Sei sulla pagina 1di 560

Particulate-Filled

Particulate-Filled

Polymer

Polymer

Composites

Composites

Second

Edition

Editor R.N. Rothon
Editor
R.N. Rothon

Rapra Technology Limited

Particulate-Filled Polymer Composites

2nd Edition

Editor: Roger N. Rothon

rapra

TECHNOLOGY

Rapra Technology Limited

Shawbury, Shrewsbury, Shropshire, SY4 4NR, United Kingdom Telephone: +44 (0)1939 250383 Fax: +44 (0)1939 251118 http://www.rapra.net

First Published in 2003 by

Rapra Technology Limited

Shawbury, Shrewsbury, Shropshire, SY4 4NR, UK

©2003, Rapra Technology Limited

All rights reserved. Except as permitted under current legislation no part of this publication may be photocopied, reproduced or distributed in any form or by any means or stored in a database or retrieval system, without the prior permission from the copyright holder.

A catalogue record for this book is available from the British Library.

Every effort has been made to contact copyright holders of any material reproduced within the text and the authors and publishers apologize if any have been overlooked.

Cover micrograph of aramid fibre-reinforced polyamide 6,6, reproduced with permission from P.R. Hornsby, Brunel University, UK.

ISBN: 1-85957-382-7

Typeset by Rapra Technology Limited Cover printed by The Printing House, Crewe, Cheshire, UK Printed and bound by Rapra Technology Limited, Shrewsbury, UK

Contents

Preface

 

1

Contributors

3

1

General Principles Guiding Selection and Use of Particulate Materials

5

1.1 Introduction

5

1.2 Basic Characteristics of Particulate Fillers

5

 

1.2.1 Cost

6

1.2.2 Chemistry, Composition and Impurities

7

1.2.3 Density or Specific Gravity

9

1.2.4 Hardness

10

1.2.5 Abrasiveness

11

1.2.6 Optical Properties

11

1.2.7 Thermal Properties

14

1.2.8 Particle Shape and Size

16

1.2.9 Shape

17

1.2.10 Particle Size

20

 

1.3 Surface Modification

26

 

1.3.1 Stearic Acid and Stearates

26

1.3.2 Coupling Agents

27

1.3.3 Polymer Modifications

27

1.3.4 Direct Bonding

28

 

1.4

Particle Packing and the Maximum Packing Fraction

29

 

1.4.1 Introduction

29

1.4.2 Determination of Maximum Packing Fraction (P f ) by Oil Absorption Procedures

29

1.4.3 Particle Packing Theory

30

Particulate-Filled Polymer Composites

 

1.4.4

Applications of Packing Principles to Particulate Filled Composites

33

1.5 Interparticle Spacing

34

1.6 Particle Effects on the Structure of Polymers

36

1.6.1

Introduction

36

1.6.2

Molecular Weight Reduction During Processing

36

1.6.3

Molecular Weight and Crosslinking Changes due to Cure Modifications

37

1.6.4

Preferential Adsorption of Polar Species

37

1.6.5

Formation of an Interphase of Immobilised Polymer

38

1.6.6

Effects on Polymer Conformation due to the Presence of Particle Surfaces and Interparticle Spacing

42

1.6.7

Effects on Crystallinity

42

References

45

2

Principal Types of Particulate Fillers

53

2.1 Introduction

53

2.2 Particulate Fillers from Natural Origins (Mineral Fillers)

53

2.2.1 Introduction

53

2.2.2 Minerals and Rocks

54

2.2.3 Rocks

55

2.2.4 Calcium Carbonate Minerals

57

2.2.5 Dolomite

61

2.2.6 China Clay or Kaolin

61

2.2.7 Calcined Clay

66

2.2.8 Mica

69

2.2.9 Talc

70

2.2.10 Montmorillonite (AlMg) 8 (Si 4 O 10 ) 3- (OH) 10 .12H 2 O

72

2.2.11 Barite (BaSO 4 )

73

2.2.12 Calcium Sulfate Products

74

2.2.13 Wollastonite (CaSiO 3 )

74

2.2.14 Crystalline Silicas

76

Contents

 

2.3

Synthetic Particulate Fillers

78

 

2.3.1 Carbon Black

78

2.3.2 Synthetic Silicas

81

2.3.3 Hydroxides and Basic Carbonates

84

2.3.4 Precipitated Calcium Carbonate (PCC)

96

 

Acknowledgements

96

References

97

3

Analytical Techniques for Characterising Filler Surfaces

101

3.1 Introduction

101

3.2 Acid-Base Theory

104

 

3.2.1 Introduction

104

3.2.2 Gutmann Approach

105

3.2.3 Drago Approach

106

3.2.4 Use in Characterising Fillers

106

 

3.3 Analytical Techniques

108

3.4 Reactive Techniques

109

 

3.4.1 Flow Microcalorimetry

109

3.4.2 Inverse Gas Chromatography

119

 

3.5

Spectroscopic Techniques

124

 

3.5.1 Introduction

124

3.5.2 X-Ray Photoelectron Spectroscopy

124

3.5.3 Secondary Ion Mass Spectrometry

130

3.5.4 Diffuse Reflectance Fourier Transform Infrared Spectroscopy . 134

3.6

Methods for Examining Structural Order in Filler Coatings

145

3.6.1 Wide Angle X-Ray Diffraction

145

3.6.2 Differential Scanning Calorimetry

146

3.7

Summary

147

References

148

Particulate-Filled Polymer Composites

4

Surface Modification and Surface Modifiers

153

4.1 Introduction

153

4.2 Reasons for Using Surface Modifiers

153

4.3 General Principles of Surface Modification

154

4.4 Methods of Using Surface Modifiers

155

4.5 Choice of Coating Level

156

4.6 Techniques for Determining the Amount of Coating Present, and Assessing the Amount Needed for Mono-layer Coverage

158

 

4.6.1 Determination of Amount of Additive and it’s Distribution

158

4.6.2 The Monolayer and it’s Determination

160

4.6.3 Effects of Processing on the Coating Structure

163

 

4.7

Surface Modifier Types

163

 

4.7.1 Monomeric Organic Acids and their Salts

163

4.7.2 Stearic Acid (CH 3 (CH 2 ) 16 COOH

166

4.7.3 Other Saturated Fatty Acids and Related Substances

170

4.7.4 Effects of Stearic Acid Coating in Composites

171

4.7.5 Fatty Acid Salts

173

4.7.6 Unsaturated and other Functional Organic Acids in Composites

173

4.7.7 Polymeric Acids and Anhydrides

175

4.7.8 Organo-silicon Compounds

177

4.7.9 Examples of Silane Coupling Agent Effects in Filled Polymers

190

4.7.10 Organo-Titanates and Zirconates

191

4.7.11 Aluminates and Zircoaluminates

198

4.7.12 Phosphates and Borates

199

4.7.13 Organic Amines and Amino-acids

200

 

4.8

Conclusions

200

References

201

Contents

5

Compound Preparation, Mixture Characterisation and Process Enhancement of Particulate-Filled Polymer Compounds

207

5.1 Introduction

207

5.2 Functional Characteristics of Compounding Machinery

209

5.2.1 Transport of Feedstock

210

5.2.2 Melting and Shear Heating

213

5.2.3 Mixing

214

5.2.4 Melt Devolatilisation

217

5.2.5 Melt Pumping and Pressurisation

218

5.3 Constructional Design of Compounding Plant

219

5.3.1 Low and Medium Intensity Premixing Procedures

220

5.3.2 High-Intensity Compounding Machinery

221

5.4 Characterisation of Filled Compounds

228

5.4.1 Introduction

228

5.4.2 Residence Time Distribution

229

5.4.3 Specific Energy Input

231

5.4.4 Screen Pack Analysis

231

5.4.5 Rheological Analysis

232

5.4.6 Ultrasonic Measurement

233

5.4.7 Microstructural Analysis

235

5.4.8 Miscellaneous Methods of Analysis

239

5.5 Process Enhancement of Particulate Polymer Composites

240

5.5.1 Addition of Rigid Particulate Fillers

241

5.5.2 Effects on Polymer Molecular Weight

243

5.5.3 Short Fibre-Reinforced Thermoplastics Composites

244

5.6 Woodflour and Natural Fibre-Filled Thermoplastics

246

5.7 Supercritical Fluid Assisted Processing of Filled Compounds

249

5.8 Processing of Thermoset Recyclate Waste Materials

250

5.9 Preparation of Silicate Layer Polymer Nanocomposites

251

Particulate-Filled Polymer Composites

5.10

Conclusions

254

References

 

254

6 Effects of Particulate Fillers on Flame Retardant Properties of Composites

263

6.1 Introduction

 

263

6.2 General Effects of Fillers on Polymer Flammability

263

6.3 Fire Retardant Testing

264

 

6.3.1 Oxygen Index Test (ASTM D2863-87)

266

6.3.2 Underwriters Laboratory Vertical Burn Test (UL94 -1980)

266

6.3.3 Horizontal Burn Test (ASTM D635-88)

267

6.3.4 Ignitability Test (ISO5657 1986)

267

6.3.5 Cone Calorimeter (ASTM E1354, ISO 5660)

267

6.3.6 Smoke and Corrosive Gas Tests

268

6.4

Fire Retardant Fillers that Rely on Endothermic Decomposition

269

6.4.1 Historical Background

269

6.4.2 Potential Endothermic Flame Retardant Fillers

270

6.4.3 Performance of Endothermic Flame Retardant Fillers

273

6.4.4 Smoke and Corrosive and Toxic Gases

290

6.5 Nano-Clays

 

296

6.6 Ammonium Polyphosphate (APP)

297

6.7 Fillers for Use in Conjunction with Halogens

297

References

 

298

7 Particulate Fillers in Elastomers

303

7.1 Introduction

 

303

7.2 Uses of Elastomers

303

7.3 Elasticity of Rubber

303

7.4 Formulation of Elastomers

306

 

7.4.1

General

306

Contents

 

7.4.2 Selection of Polymer

306

7.4.3 The Curing System

308

7.4.4 Antioxidants and Antiozonants

310

7.4.5 Coupling Agents

311

7.4.6 Process Oils and Plasticisers

311

7.4.7 Fillers

313

7.4.8 Specialty Additives

314

7.5

The Performance of the Polymer

317

7.5.1 Specification of the Polymer

317

7.5.2 Processing Considerations

321

7.5.3 Strength Characteristics of Polymers

321

7.5.4 Compounding Considerations

322

7.6

The Performance of Fillers

326

7.6.1 Reinforcement of Rubber by Fillers

326

7.6.2 Processing Considerations

330

7.6.3 Compounding Considerations

333

7.7

Filler Types

338

7.7.1 Specification of Fillers for Elastomers

338

7.7.2 Carbon Black

340

7.7.3 Synthetic Silicas and Silicates

343

7.7.4 Clay Minerals

344

7.7.5 Calcium Carbonates

346

7.7.6 Alumina Trihydrate

348

7.7.7 Talcs

349

7.7.8 Natural Silicas

349

7.7.9 Barytes and Blanc Fixe

350

7.7.10

Miscellaneous

350

Acknowledgements

350

References

351

Particulate-Filled Polymer Composites

8

Filled Thermoplastics

357

8.1 Introduction

357

8.1.1 Thermoplastics and Typical Applications

357

8.1.2 Thermoplastic Composites

359

8.2 Bulk and Process Related Properties

361

8.2.1 Specific Gravity or Relative Density

361

8.2.2 Acoustic Properties

362

8.2.3 Melt Viscosity (MFI)

363

8.2.4 Compounding and Extrusion

364

8.2.5 Thermal Conductivity and Specific Heat Capacity

367

8.2.6 Thermal Expansion

368

8.2.7 Electrical Properties

369

8.2.8 Barrier Properties

370

8.3 Mechanical Properties

371

8.3.1 Introduction

371

8.3.2 Modulus – Tensile and Flexural

372

8.3.3 Heat Deflection Temperature (HDT)

374

8.3.4 Yield Strength

375

8.3.5 Impact Strength (Toughness)

376

8.4 Effects of Filler on the Polymer Phase

380

8.4.1 Introduction

380

8.4.2 Nucleation

380

8.4.3 Transcrystallinity

381

8.4.4 Interphase

382

8.5 Surface Science Aspects

383

8.5.1 Introduction

383

8.5.2 Surface Energy and Surface Tension

383

8.5.3 Wetting and Spreading

384

8.5.4 Adhesion

384

8.5.5 Dispersion and Agglomeration

387

8.5.6 Surface Treatments – Dispersants and Coupling Agents

388

Contents

 

8.6

Aesthetics

390

 

8.6.1 Introduction

390

8.6.2 Colour/Pigmentation

390

8.6.3 Surface Finish and Gloss

390

8.6.4 Scratch and Abrasion Resistance

391

 

8.7

Stabilisation and Recycleability

392

 

8.7.1 Introduction

392

8.7.2 The Effect of Filler Chemistry and Impurities on Stability

393

8.7.3 The Effect of Antioxidant Adsorption on Stability

394

8.7.4 Recycleability

395

 

8.8

Uses of Filled Thermoplastics

396

 

8.8.1

Uses of Fillers

396

8.8.2

Fillers in PVC

397

8.8.2

Uses of Fillers in Unplasticised PVC

401

8.8.3

Uses of Fillers in Polypropylene

403

8.8.4

Uses of Fillers in Polyethylene

406

8.8.5

The Use of Fillers in Polyamides

408

8.8.6

Polybutylene Terephthalate

409

8.8.7

Polyethylene Terephthalate

410

 

8.9

Conclusions

411

Acknowledgements

 

413

References

413

9

Filled Thermosets

425

9.1 Introduction

425

9.2 Brief Chemistry of Thermoset Polymers

427

 

9.2.1 Free-Radical Chain-Growth Curing Resins

427

9.2.2 Step Addition Curing Resins

435

9.2.3 Condensation Resins

440

 

9.3 Mechanical Properties

444

Particulate-Filled Polymer Composites

 

9.3.1 Modulus

444

9.3.2 Fracture Toughness and Fracture Energy

450

9.3.3 Failure Stress

463

9.3.4 Fatigue

469

 

9.4

Applications

477

 

9.4.1 Cost Reduction

477

9.4.2 Modified Mechanical Properties

477

9.4.3 Exotherm Control

478

9.4.4 Shrinkage Control

480

9.4.5 Processing Aids

481

9.4.6 Flame Retardants

482

9.4.7 Metal Fillers

482

9.4.8 Structural Adhesives

483

 

Acknowledgements

483

References

484

10

Composites Using Nano-Fillers

489

10.1 Introduction

489

10.2 Scope

489

10.3 General Comments

490

10.4 Nano-Filler Forms

490

 

10.4.1 Regular Shapes

491

10.4.2 Rods, Fibres,

492

10.4.3 Platy Nano-Fillers (Nano-Clays and Related Materials)

493

 

10.5 Summary and Future Perspectives

510

Acknowledgements

511

References

511

Abbreviations and Acronyms

515

Author Index

521

Index

527

Preface

I was delighted to accept the chance offered by Rapra Technology to produce this updated second edition of Particulate Filled Polymer Composites, a book first published in 1995. The first edition had been very well received but has been out of print for some time. Despite the relatively short time since the publication of the first edition, much has changed and hence considerable new material has been introduced, including a completely new chapter covering the latest developments in nano-filler technology.

