength. A variety of thermoplastics, such asPS, nylon, SAN, ABS, PC, and PVC, containing solid glass spheres or glass fiber/sphere combinations, are used in automotive, appliances, and connectors applica-tions. An increase in extrusion throughput with increasing glass bead loading hasbeen reported for nylon compounds [6], and increased scratch and abrasion resist-ance in nylon and HDPE have been documented [6,7]. Thermoset applications in-clude incorporation into epoxies, polyesters, polyurethanes, and silicones, and as par-tial replacement of the fibers in fiberglass-reinforced materials.

For hollow spheres, in addition to the effects imparted by their spherical shape,their most important function is density reduction. The effects on mechanical prop-erties are strongly dependent on loadings and wall thickness of the spheres, but a cer-tain modulus increase is usually accompanied by lower tensile and impact strengths.High dielectric strength, reduced dielectric constant, and good thermal insulation areadditional attributes imparted by hollow spheres. A significant application of hollowspheres is in thermoset syntactic foams based on liquid epoxy, polyurethane, andpolyester materials, as well as in vinyl plastisols and foams and in other materialsprocessed at low pressures. Applications are found in cultured marble, the automo-tive industry, recreational items, sports goods, the electronics industry, ablative com-posites, and in f lotation and buoyancy [8]. In injection-moldable thermoplastics suchas nylon or PP, high crush strength hollow spheres with a density of 0.6 g cm–3 haveproduced composite density reductions very close to those theoretically predicted onthe basis of Eq. 1-4 [4]. This is an indication of the very small degree of breakage dur-ing processing. Considering the fairly high cost of the high strength glass bubbles(about US$ 6.00/kg, 2003 prices), overall cost reductions calculated from Eq. 1-5 areonly possible for some higher priced engineering thermoplastics such as PEEK.

21.4Suppliers

Major suppliers of solid glass spheres are: Potters Industries, Inc., Valley Forge, PA;Sovitec France, Florange, France. Major suppliers of hollow glass and ceramicspheres are: 3M Specialty Additives, St. Paul, MN; PQ Corp. Valley Forge, PA; Gref-co Minerals, Inc., Torrance, CA; Potters Industries, Inc., Valley Forge, PA; Enviros-pheres Pty. Ltd., Australia; Trelleborg Fillite Ltd, U.K.; Advanced Minerals Ltd., U.K.,Cenosphere Co., TN; Zeelan Industries, St. Paul, MN. Additional suppliers/distribu-tors are listed in refs. [3,9,10].

21 Glass and Ceramic Spheres

385

References

1 Katz, H. S., Milewski, J. V. (Eds.), Handbookof Fillers and Reinforcements for Plastics,Chapters 18 and 19, Van Nostrand ReinholdCo., New York, 1978.

2 Potters Industries Inc., http://www.potters-beads.com/markets/polyspheriglass.asp;accessed March 28, 2004.

3 Wypych, G., Handbook of Fillers, ChemTecPubl., Toronto, Ont., Canada, 2000, pp72–75 and 87–91.

4 Israelson, R., Proc. Functional Fillers forPlastics 2003, Intertech Corp., Atlanta, GA,Oct. 2003.


6 Anonymous, “Solid glass beads offer majorbenefits for polyamide compounds”, Plas-tics Additives & Compounding 2002, 4(6),32–33.

7 Beatty C. L., Elrahman, M. A., “Fillers(Glass Bead Reinforcement)”, in ConcisePolymeric Materials Encyclopedia (Ed.: Sala-mone, J. C.), CRC Press, Boca Raton, FL,1999, pp 475–476.

8 Muck, D. L., Ritter, J. R., Plastics Compound-ing, Jan.–Feb. 1979, pp 12–28.

9 World Buyers’ Guide 2004, Plastics Additives& Compounding 2004, 5, 7.

10 Plastics Compounding, Redbook Directory,1999, Advanstar Communications, Inc.,Cleveland, Ohio.

References

22Bioactive Fillers

Georgia Chouzouri and Marino Xanthos

22.1Introduction

During the last decades, the need for biomaterials in dental, craniofacial, and ortho-pedic applications has increased and so has the necessity for further development ofnew engineering composite materials. These types of materials are required to pro-vide distinctive mechanical performance in bone growth applications, as well as bio-compatibility and biological active response, known as bioactivity. According toHench [1], bioactivity is the ability of a material to elicit a specific biological responseat its interface with a living tissue, which results in the formation of a bond betweenthe tissue and the material. It is essential to understand that no single biomaterial isappropriate for all tissue engineering applications, and also that the mechanical andbiological behavior can be “tailored” for a given application. The so-called biomedicalcomposites that are classified as bioinert, bioactive, and bioresorbable consist of amatrix, which can be metallic (e.g., titanium), inorganic (e.g., glass) or polymeric(e.g., HDPE), in combination with miscellaneous fillers [2]. Usually, these function-al fillers are responsible for the in vivo bioactive response, although the matrix mayalso exhibit a biological response in certain bioresorbable compositions containingdegradable synthetic (e.g. polylactic acid) or natural (e.g. collagen, polysaccharides)polymeric matrices. Certain types of glasses, ceramics, and minerals have been re-ported in the literature to act as bioactive fillers. They fall under the generic term ofbioceramics [3], which encompasses all inorganic, non-metallic materials made to beused in the human body. Such materials have been used on their own as implants,have been dispersed in matrices such as polymers and inorganics, or have been ap-plied as coatings on metallic implants.


388

22.2Bone as a Biocomposite

Bone is a natural composite material, the major components of which are type I col-lagen, calcium phosphate minerals (hydroxyapatite is the predominant one), carbon-ate-substituted apatite, and water [4]. Bone is a brittle anisotropic material with lowelongation at fracture (3–4%) and its properties may vary broadly. Tensile modulusand strength for a long human bone are reported to be 17.4 GPa and 135 MPa, re-spectively, in the axial direction, and much lower in the radial direction: 11.7 GPa and61.8 MPa, respectively [5]. Compressive strength, a property more relevant to actualuse of bone, is higher, approaching 196 MPa and 135 MPa in the axial and transversedirections, respectively. In biomedical polymer composites, attempts are made toreach these high modulus/strength levels through the introduction of high volumeloadings (as high as 45%) of bioactive fillers.

Bone has a complex structure with several levels of organization. In developingbone substitutes, two structure levels are considered. The first is a bone apatite-rein-forced collagen that forms lamellae at the nm to µm scale, and the second is the os-teon-reinforced interstitial bone at the µm to mm scale. Figure 22-1 depicts the struc-tural organization of the bone in the human body [2].

The apatite–collagen composite prompted researchers to investigate composites ofbioactive ceramics in polymer matrices as alternatives for bone replacement [2]. Usu-

22 Bioactive Fillers

Fig. 22-1 Structural organization of the bone in the human body. (Reprinted from ref. [2], Copyright 2003, with permission from Elsevier).

389

ally, these polymers can be either biostable or biodegradable, depending on the in-tended application. In composites containing bioactive ceramics in biodegradablepolymers, the rate of degradation needs to be controlled so that mechanical integrityis retained in the early stages of bone healing. Table 22-1 lists some polymers thathave been reported in the literature to be widely used for the production of biocom-posites for tissue engineering applications.

Tab. 22-1 Examples of polymers used in tissue engineering applications.

Biostable polymers Biodegradable polymers

polyethylene (PE, HDPE) polylactic acid (PLA)polyether ether ketone (PEEK) polyglycolic acid (PGA)polysulfone (PSU) poly-ε-caprolactone (PCL)polyurethane (PU) poly-β-hydroxybutyrate (PHB)poly(methyl methacrylate) (PMMA) poly-δ-valerolactonebisphenol-A-glycidyl methacrylate (bis-GMA) blends of starch with ethylene vinyl alcohol (SEVA)

Bone regeneration through bioactive fillers is accelerated by the formation of abonding layer with apatite structure that is reinforced with collagen. This apatitestructure can be formed when the material comes into contact with human bloodplasma. Nucleation of the apatite layer may be promoted by a variety of functionalgroups (carboxyl, hydroxyl) that are present on the filler surface or are formedthrough contact with physiological f luids. In order to screen the bioactivity of a ma-terial in vitro, researchers have used simulated body f luid (SBF), which contains sim-ilar ion concentrations and has similar pH to human blood plasma [6]. For certainsystems, the concentrations of calcium and phosphorus in the biological or simulat-ed body f luid can support the formation of the apatite layer needed for bone in-growth, making the presence of these elements in the filler structure unnecessary.

22.3Bioceramics Suitable for Tissue Engineering Applications

Bioceramics is a generic term that covers all inorganic, non-metallic materials thathave been used in the human body as implants or prostheses. According to the typeof tissue attachment, there are three types of bioceramics: bioinert, bioactive, andbioresorbable [1,7]. Inert bioceramics are biologically inactive, with a characteristiclack of interaction between the tissue and the bioceramic and vice versa. Common ex-amples are alumina and zirconia. Bioactive ceramics are surface-reactive ceramicsthat form a bonding layer with the tissue in order to accelerate the bone growth. Typ-ical examples are hydroxyapatite, bioactive glasses, and bioactive glass ceramics.Bioresorbable active ceramics are designed to gradually degrade and be slowly re-placed by the host tissue. Tricalcium phosphates, calcium phosphate salts, and calci-um carbonate minerals are common bioresorbable ceramics. Table 22-2 summarizes

22.3 Bioceramics Suitable for Tissue Engineering Applications

390

the most significant ceramics that have been reported to show bioactivity in vitro andin most cases in vivo.

Tab. 22-2 Bioactive fillers used in tissue engineering applications.

Chemical name Chemical formula

hydroxyapatite Ca10(PO4)6(OH)2

dicalcium phosphate CaHPO4·2H2Oβ-tricalcium phosphate Ca3(PO4)2

tetracalcium phosphate Ca4P2O9

calcium carbonate CaCO3

wollastonite CaSiO3

bioactive glasses (see Table 22-3)A-W glass ceramics Ca10(PO4)6(OH,F)2 – CaSiO3 in MgO/CaO/SiO2 matrix

Similarly to other functional fillers, the shape, size, size distribution, pH, and vol-ume percentage of the bioactive filler, as well as the type and level of bioactivity, andthe filler distribution in the matrix play important roles in determining the proper-ties of the composites. In addition, the matrix properties, the filler–matrix interfacialstate, as well as the processing parameters are of great importance with regard to theperformance of the final biomaterial [2]. In the following sections, some examples ofcomposites containing specific bioactive and some bioresorbable active ceramics arepresented. Matrices to be considered are mostly polymers. The majority of such com-posites are prepared by conventional melt processing methods (extrusion com-pounding followed by injection or compression molding), although some compositesare prepared by solution casting techniques. Attempts have been made to simulatebone structure and properties through specialized forming technologies includingshear-controlled orientation injection molding (SCORIM® ) [8] and hydrostatic ex-trusion [2]. The state-of-the-art and recent developments in bioinert, biodegradable,and injectable polymer composites for hard tissue replacement have recently been re-viewed by Mano et al. [8].

