Nanofibers and Nanotechnology in Textiles

i

Nanofibers and nanotechnology in textiles

© 2007, Woodhead Publishing Limited

ii

The Textile Institute and Woodhead Publishing The Textile Institute is a unique organisation in textiles, clothing and footwear. Incorporated in England by a Royal Charter granted in 1925, the Institute has individual and corporate members in over 90 countries. The aim of the Institute is to facilitate learning, recognise achievement, reward excellence and disseminate information within the global textiles, clothing and footwear industries. Historically, The Textile Institute has published books of interest to its members and the textile industry. To maintain this policy, the Institute has entered into partnership with Woodhead Publishing Limited to ensure that Institute members and the textile industry continue to have access to highcalibre titles on textile science and technology. Most Woodhead titles on textiles are now published in collaboration with The Textile Institute. Through this arrangement, the Institute provides an Editorial Board which advises Woodhead on appropriate titles for future publication and suggests possible editors and authors for these books. Each book published under this arrangement carries the Institute’s logo. Woodhead books published in collaboration with The Textile Institute are offered to Textile Institute members at a substantial discount. These books, together with those published by The Textile Institute that are still in print, are offered on the Woodhead website at: www.woodheadpublishing.com. Textile Institute books still in print are also available directly from the Institute’s website at: www.textileinstitutebooks.com.

© 2007, Woodhead Publishing Limited

iii

Nanofibers and nanotechnology in textiles Edited by P. J. Brown and K. Stevens

CRC Press Boca Raton Boston New York Washington, DC

WOODHEAD

PUBLISHING LIMITED Cambridge, England

© 2007, Woodhead Publishing Limited

iv Published by Woodhead Publishing Limited in association with The Textile Institute Woodhead Publishing Limited, Abington Hall, Abington Cambridge CB21 6AH, England www.woodheadpublishing.com Published in North America by CRC Press LLC, 6000 Broken Sound Parkway, NW, Suite 300, Boca Raton, FL 33487, USA First published 2007, Woodhead Publishing Limited and CRC Press LLC © 2007, Woodhead Publishing Limited The authors have asserted their moral rights. This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. Reasonable efforts have been made to publish reliable data and information, but the authors and the publishers cannot assume responsibility for the validity of all materials. Neither the authors nor the publishers, nor anyone else associated with this publication, shall be liable for any loss, damage or liability directly or indirectly caused or alleged to be caused by this book. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming and recording, or by any information storage or retrieval system, without permission in writing from Woodhead Publishing Limited. The consent of Woodhead Publishing Limited does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing from Woodhead Publishing Limited for such copying. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. Library of Congress Cataloging in Publication Data A catalog record for this book is available from the Library of Congress. Woodhead Publishing ISBN 978-1-84569-105-9 (book) Woodhead Publishing ISBN 978-1-84569-373-2 (e-book) CRC Press ISBN 978-1-4200-4449-2 CRC Press order number: WP4449 The publishers’ policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp which is processed using acidfree and elementary chlorine-free practices. Furthermore, the publishers ensure that the text paper and cover board used have met acceptable environmental accreditation standards. Typeset by Replika Press Pvt Ltd, India. Printed by TJ International Limited, Padstow, Cornwall, England

© 2007, Woodhead Publishing Limited

v

Contents

Contributor contact details

xiii

Part I Nanofiber production 1

Electrospinning of nanofibers and the charge injection method

3

D. R. SALEM, Charge Injection Technologies Inc., USA

1.1 1.2 1.3 1.4 1.5 2

Introduction Principles of electrostatic atomization Electrospraying and electrospinning by the capillary method Electrospraying and electrospinning by the charge injection method References

3 3

12 20

Producing nanofiber structures by electrospinning for tissue engineering

22

5

F. K. KO, The University of British Columbia, Canada and M. R. GANDHI, Drexel University, USA

2.1 2.2 2.3 2.4 2.5 2.6 2.7

Introduction Fabrication of nanofibrous scaffolds Characterization of nanofibrous scaffolds Cell–scaffold interaction Summary and conclusion Acknowledgments References

22 28 30 36 42 43 43

3

Continuous yarns from electrospun nanofibers

45

E. SMIT, U. BÜTTNER and R. D. SANDERSON, Stellenbosch University, South Africa

3.1

Introduction

© 2007, Woodhead Publishing Limited

45

vi

Contents

3.2 3.3 3.4 3.5 3.6 3.7 3.8

Using electrospun nanofibers: background and terminology Controlling fiber orientation Producing noncontinuous or short yarns Producing continuous yarns Summary and future trends Sources of further information and advice References

45 48 49 52 66 67 68

4

Producing polyamide nanofibers by electrospinning

71

M. AFSHARI, R. KOTEK and A. E. TONELLI, North Carolina State University, USA and D.-W. JUNG, Hyosung Corporation, South Korea

4.1 4.2 4.3 4.4

4.5 4.6 4.7 5

Introduction The electrospinning process Properties of electrospun nanofibers Measuring the effects of different spinning conditions and the use of high molecular weight polymers on the properties of electrospun nanofibers Improving the properties of electrospun nanofibers: experimental results Conclusions References

71 71 73

77 85 87

Controlling the morphologies of electrospun nanofibres

90

75

T. LIN and X. G. WANG, Deakin University, Australia

5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9

Introduction The electrospinning process and fibre morphology Polymer concentration and fibre diameter Fibre bead formation and fibre surface morphology Controlling fibre alignment and web morphologies Bicomponent cross-sectional nanofibres Future trends Acknowledgements References

Part II Carbon nanotubes and nanocomposites 6

Synthesis, characterization and application of carbon nanotubes: the case of aerospace engineering

90 91 93 96 100 103 107 108 108 111 113

M. REGI, University of Rome ‘La Sapienza’, Italy

6.1

Introduction

© 2007, Woodhead Publishing Limited

113

Contents

6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 6.10 6.11 6.12 6.13 7

The development and structure of carbon nanotubes Synthesis of carbon nanotubes Characterization techniques Purification techniques The use of carbon nanotubes in aerospace engineering Nanostructured composite materials for aerospace applications Nanostructured solid propellants for rockets Frequency selective surfaces for aerospace applications Other aerospace applications of carbon nanotubes Conclusions Acknowledgments References Carbon nanotube and nanofibre reinforced polymer fibres

vii

115 124 140 152 157 162 170 175 182 184 184 185 194

M. S. P. SHAFFER, Imperial College London, UK and J. K. W. SANDLER, University of Bayreuth, Germany

7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 8

Introduction Synthesis and properties of carbon nanotubes Developing nanotube/nanofibre–polymer composites Adding nanotubes and nanofibres to polymer fibres Analysing the rheological properties of nanotube/nanofibre–polymer composites Analysing the microstructure of nanotube/nanofibre– polymer composites Mechanical, electrical and other properties of nanocomposite fibres Future trends References Structure and properties of carbon nanotube-polymer fibers using melt spinning

194 197 201 206 208 212 216 221 222 235

R. E. GORGA, North Carolina State University, USA

8.1 8.2 8.3 8.4 8.5 8.6 8.7 8.8 8.9

Introduction Producing carbon nanotube-polymer fibers Thermal characterization Fiber morphology Mechanical properties of fibers Conclusions and future trends Sources of further information and advice Acknowledgments References

© 2007, Woodhead Publishing Limited

235 236 237 238 245 251 252 252 253

viii

Contents

9

Multifunctional polymer nanocomposites for industrial applications

256

S. J. BULL, University of Newcastle, UK

9.1 9.2 9.3

256 257

9.7 9.8 9.9 9.10

Introduction The development of functional polymer nanocomposites Improving the mechanical properties of polymer nanocomposites Improving the fire-retardant properties of polymer nanocomposites Improving the tribological properties of polymer nanocomposites Case-study: development of a nanocomposite sliding seal ring Enhancing the functionality of polymer nanocomposites Conclusions Acknowledgements References

10

Nanofilled polypropylene fibres

281

9.4 9.5 9.6

258 260 262 265 273 275 275 275

M. SFILIGOJ SMOLE and K. STANA KLEINSCHEK, University of Maribor, Slovenia

10.1 10.2 10.3 10.4 10.5 10.6 10.7

Introduction Polymer layered silicate nanocomposites The structure and properties of layered silicate polypropylene nanocomposites Nanosilica filled polypropylene nanocomposites Calcium carbonate and other additives Conclusion References

281 282 284 289 291 293 293

Part III Improving polymer functionality

299

11

301

Nanostructuring polymers with cyclodextrins A. E. TONELLI, North Carolina State University, USA

11.1 11.2 11.3 11.4 11.5

Introduction Formation and characterization of polymer–cyclodextrin– inclusion compounds Properties of polymer–cyclodextrin–inclusion compounds Homo- and block copolymers coalesced from their cyclodextrin–inclusion compounds Constrained polymerization in monomer–cyclodextrin– inclusion compounds

© 2007, Woodhead Publishing Limited

301 302 304 308 310

Contents

11.6

11.8 11.9 11.10 11.11

Coalescence of common polymer–cyclodextrin–inclusion compounds to achieve fine polymer blends Temporal and thermal stabilities of polymers nanostructured with cyclodextrins Cyclodextrin-modified polymers Polymers with covalently bonded cyclodextrins Conclusions References

12

Dyeable polypropylene via nanotechnology

11.7

ix

311 312 313 314 316 316 320

Q. FAN and G. MANI, University of Massachusetts Dartmouth, USA

12.1 12.2 12.3

12.7 12.8 12.9

Introduction Dyeing techniques for unmodified polypropylene Modified polypropylene for improved dyeability using copolymerization and other techniques Polyblending and other techniques for improving polypropylene dyeability Dyeing polypropylene nanocomposites Using X-ray diffraction analysis and other techniques to assess dyed polypropylene nanocomposites Conclusions Acknowledgments References

334 345 346 346

13

Polyolefin/clay nanocomposites

351

12.4 12.5 12.6

320 321 323 324 326

R. A. KALGAONKAR and J. P. JOG, National Chemical Laboratory, India

13.1 13.2 13.3 13.4 13.5 13.6 13.7 13.8

Introduction Organomodification of clays Polymer/clay nanocomposites Polypropylene/clay nanocomposites Polyethylene/clay nanocomposites Higher polyolefin/clay nanocomposites Conclusions References

351 354 356 360 367 372 374 381

14

Multiwall carbon nanotube–nylon-6 nanocomposites from polymerization

386

Y. K. KIM and P. K. PATRA, University of Massachusetts Dartmouth, USA

14.1 14.2 14.3

Introduction Nanocomposite synthesis and production Characterization techniques

© 2007, Woodhead Publishing Limited

386 387 388

x

Contents

14.4

Properties of multiwall carbon nanotube–nylon-6 nanocomposite fibers Conclusions Acknowledgments References

14.5 14.6 14.7

Part IV Nanocoatings and surface modification techniques 15

Nanotechnologies for coating and structuring of textiles

391 404 405 406 407 409

T. STEGMAIER, M. DAUNER, V. VON ARNIM, A. SCHERRIEBLE, A. DINKELMANN and H. PLANCK, ITV Denkendorf, Germany

15.1 15.2 15.3 15.4 15.5 15.6 15.7 16

Introduction Production of nanofiber nonwovens using electrostatic spinning Anti-adhesive nanocoating of fibers and textiles Water- and oil-repellent coatings by plasma treatment Self-cleaning superhydrophobic surfaces Sources of further information and advice References

409 410 417 418 421 427 427

Electrostatic self-assembled nanolayer films for cotton fibers

428

G. K. HYDE and J. P. HINESTROZA, Cornell University, USA

16.1 16.2

16.8

Introduction Principles of electrostatic self-assembly for creating nanolayer films Advantages and disadvantages of electrostatic self-assembly Substrates used for electrostatic self-assembly Polyelectrolytes used for electrostatic self-assembly Analyzing self-assembled nanolayer films on cotton Conclusions: functional textiles for protection, filtration and other applications References

17

Nanofabrication of thin polymer films

16.3 16.4 16.5 16.6 16.7

428 428 431 432 434 436 439 440 448

I. LUZINOV, Clemson University, USA

17.1 17.2 17.3 17.4

Introduction Macromolecular platform for nanofabrication ‘Grafting from’ technique for synthesis of polymer films ‘Grafting to’ technique for synthesis of polymer films

© 2007, Woodhead Publishing Limited

448 449 451 455

Contents

xi

17.5 17.6 17.7 17.8 17.9

Synthesis of smart switchable coatings Synthesis of ultrahydrophobic materials Conclusions Acknowledgments References

458 464 466 466 467

18

Hybrid polymer nanolayers for surface modification of fibers

470

S. MINKO and M. MOTORNOV, Clarkson University, USA

18.1 18.2 18.3 18.4 18.5 18.6 18.7 18.8 18.9

Introduction: smart textiles via thin hybrid films Mechanisms of responsive behavior in thin polymer films Polymer–polymer hybrid layers Polymer–particles hybrid layers Hierarchical assembly of nanostructured hybrid films Future trends Sources of further information and advice Acknowledgment References

470 471 478 484 485 489 490 490 490

19

Structure–property relationships of polypropylene nanocomposite fibres

493

C. Y. LEW, University of Oxford, UK and G. M. MCNALLY, Queen’s University Belfast, UK

19.1 19.2 19.3 19.4 19.5 19.6 19.7 19.8 19.9

Introduction Materials, processing and characterisation techniques Structure and morphology Phase homogeneity and spinline stability Optical birefringence and infrared activation Crystallisation behaviour and mechanical performance Exfoliation by extensional flow deformation Conclusions References

© 2007, Woodhead Publishing Limited

493 495 497 502 505 509 513 514 515

xiii

Contributor contact details

(* = main contact)

Editors

Chapter 2

P. J. Brown and K. Stevens 265 Sirrine Hall School of Materials Science and Engineering Clemson University Clemson SC 29634 USA

F. K. Ko* Advanced Materials and Process Engineering Laboratory (AMPEL) Department of Materials Engineering The University of British Columbia 113–2355 East Mall Vancouver BC, Canada V6T 1Z4

e-mail: [email protected] e-mail: [email protected]

Chapter 1 David R. Salem Charge Injection Technologies Inc. Present address: Nanoproducts Corporation 14330 Longs Peak Court Longmort CO 80504 e-mail: [email protected] [email protected]

© 2007, Woodhead Publishing Limited

e-mail: [email protected]

M. R. Gandhi School of Biomedical Engineering Sciences and Health System Drexel University Philadelphia PA 19104 USA e-mail: [email protected]

xiv

Contributor contact details

Chapter 3

Chapter 5

E. Smit*, U. Büttner and R. D. Sanderson UNESCO Associated Centre for Macromolecules & Materials Department of Chemistry and Polymer Science Stellenbosch University Private bag X1 Matieland 7602 South Africa

T. Lin* and X. G. Wang Centre for Material and Fibre Innovation Faculty of Science and Technology Deakin University Geelong Victoria 3217 Australia

e-mail: [email protected] [email protected]

Chapter 4 M. Afshari*, R. Kotek and A. E. Tonelli College of Textiles North Carolina State University Raleigh NC 27695-8301 USA e-mail: [email protected] [email protected] [email protected]

Dong-Wook Jung Hyosung Corporation South Korea

e-mail: [email protected]

Chapter 6 M. Regi (PhD) University of Rome ‘La Sapienza’ Department of Aeronautics and Astronautics Engineering Via Eudossiana 18 00184 Roma Italy e-mail: [email protected]

Chapter 7 M. S. P. Shaffer* Department of Chemistry Imperial College London London SW7 2AZ UK e-mail: [email protected]

J. K. W. Sandler Polymer Engineering University of Bayreuth D-95447 Bayreuth Germany e-mail: [email protected] [email protected]

© 2007, Woodhead Publishing Limited

Contributor contact details

xv

Chapter 8

Chapter 11

R. E. Gorga Fiber and Polymer Science Program Department of Textile Engineering, Chemistry and Science Campus Box 8301 North Carolina State University Raleigh NC 27695-8301 USA

A. E. Tonelli Fiber and Polymer Science Program North Carolina State University Campus Box 8301 Raleigh NC 27695-8301 USA

e-mail: [email protected]

Chapter 9 S. J. Bull School of Chemical Engineering and Advanced Materials University of Newcastle Newcastle upon Tyne NE1 7RU UK

e-mail: [email protected]

Chapter 12 Q. Fan* and G. Mani Department of Materials and Textiles University of Massachusetts Dartmouth North Dartmouth MA 02747 USA e-mail: [email protected]

e-mail: [email protected]

Chapter 13

Chapter 10

R. A. Kalgaonkar and J. P. Jog* Polymer Science and Engineering Division National Chemical Laboratory Dr Homi Bhabha Road Pasahan Pune – 411008 India

M. Sfiligoj-Smole* and K. Stana Kleinschek University of Maribor Faculty of Mechanical Engineering Smetanova 17 SI 2000 Maribor Slovenia e-mail: [email protected]

© 2007, Woodhead Publishing Limited

e-mail: [email protected] [email protected]

xvi

Contributor contact details

Chapter 14

Chapter 17

Y. K. Kim and P. K. Patra* Department of Materials and Textiles College of Engineering University of Massachusetts Dartmouth 285 Old Westport Road North Dartmouth MA 02747 USA

I. Luzinov School of Materials Science and Engineering 161 Sirrine Hall Clemson University Clemson SC 29634-0971 USA

e-mail: [email protected] [email protected]

Chapter 18

Chapter 15 T. Stegmaier*, M. Dauner, V. Von Arnim, A. Scherrieble, A. Dinkelmann and H. Planck Institut für Textil- und Verfahrenstechnik Denkendorf Koerschtalstrasse 26 D-73770 Denkendorf Germany e-mail: [email protected]

Chapter 16 G. K. Hyde and J. P. Hinestroza* Cornell University Fiber Science Program Ithaca NY 14850 USA e-mail: [email protected]

© 2007, Woodhead Publishing Limited

e-mail: [email protected]

S. Minko* and M. Motornov Clarkson University Department of Chemistry 8 Clarkson Ave Potsdam NY 13699 USA e-mail: [email protected]

Chapter 19 C. Y. Lew* Department of Engineering Science University of Oxford Engineering and Technology Building Parks Road Oxford OX1 3PJ UK e-mail: [email protected]

G. M. McNally Polymer Processing Research Centre Queen’s University Belfast Ashby Building Stranmillis Road Belfast BT9 5AH UK

Part I Nanofiber production

1

© 2007, Woodhead Publishing Limited

1 Electrospinning of nanofibers and the charge injection method D. R. S A L E M, Charge Injection Technologies Inc., USA

1.1

Introduction

The use of electric charge to break up liquids into small particles has been well known and extensively studied for over a century, but commercial applications have been constrained by difficulties in surmounting flow rate limitations associated with the underlying physics of the process. This is true for both electrospraying, in which low-viscosity liquids can be atomized into droplets, and electrospinning, in which viscoelastic liquids can be transformed into filaments of submicrometer and nanometer dimensions. In this chapter, we will start by reviewing the principal forces involved in electrostatic atomization, which also form the basis of the electrospinning process, and then discuss the development of the science and technology of electrospraying and electrospinning, with particular emphasis on efforts to increase the rate at which nanofibers can be electrospun. After reviewing advances in the conventional approach to charging liquids in electrospraying and electrospinning (usually referred to as the capillary or needle method) we will highlight an alternative charging technology, known as the charge injection method, which is being developed for the production of nanometer and submicrometer fibers at exceptionally high output rates.

1.2

Principles of electrostatic atomization

It has long been known that application of electric charge to a liquid droplet causes instability of the liquid, resulting in distortion of the droplet or meniscus and in the ejection of liquid filaments and/or satellite droplets.1–4 The effect is explained as a competition between the Coulomb repulsion of like charges favoring droplet distortion/partitioning and surface tension opposing droplet division. For example, in the case of a droplet of a conductive fluid in an electric field (where the charge accumulates at the droplet surface and there is no electric field inside the droplet) the pressure balance is given by: 3

© 2007, Woodhead Publishing Limited

4

Nanofibers and nanotechnology in textiles

e2 ∆P = 2σ – R 32 π 2 ε 0 R 4

[1.1]

where e is the total droplet charge, R is the droplet radius, σ is the surface tension and ε0 is the vacuum permittivity. It is informative that the relationship between the pressure drop and droplet radius is not monotonic (Fig. 1.1) – the electrostatic pressure, e2/(32π 2ε0R4), becomes dominant as droplet radius becomes smaller (charge density increases), so that the function passes through a maximum and then reaches a point at which the pressure in the atmosphere and the pressure in the droplet are the same (p = 0). This point is associated with the electrostatic Rayleigh criterion, and can be interpreted as the maximum charge density that a droplet of a given diameter can withstand. Rewritten as the charge per mass, the Rayleigh relation takes the more familiar form:5 e = M

288ε 0 σ d 3ρ 2

[1.2]

The non-monotonic relationship between pressure drop and droplet radius has important consequences for understanding and predicting droplet/vapor coexistence and the behavior of an evaporating charged droplet, for which the pressure balance can be expressed as:6 e2   2σ ln Pv / P0 = v ∆ P = v  R –  kT kT  32 π 2 ε 0 R 4 

[1.3]

p 0.5

0

1

2

3

4

5

6

X –0.5

–1.0

–1.5

–2.0

1.1 Dimensionless pressure drop p = ∆Pᐉ/2σ as a function of the dimensionless droplet radius X = R/ᐉ, where ᐉ is the characteristic length scale.

© 2007, Woodhead Publishing Limited

Electrospinning of nanofibers and the charge injection method

5

where P0 is the saturation pressure for a planar vapor/liquid uncharged surface. Kornev et al. have employed this relationship to anticipate the destiny of charged droplets surrounded by their own vapor under a range of pressure conditions.6 It is noteworthy, especially in relation to our later discussions on electrospinning, that cylindrical liquid columns are also subject to the Rayleightype instability, in which case the pressure balance is given by:6

κ2 ∆P = σ – 2 R 8π ε 0 R 2

[1.4]

where κ is the charge per unit length of the filament. Written in terms of charge density, the Rayleigh criterion for a charged liquid becomes: e M

= column

64σε 0 d 3ρ 2

[1.5]

It is immediately apparent from Equations [1.2] and [1.5] that the charge required to reach the Rayleigh limit is about two times smaller for a column of liquid than for a droplet of the same radius. The above analysis relates to charge-induced liquid break-up under static conditions, in order to provide an understanding of the primary forces involved, but the charge-induced break-up of flowing liquids is complicated by the superposition hydrodynamic perturbations and electrostatic instabilities that result in a variety of disruption behaviors, some of which will be discussed below.