The use of particulate fillers in polymers has a long history, and they continue to play a very important role today. Despite the apparent commodity status, the area is still very dynamic and considerable changes have occurred in applications over the last few years. The most significant of these have been the dramatic growth in the use of precipitated silica in energy efficient tyre applications and the emergence of ‘nano-filler’ technologies.

The multidisciplinary nature of this topic is well illustrated by the scope of the references in the book. These are drawn from many fields, including: mineralogy, crystallography, precipitation and crystal growth, powder technology, surface and colloid science, organic and organo-metallic chemistry, optics, materials science, and polymer science and technology. It also brings together people from many backgrounds, such as filler producers, machinery manufacturers, polymer compounders and suppliers of processing aids and surface modifiers.

Together with my co-authors, I have set out to describe the fundamentals involved in producing, characterising and using particulate fillers in as clear and concise a way as possible. The authors have been encouraged to pay particular attention to those areas of their subjects which their experience shows cause the most confusion, or which are poorly covered elsewhere. Most importantly, where a topic is controversial, or poorly understood, they have been encouraged to interpret the existing literature, not merely quote opposing views, as is often the case. Where possible, use has been made of previously unpublished work to illustrate key points and extensive literature references are provided so that any subject can be followed up in depth, if needed.

Chapter 4, on surface modification, is a good example of the approach taken. Most works concentrate on the organo-silanes. While they are still treated in detail here, significant space is also given to fatty acids and organo-titanates. Fatty acids are of considerable commercial importance and exhibit complexities that can cause problems

Particulate-Filled Polymer Composites

for the unwary. Considerable controversy exists over the mode of action of the organo- titanates and the available scientific literature is critically reviewed here.

As with any editor, I am very much indebted to my co-authors for their hard work. I am also indebted to the staff at Rapra Technology, especially Frances Powers, for their help and support in turning the various contributions into a very well presented composite.

Finally, I would like to express my particular thanks to the late Professor Derek Birchall, FRS, who first aroused my interest in this area and to the late Hugh Olmstead founder of Intertech Corporation and to my many friends and colleagues, especially those at Manchester Metropolitan University, Avecia Limited, Colin Stewart Minchem, CSIRO Division of Minerals, Electrolux and the Intertech Corporation, who have helped me to continue this interest in recent years.

Roger N. Rothon

July 2003

Contributors

David Ashton ICI Acrylics PO Box 34 Darwen Lancashire BB3 1QB

David Briggs Siacon Consultants Ltd. 21 Woodfarm Road Malvern Worcestershire WR14 4RL

Chris DeArmitt BASF Aktiengesellschaft

KSP/S-E100

67056 Ludwigshafen Germany

Michael Hancock 6 Poltair Avenue St Austell Cornwall PL25 4LY

Graham Jackson LGC Ltd. The Heath Runcorn Cheshire WA7 4QD

Peter Hornsby The Wolfson Centre Brunel University Uxbridge Middlesex UB8 3BH

Christopher Liauw Manchester Metropolitan University Centre For Materials Science Research John Dalton Building Chester Street Manchester M1 5GD

Michael Orton 24, School Lane Hartford Cheshire CW8 1PE

Roger Rothon Rothon Consultants and Manchester Metropolitan University 3 Orchard Croft Guilden Sutton Chester CH3 7SL

David Skelhorn Performance Materials Division Imerys 100 Mansell Court East Suite 300 Roswell Georgia 30076 USA

Howard Taylor Manchester Metropolitan University Department of Engineering and Technology John Dalton Building Chester Street Manchester M1 5GD

Particulate-Filled Polymer Composites

1 General Principles Guiding Selection and Use of Particulate Materials

Roger N. Rothon and Michael Hancock

1.1 Introduction

Particulate-filled polymer composites have a long history and consequently newcomers to the field usually expect to find an area of well-understood science with few intellectual challenges remaining. However, they are usually amazed to find that this is far from the truth in many areas, with few reliable generalisations (but several unreliable ones) available and much basic information yet to be established. This is largely due to the way in which the technology has developed, with different filler and polymer combinations tending to be developed largely piecemeal to meet the specific demands of various industries. The initial, and to some extent continuing, emphasis on cost reduction has also meant that many fillers have been poorly characterised.

The purpose of this chapter is to give readers sufficient background so that they may approach the subject with an open mind and see the whole area as one with stimulating intellectual challenges, not as an area of mystery and witchcraft. This chapter discusses how fillers can be characterised and defined with the least ambiguity, and hence how their behaviour in composites can be understood and hopefully predicted. We also highlight the principal sources of misunderstandings and the limitations imposed by commonly used methods of measurement, and explain why and where deviations from general rules are likely to occur.

1.2 Basic Characteristics of Particulate Fillers

In developing a particulate-filled composite, the formulator needs to be able to answer the following questions:

1. What property benefits are being sought?

2. What deleterious changes may also occur and can they be tolerated?

3. How easy is the filler to handle and how might it affect processing?

Particulate-Filled Polymer Composites

4. Are special additives needed?

5. What is the true cost of using the filler, is it justifiable, and are there more cost effective alternatives?

Important information for answering the above questions includes, cost, purity, particle size and shape, density, hardness, optical and thermal properties and chemistry. The primary source of this information should be the filler supplier, although frequently data may be sparse. A brief description of each factor, its measurement and significance follows.

1.2.1 Cost

In theory this is a fairly simple topic, but one that causes considerable problems for the unwary. It is widely assumed that polymers are expensive and fillers are cheap, and many articles on filled polymers start with the statement that fillers are used primarily to reduce costs. While this can often be the case, these savings are frequently not as great as anticipated and, in quite a few instances, compound costs, even with the lowest cost filler, can be higher than the unmodified polymer. There are two principal reasons for this.

Firstly, the process of compounding filler into polymer costs money in the form of capital investment, manpower and energy [1]. In cases where compounding is essential, because other additives such as stabilisers or curatives have to be added to the polymer, then the cost of incorporating a filler is markedly reduced. In these cases, exemplified by elastomers, polyvinyl chloride (PVC) and thermosets, the use of fillers is the rule rather than the exception, unlike the case with, say, polyethylene.

Secondly, prices of fillers and polymers are usually quoted in weight terms, while the majority of their composites are used on a volume basis. The specific gravity of most fillers is two to three times that of common polymers and, while raw material savings may accrue on a weight basis, more mass of the final compound will have to be used to achieve the same volume than would be the case with the unfilled polymer. When looking at potential cost savings, comparison of polymer and filler costs must therefore be made on a volume basis. On this basis one should generally regard the effective raw material costs of mineral fillers as two to three times their price (by weight). As a consequence many fillers will not give significant cost savings in the commodity plastics, such as polypropylene (PP) and this explains why the emphasis today is on achieving specific effects. However, while fillers may not ‘cheapen’ a filled composite they may well produce a cost-property performance that allows the composite to compete against and often replace more expensive systems.

General Principles Guiding Selection and Use of Particulate Materials

1.2.2 Chemistry, Composition and Impurities

While the chemical nature of the filler is frequently of little direct importance to its use in composites, it plays two important roles, in that it determines the structure of the mineral, and also the nature of the interaction between polymer and filler. However, away from these a priori considerations, only in a few cases does the chemical reactivity of the filler play a significant role in the properties of a composite.

1.2.2.1 Bulk Chemistry

Intrinsically fillers can be divided into two types, reactive and inert. Reactive fillers will react with their environment. A good example of this is gibbsite (aluminium hydroxide), which will react with both acidic and basic substances. Aluminium hydroxide also loses its water of crystallisation at around 200 °C and this enables it to provide fire retardancy in polymer formulations. The silicate minerals: (kaolin, mica, talc, quartz, etc.), are, in classical chemical terms, virtually inert, only being attacked by very strong acids and alkalis. The carbonate minerals and the hydroxide minerals are very reactive to acids.

The interactions between the constituent elements in a filler determine its molecular structure and hence crystallinity, which then dictates all the intrinsic properties of the filler.

1.2.2.2 Surface Chemistry

Filler surface chemistries are of more significance than the bulk ones, as they determine both the rate of wetting and the strength of interaction with polymers. They are invariably different from bulk chemistry but, unfortunately, they are poorly characterised for many fillers. Because of the interest in this very important topic, techniques for surface characterisation are covered in detail in Chapter 3.

1.2.2.3 Surface Interactions

Polymers have a very much higher (20-30 times) thermal expansion coefficient than mineral fillers. Thus, in many well-dispersed, hot processed composites, a radial compressive stress develops as the polymer cools leading to an intimate interaction between matrix and filler. The value of this stress can be calculated, and depends on both polymer and filler [2]. Because of this, the interaction will at the very least be mechanical, but other types will exist depending on the surface chemistry of the filler and also the chemistry of the polymer.

Particulate-Filled Polymer Composites

Wettability of filler by polymer is an indication of compatibility between the two. Wettability by polymer is not readily measured directly, but the effect that the material has on surface tension of a liquid is a measure of its surface energy. Usually water is chosen as the medium. Schlumf [3] has reported surface energies for a variety of fillers (Table 1.1). There have been a wide number of claims that surface energies determine the forces between polymer chains and different phases, determining mechanical properties of the composite. Adhesion at a polymer-filler interface has been shown to exert a considerable influence on mechanical responses, and a correlation between acid-base characteristics of filler and polymer (as determined by inverse gas chromatography) and properties has been established [4]. Lewis acid-base interactions between filler and polymer are claimed to be the most important component in the adhesion of a filler to polymers and thus in determining properties of composites [5]. For some filler-polymer systems, the strength of interfacial acid-base bonding may be appreciably enhanced by surface modification of the filler, or by modification of the polymer, giving large increases in properties.

Table 1.1 Surface energies of fillers and plastics

Material

Surface energy (m Jm -2 )

Diamond

10,000

Mica

240-500

Glass

1,200

Titanium dioxide

650

Kaolin

500-600

Calcium carbonate

65-70

Stearate coated calcium carbonate

25-30

Talc

65-70

Polymers

15-60

Polypropylene

31

1.2.2.4 Chemical Analysis and Impurities

The purity of a filler is of importance both commercially and technically, i.e., the users need to know what they are buying, whether it contains components that may be detrimental to the properties of their product and whether it will cause environmental problems. Impurities include trace elements that may be on the filler surface or in the structure of the filler and ancillary minerals, which will have been formed at the same

General Principles Guiding Selection and Use of Particulate Materials

time as the main mineral (see Chapter 2). Sometimes additives from the filler manufacture may be present, (e.g., surfactants). The form of the impurity can be very important in determining it’s importance. Sometimes a potentially detrimental impurity can be tolerated in relatively high levels, if present in an inert form, but small traces of the same impurity, if present on the surface in an active form, can be very deleterious. Examples are certain transition metals, (e.g., iron, manganese, copper), which can seriously affect colour and thermal stability if present in an active form, even at levels of a few parts per million. Thus some micas and talcs can contain high levels of iron in an inert form, which is not detrimental to polymer ageing, while others can have lower levels of more active material. Some impurities may also pose health hazards, examples being crystalline silica and asbestos. Fillers are not always analysed for such low-level impurities and it is a mistake to believe that, if a possible deleterious impurity is not quoted in a data sheet, it is not present in the product. Conversely, it must be stressed that, if an element or substance is present in a non-active form, it may have little effect on polymer properties [6].

It is also necessary to point out that the way chemical composition is presented, can, and does, cause misunderstandings. Because of the methods of analysis used, many mineral compositions are quoted as a list of oxides rather than the actual constituent chemical. Thus a calcium carbonate may be specified as so many percent CaO or a clay may be reported in terms of SiO 2 and Al 2 O 3 . This does not imply that any of the oxides are present in the free state. It should also be recognised that some standard methods of analysis do not detect very light elements.

1.2.3 Density or Specific Gravity

This is one of the simplest of properties to define, understand and use. Density is mass per unit volume with SI (Système Internationale) units of kilograms per cubic meter. Specific gravity and relative density, the weight of the substance in relation to the weight of an equivalent volume of water, are often used synonymously. They are dimensionless but have the same numerical values as density.

Density is determined by the size of the atoms forming the mineral, the closeness with which they are packed together and by impurities present in the crystal lattice: the heavier the atoms and the more tightly they are packed together, the higher the density. Thus, smithsonite, ZnCO 3 , has a higher specific gravity (4.3) than calcite, CaCO 3 (2.7) because zinc has a much higher atomic mass than calcium, but magnesite, MgCO 3 , has a higher specific gravity (3.0-3.2) than calcite despite the fact that magnesium is lighter than calcium. This is because its ions pack much more closely together. In some fillers air is deliberately introduced into a matrix to reduce its density.

Particulate-Filled Polymer Composites

True specific gravities of particulate additives can range from less than 1.0 for hollow glass beads to over 6.0 for some metallic fillers, but most lie in the range 1.6-2.8, while common polymers range from 0.9 to 1.4.

Powder densities are of commercial importance, indicating ease of handling, and polymer compounding, and even affecting plant design. Values for bulk and packing densities are usually supplied by filler producers and are a measure of how the powder packs under various conditions. They are determined to some extent by the true density, particle shape and size of the filler, but the processing involved in its production also plays an important role. Surface treatments, and the methods used to apply them, can also be important.

A powder with a low bulk density relative to its true density will contain a great deal of air. It will flow easily but will be difficult to incorporate into polymers, especially in equipment such as internal mixers and screw compounders.

1.2.4 Hardness

Hardness is the resistance of a mineral to scratching. It is related to the structure of the mineral, the strength of the chemical bonds and the density of packing of its constituent atoms. F. Mohs, a mineralogist drew up a table of hardness in 1812 called the Mohs Hardness Scale (Table 1.2), which is still used to rank minerals by their resistance or susceptibility to scratching by other minerals. This scale is widely used for mineral fillers, but care must be taken in using the data because the scale is not linear, with the differences becoming greater at the high end. The hardness of most common particulate fillers range from 1 to 4. Talc is the softest, with a value of one.