22.4Bioceramics as Functional Fillers

22.4.1Hydroxyapatite (HA)

HA is considered to be a biocompatible and osteoconductive material, exhibiting on-ly an extracellular response leading to bone growth at the bone–filler interface [2]. Bycontrast, osteoproductive fillers such as bioactive glasses (see Section 22.4.4) elicitboth an extra- and an intracellular response at the interface [3]. None of these fillersare osteoinductive, since the presence of bone morphogenic proteins, BMP, and/orother growth factors (e.g. insulin growth factor IGF-I) are required. Researchers have


391

used several bioceramics along with osteoinductive materials in order to promotefaster bone regeneration, but this is beyond the scope of this chapter. One of the mainreasons why hydroxyapatite (HA) has been investigated as a bioactive filler is its sim-ilarity to the biological hydroxyapatite in impure calcium phosphate form found inhuman bone and teeth. HA has a Ca:P ratio of 10:6 and its chemical formula isCa10(PO4)6(OH)2. The biological HA additionally contains magnesium, sodium,potassium, and a poorly crystallized carbonate-containing apatite phase, as well as asecond amorphous calcium phosphate phase [9].

Several biocomposites containing HA as a filler have been described in the litera-ture, although not all of them have achieved clinical success. The matrix in the mostwidely known and investigated HA composites is high density polyethylene (HDPE),a biocompatible and biostable polymer broadly used in orthopedics. The compositeknown as HAPEX, first introduced by Smith & Nephew Richards in 1995 [2], wasthe result of pilot studies, laboratory testing, clinical trials, and pilot plant productionefforts spanning a period of about 15 years until regulatory approval was attained [3].A range of 0.2 to 0.4 volume fraction HA was determined to be optimal. HAPEXwas the first composite designed to mimic the structure and retain the properties ofbone, and is mainly used for middle ear implants. It has mechanical properties sim-ilar to those of bone and it is easy to trim, which allows surgeons to precisely fit it atthe time of implantation. The goal of the inventors is to produce similar compositematerials that can carry greater loads for other parts of the body. Both commerciallyavailable and “in house” synthesized hydroxyapatites have been evaluated, along withdifferent polyethylene types. By varying the amount and particle size of HA, a rangeof mechanical properties approaching those of bone, and different degrees of bioac-tivity can be obtained depending on the application [2]. The in vitro and in vivo re-sponses have also been extensively assessed. In human osteoblast cell primary cul-tures used for in vitro experiments, the osteoblast cells appeared to attach to HA; cellproliferation followed, thus confirming the bioactivity of the composites. In in vivoexperiments with adult rabbits, the composite implant surface was covered by newlyformed bone.

Sousa et al. [10] investigated HDPE filled with 25 wt. % commercially available HAwith average particle size of 10 µm, produced by melt mixing and followed by shear-controlled orientation injection molding (SCORIM®) to simulate bone structure.Sousa et al. [11] also produced composites of blends of starch with ethylene vinyl al-cohol (SEVA-C) with 10, 30, and 50 wt. % hydroxyapatite by twin-screw extrusioncompounding followed by SCORIM®, as well as by conventional injection molding.SCORIM® processing appeared to improve the stiffness of the composites, givingbetter control over their mechanical properties compared to conventional injectionmolding. No data were reported regarding the bioactivity of these composites. Simi-larly, SEVA-C filled with 30 wt. % commercially available HA was produced by Leonoret al. [12] by means of melt mixing followed by injection molding to create circularsamples in order to study the formation of a calcium phosphate layer when immersedin a simulated body f luid (SBF) solution. Figures 22-2 and 22-3 are scanning electronmicrographs (SEMs) of the unfilled and filled matrix before and after immersion in

22.4 Bioceramics as Functional Fillers

392

the SBF and clearly show bonding to the tissue apatite layer in the case of the filledpolymer.

Biocomposites of polysulfone (PSU) filled with 40 vol. % HA have also been pro-duced for hard tissue replacement [2]. PSU is a better matrix candidate than HDPEdue to its higher strength and modulus, which can provide better performancein load-bearing applications. PSU/HA composites were produced similarly toHA/HDPE composites by using conventional compounding methods, followed bycompression or injection molding in order to give the desired shape. By increasingthe HA content, the stiffness of the composite was increased to levels close to the low-


Fig. 22-2 Scanning electron micrographs of an unfilled SEVA-C sur-face before (a) and after 7 days (b) immersion in SBF. (Reprintedfrom ref. [12], Copyright 2003, with permission from Elsevier).

393

er limit for human bone. Of particular importance in this and other composites con-taining HA and bioactive glass is the control of the polymer/filler interfacial strength,a complex problem as bioactivity is a surface-related phenomenon.

22.4.2Calcium Phosphate Ceramics

Different phases of calcium phosphate ceramics have been used depending on theapplication. The stability of these ceramics is subject to temperature and the presenceof water. In the body (T = 37 °C and pH 7.2–7.4), calcium phosphates are convertedto HA. At lower pH (< 4.2), dicalcium phosphate (CaHPO4·2H2O) is the stable phase.At higher temperatures, other phases of phosphate minerals, such as β-tricalciumphosphate (Ca3(PO4)2 ), which is chemically similar to HA with a Ca:P ratio of 3:2, andtetracalcium phosphate (Ca4P2O9) exist. Tricalcium phosphate (TCP) is not a naturalbone mineral component, although it can be partly converted into HA in the body ac-cording to the following reaction [1]:

4 Ca3(PO4)2 (solid) + 2 H2O → Ca10(PO4)6(OH)2 (surface) + 2 Ca2+ + 2 HPO42–

TCP is an osteoconductive and resorbable material, with a resorption rate depend-ent on its chemical structure, porosity, and particle size [9].

Composites of polyhydroxybutyrate (PHB), a natural biodegradable thermoplasticβ-hydroxy acid, with TCP have been prepared by conventional melt processing tech-nologies (extrusion-, injection-, or compression molding) [2,13,14]. In vitro experi-ments in SBF produced an apatite-like structure on the composite surface suggestingbioactivity. When immersion in SBF was extended to two months or more, the onsetof matrix degradation could be followed by the decrease in storage modulus.


Fig. 22-3 Scanning electron micrographs ofSEVA-C + 30% HA composite before (a) andafter 7 days (b) immersion in SBF, and at dif-ferent magnification (c) and of a different

cross-section (d). (Reprinted from ref. [12],Copyright 2003, with permission fromElsevier).

394

Composites of chitosan and β-TCP with improved compressive modulus andstrength have been prepared by a solid–liquid phase separation of the polymer solu-tion and evaporation of the solvent [8]. The composites exhibited bioactivity when im-mersed in SBF solution, as a result of the nucleation sites provided by the bioceram-ic. Variation of the polymer/filler ratio and development of different macroporousstructures resulted in products with potential applications in tissue engineering.

22.4.3Calcium Carbonate

Calcium carbonate (CaCO3) minerals can exist in the forms of vaterite, aragonite, andcalcite (see also Chapter 16). All forms have the same chemistry, but different crystalstructures and symmetries. Aragonite is orthorhombic, vaterite is hexagonal, and cal-cite is trigonal. Natural coral is calcium carbonate in the aragonite form (>98% Ca-CO3). It is a porous, slowly resorbing material with an average pore size of 150 µmand very good interconnectivity. For use in periodontal osseous defects, it can be sup-plied with an average particle size of 300–400 µm. The major advantage of calciumcarbonate is that while other bioactive materials such as HA have to go through theformation of carbonate-containing structures, calcium carbonate can pass over thisstep; consequently, this can result in a more rapid bone ingrowth [9].

An application of calcium carbonate as a bioactive filler was discussed by Kasugaet al. [15], who incorporated powders consisting mainly of vaterite prepared by a car-


Fig. 22-4 Cross-sectional scanning electronmicrographs of PLA-CaCO3 composites(CCPC) following exposure to SBF; (a, b) 20%

filler; (c, d) 30% filler; (a, c) 1 day; (b, d) 3 daysof immersion. (Reprinted from ref. [15], Copy-right 2003, with permission from Elsevier).

395

bonation process in methanol into a polylactic acid (PLA) matrix. Composites con-taining 20–30 wt. % vaterite were prepared through solution mixing and immersedin SBF at 37 °C. Scanning electron micrographs (Figure 22-4) showed the formationof a thick apatite layer even after a short period of 1–3 days.

22.4.4Bioactive Glasses

Special compositions of glass appear to have the ability to develop a mechanicallystrong bond to bone. The so-called bioactive glasses contain SiO2, Na2O, CaO, andP2O5 in specific ratios [1,9,16]. Bioactive glasses differ from the traditional soda-lime/silica glasses (see also Chapter 7) as they have to contain less than 60 mol. %SiO2, have high Na2O and CaO contents, and have a high CaO/P2O5 ratio. As a result,when these glasses are exposed to physiological liquids they can become highly reac-tive. This feature distinguishes the bioactive glasses from bioactive ceramics such asHA. When the latter comes into contact with physiological f luids, both its composi-tion and physical state remain unchanged, in contrast to the bioactive glass, whichundergoes a chemical transformation. A slow exchange of ions between the glass andthe f luid takes place [7], resulting in the formation of a biologically active carbonatedHA layer that provides bonding to the bone and also to soft connective tissues. Sili-con and calcium slowly dissolved from the glasses activate families of genes in oldbone cells, which then form new bone cells [3].

Most of the bioactive glasses are based on bioglass designated as 45S5, which im-plies 45 wt. % SiO2 and a CaO/P2O5 molar ratio of 5:1. Glasses with a lower CaO/P2O5

ratio will not bond to bone. Nevertheless, based on modifications of the 45S5 bio-glass, a series of other bioactive glasses has been investigated by substituting, forinstance, 5–15 wt. % SiO2 with B2O3 or 12.5 wt. % CaO with CaF2 [1,7,9,17,18].Table 22-3 provides typical compositions of bioactive glasses.

Tab. 22-3 Bioactive glasses and their compositions in weight percent [1]

Glass SiO2 P2O5 CaO CaF2 Na2O B2O3 MgOdesignation

45S5 45 6 24.5 – 24.5 – –45S5F 45 6 12.25 12.25 24.5 – –45S5.4F 45 6 14.7 9.8 24.5 – –40S5B5 40 6 24.5 – 24.5 5 –45S5.OP 45 – 24.5 – 30.5 – –45S5.M 48.3 6.4 – – 26.4 – 18.5

A large variety of bioactive glass polymer composites has been investigated. For ex-ample, Narhi et al. [19] explored the biological behavior of a composite based on acopolymer of degradable poly(ε-caprolactone-co-DL-lactide) filled with glass S53P4 inexperimental bone defects in rabbits. The glass granules were varied in diameterfrom less than 45 µm to 90–315 µm. Bone ingrowth was mainly observed in the su-


396

perficial layers of the composites containing filler of larger particle size and at high-er concentrations. In another example, Bioglass®-reinforced polyethylene containing30 vol. % filler prepared by melt processing methods exhibited excellent biocompat-ibility and enhanced osteoproductive properties compared with the HAPEX mate-rial containing HA. Microscopic examination of the interface between human os-teoblast-like cells and the composite indicated direct bonding between the hydroxy-carbonate apatite layer formed on the filler particles in vitro and the cells [20–22]. Bio-glass®/polysulfone composites have been shown to provide a closer match to themodulus of cortical bone, with an equivalent strain-to-failure ratio [3].