1.3

Electrospraying and electrospinning by the capillary method

1.3.1

Operating modes

The earliest, and still the most widespread, practical use of electrostatic instabilities in liquids is electrospraying. It should be pointed out, however, that the term electrospraying is frequently applied to processes in which the primary liquid break-up is not generated by electrostatic forces, but by high pressure or some other mechanical method. In this case, the applied electric field mainly serves to charge the droplets so that they can be efficiently attracted to a grounded target and the technology is better described as electrostatically assisted spraying. Important commercial examples include electrostatic paint guns and agricultural sprayers, where large volumes of charged particles must be generated. Electrospraying in which the primary break-up process (as well as any subsequent droplet division) occurs as a direct result of electrostatic forces is often referred to as electrohydrodynamic (EHD) atomization, and tends to

© 2007, Woodhead Publishing Limited

6

Nanofibers and nanotechnology in textiles

find application where flow rates can be low or minuscule. This is because EHD atomization using conventional charging technologies (often referred to as the capillary method) cannot operate at high rates of liquid delivery. In a common set-up, a conductive liquid is delivered to the tip of a metal capillary, which is at high negative or positive potential (Fig. 1.2). As a result of the electric field generated, charge accumulates at the surface of the pendant droplet formed at the tip of the capillary and creates an instability that deforms the hemispherical droplet into a cone shape, often referred to as a Taylor cone.3, 4, 7 At a sufficiently high field strength, a jet of liquid is continuously ejected from the apex of the cone and breaks up into charged particles. In this cone-jet mode of operation,8 a stable, continuous stream of charged particles can be generated. The break-up of the jet may be via an axisymmetric varicose instability, a bending/whipping instability or, more rarely, a ramified mode involving distortion of the jet’s circular cross-section and emission of lateral sub-jets (Fig. 1.3).10, 11 The varicose instability occurs at relatively low surface charge and is similar in manner to the break-up of a neutral jet. This mode can produce charged sprays with highly monodisperse droplet diameters and mean diameters ranging from a few nanometers to hundreds of micrometers, depending on field strength and fluid properties such as conductivity and viscosity. As surface charge on the jet increases (by raising the flow rate11–14 or the applied voltage11, 13, 15 to increase current), the axisymmetric break-up mode

Syringe

Liquid

Capillary

Voltage source

Taylor cone and liquid jet Whipping filament (electrospinning) Charged droplets (electrospraying) Collector electrode

1.2 Typical set-up for electrospraying/electrospinning by the capillary method. The inset is an example of a pendant droplet, distorted by the electric field, and the emitted jet (adapted from Ref. 9).

© 2007, Woodhead Publishing Limited

Electrospinning of nanofibers and the charge injection method

Axisymmetric

Bending/whipping

7

Ramified

1.3 Principal jet break-up modes (adapted from Ref. 11).

transitions to the bending/whipping instability.10, 11, 16 The whipping motion rapidly thins the jet and breaks it into a spray with polydisperse droplet diameters having mean values usually of the order of tens of micrometers. If the jet is highly charged, the electric stresses can overcome surface tension, causing the cross-section of the jet to deform or bulge in one or more locations, from which fine sub-jets are released.10, 11 This ramified mode is of course related to the electrostatic Rayleigh break-up mode anticipated by Equation [1.4], although this equation cannot be directly used to indicate the charge threshold for Coulombic rupture in a column of liquid that is flowing. For example, it has been shown that the stretching of a charged liquid column, as in an accelerating jet, not only introduces hydrodynamic perturbations, but also modifies (compared with a static liquid column) the relationship between Laplacian pressure and electrostatic pressure in a way that tends to stabilize the column against Coulombic disruption.6 Ramified jet break-up is seldom observed in the capillary method of electrospraying because corona discharge prevents reaching the required field strength. However, it may be noted that dramatic Coulombic explosion of a liquid helium jet was observed by Tsao et al. using capillary electrospraying.17 No Taylor cone was formed, and the shattering of the helium jet into droplets of 1–10 µm diameter was attributed to charge densities that – owing to high current and exceptionally low surface tension – were computed to be 50 times the Rayleigh limit (for a stationary liquid cylinder). In this case, the ratio of electric stress to surface tension was evidently sufficient to overwhelm any stabilizing effects of the accelerating jet. If the charged droplets from any electrospraying process evaporate sufficiently rapidly, they may undergo further disruption and division after the initial break-up, since the shrinking droplets (both parent and daughter droplets) will repeatedly attain the threshold charge for electrostatic Rayleigh

© 2007, Woodhead Publishing Limited

8

Nanofibers and nanotechnology in textiles

division.18–21 This phenomenon led to the ground-breaking application of cone-jet electrospraying in mass spectroscopy, where the droplets become so small from repeated division that the electric field due to the high surface charge density is strong enough to desorb ions from the droplets into the ambient gas.22 ‘Quasi-molecules’ suitable for mass analysis are produced when solute species are attracted to desorbing cations and anions. Single and few-chain polymer particles have been produced by this multiple-fission method.23 Electrospinning of fibers from polymer solutions is usually carried out by the capillary method of Fig. 1.2. In a typical experiment, a pendant droplet of the polymer solution at the capillary tip is subject to the electric field created by the potential difference between the capillary and a grounded collector.9, 24, 25, 26 Although polymer solutions are subject to the same competition between axisymmetric and whipping instabilities described for electrospraying,27, 28 the viscoelastic properties of these solutions delay or completely suppress the break-up of the fluid into droplets. Consequently, the jet rapidly thins as it is accelerated from the cone apex and, at high enough field strengths, is further attenuated as it undergoes the whipping instability.9, 27–30 Evaporation of the solvent in the course of this draw-down process results in the accumulation of solid filaments at the collector with mean diameters typically in the range 100–500 nm, and sometimes well below 100 nm.31, 32 An example of nanofibers produced by capillary electrospinning is shown in Fig. 1.4.

5 µm

1.4 SEM image of poly(vinyl pyrrolidone) nanofibers electrospun by the ‘capillary’ method. (Adapted from D. Li and Y. Xia, Adv. Mater. (2004) 16, p. 1151.)

© 2007, Woodhead Publishing Limited

Electrospinning of nanofibers and the charge injection method

9

Nevertheless, droplet creation – both isolated droplets and ‘beads on a string’9, 31–34 – can occur in electrospinning when the solution is insufficiently elastic. Inadequate solution elasticity usually arises from lack of molecular entanglements in solutions that are too dilute, and the polymer concentration generally needs to be well above the critical concentration for chain overlap.34 The essential importance of elasticity in electrospinning has been confirmed recently in a systematic study by Yu et al. using model fluids with different degrees of elasticity.35 They demonstrated that a critical value of elastic stress indicates the complete suppression of the instability responsible for jet break-up into droplets, and marks a transition from bead-on-a-string morphology to the formation of uniform fibers. On the basis that the whipping of the jet arises when surface tension is insufficient to stabilize the jet against perturbations that grow with increasing surface charge, Fridrikh et al.36 have shown that the terminal diameter ht of a polymer solution undergoing the whipping instability is controlled by the flow rate Q, electrical current I and surface tension γ. Thus, at a given polymer concentration:

Q2   2 ht =  γ ε 2 2  π χ I (2 ln – 3)  

1/3

[1.6]

where ε is the dielectric constant of the outside medium and χ is the dimensionless wavelength of the instability responsible for the normal displacements. This relationship (which assumes negligible solvent evaporation during draw-down and neglects elastic effects) confirms that a valuable strategy to minimize fiber diameter is to increase the charge-carrying capacity of the fluid. This has been demonstrated in a number of experimental studies where conductive additives were introduced into the solution.26, 32, 37 An interesting modification of the capillary method, explored recently, involves the application of AC voltage instead of DC,38 or, more optimally, AC voltage with biased DC voltage.39 As a result of short charge segments of alternating polarity along the jet, the net charge was reduced and the whipping instability was thereby diminished or suppressed. Yet under optimized AC frequency and DC bias conditions, draw-down from the straight jet was sufficient to produce aligned fibers with diameters ranging from 100 to 200 nm.

1.3.2

Output limitations and recent developments

The greatest drawback of both electrospraying and electrospinning by the standard capillary method is the very low rates at which they can usefully generate sprays or nanofibers. Typically, output of fluid from a single capillary is in the range 1–5 milliliters per hour.

© 2007, Woodhead Publishing Limited

10

Nanofibers and nanotechnology in textiles

The rate of fluid flow is largely determined by the strength of the electric field, since the field is responsible for accelerating the jet from the cone apex. However, the field strength that can be applied is limited by the electrical breakdown strength of the atmosphere (usually air) in which the droplets or fibers are forming, in addition to which the space charge field from the charged droplets or filament can increasingly attenuate the applied electric field as flow rate (and current) increases. It is also well documented that with the standard capillary method, the diameters of the droplets or fibers increase with flow rate, so that – even before reaching the limiting flow rate imposed by electrical breakdown – the dimensions of these entities may be too large to be of practical interest. Furthermore, it is noteworthy that despite volume flow rates of milliliters per hour, the jet velocity must be in the range 100– 300 m/s in existing electrospinning practice for fibers to attain dimensions of a few hundred nanometers. It is therefore improbable that a substantial increase of these, already high, velocities is likely to arise from manipulation of electric fields alone. For these reasons, most efforts to increase output from electrospraying and electrospinnng have centered on multiplexing the liquid output source, either by stacking an array of capillaries as closely together as possible, or by attempting to operate in the so-called multi-jet mode. Both these approaches are complex, and are difficult to implement robustly, although some significant improvements in output have been reported recently. The multi-jet mode, in which several cone-jets emanate from a single capillary tube, appears within a narrow range of voltages and flow rates, and the individual jets are prone to interruption and positional instability. Duby et al.40 have recently demonstrated, however, that by introducing a number of grooves in the tube circumference to intensify the electric field at these locations, the stability of the multi-jet mode in electrospraying was much improved. Fifty cone-jets per cm2 could be achieved, capable of delivering sprays with mean droplet diameters in the range 7–20 µm at rates between about 0.2 and 1.5 ml/min per cm2 of spray-head area. This delivery rate remains far too low for most commercial applications, and in any case the multi-jet approach has not, to our knowledge, been successfully applied to electrospinning. Assembling an array of capillaries or orifices to increase output in electrospraying and electrospinning has been explored by several groups, but the fundamental difficulty with this tactic is that each jet is subject to electric-field shielding and space-charge interference from adjacent jets.41–45 Consequently, at orifice densities that would significantly improve output, only the outermost jets in an array will function effectively, and addressing this problem requires careful configuration of the electric field at each jet. In the case of electrospraying, a systematic attempt to configure the electric fields was made by Deng et al.41 who microfabricated an array of orifices

© 2007, Woodhead Publishing Limited

Electrospinning of nanofibers and the charge injection method

11

etched in silicon with a density of 250 orifices/cm2. Reduction of cross-talk between jets and minimization of space-charge feedback were achieved by positioning an extractor electrode (consisting of a plate with holes concentric with the output orifices) at a distance from the orifices that was comparable to the inter-orifice spacing of 0.5 mm. It would appear that, under optimal conditions, monodisperse sprays with flow rates of the order of 5 ml/min per cm2 of spray-head area can be delivered by this type of device. For electrospinning, similar attempts have been made to configure the electrical fields using secondary external electrodes42, 43 but so far the achievable packing density of spinneret orifices for successful nanofiber spinning appears to be considerably less than in the case of electrospraying. Reasonable jet stability in electrospinning seems to require inter-orifice spacing in the range of several millimeters to centimeters with packing densities of the order of two jets/cm2. Consequently, in these multiplexed systems, output per unit area of the spinneret head remains very modest. Electrospinning apparatus appearing to provide a relatively high nanofiber output is, however, described in a recent US patent.46 In this disclosure, block-type nozzles emit charged jets of a polymer solution from an array of pins. In one example, 200 blocks were deployed, comprising a total of 40 000 pins. However, no information is given on inter-pin spacing or strategies used for minimizing electrical interference between jets: rather, the patent focuses on the application of a device that intermittently delivers solution to each nozzle block (held at high voltage), thereby charging discrete quantities of the fluid consecutively. This is claimed to reduce the deterioration of electric forces that occur when the high voltage is applied to the whole spinning solution. In the case where 40 000 pins were used (delivering ~0.003 g/min of solid polyamide fiber per pin), a 0.6 m wide web was produced at 60 m/min and had a base weight of 3 g/m2.

1.3.3

Viscosity limitations and recent developments

It would be preferable to spin nanofibers from molten polymer than from polymer solution because the solvents used are generally flammable and/or toxic. For nanofiber production on an industrial scale, the use of solvents presents a significant environmental hazard, and the need for expensive safety precautions and solvent recovery equipment provides an additional obstacle to commercial scale-up of the technology. Whereas fibers can certainly be electrospun from the melt, they generally have diameters in the micrometer range. Resistance to nanofiber formation is largely due to the high viscosity of the polymer melt, being one to two orders of magnitude higher than typical solution viscosities, and this problem is no doubt compounded by the rapid increase in viscosity taking place as the thinning (and whipping) filament cools to ambient temperature. Furthermore,

© 2007, Woodhead Publishing Limited

12

Nanofibers and nanotechnology in textiles

melt electrospinning does not benefit from the contribution of solvent evaporation to the thinning process. Electrospinning into a heated chamber is likely to reduce the tendency for premature solidification of the molten filament during the electrically driven draw-down process, and this approach has been investigated recently by Zhou et al.47 Under optimized operating conditions, they electrospun molten polylactic acid (PLA) from a single capillary into fibers with a mean diameter of about 800 nm. The flow rate was 0.01 ml/min, and the spinning mode was characterized by a typical whipping instability. Another method to assist draw-down of high-viscosity polymer fluids in electrospinning involves subjecting the electrified polymer jet to a hot air stream, thereby decreasing the viscosity of the fluid and adding a mechanical pulling force to the jet. This has been developed into a multiple jet ‘electroblowing’ (or ‘blowing-assisted electrospinning’) system and is considered by its inventors to be suitable for molten polymers.43 The only published data to date, however, are from electroblowing of hyaluronic acid (HA) solutions, which exhibit unusually high viscosity at relatively low concentrations and could not be spun by unassisted electrospinning.48, 49 Via the electroblowing method, consistent fabrication of HA fiber membranes was achieved with mean fiber diameter of about 70 nm. It may be noted, however, that despite the pulling force of the heated air stream on the jet, the output of this process is not appreciably higher than an unassisted multiplejet electrospinning system, even when applied to lower-viscosity polymer solutions.

1.4

Electrospraying and electrospinning by the charge injection method

1.4.1

Principle of operation

Electrohydrodynamic atomization of liquids by charge injection is an alternative approach to the capillary method, and offers distinct advantages in terms of output and efficiency. One configuration of the charge injection method, shown in Fig. 1.5, comprises two electrodes immersed in a (non-conducting) fluid. The sharpened point of the ‘emitter electrode’, held at high electric potential, is centered over a grounded orifice (‘blunt electrode’). Owing to the small distance between electrodes (typically one to three orifice diameters), an intense electric field is set up in the fluid, which is much larger than that provided by the capillary method. The fluid is continuously forced through the orifice under high pressure and becomes highly charged as it passes between the electrodes. Most of this charge remains in the liquid because of the low mobility of electrons in an insulating fluid and due to the very short residence time of the fluid prior to exiting the orifice, the applied flow rate

© 2007, Woodhead Publishing Limited

Electrospinning of nanofibers and the charge injection method

13

Emitter electrode

Voltage source

Pressurized liquid

R

2nd electrode

Collector electrode

1.5 An example of the ‘triode’ configuration of the charge injection method.

being typically between 0.2 and 5 ml/s (three orders of magnitude greater than in the capillary method). After emerging from the orifice, the charged droplets or filaments can be attracted towards a third (collector) electrode in the form of a grounded (or oppositely charged) object, or they may be allowed to disperse freely in the environment. Unlike the capillary method, the charge injection system does not involve the formation of a Taylor cone, and the velocity of the fluid stream is determined by the (mechanical) pressure applied, not by the strength of the external electric field. In this way, charge injection electrospraying/electrospinning overcomes throughput limitations by decoupling the fluid flow from the field strength. Whereas the capillary method is applicable to conductive liquids, it is clear that charge injection technology can only be used with insulating or weakly conducting fluids, and much of the development of this technology has been related to the atomization of hydrocarbon fuels for combustion applications. For electrospinning fibers, however, there is no shortage of suitable polymer–solvent combinations with dielectric properties, and the non-conducting nature of polymer melts should make them particularly amenable to electrospinning by charge injection.

1.4.2

Operating regimes and limits

Spray behavior from charge injection atomizers has been studied and described in some detail by Kelly,50–52 a pioneer of this technology, and it will be

© 2007, Woodhead Publishing Limited

14

Nanofibers and nanotechnology in textiles

useful to review the series of events that occur to a stream of (low molecular weight) fluid issuing from the orifice of such a device as the applied voltage is increased. Starting as a uniform column, the stream starts to disrupt into droplets at some distance from the orifice once the voltage attains a threshold value (dependent on fluid properties, flow rate, device configuration, etc.). As the voltage is raised higher (further increasing the spray current and charge density of the fluid) column disruption occurs closer and closer to the orifice, atomization progressively intensifies, the droplets become smaller and the lateral dispersion of the spray increases. A voltage level is finally reached, however, at which the spray current falls precipitously due to electrical breakdown phenomena, causing the atomization to halt abruptly and the spray to collapse back to a flowing column of liquid (Fig. 1.6). Since the liquid breaks up most energetically at the maximum charge density, just prior to the breakdown voltage, a feedback controller for optimization of performance has been designed which detects trichel discharge events that occur as the breakdown limit is approached and then automatically adjusts the voltage back to stay just below this limit, providing stable operation for unlimited run times.50f Shrimpton and Yule54 have identified subcritical and supercritical flow regimes delineated by different types of electrical breakdown. At subcritical flow rates, the spray is not finely atomized and it is limited by electrical breakdown of the fluid within the atomizer. At supercritical flow rates, fine atomization occurs and spraying is limited by electrical breakdown of the air outside the orifice, which results from the high surface charge density of the exiting liquid column. It has been found, however, that the maximum achievable charge density increases with increasing stream velocity and decreasing orifice diameter.51, 54 The break-up behavior of a charge-injected columnar stream is depicted in Fig. 1.7. Although the orifice diameter is relatively large and the liquid is Optimal performance Charge density

Corona-induced breakdown

Voltage

1.6 Charging behavior of a fluid in a triode-type charge injection device (adapted from Ref. 53).

© 2007, Woodhead Publishing Limited

Electrospinning of nanofibers and the charge injection method

15

Orifice

1.7 Break-up of a liquid column of mineral oil just after exiting the orifice of a triode-type charge injection device (adapted from Ref. 52). Flow direction is from top to bottom of the image.

weakly charged in this example (0.15 C/m3), the column disruption is quite dramatic and revealing. Just after issuing (downwards) from the orifice, the column distorts laterally due to charge repulsion. The charge, and the fluid that carries it, tend to migrate to the edges and concentrate there, forming a thick rim around the extended core region. The high charge density in the rim creates additional instabilities, and lateral jets are emitted from localities at which Coulombic forces have overcome surface tension. This behavior is clearly reminiscent of the ramified jet mode, observed in cone-jet electrospraying at very high electric fields, and is indicative of Rayleightype Coulombic rupture. As charge density increases, this break-up mode becomes more vigorous and, in the absence of electrical breakdown phenomena, would be expected to progress towards explosive disruption at the fluid exit point. The observed form of Coulombic rupture appears to be favored by the charge injection method in both electrospraying and electrospinning, and the whipping mode is rarely observed. In the quest to improve the productivity of electrospinning, it is clear that stream-splitting instabilities in which multiple filaments are generated from a single liquid column are much more desirable than instabilities that generate a single, thinning filament from an intact column. Optimization

© 2007, Woodhead Publishing Limited

16

Nanofibers and nanotechnology in textiles

of column splitting or bursting in charge injection electrospinning is therefore of significant interest. Nanofiber webs spun by the charge injection system were first produced by A. J. Kelly in 2000. Research and development work has shown that the morphology of charge injection fiber webs is dependent on the polymer/ solvent system and on process variables (which for proprietary reasons cannot be detailed in this chapter). For example, polyurethane fibers produced at 1 ml/s under one set of conditions have significantly more network junctions than is typically found in webs produced by the capillary method, but show little direct evidence of being generated by a column-bursting mechanism (Fig. 1.8). On the other hand, fibers spun from the same polyurethane at the same flow rate under different operating conditions, display a distinct hierarchical structure in which interconnected networks of larger diameter fibers are further interconnected by networks of smaller diameter fibers (Fig. 1.9). The latter morphology is consistent with other evidence that the predominant mechanism of fiber formation in charge injection electrospinning involves expansion and splitting of two-dimensional films, propagated from a stretched and disrupted liquid column. These bimodal, interconnected membrane structures should be of significant commercial value, since the majority of filaments are nanofibers (providing, for example, ultra-efficient filtration properties) and the larger fibers are interconnected in such a way as to provide structural integrity and strength to the web. To our knowledge, however, the charge injection method has so

600 nm

1.8 SEM image of polyurethane nanofibers electrospun by the charge injection method.

© 2007, Woodhead Publishing Limited

Electrospinning of nanofibers and the charge injection method

17

6 µm (a)

2 µm (b)

1.9 SEM images of polyurethane nanofibers electrospun by the charge injection method, illustrating highly interconnected networks of fibers with smaller and larger diameters.

far been unable to produce webs in which the majority of the polymer has been converted into nanofibers. The membranes either contain nanofibers with unconverted polymer or they are fully fibrillar but comprise a mixture of nanofibers and microfibers.

© 2007, Woodhead Publishing Limited

18

Nanofibers and nanotechnology in textiles

Under yet another set of operating conditions, electrospinning from a solution of polypropylene and decalin at elevated temperature (below the solvent’s boiling point), we observed scattered instances of hollow droplets that appear to have partially transformed into nanofibers as a result of explosive rupture (Fig. 1.10). As discussed by Kornev et al.,55 we attribute this to a charge-induced cavitation process that can be predicted from theory.

2 µm (a)

600 nm

(b)

1.10 SEM images of nanofibers produced from ‘exploded’ droplets in charge injection electrospinning of a polypropylene/decalin solution.55 (The lower image is a section of the upper image at higher magnification.)

© 2007, Woodhead Publishing Limited

Electrospinning of nanofibers and the charge injection method

19

Efforts to increase nanofiber yield in solution and to achieve nanofiber production from molten polymer using charge injection are focusing on increasing the charge density beyond the limits (of about 6 C/m3) currently imposed by electrical breakdown phenomena, and improving the uniformity of charge distribution across the orifice diameter (since it is known that charge density is generally lower at the center of the exiting fluid column than in the outer regions).51 Some strategies are outlined below.

1.4.3

Strategies for further development

A patented strategy for increasing charge density above the corona-induced breakdown point is to pulse the voltage of the charge injection device from a base value below the breakdown limit to a value above the breakdown limit, for a time shorter than that required for breakdown.50g It has been shown that under typical spraying condition, pulses of a few milliseconds duration should allow voltage to exceed the threshold level without a corona discharge effect occurring. Another approach is to use a sealed electron beam gun (Fig. 1.11).50d The fluid is passed over an electron-transparent window of the sealed gun, which injects energetic electrons into the fluid as it sweeps by. This method could provide highly efficient charging of the fluid and may be the key to obtaining nanofibers from the melt at high throughput. Implementation is difficult in practice, however, because small, robust e-beam guns must be custom made, and the electron-permeable window must be strong enough to withstand the fluid pressure and thin enough to be highly transparent to electrons.

High vacuum Polymer fluid

Sealed tube Cathode

Nanofiber formation Electron transparent window

Voltage

1.11 Schematic representation of charge injection into a flowing polymer fluid by means of a miniature electron beam device.

© 2007, Woodhead Publishing Limited

20

1.5 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.