Table 1.2 Mohs scale of mineral hardness

 

1 Talc (softest)

5eApatit

8

Topaz

2 Gypsum

6eOrthoclas

9mCorundu

3 Calcite

7

Quartz

10

Diamond

4 Fluorite

 

Hardness must not be confused with toughness, which relates to the ability of the mineral to resist fracture. Hard minerals can often fracture very easily, for example, by cleavage along crystal planes.

General Principles Guiding Selection and Use of Particulate Materials

Where mixed fillers are used, if one is much harder than the other, then abrasion of the softer one may occur. This is of particular importance when hard particulates are used with easily damaged glass fibres.

Minerals such as talc are soft because they readily delaminate due to weak structural features (see Chapter 2). Such soft particles can fracture or cleave when polymer composites are deformed, thus limiting their reinforcing properties [7].

1.2.5 Abrasiveness

Abrasion due to the filler has quite an important role in the processing of filled polymers.

It has several origins. Mohs hardness is a major factor with the harder minerals, such

as quartz and wollastonite (CaSiO 3 ), approaching, or even exceeding the hardness of the metals used in processing. Large particles and angular particles with sharp edges are particularly detrimental, probably due to a scratching mechanism, followed by attrition of the exposed edges. Sometimes the coarse component of a filler may be predominantly a harder impurity, which leads unfortunately to a combination of the two mechanisms.

Chemical attack may also be caused by some reactive fillers, leading to etching and progressive erosion.

1.2.6 Optical Properties

The colour, opacity and gloss of a composite are very important, both aesthetically and functionally, and will be strongly affected by the incorporation of fillers. When light strikes any surface, it is subjected to several effects: absorption, scattering, reflection, polarisation and interference. The extent of each effect is dependent on the nature of both exit and entrance media, but for all practical purposes only three media, air, filler and polymer, will be considered here in a superficial treatment. The reader is referred to works such as by Bohren and Huffman for a detailed explanation [8].

A light ray incident upon an air-solid interface at an angle, i, will be bent or refracted

somewhat normal to the interface, at an angle of refraction, r. A relationship of the

incident angle, i, and the refracted angle, r, sin i/sin r, is constant for a given isotropic solid. The constant always denoted as n is called the refractive index. The refractive index of a filler is one of the most important parameters affecting the optical properties

of

a filled polymer system and is always quoted relative to air. It is, in fact, a manifestation

of

the ratio of the speed of light in air to its speed in the solid.

Particulate-Filled Polymer Composites

The refractive index of a mineral is the sum of the specific refractive indices of all individual components normalised for concentrations. Simplifying, it can be said that the refractive index of a solid is directly related to its density. When light strikes a surface, some will be reflected with the amount reflected being determined by the density of the constituent atoms and is thus affected by the refractive index, n. Reflection (r) and transmission (t) coefficients are given by:

r =

1 n

1 + n

t =

2

1 + n

(1.1)

To a first approximation this leads to the fraction of incident light being reflected as simply (n 1) 2 /(n + 1) 2 and approaches zero as the refractive index approaches unity. The response of a plane surface to an incident beam of light is complicated, as briefly described previously. Mineral fillers, however, complicate the subject even more deeply in that they are a priori made up of particles that usually have sizes in the order of the wavelengths of visible light. Thus large surfaces will reflect and transmit the light waves but particles will scatter them.

Therefore, light entering a typical powdered filler, comprising many small particles, will experience reflection, refraction, diffraction and interference at each air - particle interface. Refraction into the filler particles can lead to absorption of the light (the radiation absorbed will be converted into another form such as heat) and a measure of this is the absorption coefficient. The remainder of the light is diffused by scattering and the measure of this is its scattering coefficient. The colour (sometimes referred to as brightness, reflectance or whiteness) is related to absorption and scattering coefficients by Kubelka-Munk equations [9]. The reader is referred to the original paper and to some of the speciality books on the subject [10] for more details. The level of scattering is very dependent on particle size, so that finer fillers appear whiter than coarser ones: as an approximation, scattering in air is at a maximum when the particle size of powder is one-third to one-half the wavelength of the radiation.

Levels of reflectance and refraction, and hence scattering, are directly related to the refractive index of the mineral. When incorporated into a polymer matrix, the ratio of its refractive index to that of the polymer will strongly affect the optical properties of the composite. As the ratio approaches unity (which is the case for most common polymers and filler): reflection at the interfaces between filler and polymer approaches zero. This means that light scattering is reduced dramatically and light absorption becomes obvious. Mie’s theory [11] now widely used for particle size determinations from light-scattering data, predicts that as differences in refractive indices become small, not only does scattering become less but particle size effects become less and maximum scattering moves to larger sizes, often several times greater than the wavelength of the light.

General Principles Guiding Selection and Use of Particulate Materials

A filler such as titanium dioxide, having a refractive index much higher than that of the polymer, will therefore exhibit considerable scattering and will give white products, unless its absorption coefficient is high in which case dark, opaque composites will be produced (carbon black or black iron oxide are the ultimate in this instance).

Most fillers are white or off-white powders due to the reflection of visible light. This reflection consists of two components, absorption and scattering, and, as discussed previously, most fillers have low absorption but high scattering coefficients. However, the amount of light scattered depends on the refractive index of the mineral relative to the medium in which it is measured. In air the differences in refractive indices mean that most minerals scatter light and appear white. Most plastics, on the other hand, have refractive indices close to those for minerals (1.5-1.6) and much of the scattering is lost (the light, in fact cannot ‘see’ the boundary between filler and polymer). There are a few cases in which mineral and polymer have almost identical refractive indices. In these instances, scattering virtually vanishes and the composite becomes almost transparent until loadings are increased to a level at which particle-particle interaction occurs [12]. This transparency is of commercial importance in several applications such as silica-filled styrene-butadiene rubber (SBR) shoe soles, and calcined clay- and glass fibre-filled unsaturated polyester roof panels.

On the other hand, for example, calcite has a refractive index of 1.6, which is sufficiently different from most of the common plastics for some scattering and hence pigmentation to occur, but whiteness and opacity will only become very noticeable at high loadings and with thick products.

With the loss of light scattering, light absorption becomes the dominant optical property of a mineral-filled plastic. Light is absorbed by transition metal ions, with different wavelengths being absorbed by different species of ion, and by naturally occurring organics, which are poly-aromatic species with strong absorption across the whole spectrum (that is they appear brown or black). The impurities in mineral fillers are mainly transition metal ions, particularly iron, and humates or lignates. Colours due to these become frequently more marked in a plastic.

Some minerals exhibit birefringence with refractive indices different in differing directions, due to an asymmetric crystal structure. Plastics filled with these are dichroic and exhibit unusual visual appearances.

As well as being a direct cause of the colour in a filled plastic, a filler may indirectly generate colour by either degrading the polymer (forming radicals and unsaturated species) or by deactivating antioxidants/stabilisers, etc., used in the polymer. In some cases, breakdown of filler surface treatments can also generate colour problems. Filler effects can be observed in most thermoplastics but are particularly relevant in PVC where dehydrochlorination, giving a polyene conjugated structure, is readily promoted by Lewis acid sites on a mineral surface.

Particulate-Filled Polymer Composites

During the processing of filled thermoplastics, but particularly during their injection moulding, white markings are observed at points where high melt strain rates are encountered. These are especially apparent around the sprue and are very much more noticeable than for unfilled plastics. Some may be due to water or gas bubbles (splash markings), some to poor dispersion or filler separation, but most seem to result from flow patterns. These have been frozen in a strained state before they have had time to relax due to the high thermal conductivity of filled polymers [13].

When filled polymers are strained, they whiten to some extent. This is partly due to orientation of polymer chains around filler particles but also to light scattering from vacuoles formed as the polymer is pulled away from the particle surfaces.

Optical properties can also be affected by the fact that filler particles can affect the morphology of the plastic by nucleating its crystallisation at the interface. Thus a composite with different light scattering properties may be obtained.

1.2.6.1 Spectral Absorption

The ultraviolet and visible spectra of the common fillers are largely featureless, unless there are significant levels of transition metal ions in the mineral, when strong absorptions will occur. However, the most important practical aspect is the infra-red (IR) spectrum of a filler. This arises from the nature of the chemical bonds present and will be unique for each filler type. A common use of IR spectroscopy is as an analytical tool to identify a filler. An important practical consequence of a filler’s IR absorption is the use of certain fillers in horticultural and agricultural films (see Chapter 8). Spectra of some of the more common minerals incorporated in a low- density polyethylene (PE) film are shown in Figure 1.1. Calcium carbonate shows only two strong absorption bands because the only energy absorptions occur in the CO 3 ; ion (bending, stretching and deformation modes). Kaolin, a complex alumino- silicate, shows a much more complicated IR spectrum: after calcining, the —Al—O—Si— bonds rearrange and the resultant product has a very complex IR pattern with several broad, overlapping bands.

1.2.7 Thermal Properties

Leaving aside the special cases of phase change and decomposition, the principal thermal properties of interest are specific heat, thermal conductivity and coefficient of expansion.

General Principles Guiding Selection and Use of Particulate Materials

Guiding Selection and Use of Particulate Materials Figure 1.1 Infra-red spectra of mineral filled polyethylene

Figure 1.1 Infra-red spectra of mineral filled polyethylene films

1.2.7.1 Specific Heat

This is usually defined as the energy required to raise 1 gram of material by 1 K with SI units of J.kg -1 .K -1 . Specific heats are easily measurable and available for most particulates [14]. Where they are not, rules exist for estimating the value [7]. The volume specific heats of most inorganic fillers are similar to the common polymers and the rule of mixtures gives a good approximation to the values found in composites.

1.2.7.2 Thermal Conductivity

This is the rate at which heat energy is transmitted through a substance, SI units are W m -1 K -1 but it is usually quoted as cal cm -1 K -1 . Mineral fillers are about an order of magnitude more conducting than most polymers, while metallic fillers are even more so. Filled composites are therefore more conductive in general than the base polymer. The simple law of mixtures is only applicable at low concentrations because, as with electrical conductivity, particle-particle contact will occur at high loadings leading to a sudden

Particulate-Filled Polymer Composites

sharp increase in conductivity, the so-called percolation threshold [15], which is dependent on particle shape and packing.

This increased conductivity is of considerable importance in some processing operations, especially injection moulding.

1.2.7.3 Coefficient of Thermal Expansion

As the name implies, this is the rate at which a material changes volume on heating or cooling. SI units are K -1 . Most polymers have coefficients of expansion at least an order of magnitude greater than mineral fillers: thus mineral-filled composites have lower coefficients of expansion than unfilled polymers and, in well-annealed systems, the rule of mixtures applies unless strong bonding occurs between filler and polymer. Some fillers have a negative coefficient of expansion and are used to achieve composites with zero expansion. Other fillers show anisotropy, that is they have coefficients that are direction dependent.

The differences in expansion between particle and matrix can result in intimate contact between them in a composite and consequently considerable stresses occur. These can have the same properties as chemical bonding at low strains and can mask any effects of changes in polymer-filler interactions in some tests.

1.2.8 Particle Shape and Size

The importance of these factors is felt at all stages of composite production and use. They affect powder flow, compounding behaviour, composite viscosity, and mechanical, thermal and optical properties. Indeed, most of the current predictive equations for the properties of filled composites use shape and size factors, often determined using model particles such as glass spheres and flakes. Unfortunately such particles are rare in the real world and it is important in applying equations to appreciate the limitations imposed by problems of adequately measuring and describing the morphology of fine particles.

One of the principal difficulties is due to the ability of many fillers to exhibit a variety of particle shapes and sizes depending on the work done in dispersing them. Their ‘effective’ shape and size can therefore vary at any stage of composite formation and use. In principle one would like to characterise them in situ. This is, however, far less easy than characterising the initial particulate material itself and one is usually reduced to trying to carry out measurements under conditions that will represent as near as possible those encountered in use. In this context, the concept of ‘effective’ particle, which is the size and shape achieved in the actual application, is a very useful one to keep in mind and is returned to later in this chapter.

General Principles Guiding Selection and Use of Particulate Materials

1.2.9 Shape

1.2.9.1 Introduction

Particle shape (as will be discussed in following chapters) is very important in determining the stiffness, or rigidity, of a composite, the flow and rheology of a melt or liquid, tensile and impact strength, and the surface smoothness of a component, i.e., many of the important properties of a composite.

Shape is determined by the genesis of the filler, by its chemistry, its crystal structure and by the processing it has undergone. Unfortunately, it is usually poorly defined, the literature abounding with vague terms such as roughly spherical, blocky, irregular, platy, acicular, etc. Some typical particles likely to be found in fillers are shown in Figure 1.2. All the fillers commonly used are microscopic in size imposing major difficulties both in how to measure, and then how to describe and quantify shape in any simple yet meaningful way. In addition, there are problems of distinguishing between primary particles, agglomerates and aggregates that occur in a filler, especially in synthetic materials, in which a ‘structure’ is sometimes deliberately designed. The subject is complicated further because aggregates and even particles may break down during processing (see Section 1.2.10.3 for a definition of these various terms).

Section 1.2.10.3 for a definition of these various terms). Figure 1.2 Some particle types likely to

Figure 1.2 Some particle types likely to be found in common fillers

Particulate-Filled Polymer Composites

Despite the complex problems outlined previously, fillers are used because of their shape in a very wide range of polymers to give specific properties. For example, conventional clays are used in hose and chemical lining because their shape reduces permeability to fluids; platy talcs give rigidity to PP; the complex aggregate structure of precipitated calcium carbonates contributes to the structure of liquid polysulfides; and the special structured shape of many carbon blacks and synthetic silicas is important to their performance in elastomers. A further example is the emerging use of very high aspect ratio, nano-clays as reinforcing, fire retarding and gas and fluid barrier fillers.

1.2.9.2 Origins of Particle Shape

Synthetic products will have their shape determined by their chemical composition and by the production conditions. For example, precipitated calcium carbonate (CaCO 3 ) can be produced with different shapes by changing precipitation conditions to produce aragonite, calcite or vaterite. These conditions can be chosen to produce single crystals or complicated aggregates, which can be modified during drying and milling. Another example is the very complex shape of most carbon black particles arising from the partial fusing and solidification of pyrolysing droplets into three-dimensional chain-like aggregates during manufacture.