22.4.5Apatite–Wollastonite Glass Ceramics (A-W)

A-W glass ceramics (AWGC) consist of crystalline apatite [Ca10(PO4)6(OH)F2] andwollastonite (CaSiO3) (see also Chapter 14) in an MgO/CaO/SiO2 glassy matrix. Thenominal composition by weight is MgO 4.6%, CaO 44.7%, SiO2 34.0%, P2O5 16.2%,CaF2 0.5% [23]. This composition has been used as a bone replacement material dueto its high bioactivity and its ability to instantaneously bond to living tissue withoutforming a fibrous layer. The mechanical properties of AWGCs are better than thoseof both bioactive glass and HA [1,23–25]. In addition, AWGCs appear to have long-term mechanical stability in vivo, as they chemically bond to living bone within 8–12weeks after implantation [24,25]. According to Hench [1], the addition of Al2O3 orTiO2 to the AWGC may inhibit bone bonding.

Shinzato et al. [23] evaluated AWGCs as fillers in bisphenol A/glycidyl methacry-late (bis-GMA) composites. An AWGC filler with an average particle size of 4 µm wassynthesized and incorporated into the polymer at 70 wt. %. The composite had acured surface on one side and an uncured surface on the other in order to evaluate itsbone bonding ability. Such composites were implanted into the tibiae of male whiterabbits. Direct bone formation through a Ca/P-rich layer was observed histologicallyonly for the uncured surfaces, presumably due to enhanced diffusion in the non-cross-linked state and faster exposure to the filler surface [23]. In another study [26],Juhasz et al. investigated composites of HDPE filled with AWGC of average particlesize in the range 4.4–6.7 µm at filler contents ranging from 10 to 50 vol. %. With anincrease in AWGC volume fraction, increases in Young’s modulus, yield strength,and bending strength were achieved, while the fracture strain decreased. Specifical-ly, a transition in fracture behavior from ductile to brittle was observed at certain fillerconcentrations. Based on mechanical and bioactivity test data, composites with50 vol. % AWGC appear to have potential as implants for maxillofacial applications.

22.4.6Other Bioactive Fillers

Within the class of bioceramics, wollastonite (CaSiO3) has also been shown to bebioactive and biocompatible. The structure of wollastonite is similar to that of bioac-tive glasses and glass ceramics and, consequently, the formation of an apatite layer


397

when it is in contact with biological f luids can be similarly explained. Liu et al. [27]used commercially available wollastonite of particle size in the range 10–60 µm forcoating Ti-6Al-4V substrates through a plasma-spayed method. The obtained speci-mens were further soaked in a lactic acid solution to activate surface functionalgroups, and then rinsed and immersed in SBF. As expected, an apatite layer wasformed through surface reactions.

In the family of glass ceramics, Ceravital™, a low-alkali, bioactive, silica glass ce-ramic, is extensively used [28]. It shows similar surface activity towards biological f lu-ids as bioactive glass. Gross and Strunz [1] observed, however, that even small addi-tions of Al2O3, Ta2O5, TiO2, Sb2O3 or ZrO2 could inhibit bone bonding. Anotherbioactive filler produced by combining carbonate-containing amorphous calciumphosphate, as a basic ceramic material, as well as crystalline carbonate (calcite) wasused by Schiller et al. [29] to formulate composites through solution mixing with poly-lactides and poly(D,L-lactide-co-glycolide). The filler was found to increase bioactivityand also maintained pH in the physiological range for long-term applications. Basedon preliminary bone growth results, such composites were seen to have future po-tential as skull implants with specific geometries.

22.5Modif ication of Bioceramic Fillers

In order to further improve their properties, bioactive ceramics have been modifiedby incorporating various elements. Yamasaki et al. [30] added magnesium, an im-portant element controlling biological functions, during the synthesis of functional-ly graded carbonate apatite crystals. The composites prepared with magnesium-con-taining filler appeared to promote higher bone density than those without it. Simi-larly, Blaker et al. [31] incorporated Ag2O into bioactive glasses. The silver-dopedglasses were used as coatings for surgical sutures and appeared to minimize the riskof microbial contamination without in any way compromising their bioactivity. In an-other study, Ito et al. [32] used zinc, a trace element found in human bone, to modifyTCP ceramics. TCP is an appropriate zinc carrier, since its crystal structure has sitesthat can accommodate divalent cations with ionic radius similar to that of zinc. Zincwas found to stimulate bone formation in vitro as well as in vivo when its concentra-tion was within non-cytotoxic levels.

22.6Fillers Formed In Situ

Surface area, pore volume, and pore size distribution are very significant factors inrelation to the surface reactivity of bioceramic materials. In situ filler formation bysol-gel methods (see also Chapter 23) has been used to prepare bioactive glasses withhigh surface area and porosity; as a result, for a given glass composition, an increasein growth rate of the interfacial apatite layer can be obtained.

22.6 Fillers Formed In Situ

398

Pérez-Pariente et al. [33] investigated the effect of composition and textural prop-erties on the bioactivity of glasses prepared by the sol-gel method by hydrolysis andpolycondensation of appropriate amounts of tetraethyl orthosilicate (TEOS), triethylphosphate (TEP), calcium nitrate (Ca(NO3)2·4H2O), and magnesium nitrate(Mg(NO3)2·4H2O) with 1 M HNO3 as a catalyst. After immersion of the formed gelin SBF, glasses with higher CaO content developed higher porosity, which led to ap-atite nucleation on the surface from the very first stages; in contrast, glasses withhigher SiO2 content were found to have increased surface area and as a result showedan increased growth rate of the Ca-P layer on their surfaces. Similar results were ob-tained by Peltola et al. [34], who prepared sol-gel derived SiO2 and CaO/P2O5/SiO2

compositions and examined their bioactivity in SBF. In contrast to a higher phos-phorus concentration, which was ineffective, a higher calcium concentration ap-peared to favor apatite nucleation.

In another study by Rhee [35], silanol groups appeared to provide nucleation sitesto favor the formation of apatite crystals in organic polymer/silica hybrids of low andhigh molecular weight polycaprolactone (PCL) prepared by the sol-gel method. Whenimmersed in SBF, fast and uniform nucleation and growth occurred in the case of thelow molecular weight hybrid, due to an increase in the number of interaction pointswith the silica and the decreased size of the silica domain. Additionally, the lower mo-lecular weight of PCL means faster degradation and faster exposure of the silicaphase to the SBF.


Considering the wide range of materials available for evaluation as fillers and matri-ces and the great variety in the available methods for composite preparation, it is clearwhy research in the field of tissue engineering is showing significant growth. It isvery important to recall that although there are many methods for producing bio-composites, each one is unique with a very specific application in its field. The ma-jor advantage of these materials, but also the biggest challenge in such research anddevelopment efforts, is the ability to “tailor” the properties of the resulting bioactiveand bioresorbable composites in order to enhance the process known as osseointe-gration. Particularly important is the need to balance mechanical properties (modu-lus, strength, fracture toughness) and, in the case of erodable matrices, degradationrate with biological response. Time for bone growth, the possibility of tissue inf lam-mation, and compensation for pH decrease due to the formation of acidic degrada-tion components, are important parameters that need to be taken into account. Con-tinued research aimed at identifying suitable bioactive functional fillers will un-doubtedly assist these efforts. It is recognized, however, that regulatory issues, andlong term performance and liability issues may extend the time interval between suc-cessful clinical trials and market introduction of new biocomposites.


399

References

1 Hench, L. L., “Classes of Materials used inMedicine”, Chapter 2 of Biomaterials Sci-ence. An Introduction to Materials in Medi-cine (Eds.: Ratner, B. D., Hoffman, A. S.,Schoen, F. J., Lemons, J. E.), AcademicPress, San Diego, 1996, pp. 73–84.

2 Wang, M., Biomaterials 2003, 24, 2133–2151.3 Hench, L. L., in http://www.in-

cites.com/papers/ProfLarryHench.htmlandhttp://www.bg.ic.ac.uk/Lectures/Hench/BioComp/Chap3.shtml

4 Lutton, C., et al., Aust. J. Chem. 2001, 54,621–623.

5 Callister, Jr., W. D., Materials Science and En-gineering, An Introduction, 6th Ed., John Wi-ley & Sons, Hoboken, NJ, 2003, p. 598.

6 Kokubo, T., et al., J. Biomed. Mater. Res.1990, 24, 721–734.

7 Krajewski, A., Ravaglioli, A., “Bioceramicsand Biological Glasses”, Chapter 5 of Inte-grated Biomaterials Science (Ed.: Barbucci,R., et al.), Kluwer Academic/Plenum Pub-lishers, New York, 2002, pp 208–254.

8 Mano, J. F., et al., Composites Science andTechnology 2004, 64, 789–817.

9 Ashman, A., Gross, J. S., “Synthetic Os-seous Grafting”, Chapter 8 of BiomaterialsEngineering and Devices: Human Applica-tions (Ed.: Wise, D. L., et al.), HumanaPress, New Jersey, 2000, pp 140–154.

10 Sousa, R. A., et al., Proc. 59th SPE ANTEC2001, 47, 2001.

11 Sousa, R. A., et al., Polym. Eng. Sci. 2002,42, 1032–1045.

12 Leonor, I. B., et al., Biomaterials 2003, 24,579–585.

13 Wang, M., et al., Bioceramics 2000, 13,741–744.

14 Wang, M., et al., Bioceramics 2001, 14,429–432.

15 Kasuga, T., et al., Biomaterials 2003, 24,3247–3253.

16 Hench, L. L., “Bioactive Ceramics”, Part IIof Bioceramics: Material Characteristics Ver-sus In Vivo Behavior (Eds.: Ducheyene, P.,Lemons, J. E.), The New York Academy ofSciences, New York, 1988, pp 54–71.

17 Fujibayashi, S., et al., Biomaterials 2003, 24,1349–1356.

18 Brink, M., et al., J. Biomed. Mater. Res. 1997,37, 114–121.

19 Närhi, T. O., et al., Biomaterials 2003, 24,1697–1704.

20 Huang, J., et al., Bioceramics 1997, 10,519–522.

21 Huang, J., et al., J. Mater. Sci.: Mater. Med.1997, 8, 809–813.

22 Huang, J., et al., Bioceramics 2000, 13,649–652.

23 Shinzato, S., et al., J. Biomed. Mater. Res.2000, 53, 51–61.

24 Yamamuro, T., et al., “Novel Methods forClinical Applications of Bioactive Ceram-ics”, Part II of Bioceramics: Material Charac-teristics Versus In Vivo Behavior (Eds.:Ducheyene, P., Lemons, J. E.), The NewYork Academy of Sciences, New York, 1988,pp 107–114.