Nanofibers and nanotechnology in textiles

References Rayleigh, F.R.S., Phil. Mag. (1882) 44, p. 184 Zeleny, J., Phys. Rev. (1917) 10, p. 1 Taylor, G.I., Proc. R. Soc. London (1966) A291, p. 145 Taylor, G.I., Proc. R. Soc. London (1969) A313, p. 453 Schweizer, J.W. and Hanson, D.N., J. Colloid Interface Sci. (1971) 35, p. 417 Kornev, K. Kurtz, S. and Salem, D.R., To be published Pantano, C., Ganan-Calvo, A.M. and Barrero, A., J. Aerosol Sci. (1994) 25, p. 1065 Cloupeau, M. and Prunet-Foch, B., J. Electrostatics (1989) 22, p. 135 Fong, H. and Reneker, D.H. in Structure Formation in Polymeric Fibers, D.R. Salem (Ed) (2001) p. 225 Cloupeau, M. and Prunet-Foch, B., J. Aerosol Sci. (1994) 25, p. 1021 Hartman, R.P.A., Brunner, D.J., Camelot, D.M.A., Marijnissen, J.C.M. and Scarlett, B., J. Aerosol Sci. (2000) 31, p. 65 Fernandez de la Mora, J. and Loscertales, I.G., J. Fluid Mech. (1994) 260, p. 155 Ganan-Calvo, A.M., Davila, J. and Barrero, A., J. Aerosol Sci. (1997) 28, p. 249 Ganan-Calvo, A.M., Phys. Rev. Lett. (1997) 79, p. 217 Gomez, A. and Tang, K., Proc. 5th Int. Conf. on Liquid Atomization and Spray Systems (1991) p. 805 Magarvey, R.H. and Outhouse, L.E., J. Fluid Mech. (1962) 13, p. 151 Tsao, C.C., Lobo, J.D., Okumura, M. and Lo, S.Y., J. Phys. D: Appl. Phys. (1998) 31, p. 2195. Doyle, A., Moffett, D.R. and Vonnegut, B., J. Colloid Sci. (1964) 19, p. 136 Roth, D.G. and Kelly, A.J., IEEE Trans. Ind. Appl. (1983) IA-19, p. 771 Duft, D., Achtzehn, T., Muller, R., Huber, B.A. and Leisner, T., Nature (2003) 421, p. 128 Lopez-Herrera, J.M. and Ganan-Calvo, A.M., J. Fluid. Mech. (2004) 501, p 303 Fenn, J.B., Mann, M., Meng, C.K., Wong, S.F. and Whitehouse, C.M., Science (1989) 246, p. 64 Festag, R., Alexandratos, S.D., Cook, K.D., Joy, D.C., Annis, B. and Wunderlich, B., Macromolecules (1997) 30, p. 6238 Formhals, A., US Patent 2,349,950 (1944) Baumgarten, P.K., J. Colloid and Interface Sci. (1971) 36, p. 71 Bornat, A., US Patent 4,323,525 (1982) Hohman, M.M., Shin, M., Rutledge, G. and Brenner, M.P., Phys. Fluids (2001) 13, p. 2201 Hohman, M.M., Shin, M., Rutledge, G. and Brenner, M.P., Phys. Fluids (2001) 13, p. 2221 Reneker, D.H., Yarin, A.L., Fong, H. and Koombhongse, S., J. Appl. Phys. (2000) 87, p. 4531 Yarin, A.L., Koombhongse, S. and Reneker, D.H., J. Appl. Phys. (2001) 89, p. 3018 Fong, H. and Reneker, D.H., Polymer (1999) 40, p. 4585 Zong, X., Kim, K., Fang, D., Ran, S., Hsiao, B.S. and Chu, B. Polymer (2002) 43, p. 4403 Lee, K.H., Kim, H.Y., Bang, H.J., Jung Y.H. and Lee, S.G., Polymer (2003) 44, p. 4029 McKee, M.G., Wilkes, G.L., Colby, R.H. and Long, T.E., Macromolecules (2004) 37, p. 1760

© 2007, Woodhead Publishing Limited

Electrospinning of nanofibers and the charge injection method

21

35. Yu, J.H., Fridrikh, S.V. and Rutledge, G.C., Polymer (2006) 47, p. 4789 36. Fridrikh, S.V., Yu, J.H., Brenner, M.P. and Rutledge, G.C., Phys. Rev. Lett. (2003) 90, p. 144502-1. 37. Hou, H., Jun, Z., Reuning, A., Schaper, A., Wendorff, J.H. and Greiner, A., Macromolecules (2002) 35, p. 2429 38. Kessik, R., Fenn, J.B. and Tepper, G., Polymer (2004) 45, p. 2981 39. Sarkar, S. and Tepper, G., The Fiber Society Fall Annual Meeting and Technical Conference, October 2006, 5A 40. Duby, M.H., Deng, W., Kim, K., Gomez, T. and Gomez, A., J. Aerosol Sci. (2006) 37, p. 306 41. Deng, W., Klemic, J., Li, X., Reed, M. and Gomez, A., J. Aerosol Sci. (2006) 37, p. 696 42. Chu, B., Hsiao, B.S. and Fang, D., US Patent 6,713,011 (2004) 43. Burger, C., Hsiao, B.S. and Chu, B., Annu. Rev. Mater. Res. (2006) 36, p. 333 44. Theron, S.A., Yarin, A.L., Zussman, E. and Kroll, E., Polymer (2005) 46, p. 2889 45. Tomaszewski, W. and Szadkowski, M., Fibers Textiles E. Eur. (2005) 13, p. 22 46. Kim, H.Y., US Patent 6,991,702 (2006) 47. Zhou, H., Green, T.B. and Joo, Y.L., Polymer (2006) 47, p. 7497 48. Um, I.C., Fang, D., Hsiao, B.S., Okamoto, A. and Chu, B., Biomacromolecules (2004), 5, p. 1428 49. Wang, X.F., Um, I.C., Fang, D., Okamoto, A. and Chu, B., Polymer (2005) 46, p. 4853 50. Kelly, A.J., US Patents (a) 4,255,777 (1981), (b) 4,380,786 (1983), (c) 4,630,169 (1986), (d) 5,093,602 (1992), (e) 6,206,307 (2001), (f) 6,227,465 (2001), (g) 6,656,394 (2003), (h) 6,964,385 (2005) 51. Kelly, A.J., Aerosol Sci. Tech. (1990) 12, p. 526 52. Kelly, A.J., J. Aerosol Sci. (1994) 25, p. 1159 53. Lehr, W. and Hiller, W., J. Electrostatics (1993) 30, p. 433 54. Shrimpton, J.S. and Yule, A.J., Exp. Fluids (1999) 26, p. 490 55. Kornev, K., Kurtz, S. and Salem, D.R. To be published

© 2007, Woodhead Publishing Limited

2 Producing nanofiber structures by electrospinning for tissue engineering F. K. K O, The University of British Columbia, Canada and M. R. G A N D H I, Drexel University, USA

2.1

Introduction

Biocompatible and biodegradable polymeric biomaterials are used to develop biological matrices or scaffolds not only for tissue engineering but also for various biomedical applications including wound dressings, membrane filters and drug delivery. These materials include synthetic polymers such as poly(lactide-co-glycolide), poly-(lactic acid), poly-(glycolic acid) and poly(caprolactone) and natural biopolymers such as silk, mussel adhesive protein, keratin, elastin and collagen. The natural materials are of considerable interest due to their structural properties and superior biocompatibility. Electrospinning is a unique method capable of producing nanoscale fibers. We have utilized this technique in our laboratory to fabricate nanofibrous scaffolds from both synthetic as well as natural polymers. The strength of a biomaterial is very important for the design of scaffolds meant for biomedical applications. They should provide the initial framework for the cells to withhold the natural loading conditions. The architecture of an engineered scaffold is also important in the design of a synthetic tissue replacement. Thus scaffolds should mimic the mechanical and geometrical properties of the replacing tissue. We have generated various structures including nonwoven mesh, 3D braided fibers, microspheres and foams as scaffolds for tissue engineering.

2.1.1

Tissue engineering concept

Tissue engineering was identified by the United States National Science Foundation some ten years ago as an emerging area of national importance and defined as:1 ‘Tissue Engineering is the application of principles and methods of engineering and the life sciences towards the fundamental understanding of structure/function relationships in normal and pathological mammalian tissues and the development of biological substitutes to restore, maintain or improve functions.’ 22

© 2007, Woodhead Publishing Limited

Producing nanofiber structures by electrospinning

23

Fueled by the exciting progress made in biotechnology in recent years, tissue engineering is quickly becoming a method of choice for the development of implants in surgery. It is expected that it will become a viable option in the healthcare industry and it is on the verge of breaking into a rapid growth mode in the next decade. For example, in orthopedic reconstruction, surgeons often replace damaged tissue resulting from trauma, pathological degeneration or congenital deformity with autogenous grafts.1 Reconstructive surgery is based upon the principle of replacing these types of defective tissues with viable, functioning alternatives. The grafting of bone in skeletal reconstruction has become a common task of the orthopedic surgeon, with over 863 200 grafting procedures performed each year in the United States. For cartilage replacement, there are over 1 000 000 procedures of various types performed each year, and for ligament repairs, there are approximately 90 000 procedures performed per year.1 Currently, autograft2,3 (tissue taken from the patient) and allograft4–6 (tissue taken from a cadaver) are the most common replacement sources for the treatment of musculoskeletal problems. In repair of anterior cruciate ligament injuries, a segment of the patellar tendon has frequently been used.3 For cartilage and bone repair, transplantation of autogenous grafts has been the current treatment of choice. Unfortunately, these gold standards possess certain disadvantages. For any type of autogenous tissue, the key limitations are donor site morbidity where the remaining tissue at the harvest site is damaged by removal of the graft, and the limited amount of tissue available for harvesting. The use of allograft attempts to alleviate these problems. However, this type of graft is often rejected by the host body because of an immune response to the tissue. Allografts are also capable of transmitting disease. Although a thorough screening process eliminates most of the disease-carrying tissue, this method is not 100% effective.4 As a result of the limitations with conventional reconstructive graft materials, surgeons have looked to the field of tissue engineering for synthetic alternatives.1,6–12 As articulated succinctly by Professors Vacanti and Mikos,13 the key challenges in tissue engineering are the synthesis of new cell adhesionspecific materials and the development of fabrication methods to produce reproducible 3D synthetic or natural biodegradable polymer scaffolds with tailored properties. These properties include porosity, pore size distribution and connectivity, mechanical properties for load-bearing applications and rate of degradation. Scientists around the globe have tried to address these issues by fabricating a 3D surface having the properties equivalent to the replacing tissue. These artificial 3D matrices, called scaffolds, provide the structural integrity similar to the natural extracellular matrix in the body. The scaffold is then seeded with the cells taken from the patient’s normal tissue or from the donor. The cells used for seeding tissue engineered scaffolds can be stem cells or mature adult cells. The biochemical and/or mechanical

© 2007, Woodhead Publishing Limited

24

Nanofibers and nanotechnology in textiles

Patient biopsy

Tissue engineering scaffold

Isolated and expanded cells

Growth factors

Cells cultured Cell–scaffold on scaffold constructs in bioreactors

New tissue

2.1 Concept of tissue engineering.

signals are then provided for the differentiation of the cells into tissues. These signals are mostly in the form of growth factors. Ideally, the tissue will form and the scaffolds will degrade, leaving behind the regenerated tissue. Thus, the classic ‘triad’ of tissue engineering is based on the three basic tissue components: a scaffold on which cells are incorporated and signals provided to build and differentiate the tissue. Figure 2.1 describes the concept of tissue engineering.

2.1.2

Scaffolds for tissue engineering

It is well known that biological tissues consist of well-organized hierarchical fibrous structures14 ranging from nanometer to micrometer scale as described in Fig. 2.2. The successful regeneration of biological tissue and organs calls for the development of fibrous structures with fiber architectures conducive to cell deposition and cell proliferation. Of particular interest in tissue engineering is the creation of reproducible and biocompatible 3D scaffold for cell ingrowth, resulting in bio-matrix composites for repair and replacement of various tissues. A large family of fiber architectures is available for surgical implants (Fig. 2.3). The design and selection of these fiber architectures for tissue engineering can be carried out on the fiber and structural levels resulting in a wide range of dimensional scale, fiber tortuosity and fabric porosity as characterized by the fiber volume fraction-orientation map.15 Since the 1950s, many of the textile structures illustrated in Fig. 2.3 have been used for surgical implants with considerable success. Some of the structures and their respective applications are summarized in Table 2.1. These fibrous structures are mostly from non-absorbable polymers consisting of fibers larger than 10 µm in diameter. Unfortunately, these fibers were developed primarily for clothing rather than for medical applications. It is

© 2007, Woodhead Publishing Limited

Producing nanofiber structures by electrospinning

25

Evidence: X-ray EM X-ray

X-ray EM

Microfibril

X-ray EM SEM

Subfibril

SEM OM

EM SEM OM

Fibril Tendon Fascicle

Tropocollagen

3.5 nm staining sites

64 nm periodicity Waveform Fascicular Fibroblasts or crimp structure membrane

1.5 nm 3.5 nm 10–20 nm

50–500 nm

50–300 µm

Reticular membrane

100–500 µm

Size scale

2.2 Hierarchical fiber architecture of tendons (EM, electron microscope; SEM, scanning electron microscope; OM, optical microscope; see Ref. 14 for more details).

Biaxial woven

High modulus woven

Multilayer woven

Triaxial woven

Weft knit Weft knit Weft knit Weft knit laid in weft laid in weft laid in warp laid in warp

Warp knit

Warp knit Weft laid in inserted warp warp knit

Tubular braid

Tubular Flat braid Flat braid braid laid laid in in warp warp

Square 3D braid Square braid braid laid in warp

Fiber Weft mat inserted warp knit laid in warp

Stitchbonded Biaxial laid in warp bonded

3D braid laid in warp

XYZ laid in system

2.3 Fiber architecture for surgical implants (see Ref. 15 for more details).

necessary to build scaffolds that mimic the mechanical properties of the native tissue as they provide the basic 3D mechanical framework for the cells to attach to and proliferate before they can differentiate into a tissue. Not only the mechanical properties but also the hierarchical geometry of the

© 2007, Woodhead Publishing Limited

26

Nanofibers and nanotechnology in textiles

Table 2.1 Textile structures for surgical implants (see Ref. 16 for more details) Application

Material

Structure Yarn

Fabric

Arteries

Polyester Dacron 56 Teflon

Textured

Weft/warp knit Straight tube bifurcation Plain woven Straight tube, nonwoven

Tendon

Polyester Dacron 56 Kevlar

Low twist filament

Plain woven narrow tape coated with silicon rubber

Hernia repair

Polypropylene

Monofilament

Tricot jersey knit

Esophagus

Regenerated collagen

Multifilament

Plain weave

Heart valve

Polyester Dacron 56

Multifilament

Plain weave Knit

Patches

Polyester Dacron 56

Textured

Knitted velour

Sutures

Polyester Nylon Collagen Silk

Monofilament Multifilament

Braid Woven tapes

Ligament

Polyester Teflon Polyethylene

Multifilament

Braid

Bone and joints

Carbon in thermoset or thermoplastic matrix

Multifilament

Woven Braid

tissue plays a major role in engineering the design for scaffolds. In the natural tissues the cells are embedded in the extracellular matrix mainly made up of collagen and elastin. The scaffolds for the tissue engineering application should mimic the mechanical properties and hierarchical organization of the natural extracellular matrix. The scaffolds replacing a tissue should be fibrous as well as nanoscale to mimic the natural extracellular matrix. As stated earlier not only the mechanical properties but also the geometry of the extracellular matrix plays a role in scaffold fabrication. In order to take the 3D and nanofibrous nature of biological tissue into consideration, we have developed an integrated hierarchical design methodology for the manufacturing of scaffolds for tissue engineering. A schematic of the hierarchical structural design of a braided structure fabricated in our laboratory for anterior cruciate ligament is shown in Fig. 2.4.

© 2007, Woodhead Publishing Limited

Producing nanofiber structures by electrospinning

Fiber drawing effect

27

Crimp effect β

θ

Braid angle effect

Yarn twist effect

φ

Material property

Fiber and yarn packing

Braid helix angle

2.4 Structural hierarchy of fibrous assemblies.

2.1.3

Scaffold fabrication and electrospinning procedure

Several methods have been developed to fabricate highly porous biodegradable scaffolds, including fiber bonding, braiding, solvent casting, particle leaching, phase separation, emulsion freeze drying, gas foaming and 3D printing techniques. Using these methods in our laboratory, we have successfully made scaffolds with various geometries including sponges, microspheres, 3D braid and nanofibers. However the simplicity of the electrospinning process to generate nanofibers makes it an ideal process for scaffold fabrication. The further discussion in this chapter will be focused strongly on nanofibrous scaffold fabrication and characterization. In the next sections more specific examples on the use of nanofibrous scaffolds are mentioned followed by a brief description of various experiments conducted in our laboratory. The electrospinning technique is a process of generating ultrafine fibers in the nanometer to micrometer scale.17–19 In the electrospinning process an electric field is generated between an oppositely charged polymer fluid and a collection screen, the electrode. A polymer solution is added to a glass syringe with a capillary tip. An electrode is placed in the solution with another connection made to a metal screen. As the power is increased, the charged polymer solution is attracted to the screen. Once the voltage reaches a critical value, the charge overcomes the surface tension of the polymer cone formed on the

© 2007, Woodhead Publishing Limited

28

Nanofibers and nanotechnology in textiles

Reservoir containing polymer

Spinning distance Nanofibers Power supply Collecting plate

2.5 Schematic of the electrospinning process.

capillary tip of the syringe and a jet of ultrafine fibers is produced. As the charged fibers are splayed, the solvent quickly evaporates and the fibers are accumulated randomly on the surface of the collection screen. This results in a nonwoven mesh of nano- to micrometer scale fibers. A schematic drawing of the electrospinning process is shown in Fig. 2.5.

2.2

Fabrication of nanofibrous scaffolds

2.2.1

Polymeric nanofibers

Tissue engineering scaffolds should have the following characteristics: • porosity for cell migration; • balance between surface hydrophilicity and hydrophobicity for cell attachment; • mechanical properties comparable to natural tissue to withstand natural loading conditions; • degradation capability so that it gets completely reabsorbed after implantation; • nontoxic byproducts; • 3D matrix. Scientists around the globe have used various biocompatible and biodegradable synthetic polymeric biomaterials including polylactic (PLA) and glycolic acids (PGA), and their copolymers (PLAGA), polycaprolactone, polydioxanone, polyanhydrides, polyorthoester, polytrimethylene carbonate and polyphosphazene for scaffold fabrication. Various nondegradable materials such as polyester, polypropylene, polytetrafluoroethylene, polyethylene and polycarbonates are used as well. The main advantages of synthetic polymers

© 2007, Woodhead Publishing Limited

Producing nanofiber structures by electrospinning

29

are they are available in bulk and their properties are tailorable. However they lack cell recognition signals and sometimes their degraded products may be toxic. PLA, PGA and PLAGA are the most abundantly used polymers for scaffold fabrication. These are US Food and Drug Administration (FDA) approved polymers and their mechanical and degradable properties can be tailored by varying the ratio of PLA to PGA. We have electrospun a number of polymers including PLA, PGA and PLAGA in our laboratory. We have optimized the spinning process by varying the parameters such as choice of solvent, charge density, spinning distance, viscosity, concentration and molecular weight of polymer. Continuous, uniform and nanoscale fibers were produced by the electrospinning process. These electrospun nanofibrous scaffolds have a large surface area to volume ratio, high porosity and a variety of pore size distribution, giving all the important features necessary for ideal tissue engineering scaffolds.

2.2.2

Protein nanofibers

Natural materials such as collagen, elastin, keratin, silk, fibrin clot, chitosan and mussel proteins are ideal candidates for scaffold fabrication. Of particular interest are protein material from silkworm silk and spider silk because of their superior biocompatibility and unique mechanical properties of combined strength and toughness. The big advantages of natural materials over synthetic ones are favorable cell interaction and nontoxic degradation products. The disadvantages include limited supply and restricted design flexibility. Various proteins that we have electrospun are collagen, elastin and silk. Recently our laboratory has done extensive study on silkworm silk and spider silk to fabricate nanofibrous scaffolds.20–23 Silkworm silk fibers from Bombyx mori cocoons were utilized. The silk protein from silkworm silk contains two fibroin proteins held together by a glue-like protein called sericin. Sericin causes T-cell mediated hypersensitivity if introduced in the body. Sericin is removed from the cocoon fibers by a process called degumming. The degummed Bombyx mori silk fibers are biocompatible. Silkworm silk, unlike spider silk, is readily available from cocoons but one has to go through a procedure to make it spinnable by the electrospinning process. We have established a protocol to make a spinnable dope from Bombyx mori cocoon fibers. Briefly the procedure includes dissolving degummed Bombyx mori silk fibers in 50% aqueous CaCl2 and dialyzing it against deionized water. The dialyzed fibroin solution should then be frozen for 24 h at –20 °C and lyophilized to obtain regenerated sponge. This regenerated silk fibroin sponge can be dissolved in the appropriate solvent to carry out electrospinning. As described earlier, silkworm silk is available from cocoons and can be obtained by sericulture. Spider silk has better properties than silkworm silk but it is not possible to grow spiders in a farm because of their cannibalistic nature.

© 2007, Woodhead Publishing Limited

30

Nanofibers and nanotechnology in textiles

Dragline silk from major ampullate glands of the spider Nephila clavipes is a fibrous protein having crystalline regions of anti-parallel β-sheet interspersed with elastic amorphous segments. These two segments are represented by two different proteins, MaSp1 (Major Ampullate Spidroin 1) and MaSp2 (Major Ampullate Spidroin 2), coded by different genes. Nexia Biotechnologies Inc. in 1999 introduced an innovative technology of using a transgenic approach for large-scale production of spider silk. They produced recombinant spider silk, BiosteelR in the BELER (Breed Early Lactate Early) goat system. The milk produced by transgenic goats contained MaSp1 and MaSp2 proteins which can be isolated and purified to homogeneity. We have electrospun MaSp1 and MaSp2 in our laboratory to generate nanofibrous scaffolds.

2.2.3

Nanocomposite nanofibers

Sometimes it is necessary to have more than one material in nanofibers to mimic the structural and mechanical properties of the natural extracellular matrix. This can be done by a process called co-electrospinning, whereby blends of two different materials are electrospun to fabricate the scaffolds. Depending on the application one can make blends of different polymers or proteins or a combination of both. In our laboratory we have carried out extensive work on nanocomposite nanofibers fabricated from carbon nanotubes (CNTs). CNTs are a layer of graphite, one atom thick, rolled into a cylinder. CNT has a Young’s modulus in the order of 1 TPa.24–26 The toughness of CNT ranges from 6 to 30%. Incorporation of CNTs in polymeric and/or protein scaffolds not only improves the mechanical properties but gives unique electrical conductivity as well. We have fabricated co-electrospun scaffolds from various proteins, polymers and CNT.

2.3

Characterization of nanofibrous scaffolds

2.3.1

Porosity and pore size distribution

In the remaining sections we will discuss the findings of different experiments carried out in our laboratory to demonstrate various concepts. In one such experiment we fabricated three scaffolds from PLAGA. These were: (1) 150–300 µm PLAGA sintered spheres; (2) 3D braided structure consisting of 20 bundles of 20 µm filaments of PLAGA; and (3) electrospun nanofibrous scaffolds from PLAGA. The porosity of these scaffolds was characterized by mercury porosimetry measurement to characterize their pore size distributions. The detailed description of fabrication techniques of these scaffolds can be obtained from various papers.18,27–29 The concept we want to demonstrate here is the differences in porosity and pore size distribution of these scaffolds having different architectures. The pore size and distribution were characterized using a Micrometrics Autopore III porosimeter. As shown in Fig. 2.6, a

© 2007, Woodhead Publishing Limited

Producing nanofiber structures by electrospinning

31

Log differential intrusion vs diameter 1.1

+ Intrusion for cycle 1

1.0

Log differential intrusion (ml/g)

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0

50

100

150

200 250 300 Diameter (µm)

350

400

450

500

2.6 Pore size distribution of braided structure.

bimodal distribution of pores was observed for the 3D braided structures showing the existence of large interstitial pores of the order of 250 µm and interfiber pores having a diameter of the order of 30 µm. On the other extreme, as shown in Fig. 2.7, the nanofiber structures show a predominant concentration of pores with an average pore size of 14 µm. The pore surface of the nanofibrous structures was 0.823 m2/g, an order of magnitude greater than that in the 3D braided structure at 0.0045 m2/g. The sintered spheres show a single mode pore diameter distribution over the range of 200–600 µm as shown in Fig. 2.8, depending on the sphere diameter.