The external shape of a natural mineral is a manifestation of its crystal structure, which will be briefly discussed in Chapter 2. It is also dependent on the environmental conditions in which the mineral was formed. If allowed to grow without constraint, then the particles are bounded by crystal faces, which are disposed in a regular way such that there is a particular relationship between them in any one mineral species, which is derived from regular atomic arrangement. However, under pressure, temperature or the effects of impurities, the crystal may adopt different shapes or habits. These include cubic, fibrous (fine, long, needles), acicular (needle-like), lamellar (plate-like) and prismatic. It is very unusual for perfect crystals to be found, but even poorly formed ones will always show evidence of their intrinsic symmetry.

Many fillers (as will be discussed later) being extracted from the earth or rocks and processed by fairly simple methods will exhibit the same basic original shapes. Others, however, may be extracted by complicated routes or involve intensive comminution, thus altering particle shape. This is determined by the chemical nature and strength of the bonds between the atoms and groups in the mineral. These depend on electronegative differences between atoms and the bonding can range from covalent through to ionic. Two other types of interaction or bonding are important.

Hydrogen forms co-ordinate bonds with atoms with very high electro-negativities, such as oxygen. This is very important in hydrated minerals such as kaolinite and gibbsite,

General Principles Guiding Selection and Use of Particulate Materials

with layers being bonded together by these relatively weak hydrogen bonds. In ‘neutral units’ another weak force, known as Van der Waals bonding, due to non-uniform charge distributions, bonds group together. The types and strengths of bonds present in the crystal determine the methods by which the minerals break down on processing. Thus in the calcite crystal, the weakest bonds exist between the calcium ion and the carbonate group and these cleave invariably, giving rise to a regular rhombohedral shape. Kaolin platelets, held by hydrogen bonds, separate more readily from each other rather than fracture. In talc, magnesium silicate layers are held together by weak Van der Waals forces and very readily undergo cleavage into progressively thinner plates.

1.2.9.3 Assessment and Measurement of Shape

The shapes outlined previously are very difficult to specify precisely using Euclidean geometry, but various approaches have been attempted to describe shape, anisotropy and ‘structure’ as numbers. Structure is particularly interesting and, while only really recognised currently in the elastomer field, it probably has significance, as yet overlooked, in other areas.

In synthetic fillers it is sometimes difficult to separate fundamental from aggregate shape but, where there is sufficient incentive, then ways will usually be found to overcome such difficulties. Such incentives arose in carbon blacks and more recently in precipitated silica, where shapes are very complex, but an understanding is critical to their high value usage in the tyre industry. Much work has been done, especially by Medalia and Heckman [16], and by Hess and co-workers [17], to develop automatic image analysis procedures. Using such procedures, all the aspects described previously have been investigated. This work has much to teach us about other filler particles.

Fractal geometry is now showing great promise for describing the shape of various complex, irregular particles. As usual, the carbon black industry is leading the way in exploring the potential of fractals for describing fillers and promising results are already emerging. A good introduction to fractal geometry and its applications to particles can be found in the excellent work by Kaye [18], and a number of recent papers have explored the measurement and significance of fractal dimensions of filler particles including carbon black [19-21].

The importance of measuring ‘structure’ results from its role in describing the ability of a particle to trap and partly shield a portion of the polymer matrix from deformation. This is generally known as an occluded polymer and its presence is important, especially in many of the unique properties of carbon and silica-filled elastomers. A simplified picture is given in Figure 1.3 and structure can be thought of as increasing the effective filler volume. While the carbon black industry is leading the way in using sophisticated techniques to measure shape and assess structure, they are also prepared to use fairly simple measurements, and in particular the oil absorption procedure with certain refinements, to give a quantification of this parameter [22].

Particulate-Filled Polymer Composites

Particulate-Filled Polymer Composites Figure 1.3 Schematic illustration of the effect of particle structure and occluded

Figure 1.3 Schematic illustration of the effect of particle structure and occluded polymer

The shape of simple particulate fillers is usually expressed as its aspect ratio (AR), which is the ratio of the particle’s diameter to its thickness. Measurement of aspect ratio is relatively simple for large particles, although time consuming. But, for the micrometer-sized particles that are normally encountered in fillers used in polymers, it is difficult, needs expensive equipment (such as scanning electron microscopes), and is very time consuming, depending on the spread of aspect ratios normally found.

Other techniques are being sought to give an easier, more rapid measurement. Fourty reports that a comparison of particle-size distribution curves measured by sedimentation and laser diffraction techniques can give a shape factor [23].

Electrical conductivity of an aqueous suspension of a mineral under flow compared with it at rest (after the particles have randomised) gives a mean value for AR or a shape factor. Reasonable agreement with electron microscopy has been found [24].

This complex situation is further complicated by the fact that not only is shape often dependent on size, but measurement of size is affected by shape. This interdependence is discussed next and also in the descriptions of individual minerals given in Chapter 2.

1.2.10 Particle Size

1.2.10.1 Introduction

Particle size is a very important, property of a filler, but is a particularly complex area where great confusion still occurs. For an individual, naturally occurring, filler it will have been determined by the origin and mineralogy of the deposit from which it has been extracted, by the method used in mining, and by separation procedures used during processing. For synthetic fillers, size will be determined by the conditions used

General Principles Guiding Selection and Use of Particulate Materials

in its synthesis such as precipitation and possibly by the drying and any coating procedures. Size is one variable that can be controlled (in theory at least), and its importance is felt at all stages of composite production and use. Hence, there is considerable interest in its measurement. Particle size distributions are more useful than single average values, although the latter have the merit of simplicity. Size is an easy property to measure reproducibly using a variety of techniques including sieving, sedimentation, optical scattering and diffraction from particulate suspensions. Each, however, has a different dependence upon particle dimension and varies with particle shape. Thus correlation between different minerals measured by different techniques, in different laboratories is difficult. Again the reader is referred to one of the specialised books on the subject [25].

The problems arise from the fact that most filler particles are irregular in shape and contain a wide range of sizes, some of which will be individual particles and others agglomerates (also with a different shape from the individual particles). In such instances it is not possible to describe fully the particle size by a few numbers, as is often attempted. Commercially, most producers will quote two or three values to indicate the size of their fillers (see Section 1.2.10.2 and Chapter 2). Thus the data used in published studies are often a poor approximation of the real situation and can be very misleading. Wherever possible, the reader should try to determine and to use the full particle size distribution curve and be very cautious in interpreting literature studies in which only limited data such as average particle sizes or surface areas are given. In particular very small amounts of particles above or below a critical size can sometimes dominate certain properties, (e.g., viscosity, fatigue, tensile and impact strength) but will not be apparent in many particle size determinations. An interesting example of the effect of small amounts of oversize particles on the properties of carbon black-filled elastomers has been given by Gent [26].

1.2.10.2 Particle Size Measurement

As already stated, filler suppliers usually quote three or four properties as a measure of the particle size:

1. Top cut. This is the size below which the majority (usually 99% or 99.9% by weight) of the particles are finer.

2. Coarse particles may also be reported as that percentage above a certain size.

3. Average particle size is obtained from particle size distribution curves and is the size that 50% of the particles are below. This can be defined in a variety of ways, e.g., by weight, volume or number, and care must be taken to understand the definition being used.

Particulate-Filled Polymer Composites

4. Specific surface area. This is determined by a number of adsorption techniques and is affected by size, shape and structure of the filler particles.

Coarse particles, i.e., particles above 40 μm in size are usually quoted separately and are measured by screening the filler through standard mesh sieves. This may be carried out in aqueous suspension or as a dry powder. If the filler is heavily agglomerated then erroneous results may be obtained especially in dry screening. For accuracy, aqueous measurements are preferred with the filler being completely dispersed both by mechanical means and also using a dispersing agent. An idea of the structure and ease of dispersion of a filler may be obtained by comparing screen residues obtained by both wet and dry screening. If the filler is hydrophobic, then difficulties will be encountered in wet screening. These may be overcome by dispersing the filler in an organic liquid such as butanol, but a wetting agent and a dispersing agent are often both required.

For some applications it is important to know the amount of particles that exceed 10 μm. This may be obtained by sieving but practical difficulties due to sieve screen blinding can occur.

The most widely used techniques in this case involve sedimentation in a fluid, usually water and applying Stokes’ law. This states that, under standard equilibrium conditions, the time, t, taken for a particle to settle to a fixed depth is inversely proportional to the square root of its spherical diameter, d. This is given in Equation 1.2:

d = 10

4

gt

18

η l

(

ρ

s

ρ

f

)

1

/

2

(1.2)

where d is the particle diameter, η is the viscosity of the fluid, ρ s and ρ f are the densities of the solid and the fluid, respectively, g is the gravitational constant and t is the time for the particle to settle a distance, l.

Diameters obtained in this manner are precise descriptions of the particles but will compare with the results of other techniques only when the particles are perfectly spherical. In practice, most fillers have irregularly shaped particles and it is common to interpret experimental data in terms of theories applicable to spherical particles. Dimensions obtained are then for ‘equivalent spherical diameters’ (esd), which are the diameters of spheres that would give the same behaviour as that obtained from the sample by the method in question.

Specific surface area is usually measured by the quantitative adsorption of nitrogen following the procedure originally described by Brunauer, Emmett and Teller, and known as the BET method (Equation 1.3) [27]:

General Principles Guiding Selection and Use of Particulate Materials

P

1

C

1 P

VP (

o

P

VC

m

V C P

mo

) =

+

(1.3)

where V is the volume of gas, under standard conditions, adsorbed at equilibrium pressure, P; V m is the volume of gas necessary for monolayer coverage; P o is the saturation pressure; and C is the BET constant. Often only a single point measurement is made to obtain surface area, but it is much more reliable to measure adsorptions at several pressures and then plot P/ V(P o P) against P/P o .

Permeability of a bed of packed particles to a fluid can also be used to give a surface area [28].

However, there are major sources of difficulty and confusion in describing particle size and shape. These include adequately defining what subdivision of material is to be described as the particle, in relying on particle size measurements carried out on the powdered particulate as an adequate description of it in a composite, and how shape and size interact.

1.2.10.3 Primary Particles, Agglomerates and Aggregates

It surprises many people that a variety of ‘particles’ can exist in a powder, each of which can be determined as the particle size under certain conditions of dispersion and measurement. Moreover, the appropriate particle size for consideration can itself vary according to whether one is dealing with powder flow, behaviour during compounding and dispersion, or the properties of the final composite. These different types of particles are generally described as primary or ultimate particles, agglomerates and aggregates. In some instances these particle types are readily distinguishable but in others there can be appreciable overlap. Before discussing the types of particle in detail it is necessary to clarify the terminology as two contradicting conventions are widely used, and this in itself is a cause of considerable confusion. The need is to distinguish between collections of particles that are weakly and strongly bonded together. In this book we shall use the term agglomerate for weakly bonded particle collections and aggregate for strongly bonded ones. This is opposed to the views of Kaye [18] based on the derivations of the two words. The chosen terminology is, however, at least as widely used and is especially prevalent in the carbon black industry. The reader should always ascertain which terminology is in use when reading articles in the literature.

An idealised view of particle type and breakdown with work during composite formation is presented in Figure 1.4. This goes beyond primary particles and considers the case where fragmentation of these may occur, e.g., hollow glass beads during thermoplastic compounding. With such simple systems it is usually fairly easy to match the particle sizing method to the degree of dispersion expected in the composite and obtain realistic answers.

Particulate-Filled Polymer Composites

Particulate-Filled Polymer Composites Figure 1.4 Idealised view of the way filler particles disperse and of the

Figure 1.4 Idealised view of the way filler particles disperse and of the different forms of particle types that might be encountered

Unfortunately, many filler systems do not exhibit such simple profiles, and the steps shown are often less sharp and overlap. Some possibilities are shown in Figure 1.5. A further complication arises when agglomerates form from initially well-dispersed systems. These agglomerates are sometimes referred to as flocs and can arise due to loss of colloidal stability in polymerising systems, or to reticulation (filler network formation) above the glass transition, especially in cured elastomers, an effect often observed with carbon blacks. The most difficult situation to deal with is quite frequently met with synthetic products, especially those formed by precipitation. This is where quite strong, complex aggregates are present, in addition to agglomerates. These aggregates often break down slowly, leading to a drawn-out step in the effective size profile. Frequently, they do not fully break down to primary particles under normal processing conditions. The effective particle size will then be critically dependent on the exact processing conditions and will be very difficult to predict in advance. Precipitated calcium carbonates are a good example where this type of situation is encountered.

Most particle sizing techniques attempt to break down powders to their primary sizes by use of intensive mixing, ultrasonics and dispersants, but even these can be insufficient.

Curves of the type shown in Figures 1.4 and 1.5 would be very useful for predicting filler behaviour, especially if the work input could be related to different compounding procedures. In principle, this should be possible, at least in a semi-quantitative way, by varying the energy used in dispersing the powder. Such procedures have been found

General Principles Guiding Selection and Use of Particulate Materials

Guiding Selection and Use of Particulate Materials Figure 1.5 Complex particle dispersion behaviour, as often

Figure 1.5 Complex particle dispersion behaviour, as often encountered with fine, precipitated fillers

valuable by one of the authors and some very useful preliminary work of this type has been reported by Thoma and co-workers [29]. Some typical results obtained by the author (RR) with a coated precipitated calcium carbonate are given in Table 1.3 (Note only average particle size values are given for illustrative purposes, the earlier comments about

Table 1.3 Effect of measuring conditions on the apparent particle size of a coated precipitated calcium carbonate

 

Average

 

Measuring Condition

particle

Comments

size

 

(μm)

 

Laser diffraction of organic dispersion using weak ultrasonics

20

Detecting agglomerates

Laser diffraction of organic dispersion using medium ultrasonics

4

Detecting basic aggregate size

Laser diffraction of organic dispersion using strong ultrasonics

0.2

Detecting some form of sub-aggregate structure

Electron microscopy

0.07

Detecting primary crystallite size

X-ray line broadening

0.07

Detecting primary crystallite size

Particulate-Filled Polymer Composites

the dangers of using a single value for a distribution still apply). Furnace-type carbon blacks provide a good example of the problems involved with concepts and terminology in this area. Such carbon black particles are formed by partial fusion and solidification of very small spherical units. These spherical units are generally referred to as the primary particles. Under no compounding or particle-sizing conditions, however, is it possible to break down the actual particles completely into these ultimate units. Unfortunately, some breakdown of the fused structures does occur, depending on the processing conditions, so that it is not appropriate to consider these fused structures as the ultimate particle either. In view of these complexities, the best rule is to keep the basic principles as outlined previously firmly in mind when trying to relate particle size as determined on a powder to what happens in an actual composite.