25 Juhasz, J. A., et al., J. Biomed. Mater. Res.2003, 67A, 952–959.

26 Juhasz, J. A., et al., Biomaterials 2004, 25,949–955.

27 Liu, X., et al., Biomaterials 2004, 25,1755–1761.

28 Reck, R., et al.,“Bioactive Glass-Ceramics inMiddle Ear Surgery”, Part II of Bioceramics:Material Characteristics Versus In Vivo Be-havior (Eds.: Ducheyene, P., Lemons, J. E.),The New York Academy of Sciences, NewYork, 1988, pp 100–106.

29 Schiller, C., et al., Biomaterials 2004, 25,1239–1247.

30 Yamasaki, Y., et al., Biomaterials 2003, 24,4913–4920.

31 Blaker, J. J., et al., Biomaterials 2004, 25,1319–1329.

32 Ito, A., et al., Mat. Sci. Eng. 2002, 22, 21–25.33 Pérez-Pariente, J., et al., J. Biomed. Mater.

Res. 1999, 47, 170–175.34 Peltola, T., et al., J. Biomed. Mater. Res. 1999,

44, 12–21.35 Rhee, S., Biomaterials 2003, 24, 1721–1727.

References

23In Situ Generated Fillers: Organic–Inorganic Hybrids

Leno Mascia

23.1Introduction

Over the last ten years or so there has been considerable research interest in academicinstitutions in producing new materials from supermolecular combinations of anorganic and an inorganic component to form separate domains with dimensions be-tween 5 and 100 nm. These materials are generally known as organic–inorganic hy-brids, ceramers or nanocomposites. The latter term, however, is more widely used forsystems in which the discrete inorganic nanoparticles are produced by exfoliation ofcertain inorganic minerals, such as montmorillonite [1] (see also Chapter 9).

Organic–inorganic hybrid materials are designed to achieve the maximum possi-ble enhancement of properties for macromolecular materials through the in situ for-mation of nanostructured inorganic oxide domains. These materials have evolvedfrom the sol-gel technology used for the production of ceramic coatings, powders,and fibers. The diagram in Figure 23-1, adapted from ref. [2], illustrates both themechanism leading to the formation of a gel and the way in which the growth of solparticles controls the morphology and level of porosity in the final dried product (xe-rogel). Although several metal alkoxides have been studied for the production of hy-brid materials, such as those forming titania or zirconia as the inorganic phase, themost widely used precursors are based on tetraalkoxysilanes and produce silica do-mains.

From an examination of the gelation mechanism depicted in Figure 23-1, it can bededuced that it should be possible to produce a nanostructured composite materialby entrapping the organic component between the gel particles. Indeed, the additionof an organic component to an inorganic oxide precursor solution was originally usedsimply to prevent cracking during drying of the gel. The organic matter was drivenoff in the subsequent desiccation and sintering steps.

The concept of in situ generated fillers derives from the fact that the inorganicphase is produced during processing, unlike in the case of conventional fillers, whichare added as pre-formed particles, even though these may have been produced by the


402

sol-gel method. This approach has subsequently been adopted to produce non-sin-tered silica products, known as ormosils (organic modified silicas). In these materials,the organic component forms part of a heterogeneous network surrounding the sili-ca domains and provides the necessary conditions for the development of mechani-cal strength [3]. Gross phase separation is prevented by chemically binding the twophases so that the final material contains domains of the two components as co-con-tinuous phases, irrespective of the relative amounts present.

The organic–inorganic hybrids that are discussed in this chapter are those in whichthe organic component constitutes the major phase and is present in quantities in theregion of 50 to 85 wt. % [4]. Due to the continuity of the phases, the final product re-tains many of the properties of the two corresponding macromolecular components,hence the reason for the use of the adjective “hybrid” to describe the nature of thesematerials. This type of morphology, in fact, creates the conditions to achieve the mostefficient mechanism for the transfer of external excitations, thereby enhancing thecontribution of each phase to the properties of the hybrid material.

23 In Situ Generated Fillers: Organic–Inorganic Hybrids

Fig. 23-1 Mechanism of the formation of silica from hydrolyzedtetraethoxysilane by the sol-gel process.

403

23.2Methodology for the Production of Organic–Inorganic Hybrids

There are basically two approaches used for the production of organic–inorganic hy-brids.

In one case (Method I), the organic component forms a phase-separated macro-molecular network, which is chemically bound to the surrounding inorganic do-mains. In the second case, the organic component consists of a linear polymer inter-dispersed with inorganic domains. These two methods are outlined below, high-lighting the main advances that have been made in recent years [5].

Method IA suitably functionalized organic oligomer, bearing alkoxysilane terminal groups, ismixed with a solution of a “ceramifiable” metal alkoxide, usually in an alcohol andwater, although other polar aprotic solvents can be used. The organic and inorganiccomponents can react both intramolecularly and intermolecularly to form two chem-ically bonded nanostructured co-continuous domains, one composed of the metal ox-ide network and the other composed of the organic component. This morphology isachieved by controlling the rates of the competing hydrolysis and condensation reac-tions. In the early days of the development of these materials, the morphology wasfrequently controlled by allowing gelation to take place very slowly in a closed envi-ronment in order to prevent escape of the solvent [6–8].

The very early systems were obtained from hydroxyl-terminated low molecularweight (MW) polydimethylsiloxanes. These were reacted with tetraethoxysilane(TEOS) in the presence of water and an acid catalyst to form separate macromolecu-lar aggregates as depicted in Figure 23-2 (reconstructed from ref. [7]).

Other oligomers that have been used for this purpose include polytetramethyleneoxide and polycaprolactones, both of which bear alkoxysilane functional groups at

23.2 Methodology for the Production of Organic–Inorganic Hybrids

Fig. 23-2 Nanostructure of polytetramethylene oxide (PTMO)–silicahybrid.

404

their chain ends [9,10]. An interesting area with considerable commercial potentialis the use of an epoxy resin as the organic oligomer. The organic component can alsobe an unsaturated polymerizable monomer attached to an alkoxysilane. Thesemonomers will give rise to the formation of two separate networks as closely spacedco-continuous phases, one consisting essentially of a metal oxide network and theother containing the organic component of the hybrid material.

This is the traditional route used to produce ormosils and constitutes one of the ear-liest approaches to the production of hybrid materials. Particularly useful are themethacrylate alkoxysilanes due to the possibility of inducing the polymerization ofthe organic component by exposure to UV light [8]. There is a wide range of such sys-tems used commercially, particularly for coatings on organic glasses and as dentalmaterials [9].

Method IIThis method involves mixing solutions of high MW polymers and a metal alkoxide,often in combination with conventional functionalized alkoxysilane coupling agents.This is the preferred method to produce materials with a higher level of ductility thanwould be achievable by Method I. It is also the more difficult system due to the lowmiscibility of high molecular weight polymers and the associated propensity to ex-clude the inorganic component with the formation of large particulate domains[10–12].

23.3General Properties of Organic–Inorganic Hybrid Materials

The most important characteristics of hybrid materials, which arise from the co-con-tinuity of the two phases, are:

1. a large increase in modulus, strength, and hardness, particularly around the glasstransition temperature of the organic phase;

2. a considerable reduction in thermal expansion coefficient;3. enhanced barrier properties;4. notable improvement in thermal oxidative stability.

In the case of the properties listed as points 1 and 2, one notes that the co-continu-ity of phases produces conditions under which the strain in the two phases is equal.Hence, the global stress acting on the “composite” structure is the sum of the stress-es on the two phases weighted by the respective volume fractions, i.e.:

σh = K [Viσi + (1 – Vi)σo] (23-1)

where σ = stress, V = volume fraction, and K = phase efficiency factor, and the sub-scripts h, i, and o denote the hybrid, the inorganic phase, and the organic phase, re-spectively.


405

For conditions under which the strains in the two phases are equal, Eq. 23-1(known as the “law of mixtures”) can also be written in terms of Young’s modulus, E(see Chapter 2). The modulus obtained from the law of mixtures corresponds to themaximum value achievable in composite materials. For co-continuous phase com-posites it is likely to be in the region of 0.6–0.8, considering that it approaches unityonly for continuous fiber composites in the direction of the fibers. From this it canbe deduced that materials with very high modulus can be achieved in organic–inor-ganic hybrids, owing to the high modulus of the inorganic component.

The maximum stress that a material will withstand corresponds to the strength, i.e.the value of the stress at which fracture occurs. Due to the isostrain conditions with-in the two phases, it can be stipulated that fracture in the hybrid materials takes placewhen the most brittle phase, in this case the inorganic phase, reaches its maximumachievable strain value, σfracture. When this condition is reached, fracture will propa-gate rapidly through the organic phase without any further increase in strain. Ex-pressed in terms of fracture conditions, Eq. 23-1 becomes:

σh(fracture) = σfracture K [ViEi + (1 – Vi)Eo] (23-2)

where Ei = modulus of inorganic phase and Eo = modulus of organic phase.The strength of hybrids is limited, therefore, by the low strains to fracture exhibit-

ed by the inorganic phase.Using similar arguments for the prediction of hardness, it can be deduced that the

level of enhancement achievable is expected to be between that achievable in theYoung’s modulus and that in strength. The way in which these features can be ex-ploited to enhance the properties of microstructured composites will be illustratedlater.

The reasons for the improvements in diffusion-related properties, listed as points3 and 4 above, can be related to the tortuosity of the two co-continuous domains. Dif-fusion of gases occurs primarily through the organic phase, owing to the high densi-ty of the inorganic phase. Although it is difficult to make reliable quantitative esti-mates for such systems, it can be expected that the very small dimensions of the twophases constitute an additional factor for the enhancement of barrier propertiesthrough interfacial adsorption of the diffusing species. The improved thermal oxida-tive stability of the organic polymer in a hybrid material can also be related to the en-hancement in barrier properties provided by the inorganic phase. This reduces therate of inward diffusion of oxygen and outward diffusion of volatiles formed as a re-sult of degradation reactions.

For permeation phenomena involving the absorption of diffusing species throughswelling, the very large reduction in absorption of liquids achievable with organic–in-organic hybrids can be related to the isostrain conditions within the two phases. Sincethe inorganic component is impervious to the diffusing species and has a very highYoung’s modulus, it will severely restrict the swelling of the organic phase and dra-matically reduce the total amount of solvent being absorbed (see later).

Systems that have been investigated by the author’s research group over the lastfew years will be used to illustrate the above principles and to provide additional in-

23.3 General Properties of Organic–Inorganic Hybrid Materials

406

sights into the properties achievable with organic–inorganic hybrids and their po-tential applications.

23.4Potential Applications of Hybrids in Polymer Composites and Blends

23.4.1Use of Polyimide–Silica Hybrids in Polymer Composites

23.4.1.1 GeneralThere are two main areas in which hybrid materials have been explored to enhancethe performance of composites:

1. for continuous fiber/epoxy composites as fiber coatings, particularly for glassfibers to replace existing sizes;

2. as matrices for continuous carbon-fiber composites, replacing the polyimide ma-trix with the corresponding silica hybrid.