2.3.2

Morphology and fiber diameter distribution

In this section we will discuss our experiment with electrospun spider silk proteins. The MaSp1 and MaSp2 proteins obtained from Nexia Biotechnologies were dissolved in various ratios (1:0, 1:1, 1:3, 3:1, 0:1) in an appropriate solvent to prepare the spinning dope for electrospinning. The spinning dope was placed in a 3 ml syringe (18-G and spinning angle 45°). The tip-to-

© 2007, Woodhead Publishing Limited

32

Nanofibers and nanotechnology in textiles Log differential intrusion vs diameter + Intrusion for cycle 1

4.0

Log differential intrusion (ml/g)

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0.0 0

50

100

150

200 250 300 Diameter (µm)

350

400

450

500

2.7 Pore size distribution of nanofibrous structure.

collection plate (covered with aluminum foil) distance was varied from 5 to 10 cm. The electric field applied between the collecting plate (cathode) and the needle tip (anode) ranged from 1 to 6 kV/cm. The morphology of the gold sputtered electrospun fibers was examined and their diameters were determined by field emission environmental scanning electron microscope (Phillips XL-30 ESEM). The average fiber diameter and its distribution were determined based on 100 random measurements. Figure 2.9 shows the morphology and fiber diameter distribution of silk fibers at 10 cm spinning distance and electric field of 3 kV/cm. A 12% concentration of MaSp1 at these conditions produced continuous uniform fibers with diameter 100.7 ± 36.43 nm. MaSp2 was spinnable at these conditions but produced beads at 12% with fiber diameter of 28.7 ± 17.48. The 3:1 ratio of MaSp1 to MaSp2 was spinnable at 12% but for 1:1 and 1:3, the viscosity was too high and the concentration had to be reduced to 10% and 6% respectively to obtain continuous and uniform fibers. A 3:1 ratio at 12% concentration, 10 cm spinning distance and charge density of 3 kV/cm produced fibers with diameter

© 2007, Woodhead Publishing Limited

Producing nanofiber structures by electrospinning

33

Log differential intrusion vs diameter 1.4

MS-25–590 µm

1.3 1.2

MS 300–350 µm

Log differential intrusion (ml/g)

1.1 1.0 0.9

MS 180–212 µm

0.8 0.7

C:\Win9420\Data\0590_2.SMP C:\Win9420\Data\0355_2.SMP C:\Win9420|Data|0212_2.SMP

0.6 0.5 0.4 0.3 0.2 0.1 0.0 0

50

100

150

200 250 300 Diameter (µm)

350

400

450

500

2.8 Pore size distribution of sintered spheres.

of 81.1 ± 29.73 nm. A 1:1 ratio at 10% concentration, 10 cm spinning distance and electric field of 3 kV/cm generated 63.3 ± 25.17 nm size fibers and 1:3 ratio at 6% concentration, 10 cm spinning distance and electric field of 3 kV/ cm delivered fibers with diameter of 57.4 ± 25.75 nm. This experiment demonstrates the importance of optimization techniques that need to be carried out during electrospinning in order to generate continuous, uniform nanoscale fibrous scaffolds. The fiber diameter can be controlled by varying the processing parameters such as solution concentration, viscosity, applied charge and charge density, type of solvent employed, distance from tip of capillary to the collection plate, flow rate, diameter and angle of spin of the spinneret.

2.3.3

Tensile properties

We will again discuss our experiment with spider silk. For this experiment we co-electrospun spider silk protein and 1% CNT. Aligned nanofibers were generated using MaSp1 and 1% CNT. We calculated tensile properties of aligned MaSp1 nanofibers with and without CNT. The mechanical properties of fibers were determined by KES-G1 Kawabata micro tensile tester at the

© 2007, Woodhead Publishing Limited

bMaSp1: MaSp2 (3:1): 12% 10 cm 30 kV Average diameter: 81.1±29.73 nm

bMaSp2: 12% 10 cm 30 kV Average diameter: 28.7±17.48 nm

bbMaSp1: MaSp2 (1:1): 10% 10 cm 30 kV Average diameter: 63.3±25.17 nm

14

50

MaSp1: MaSp2 (1:3): 6% 10 cm 30 kV Average diameter: 57.4±25.75 nm

30

12 40

10

20

20

8 6

10

4

4 10

2 0 10 30 50 70 90 110 130150 170190 Fiber diameter (nm)

20 Frequency

6

30

Frequency

8

Frequency

Frequency

Frequency

10

0

0 10

30

50 70 90 110 130 Fiber diameter (nm)

0 10 30 50 70 90 110 130 150 170 190 Fiber diameter (nm)

0

10 30 50 70 90 110130150170 190 Fiber diameter (nm)

× 1000

× 5000

× 35000

2.9 Fiber morphology and diameter distribution of electrospun spider silk.

© 2007, Woodhead Publishing Limited

10

2

10 30 50 70 90 110 130 150 170 190 Fiber diameter (nm)

Nanofibers and nanotechnology in textiles

30

14 12

34

bMaSp1: 12% 10 cm 30 kV Average diameter: 100.7±36.43 nm

Producing nanofiber structures by electrospinning

35

elongation rate of 0.2 mm/s. The aligned fibers were rolled into a yarn. The 4 cm long sample was glued from both sides on a paper frame having 1 cm length. This gives us the gauge length of about 3 cm. These samples were then mounted on a Kawabata micro tensile machine and the tensile properties were measured from the average of five samples. The modulus of MaSp1 nanofibers was 123.29 MPa. The ultimate tensile stress was found to be 9.59 MPa and the elongation at break was 14.33%. The incorporation of 1% CNT improves the mechanical properties significantly. The modulus of MaSp1 with 1% CNT was 1004.36 MPa. The ultimate tensile stress increased to 40.74 MPa; however elongation at break was reduced to 7.39%. Figure 2.10 shows the stress–strain curve generated from the average of five samples. This experiment demonstrates the concept of co-electrospinning protein with CNT and thereby improving the mechanical properties. In spite of the significant improvement of mechanical properties of the pristine MaSp1 silk by the addition of 1% CNT, we feel that we still have not made use of the expected potential of CNT. Considering the modulus of MaSp1 as 100 MPa and CNT as 1 TPa, if we apply the rule of mixture for the properties of composite material then the modulus of MaSp1 with 1% CNT should be ~10 000 MPa. Our experimental results indicated that the modulus of the composite silk was 1000 MPa. This means that the silk could be made even stronger. We can improve the efficiency of property translation from the CNT to the silk matrix by improving the interfacial bonding between the CNT and the matrix. 45 MaSp1 + 1% CNT nanofibers

40 35

Stress (MPa)

30 25 20 15 MaSp1 nanofibers 10 5 0 0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

Strain

2.10 Tensile properties of aligned MaSp1 with and without CNT.

© 2007, Woodhead Publishing Limited

36

Nanofibers and nanotechnology in textiles

Further improvement can be made by improving the alignment of CNTs within the MaSp1 matrix and the alignment of the nanofibrils. A better dispersion of the CNT will facilitate the alignment of the CNT, and the functionalization of the CNT will promote a higher level of bonding between the CNT and the silk matrix.

2.3.4

Other specific characterization techniques

Characterization techniques to be employed for nanofibrous scaffolds is one of the most important and rapidly growing fields of tissue engineering. Morphology and fiber diameter distribution can be characterized by environmental scanning electron microscope. Porosity can be determined by mercury porosimeter. We need to know the mechanical properties of our scaffolds. Thus our scaffolds should pass through a battery of characterization before we can test them with cell culture. Apart from the characterization methods described here there are many other investigations that we need to carry out depending on the scaffolds. For example if we co-electrospin our material with CNT we need to know whether these CNTs are inside the fiber or on the surface. Raman spectroscopy provides a good way of determining this. Transmission electron microscopy can also be used to locate CNTs in the nanocomposite. The structure and composition of a material can be characterized by Fourier transform infrared spectrometry and X-ray diffraction study. Figure 2.11 shows the Raman spectroscopy of electrospun MaSp1 with and without CNT. The electrospun MaSp1 having 1% CNT showed characteristic Raman peaks of CNT between 195 and 270 cm–1, indicating radio breathing mode, and 1570 and 1593 cm–1, indicating tangential mode for CNT. This confirms successful incorporation of CNT in nanocomposite fibers.

2.4

Cell–scaffold interaction

To study the effect of CNT and fiber size on cell proliferation, silkworm silk natural micrometer sized fibers, silkworm silk nanofibers with and without CNT and spider silk nanofibers with and without CNT were used. Human chondrosarcoma cells (ATCC HTB94) were maintained in culture using DMEM (Dulbecco’s modified Eagle’s medium; Mediatech) supplemented with 10% fetal bovine serum (FBS), 1% penicillin-streptomycin and 1% L-glutamine. For cell seeding on scaffolds, cells from the tissue culture flasks were trypsinized for 5 minutes at 37 °C, neutralized with DMEM, centrifuged for 5 minutes at 1200 rpm and resuspended in DMEM. A sum of 1 000 000 cells was seeded per scaffold over 48 hours on a shaker at 37 °C. The culture medium was supplemented with ascorbic acid (40 µg/ml) on the first day. The cell–scaffold constructs were maintained in the culture environment and

© 2007, Woodhead Publishing Limited

Producing nanofiber structures by electrospinning Radial breathing mode 195

270

37

Disorder induced Tangential mode mode 1299

1570

1593

Intensity (arbitrary units)

Electrospun MaSp1 fibers

Electrospun MaSp1 + 1% CNT fibers

Pure CNT (in powder form) 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000

Raman shift (cm–1)

2.11 Raman spectra of electrospun MaSp1 fibers with and without CNT confirming successful incorporation of CNT in the composite fiber.

were studied for morphology and proliferation on days 3, 7 and 14. They were removed from the culture media on 3, 7 and 14 days. They were washed twice with 1× phosphate buffered saline (PBS) and then subjected to a gradient of ethanol (20%, 50%, 70%, 90%, 100%), each for 10 min. They were refrigerated overnight at 4 °C. The scaffolds were immediately coated with palladium. The morphology of cells on scaffolds was examined by Phillips XL-30 ESEM. A cell proliferation assay was carried out at 3, 7 and 14 days. Scaffolds being assayed were fed incubated for 2.5 h with serumfree media supplemented with 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide solution (MTT). The concentration used was 2.5 mg MTT/ml of 1× PBS. The scaffolds were then vortexed with 500 µl acidic isopropanol (0.04 M HCl in absolute isopropanol). The intensity of 200 µl of this solution was measured at 595 nm.

2.4.1

Co-electrospinning effect

Figure 2.12 shows the morphology of cells after 3, 7 and 14 days on spider silk nanofibers. Figure 2.13 shows the morphology on spider silk nanofibers with 1% CNT. On day 3 the cells were attached to the nanofibrous scaffolds

© 2007, Woodhead Publishing Limited

38

Nanofibers and nanotechnology in textiles

Day 3

Day 7

Day 14

2.12 Morphology of cells after 3, 7 and 14 days on spider silk nanofibers.

Day 7

Day 3

Day 14

2.13 Morphology of cells after 3, 7 and 14 days on spider silk nanofibers with 1% CNT.

which acted as a support, providing a 3D framework for cells. The cells migrated and proliferated over time and that was evident by the production of extracellular matrix. By day 7 there were more cells and they formed bridges over the nanofiber scaffolds. The entire scaffold was covered with the cells by day 14. Overall, there were fewer cells on the scaffolds having CNT but the morphology remained the same. Figure 2.14 shows the morphology of cells after 3, 7 and 14 days on silkworm nanofibers. Figure 2.15 shows the morphology on silkworm nanofibers with 1% CNT.

© 2007, Woodhead Publishing Limited

Producing nanofiber structures by electrospinning

Day 3

39

Day 7

Day 14

2.14 Morphology of cells after 3, 7 and 14 days on silkworm nanofibers.

Day 3

Day 7

Day 14

2.15 Morphology of cells after 3, 7 and 14 days on silkworm nanofibers with 1% CNT.

2.4.2

Size effect

The cells were able to proliferate over the micrometer-sized natural cocoon fibers but the morphology was significantly different (Fig. 2.16). The cells were not able to attach and spread well on microfibers. They were seen rotated around the fiber whereas on nanofibers they were attached over the number of fibers and were able to spread and make contacts with underlying

© 2007, Woodhead Publishing Limited

40

Nanofibers and nanotechnology in textiles

Natural silk fibers (micrometer size)

Nanofibers

2.16 Morphology of cells on natural silkworm silk fibers and nanofibers after 7 days in culture. MTT assay 0.25 A

Intensity

0.20

B

0.15

C

0.10

A B

D B C

0.05 A

C D E

0

F

F

F

3

E

D E

7 Days in culture

14

(A) Spider nanofibers

(B) Spider nanofibers +1% CNT

(C) Silkworm nanofibers

(D) Silkworm nanofibers + 1%CNT

(E) Natural silkworm silk fibers

(F) Blank (Control)

2.17 Cell proliferation (MTT) assay.

nanofibers. Figure 2.17 shows the result of MTT assay. It can be seen that the nanofibers had the maximum number of cells after 14 days in culture compared with the micrometer-sized fibers. Also, incorporation of CNT did not change the cell proliferation significantly. This study shows the importance of nanofibers over microfibers. The tissue engineering scaffolds act as an artificial extracellular matrix until the cells start producing their own. During

© 2007, Woodhead Publishing Limited

Producing nanofiber structures by electrospinning

41

this time cells should attach, migrate and proliferate over the scaffolds. The nanofiber scaffolds from silk have a greater advantage over the micrometersized natural fibers. The study also proved that nanofibers with 1% CNT supported cell attachment and proliferation.

2.4.3

Architecture effect

The three types of scaffolds described in Section 2.3.1 were used to study the cell–scaffold interaction on different architectures of scaffolds. Osteoblasts isolated from neonatal rat calvarias and grown to confluence in Ham’s F-12 medium (GIBCO), supplemented with 12% Sigma’s fetal bovine were seeded on all three scaffolds: (1) 150–300 µm PLAGA sintered spheres, (2) 3D braided structure consisting of 20 bundles of 20 µm filaments of PLAGA and (3) electrospun nanofibrous scaffolds from PLAGA. Briefly, cells were seeded on the UV sterilized PLAGA matrices at a density of 100 000 cells/ cm2. The osteoblasts were cultured on the scaffolds for durations ranging from 1 to 21 days. They were prepared according to established procedures by fixing in gluteraldehyde and dehydrated through a series of ethanol dilutions. The seeded cells were labeled with [3H]-thymidine and the thymidine uptake were measured at 1, 3, 7, 10 and 11 day intervals. The cell proliferation, as shown in Fig. 2.18, is expressed in terms of the amount of [3H]-thymidine uptake as a function of time. It can be seen that there is a consistent increase in cell population with time. The nanofibrous structure demonstrated the most cell growth whereas the tightly woven 3D braided structure showed the least proliferation. On the other hand, the cell growth on fused microsphere structure is between that of the 3D braid and the nanofibrous structures showing a surprising drop after the 10th day. Scanning electron microscopy pictures are shown in Fig. 2.19. It can be seen that, in responding to the large spheres wherein the cells are more than 10× smaller than the spheres (Fig. 2.19a), the cells tend to spread over the

Thymidine uptake (µCi)

0.007

3D braid

0.006

Microsphere

0.005

Nano nonwoven

0.004

Control

0.003 0.002 0.001 0 1

3

7 Time (days)

2.18 Cell proliferation of fibrous scaffolds.

© 2007, Woodhead Publishing Limited

10

11

42

Nanofibers and nanotechnology in textiles

(a) Microspheres

(b) 3D braid

(c) Nanofibers

2.19 Cell–fiber architecture interaction.

surface of the sphere before connecting to the adjacent spheres and eventually forming an interconnected cellular network. In the case of 20 µm filaments in unidirectional bundles and 3D braid (Fig. 2.19b) wherein the cells are about the same order of magnitude in dimension, the cell–matrix reaction appears to be similar. The cells tend to slide off the matrix at the moment of seeding. Those cells that remain on the surface of the substrates tend to grow around the filaments and bridge onto the adjacent filaments along the length. The most intensive cell deposition was seen in the nanofibrous structure (Fig. 2.19c). Extensive cell spreading was observed along the length of the fibrils and through the thickness.

2.5

Summary and conclusion

Tissue engineering is a rapidly growing field wherein cells either from the patient or a donor are seeded on an artificial scaffold made up of protein or polymer. It is a good alternative to traditional options. Nanoscale fibrous materials are the ideal candidates for tissue engineering scaffolds because of their high surface area to volume ratio, greater porosity and pore size distribution. Electrospinning is a simple method to generate nanofibrous scaffolds. The scaffolds from natural protein material have better biocompatibility as the degraded products are nontoxic. Silk protein has a clear advantage over other natural resources because of a unique combination

© 2007, Woodhead Publishing Limited

Producing nanofiber structures by electrospinning

43

of strength and toughness. One can incorporate carbon nanotubes in the scaffolds and increase mechanical properties and electrical conductivity without much harm to cell proliferation.

2.6

Acknowledgments

The research on PLAGA was supported in part by NIH Grant 5 F31 GM1890502, NIH Grant AR46117 and NSF Presidential Grant BES9553162/BES981782. The work of silk scaffolds was funded in part by Pennsylvania Nanotechnology Institute (NTI) and Taiwan Textile Research Institute (TTRI). We would like to thank Nexia Biotechnologies for providing transgenic spider silk.

2.7

References

1. R. Skalak, C.F. Fox and Y.C. Fung. Preface, in Tissue Engineering, edited by R. Skalak and C.F. Fox. Proceedings of NSF workshop on Tissue Engineering, Granlibakken, Lake Tahoe, California, Feb. 26–29, 1988. 2. R. Langer and J.P. Vacanti. Science 260, 920 (1993). 3. O.H. Friedman, J.M. Sherman, et al. Clin. Ortho. 196, 9 (1985). 4. G.E. Jackson, R. Windler and T.M. Simon. Amer. J. Sports Med. 18, 1 (1990). 5. A.R. Gadzag, J.M. Lane, D. Glaser and R.A. Forster. J. Amer. Acad. Ortho. Surg. 3, 1 (1995). 6. M. Shino, S. Inoue, et al. J. Bone Joint Surg. 70B, 556 (1988). 7. D.W. Jackson, J.T. Heinrich, M. Timothy and M.S. Simon. Arthroscopy 10, 442 (1994). 8. M.A. Attawia, J.E. Devin and C.T. Laurencin. J. Biomed. Mater. Res. 29, 843 (1995). 9. C.N. Cornell. Tech. Orthop. 7, 55 (1992). 10. D.W. Jackson, Ed. The Anterior Cruciate Ligament: Current and Future Concepts. Raven Press: New York; 1993. 11. C.B. Frank, S.L.-Y. Woo, T. Andriacchi, et al. Injury and Repair of the Musculoskeletal Soft Tissues, edited by S.L.Y. Woo and J.A. Buckwalter. American Academy of Orthopaedic Surgeons: Park Ridge; 1988, p. 45. 12. R.P. Lanza, R. Langer and W.L. Chick, Eds. Principles of Tissue Engineering. R.G. Landes Company and Academic Press, Inc.: San Diego, CA; 1997. 13. C.A. Vacanti and A.G. Mikos. Letter from the Editors, Tissue Engineering, Vol. 1, (1) (1995). 14. J. Kastelic, A. Galeski and E. Baer. J. Connective Tissue Res. 6, 11–23 (1978). 15. F.K. Ko. Ceramic Bull. February (1989). 16. F.K. Ko. Textile Asia April (1997). 17. A. Formhals. US Patent 1,975,504 (1934). 18. F.K. Ko, C.T. Laurencin, M.D. Borden and D. Reneker. ‘The dynamics of cell–fiber architecture interaction,’ Proceedings, Annual Meeting, Biomaterials Research Society, San Diego, April, 1998. 19. J. Doshi and D. Reneker. J Electrostatics 35, 151 (1995). 20. J. Ayutsede et al. Biomacromolecules 7, 208 (2006). 21. J. Ayutsede et al. Polymer 46, 1625 (2005).

© 2007, Woodhead Publishing Limited

44

Nanofibers and nanotechnology in textiles

22. S. Sukigara, M. Gandhi, J. Ayutsede, M. Micklus and F. Ko. Polymer 44, 5721 (2003). 23. S. Sukigara, M. Gandhi, J. Ayutsede, M. Micklus and F. Ko. Polymer 45, 3701 (2004). 24. J.P. Lu et al. Phys. Rev. Lett. 79, 1297 (1997). 25. A. Krishnan et al. Phys. Rev. B. 58, 14013 (1998). 26. E.W. Wong et al. Science 277, 1971 (1997). 27. F.K. Ko. Three dimensional fabrics for composites, in Textile Structural Composites, Edited by T.W. Chou and F.K. Ko. Elsevier: Amsterdam; 1989. 28. M. Borden, S.F. El-Amin, M. Attawia and C.T. Laurencin. Biomaterials 24, 597 (2003). 29. W.J. Li, C.T. Laurencin, E.J. Caterson, R.S. Tuan and F.K. Ko. J. Biomed. Mater. Res. 60, 613 (2002).

© 2007, Woodhead Publishing Limited

3 Continuous yarns from electrospun nanofibers E. S M I T, U. B Ü T T N E R and R. D. S A N D E R S O N, Stellenbosch University, South Africa

3.1

Introduction

Electrospinning is a simple but extremely versatile method for obtaining continuous nano- and microfibers of natural and synthetic polymers, as well as of inorganic oxide materials. The process itself has been discussed in the first two chapters. Additional information for the interested reader can be found in reviews on the topic.1,2 The focus of this chapter is on the manufacture of continuous yarns from electrospun fibers. Section 3.2 provides a brief overview of the potential applications of electrospun fiber yarns and some important textile terms related to this field. The importance of, and the underlying principles related to, controlling fiber orientation during electrospinning will then be discussed in Section 3.3. Section 3.4 describes how some of the principles of controlling fiber orientation during electrospinning have been applied for obtaining short or noncontinuous yarns. The different approaches taken to obtain continuous yarns from electrospun fibers are discussed in Section 3.5. Section 3.6 includes a discussion of some likely future trends in the field and a short review of some sources of further information is given in Section 3.7.