As mentioned before, many of the previous problems would be removed if there were simple and reliable methods for assessing the ‘effective’ particle size in the polymer matrix. Various methods can be used including microscopy and contact radiology, both coupled with image analysis, and recovery from the polymer followed by conventional sizing techniques, although none are completely satisfactory. A more detailed discussion of some of these aspects will be found in Chapter 5. In the authors’ opinions the understanding of filled composites would be greatly improved by advances in this area.

1.3 Surface Modification

Surface modification of fillers to give improved properties to a polymer composite is a topic that has received enormous attention over the last 30 years. Improvements in mechanical properties, dispersion of the filler (which leads to improved properties), improved rheology and higher filler loading have all been reported to accrue from rendering the surface more hydrophobic and hence compatible with the polymer or by enabling the filler to bond covalently, through hydrogen or ionic bonds to the polymer; or by changing the physical nature of the interface so that energy absorption can occur. This section will deal mainly with mineral surface interactions. The use of surface modifiers is dealt with at length in Chapter 4 and specific examples are described in Chapters 2, 7, 8 and 9.

1.3.1 Stearic Acid and Stearates

The most widely used surface modification is treatment with stearic acid. This is believed to result in a stearate salt coating on most fillers and metal stearates are also used. Stearic acid will react with basic minerals to give a surface that is covered with strongly bonded long organic ions (this is discussed in more detail in Chapter 4). Stearic acid- modified silicates are commercially available but in these cases the stearic acid is almost

General Principles Guiding Selection and Use of Particulate Materials

certainly weakly adsorbed and probably desorbs during melt compounding. Similarly, metal stearates will form weak bonds with mineral surfaces and desorb from them. Organo-amines have been used to render silicate surfaces hydrophobic, bonding probably through strong coordinate bonds. These relatively labile coatings will help improve dispersion in the initial stages of a compounding operation, before they desorb. They also give a protective coating to the filler, minimising any polymer degradation that may occur before stabilisers, antioxidants, etc., are fully dispersed.

1.3.2 Coupling Agents

Surface treatments with bi-functional additives, which form very strong covalent bonds to the filler and then bond to a polymer by a variety of mechanisms, are widely available. They are based on organo-metallic compounds with the general formula:

(R I ) a M(R II ) b

(1.4)

where M is a metal ion with valency a + b; R I is an organic group, the choice of which depends on the polymer in which it is to be used and R II is a group designed to react with a mineral surface.

The most commonly encountered compounds are organo-silanes [30] and organo- titanates [31], but others, including organo-zirconates [32], organo-borates [33] and complexes of chromium [34] have been proposed. R II in most of the coupling agents is an alkoxy group (methoxy, ethoxy or 2-methoxyethoxy are common), which is reactive to hydroxyl groups and to water. The preferred reaction mechanism is for hydrolysis to occur, initially by reaction with environmental or adsorbed water, with the kinetics being affected by the nature of both the alkoxy and other reactive groups on the molecule. The organo-metallic hydroxide then condenses with hydroxyls on the mineral surface or it can form oligomers followed by polymers by self-condensation reactions [35].

Multiple layers of many coupling agents can be adsorbed onto a mineral surface, with the packing of the coupling reagent on the surface being determined by the size of the R I organic group [36]. The structure of these layers can be very complex, with both strongly and weakly adsorbed species being present. This is discussed in detail in Chapter 4.

1.3.3 Polymer Modifications

To try to improve energy absorption at the interface, and hence improve composite toughness, modification of the filler with a polymeric, low modulus, interlayer has

Particulate-Filled Polymer Composites

been reported [37]. There have been many papers published on this technique. Aivazyan and co-workers [38] review the surface modification of mineral fillers by polymer deposition, using polymer adsorption from solution, mixing of polymer with dispersed filler and hetero-coagulation of latex polymer particles on the filler. Good bonds between coating and filler are often achieved by grafting a reactive group on to the interfacial polymer before or during filler treatment. Thus, kaolin can be coated with a styrene- diacetone acrylamide copolymer by dispersing the two monomers with the clay in an aqueous medium and polymerising with a persulfate catalyst [39].

Unsaturated monomers have been adsorbed on to fillers and then polymerised to give

encapsulated products. The modulus of the polymer could be modified by selecting the monomer [40]. Acrylic acid-vinyl chloride (1:99) has been polymerised on to calcium carbonate [41]; reports of the use of 3,5-triacryloxyhexahydro-S-triazine [42], bis-phenol

A and epichlorohydrin [43], methyl methacrylate [44], and acrylic acid [45] have also

been published.

Graft copolymers between unsaturated acids, especially acrylic acid and maleic anhydride (MA), and polyolefins (PE and PP) are widely used as surface modifiers and compatibilisers, sometimes in combination with bi-functional coupling agents [46], for talc, calcium carbonate and calcined clays. Such polymer coatings include polypropylene-maleic anhydride [47], polypropylene cis-4-cyclohexene-1,2 dicarboxylic acid [48], polystearyl or polylauryl acrylate [49], polypropylene-acrylic acid, partially oxidised poly(butane diol) [50] and ethylene-vinyl acetate copolymers [51]. Acid- containing products can react with basic fillers. With most other types, they will simply adsorb on to the mineral surface, but they can form esters with some non-basic metal hydroxyls, notably silanols.

1.3.4 Direct Bonding

Polymers themselves can be grafted on to a mineral surface and the resulting composites have been reported to be significantly better than simple mixes because of strong covalent polymer-filler bonds. Composites of clay with ethylene-MA graft copolymer, in which the anhydride groups interact with the hydroxyl groups on the surface of the clay, give increased impact strength and flexural modulus compared with a physical blend of clay and ethylene-MA random copolymers [52]. Polymerisation of monomers can be induced on the mineral by natural catalytic sites

on its surface [53] or by adsorbing polymerisation catalysts on to the mineral prior

to treatment with monomers [54, 55].

General Principles Guiding Selection and Use of Particulate Materials

1.4 Particle Packing and the Maximum Packing Fraction

1.4.1 Introduction

The packing behaviour of particles in a polymer matrix determines at what loading particle/particle effects become important and is a critical factor in the understanding and design of polymer composites, especially when highly filled systems are involved.

The maximum packing fraction, Pf, is a particularly useful concept. This is the maximum volume fraction of particulate that can be incorporated before a continuous network is developed and voids begin to appear in the composite. In addition to indicating the maximum practical loadings obtainable, Pf is also useful in understanding and describing the effect of filler loading on composite properties. Many properties change rapidly as one approaches Pf, almost in a percolation way, and this parameter is now incorporated into many mathematical treatments of property-loading dependence [56]. Indeed, it often goes a long way to explaining why different physical forms of the same filler material can give markedly differing results at the same loading.

1.4.2 Determination of Maximum Packing Fraction (P f ) by Oil

Absorption Procedures

Despite its utility, Pf is difficult to measure or predict with great precision. This is due to complex interactions between the particles and the polymer matrix, and the influence of the fabrication method used. Both of these factors affect the way the particles pack and make conventional measurements of packing, such as the tap density of the powder, of limited quantitative value. This is especially true for many typical particulate fillers, which, due to their fine size and irregular shapes, exhibit particle agglomeration and do not pack well in the dry state. Such powders often pack considerably better when wetted by suitable liquids, especially if dispersants are present. These wetting and dispersing effects lie behind the pragmatic approach of using oil absorption procedures to determine an effective Pf. Originally introduced for application in putty and paint technology, oil absorption methods rely on titrating a sample of powder with an oil or other liquid, while continually rubbing and mixing the mass. End-points are readily detected in these procedures, often defined as the point at which a putty of a certain consistency is obtained. It is generally believed that at this point all the particles are dispersed and wetted, and all the gaps between the particles are just filled with liquid. In practice, only two liquids are used to any extent, linseed oil and phthalate esters. Both simple manual and instrumental methods are available for determining oil absorption values. It is important that standard procedures are closely followed if reproducible results are to be obtained and a number of standards have been issued. The simplest procedure is the spatula rub-out method, which is embodied in ASTM D281 [57]. A useful description of its application to particulate fillers has been given by Ferrigno [58].

Particulate-Filled Polymer Composites

Oil-absorption results are usually quoted as cubic centimetres of oil per 100 grams of particulate at the end point. Simple mathematical treatment using oil and particle densities can then convert this to a volume fraction. As a guide, linseed oil absorptions for calcium carbonate of 20 cm 3 100 g -1 and 50 cm 3 100 g -1 correspond to filler volume fractions of 0.64 and 0.41, respectively.

The attraction of the oil absorption methods lies in their simplicity and their crude approximation to polymer processing. Different liquids are also observed to give markedly different values and this is believed to be due to their respective abilities to wet out and disperse the particles. Thus in theory, by choosing a liquid of similar properties to the polymer in question, one should be able to improve the usefulness of the data, although this does not seem to have been greatly used to date. As mentioned previously, only two liquids are commonly used and, of these, linseed oil generally gives lower values suggesting that it wets and disperses particles better than phthalates. The phthalates are popular because they are widely used as plasticisers in PVC compounds. Liquid paraffins have been used as model systems for olefinic polymers, and squalane could be used as a model system for PP.

A detailed study of the oil absorption test and its meaning in terms of particle packing has been reported by Huisman [59]. He found that the agglomerates present in the original powder are frequently not fully dispersed during the test, but may be compressed and reduced in porosity. The degree to which this occurs critically affects the packing and hence the oil-absorption value. Differences in the degree of compaction obtained go a long way to explaining variation in oil-absorption values between different operators, and between manual and instrumental methods, as well as between different liquids. While oil absorption serves a useful purpose as a guide to the effective value of Pf, great care must be taken in applying it too literally. In some polymer systems there may be a good match with the polymer wetting properties and the dispersion procedures. In others this may be very poor. Oil absorption values are widely used for assessing carbon black properties in the rubber industry. In this field, special procedures have been developed to attempt to reproduce the effect of rubber compounding conditions on particle morphology [22].

1.4.3 Particle Packing Theory

Much effort has been devoted to studying the effect of particle size and shape on the packing properties of powders. Special attention has been given to identifying those factors that allow high packing fractions to be obtained. This subject is very relevant to the design of filler systems and the general principles, which often cause considerable confusion, are outlined here. For greater detail specialist publications such as that by German should be consulted [60].

General Principles Guiding Selection and Use of Particulate Materials

The packing behaviour of particulate materials depends largely on their particle size, shape and surface characteristics. The behaviour of model systems with closely defined size and shape distributions is now well understood. Real particulate materials are much harder to treat, largely due to the difficulty in determining and describing their size and shape distributions accurately. Nevertheless, the principles derived for the model systems can be applied in a semi-quantitative way and appear to work reasonably well.

The aims of particle packing theory are to predict the maximum volume fraction that can be obtained under a given set of circumstances and the structure of the particle assembly at this point. The first point to appreciate is that there are several ways of

packing the same collection of particles, each one resulting in a different maximum volume fraction. The circumstances under which the particles are assembled will determine which of these structures is obtained in practice. In the extreme case consider a three-dimensional jigsaw puzzle. With care this can be assembled to give perfect packing with no free space. This is known as ordered packing. However, if the pieces are put in a container and shaken, it is extremely unlikely that they will ever discover this perfect packed structure. Nevertheless, it will be found that, providing there are enough particles, a certain packed density can be reproducibly obtained implying that some statistically balanced structure

is obtainable. This is known as a random packing situation. More than one ordered and

random packing fraction may be observed with many systems.

Appreciation of the nature of ordered and random packing is important in understanding

the application of packing theory. Ordered packing results in long-range structure as in

a crystal lattice and, although it can sometimes be observed in suspensions of mono-

disperse spherical particles, is not of great importance in our context. As its name implies,

random packing is a statistical process with no long-range structure and it is more relevant to particulate-filled composites. While some ordered packing configurations can be of very low density, ordered packing will always be capable of giving higher packing densities than random packing.

Two classes of random packing are recognised, loose and dense random packing. Loose packing refers to the sort of packing obtained when particles are randomly assembled under conditions where they cannot easily move past each other, while in dense random packing, conditions are such that movement is possible. In powder technology terms, loose random packing corresponds with a pour density and dense random packing to a tapped density.

The simplest case to consider first is the packing of smooth, regular, mono-sized particles. This has been well studied for smooth spherical particles, which can readily move past one another. With these particles a maximum ordered packing fraction of 0.74 has been established, although other, less dense packings are feasible. Random loose packing fraction is difficult to predict accurately but is about 0.60. Random dense packing is more readily predicted and is about 0.64.

Particulate-Filled Polymer Composites

The theory has been extended quantitatively to a number of regular, non-spherical shapes. Many of these such as cubes, rectangles and plates can have ordered packing fractions of unity but most have less dense random packing than spheres, largely due to the difficulty of the particles moving with respect to one another to optimise packing. Cubic particles do, however, seem to have a slightly higher random packing density than spheres (about 0.75). Quantification of the packing behaviour of irregular particles is poorly developed at present but the enhanced interparticle friction soon causes the packing fraction to decrease as the particles depart from sphericity. Values of 0.50 are not uncommon for three-dimensional irregular particles. Very low packing fractions (<0.1) are obtained with high aspect ratio fibres.

The general approach to obtaining high packing fraction powders is to use small particles to fill in the pores in the packed structure obtained from large particles and then to use even smaller particles to fill in the remaining pores and so on. A simple view of this is presented in Figure 1.6. The approach is known as multimodal packing, and much effort has been devoted to defining the optimum particle size ratios and quantities of the various modes and the resulting packing fraction obtained.

The case of spherical particles provides the simplest example. As we have already seen, the maximum ordered packing fraction is 0.74. From geometric principles it can be shown that this structure has two sizes of pores present, which can be filled by spheres with 22.5% and 41.4% of the diameter of the largest spheres. This results in an increase in packing fraction to 0.81.

Unfortunately, such simple geometric calculations are not possible for the more relevant random packing situation, even when spherical particles are involved. This is because there will be a random distribution of pore sizes and the small particles will not spread out evenly among them. In particular, the small particles show a tendency to cluster around the

In particular, the small particles show a tendency to cluster around the Figure 1.6 Classical bimodal

Figure 1.6 Classical bimodal packing effect

General Principles Guiding Selection and Use of Particulate Materials

surface of the larger ones, increasing their spacing [61]. It can be shown, however, that the limiting case for random packing of a bimodal distribution of spherical particles of a single material is achieved at a packing fraction of 0.86. This is obtained with 73% large and 27% small particles and a size ratio of at least 7:1 and preferably up to 20:1 [62]. As with ordered packing, increasing the number of modes improves the packing, but similar ratios are needed for each mode. This soon results in an excessively wide particle size distribution accompanied by ever decreasing benefits. Three modes give a limiting random packed fraction of 0.91 and the ultimate limiting value appears to be 0.96.