The chemical structures of the materials used to prepare the hybrids are shown inFigure 23-3. In both cases, the hybrid material is a polyimide–silica system obtainedby the sol-gel processing of a polyamic acid and an appropriate alkoxysilane solutionto create the silica phase, using γ-glycidyloxypropyltrimethoxysilane as a dispers-ing/coupling agent. A high MW polyamic acid (Pyre ML-RK692, DuPont) was usedto prepare the precursor solution in N-methylpyrrolidone for the coating of glassfibers, and a low MW polyamic acid (Skybond 703, Monsanto) was used for the pro-duction of the hybrid matrix of carbon fiber composites. The silica contents of thepolyimide–silica hybrids used were about 25 wt. % for the coating formulation andca. 8 or 16 wt. % for hybrids used as the matrix for carbon-fiber composites.

The scanning electron micrographs (SEMs) in Figure 23-4 [5] show the morpholo-gies of typical polyimide–silica hybrids based on Skybond 703 and Pyre ML-RK692,respectively. Both micrographs reveal the presence of co-continuous domains. Thehybrid based on the low MW polyimide (Skybond) displays fractal topography, whichis indicative of the existence of highly diffused nanostructured phases. The hybridbased on the high MW polyimide (Pyre ML), on the other hand, has a coarser andmore nodular structure, but its films are very transparent. Tunneling electron mi-crographs (TEMs) of the latter systems (see later) show that the dispersed phase con-sists of co-continuous domains of the two components, and is probably richer in sil-ica than the surrounding areas [13]. Not only the high molecular weight but also thehigher chain rigidity of Pyre ML, relative to Skybond, may be responsible for the for-mation of a coarser morphology.

Polyimides were selected for these systems as the manufacturing processes fortheir applications are solution-based, and the solvent used for the polyamic acid (pre-cursor for polyimides) is miscible with water. Both aspects are compatible with theprocesses used in sol-gel applications. Furthermore, polyimides represent a class of


40723.4 Potential Applications of Hybrids in Polymer Composites and Blends

Fig. 23-3 Polyamic acids for the formation of the polyimide, andalkoxysilane components for the in situ generation of the SiO2

phase.

408

polymeric materials with a very high service temperature rating, due to their excel-lent thermal oxidative stability and retention of properties up to very high tempera-tures. The incorporation of silica domains through the formation of hybrids repre-sents a natural approach for further extending the use of polymers into high tem-perature applications.


Fig. 23-4 Scanning electron micrographs of fractured surfaces ofpolyimide–silica hybrids (25 wt. % SiO2).

409

23.4.1.2 Coatings for Glass FibersIn research by the author’s group, commercially available glass fibers, for use inepoxy resin composites, were washed with hot xylene to remove the commercial“size” and subsequently recoated with the hybrid precursor solution. After curing,the coating thickness was in the region of 0.5 µm. Films were cast from similar solu-tions using different amounts of alkoxysilane precursor and their properties wereevaluated as a basis for selecting the most appropriate composition for the fiber coat-ings. The drying and curing schedule was the same as for the fiber coatings. Unidi-rectional composites were then prepared by winding and impregnating the fibers ona frame, which was then fitted in a “leaky mold” and pressed at the appropriate tem-perature to cure the bisphenol A/epoxy resin matrix with hexahydrophthalic anhy-dride. The glass concentration was 65 wt. %.

The force–displacement curves shown in Figure 23-5 [14] were obtained in f lexureat a 5:1 span-to-thickness ratio; such short spans are normally used to measure theinterlaminar shear strength of composites. The results suggest that the use of thepolyimide–silica coating maintains the rigidity and the interlaminar shear strength


Fig. 23-5 Comparison of force/deflectioncurves, recorded in interlaminar shearstrength tests, between composites made with

commercial glass fibers (as received) andfibers coated with the Pyre ML-RK692 (PI)–silica (25 wt. %) hybrid.

410

achieved with commercial sizes. The ductility associated with interlaminar shear fail-ures, however, is substantially increased over that obtained with conventional sizes,due to a change in failure mechanism from a rapid growth of a few large cracks in thecentral planes to slow growth of multiple fine cracks formed in several parallel planesin the middle regions of the specimen. The polyimide–silica hybrid at the interphasebetween the glass fibers and the epoxy resin matrix acts a “crumbling” zone, whicharrests the propagation of the cracks in the planes along the fibers.

The hitherto studied methods for increasing the interlaminar shear ductility havebeen based on coating fibers with rubbery polymers or very ductile thermoplasticsfrom solutions. In these cases, the enhancement in ductility occurs at the expense ofthe interlaminar shear rigidity and strength [15].

Figure 23-6 shows an example of a variation of interlaminar shear rigidity withtemperature, measured by dynamic mechanical tests at 1 Hz, using a cantilever bend-ing mode with loading distance equal to half the span for the corresponding inter-laminar shear tests in three-point bending [14]. A comparison of the two curves indi-cates that the interlaminar shear rigidity becomes increasingly higher than for con-ventional sized fiber systems upon traversing the glass transition temperature (Tg) ofthe polymeric matrix, due to the much higher Tg of the interphase material(280–300 °C) relative to that of the matrix (around 150 °C). The higher interlaminarshear rigidity above the Tg of the matrix suggests that the use of polyimide–silica hy-


Fig. 23-6 Effect of replacing a commercial sizing with a poly-imide–silica hybrid coating on the interlaminar shear rigidity of aglass fiber/epoxy composite as a function of temperature.

411

brids as coatings for fibers can provide better safety margins for the designers againstpossible excursions of the product to high temperatures.

23.4.1.3 Polyimide–Silica MatricesThe precursor solution for these polyimide–silica hybrids is prepared in the sameway as for coating formulations. The only differences are in the type and molecularweight of the polyamic acid (see Figure 23-3) and the amounts of alkoxysilane usedfor the formation of the silica domains. Films have been produced using the samecuring schedule as described for the coatings. Composites have been prepared as de-scribed previously using commercial carbon fibers.

Scanning electron micrographs of fractured surfaces reveal a fractal topography,which suggests that the two phases are not segregated but exhibit a graded supramol-ecular structure. The mechanical properties have been evaluated at both room tem-perature and at high temperatures, close to the Tg of the polyimide. Load–def lectioncurves obtained from f lexural tests (span-to-depth ratio = 20:1) are compared in Fig-ure 23-7 [16]. The curves show that for systems in which a polymeric matrix is re-placed by its corresponding silica hybrid the resulting increase in modulus at highertemperatures is even more dramatic.

Finally, the thermograms in Figure 23-8 [17] show that the presence of the co-con-tinuous silica domains within the polyimide matrix increases the thermal oxidativestability by shifting the weight loss curve to higher temperatures. This is associated


Fig. 23-7 Flexural load/deflection curves for carbon fiber compos-ites measured at a) room temperature, b) 150 °C, c) 280 °C [——— polyimide; – – – – – hybrid with 7.5% SiO2; – · – · – hybridwith 15% SiO2].

412

with the improved barrier properties provided by the silica phase, both for the infu-sion of oxygen and the outward diffusion of gases resulting from pyrolysis of the poly-mer chains. However, when large amounts of coupling agents (above the optimum)are used, the resulting increase in the amount of aliphatic matter present bringsabout a deterioration in thermal stability.


Fig. 23-7 (continued)

413

23.4.2Utilization of Hybrids in Thermosetting Resins

23.4.2.1 Epoxy–Silica Hybrids for Coating ApplicationsAlthough the oligomeric products resulting from hydrolysis and condensation reac-tions of TEOS can react with epoxy resins to produce compatibilizing species, the re-actions are difficult to control and can cause severe embrittlement in the cured prod-ucts. For this reason, some authors have used aliphatic resins and/or hardeners sothat the final cured products are rubbery in nature rather than glassy [11]. With aro-matic resins, such as those based on bisphenol A, cured with aromatic or cyclicaliphatic hardeners, it is difficult to prevent gross phase separation without intro-ducing an adequate level of alkoxysilane functionality in the epoxy resin [12].

Many commercially available silane coupling agents can be used for this purpose,particularly γ-mercaptopropyltrimethoxysilane (MPTS) and amino-bis(γ-propyl-trimethoxysilane), known as A-1170. Much less effective compatibilizing behaviorwas found when γ-glycidyloxytrimethoxysilane (GOTMS) was added to TEOS in thepreparation of the silica precursor solution. It was anticipated that the reaction of the


Fig. 23-8 TGA thermograms forpolyimide–silica (Skybond-based) containingvarious amounts of coupling agent. S = poly-imide; S/A(0) = polymide–silica hybrid (no

coupling agent) S/A(G0.12) = polymide–silicahybrid (0.12 molar ratio coupling agent/TEOS)S/A(G0.48) = polymide–silica hybrid(0.48 molar ratio coupling agent/TEOS)

414

epoxy groups in GOTMS with the epoxy resin hardener would produce compatibiliz-ing species for the epoxy resin and alkoxysilane component, but this was not foundto be the case.

In order to achieve a high level of miscibility between the two precursor compo-nents, a mixture of low MW (350) and high MW (5000) bisphenol A type resins in aweight ratio of 9:1 was used. Miscibility would be attributed to the presence of a largenumber of hydroxyl groups in the high MW component, which would strongly inter-act with the silanol groups of the alkoxysilane solution, possibly through a series ofequilibrium ester exchange reactions.

A comparison was made between the functionalization with MPTS and A-1170 ata level of 10 mol. % with respect to the epoxy groups present in the resin mixture. Thefunctionalized epoxy resin was dissolved in a butanol/xylene mixture, subsequentlymixed with a pre-hydrolyzed alkoxysilane solution containing 12 mol. % GOTMS,and then the p-dicyclohexyl aminomethane hardener was added. The solutions werecast in PTFE molds to produce thin plaques, about 0.75 mm thick, and cured first for24 h at room temperature and then for 1 h at 120 °C. In order to demonstrate themuch higher efficacy of co-continuous domains over particulate domains of similarsize, similar systems to the above were produced using dispersions of silica sols inisopropanol (particle diameter = 7–9 nm). The difference in morphology of the twosystems is illustrated by the tunneling electron micrographs in Figure 23-9 [18], fromwhich it is clear that the use of TEOS produces nanostructured co-continuous do-mains whereas the silica sol dispersion produces the expected particulate structure.

The linear expansion as a function of temperature and the solvent absorption be-havior in tetrahydrofuran (THF) for samples containing 15% silica prepared by dif-ferent methods are shown in Figures 23-10 and 23-11, respectively [18]. Not only dothese demonstrate the superiority of systems containing co-continuous domains, but


Fig. 23-9 Tunneling electron micrographs ofsystems containing 15 wt. % SiO2, both pro-duced from a bisphenol A epoxy resin graftedwith an amine alkoxysilane coupling agent: a)

epoxy–silica hybrid (co-continuous domains),b) epoxy–silica nanocomposite (particulate sil-ica domains).


Fig. 23-10 Comparison of linear expansion as a function of temper-ature of an epoxy resin control with an epoxy–silica nanocompositeand an epoxy–silica hybrid.