3.2

Using electrospun nanofibers: background and terminology

The three inherent properties of nanofibrous materials that make them very attractive for numerous applications are their high specific surface area (surface area/unit mass), high aspect ratio (length/diameter) and their biomimicking potential. These properties lead to the potential application of electrospun fibers in such diverse fields as high-performance filters, absorbent textiles, fiber-reinforced composites, biomedical textiles for wound dressings, tissue scaffolding and drug-release materials, nano- and microelectronic devices, electromagnetic shielding, photovoltaic devices and high-performance electrodes, as well as a range of nanofiber-based sensors. 45

© 2007, Woodhead Publishing Limited

46

Nanofibers and nanotechnology in textiles

In many of these applications the alignment, or controlled orientation, of the electrospun fibers is of great importance and large-scale commercialization of products will become viable only when sufficient control over fiber orientation can be obtained at high production rates. In the past few years research groups around the world have been focusing their attention on obtaining electrospun fibers in the form of yarns of continuous single nanofibers or uniaxial fiber bundles. Succeeding in this will allow the processing of nanofibers by traditional textile processing methods such as weaving, knitting and embroidery. This, in turn, not only will allow the significant commercialization of several of the applications cited above, but will also open the door to many other exciting new applications. Incorporating nanofibers into traditional textiles creates several opportunities. In the first instance, the replacement of only a small percentage of the fibers or yarns in a traditional textile fabric with yarns of similar diameter, but now made up of several thousands of nanofibers, can significantly increase the toughness and specific surface area of the fabric without increasing its overall mass. Alternatively, the complete fabric can even be made from nanofiber yarns. This has important implications in protective clothing applications, where lightweight, breathable fabrics with protection against extreme temperatures, ballistics, and chemical or biological agents are often required. On an aesthetic level, nanofiber textiles also exhibit extremely soft handling characteristics and have been proposed for use in the production of artificial leather and artificial cashmere. In biomedical applications the similarity between certain electrospun polymeric nanofibers and the naturally occurring nanofibrous structures of connective tissues such as collagen and elastin gives rise to the opportunity of creating artificial biomimicking wound dressings and tissue engineering scaffolds. Several studies on nonwoven nanofiber webs of biocompatible polymers such as poly(ε-caprolactone),3 poly(lactide-co-glycolide),4 poly(Llactic acid),5 collagen6 and regenerated silkworm silk7 have already shown potential in this area. Simple three-dimensional constructs for vascular prostheses have also been manufactured by electrospinning onto preformed templates.8 Although these initial studies show that enhanced cell adhesion, cell proliferation and scaffold vascularization can be obtained on porous, nonwoven nanofiber webs, the simplicity of the constructs and the fragile nature of nonwoven webs still limit their applicability to small areas. Creating complex three-dimensional scaffold structures with fibers aligned in a controlled fashion along the directions of the forces that are usually present in dynamic tissue environments, as for instance in muscles and tendons, will lead to significant improvements in the performance of tissue engineering scaffolds. With continuous nanofiber yarns it will become possible to create such aligned fiber structures on a large scale, simply by weaving. In addition, the age-old techniques of knitting and embroidery can then be applied to

© 2007, Woodhead Publishing Limited

Continuous yarns from electrospun nanofibers

47

create very complicated, three-dimensional scaffolds, with precisely controlled porosity, and yarns placed exactly along the lines of dynamic force. Several other fields will also benefit from the availability of continuous yarns from electrospun fibers. Owing to the high fiber-aspect ratios and increased fiber–matrix adhesion caused by the high specific surface areas, aligned nanofiber yarns can lead to stronger and tougher, lightweight, fiberreinforced composite materials. The incorporation of nanofiber-based sensors into textiles can lead to new opportunities in the fields of smart and electronic textiles. Aligned nanofiber yarns of piezo-electric polymers and other microactuator materials may lead to better performance in advanced robotics applications. Since the revival of electrospinning in the early 1990s, several research groups have worked on controlling the orientation of electrospun fibers. This is an important step on the road towards obtaining aligned nanofiber yarns and will be discussed in more detail in Section 3.3. A number of approaches to obtaining yarns from electrospun fibers have already been proposed in the open literature and will be discussed in greater detail in Sections 3.4 and 3.5. First, however, a brief discussion on terminology is required. Those who have worked in the field of electrospinning over the past decade have come from various disciplinary backgrounds, including physics, chemistry and polymer science, chemical and mechanical engineering, and also from the traditional textiles field. The result of this has been that literature on the topic of electrospinning, and especially yarns from electrospun fibers, is plagued with terminology from different disciplines, which often leads to misunderstanding and even self-contradictory statements. So, for instance, in a paper on the electrospinning of individual fibers of a novel polymer some authors might use the term yarn when they are actually referring to an individual fiber, or authors might refer to spinning a filament when they are actually spinning a yarn. In an attempt to avoid this kind of confusion, the authors propose the use of generally accepted terminology from the textiles industry. Some of the terms used in the following sections are defined below. • Fiber – a single piece of a solid material, which is flexible and fine, and has a high aspect ratio (length/diameter ratio). • Filament – a single fiber of indefinite length. • Tow – an untwisted assembly of a large number of filaments; tows are cut up to produce staple fibers. • Sliver – an assembly of fibers in continuous form without twist. The assembly of staple fibers, after carding but before twisting, is also known as a sliver. • Yarn – a generic term for a continuous strand of textile fibers or filaments in a form suitable for knitting, weaving or otherwise intertwining to form a fabric.

© 2007, Woodhead Publishing Limited

48

Nanofibers and nanotechnology in textiles

• Staple fiber – short-length fibers, as distinct from continuous filaments, which are twisted together (spun) to form a coherent yarn. Most natural fibers are staple fibers, the main exception being silk which is a filament yarn. Most artificial staple fibers are produced in this form by slicing up a tow of continuous filaments. • Staple fiber yarn – a yarn consisting of twisted together (spun) staple fibers. • Filament yarn – a yarn normally consisting of a bundle of continuous filaments. The term also includes monofilaments. • Core-spun yarn – a yarn consisting of an inner core yarn surrounded by staple fibers. A core-spun yarn combines the strength and/or elongation of the core thread and the characteristics of the staple fibers that form the surface. • Denier – a measure of linear density: the weight in grams of 9000 meters of yarn. • Tex – another measure of linear density: the weight in grams of 1000 meters of yarn.

3.3

Controlling fiber orientation

As stated in the previous section, achieving control over the orientation of electrospun fibers is an important step towards many of their potential applications. However, if one considers the fact that fiber formation occurs at very high rates (several hundreds of meters of fiber per second) and that the fiber formation process coincides with a very complicated three-dimensional whipping of the polymer jet (caused by electrostatic bending instability), it becomes clear that controlling the orientation of fibers formed by electrospinning is no simple task. Various mechanical and electrostatic approaches have been taken in efforts to control fiber alignment. The two most successful methods are the following: • Spinning onto a rapidly rotating surface – several research groups have been routinely utilizing this technique to obtain reasonably aligned fibers.9–12 The rapid rotation of a drum or disk and coinciding high linear velocity of the collector surface allows fast take-up of the electrospun fibers as they are formed. The ‘point-to-plane’ configuration of the electric field does, however, lead to fiber orientations that deviate from the preferred orientation. A special instance of the rotating drum set-up involves spinning onto a rapidly rotating sharp-edged wheel, which utilizes an additional electrostatic effect, since the sharp edge of the wheel creates a stronger converging electrostatic field, or a ‘point-to-point’ configuration, which has a focusing effect on the collected fibers (see Fig. 3.1). This in turn leads to better alignment of the fibers.13

© 2007, Woodhead Publishing Limited

Continuous yarns from electrospun nanofibers

49

Syringe Pendant drop

Envelope cone

Inverted envelope cone

θ = 26.6° 5mm θ 5mm

Sharpened edge

Rotating disk collector

Axis

3.1 Converging electrostatic field on sharp-edged wheel electrode. Reprinted from reference 1. Copyright (2003), with permission from Elsevier.

• The gap alignment effect – uniaxially aligned arrays of electrospun fibers can be obtained through the gap alignment effect, which occurs when charged electrospun fibers are deposited onto a collector that consists of two electrically conductive substrates, separated by an insulating gap. This electrostatic effect (see Fig. 3.2) has been observed by various groups.1,14,15 Recently this was investigated in more detail by Li et al.16–18 Briefly, the lowest energy configuration for an array of highly charged fibers between two conductive substrates, separated by an insulating gap, is obtained when fibers align parallel to each other.

3.4

Producing noncontinuous or short yarns

Both spinning onto a rapidly rotating collector and the gap alignment effect have been used to obtain short yarns for experimental purposes.

© 2007, Woodhead Publishing Limited

50

Nanofibers and nanotechnology in textiles

3.2 Aligned fibers obtained through the gap alignment effect. Reprinted from reference 14 and kindly supplied by R. Dersch. Copyright (2003), with permission from John Wiley & Sons Inc. Rings of aligned nanofibers

3.3 Aligned fiber tows on rotating disk collector. Reprinted with permission from reference 19 with kind permission from the author.

3.4.1

Rotating collector method

In work performed at Drexel University19 poly(ethylene oxide) (PEO) fibers were spun onto a rapidly rotating disk (Fig. 3.3), where the shearing force of the rotating disk led to aligned fibrous assemblies with good orientation. These oriented fibers could then be collected and manually twisted into a yarn (Fig. 3.4). Fennessey and Farris20 collected tows of aligned polyacrylonitrile (PAN) fibers using a rotating drum set-up. The tows, measuring ca. 32 cm × 2 cm, were then linked together and twisted using a Roberta-type electric twister.

© 2007, Woodhead Publishing Limited

Continuous yarns from electrospun nanofibers

51

3.4 Twisted yarns obtained from tows spun on a rotating disk collector. Reprinted with permission from reference 19 with kind permission from the author.

Twisted yarns of PAN nanofibers with twist angles of between 1.1° and 16.8° were prepared, with a denier between 326 and 618 and an average denier of 446. The stress–strain behaviour of the yarns was examined and the modulus, ultimate strength and elongation at the ultimate strength were measured as a function of twist angle.

3.4.2

Gap alignment method

Deitzel et al.21,22 made short yarns by quickly passing a wooden frame through the electrospinning jet several times (for up to an hour), in a process also known as ‘combing’, resulting in a tow of reasonably aligned fibers, which were then ‘gently twisted’ to form a yarn. Fong et al.23 also used the combing technique, by rapidly oscillating a grounded frame within the jet. Dalton et al.24 used a modified version of the gap method of alignment by electrospinning poly(ε-caprolactone) onto two parallel grounded rings that

© 2007, Woodhead Publishing Limited

52

Nanofibers and nanotechnology in textiles

2 cm

3.5 Fibers aligned between two parallel ring collectors. Reprinted from reference 24 and kindly supplied by P. D. Dalton. Copyright (2005), with permission from Elsevier.

2 µm 50 µm

3.6 Twisted poly(ε-caprolactone) yarn. Reprinted from reference 24 and kindly supplied by P. D. Dalton. Copyright (2005), with permission from Elsevier.

were placed 80 mm horizontally apart (Fig. 3.5), resulting in 1.26 µm diameter fibers, neatly aligned between the two rings. When one of the collection rings was rotated at high speed, a wound multifilament yarn with a diameter of less than 5 µm and a length of 50 mm was obtained (Fig. 3.6).

3.5

Producing continuous yarns

A common misconception in recent electrospinning literature is that the first literature on electrospinning dates back to 1934 when Formhals patented a

© 2007, Woodhead Publishing Limited

Continuous yarns from electrospun nanofibers

53

method for manufacturing yarns from electrospun fibers.25 Some of the first publications on electrospinning date back as far as 1902, when first Cooley26 and then Morton27 patented processes for dispersing fluids. In both patents, the authors describe processes for producing very fine artificial fibers by delivering a solution of a fiber-forming material, such as pyroxylin, a nitrated form of cellulose, dissolved in alcohol or ether, into a strong electric field. The reason for this oversight is unclear, but it could possibly be blamed on differences in terminology since the term ‘electrospinning’ has only become popular with the revival of the process in the mid-1990s. Another more puzzling oversight, which has recently led several authors to bemoan the lack of processes for making continuous yarns from electrospun fibers, is that Formhals actually registered a series of seven patents over a period of ten years between 1934 and 194425,28–33 and that all these patents describe processes and/or improvements to processes for the manufacture of continuous yarns from electrospun fibers. Since the youngest of these patents is more than 60 years old, one could speculate that these processes did not really work, which would explain the absence of commercially available electrospun fiber yarns. An alternative explanation could be that, since Formhals lived in Mainz, Germany, and since the last patent application was filed in 1940, the disruption of World War II and the ensuing years simply led to the processes being forgotten. Closer inspection of the patents, aided by more recent knowledge of the electrospinning process, also leads us to believe that at least some of the described processes are viable and that they deserve further consideration. The patents of Formhals show a gradual evolution of his yarn production process over time and in many instances he applied the same fundamental practical aspects of electrospinning that have re-occurred in the recent literature. These included obtaining aligned fibers by spinning onto conductive strips or rods that were separated in space from each other by an insulating material (gap alignment effect), increasing production rates by using multiple spinnerets, regulating the electric field between the source and the collector by adding additional electrostatic elements, using corona discharge to discharge the electrospun fibers, and post-treating the electrospun fibers by submerging them in a liquid bath. The fact that a recent patent34 (discussed in Section 3.5.2) closely resembles the first process patented by Formhals in 1934 also supports the viability of his patents and so our group is currently re-evaluating the processes disclosed in the patents; results will be forthcoming in future publications. For the sake of comprehensiveness, the Formhals processes are included in the following discussion. At present, there are 13 different methods for making continuous yarns from electrospun fibers discussed in the open literature. Formhals patented four of these methods between 1934 and 1944, while the other nine appeared only after 2001.

© 2007, Woodhead Publishing Limited

54

3.5.1

Nanofibers and nanotechnology in textiles

Rotating dual-collector yarn

Formhals’s original patent25 relates to the manufacture of slivers of cellulose acetate fibers by electrospinning from a cogwheel source onto various collector set-ups. In these collector set-ups fibers are first spun onto a rotating collector and then removed in a continuous fashion, onto a second take-up roller. The first of these collector set-ups consisted of a solid conductive wheel or ring with string attached to the edge. In this set-up the wheel was rotated for a short period while fibers were spun onto its edge. The process was then stopped, the string was loosened and then drawn over rollers and/or through twist-imparting rings to a second take-up roller, and the spinning process restarted. The newly spun sliver or semi-twisted yarn was then drawn off continuously onto the second take-up roller. Another collector set-up consisted of a looped metal belt with fixtures to push or blow the fiber sliver off the belt before the fiber sliver was collected on a second take-up roller. The concept of using multiple spinnerets for increasing production rates was also introduced in this patent. Formhals later identified several problems related to his first design and hence in subsequent patents he made various additions and/or alterations to the original design, which were intended to eliminate these problems. These problems and their solutions included the following: • Problem: Fibers flying to-and-fro between the source and the collector. Solution: In his second patent,28 Formhals claimed that one cause of the fibers flying to-and-fro between the jet and collector was that the collector was at too high a voltage of opposite polarity and that resulting corona discharge from the collector reversed the charge of the fibers while they were passing between the jet and the collector. This in turn caused them to change direction and fly back to the source. He proposed to eliminate this problem by adding a voltage regulator on the collector-side of the circuit in order to down-regulate the voltage of the collector. • Problem: Fibers not drying sufficiently between source and collector. Solution: Formhals later designed various additions to his spinning system for regulating the shape and intensity of the electrical field in the vicinity of the spinning source.30 This was done in order to direct the formed fibers along a longer, predetermined and constant path towards the collecting electrode and was achieved by placing, in close proximity to the fiberstream, conductive strips, wires, plates and screens, which were connected to the same potential as the fiber source. These additions allowed a more thorough drying of fibers before they were deposited on the collector. An additional problem, which is not specifically discussed in Formhals’s patents, but which can be foreseen when examining his first collector design,25 is that fiber alignment would be less than ideal when spinning onto a solid

© 2007, Woodhead Publishing Limited

Continuous yarns from electrospun nanofibers

55

wheel or belt. It appears, however, that Formhals did encounter this problem and that he overcame it by utilizing the gap alignment effect in the design of subsequent collectors. The design consisted of a picket-fence-like belt, with individual, pointed electrodes, separated from each other by an air gap.

3.5.2

Multi-collector yarn

In this patented process from the Korea Research Institute of Chemical Technology,34 continuous slivers or twisted yarns of different polymers, but especially of polyamide–polyimide copolymers, are claimed to be obtained by electrospinning first onto one stationary or rotating plate or conductive mesh collector, where the charges on the fibers are neutralized, and then continuously collecting the fibers from the first onto a second rotating collector. A diagram depicting the process is given in Fig. 3.7. The underlying principle of this process closely resembles the rotating dual-collector yarn process patented by Formhals in 1934.25

3.5.3

Core-spun yarn

In his 1940 patent,31 Formhals described a method for making composite yarns by electrospinning onto existing cotton, wool or other pre-formed yarn. It was also proposed that a sliver of fibers, such as wool, could be coated with the electrospun fibers before twisting the product into an intimately blended yarn. High voltage

Spinneret

Motor

x

First collector disk

Second collection roller

y Nanofiber yarn Motor

3.7 Multi-collector yarn process diagram. Reprinted from reference 34. Copyright (2005), with permission from Korea Research Institute of Chemical Technology.

© 2007, Woodhead Publishing Limited

56

3.5.4

Nanofibers and nanotechnology in textiles

Staple fiber yarn

Formhals developed a method for controlling the length of the electrospun fibers,29 with the main objective being to manufacture fibers with a controlled and comparatively short length. This goal was achieved by modulating the electric field using a spark gap. In this modulation, it was preferable to periodically switch the field strength to at least 35% and preferably 20% of its original voltage in order to interrupt the electrospinning process for a short period and thereby create a sliver of short fibers, which could then be spun into a staple fiber yarn.

3.5.5

Continuous filament yarn

Instead of spinning directly onto the counter-electrode, Formhals altered the process33 so that the polarity of the charge on the fibers was changed before reaching the counter-electrode. This was achieved by using high voltage of opposite polarity on a sharp-edged or thin wire counter-electrode. The high voltage led to corona discharge, which initially reduced and eventually inverted the charge on the fiber while it was travelling from the source to the collector. This caused the fibers to turn away from the collector electrode and they could then be intercepted at a point below the counter-electrode and rolled up as a continuous filament yarn on a take-up roller. In a final improvement on the system, the entire spinning apparatus was encased in a box with earthed conductive siding.32 This avoided build-up of charges in the panels, which could lead to disturbance of the electric field inside the spinning chamber and disruption of the spinning process. In addition, variable voltage power on the source and collector electrodes allowed the tuning and moving of the position of the neutral zone in which the yarn formation process takes place, which in turn allowed better control over the continuity of the spinning process.

3.5.6

Self-assembled yarn

The self-assembled yarn process was developed by Ko et al. at Drexel University.19, 35, 36 When a solution of pure PAN, or a PAN-containing polymer blend, was electrospun onto a solid conductive collector under appropriate conditions, the fibers did not deposit on the collector in the form of a flat nonwoven web as is usually observed. Instead, initial fibers deposited on a relatively small area of the collector and then subsequent fibers started accumulating on top of them and then on top of each other, forming a selfassembled yarn structure that rapidly grew upwards from the collector towards the spinneret. The formation of a self-assembled yarn is illustrated in Fig. 3.8. The self-assembling yarn, suspended in the space between the spinneret

© 2007, Woodhead Publishing Limited

Continuous yarns from electrospun nanofibers

57

3.8 Self-assembled yarn formation. Reprinted with permission from reference 19 with kind permission from the author.

and the collector, continued to grow in this fashion until it reached a critical point somewhere in the vicinity of the spinneret. At this critical point, a branched tree-like fiber structure formed and newly formed fibers deposited on the branches of the tree. The yarn could then be collected by slowly taking up the fibers collected on one of the tree branch structures, or by slowly moving the target electrode away from the spinneret. Post-processing of the yarn, including twisting, could be done in a second step. Ali36 proposed that the charge on the electrospun PAN fibers, which is induced through the high voltage in the spinneret, is dissipated through the evaporation of the solvent during the electrospinning process, so that the fibers are essentially neutral when they reach the collector electrode. This could explain why the initial fibers deposit on such a small area on the collector. If the fibers on the collector are charged, they repel incoming fibers leading to an expanding random web. Neutral fibers would not have the same repelling effect on incoming fibers and so the fibers would collect on a smaller area. Neutral fibers, deposited on top of each other, and therefore closer to the spinneret than the target electrode surface, also form an attractive

© 2007, Woodhead Publishing Limited

58

Nanofibers and nanotechnology in textiles

target for incoming fibers. This would explain why subsequent fibers selectively deposit on the tip of the self-assembling yarn.

3.5.7

Conical collector yarn

A method for the production of poly(ε-caprolactone) hollow fibers by the electrospinning process was reported by Kim et al.37 The conventional electrospinning device was modified to include a conical collector and an air-suction orifice to generate hollow and void-containing, uniaxially aligned electrospun fibers. Use of the conical collector allowed for the collection of aligned yarns with diameters of approximately 157 µm.

3.5.8

Spin-bath collector yarn

In this recently published method, developed by our group at Stellenbosch University,38 continuous uniaxial fiber bundle yarns are obtained by electrospinning onto the surface of a liquid reservoir counter-electrode. The web of electrospun fibers, which forms on the surface of the spin-bath, is drawn at low linear velocity (ca. 0.05 m/s) over the liquid surface and onto a take-up roller. A diagrammatic representation of the electrospinning set-up is given in Fig. 3.9. All the yarns obtained using this method exhibit very high degrees of fiber alignment (see Fig. 3.10) and bent fiber loops are observed in all the yarns. The process of yarn formation is illustrated in Fig. 3.11. It can be described in three phases. In the first phase, a flat web of randomly looped fibers forms

Power supply

+ 15kV Pasteur pipette spinneret Take-up roller

Jet

Nanofiber yarn

Shallow water bath

0 kV Ground electrode inside bath

3.9 Yarn-spinning set-up with grounded spin-bath collector electrode. Reprinted from reference 38. Copyright (2005), with permission from Elsevier.

© 2007, Woodhead Publishing Limited

Continuous yarns from electrospun nanofibers

59

10 µm

3.10 Electrospun fiber yarn of poly(vinyl acetate) showing high degree of fiber alignment.

Web formation on water surface

Web elongation and fiber alignment

Alligned fiber yarn Draw direction

Water bath Bath edge

3.11 Top view of the yarn formation process. Reprinted from reference 38. Copyright (2005), with permission from Elsevier.

on the surface of the liquid. In the second phase, when the fibers are drawn over or through the liquid, the web is elongated and alignment of the fibers takes place in the drawing direction. The third phase consists of drawing the web off the liquid and into air. The surface tension of the remaining liquid on the web pulls the fibers together into a three-dimensional, round yarn structure. The average yarn obtained in a single-spinneret electrospinning set-up contains approximately 3720 fibers per cross-section and approximately 180 m of yarn can be spun per hour. The yarns obtained are very fine, with calculated linear densities in the order of 10.1 denier. Figure 3.12 shows a comparison

© 2007, Woodhead Publishing Limited

60

Nanofibers and nanotechnology in textiles Spider silk fiber

Nanofiber yarn

Human hair

10 µm

3.12 Relative sizes of nanofiber yarn, spider silk fiber and human hair.

between the nanofiber yarn, a spider silk fiber and a human hair. Although higher linear densities can be obtained by reducing the yarn take-up rate, this is accompanied by a decrease in fiber alignment within the yarn. Currently investigations are focused on various options to overcome these challenges by, for instance, combining aligned yarns from multiple spinnerets into single yarns.

3.5.9

Twisted nonwoven web yarn

This method, patented by Raisio Chemicals Korea Inc., 39 involves electrospinning nanofibers through multiple nozzles to obtain a nonwoven nanofiber web, either directly in a ribbon form or in a larger form, which is then cut into ribbons, and subsequently passing the nanofiber web ribbons through an air twister to obtain a twisted nanofiber yarn. A diagram depicting the process is given in Fig. 3.13. Table 3.1 shows process variables and yarn characteristics for three polymer yarns that have been spun through this process.

3.5.10 Grooved belt collector yarn In a recent patent by Kim and Park,40 a ribbon-shaped nanofiber web is prepared by electrospinning onto a collector consisting of an endless belt-

© 2007, Woodhead Publishing Limited

Continuous yarns from electrospun nanofibers

61

(–) Air twister (+)

Multiple spinneret nozzle block

Nanofiber web

Nanofiber Drawing yarn rollers

Take-up roller

3.13 Spinning process used to prepare a twisted nonwoven web yarn. Reprinted from reference 39. Copyright (2005), with permission from inventor. Table 3.1 Process variables and yarn characteristics of twisted nonwoven web yarns Polymer

Poly(ε-caprolactone) 70% Polyurethane Mw = 80 000 (Mw = 80 000), 30% PVC (n = 800)

Nylon-6

Solvent

75:25 (v/v) methylene chloride/DMF

5:5 (v/v)

Formic acid

Concentration (wt%)

13

12.5

15

Unit width of web (cm)

2.5

60 cm wide web cut to 2 cm ribbons

1.8

Number of nozzles in block

800

400

1000

Throughput rate per nozzle (mg/min)

1.6

2.0

1.2

Yarn twist rate (turns/m)

60

45

80

Drawing elongation (times)

2

1.2

2

Yarn production rate (m/min)

64.2

30

50

DMF/THF

Fineness (denier)

75

120

75

Strength (g/denier)

1.3

1.4

3.0

Elongation (%)

32

50

36

DMF, N, N-dimethyl formamide; THF, tetrahydrofuran.

type nonconductive plate with grooves formed at regular intervals along a lengthwise direction and a conductive plate inserted into the grooves of the nonconductive plate. The nanofiber webs are electrospun onto the conductive

© 2007, Woodhead Publishing Limited

62

Nanofibers and nanotechnology in textiles Table 3.2 Characteristics of grooved belt collector yarns Polymer

Nylon

70% Polyurethane (Mw = 80 000), 30% PVC (n = 800)

Linear density (denier) Strength (g/denier) Elongation (%) Nanofiber diameter (nm)

75 4.5 42 186

75 3.4 45 480

plates in the grooves and later separated from the collector, focused, drawn and wound into a yarn. Table 3.2 shows yarn characteristics of two yarns made through this process.