The packing behaviour of multimodal mixtures of irregular particles is very poorly understood at present, although some empirical studies have been made on simple systems. Where different sizes of similar shapes are mixed, then improvements in packing similar to those observed with spheres are usually obtained but starting from a lower level. It also appears that a wide difference in sizes and a large proportion (by weight) of the bigger particles is advantageous. Where different shapes are involved then the picture is very unclear and the packing density may even decrease.

While the multimodal approach is theoretically attractive and is useful for understanding the general principles, it has some practical limitations. Continuous particle size distributions are often more readily achieved and less costly and so more relevant. Much work has gone into trying to predict the optimum shape of distribution for maximum packing. This is a very complex subject but it has been shown that packing fractions as good as those obtained from the multimodal approach can be obtained from carefully controlled continuous particle size distributions. These distributions are hard to define and are usually developed empirically. They can, however, be looked upon as being derived from multimodal principles. In general, such distributions are skewed having an abundance by weight of large particles and a long tail to the fine end.

1.4.4 Applications of Packing Principles to Particulate Filled Composites

The packing principles outlined in the previous section are of great value in preparing filler size distributions capable of being processed at high loadings in polymeric systems. In the extreme case monomodal spherical particles can be specially synthesised and mixed together but this is only worthwhile in a few high value applications, such as light curable dental resins. More often adjustments are made to the natural particle size distribution of standard fillers composed of irregular particles to improve the packing. These adjustments may be by classification to remove certain fractions or by blending in extra material in specific size ranges. In some instances narrow sieve fractions are prepared and recombined in set proportions. Where comminution is used to produce fillers then the processing can be controlled to some extent to produce favourable size distributions. Practical procedures for producing maximum packing systems were originally worked

Particulate-Filled Polymer Composites

out by Furnas [63] and have recently been described in some detail by Ferrigno [64]. It must be borne in mind, however, that other considerations may restrict the particle size range that can be tolerated in a composite.

While particle distributions prepared according to packing principles generally give the highest processable loadings in polymers, the prediction of these loadings and the measurement of the effective maximum packing fraction, is not very easy. As mentioned earlier a number of different random packing fractions can exist for the same collection of particles depending on the assembly conditions. Particle packing is usually determined on a powder, by procedures such as pour or tap density, or by some compression process. These are not very relevant as the particles used as fillers tend to be relatively fine and irregular and do not pack well under these conditions. When processed into polymers, the matrix itself, or additives, together with the compounding procedures, can drastically alter particle packing behaviour. Different types of packing are also relevant to different processing conditions. Where flow is required then some sort of random loose packing is probably the maximum achievable, whereas if the dispersion is subsequently compression moulded with squeezing out of excess matrix, then a denser random packing becomes more relevant. Where particles can readily move with respect to one another and a dense packed arrangement is readily obtained, then the phenomenon of dilatancy, or shear thickening, may be observed. This is caused by the need for the particles to move to a more open structure before flow can occur. In such cases processing difficulties may rule out use of compositions based on the dense packing configuration.

As has been discussed earlier, procedures such as oil absorption, although imperfect, probably remain the best way of determining a relevant Pf.

1.5 Interparticle Spacing

The distance between filler particles in a polymer matrix is an important consideration, which often does not receive the attention it deserves. It can be appreciated that composite properties may be markedly affected when this distance becomes very small. This is partly because the polymer chains may not now have the full freedom of movement they would have in the bulk polymer. Equally importantly, if there is a shell of modified polymer adsorbed on the particles, then at some limiting spacing this shell will overlap with that on adjacent particles leading to a percolation effect and to this phase dominating some properties (Figure 1.7). It can also be shown that the critical distance of separation is likely to be a small number of radii of gyration of polymer chains (20-200 nm). Interparticle spacing is directly determined by the number of particles present and this changes markedly with particle size. This is dramatically illustrated in Figure 1.8 by the micrographs of two poly(methylmethacrylate) composites containing equal volume fractions of aluminium hydroxide of differing particle size.

General Principles Guiding Selection and Use of Particulate Materials

Guiding Selection and Use of Particulate Materials Figure 1.7 Idealised view of the effect of interparticle

Figure 1.7 Idealised view of the effect of interparticle distance and of absorbed polymer on composite structure and properties

of absorbed polymer on composite structure and properties Figure 1.8 Scanning electron microscope images of
of absorbed polymer on composite structure and properties Figure 1.8 Scanning electron microscope images of

Figure 1.8 Scanning electron microscope images of cross-sections of poly(methylmethacrylate) filled with equal volume fractions of two different particle size aluminium hydroxide fillers: a) 55% coarse particles, b) 55% fine particles

Particulate-Filled Polymer Composites

1.6 Particle Effects on the Structure of Polymers

1.6.1 Introduction

There is now considerable evidence that particulate fillers, especially inorganics, can significantly affect the structure of the matrix polymer itself, and hence the properties of the final composite. Such effects can vary enormously from system to system and have largely been overlooked in many studies of filled composites. The existence of these effects may explain many of the anomalies reported in trying to develop and apply generalised predictive equations [65].

Matrix modifications can occur in a number of ways and the effect on properties can sometimes be at least as important as that arising from the much more widely recognised interfacial bonding effects. Unfortunately, no systematic study of this area has yet been made.

The principal ways in which a particulate phase may affect polymer structure are:

1. Molecular weight reduction during processing.

2. Molecular weight and crosslink modifications due to interference with curing processes.

3. Adsorption of polar, low molecular weight species such as surfactants, plasticisers, stabilisers and antioxidants and oxidation products.

4. Formation of an immobilised shell of polymer around the particles. Often this shell is rich in a certain molecular weight fraction, which is then depleted in the matrix.

5. Modification of crystallinity in semi-crystalline polymers.

6. Effects on polymer conformation due to particle surfaces and inter-particle spacing.

A brief description of the various effects together with examples of their importance is

given in the next section.

1.6.2 Molecular Weight Reduction During Processing

It is well known that polymers may undergo mechanochemical degradation during

compounding and moulding operations [66]. Indeed this is an important feature of natural rubber processing. As most properties are dependent on molecular weight, this can play an important role in determining final properties and conditions (including the use of stabilisers) are usually chosen to limit any degradation.

General Principles Guiding Selection and Use of Particulate Materials

As has already been discussed, and will be returned to later, the introduction of a particulate phase may markedly affect the amount of degradation. The usual effect is for degradation to be increased due to increased mechanical work, to chemical interactions at the interface and possibly to deactivation of stabilisers and antioxidants. The effect is most pronounced under conditions (especially high loadings) where viscosities are high, leading to excessive shear forces being developed. Such effects have been reported in filled elastomers [67] and polypropylene [68].

1.6.3 Molecular Weight and Crosslinking Changes due to Cure Modifications

The detailed microstructure of free-radical cured polymers determines their overall properties, and is governed by the rate of cure and relative rates of various competitive polymerisation processes such as: initiation, propagation, combination and termination.

Inorganic particles can modify the cure process in several ways. In particular, by lowering the overall exotherm, they can significantly reduce the cure temperature and kinetics, thus altering the relative rates of the various processes described above. The surface of the particles can also play a role by absorbing and altering the stability of the growing polymers, or by absorbing and altering the activity of the initiator and curative. Most of these effects have been reported but, unfortunately, structure determination is very difficult with crosslinked systems and little or no direct evidence for resulting structure changes has been reported. An example of the effect of different filler particles on the curing of unsaturated polyesters (and by inference on their structure), as measured by the cure exotherm, has been given by Plueddemann and Stark [69].

1.6.4 Preferential Adsorption of Polar Species

Polar materials are usually more strongly adsorbed on particulates than is the matrix polymer and, where present, can displace the polymer altogether from the surface. This can modify the composite properties in two ways. Firstly, it may lead to a weak interfacial bond. Secondly, where the polar species was providing a significant matrix function such as plasticisation, cure modification or flow enhancement, this can be lost.

Good examples of the preferential adsorption of polar materials have been given by Barnett and Jones [70], Evans and co-workers [71] and by Karlivan and co-workers [72]. Barnett and Jones carried out some elegant and painstaking work on the effect of mineral fillers on the reinforcement of emulsion-polymerised SBR. This type of elastomer contains high levels of organic surfactants remaining from the polymerisation process. Certain mineral fillers gave exceptional and unexpected reinforcement properties, which were

Particulate-Filled Polymer Composites

demonstrated to be due to the preferential adsorption of these surfactants, resulting in a significant improvement in the strength of the matrix. This strength improvement was, at least in part, due to modification of the curing process showing how complex and inter-related the effects of adding particulates can sometimes be. More recently, Evans and Rothon have shown how fatty acids prevent the formation of bound rubber when precipitated calcium carbonate is used as a filler in SBR elastomers. When they are absent, high levels of bound rubber are formed, accompanied by a marked increase in reinforcement. Karlivan demonstrated how certain inorganic fillers could be used to preferentially adsorb low molecular weight polar species produced by thermo-oxidation of polyolefins. These species tend to form during heat welding of polyolefins to metals and, by accumulating at the interface, limit the adhesive strength. Their removal markedly improved adhesion.

1.6.5 Formation of an Interphase of Immobilised Polymer

It is now well established that, in the absence of competing polar species, molecules of the matrix polymer can be effectively adsorbed on to the surface of a particulate phase, even when this is an inorganic mineral. The strength of this adsorption can play a key role in determining composite properties as will be discussed in later chapters.

It is also generally recognised that this adsorbed polymer may have different properties from the bulk, but the consequences of this for composite properties are often overlooked. This is partly because, as we shall see, the effects only become significant at high loadings and with small particles, but also because the adsorbed layers are hard to detect and characterise. Where they are observed, the effects can be at least as great as those associated with particle matrix bonding. The principal effect of this adsorbed layer is to increase the effective size and hence volume fraction of the particulate phase. A complication that has caused much confusion in the past, and is only just becoming recognised and understood, is the apparent variability in the thickness of the immobilised layer according to the method of measurement and the composite property being studied [73].

Once this difficulty is appreciated, its origins can be clearly understood. Thus, in the absence of other adsorbing species, there will be a primary layer of strongly adsorbed polymer at the filler surface. The conformation and properties of this are likely to be severely modified and probably in all instances it can be regarded as immobilised and contributing to particle volume. As one moves beyond this layer, the effects due to the particle do not suddenly disappear but will gradually tail away. This outer region will be less strongly associated with the particle and may only appear to be immobilised under certain test conditions.

General Principles Guiding Selection and Use of Particulate Materials

As we shall see, while effects of particles can sometimes extend for large distances, the immobilised layer thickness generally appears to range from 1 to 50 nm and in many instances is much less than 10 nm. The effect of various layer thicknesses on the effective volume of different sized cubic particles is presented in Table 1.4. Only when the layer is of the order of 50 nm does it have much effect on particles over 1 μm in size, but even thin layers of 10 nm or less have a significant effect on particles about 0.5 μm, and this becomes dramatic as very small particles (0.1 μm) are reached. Because of their increased surface to volume ratio, anisotropic particles such as plates will show greater effects than those recorded in Table 1.4. Determination of the thickness of immobilised layers is far from straightforward and can give a variety of answers, as mentioned earlier.

Table 1.4 Calculated effect of surface layers on the apparent volume of various sized cubic particles

Particle size (μm)

% volume increase for a layer thickness of:

1 nm

10 nm

100 nm

1000 nm

0.1

 

637

2600

-

0.5

1.2

12

174

-

1.0

 

063.6

7

2600

5.0

< 0.1

1.2

12

174

10.0

< 0.1

0.6

6

73

The adsorption and conformation of polymer molecules on particle surfaces is of great interest to colloid scientists, especially with regard to steric stabilisation of dispersions, and general guidelines can be obtained from their studies. A good description of the subject can be found in treatise such as that by Napper [74]. It is found that single point adsorption results in the molecule extending preferentially in a direction perpendicular to the surface and a layer thickness that approaches a value of twice the radius of gyration of the free polymer. The radius of gyration is dependent on molecular weight but for our purposes the effective layer thickness due to this type of conformation will be in the range of 20-40 nm. With most polymers of interest to us, multipoint attachment is more likely. This will significantly reduce the layer thickness until, in the limit of flat adsorption, this can be as low as 1 nm or less.

The most direct method of determining the thickness of adsorbed polymer is by carrying out adsorption studies, either from a suitable solution of the polymer, or by extracting an actual composite and determining the residual or ‘bound’ polymer. Both approaches have

Particulate-Filled Polymer Composites

been used and the layer thickness is calculated from the volume of adsorbed polymer by dividing by the surface area of the particulate phase. The bound polymer approach has been frequently used in studies on carbon black-filled elastomers and several useful reviews have been published [75]. Unfortunately, the origin of bound polymer is complex and may be of limited use in characterising effects in the composite itself. Bound layer thicknesses determined in this way for carbon black-elastomer systems are usually of the order of 1-5 nm. There has been only limited work on determining bound polymer with mineral fillers and in polymers other than elastomers. Several studies have shown that bound rubber readily forms with high surface area silicas in elastomers [76], where calculated layer thicknesses are similar to those reported for carbon blacks. Evans and Rothon have reported on bound rubber formation with precipitated calcium carbonate in an elastomer system [71] and again the calculated layer thickness based on their data is less than 5 nm.

Similar low values have been reported by Kosfeld and co-workers for kaolin and zinc sulfide [77] in a rubber-modified polypropylene and by Lipatov [78] and Boonstra [79] for silica-filled polyethylene. It is probable that these determinations only detect the polymer molecules that are strongly attached to the particles, probably in multipoint fashion. This explains the low values of thickness observed, but it does not necessarily follow that the immobilised, or modified thickness in composites is not significantly greater than this.

In addition to bound polymer of the type described previously, polymer may also be shielded from some deformation processes by being occluded within the particle shape (See Section 2.8.2) and this is especially important with some carbon blacks.

Viscosity measurements provide another fairly simple method for determining a layer thickness. For non-agglomerated particles, a fairly simple equation (Equation 1.5) exists relating relative viscosity of a suspension to the effective volume fraction of the solid phase:

η s η = 1 + K Φ + K´ Φ + …

(1.5)

Where η s and η are the viscosities of the suspension and the solvent, respectively, K is a coefficient depending on particle size, shape and concentration, and Φ is the effective volume concentration. K can be determined in pure solvent without polymer being added and then used to determine the change in Φ resulting from addition of polymer.