Fig. 23-11 Comparison of room temperature THF absorption as afunction of immersion time of an epoxy resin control with anepoxy–silica nanocomposite and an epoxy–silica hybrid.

416

they also illustrate the very high efficiency of the latter type of morphology in en-hancing both the mechanical and barrier properties of the base macromolecularcomponent of the hybrid, which is particularly useful in applications such as surfacecoatings.

23.4.2.2 Toughening of Epoxy Resins with Organic–Inorganic HybridsOver the last 30 years or so, there has been an enormous interest in the tougheningof thermosets (particularly epoxy resins) by the in situ generation of rubbery particles.This approach, however, brings about a deterioration of stiffness and strength-relat-ed properties. The rubbery particles appear to contain aggregated rigid inclusions,but their true role in the toughening mechanism is not generally known.

The author has found a means by which rubbery particles can be formed from aperf luoroether oligomer/alkoxysilane-based “hybrid precursor solution” dissolved inan epoxy resin/hardener mixture. The precipitation of the rubbery particles, whichtakes place during gelation of the resin, occurs with co-precipitation of the silicaphase component in the form of a rubber–silica hybrid. By tailoring the functional-ization of the oligomer it is possible to control the precipitation of particles in such away as to produce a ductile interlayer between the silica hybrid particles and the epoxymatrix. To achieve this type of morphology, a mixture of two perf luoroetheroligomers is used, one bearing multiple siloxane groups and the other bearing a larg-er number of carboxylic acid functional groups. A combination of acid and alkoxysi-lane functionalities is always present to ensure that reactions can always take placebetween the modifying oligomer and the two other components, i.e. the epoxy resinand silica.

The structure of the perf luoroether oligomer is shown in Figure 23-12. Function-alization was carried out in three consecutive steps: 1) reaction with chlorendic an-hydride in a 1:1 molar ratio to convert one terminal hydroxyl group into a carboxylicacid group; 2) reaction with glycidylpropyltrimethoxysilane in a 1:1 molar ratio to con-vert the carboxylic acid groups into trimethoxysilane functionalities; 3) reaction withchlorendic anhydride in a molar ratio of 1:1.6 to convert the residual hydroxyl groupsfrom step 1, and the additional groups formed in step 2, into carboxylic acid groups(nominally to a level of 80%). Although side reactions were identified from the out-lined reaction scheme, the main products contained the specified carboxylicacid/trimethoxysilane dual functionality. The choice of a perf luoroether oligomer forthe production of particulate hybrids within the epoxy resin [19] was made on the ba-sis of the outstanding thermal oxidative stability that they can confer upon curedepoxy resins, which allows the products to be used in high temperature applications.

An example of the scheme used for the introduction of dual functionality in theoligomer, in the form of carboxylic acid groups and trialkoxysilane groups, respec-tively, is shown in Figure 23-12 [20,21].

The resin used was a low MW bisphenol A type epoxy cured with an aromatic di-amine. To ensure that the silica produced by the sol-gel reaction resided preferential-ly within the precipitated particles, the TEOS solution, containing an appropriateamount of coupling agent, was first hydrolyzed using p-toluenesulfonic acid as cata-lyst and then mixed with the perf luoroether oligomer containing a high level of silox-


417

ane functionalities. After mixing with the epoxy resin, a perf luoroether oligomer con-taining a predominance of carboxylic acid groups was added, followed by the hard-ener. Upon curing, the products of the reaction of the latter oligomer with the epoxyresin formed the interphase layer through a sequential phase-separation process, inwhich the primary particles were formed by the faster gelling perf luoroether-silox-ane reaction products.

An interesting aspect of these systems is that the silanol groups from the pre-hy-drolyzed TEOS can also react with the epoxy resin, particularly through condensationreactions with the hydroxyl groups formed as a result of the opening of the oxiranering. This will provide additional cross-linking within the epoxy matrix, which wouldnot otherwise be available when using amine hardeners. When carefully controlledto avoid excessive embrittlement, these additional reactions can increase both the Tg

and the strength of the matrix.The tunneling electron micrographs in Figure 23-13 [21] show a typical structure of

a toughened epoxy resin. The precipitated particles display the nanostructured mor-phology of organic–inorganic hybrids. A diffuse interphase region between the par-ticles and the matrix is also discernible in these micrographs.

The mechanical properties, namely Young’s modulus, f lexural strength, and frac-ture toughness (critical strain energy release rate, Gc), all as a function of the total per-f luoroether content with a fixed amount of silica with respect to the perf luoroethercomponent, equivalent to 25%, are shown in Figure 23-14 [21]. The notable features


Fig. 23-12 Structure of perf luoroether oligomers used as the start-ing materials for hybrid rubber–silica toughening of bisphenol-Aepoxy resin, where p/q = 0.7, z = 1.5, and n = 10 (approximately).

Fig. 23-13 Tunneling electron micrographs ofepoxy–perf luoroether–silica hybrids showingthe dispersed phase (low magnification – left

image) and the interfacial region (high magni-fication – right image); primary particles arebetween 5 and 10 µm in diameter.

418

of the data are the increase in toughness with increasing particulate perf luo-roether–silica hybrid content, which is accompanied by concomitant increases inYoung’s modulus and f lexural strength. The increase in modulus seen in our systemrepresents a unique achievement in the toughening of epoxy resins, which is not at-tained by the use of conventional functionalized oligomers, known as “liquid rub-bers”, typified by carboxyl-terminated butadiene–acrylonitrile (CTBN) and amine-terminated butadiene–acrylonitrile (ATBN) oligomers.

Acknowledgements

The author is grateful to the many colleagues who have contributed to the contentsof this chapter, in particular Drs. H. Demirer, C. Xenopoulos, S.-Y. Ng, and L. Prezzi.


Fig. 23-14 Effects of perf luoroether content on mechanical proper-ties of epoxy–perf luoroether–silica hybrids (f lexural modulus, f lexur-al strength, and critical strain energy release rate, Gc)

419

References

1 Yan, K., et al. J. Polym. Sci., Part A: Polym.Chem. 1993, 31, 2493.

2 Brinker, C. J., Sol-Gel Science: The Physicsand Chemistry of Sol-Gel Processing, Chapter3, London Academic Press, London, U.K.,1990.

3 Nass, R., et al., J. Non-Crystalline Solids1990, 121, 370.

4 Mascia, L., Trends in Polymer Science 1995, 3,61.

5 Mascia, L., La Chimica e L’ Industria 1998,80, 623.

6 Clarson, J. S., Mark, J. E., Polym. Commun.1989, 30, 275.

7 Wang, B., et al., Macromolecules 1991, 24,3449.

8 Wen, J., Mark, J. E., J. Appl. Polym. Sci.1991, 85, 1135.

9 Brennan, A. B., et al., J. Inorg. andOrganomet. Polym. 1991, 1, 167.

10 Mark, J. E., et al., Macromol. Symp. 1995, 98,731.

11 Matejka, L., et al., J. Non-Crystalline Solids1998, 226, 114.

12 Mascia, L., Tang, T., Polymer 1998, 39, 3045.13 Mascia, L., Kioul, A., J. Crystalline Solids

1994, 175, 169.14 Demirer, H., Ph.D. Thesis, Loughborough

University, U.K., 1999.15 Mascia, L., et al., Plastics, Rubber and Com-

posites: Processing and Applications 1993, 19,221.

16 Mascia, L., et al., Composites, Part A 1996,27A, 1211.

17 Xenopoulos, C., et al., High PerformancePolymers 2001, 13, 1.

18 Prezzi, L., Ph.D. Thesis, LoughboroughUniversity, U.K., 2004.

19 Mascia, L., et al., J. Appl. Polym. Sci. 1994,51, 905.

20 Mascia, L., Heath R. J., Ng, S.-Y., unpub-lished data.

21 Ng, S.-Y., Ph.D. Thesis, Loughborough Uni-versity, U.K., 2002.


421

Index

aacetylene black 319acrylic acid 122

– functionalized polyolefin 122acrylic acid modified polypropylene 218additive 6, 8 f, 105

– continuous 6, 8– coupling 105– discontinuous 6, 8 f

adhesion 18adsorption 18

– interfacial force 18– molecular orientation 18– polymer mobility 18

af-glass 136agglomerate formation 45 ffa-glass 136, 381

– chemical composition 136– property 136

alkyl borate 121alkyl organophosphate 121alkyl sulfonate 121alumina 389aluminium trihydrate 289, 293 f

– thermal degradation 294aminosilane 75 f, 78 ff

– ATH 79 f– calcined clay 78– clay 75– reaction 76– talc 75

ammonium pentaborate 307ammonium polyphosphate 290, 304anti-blocking agent 339anti-blocking filler 356 ff

– type 356anti-fogging agent 339antimony oxide 290antimony trioxide 309

– toxicity 309

anti-static agent 339apatite 283aragonite 394aramid 340 aramid fiber 136ar-glass 136asbestos 10, 208, 214aspect ratio 9 f, 28, 225, 241, 245, 271, 277, 381ATH 76 f, 79 f, 96, 113 ff, 120, 123, 289, 294, 296, 298,

– application 296– cross-linked LDPE filled 96– filled EVA 76 f, 79 f, 113 ff, 120– filled HDPE 116– filled PP 123– morphology 294– PP filled 96– surface modified 116– surface treatment 298

ATH filled EPDM 295

bbanbury mixer 40barium ferrite 100

– titanate/zirconate effect 100barium sulfate 10BaSO4 95, 100

– filled 95– polybutadiene 95– titanate/zirconate effect 100

bast fiber 196 fbentonite 165, 168bioacitve compsite 283bioactive filler 283, 387 ffbioactive glass 389, 395, 397

– composition 395– silver-doped 397

bioactive glass ceramic 389bioactivity 283, 387 ffbioceramic 387 ffbiocompatibility 387 ff


422

biodegradable polymer 389bioglass® 395biomaterial 387 ffbiomedical 387 ffbiostable polymer 389biotite 151bismaleimide 118 ffBN 345, 352 f, 355

– application 355– cost/availability 353– manufacturer 352– property 345– supplier 352

bone 6, 283, 388 ff– component 388– compressive strength 388– regeneration 389– strength 388– structur 388– tensile modulus 388

bone growth 387 ffboron nitride (BN) 340 ff, 343, 349, 355

– environmental consideration toxicity 355– graphite 355– polytetraf luoroethylene 355– production 343– structure property 349

boron-containing fire retardant 311– toxicity 311

brominated fire retardant 309– toxicity 309

bromine-containing fire retardant 300bulk molding compound 277bulk resistivity 335

cCaCO3 96 f, 100

– filled HDPE 97– filled PP 97– LLDPE filled 96– titanate/zirconate effect 100

CaCO3 coated 278CaCO3 uncoated 278CaCO3-filled PP 281cadmium 100

– titanate/zirconate effect 100calcination 227calcined kaolin 227 ff

– benefit 230calcite 271, 394calcium carbonate 10, 53, 89, 109 ff, 113, 117, 122 ff,