3.5.11 Vortex bath collector yarn In this patent pending process developed at the National University of Singapore,41 a basin with a hole at the bottom is used to allow water to flow out in such a manner that a vortex is created on the water surface. Electrospinning is carried out over the top of the basin so that electrospun fibers are continuously deposited on the surface of the water. Owing to the presence of the vortex, the deposited fibers are drawn into a bundle as they flow through the water vortex. Generally, a higher feed rate or multiple spinnerets are required to deliver sufficient fibers on the surface of the water so that the resultant fiber yarn has sufficient strength to withstand the drawing and winding process. Figure 3.14 shows the set-up used for the yarn drawing process. Yarn drawing speeds as high as 80 m/min have been achieved and yarns made of poly(vinylidene fluoride) (PVDF) and polycaprolactone have been fabricated using this process. Figure 3.15 shows a PVDF yarn formed through this process. The diameter of the yarn is dependent on the amount of fibers deposited on the surface of the water, the diameter of the fibers that make up the yarn as well as the speed of water flowing through the hole.

3.5.12 Gap-separated rotating rod yarn This method developed by Doiphode and Reneker at the University of Akron42 utilizes the gap alignment effect in a very similar way to the work published by Dalton et al.24 discussed in Section 3.4. The process is described with reference to Fig. 3.16. Fibers are electrospun between a 2 mm metal rod on the right and a hollow 25 mm metal rod with a hollow hemisphere attached to its end on the left. Both the geometries are grounded and placed at a distance of a few centimeters. Fibers are collected across the gap between

© 2007, Woodhead Publishing Limited

Continuous yarns from electrospun nanofibers

63

High-voltage supply Spinneret

Water supply

Electrospun fiber mesh deposited on water

Hole at the bottom of basin Roller Electrospun fiber yarn Container for excess water

3.14 Vortex bath collector yarn process. Reprinted with permission and kindly supplied by W. E. Teo and S. Ramakrishna.

3.15 PVDF yarn obtained through the vortex bath collector process. Reprinted with permission and kindly supplied by W. E. Teo and S. Ramakrishna.

© 2007, Woodhead Publishing Limited

64

Nanofibers and nanotechnology in textiles Syringe filled with polymer solution and connected to positive potential

Hollow rotating metal rod with metal hemisphere at one end

Twisting zone

Ground metal rod Translation

3.16 Schematic diagram for gap-separated rotating rod yarn set-up. Figure kindly supplied by Sphurti V. Doiphode and Darrell H. Reneker.

these two collector surfaces and are given a twist by rotating the hemispherical collector. Yarn collected in this manner on the tip of the metal rod can be translated away from the rotating collector, thereby drawing the yarn and producing yarn continuously. Yarns with lengths up to 30 cm were produced by this method and the creators of the process believe that optimizing the winding mechanism can lead to production of continuous yarns. An electron microscope image of a Nomex® yarn spun through this process is given in Fig. 3.17.

3.5.13 Conjugate electrospinning yarn Methods for making continuous nanofiber yarns based on the principle of conjugate electrospinning were recently published and patented by Xinsong Li et al. at Southeast University in Nanjing43–45 as well as Luming Li and coworkers at Tsinghua University in Beijing.46,47 In conjugate electrospinning, two spinnerets or two groups of spinnerets are placed in an opposing configuration and connected to high voltage of positive and negative polarity respectively. The process is presented diagrammatically in Fig. 3.18. Oppositely charged fiber jets are ejected from the spinnerets and Coulombic attraction leads the oppositely charged fibers to collide with each other. The collision of the fibers leads to rapid neutralization of the charges on the fibers and rapid decrease in their flying speeds. In the processes described by both groups, the neutralized fibers are then collected onto take-up rollers to form yarns like the one depicted in Fig. 3.19. Each continuous yarn contains a large quantity of nanofibers, which are well aligned along the longitudinal axis of the yarn. Conjugate electrospinning works for a variety of polymers, composites and ceramics.

© 2007, Woodhead Publishing Limited

Continuous yarns from electrospun nanofibers

65

10 µm

3.17 Nomex® yarn made with the gap-separated rotating rod process. Figure kindly supplied by Sphurti V. Doiphode and Darrell H. Reneker.

3.18 Conjugate electrospinning set-up. Image kindly supplied by Xinsong Li.

© 2007, Woodhead Publishing Limited

66

Nanofibers and nanotechnology in textiles

3.19 Yarn obtained through conjugate electrospinning. Image kindly supplied by Xinsong Li.

3.6

Summary and future trends

With only a few exceptions, most of the open literature on yarns from electrospun fibers focuses on the process of yarn formation, rather than on the yarns obtained. Although the process is certainly an important aspect, researchers should bear in mind that ultimately the intended end-user of their results will be the fibers and textiles manufacturing industry. With this in mind, future research should pay more attention to the properties of the yarns obtained, and report more on these with specific focus on tenacity, elasticity and linear density values. To the best of our knowledge, there is currently no commercially available continuous nanofiber yarn produced through electrospinning. This is likely to change in the very near future and will lead to rapid worldwide evaluation of the product for numerous potential applications. This will also lead to evaluation of nanofiber yarn properties under ‘real world’ circumstances and results should indicate where further work is required. Although electrospinning is more than a hundred years old and processes for producing electrospun nanofiber yarns have existed for more than 60 years, little is known about the mechanical and other properties of nanofibers, and especially their twisted yarns. Some recent work has focused on some of the properties of twisted yarns of specific polymers, as discussed in Sections 3.4 and 3.5, but many of the unknowns still need to be investigated. There are certain drawbacks of the electrospinning process, such as low production rates, the requirement for proportionately large quantities of

© 2007, Woodhead Publishing Limited

Continuous yarns from electrospun nanofibers

67

industrial solvents, and the necessity for electrospinning through small, needlelike capillaries that tend to get blocked over time – through precipitation of small amounts of the spinning polymer at the capillary mouth. There are also certain challenges related to repeatably electrospinning core–shell and hollowcore fibers and obtaining sub-100 nm fiber diameters of all spinnable polymers without the formation of bead defects. Future work will undoubtedly be aimed at eliminating these drawbacks. Possible solutions to these problems could include multiple-jet, needleless spinning to overcome low production rates and needle blockage. The amounts of industrial solvents used in electrospinning could possibly be reduced or eliminated through application of supercritical fluid techniques or water-based emulsion chemistry. In the field of tissue engineering, more information on and a better understanding of the wettability and permeability of nanofiber yarns, as well as their structural properties as a function of biodegradation, should lead to the development of highly functional tissue scaffolds and wound dressings. Successful electrospinning of other materials, such as metals and non-oxide ceramics, and better control over the crystallinity of electrospun polymer fibers will lead to significant advances in nanofiber reinforced composite materials. Quality control methods for nanofiber yarn production processes will need to be developed for commercialization purposes, and other new problems, which will also arise once nanofiber yarns are spun on an industrial scale, will have to be dealt with. On a purely aesthetic level, once nanofibers are incorporated into wearable textiles, the question of coloration will arise. This might pose some problems, since fibers with diameters smaller than the optically visible wavelength range are seen through diffraction of light, not reflection, and therefore they usually appear white under normal circumstances.

3.7

Sources of further information and advice

As stated in the introduction, more general information on the topic of electrospinning can be obtained from some excellent review articles.1,2 The reader is also referred to a recently published book ‘An Introduction to Electrospinning and Nanofibers’ by Ramakrishna et al.48 The book gives an introduction to electrospinning and also a basic background to a wide range of subdisciplines, such as polymer science, rheology and electrostatics, that someone new to the field might want to learn more about. For more information on yarn terminology and definitions of textiles terms, the reader is referred to the Textiles Intelligence glossary of textiles terms (http://www.textilesintelligence.com/glo/). For more information on specific yarn formation processes discussed in the previous sections, the reader is referred to the specific publications or patents as cited.

© 2007, Woodhead Publishing Limited

68

3.8

Nanofibers and nanotechnology in textiles

References

1. Huang, Z.-M., Zhang, Y.-Z., Kotaki, M., Ramakrishna, S. ‘A review on polymer nanofibers by electrospinning and their applications in nanocomposites’, Composites Science and Technology, 2003, 63, 2223–2253. 2. Frenot, A., Chronakis, I. S. ‘Polymer nanofibers assembled by electrospinning’, Current Opinion in Colloid and Interface Science, 2003, 8, 64–75. 3. Yoshimoto, H., Shin, Y. M., Terai, H., Vacanti, J. P. ‘A biodegradable nanofiber scaffold by electrospinning and its potential for bone tissue engineering’, Biomaterials, 2003, 24, 2077–2082. 4. Zong, X., Ran, S., Kim, K. S., Fang, D., Hsiao, B. S., Chu, B. ‘Structure and morphology changes during in vitro degradation of electrospun poly(glycolide-colactide) nanofiber membrane’, Biomacromolecules, 2003, 4, 416–423. 5. Zong, X., Kim, K., Fang, D., Ran, S., Hsiao, B. S., Chu, B. ‘Structure and process relationship of electrospun bio-absorbable nanofiber membranes’, Polymer, 2002, 43(16), 4403–4412. 6. Huang, L., McMillan, R. A., Apkarian, R. P., Pourdeyhimi, B., Conticello, V. P., Chaikof, E. L. ‘Generation of synthetic elastinmimetic small diameter fibers and fiber networks’, Macromolecules, 2000, 33, 2989–2997. 7. Jin, H.-J., Fridrikh, S. V., Rutledge, G. C., Kaplan, D. L. ‘Electrospinning Bombyx mori silk with poly(ethylene oxide)’, Biomacromolecules, 2002, 3, 1233–1239. 8. Kidoaki, S., Kwon, Il. K., Matsuda, T. ‘Mesoscopic spatial designs of nano- and microfiber meshes for tissue-engineering matrix and scaffold based on newly devised multilayering and mixing electrospinning techniques’, Biomaterials, 2005, 26, 37– 46. 9. Jiang, H., Fang, D., Hsiao, B. S., Chu, B. and Chen, W. ‘Optimisation and characterization of dextran membranes prepared by electrospinning’, Biomacromolecules, 2004, 5, 326–333. 10. Bhattarai, S. R., Bhattarai, N., Yi, H. K., Hwang, P. H., Cha, D. Il., Kim, H. Y. ‘Novel biodegradable electrospun membrane: scaffold for tissue engineering’, Biomaterials, 2004, 25, 2595–2602. 11. Fong, H., Reneker, D. H. ‘Elastomeric nanofibers of styrene–butadiene–styrene triblock copolymer’, Journal of Polymer Science: Part B: Polymer Physics, 1999, 37, 3488–3493. 12. Laffin, C., McNally, G. M., Sanderson, R. D., Greyling, C. J. ‘The manufacture of aligned poly(acrylonitrile) fibres by electrospinning’, Proceedings of ANTEC 2005, Boston, pp. 1825–1829, 2005. 13. Zussman, E., Theron, A., Yarin, A. L. ‘Formation of nanofiber crossbars in electrospinning’, Applied Physics Letters, 2003, 82, 973–975. 14. Dersch, R., Liu, T., Schaper, A. K., Greiner, A., Wendorff, J. H. ‘Electrospun nanofibers: internal structure and intrinsic orientation’, Journal of Polymer Science Part B – Polymer Physics, 2003, 41, 545–553. 15. Katta, P., Alessandro, M., Ramsier, R. D., Chase, G. G. ‘Continuous electrospinning of aligned polymer nanofibers onto a wire drum collector’, Nano Letters, 2004, 4, 2215–2218. 16. Li, D., Wang, Y., Xia, Y. ‘Electrospinning of polymeric and ceramic nanofibers as uniaxially aligned arrays’, Nano Letters, 2003, 3, 1167–1171. 17. Li, D., Wang, Y., Xia, Y. ‘Electrospinning nanofibers as uniaxially aligned arrays and layer-by-layer stacked films’, Advanced Materials, 2004, 16, 361–366.

© 2007, Woodhead Publishing Limited

Continuous yarns from electrospun nanofibers

69

18. Li, D., Ouyang, G., McCann, J. T., Xia, Y. ‘Collecting electrospun nanofibers with patterned electrodes’, Nano Letters, 2005, 5, 913–916. 19. El-Aufy, A. K. ‘Nanofibers and nanocomposites poly(3,4-ethylene dioxythiophene)/ poly(styrene sulfonate) by electrospinning’, PhD Thesis, March 2004, Drexel University Online Library, URL: http://dspace.library.drexel.edu/handle/1860/282, Access date: 10 May 2005. 20. Fennessey, S. F., Farris, R. J. ‘Fabrication of aligned and molecularly oriented electrospun polyacrylonitrile nanofibers and the mechanical behaviour of their twisted yarns’, Polymer, 2004, 45, 4217–4225. 21. Deitzel, J. M., Kleinmeyer, J. D., Hirvonen, J. K., Beck Tan, N. C. ‘Controlled deposition and collection of electro-spun poly(ethylene oxide) fibers’, US Army Research Laboratory Report - ARL-TR-2415, 2001. 22. Deitzel, J. M., Kleinmeyer, J. D., Hirvonen, J. K., Beck Tan, N. C. ‘Controlled deposition of electrospun poly(ethylene oxide) fibers’, Polymer, 2001, 42, 8163– 8170. 23. Fong, H., Liu, W., Wang, C.-S., Vaia, R. A. ‘Generation of electrospun fibers of nylon 6 and nylon 6-montmorillonite nanocomposite’, Polymer, 2002, 43, 775–780. 24. Dalton, P. D., Klee, D., Möller, M. ‘Electrospinning with dual collection rings’, Polymer, 2005, 46, 611–614. 25. Formhals, A. ‘Process and apparatus for preparing artificial threads’, US Patent 1,975,504; 1934. 26. Cooley, J. F. ‘Apparatus for electrically dispersing fluids’, US Patent 692,631; 1902. 27. Morton, W. J. ‘Method for dispersing fluids’, US Patent 705,691; 1902. 28. Formhals, A. ‘Method and apparatus for the production of fibers’, US Patent 2,123,992; 1938. 29. Formhals, A. ‘Artificial fiber construction’, US Patent 2,109,333; 1938. 30. Formhals, A. ‘Method and apparatus for spinning’, US Patent 2,160,962; 1939. 31. Formhals, A. ‘Artificial thread and method for producing same’, US Patent 2,187,306; 1940. 32. Formhals, A. ‘Production of artificial fibers from fiber forming liquids’, US Patent 2,323,025; 1943. 33. Formhals, A. ‘Method and apparatus for spinning’, US Patent 2,349,950; 1944. 34. Lee, J.-R., Jee, S.-Y., Kim, H.-J., Hong, Y.-T., Kim, S., Park, S.-J. ‘Filament bundle type nano fiber and manufacturing method thereof’, PCT Application, WO 2005/ 123995 A1. 35. Ko, F., Gogotsi, Y., Ali, A., Naguib, N., Ye, H., Yang, G., Li, C., Willis, P. ‘Electrospinning of continuous carbon nanotube-filled nanofiber yarns’, Advanced Materials, 2003, 15, 1161–1165. 36. Abd El-Fattah Ali, A. ‘Carbon nanotube reinforced carbon nano composite fibrils by electro-spinning’, PhD Thesis, October 2002, Drexel University Online Library, URL: http://dspace.library.drexel.edu/handle/1860/17, Access date: 19 May 2005. 37. Kim, H. Y., Kim, K. W., Lee, K. H., Yoo, E.-S., Farris, R. J. Fennessey, S. F. ‘Electrospun hollow fibers of poly(e-caprolactone)’, Abstracts of Papers, 228th ACS National Meeting, Philadelphia, PA, United States, August 22–26, 2004. American Chemical Society, Washington, DC, 2004. 38. Smit, E., Bűttner, U., Sanderson, R. D. ‘Continuous yarns from electrospun fibers’, Polymer, 46, 2005, 2419–2423. 39. Kim, H.-Y. ‘A process of preparing continuous filament composed of nanofibers’, PCT Application, WO 2005/073442 A1.

© 2007, Woodhead Publishing Limited

70

Nanofibers and nanotechnology in textiles

40. Kim, H.-Y., Park, J.-C., ‘A process of preparing continuous filament composed of nanofibers’, PCT Application, WO 2006/052039 A1. 41. Teo, W. E., Ramakrishna, S. National University of Singapore, personal communication. 42. Doiphode, S. V., Reneker, D. H. University of Akron, personal communication. 43. Li, X., Sun, F., Yao, C., Song, T. ‘Conjugate electrospinning: continuous yarns from oppositely charged nanofibers’, Preprint of Polymeric Materials Science and Engineering (PMSE): Electrostatic Polymer Processing, 231st American Chemical Society National Meeting, Atlanta, GA, March 26–30, 2006. 44. Xinsong, L., Chen, Y., Tangyin, S. ‘Method for the preparation of continuous nanofiber yarns’, Chinese Patent No. CN1687493, 2005. 45. Xinsong, L., Chen, Y., Fuqian, S. ‘Apparatus and methods for the preparation of continuous nanofiber yarns’, Chinese Patent No. CN 1776033, 2005. 46. Pan, H., Li, L., Hu, L., Cui, X. ‘Continuous aligned polymer fibers produced by a modified electrospinning method’, Polymer 2006, 47, 4901–4904. 47. Li, L., Pan, H., Hu, L, ‘Device and method for electrospinning and fiber collecting’, Chinese Patent No. CN 1766181, 2006. 48. Ramakrishna, S., Fujihara, K., Teo, W., Lim, L., Ma, Z. An Introduction to Electrospinning and Nanofibers. Singapore, World Scientific Publishing, 2005.

© 2007, Woodhead Publishing Limited

4 Producing polyamide nanofibers by electrospinning M. A F S H A R I, R. K O T E K and A. E. T O N E L L I, North Carolina State University, USA and D.-W. J U N G, Hyosung Corporation, South Korea

4.1

Introduction

Conventional fiber spinning techniques such as wet spinning, dry spinning, melt spinning and gel spinning usually produce polymer fibers with diameters down to the micrometer range. If the fiber diameter is reduced from micrometers to nanometers, very large surface area to volume ratios are obtained and flexibility in surface functionalities and better mechanical performance may be achieved. These unique characteristics make polymer nanofibers optimal candidates for many important applications.1 Polymer fibers can be generated from an electrostatically driven jet of polymer solution or polymer melt. This process, known as electrospinning, has received a great deal of attention in the past decade because of its ability to consistently generate polymer fibers that range from 5 to 500 nm in diameter.2–5 The idea of electrospinning dates back more than 60 years. Formhals6–10 introduced electrospinning methods and described fiber formation during the spinning process. Because of the small pore size and high surface area inherent in electrospun textiles, these fabrics show promise for use in protective clothing for soldiers (to protect against extreme weather conditions, ballistics and nuclear, biological and chemical warfare), filtration applications, membranes, reinforcing fibers in composite materials, optical and electronic applications (piezoelectric, optical sensors), biomedical devices (cosmetics, skin healing and skin cleansing, wound dressing, drug delivery and pharmaceuticals, supports for enzymes or catalysts, scaffolds for tissue engineering, and templates for the formation of hollow fibers with inner diameters in the nanometer range.1, 11–19

4.2

The electrospinning process

Figure 4.1 shows a typical electrospinning apparatus. A high electrical potential, typically 5–30 kV, is applied to a polymer solution contained in a syringe. In electrospinning the tensile force is generated by the interaction of an applied electric field with the electrical charge carried by the jet rather than by the 71

© 2007, Woodhead Publishing Limited

72

Nanofibers and nanotechnology in textiles

Splay

Collect

Base Jet

HV

4.1 Schematic of electrospinning apparatus.

spindles and reels used in conventional spinning.12 A stable electrospinning jet has four distinct regions (see Fig. 4.1). The jet emerges from the charged surface at the base region, travels through the jet region, divides into many fibers in the splaying region, and stops in the collection region. The base is the region where the jet emerges from the liquid polymer. The geometry of the jet near the base is a tapered Taylor cone in which the axial velocity of the liquid increases as the polymer is accelerated along the axis of the jet. The base may have a circular cross-section, or it may have some other shape if the surface tension of the liquid anchors the jet to the lip of a hole or some other stationary object. An electric field at the surface of a liquid produces a force that, if the electric field is strong enough, causes a jet of liquid to be ejected from a surface that was essentially flat before the field was applied. The jet is the region beyond the base where the electrical forces continue to accelerate the polymer liquid and then stretch the jet. In this region, the diameter of the jet decreases and the length increases in a way that keeps constant the amount of mass per unit time passing any point on the jet axis. Larrondo and Manley,20, 21 through analysis of the flow field in an electrically driven jet, showed that the flow is a combination of parabolic and purely extensional flow. It was found, however, that the region about the symmetry axis of the jet is free of rotational components and is thus an area of pure extensional flow.21 Splaying occurs in a region in which the radial forces from the electrical charges carried by the jet become larger than the cohesive forces within the jet, and the single jet divides into many charged jets with approximately equal diameters and charge per unit length. The collection region is where the jet is stopped. The polymer fibers that remain after the solvent has evaporated may be collected on a metal screen. The initiation and formation of the jet constitute a complex process with many variations.12 In the electrospinning process the morphology of the fibers depends on various parameters, such as solution concentration, applied electric field

© 2007, Woodhead Publishing Limited

Producing polyamide nanofibers by electrospinning

73

strength and tip-to-collector distance.22–24 Although research has provided much fundamental understanding of the process, several difficulties remain. Many parameters can influence the transformation of polymer solutions into nanofibers through electrospinning. These parameters include: (a) solution properties, such as viscosity, elasticity, conductivity and surface tension; (b) governing/operating variables, such as hydrostatic pressure in the capillary tube, electric potential at the capillary tip and the gap (distance between the tip and the collecting screen); and (c) ambient parameters such as solution temperature, humidity and air velocity in the electrospinning chamber.14 Sometimes electrospun fibers exhibit bead-on-string structures, which have been generally considered to be undesirable by-products or defects. Theoretical analyses in the literature predicted three types of instabilities for an electrically driven jet: the axisymmetric Rayleigh instability, the electric field-induced axisymmetric instability and whipping. The process of bead formation revealed that the formation of a beaded structure resulted from axisymmetric instabilities and flow of the electrospun jet. Applied voltage, solution surface tension and conductivity influenced the formation of beaded electrospun fibers.25 Zuo and coworkers showed for poly(hydroxybutyrateco-valerate) (PHBV) electrospun fibers that higher applied voltage favors formation of smooth fibers, and beads are likely to be formed at high solution feed rates. High surface tension promotes the formation of PHBV electrospun fibers with beads, whereas increased conductivity achieved with mixed solvents favors uniform smooth fibers.25