Viscometric determinations of this type tend to give quite high layer thicknesses of the order of several tens of nanometres. An example of this approach can be found in the work of Chibowski who studied the adsorption of polyacrylamide of different molecular weights on to the surface of titanium dioxide from aqueous solution [80]. He found effective layer thicknesses of 30-70 nm increasing with polymer molecular weight. The results obtained are, of course, influenced by the solvent. If possible, the measurements

General Principles Guiding Selection and Use of Particulate Materials

should be carried out in a theta solvent, (i.e., one in which the dimensions of the polymer coil are the same as in the melt).

Various procedures have been applied to composite materials to attempt to determine the amount and properties of the immobilised layer in situ. Much of the interest in this has been in the former eastern block countries of Europe and has been reviewed by Enikolopyan [81]. Among the measurements used have been density, specific heat, dielectric constant and dynamic mechanical analysis. The general conclusions are that, as one might expect from reduced configurational entropy, the surface layers are denser and more rigid than the bulk polymer.

Again most work has been carried out on carbon black-filled elastomers. This has shown that a slight broadening can be observed in the glass transition of the polymer, which is interpreted as being that of the surface layers. Deconvolution indicates that the glass transition of these layers is about 10 °C higher than the matrix and the calculated volume of modified material approximates to a 3 nm layer [82].

Only isolated studies have been reported in other systems. Galeski has carried out thermal analysis of chalk-filled isotactic polypropylene [83]. Uncoated chalk was found to modify all the transitions present in the unfilled polymer and to lead to two new transitions assigned to restricted motion in the interface layers. These effects were largely eliminated when the chalk was surface treated, confirming their surface origin. No attempt was made to calculate layer thickness. Gerard and co-workers have reported on the effect of glass beads in a crosslinked epoxy matrix [84]. They used dynamic mechanical analysis and found the uncoated beads to reduce the mobility of the polymer, an effect that could be largely overcome by coating the beads. Analysis of this sort has much to offer and hopefully will be more widely applied in future work.

Workers such as Pukansky and Tudos have developed procedures for calculating effective layer thicknesses from the mechanical properties of composites [85]. They have applied this to calcium carbonate-filled polypropylene and found the effective layer thickness to vary according to the property being measured. A value of 5 nm was obtained from Young’s moduli in agreement with the thicknesses expected for strongly bound polymer.

A

value of 150 nm was, however, calculated from tensile yield and strength calculations.

It

was postulated that the size of the effective interlayer depends on the elongation at

which it is being determined.

The aspect of bound or modified polymer still has many secrets to reveal and its existence and importance should always be assessed when new polymer/filler combinations are being developed, especially with high filler loadings and high surface area particles.

Particulate-Filled Polymer Composites

1.6.6 Effects on Polymer Conformation due to the Presence of Particle

Surfaces and Interparticle Spacing

The properties of polymers depend to a large degree on the conformations available to their chains. The presence of solid, particulate material can significantly modify these conformations, even in the absence of any particle-polymer interaction or other effects such as already discussed.

Firstly, the presence of a solid surface will restrict the conformation of polymer molecules in its vicinity. This type of effect is widely recognised in colloid science where, as already mentioned, it has been extensively studied with regard to steric stabilisation of suspensions. Detailed treatment can be found in works such as those of Napper [74]. The basic findings are that, in the absence of multipoint attachment, chain molecules respond to this restriction by expanding in a direction parallel to the surface. This expansion can double the effective molecular size in that direction.

Secondly, polymer molecules occupy quite considerable volumes, depending on their molecular weight and temperature. At high filler loadings, especially with ultrafine particles, the interparticle spacing may be of similar size or smaller. This must mean that the polymer molecules will be forced to take up a strained conformation and thereby be considerably modified in their properties.

While these effects are undoubtedly present in particulate composites they have not, to the authors’ knowledge, been addressed in any detail so that their significance is still largely unknown.

1.6.7 Effects on Crystallinity

There is sufficient evidence available to show that particulate fillers can affect the crystallinity of some polymers and that this may in turn affect mechanical properties. Despite this, the area has received surprisingly little attention and many studies of filler/ property effects in semi-crystalline polymers do not consider this possibility.

Many thermoplastic materials are semi-crystalline and the precise nature of this crystallinity plays a major role in determining their properties. In addition to the level of crystallinity, factors such as crystal phase, crystallite numbers, orientation, size and shape are all important, but the exact way in which they affect properties is far from understood.

In unfilled polymers the crystallites generally seem to be present in spherulitic aggregates [86]. Because of hindered motion, thermoplastics have a marked tendency to super cool before crystallising. The degree of super cooling will itself determine the number of

General Principles Guiding Selection and Use of Particulate Materials

crystallites and will vary according to forming conditions. As a consequence of this super cooling, thermoplastics are very prone to nucleating effects and specially chosen nucleating agents are frequently added to ensure rapid crystallisation during forming. Among other things, this reduces cycle times. In some cases, nucleants may be used to favour certain crystal phases.

Such nucleating agents are often salts of organic acids and sometimes fine fillers. Their use does not generally alter the spherulitic nature of the crystallinity, only the size and number of crystallites. Even so, there can be marked changes in properties accompanying the use of such nucleants, examples of which are given by Jansen [87]. In particular their use generally seems to increase impact strength due to the reduction in crystallite size.

Given the importance of crystallinity in determining mechanical and other properties and the susceptibility of polymers to nucleation, it is pertinent to enquire whether normal filler particles have any significant effects in this context. This area is surprisingly little studied, but there is sufficient evidence to show that marked nucleation can occur, although this may vary from filler to filler, with surface treatment also playing a role. There are also indications that the type of crystallinity may be affected and thtat there is some circumstantial evidence for a link between nucleation effects and mechanical properties. This area is, however, not yet well understood and we are a long way from having a good generalised theory.

The subject is probably best advanced for polymers in contact with glass fibres and certain platy particles such as sheet silicates. As well as acting as nucleants, this type of material can alter the type of crystal structure present. Thus, it is now well established that glass-fibre surfaces can lead to trans-crystallinity. This is the growth of radially oriented lamellae extending for a few hundred micrometres from the fibre surface. This topic has been discussed in detail by Thomason [88]. Transcrystallinity is also discussed in Chapter 8 of this work.

Outside of these areas the picture is much less clear. Straight nucleation affects are easily detected by thermal analysis techniques. These generally manifest themselves in an earlier onset temperature and peak temperature, but not necessarily in any increase in total crystallinity. There are many reports of significant effects for various polymer/filler systems. Thus, talc has been reported as nucleating PE [89], PP [90] and Nylon 6 [89]. Calcium carbonate has been reported as nucleating PP [91-93] and PE [94], while sepiolite (Mg 4 Si 6

O 15 (OH) 2 . 6H 2 O) has been reported as nucleating PP [95]. Surface treatment of the filler

particles has also often been found to reduce or eliminate these nucleating effects [89, 96, 97]. One of the authors (RR), has observed similar nucleating effects for aluminium and

magnesium hydroxide in polypropylene and also confirmed the reduction of this effect by surface treatment. Typical results illustrating these effects are presented in Figure 1.9. What

is far less clear is whether and how these nucleation effects alter the crystal structures

present and whether, like organic nucleants, they alter mechanical properties.

Particulate-Filled Polymer Composites

Particulate-Filled Polymer Composites Figure 1.9 Effect of magnesium hydroxide fillers on the crystallisation of a

Figure 1.9 Effect of magnesium hydroxide fillers on the crystallisation of a polypropylene copolymer (DSC experiment, cooling at 10 °C/min)

There is some evidence for a link with mechanical properties in the work of Hutley and Darlington [96, 97]. They found a good inverse correlation between the nucleating ability of fillers such as calcium carbonates and falling weight impact strength in PP. Surprisingly, the organic nucleants were also found to decrease the impact strength in line with their nucleating ability, although the opposite effect is usually observed, as discussed earlier. According to these workers, other mechanical properties were little affected. Mitsuishi and co-workers have also studied the effect of calcium carbonate on the nucleation and mechanical properties of PP [91]. They reported significant correlation with the loss modulus, tensile modulus and tensile impact strength. They confirmed the decrease of impact strength with increasing nucleating effect reported by Hutley and Darlington.

Quantifying the effects of the fillers on crystal structure and hence explaining the mechanical property effects described above is very difficult. This is largely because of the problems of carrying out direct observations using traditional microscopy procedures with such highly filled systems. Among the effects that could be looked for are: changes in total level of crystallinity, changes in crystal phase and its orientation, changes in the number and size of crystalline particles, and changes in the type of crystalline particle. In none of these areas is the picture completely clear and the effects obtained probably vary from system to system. One of the more detailed studies in this field has been by McGenity and co-workers on calcium carbonate and talc-filled polypropylene [98].

General Principles Guiding Selection and Use of Particulate Materials

While there are some reports of significant effects on the overall level of crystallinity, most studies show little or no effect [92, 98]. Chacko and co-workers reported that the effect of calcium carbonate on the crystallinity of polyethylene was dependent on molecular weight with reduced crystallinity being observed at low molecular weight and no effect at higher molecular weight [94]. Multiple peaks are often observed in melt cooling studies, (e.g., Figure 1.9) and have been claimed as evidence for the nucleation of new crystal phases [92]. When X-ray studies have been carried out, they have usually failed to confirm this [99], although Murthy did report finding that talc altered the crystal phase present in Nylon 6 [89]. The present author (RR), also found that reheating of magnesium hydroxide-filled polypropylene displaying the twin peaks in Figure 1.9 only gave a single melting peak, suggesting that only one phase was present. On the other hand, fillers have been reported to increase the amounts of beta-phase in polypropylene [98] and to affect the crystal phase present in Nylon 6 [89].

The few studies that have been carried out to date strongly suggest that particulate fillers at normal usage levels markedly disrupt the normal spherulitic morphology. This is not surprising as the interparticle spacing would probably be insufficient to allow these to develop properly. There is also some evidence that new forms of crystallinity can be present associated with the particle surfaces.

Thus Chacko and co-workers [94], report lamellar growths originating from selected spots on calcite surfaces in polyethylenes and radiating outwards for up to 400 μm. Rybnikar reported similar effects for calcite in polypropylene [93]. In their work, Kowalewski and Galeski carried out a detailed study of calcium carbonate in polypropylene including work with large single crystals [92]. They reported that the effects observed depended on the crystal faces present. A form of trans-crystallinity was observed on some crystal faces and epitaxial growth was postulated for others.

In summary, it seems clear that particulate fillers can have significant nucleating effects in semi-crystalline polymers and that this may lead to effects on mechanical properties. Much work remains to be done to clarify this and provide a clear, coherent description of the effects involved, however. This would be greatly helped if simple techniques for determining polymer crystal structures in filled systems were available. It would also be highly desirable to have a better understanding of how the structure of polymer crystallinity affects composite properties.

References

1. P. Challinger, Fillers in Plastics Conference, Filplas 81, British Plastics Federation, London, UK, 1981.

Particulate-Filled Polymer Composites

2. M.E.J. Dekkers and D.J. Heikens, Journal of Materials Science, 1983, 18, 3281.

3. H.P. Schlumf, Polypropylene Compounds — Sustaining the Market Growth, Corporate Developments Consultants Conference, Brussels, Belgium, 1991.

4. H.P. Schreiber and F. St. Germain , Journal of Adhesion Science and Technology , 1990, 4, 4, 319.

5. F.M. Folkes, Proceedings of the SPE Fillers and Additives for Plastics Conference, Gothenberg, Sweden, 1988.

6. K.F. Heinisch, Dictionary of Rubber, Applied Science Publishers Ltd., London, UK, 1974.

7. R.P. Sheldon, Composite Polymeric Materials, Applied Science Publishers, London, UK, 1982.

8. C.F. Bohren and D.R. Huffman, Absorption and Scattering of Light by Small Particles, John Wiley and Sons, New York, NY, USA, 1983.

9. P. Kubelka and F. Munk, Zhurnal Tekhnicheskoi Fiziki , 1931, 12 , 593.

10. M. Kerker, The Scattering of Light and other Electromagnetic Radiation, Academic Press, New York, NY, USA, 1969.

11. G. Mie, Annalen der Physik, 1908, 25, 325.

12. P.N. Dunlap and S.E. Howe, Polymer Composites, 1991, 12, 1, 39.

13. M. Hancock, Developments in PVC Symposium, Loughborough, UK, 1986.

14. T.H. Ferrigno in Handbook of Fillers for Plastics, 2nd Edition, Eds., H.S. Katz and J.V. Milewski, Van Nostrand Reinhold, New York, NY, USA, 1987.

15. A.A. Berlin, S.A. Volfson, N.S. Enikolopian and S.S. Negmatov, Principles of Polymer Composites, Springer-Verlag, Berlin, Germany, 1991.

16. A.I. Medalia and F.A. Heckman, Carbon, 1969, 7, 567.

17. W.M. Hess, G.C. McDonald and E.M. Urban, Rubber Chemistry and Technology, 1973, 46, 204.

18. B.H. Kaye, A Random Walk through Fractal Dimensions, VCH Publishers, New York, NY, USA, 1989.

General Principles Guiding Selection and Use of Particulate Materials

19. T.W. Zerda, H. Yang and M. Gerspacher, Rubber Chemistry and Technology, 1992, 65, 130.

20. C.R. Herd, G.C. McDonald and W.M. Hess, Rubber Chemistry and Technology, 1992, 65, 107.

21. H. Barthel, F. Achenbach and H. Maginet, Proceedings of Moffis-93, Namur, Belgium, 1993, p.301.

22. R.E. Dollinger, R.H. Kallenberger and M.L. Studebaker, Proceedings of the 90th ACS Rubber Division Meeting, New York, NY, USA, Fall, 1966, Paper No.19.

23. G. Fourty, Proceedings of Moffis-91, Conference on Mineral and Organic Functional Fillers in Plastics, Le Mans, France, 1991, p.59.

24. L.F. Gate and T.W. Webb, inventors; ECC international Ltd., assignee; GB 2,240,398, 1994.

25. T. Allen, Particle Size Measurement, 3rd Edition, Chapman and Hall, London, UK, 1981.

26. A.N. Gent, Journal of Materials Science, 1980, 15, 2884.

27. S. Brunauer, P.H. Emmett and E. Teller, Journal of the American Chemical Society, 1938, 60, 309.

28. Characterisation of Powder Surfaces: with Special Reference to Pigments and Fillers, Eds., G.D. Parfitt and K.S.W. Sing, Academic Press, London, UK, 1976.

29. S.G. Thoma, M.C. Ciftioglu and D.M. Smith, Powder Technology, 1991, 68, 53.

30. R.E. Godlewski, Proceedings of the 40th Annual Conference of the Reinforced Plastics/Composites Institute, Atlanta, GA, USA, 1985, Paper 4-F.