271 ff, 389– application 275– cost/availability 274

– coupling 89– dispersion 89– dry grinding 272– environmental consideration 275– fatty acid treatment 111– filled EPDM 117– filled polypropylene 109 f, 122 ff– filled unsaturated polyester 113– primary function 271– production method 272– secondary function 271– squalane slurry 111– supplier 274– surface treatment 271– toxicity consideration 275– wet grinding 272– world production 274

calcium carbonate filled polyvinyl chloride 276calcium carbonate glass fiber reinforced thermoset

277calcium carbonate molding polyolefin 277calcium carbonate occur 273

– property 273– structure 273

calcium metasilicate 241calcium nitrate 397calcium phosphate 388 fcarbon black 317 ff, 319, 321, 326 ff

– agglomerate 317– aggregate 317– application 326– DBPA absorption 319– dispersion 326 f– distribution 326– microscopy 326– morphology 326– multiphase blend 327– source 321– structure 317– surface area 319– surface chemistry 321– production 319– safety 321– toxicity 321– volatile 319

carbon fiber 136, 175, 178, 191, 340– function 191– property 136– synthesis 178

carbon fiber composite 189– application 189

carbon fibers compound 329– fiber length 329– resistivity 329

Index

423

carbon nanofiber 175– synthesis 175– type 175

carbon nanotube 175 ff, 181, 188, 191, 328– chemical modification 181– cost/availability 188– derivatization 181– environmental consideration 188– function 191– synthesis 175– toxicity consideration 188

carbon nanotube composite 183, 189– application 189– manufacture 183

carbon nanotube-polymer nanocomposite 184carboxylic acid 107carboxylic acid anhydride 117CaSO4 100

– titanate/zirconate effect 100cavity transfer mixer 48cellulose 197, 254cellulose fiber 178 ffceramer 401ceramic sphere 381 ff

– function 382– hollow 381– production 381– property 381– solid 381– supplier 384

ceravitalTM 397c-glass 136


chalk 271channel black 317chemical vapor deposition (CVD) 175 ffchitosan 394chlorinated fire retardant 310

– toxicity 310chlorinated paraffin 119chlorine-containing f lame retardant 301chopped glass strand 138chrome complexe 120clay 75, 78 f, 222

– filler EPDM 75– filler polyamide 78

coal tar pitch 178 ffcoefficient of friction 340coefficient of thermal expansion 31

– polymer composite 31collagen 387 fcombustion 285combustion cycle 285

composite 3 ff, 11, 15, 387– cost 11– density 11– nanoclay 15– natrual fiber 15

conducting composite 332conductive filler 317conductivity 322, 324

– carbon black filled polymer 322– critical volume fraction 322– effects of carbon black type 324– effects of polymer matrix 324

cone calorimetric data 295cone calorimetry 307continuous filler 20

– modulus 20– strength 20

continuous form 138– discontinuous form 138

continuous reinforcement 22, 27coupling agent 87, 339creep 31

– polymer composite 31cristobalite 81

– filled PMMA 81critical aspect ratio 28

ddegree of fill 52delamination 225d-glass 136


dicalcium phosphate 390discontinuous reinforcement 23, 28discontinuous 20


dispersive mixing 48dissolving pulp 204distribute mixing 48double-arm sigma blade mixer 39drag f low 52

eefoliation 401e-glass 133 ff, 381


electrical property 236environmental consideration toxicity 355

– boron nitride 355– graphite 355– polytetraf luoroethylene 355

Index

424

environmental 140epoxy resin 416epoxysilane 81

– quartz f lour 81epoxy-silica hybrid 413extruder 40 ff

– single-screw 40 ff– twin-screw 40 ff

extrusion compounding 390extrusion free-form fabrication 183

ffatigue 31

– polymer composite 31fatty acid 111

– thermal stability 111fatty acid application 108

– dry coating 108– wet coating 108

Fe2O3 100– titanate/zirconate effect 100

Fe2O4 100– titanate/zirconate effect 100

fiber 8fiber extraction 196fiber glass 95

– filled PPS 95filler 8, 13 ff, 20, 35 f, 42

– aspect ratio 13– demand 14– dispersion 35– distribution 35– feeding 42– hardness 35– high aspcet ratio 13– low aspect ratio 13– market 14– moisture absorption 35– morphology 13– new application 15– orientation 35 f– pretreatment 42– shape 13– thermal property 35– thermal stability 35

filler family 12– carbon 12– graphite 12– hydroxide 12– metal 12– oxide 12– polymer 12– salt 12– silicate 12

– syntetic polymer 12filler function 13 f

– bioactivity 13– control of damping 14– control of permeability 13– degradability 14– enhancement of fire retardancy 13– enhancement of processability 14– improved dimensional stability 14– mechanical property 13– modification of electrical and magnetic

property 14– modification of optical property 14– modification of surface property 14– radiation absorption 14

filler silane treatment 66 ff– dry concentrate 68– dry procedure 66– in situ treatment 67– slurry procedure 67

filler/reinforcement 8– function 8

fire retardant 285 ff, 289, 292, 312– manufacture 312

fire retardant action 290f lake 20 f

– modulus 21– platelet 21– strength 20 f

f lake-reinforced composite 150f lame retardance 292 ff lame retardant 308

– toxicity 308f lammability of polymer 289f lax 197

– chemical composition 197FT-IR spectra 110fullerene 181fumed silica 360, 368 ff, 373, 375, 378

– application 378– environmental consideration 375– manufacturer 373– production 369– property 360– structure property 371– supplier 373– toxicological consideration 375

functional filler 12– classification 12– type 12

functional polymer 105functionalized polybutadiene 125

– maleic anhydride 125– trialkoxysilane 125

Index

425

functionalized polymer 121, 124– bonding mechanism 124– use 124

functionalized polyolefin 122furnace black 317

ggas permeability 210glass fiber 10, 123, 131 ff, 135, 138 ff, 159, 278, 409

– application 141– chemical composition 131– coating 409– cost/availability 138– environmental consideration 140– filled polypropylene 123, 159– production method 132– property 135– roving 132– silane coupling agent 133– sizing 132– strand 132– structure 135– supplier 138 f– surface free energy 132– toxicity consideration 140

glass fiber reinforced thermoplastic 141, 144 ff– adhesion 145 f– aspect ration effect 145 f– charpy impact strength 141– f lexural property 141– maleation 144 f– matrix 144 f– mechanical property 141– tensile moduli 141

glass sphere 381 ff– function 382– hollow 381– production 381– property 381 f– solid 381– supplier 384

graphite 340 ff, 345 f, 351, 354 f– amorphous 342– application 355– boron nitride 355– cost/availability 354– crystalline vein 342 ff– environmental consideration toxicity 355– f lake 342– manufacturer 351– polytetraf luoroethylene 355– production 342– property 345– structure property 346

– supplier 351ground ATH 295ground calcium carbonate (GCC) 271 ff

hhalogenated fire retardant 299 f

– mechanism of action 299– synergy with antimony trioxide 300

halpin and tsai equation 24hapexTM 391, 395hardwood 197, 253 f

– chemical composition 197heat distortion temperature 31

– polymer composite 31hemicellulose 198, 254hybrid 398, 401 ff

– production 403hybrid material 404

– property 404hydrocyapatite 388hydrostatic 390hydrotalcite 368 ff, 373, 375, 378

– application 378– environmental consideration 375– manufacturer 373– production 369– structure property 372– supplier 373– toxicological consideration 375

hydrous kaolin 230– benefit 230

hydroxyapatite 100, 390– titanate/zirconate effect 100

iimpact strength 31

– fibrous filler 31– particulate filler 31

implant 396in situ generated filler 397, 401incense cedar 255inorganic fiber 20 f

– density 21– modulus 20 f– strength 20 f

interfacial tension 18intrinsically conductive polymer 329intumescent f lake graphite 347

jjute 197

– chemical composition 197

Index

426

kkaolin 221 ff, 228, 230 ff, 236

– application 232– benefication 222 f– booklet 221– brightness 225– chemical composition 228– color 225– cost/availability 231– environmental consideration 231– mechanical property improvement 232– open-pit mining 222– primary function 232– primary processing 223– production method 222 ff– property 228– secondary function 236– structure 222– supplier 230– surface treatment 228– toxicity consideration 231

kaolin deposit 222kaolin filled HDPE 235kaolin filled polypropylene 235kaolin product 226kaolin-filled nylon-6,6 233 ffkaolinite 221 ffkenaf 196 f, 204

– chemical composition 197kneading paddle 48, 50

llaw of mixture 405lignin 197, 254limestone 271limitin oxygen index 289liquid rubber 418loblolly pine 255low melting temperature glass 306lubricant 339 ff

mmaddock 47magnesium hydroxide 110, 116, 125, 289, 294, 299

– filled EVA 110, 125– filled polypropylene 116, 119– thermal degradation 294

magnesium nitrate 397magnesium oxide 368 ff, 375 f

– application 376– burned 370– environmental consideration 375– production 368– property 370

– structure 370– thickening mechanism 376– toxicological consideration 375

magnetic filler 317, 333maleated polyethylene 278 ffmaleated polypropylene 203, 246maleated PP 172

– dispersion 172– exfoliation 172

maleic anhydride 117, 122, 125– functionalized polyolefin 122

maple 251marble 271MDH 79

– filled EVA 79mechanical property 17melamine 290, 303melt rheology 32 f

– effect of concentration and shear rate 32– effect of fiber 32– effect of filler size and shape 33– effect of filler surface treatment 33– effect of f lake 32– effect of particulate 32

metakaolin 227metal hydroxide 292, 308metal particle composite 330metallic wire 21

– density 21– modulus 21– strength 21

methacryloxy 75, 81– ATH 81– clay 75– cristobalite 81

methacryloxysilane 80– quartz f lour 80

Mg(OH)2 96– ABS filled 96– EPDM 96– polyamide 96– PP filled 96

MgO 373– manufacturer 373– supplier 373

mica 6 f, 10, 149 ff, 156 ff, 246, 278– adhesion 153– application 158– barrier property 160– composite 7– cost/availability 157– deposit 151– dry ground 149, 151, 156 f– environmental consideration 157

Index

427

– filled PP 158– f lake orientation 150– function 150 f– grade 157– high aspect ratio 150– morphology 152– mucovite 151 f– optical property 160– other function 158– planar isotropy 152– price 157– primary function 158– production method 151– production 156– property 152– structure 152– supplier 156– toxicity consideration 157– warpage control 160– weld-line strength 158– wet ground 149, 152, 156 f

mica filled thermoset 158– application 158– mechanical property 158– processing 158

mica f lake 149 ffmica reinforced thermoplastic 151, 154

– adhesion promoter 154mica/glass fiber PET 160mica-filled thermoplastic 158

– processing 158– mechanical property 158– use 158

milled glass fiber 138, 246mixing 39mixing element 47mixing enhancer 48 f

– cavity transfer 49– dulmage 49– pin 49– pineapple 49– saxton 49

modification of mechanical property 19modifier 85, 105 ff, 113, 117, 126

– carboxylic acid 107– carboxylic acid anhydride 117– coupling 106– fatty acid 107– fatty acid salt 107– non-coupling 106– organic amine 126– titanate 85– unsaturated carboxylic acid 113

modify 59, 118 ff, 125

– alkyl borate 121– alkyl organophosphate 121– alkyl sulfonate 121– bismaleimide 118– chlorinated paraffin 119– chrome complexe 120– functionalized polybutadiene 125– functionalized polymer 121– functionalized polyolefin 122– silane 59

modulus equation 23 f– continuous fiber composite 23– continuous ribbon composite 23– f lake 24– f lake composite 24– particulate composite 24– short fiber 24