4.3

Properties of electrospun nanofibers

It is difficult to measure mechanical properties of each electrospun single nanofiber with existing test techniques, because of their very small diameters. Therefore, mechanical tests were performed instead on nano-scale nonwoven webs with conventional testing methods. Lee et al.26 investigated the mechanical behavior of electrospun fiber webs of poly(vinyl chloride)/ polyurethane (PVC/PU) polyblends with different blending ratios. Ding and coworkers27 also tested mechanical properties of biodegradable nanofibrous mats comprising poly(vinyl alcohol)/cellulose acetate (PVA/CA), which were prepared by using multi-jet electrospinning methods. Mechanical properties of electrospun PU nanowebs were investigated by Pedicini and Farris.28 Among the many electrospun polymers reported in the literature are poly(pphenylene terephthalamide), tri-block copolymers, polyethylene oxide and DNA from solution; and polyethylene and polypropylene from the melt.2, 29 Nylon was the first commercialized synthetic fiber and is used throughout the world in many applications. It has been widely used as an important engineering plastic and synthetic fiber because of its good mechanical properties. It has been produced by traditional methods such as melt, wet and

© 2007, Woodhead Publishing Limited

74

Nanofibers and nanotechnology in textiles

dry spinning and is available in staple, tow, monofilament and multifilament forms.30 Fiber diameters produced by these methods range from 10 to 500 µm.31 Ryu et al.32 examined morphology, pore size, surface area and gas transport properties of nylon-6 (N6) nonwoven electrospun mats. The crystallinity of nanocomposites of nylon-6 and nylon-6–montmorillonite was studied by Fong et al.33 The ultra-large draw ratio and rapid solvent removal of electrospinning favors the formation of γ-phase nylon crystallites in pure N6 and montmorillonite–N6 fibers.33 Bergshoef and Vancso13 prepared nanocomposites with ultrathin, electrospun nylon-4,6 fibers and compared mechanical properties of nylon-4,6/epoxy composite films and epoxy films. In an effort to understand the mechanism of jet formation from polymer melts Larrond and Manley34 studied molten polymers of nylon-12 and polyethylene with the aid of an electric field. The drop formation was measured as a function of field intensity and frequency. There is qualitative agreement between the theory of Torza and experimental observation. Schreuder-Gibson et al.35 used nylon-6,6 (N6,6), polybenzimidazole (PBI) and poly(tetrafluoroethylene) membranes produced from electrospun fibers as protective layers. They measured properties of these electrospun membranes, including structural effects upon moisture transport, air convection, aerosol filtration, porosity and tensile strength. Stephen et al.36 using Raman spectroscopy showed that in the case of N6, the polymer crystalline structure was altered from α to γ form when electrospun. This, however, is not a permanent morphological conformational change and can be converted back to α form by solvent casting a film from the electrospun membrane. The ability of the electrospinning process to produce the γ form implies that the fibers are under high stress when they are being formed. Nylon-12 has only one preferred conformation, and the chain conformation is conserved after processing. Supaphol and coworkers37–39 studied the effects of electrode polarity and processing parameters (concentration, molecular weight, electrostatic field strength, solution temperature, addition of an inorganic salt and solvent system) on morphological appearance and size of the as-spun N6 fibers. An increase in the temperature of the spinning solutions decreased the size of the as-spun fibers. Addition of NaCl and increasing its concentration caused the conductivity of the spinning solutions to increase, which, in turn, caused the sizes of the as-spun fibers to increase. Fibers obtained from N6 of higher molecular weights appeared to be larger in diameter. An increase in the temperature of the solution during electrospinning resulted in a decrease in the fiber diameters. Increasing solution viscosity resulted in a reduced number of beads and increased fiber diameters. Diameters of fibers obtained under the negative electrode polarity were larger than those obtained under the positive electrode polarity. Dersch et al.40 showed that the intrinsic structure of N6 and poly(lactic acid) fibers do not differ to an appreciable extent from those found for much

© 2007, Woodhead Publishing Limited

Producing polyamide nanofibers by electrospinning

75

thicker fibers obtained by melt extrusion. The annealing of polyamide fibers at elevated temperatures resulted in a transformation from the disordered γ from to the more highly ordered α form. The orientation of the crystals along the fiber axis was strongly inhomogeneous, and on average, was very weak. In the literature there is little information on the effects of concentration, molecular weight, type of solvent, voltage and other parameters in the electrospinning process on jet stability and diameter of nylon nanofibers. In Section 4.2 we investigated the effects of voltage, distance between collector and tip of syringe, and solution concentration of N6 in formic acid on the diameters of electrospun fibers of N6 formed with stable solution jets. In Section 4.4, N6,6 nanofiber webs with different molecular weight were formed via the electrospinning method. The N6,6 solutions were produced by using the binary solvent formic acid/chloroform. Mechanical properties of two widely different molecular weight electrospun N6,6 nanowebs are compared by using conventional test methods. The main objective of this part of the study was to determine whether the use of high molecular weight N6,6 is a viable approach to improve the mechanical properties of electrospun nylon filaments.

4.4

Measuring the effects of different spinning conditions and the use of high molecular weight polymers on the properties of electrospun nanofibers

N6 chips of fiber grade from Parsilon Co., Iran, with relative viscosity 2.5– 2.6 in H2SO4 (molecular weight around 17 000 g/mol) were used. Various polymer solution concentrations ranging from 5, 6, 8 and 9 wt% were prepared by dissolving N6 in formic acid (Merck). High molecular weight N6,6 (MW = 170 000 g/mol) was prepared by using solid state polymerization of commercial N66 chips (Zytel®101, MW = 30 000 g/mol) that were supplied by Du Pont Co. Solid state polymerization of N6,6 was conducted under 0.4–5 torr at 255 °C for 8 hours, with the detailed procedure described in our previous paper.41 Ninety per cent formic acid and chloroform were obtained from Aldrich Co.

4.4.1

Viscosity of nylon 6,6 polymer solutions

The different concentration polymer solutions for electrospinning were prepared by using a solvent mixture of 90% formic acid and chloroform with the ratio of 75/25 (v/v). Solution viscosity was determined at 25 °C using a DV-II viscometer from Brookfield Co., USA.

© 2007, Woodhead Publishing Limited

76

4.4.2

Nanofibers and nanotechnology in textiles

Electrospinning conditions

For electrospinning of N6, a variable high-voltage power supply was used to apply voltages of 5 and 30 kV to the vertically oriented syringe tip. The polymer solution was placed in a 5 ml syringe to which a capillary tip of 0.8 mm inner diameter was attached. The positive electrode of the high-voltage power supply is connected to a copper wire immersed in the polymer solution. The negative electrode was connected to a metallic collector wrapped with aluminum foil. The process used in the electrospinning of N6,6 fibers is illustrated in Fig. 4.2. There are three components needed to fulfill the process: a high-voltage supply, a capillary tube with a needle of small diameter, and a metal collecting screen. A high-voltage power supply was used for creating an electrically charged polymer solution in electrospinning with voltages ranging from 0 to 30 kV. One electrode is attached to the metal capillary tip with 0.89 mm diameter and the other attached to the collector, a rotating cylindrical drum. The drum is 20 cm in diameter and 35 cm long and is covered with aluminum foil. The solution jet evaporates or solidifies and is collected as an interconnected web of small fibers. Electrospun nanowebs were produced by spinning N6,6 solutions from a 10 ml syringe with a back-pressure applied volume flow rate ranging from 90 µl/min to 0.040 ml/min.

4.4.3

Web morphology

The morphologies of electrospun low and high molecular weight N6,6 and N6 webs were observed by scanning electron microscopy (SEM, Hitachi Co., Japan) and their fiber diameters were measured with an Image J analyzer.

Pumping controller

Motor

High-voltage power source

4.2 Schematic of electrospinning process used in this study.

© 2007, Woodhead Publishing Limited

Producing polyamide nanofibers by electrospinning

4.4.4

77

Mechanical behavior

The tensile behavior of the electrospun low and high molecular weight N6,6 webs was tested on an Instron 5500R with a cross-head speed of 7 mm/min at room temperature. Rectangular-shaped tensile specimens (1 cm × 3 cm) were prepared for tensile tests. At least ten samples were tested for each fiber web.

4.5

Improving the properties of electrospun nanofibers: experimental results

It is well known that the morphologies of electrospun fibers depend on various processing parameters and environmental conditions, such as temperature and humidity.11, 12, 22, 23 Optimal electrospinning conditions, such as polymer concentration, applied electric field strength and tip-tocollector distance, were examined. It was not easy to create fibers under some electrospinning conditions. First, the solution viscosity (controlled by the solution concentration) was too low to make fibers from solutions containing less than 3–4 wt% N6. Generally, the beads-on-string morphology produced at low concentrations was regarded as an undesirable by-product of electrospinning. The viscosity, net charge density and surface tension of solution are key parameters in the formation of the beads-on-string.24 Second, the solvent was not completely evaporated owing to short tip-to-collector distances. With increased tip-to-collector distances, electrospinning did not facilitate the collection of fibers at the metallic collector, because of a relatively low electric field formed between the capillary tip and collector. The optimum tip-to-collector distance also depends on concentration of polymer solution, because as the solution concentration increased, so did the viscosity and surface tension of solution. Thus we needed to increase the voltage to overcome the solution surface tension to achieve splaying of the initial single jet into many charged jets. The type of solvent used for dissolving the polymer also has an effect on viscosity and surface tension. Figure 4.3 shows the image of a nonwoven mat of electrospun N6 fibers on a foil of aluminum. The SEM micrographs in Figs 4.4 and 4.5 show some beads in electrospun N6 fibers. The results showed that with increasing applied voltage formation of beads decreased. The beads can be removed by controlling solution viscosities and electrical conductivities of the polymer solution.22, 24 Figure 4.6 shows the distribution of electrospun (5 kV) N6 fiber diameters. The average fiber diameter increased with increasing concentration of solution. Figure 4.7 shows that increasing voltage (30 kV) decreased average diameter of electrospun fibers. By increasing the voltage, the applied electric field for splaying the single jet increased producing many jets and the diameters of the electrospun fibers decreased.

© 2007, Woodhead Publishing Limited

78

Nanofibers and nanotechnology in textiles

4.3 Nonwoven mat of N6 electrospun fibers on aluminum foil (8 wt% in formic acid at 10 kV).

4.4 Scanning electron micrograph of N6 fibers spun from 6 wt% in formic acid solution at 5 kV.

© 2007, Woodhead Publishing Limited

Producing polyamide nanofibers by electrospinning

79

4.5 Scanning electron micrograph of N6 fibers spun from 8 wt% at 10 kV.

Frequency distribution (%)

60 6 wt% 9 wt%

50 40 30 20 10 0

0–100

100–200 200–300 300–400 Diameter (nm)

4.6 Effect of solution concentration on diameter distribution of electrospun N6 fibers (voltage 5 kV).

The other important factor to the continuous production of electrospun N6 fibers is the distance between the tip of the syringe needle and the collector. To achieve splaying we needed to have an optimum distance between tip and collector. With increasing solution concentration of N6, the tip-tocollector distance has to increase. Tip-to-collector distances for 6 and 8 wt% solutions were about 3 and 5 cm, respectively. The fiber diameter generally decreased as the tip-to-collector distance increased. This is probably due to

© 2007, Woodhead Publishing Limited

80

Nanofibers and nanotechnology in textiles

Frequency distribution (%)

50 5 wt% 9 wt%

40 30 20 10 0

0–100

100–200 200–300 Diameter (nm)

300–400

4.7 Distribution of diameter of N6 electrospun fibers.

Table 4.1 Viscosity of nylon-6,6 solutions Molecular weight of nylon-6,6 (g/mol)

Concentration (%)

Viscosity (cP)

30 000 30 000 170 000 170 000

10 15 3 5

350 950 340 1000

splaying of the single jet into more charged jets or electrospun fibers. With a short tip-to-collector distance, the solvent cannot evaporate completely and the stretching of the jet is reduced.

4.5.1

Viscosity of nylon 6,6 solutions

It is well known that many parameters, such as viscosity, elasticity, conductivity, surface tension and distance between tip and collection screen, can influence the transformation of polymer solutions into nanofibers through electrospinning. Solution viscosity is one of the most important factors. Since both polymers have significantly different molecular weights, four different solutions of N6,6 with various concentrations were prepared. We used a mixed solvent consisting of formic acid and chloroform with the ratio of 75/25 (v/v). Table 4.1 shows the solution viscosities for each concentration of low and high molecular weight N6,6. N6,6 with average molecular weight of 30 000 g/mol (called low molecular weight N6,6) was used to prepare two polymer solutions at concentrations of 10% and 15%. The corresponding Brookfield viscosities were 350 and 950 cPs, respectively. In order to prepare fiber-forming solutions, the concentrations

© 2007, Woodhead Publishing Limited

Producing polyamide nanofibers by electrospinning

81

of N6,6 with average molecular weight 170 000 g/mol (called high molecular weight N6,6) were significantly lower. Correspondingly, the Brookfield viscosities were 340 and 1000 cPs for these solutions at 3% and 5%. Interestingly, as can be seen in Table 4.1, a 3% solution of high molecular weight N6,6 has comparable viscosity with that of a 10% solution of low molecular weight polymer. Similarly, a 5% solution of high molecular weight N6,6, has comparable viscosity with that of a 15% concentration solution of low molecular weight polymer. Traditionally, formic acid is used as a solvent for the dissolution of N6 or N6,6. However, its boiling point of 100 °C is relatively high and therefore it is not very suitable for electrospinning. Nevertheless, the boiling point of formic acid can be lowered by addition of a cosolvent with a relatively low boiling point (bp), namely chloroform (bp of 61 °C). One may also expect that tensile strength of as-spun nylon fibers can be improved by changing ratios of formic acid and chloroform. Indeed, Gogolewski and Pennings42 found that the morphology and tensile strength differences of dry-spun high molecular weight N6 fibers vary with different mixture ratios of formic acid and chloroform. They also documented that the ratio of formic acid to chloroform of 75/25 (v/v) gave a fiber with the highest tensile strength when a super high molecular weight N6 was used for making dry-spun fibers. Low-volatility solvents are required for both dry spinning and electrospinning processes. If solvent does not evaporate completely during electrospinning processing, it is very difficult to form a web or to remove the electrospun web from the collecting drum. Our experiments showed that the mixed solvent consisting of formic acid and chloroform is a good medium for electrospinning nylon-6,6.

4.5.2

Diameter distribution of nylon 6,6 electrospun webs

Bead formation is a very common phenomenon for electrospun fibers; therefore, we used scanning electron microscopy (SEM) to examine the appearance of our fibers. Figure 4.8 shows SEM micrographs of electrospun webs spun from low MW N6,6 at two concentrations of 10 and 15%. The webs formed with the high molecular weight N6,6 at c = 3% and c = 5% are shown in Fig. 4.9. Interestingly, no beads can be observed on all the fibers. If conditions of electrospinning, such as concentration of polymer solution, voltage power or flow rate, are not properly adjusted, beads occur on the surface of nanofilaments. Lee et al.43 found the bead morphology on electrospun polystyrene fibers. Polymer concentration, applied voltage and tip-to-collector distance strongly affect the morphology of the formed beads on polystyrene fibers. At concentrations lower than 15%, beads occurred on the electrospun fibers. No beads were observed for polystyrene fibers when the polymer concentrations

© 2007, Woodhead Publishing Limited

82

Nanofibers and nanotechnology in textiles

(a)

(b)

4.8 Scanning electron microscopy (SEM) micrograph of electrospun webs spun at: (a) c = 10%, low MW N6,6; (b) c = 15%, low MW N6,6. Left-hand scale bars are ~5 µm; right-hand scale bars are 1 µm.

(a)

(b)

4.9 Scanning electron microscopy (SEM) micrograph of electrospun webs spun at: (a) c = 3%, high MW N6,6; (b) c = 5%, high MW N6,6. Left-hand scale bars are ~5 µm; right-hand scale bars are 1 µm.

were higher than 15%. Fong and coworkers22, 44 also documented similar results from the electrospinning of polyethyleneoxide (PEO) polymer. As the concentration of PEO polymer increased, the beads disappeared. Furthermore, the bead shapes changed from spherical to spindle-like. The

© 2007, Woodhead Publishing Limited

Producing polyamide nanofibers by electrospinning

83

Frequency distribution (%)

bead formation is most likely related to the surface tension. At a low polymer concentration, the surface tension has the dominant influence and, therefore, the fibers with beads are formed, or dripping of polymer solution may even occur during spinning. At high concentrations, fibers without beads can be formed because of the cohesive nature of the high-viscosity solution. Usually, in the case of the low molecular weight N6,6, a polymer solution concentration greater than 10% is satisfactory for electrospinning, On the other hand, a polymer concentration of 3% is high enough to successfully conduct the electrospinning of the high MW N6,6. Figures 4.10 and 4.11 show the diameter distribution of electrospun filaments. As can be seen, the fiber diameters depend on the concentration of the N6,6 polymers that were used in this study. Interestingly, the distribution was narrower and the fiber diameter varied from 100 to 200 nm for the web obtained from the low molecular weight N6,6, at a concentration of 10%. However, when the concentration was increased to 15%, the fiber diameter variability increased from 100 to 300 nm. More dramatic changes with polymer c = 10%, low molecular weight N6,6

100 80 60 40 20 0 0–100

101–200 201–300 301–400 401–500 501–600 601–700 701–800 Fiber diameter (nm)

Frequency distribution (%)

c = 15%, low molecular weight N6,6 70 60 50 40 30 20 10 0 0–100

101–200 201–300 301–400 401–500 501–600 601–700 701–800 Fiber diameter (nm)

4.10 Fiber diameter distribution of the low molecular weight N6,6 electrospun webs.

© 2007, Woodhead Publishing Limited

Frequency distribution (%)

84

Nanofibers and nanotechnology in textiles c = 5%, high molecular weight N6,6

70 60 50 40 30 20 10 0 0–100

101–200 201–300 301–400 401–500 501–600 601–700 701–800 Fiber diameter (nm)

c = 3%, high molecular weight N6,6

Frequency distribution (%)

80 70 60 50 40 30 20 10 0 0–100

101–200 201–300 301–400 401–500 501–600 601–700 Fiber diameter (nm)

701–800

4.11 Fiber diameter distribution for high molecular weight N6,6 electrospun webs.

concentration were observed for the high molecular weight N6,6. At a concentration of 3% the fiber diameters were in the range 100–200 nm (Fig. 4.11). However, fibers spun from the higher molecular weight N6,6 solution at c = 5%, had a higher average diameter and a greater diameter variability. As shown in Fig. 4.11, most fibers had diameters in the range 300–400 nm, but a small fraction of fibers exhibited diameters in the range 600–700 nm. It is evident from our study that as the N6,6 concentration increases; the fiber diameter distribution of electrospun filaments becomes broader. Furthermore, the average diameter of electrospun filaments increases with increasing polymer concentration, and the increase in fiber diameters obtained between low and high concentrations of high molecular weight N6,6 solutions was larger than between those of the low molecular weight N6,6 solutions. Ryu et al.32 reported similar results for their electrospun N6 nonwoven mats. The authors also reported a broader diameter distribution in the electrospun filaments at higher polymer concentrations. An increase in the filament diameter

© 2007, Woodhead Publishing Limited

Producing polyamide nanofibers by electrospinning

85

(as the result of a higher viscosity or a higher polymer concentration) is a well-known phenomenon11, 14, 22, 45 in electrospinning. As shown in Figs 4.10 and 4.11, the electrospun fibers exhibit two different distributions at high concentrations for both low and high molecular weight N6,6. Although more detailed studies should be done to prove the generality of these observations, it appears that the bimodal distribution of fiber diameters might be common. Such behavior can be seen particularly for the filaments spun from 15 and 5% solutions of low and high molecular weight N6,6, respectively. Similar phenomena were reported by Ding et al.27 and Deitzel et al.11

4.5.3

Mechanical properties of nylon 6,6 electrospun webs

Table 4.2 shows the mechanical properties of electrospun N6,6 nonwoven webs that were made under various experimental conditions. It is evident that the polymer concentration and the molecular weight of N6,6 influence the fiber morphology and, therefore, the elongation at break and initial modulus, as well as the strength of these materials. Figure 4.12 gives the dependence of these properties as a function of N6,6 concentration. The lowest elongation at break of 25.4% can be seen for the fibers electrospun with the high MW N6,6 from a 3% solution. Interestingly, more concentrated polymer solutions (prepared from the low MW N6,6) seem to lead to an increase in the elongation at break of up to 42.4%. It is well known that most electrospun filaments are not oriented and, therefore, exhibit poor mechanical properties, namely the initial modulus and strength. Interestingly, our study clearly shows that the modulus and the strength are much higher for the webs made from the high MW N6,6 (see Table 4.2 and Fig. 4.13). Thus, it is plausible to state that the use of N6,6 with a high molecular weight provides a simple approach for improving the mechanical properties of electrospun nylon filaments.

4.6

Conclusions

The following parameters have been suggested to affect the electrospinning process: solution properties, including viscosity, polymer concentration, polymer molecular weight, conductivity and surface tension; controlled process variables including hydrostatic pressure in the capillary, electric potential at the tip of needle, emitting electrode polarity, and the distance between the tip of the needle and the collection screen; and ambient parameters including temperature, humidity and air velocity in the electrospinning chamber. There are relationships between polymer solution concentration and electric field strength and the distance between tip and collection screen for achieving a

© 2007, Woodhead Publishing Limited

86

Nanofibers and nanotechnology in textiles 45

Elongation at break (%)

Elongation at break (%) 40 35 30 25 20 0

2

4

6 8 10 Nylon 6,6 concentration (%)

12

14

16

120 Modulus (MPa) Yield stress (MPa)

Strength (MPa)

100 80 60 40 20 0 0

5

10 Nylon 6,6 concentration (%)

15

20

4.12 Mechanical properties of N6,6 electrospun webs as a function of polymer concentration. c = 3% high MW N6,6; c = 10%, low MW N6,6; c = 15%, low MW N6,6. Table 4.2 Mechanical properties of electrospun low and high molecular weight nylon-6,6 nonwovens Concentration (%)

Molecular weight (g/mol)

Modulus (MPa)

Yield stress (MPa)

Elongation at break (%)

3 10 15

170 000 30 000 30 000

106.7 65.8 23.2

12.1 10.2 3.5

25.3 27.9 42.4

steady jet of the polymer solution. Our results showed the voltage for producing electrospun N6 fibers increased with increasing concentration of polymer solution. The diameters of electrospun N6 fibers were affected by the concentration of polymer solution. The optimum tip-to-collector distance for producing a steady jet of the polymer solution depends on the concentration

© 2007, Woodhead Publishing Limited

Producing polyamide nanofibers by electrospinning

87

2.5

Load (lb)

2.0

(c)

1.5

1.0 (b) 0.5

(a)

0 0

0.05

0.1

0.15 0.2 Extension (mm)

0.25

0.3

4.13 Load–extension curves for webs made at: (a) c = 15%, low MW N6,6; (b) c = 10%, low MW N6,6; (c) c = 3%, high MW N6,6.

of polymer solution and the voltage. Increasing voltage, up to a given limit, at a specific solution concentration decreased the diameters of electrospun N6 fibers. Two N6,6 polymers having molecular weights of 30 000 (low MW polymer) and 170 000 g/mol (high MW polymer) were spun from mixed formic acid/ chloroform solutions. Electrospun nonwoven nano-webs were successfully made from 10 and 15% solutions of the low MW N6, while lower polymer concentrations of 3 and 5% were sufficient for electrospinning high MW N6,6. The morphologies and mechanical properties of these nano-scale nonwoven webs were examined. Fiber diameters increased and fiber diameter distributions were broader with increasing polymer concentration for both low and high molecular weight N6,6. Initial moduli of electrospun webs improved for high molecular weight N6,6, particularly when a low polymer concentration (3%) was used. This study demonstrates that the use of high MW N6,6 affords a simple approach for improving the mechanical properties of electrospun nylon filaments.