31. S.J. Monte, Modern Plastics Encyclopedia, 1976, 53,10A, 161.

32. L.B. Cohen, Plastics Engineering, 1983, 39, 11, 29.

33. M.M. Fein, B.K. Patnaik and F.K.Y. Chu, inventors; Dart Industries Inc., assignee; US Patent 4,073,766,1978.

34. J.A. Robertson, R.L. Adelman and H.E. Bergna inventors; E I DuPont de Nemours, assignee; US Patent 3,951,680, 1976.

Particulate-Filled Polymer Composites

35. J.D. Miller, K-P. Hoh and H. Ishida, Polymer Composites, 1984, 5, 1, 18.

36. B.D. Favis, L.P. Blanchard, J. Leonard and R.E. Prud’homme, Polymer Composites, 1984, 5, 1, 11.

37. W. Pompe, S. Völlmar and H.J. Weiss, Plaste und Kautschuk , 1974, 21, 9, 664.

38. G.B. Aivazyan and co-workers, Armianskii Khimicheskii Zhurnal, 1990, 43, 3, 186.

39. S.E. Gebura, inventor; Interpace Corporation, assignees; US Patent 3,557,038, 1971.

40. R.L. Kaas and Z.G. Gardlund, Proceedings of the 35th SPE Technical Conference, Montreal, Canada, 1977, p.333.

41. V.A. Popov and co-workers, Doklady Akademii Nauk SSSR, 1984, 275, 5, 1109.

42. F. Nitto, inventor; Japan Kokai Tokyo Koho, 80,110,138, 1980.

43. Shirashi Kogyo Co Ltd, assignee; Japan Kokai Tokyo Koho 81,100,860, 1981.

44. K. Iwaku and J. Fujimura, Journal of Applied Polymer Science, 1979, 24, 4, 975.

45. K. Choi and C.K. Kim, Pollimo, 1989, 13, 5, 426.

46. S. Cartasegna, SPE Symposium, Polymer Grafting in Twin Screw Extrusion, Antwerp, Belgium, 1990.

47. A. Ohhira and M. Tuzawa, inventors; Japan Kokai Tokyo Koho, 77,109,546, 1977.

48. S. Fujii, N. Miyazaki and I. Unno, inventors; Japan Kokai Tokyo Koho, 70,54,034, 1974.

49. K. Senda and co-workers, Japan Kokai Tokyo Koho, 74,07,336, 1974.

50. Chisso Corporation, assignee; Japan Kokai Tokyo Koho, 81,149,452, 1981.

51. N. Iguchi and T. Ono, inventors; Japan Kokai, 75,107,046, 1975.

52. N.G. Gaylord and A. Takahashi, Polymer Science and Technology, Volume 21:

Modification of Polymers, Plenum Press, New York, NY, 1983, 183.

53. M. Litvinenko and co-workers, Zhurnal Fizicheskoi Khimii, 1989, 63, 12, 284.

54. M. Bokhitski and co-workers, Plasticheskie Massy, 1988, 8, 34.

General Principles Guiding Selection and Use of Particulate Materials

55. E.G. Howard, B.L. Glazar and J.W. Collette, Proceedings of the SPE High Performance Plastics, National Technical Conference, 1976, p.36.

56. T.H. Ferrigno in Handbook of Fillers for Plastics, 2nd Edition, Eds., H.S. Katz and J.V. Milewski, Van Nostrand Reinhold, New York, NY, USA, 1987, 33.

57. ASTM D281-95, Standard Test Method for Oil Absorption of Pigments by Spatula Rub-out, 2002.

58. T.H. Ferrigno in Handbook of Fillers for Plastics, 2nd Edition, Eds., H.S. Katz and J.V. Milewski, Van Nostrand Reinhold, New York, NY, USA, 1987, 17.

59. H.F. Huisman, Journal of Coatings Technology, 1984, 56, 712, 65.

60. R.M. German, Particle Packing Characteristics, Metal Powder Industries Federation, Princeton, NJ, USA, 1989.

61. R.M. German, Particle Packing Characteristics, Metal Powder Industries Federation, Princeton, NJ, USA, 1989, p.148.

62. R.M. German, Particle Packing Characteristics, Metal Powder Industries Federation, Princeton, NJ, USA, 1989. Chapter on bimodal mixtures of spherical particles.

63. C.C. Furnas, Industrial and Engineering Chemistry, 1931, 23, 9, 1052.

64. T.H. Ferrigno in Handbook of Fillers for Plastics, 2nd Edition, Eds., H.S. Katz and J.V. Milewski, Van Nostrand Reinhold, New York, NY, USA, 1987, 15.

65. B. Pukanszky, F. Tudos and I. Kelen in Polymer Composites, Ed., B. Sedlacek, Walter de Gruyter, New York, NY, USA, 1987.

66. O.S. Kauder in Thermoplastic Polymer Additives: Theory and Practice, Ed., L.J. Lutz Jr., Marcel Dekker, New York, NY, USA, 1989.

67. B.B. Boonstra and G.L. Taylor, Rubber Chemistry and Technology, 1965, 38, 4, 943.

68. W.V. Titow and B.I. Lanham, Reinforced Plastics, Applied Science Publishers, London, UK, 1975.

69. E.P. Plueddemann and G.L. Stark, Proceedings of the 31st SPI Annual Technical Conference, Washington, DC, USA, 1976, Section 6-D.

70. C.E. Barnett and H.C. Jones, Industrial and Engineering Chemistry, 1949, 41, 7, 1518.

Particulate-Filled Polymer Composites

71. M.B. Evans, R.N. Rothon and T.A. Ryan, Proceedings of Fillers-86, BPF/PRI, London, UK, 1986, Paper No.12.

72. V.P. Karlivan, M.M. Kalnin’, L.Y. Malers, Y.Y. Avolin’sh, Y.Y. Malers and A.V. Viksne, Plasticheskie Massy, 1976, 11, 46.

73. Filled Polymers I, Ed., N.S. Enikolopyan, Advances in Polymer Science No.96, Springer-Verlag, Berlin, Germany, 1990.

74. D.H. Napper, Polymeric Stabilisation of Colloidal Dispersions, Academic Press, London, UK, 1983.

75. B. Pukansky and F. Tudos in Proceedings of the 3rd International Conference on Composite Interfaces (ICCI-III), Cleveland, OH, USA, p.691.

76. M. Ashida, K. Abe and T. Watanabe, Nippon Gomu Kyokaishi , 1976, 49, 11,

821.

77. R. Kosfeld, K. Schaefer, E.A. Hemmer, M. Hess, A Theisen and T.H. Uhlenbrich, Proceedings of the 3rd International Conference on Composite Interfaces (ICCI- III), Cleveland, OH, USA, 1990, p.385.

78. Y.S. Lipatov and F.G. Fabuliak, Vysokomolekulyarne Soedineniya, 1968, 10, 1605.

79. B.B. Boonstra, Polymer, 1979, 20, 691.

80. S. Chibowski, Powder Technology, 1990, 63, 75.

81. N.S. Enikolopyan, Filled Polymers, Advances in Polymer Science, 96, Springer- Verlag, Berlin, Germany, 1990.

82. G. Kraus, Advances in Polymer Science, 1971, 8, 155.

83. A. Galeski in Proceedings of the 3rd International Conference on Composite Interfaces (ICCI-III), Cleveland, Ohio, USA, May 1990, p.623.

84. I.F. Gerard in Proceedings of the 3rd International Conference on Composite Interfaces (ICCI-III), Cleveland, OH, USA, 1990, p.441.

85. B. Pukansky and F. Tudos in Proceedings of the 3rd International Conference on Composite Interfaces (ICCI-III), Cleveland, OH, USA, 1990, p.691.

86. A. Galeski, J. Grebowicz and M. Kryszewski, Die Makromoleculare Chemie, 1983, 184, 1323.

General Principles Guiding Selection and Use of Particulate Materials

87. I.J. Jansen, in Plastics Additives, 2nd Edition, Ed., R. Gachterand and H. Muller, Hanser Publishers, Munich, Germany, 1987.

88. J.L. Thomason and A.A. van Rooyen in Proceedings of the 3rd International Conference on Composite Interfaces (ICCI-III), Cleveland, Ohio, USA, May 1990, p. 423.

89. N.S. Murthy, A.M. Kotliar, J.P. Sibilia and W. Sacks, Journal of Applied Polymer Science, 1986, 31, 2569.

90. I. Menczel and I. Varga, Journal of Thermal Analysis, 1983, 28, 161.

91. K. Mitsuishi, S. Ueno, S. Kodama and H. Kawasaki, Journal of Applied Polymer Science, 1991, 43, 2043.

92. T. Kowalewski and A. Galeski, Journal of Applied Polymer Science , 1986, 32 ,

2919.

93. F. Rybnikar, Journal of Applied Polymer Science , 1991, 42, 2727.

94. V.P. Chacko, F.E. Karasz, R.J. Farris and E.L. Thomas, Journal of Polymer Science: Polymer Physics Edition, 1982, 20, 2177.

95. J.L. Acosta, A. Linares, M.C. Ojeda and E. Morales, Revista de Plasticos Modernos, 1985, 49, 346, 441.

96. T.J. Hutley and M.W. Darlington, Polymer Communications, 25, 226.

97. T.J. Hutley and M.W. Darlington, Polymer Communications, 1985, 26, 264.

98. P.M.McGenity, J.J. Hooper, C.D. Paynter, A.M. Riley, C. Nutbeem, N.J. Elton and J.M. Adams, Polymer, 1992, 33, 5215.

99. L. Jingjiang, W. Xiufen and G. Qipeng, Journal of Applied Polymer Science , 1990, 41, 2829.

Particulate-Filled Polymer Composites

2 Principal Types of Particulate Fillers

Michael Hancock and Roger N. Rothon

2.1 Introduction

One of the fascinations of working in the field of particulate-filled polymer composites is the wide variety of materials used as fillers, from relatively simple chalks and limestones to complex rare-earth magnetic powders. Even apparently, mundane, relatively inexpensive, fillers can vary in subtle, but important, ways according to their precise origin and method of manufacture, thus adding their interest to the scientific and technological mind.

In the space available here it is only possible to cover the major fillers and a few of the less common filler materials, the latter having been selected on the basis of lack of adequate treatment elsewhere and for illustration of important principles.

2.2 Particulate Fillers from Natural Origins (Mineral Fillers)

2.2.1 Introduction

Mineral fillers are a vital and significant part of the world’s polymer industry. Consumption in rubber and plastics is currently estimated to be over 2.5 million tonnes per year in Western Europe. In the first edition of this book growth projections were very bullish and use in plastics especially was projected to grow at 8-12% per year. On a worldwide basis, it was predicted that 20 Mt per year would be used in polymers by the year 2000 [1]. However, local and world economies have suffered several blows since 1997 and this volume has not been reached. As a best ‘guestimate’ the authors believe that filler use in polymers currently is about 15 Mt per year.

The majority of use consists of low-cost products in which price is the predominant specification property. Although this will continue to be an important area for mineral fillers, increasingly sophisticated applications for plastics, rubbers and composites mean that the requirements placed on fillers and the specific properties imparted by them will become more demanding; that is, they will be required to perform as functional additives conferring specific benefits in applications, which will be narrowly defined. It is therefore

Particulate-Filled Polymer Composites

important to understand the intrinsic properties of the minerals, how these affect bulk properties and how these then influence the final filled polymer.

The most important mineral fillers used are carbonates, clays and talcs, while other silicates are also of interest. Several carbonate minerals are known with some having potential for use as mineral fillers, although only a few are of industrial importance in plastic and rubber applications.

Calcite (calcium) and dolomite (calcium-magnesium) are the main carbonate fillers and are very widespread, being exploited in many countries. The other carbonate mineral of any importance is, in fact, a mixture of two carbonates: hydromagnesite (a hydrated basic magnesium carbonate: Mg 5 (CO 3 ) 4 (OH) 2 .4H 2 O) and Huntite (CaCO 3 .3MgCO 3 ). A large, relatively pure deposit is found in Greece and a smaller one in Texas, USA, with the former being exploited as a filler for paint and paper applications, and as a flame- retardant filler in plastics and rubber.

Clay minerals are aluminium silicates of either the two-layered kaolinite type or three- layered montmorillonite type. Only three clay minerals are commonly used in the polymer industries, kaolinite, montmorillonite and chlorite, and these will be discussed below. Chlorite, because it usually occurs with talc and has essentially the same properties, is discussed with that mineral. Talc (magnesium silicate) is widespread but is commonly found with other magnesium minerals such as magnesite.

For a complete analysis of the world’s industrial minerals, the reader is referred to publications such as the Minerals Yearbook published by the US Department of the Interior [2] and The Industrial Minerals Handbook [3].

2.2.2 Minerals and Rocks

Strictly speaking, a mineral is a substance of inorganic origin with definite chemical composition, which is found in the earth’s crust. Rocks are naturally occuring mixtures of minerals. With few exceptions, minerals possess an internal ordered arrangement of their constituent atoms or groups of atoms, that is, they are crystalline. Some lack any ordered internal structure and are said to be amorphous.

The arrangements (or motifs) of the atoms are determined by the forces between them; distances and directions from each other are also determined by these forces. These arrangements are linked together in a three-dimensional repeating network known as a lattice. From a consideration of the physics of particles packing around each other, it can be shown that there are only 14 space lattices possible. These are known as Bravais lattices. There are also limitations on how the motifs are arranged in the lattice such that they fill

Principal Types of Particulate Fillers

all space. There are only 32 possible ways of arranging atoms around a point in a lattice leading to 230 possible arrangements of objects in space, the so-called space groups. The forces between the atoms result in specific three-dimensional repeatable orders. The limiting factors depend on the shape of the unit cell and the environment in which it grows.

The external faces of a mineral are therefore partly an accident of growth: changes imposed by environmental factors do not alter the fundamental properties of the crystal, although other effects may be noticed. The regular ordered structure of the crystal means that it may have different properties in different planes and directions.

The nature of the atoms forming the mineral, the forces between them and their order determine its fundamental properties such as chemistry, density, optical properties, hardness and shape. They will also determine the mechanisms by which minerals fracture and thus play a strong role in determining shape, size and size distribution.

Impurities, which are invariably present in the mineral from its original formation, may also be present in the crystal and will play a significant role in all the above. Impurities will also play a major role in determining the surface chemistry of the minerals due to adsorption of organics or inorganics on to faces. The chemistry of the faces may also be altered, compared with the bulk, by weathering, and ‘fresh’ faces that may be formed during processing may have different adsorption behaviour and different reactivities