MOH 289mollusk 7

– shell 7molybdenite 340 ff, 344, 354

– environmental consideration 354– production 342– property 344– structure 344– toxicity consideration 354

molysulfide (MoS2) 340monomeric silane 60

– production 60– structure 60

montmorillonite 15, 164, 166, 168, 401– characteristic 168– ion-exchange 168– purification 166– structure 168– surface treatment 166

mooney equation 34MoS2 345, 351, 353, 355

– application 355– cost/availability 353– manufacturer 351– property 345– supplier 351

multiple-wall carbon nanotube (MWNT) 175 ffmuscovite 149 ff, 152, 155

– chemical analysis 152, 155– property 155

nnanoclay 95, 163 ff, 167, 170 f, 173 f

– application 171, 173– barrier application 174– concept 163– cost/availability 170

Index

428

– dispersion mechanism 171– environmental consideration 170– exfoliation 164, 171, 173– filled nylon-6 164– filled SBS 95– filled TPO 171– f lame retardancy application 174– intercalation 164, 171– mechanical property 173– production method 164– property 167– structure 167– supplier 170– surface area 163– swelling 167– synthetic 167– thermal applicaton 173– toxicity consideration 170

nanocomposite 15, 54, 163 ff, 175, 184, 401– compounding 54– melt processing 15– resistivity 184– technology 163

nanomaterial 16– safety 16

nanoporous membrane 186nanotalc 212natrual fiber 21, 195 ff, 198 ff, 203

– agent 203– application 202– consumption 195 f– cost/availability 201– degradation 200– density 21, 198– dimension 198– durability 199– environmental benefit 202 f– equilibrium moisture content 200– mechanical performance 198– modulus 21– moisture 199– polypropylene composite 203– primary function 202– production 196– secondary function 202– strength 21– structure 196– supplier 200

natural fiber composite 195 ff– growth 195– f lax 195

natrual fiber reinforced plastic 203– mechanical property modification 203

nepheline syenite 115

– filled HDPE 115Nielsen equation 34

ooak 251

– bulk density 53– loading 53

organic amine 126organic fiber 21


ormosils 402, 404osseointegration 398osteoconductive material 390

pparticulate 20, 25, 30


pearlescent pigment 10pectin 197percolation network 327percolation theory 323perf luoroether oligomer 416 ffpetroleum 178phenolic fiber 178 ffphlogopite 149 ff, 152, 155

– chemical analysis 152, 155– property 152, 155

phosphorus-containing f lame retardant 303, 311– mechansim of action 303– toxicity 311

photoluminescence 188pine 251pine wood f lour 252PMMA nanocomposite 186polyacrylonitrile (PAN) 178 ffpolyester gel coat 238polyethylene imine 182polyhydroxybutyrate 393polyimide-silica hybrid 406polylactic acid 283, 387, 394, 407polymer 21


polymer combustion 292– component 292

polymer composite 6– component 6– type 6

polyolefin film 280polyolefin microporous film 280

Index

429

– calcium carbonate 280polysaccaride 387polysulfone 392polytetraf luoroethylene (PTFE) 340 ff, 343, 348, 355

– boron nitride 355– environmental consideration toxicity 355– graphite 355– mechanical property 348– production 343– surface property 348– thermal property 348

polyvinyl pyrrolidone 182ponderosa pine 255porosity 282precipitated ATH 295precipitated calcium carbonate (PCC) 271 ffprecipitated silica 360

– property 360processing 4

– post-shaping 4– pre-shaping 4– shaping 4– thermoplastic 4– thermoset 4

processing aids 367 ffprocessing method 5

– calendering 5– casting 5– closed-mold reinforced plastic 5– compression molding 5– extrusion 5– foam molding 5– injection molding 5– open-mold reinforced plastic 5– polyurethane foam molding 5– resin injection molding 5– rotational molding 5– thermoforming 5– transfer molding 5

production method 151PTFE 345, 349, 353, 355

– application 355– cost/availability 353– electrical property 349– envrionmental resistance 349– manufacturer 353– property 345– supplier 353

pulp 204

qquaternary ammonium chloride 168

rR7 talc 212red maple 255reinforcement 8

– continous 8– discontinuous 8

reinforcing filler 9– compatibility 9– function 9– interfacial adhesion 9

repolymerization 93 f– ABS 94– neoalkoxy tianate 94– PC 94– PP 94– PS 94

resistivity measurement 334r-glass 133 ffrheological property 17rheology modifier 367ribbon 8, 20 f


rule of mixture 11

ssaponite 100

– titanate/zirconate effect 100scorim® 390 fscrew kneader 55s-glass 136


sheet molding compound 277shielding effectiveness 332silane 62 f, 74, 81, 246

– hydrolysis 63– oligomeric 62– selection 74– waterborn 62

silane coupling agent 59 ff, 413– benefit 59– fomulae 60– structure 60– world market 59

silane reactivity 65silanized surface 66silanol 398silica 358 ff

– amorphous 358– application 360– crystalline 358– environmental consideration 360

Index

430

– production 358– supplier/manufacturer 360– toxicity consideration 360

silica gel 359silylated filler analysis 68, 70 ff

– acid-base titration 73– auger electron spectroscopy 70– carbon analysis 72– colorimetric test 73– FT-IR/Raman Spectroscopy 68– hydrophobicity 73– MAS-NMR Spectroscopy 68– pyrolysis-gas chromatography 72

simulated body f luid 283, 389single-wall carbon nanotube (SWNTs) 175 ffsisal 197

– chemical composition 197sizing composition 133 ff, 135

– thermoplastic 133– thermoset 133

slip agent 339smectite 164, 167 f

– structure 167 fsmoke 287 ff

– toxicity 288– visibility 288

softwood 197, 253– chemical composition 197

sol-gel 177, 397sol-gel process 402, 406solid glass sphere 10solid surface 296solution casting 390southern red oak 255starch 391static charge decay 333stearic acid 109strength equation 27, 29

– continuous fiber composite 27– continuous ribbon composite 27– f lake composite 29– particulate composite 29– short fiber composite 29

stress relaxation 31– polymer composite 31

sulfide 100– titanate/zirconate effect 100

sulfur-containing silane 81surface modification 19surface property modifier 339surface resistivity 335surface treatment 277surface-modified wollastonite 246surface-treated ATH 296

ttalc 10, 53 f, 75, 95, 100, 207 ff, 214 ff, 221, 246, 278

– application 215– bulk density 53 f– characteristic 210– composition 208, 210– consumption 212– cost/availability 212– environmental consideration 214– filled PP 95– filler EPDM 75– grinding 208– high purity 208– loading 53– miming 207– particle shape 210– particle size 210– primary function 216– production 212– production method 207– property 208– secondary function 216– structure 208– surface chemistry 211– surface-treated 208– supplier 212– titanate/zirconate effect 100– toxicity consideration 214

talc-filled composite 210talc-filled masterbatch 215talc-filled PP 217 fftalc-filled PVC 219talc filled styrenic 219talc filled thermoplastic elastomer 219tetracalcium phosphate 390tetraethoxysilane 402tetraethyl orthosilicat 397thermoplastic 3 ff

– amorphous 3– crystalline 3– glass-transition temperature 3– melting temperature 3

thermoset 3 ffTiO2 100

– titanate/zirconate effect 100TiO2/kaolin combination 237 fTiO2-extension 237tissue engineering 389titanate 87, 246

– form designation 87titanate application 92titanate coupling agent 85 ff, 88

– chemical description 86– function 85

Index

431

– function 88titanate function 88 f, 91 f

– adhesion 91– coupling 89– dispersion 89– hydrophobicity 82

titanate function (2) 93, 95– example of system 95

titanate function (3) 95 f– example of system 96

titanate function (4) 97 ff– CaCO3-filled thermoplastic 97– carbon black (CB) filled polymer 98– example 99

titanate function (5) 99titanate function (6) 101titanate/zirconate 100

– effect 100titanium carbide 100

– titanate/zirconate effect 100tortuosity 160

– damping application 160toughness 30 f

– f lake composite 31– short-fiber composite 30

tremolite 208, 214trialkoxysilane 61

– hydrolysis 61tribological additive 340β-tricalcium phosphate 390tricalcium phosphate 389triethyl phosphat 397trusion 390tungsten carbide 100

– titanate/zirconate effect 100

uUL 94 standard 307unsaturated acid 113US standard mesh size 251

vvapor-grown carbon fiber 180vaterite 283, 394venting 52vinyl silane 75 ff

– aluminium trihydrate (ATH) 76– application 76– ATH 77– clay 75– talc 75

viscosity 33 f, 52– calcium carbonate filled polystyrene 34– polystyrene/glassfiber 33

wwater vapor transmission rate 282weldline 36wetting 18

– contact angle 18– critical surface tension 18– surface tension 18

wetting agent 339whisker 21


white oak 255wollastonite 10, 123, 241 ff, 390

– application 245– chemical analysis 242– environmental consideration 244– filled polypropylene 123– formation 241– polyamide-6 and 6,6 reinforced 245– polybutylene terephthalate reinforced 245– polypropylene reinforced 245– production 241– property 242 f– structure 242– supplier/cost 244– surface-treated 242– toxicity consideration 244

wollastonite/glass fiber reinforced polyamide-6,6 246

wollastonite-filled thermoset 246wood 6, 196, 254 ff

– chemical component 254– degradation 259– durability 259– equilibrium moisture content 257– extractive 255– specific heat 262– thermal conductivity 262– thermal diffusivity 262– thermal expansion 261

wood anatomy 253wood fiber 265

– injection-molded composite 265– maleated polypropylene 265– polypropylene 265

wood f lour 123, 204, 249 ff, 253, 256 ff, 264 ff– application 250, 263– biodegradability 267– cost/availability 263– density 256– environmental consideration 263– environmental preference 267– injection-molded composite 265

Index

432

– market size 250– moisture sorption 258– moisture 257– particle size 265– polypropylene 265– primary function 263– production method 251– property 253– secondary function 263– size classification 251– size reduction 251– structure 253– supplier 262– thermal property 260– thermoplastic 249– thermoset 249– toxicity consideration 263

wood f lour thermoplastic 264

– mechanical property modification 264wood f lour thermoset 263wood-filled polypropylene 265wood-like property 265wood-plastic composite 249, 258, 260

– weight loss 260work of adhesion 18

xxerogel 401

zzinc borate 290, 303zinc phosphate glass 307zirconate coupling agent 87

– form designation 87zirconia 389

Index

functional fillers for plastics

Documents