4.7

References

1. Huang Z., Zhang Y. Z., Kotaki M., Ramakrishna S. (2003), ‘A review of polymer nano fibers by electrospinning and their applications in nanocomposites’, Comp Sci Technol, 63, 2223–2253. 2. Pedicini A., Farris R. J. (2004), ‘Thermally induced color change in electrospun fiber mats’, J Polym Sci Part B: Polym Phys, 42, 752–757. 3. Deitze J. M., Kosik W., McKnight S. H., Beak Tan N. C., Desimone J. M., Crette S. (2002), ‘Electrospinning of polymer nano-fibers with specific surface chemistry’, Polymer, 43, 1025–1029.

© 2007, Woodhead Publishing Limited

88

Nanofibers and nanotechnology in textiles

4. Ding B., Kim H. Y., Lee S. C., Shao C. L., Lee D. R., Park S. J., et al. (2002), ‘Preparation and characterization of a nanoscale poly(vinyl alcohol) fiber aggregate produced by using electrospinning method’, J Polym Sci, Part B: Polym Phys, 40, 1261–1268. 5. Dai H., Gong J., Kim H. Y., Lee D. R. (2002), ‘A novel method for preparing ultrafine alumina-borate oxide fibers via an electrospinning technique’, Nanotechnology, 13, 674–677. 6. Formhals A. (1934), US patent 1,975,504. 7. Formhals A. (1939), US patent 2,160,962. 8. Formhals A. (1940), US patent 2,187,306. 9. Formhals A. (1943), US patent 2,323,025. 10. Formhals A. (1944), US patent 2,349,950. 11. Deitzel J. M., Kleinmeyer J., Harris D., Beck Tan N. C. (2001), ‘The effect of processing variables on the morphology of electrospun nanofibers and textiles’, Polymer, 1(42), 261–272. 12. Reneker D. H., Chun I. (1997), ‘Nanometer diameter fibers of polymer, produced by electrospinning’, Nanotechnology, 7, 216–223. 13. Bergshoef M. M., Vancso G. J. (1999), ‘Transparent nanocomposites with ultrathin, electrospun nylon-4,6 fiber reinforcement’, Adv Mater, 11, 362–365. 14. Doshi J., Reneker D. H. (1995), ‘Electrospinning process and applications of electrospun fibers’, J Electrostat, 35, 151–160. 15. Gibson P. W., Schreuder-Gibson H. L., Riven D. (1999), ‘Transport properties of porous membranes based on electrospun nanofibers’, AIChE J, 45, 190–195. 16. Fong H., Renker D. H. (2001), ‘Structure formation in polymeric fibers’, in Salem D. R., Sussman M. V., Editors. Electrospinning and Formation of Nanofibers, Munich: Hanser, 2001, Chapter 6. 17. Hou H., Jun Z., Reuning A., Schaper A., Wendorff J. H., Greiner A. (2002), ‘Poly(pxylylene) nanotubes by coating and removal of ultrathin polymer template fibers’, Macromolecules, 35, 2429–2431. 18. Kenawy E. R., Bowlin G. L., Mansfield K., Layman J., Simpson D. G., Sanders E. H. (2002), ‘Release of tetracycline hydrochloride from electrospun poly(ethyleneco-vinylacetate), polylactic acid and a blend’, J Controlled Release, 81, 57–64. 19. Scopelianos A. G. (1996), Pizoelectric biomedical device, US Patent, 5522879. 20. Larrondo L., Manley R. (1981), ‘Electrostatic fiber spinning from polymer melts. I. Experimental observation on fiber formation and properties’, J Polym Sci Polym Phys Ed, 19, 909–920. 21. Larrondo L., Manley R. (1981), ‘Electrostatic fiber spinning from polymer melts. II. Examination of the flow field in an electrically driven jet’, J Polym Sci Polym Phys Ed, 19, 921–932. 22. Fong H., Chun I., Reneker D. H. (1999), ‘Beaded nanofibers formed during electrospinning’, Polymer, 40, 4585–4592. 23. Bognitzki M., Czado W., Frese T., Schaper A., Hellwig M., Steinhart M., et al. (2001), ‘Nanostructured fibers via electrospinning’, Adv Mater, 13, 70–72. 24. Lee K. H., Kim Y. H., La Y. M., Lee D. R., Sung N. H. (2002), ‘Influence of a mixing solvent with tetrahydrofuran and N,N-dimethylformamide on electrospun polyvinyl chloride nonwoven mats’, J Polym Sci Polym Phys, 40, 2259–2268. 25. Zuo W., Zhu M., Yang W., Yu H., Chen Y., Zhang Y. (2005), ‘Experimental study on relationship between jet instability and formation of beaded fibers during electrospinning’, Polym Eng Sci, 45, 704–709.

© 2007, Woodhead Publishing Limited

Producing polyamide nanofibers by electrospinning

89

26. Lee K. H., Kim H. Y., Ryu Y. J., Kim K. W., Choi S. W. (2003), ‘Mechanical behavior of electrospun fiber mats of poly(vinyl chloride)/polyurethane polyblends’, J Polym Sci Part B Polym Phys, 41, 1256–1262. 27. Ding B., Kimura E., Sato T., Fujita S., Shiratori S. (2004), ‘Fabrication of blend biodegradable nanofibrous nonwoven mats via multi-jet electrospinning’, Polymer, 45, 1895–1902. 28. Pedicini A., Farris R. J. (2003), ‘Mechanical behavior of electrospun polyurethane’, Polymer, 44, 6857–6862. 29. Deitzel J. M., Kleinmeyer J. D., Hirvonen J. K., Beck Tan N. C. (2001), ‘Controlled deposition of electrospun poly(ethylene oxide) fibers’, Polymer, 42, 8163–8170. 30. Zimmerman J., Mark H. F., Bikales N. M. (1988), in Encyclopedia of Polymer Science and Engineering, Vol. 6, New York: Wiley, 802–839. 31. Ziabicki A. (1976), Fundamentals of Fiber Formation: The Science of Fiber Spinning and Drawing, New York: Wiley. 32. Ryu Y. J., Kim H. Y., Lee K. H., Park H. C., Lee D. R. (2003), ‘Transport properties of electrospun nylon 6 nonwoven mats’, Eur Polym J, 39, 1883–1889. 33. Fong H., Liu W., Wang C., Vaia R. A. (2002), ‘Generation of electrospun fibers of nylon 6 and nylon 6-montmorillonite nanocomposite’, Polymer, 43, 775–780. 34. Larrond O., Manley R. (1981), ‘Electrostatic fiber spinning from polymer melts. III. Electrostatic deformation of a pendant drop of polymer melts’, J Polym Sci Polym Phys, 19, 933–940. 35. Schreuder-Gibson H. L., Gibson P., Senecal K., Sennett M., Walker J., Yeomans W. (2002), ‘Protective textile materials base on electrospun nanofibers’, J Adv Mater, 34, 44–55. 36. Stephen J. S., Chase D. B., Rabolt J. F. (2004), ‘Effect of the electrospinning process on polymer crystallization chain conformation in Nylon 6 and Nylon 12’, Macromolecules, 37, 877–887. 37. Supaphol P., Uppatham C. M., Nithitanakul M. (2005), ‘Ultrafine electrospun polyamide 6 fibers: effect of emitting electrode polarity on morphology and average fiber diameter’, J Polym Sci Polym Phys, 43, 3699–3712. 38. Uppatham C. M., Nithitanakul M., Supaphol P. (2004), ‘Ultrafine electrospun polyamide 6 fibers: effect of solution conditions on morphology and average fiber diameter’, Macromol Chem Phys, 205, 2327–2338. 39. Supaphol P., Uppatham C. M., Nithitanakul M. (2005), ‘Ultrafine electrospun polyamide 6 fibers: effects of solvent system and emitting electrode polarity on morphology and average fiber diameter’, Macromol Mater Eng, 290, 933–942. 40. Dersch R., Liu T., Schaper A. K., Greiner A., Wendorff J. H. (2003), ‘Electrospun nanofibers: Internal structure and intrinsic orientation’, J Polym Sci Polym Chem, 41, 545–553. 41. Vasanthan N., Kotek R., Jung D. W., Shin D., Tonelli A. E. (2004), ‘Lewis acid-base complexation of polyamide 66 to control hydrogen bonding, extensibility and crystallinity’, Polymer, 45, 4077–4085. 42. Gogolewski S., Pennings A. J. (1985), ‘High-modulus fibres of nylon-6 prepared by a dry-spinning method‘, Polymer, 26, 1394–1400. 43. Lee K. H., Kim H. Y., Bang H. J., Jung Y. H., Lee S. G. (2003), ‘The change of bead morphology formed on electrospun polystyrene fibers’, Polymer, 44, 4029–4034. 44. Fong H., Reneker R. H. (1999), ‘Elastomeric nanofibers of styrene-butadiene-styrene triblock copolymer’, J Polym Sci Part B Polym Phys, 37, 3488–3493. 45. Demir M. M., Yilgor, I., Yilgor, E., Erman, B. (2002), ‘Electrospinning of polyurethane fibers’, Polymer, 43, 3303–3309.

© 2007, Woodhead Publishing Limited

5 Controlling the morphologies of electrospun nanofibres T. L I N and X. G. W A N G, Deakin University, Australia

5.1

Introduction

Electrospinning is a unique method of producing continuous polymer fibres. It has received a great deal of attention in recent times due to its versatility in the spinning of a wide variety of polymeric fibres, and its consistency in producing polymer fibres with the fibre diameter on submicrometre or nanometre scales. The as-spun fibres, often in the form of a non-woven mat, have an extremely high surface area to mass ratio and a highly porous structure which has potential applications in areas such as tissue scaffolds (Li et al., 2002, Matthews et al., 2002, Yoshimoto et al., 2003, Subramanian et al., 2004), wound dressings (Khil et al., 2003, Kim et al., 2004), nano-catalysis (Jia et al., 2002, Dong and Jones, 2004), protective clothing (Gibson and Schreuder-Gibson, 2000, Gibson et al., 2001), filtration (Suthar and Chase, 2001, 2002, Tsai and Schreuder-Gibson, 2003) and optical electronics (Dong et al., 2003, El-Aufy et al., 2003). Although the basic electrospinning technique was invented in the 1930s (Formhals, 1934), it is only in recent decades that fundamental research has led to great progress in this area. Studies on understanding the electrospinning process, its operating parameters and material properties have been extensive. Improved electrospinning methods, new fibre structures and potential applications are continually emerging. Professional reviews on the electrospinning process, and the advancement and applications of electrospun nanofibres in specific areas can be found in the literature (Reneker and Chun, 1996, Gibson et al., 1998, Bognitzki et al., 2001, Buer et al., 2001, Huang et al., 2003, Frenot and Chronakis, 2003, Jayaraman et al., 2004, Kameoka et al., 2004, Krishnan et al., 2004, Li and Xia, 2004b). The electrospun nanofibres usually have a regular threadlike structure and some fibres can form a ribbon-like fibrous morphology. The fibre diameter varies in range from 5 nm to 10 µm. Defectives, such as colloid beads or beads-on-string fibres, occur among the resultant fibres, or even as the major product, depending on the operating conditions and the material properties. 90

© 2007, Woodhead Publishing Limited

Controlling the morphologies of electrospun nanofibres

91

Also, the electrospun fibres can be deposited either randomly to form a nonwoven web or in an oriented manner to give an aligned nanofibre array. Twisting nanofibre bundles can form nanofibre yarns. Within the fibre web, the nanofibres can be connected physically or bonded at ‘cross-points’ to form an interconnected ‘spider web’. In addition, two polymer components have been combined to form side-by-side, or sheath–core, bicomponent electrospun fibres. The production of nanofibres in a controlled manner, so that the process gives a high-quality fibre with precise fibre morphology, is a vital task, as the fibre morphology has a significant influence on the fibre performance. The control of fibre morphology has been based on a thorough understanding of the electrospinning process and the factors that affect the electrospinning process and the fibre morphologies. The fibre alignment is controlled through the fibre deposition process and the bicomponent nanofibres are electrospun using special spinnerets. In this chapter, techniques on controlling the fibre morphology, the fibre orientation, the fibre component and the web morphology during the electrospinning process are discussed. This chapter is divided into seven sections. In Section 5.2, the electrospinning process is introduced briefly, and factors that affect the electrospinning process and the resultant nanofibres will be discussed. Possible approaches for controlling the fibre morphology in the electrospinning process will be introduced. Section 5.3 will discusses the effect of polymer concentration on the solution properties (e.g. viscosity, conductivity and surface tension) that adversely affect the fibre diameter and evenness. Tuning the fibre diameter via adjustments of the polymer concentration is described. Problems and limitations in changing polymer concentrations are discussed. In Section 5.4, fibre beads and approaches to eliminate the beads are presented. Section 5.5 discusses the orientation and assembly of nanofibres. The aligned nanofibre nonwovens and nanofibre yarns are described. The formation of interconnected nanofibre webs is also discussed. In Section 5.6, the preparation and potential applications of bicomponent nanofibres, e.g. core–sheath and side-by-side nanofibres, are given in detail. Section 5.7 discusses possibilities for further research on controlling the morphology of electrospun nanofibres.

5.2

The electrospinning process and fibre morphology

Electrospinning is an effective technique to produce continuous polymer fibres with diameters on submicrometre or nanometre scales (Reneker and Chun, 1996, Jayaraman et al., 2004, Li and Xia, 2004b). It involves a process in which a polymer solution or melt is stretched into fine filaments under the action of a high electrical voltage. The basic electrospinning system, illustrated

© 2007, Woodhead Publishing Limited

92

Nanofibers and nanotechnology in textiles

in Fig. 5.1, consists of a charged polymer solution that is fed through a small opening or nozzle (usually a needle or pipette tip). When the polymer solution is delivered to the tip of a capillary and a high-voltage difference is established between the tip and the collector electrode, charge accumulates in the polymer solution and at the tip of the capillary a droplet is electrically attracted to the collector electrode, typically 5–30 cm away, deforming into a cone shape which is referred to as a ‘Taylor cone’. With an increase in the applied voltage, the cone becomes sharper. When the voltage is larger than a critical value, the droplet overcomes the restriction of the surface tension: a jet is thus produced, forming fine filaments. Evaporation of the solvent from the filaments results in solid fibres. Although both polymer solutions and polymer melts have been electrospun into fibres, most studies have been focused on the polymer solution. Literature is scarce on the polymer melt-based electrospinning process. Therefore, in this chapter, the discussion is limited to the solution-based electrospinning system. It has been well established that both operating parameters and material properties affect the electrospinning process and the resulting fibre morphology. The operating parameters include the applied electrical field, the flow rate of the polymer solution, the distance between the nozzle and the collector (spinning distance), and the diameter of the spinneret, etc. A minute change in the operating parameters can lead to a considerable change in the fibre morphology. For example, finer nanofibres are electrospun from a nozzle of smaller diameter (Katti et al., 2004); increasing the flow rate leads to larger fibre diameter; and a higher applied voltage results in the emergence of fibre Nanofibres Polymer solution Nozzle

Polymer jet

Power supply

5.1 A basic electrospinning system.

© 2007, Woodhead Publishing Limited

Collector

Controlling the morphologies of electrospun nanofibres

93

beads, though reducing the fibre diameter (Deitzel et al., 2001, Lee et al., 2004). The material properties that affect the electrospinning process and the fibre morphology include the polymer concentration, the solution viscosity, the solution conductivity, the surface tension and other properties concerning the solvent as well as the polymer itself. Among the material properties, the solution concentration plays a major role in stabilizing the fibrous structure because it also affects other solution properties, such as the solution viscosity, the surface tension and the conductivity. The solvent used is another important factor because it mainly determines the surface tension and the evaporation process. The volatility of the solvent affects the fibre surface morphology and the web structure. The solution bulk properties come from intermolecular interactions among the solvent molecules and the polymer macromolecules. Any factors that interfere with these interactions change the solution properties that affect the electrospinning process, and the fibre morphology is thus altered accordingly. For example, the solution viscosity is closely related to the entanglement of polymer macromolecules. In a good solvent, polymer chains with a higher molecular weight tangle with each other more easily, which leads to a higher solution viscosity. Electrospinning such a polymer solution produces continuous and uniform fibres. However, when polymer of a lower molecular weight is electrospun, even at the same polymer concentration, the resultant fibre could have a colloid bead or beaded fibre morphology (Koski et al., 2003). In theory, any factor that affects the electrospinning process also provides a means to control the fibre morphology. In practice, however, only factors with a noticeable and reliable influence on fibre morphology can be employed for this purpose. Methods of controlling the fibre morphology have been reported based on either the operating parameters or the material properties.

5.3

Polymer concentration and fibre diameter

As the polymer concentration plays a major role in the electrospinning process, it is not surprising that the polymer concentration has been extensively exploited to change and control fibre morphology in electrospinning. This effort has also led to improved understanding on the effect of the polymer concentration on other solution properties, such as viscosity, surface tension and conductivity. Under the same electrospinning conditions, increasing the polymer concentration will increase the diameter of the electrospun fibres. However, a non-linear relationship between the solution concentration and the fibre diameter usually forms (Deitzel et al., 2001). The reasons for this non-linear relationship can be attributed to the non-linear relationship between the polymer concentration and the solution viscosity. As illustrated in Fig. 5.2, when the polymer concentration is low, the solution viscosity increases slightly

© 2007, Woodhead Publishing Limited

Nanofibers and nanotechnology in textiles

2000

Beads-onstring

Beads

1000

Fibres

Viscosity

800

Viscosity (cP)

1500

600 1000 400

Average fibre diameter (nm)

94

500 200 Fibre diameter 0

0 1

2

3

4

5

6 7 8 [PAN] (%, w/v)

9

10

11

12

5.2 Relationship between polymer concentration, solution viscosity and diameter of electrospun polyacrylonitrile (PAN) fibres.

with the increase in the polymer concentration. As the polymer concentration increases, the viscosity increases gradually until the concentration reaches a specific value, after which the viscosity increases considerably (Lin et al., 2005a). In the electrospinning process, the solvent evaporates from the jet/filament continuously until the jet becomes dry. Stretching the jet increases the surface area, which accelerates the solvent evaporation. From the initial jet to dry fibres, the fibre stretching process is very quick, taking only tens of milliseconds (Shin et al., 2001a). If the strength to stretch the filaments remains the same, electrospinning a polymer solution of higher viscosity could be much harder than that with a lower viscosity, because the concentration has a larger influence on the viscosity when the concentration is high. On the other hand, the polymer concentration affects the solution conductivity, which further influences the solution charge density (Shin et al., 2001b). The stretching is enhanced as a result of increased charge density. This could compensate to some extent for the difficulty in stretching a solution of higher polymer concentration. In certain cases, the strength is improved to such an extent that it neutralizes the effect of the viscosity, which results in a similarly linear relationship between the concentration and the fibre diameter. The relationship between the solution viscosity and the polymer concentration is highly dependent on the nature of the polymer (e.g. structure and molecular weight) and the intermolecular interactions within the polymer

© 2007, Woodhead Publishing Limited

Controlling the morphologies of electrospun nanofibres

95

solution (polymer–polymer, polymer–solvent, solvent–solvent, etc.). Because of these variants, it is almost impossible to establish a universal formula to include all polymer–solvent systems in a given electrospinning process. A conventional approach for controlling the fibre diameter is based on the understanding of the dependent relationship between the polymer concentration and the average diameter of the electrospun nanofibres, given the solvent system and specific operating conditions (e.g. the applied voltage, the flow rate and the spinning distance). Although reducing the polymer concentration is a straightforward way to produce finer nanofibres, electrospinning a dilute polymer solution usually leads to the emergence of colloid beads or beads-on-string fibres (also called ‘necklace fibres’). These defectives even become the main products when the polymer concentration is very low. Because of this, the electrospinning of bead-free and uniform nanofibres particularly for fibre diameters less than 100 nm still remains a great challenge. The electrospinning of an extremely dilute polymer solution to produce colloid beads is usually called ‘electrospraying’ (Kebarle and Peschke, 2000). The suggested explanation is that the viscoelastic force in the jet is too small to hold the fibrous structure. The jet gets dissociated into individual charged sections, and these sections usually turn into droplets owing to the action of the surface tension. With the evaporation of the solvent, the droplet reduces its size. This leads to an increase in the surface charge density. The droplet can further split into smaller droplets due to the higher electrostatic repulsion (Fig. 5.3a). As the polymer concentration increases, the solution viscosity also increases. The jet under the higher viscoelastic force is more difficult to break into (a)

(b)

5 µm

(d)

(c)

5 µm

5 µm

300 nm

(e)

(f)

5 µm

5 µm

5.3 Product morphologies when electrospinning polyacrylonitrile/ DMF(N, N-dimethylformamide) solution at different concentrations of (a) 2%, (b) 3%, (c) 4%, (d) 5%, (e) 6%, (f) 7% (w/v).

© 2007, Woodhead Publishing Limited

96

Nanofibers and nanotechnology in textiles

individual sections. Instead, the electrostatic repulsion among the like-charged sections elongates the thin ‘links’ between the sections, forming thinner filaments. As a result of the orientation of the macromolecular polymer, these filaments become hyper-stabilized (Yarin, 1993). Meanwhile, the relatively thicker sections get stretched thinner as well, although to a lesser extent than the ‘links’. Under the action of surface tension, they tend to take the shape of a droplet or bead. With the evaporation of the solvent, an apparent beads-on-string structure is thus formed (Fig. 5.3b–d). Further increasing the solution viscosity provides a larger viscoelastic force to resist rapid changes in shape. This allows for more uniform stretching, resulting in a continuous and homogeneous fibre structure (Fig. 5.3e and f). It has been well established that the solution viscosity is highly dependent on the intermolecular interaction of the polymer (Bercea et al., 1999). Generally, in a dilute polymer solution, the intermolecular distance is so large that the intermolecular interactions are very weak. The intermolecular interactions become predominant gradually with the increase in polymer concentration. At a certain concentration, c*, the domains of the polymer molecules begin to overlap and eventually an entanglement may develop. Thus, c* is a critical concentration to distinguish whether an obvious intermolecular interaction will happen in the polymer solution. The higher concentration regime, named as semi-dilute, is characterized by the intermolecular interaction and eventual entanglements. In addition, it is worth noting that a high applied voltage also leads to the formation of a beaded fibre (Deitzel et al., 2001).

5.4

Fibre bead formation and fibre surface morphology

The formation of beaded fibres has been attributed to a low solution viscosity (Fong et al., 1999, Lee et al., 2003). Increasing the polymer concentration, which results in an increase in the solution viscosity, has been the conventional approach to prevent the formation of the beads. However, increasing the solution viscosity is not always effective. With some polymers, polystyrene for instance, even a high concentration of polymer solution does not guarantee bead-free electrospun fibres. Our research indicates that for a given solution viscosity, the solution conductivity has a significant effect on the formation of beaded fibres. The addition of a small amount of ionic surfactant into the polymer solution is able to suppress the formation of beads effectively (Lin et al., 2004). As shown in Fig. 5.4a, beads proliferate as a result of electrospinning a normal polystyrene solution. Increasing the flow rate increased the number of beads. Efforts to eliminate the beaded fibres by adjusting the operating conditions and changing the polymer concentrations were unsuccessful. However, when a small amount of cationic surfactant was added to the same polymer solution,

© 2007, Woodhead Publishing Limited

Controlling the morphologies of electrospun nanofibres (a)

97

(b)

5.4 Electrospun polystyrene fibres: (a) non-uniform beaded fibres electrospun from an ordinary electrospinning process; (b) bead-free fibres electrospun from the same polymer solution with the addition of a cationic surfactant.

the same electrospinning process produced non-beaded fibres. The SEM image (Fig. 5.4b) reveals bead-free and uniform fibres as a result of surfactants added to the polymer solution. No isolated beads and beads-on-string structures were found. The surfactant was so effective that a concentration as low as 10–6 mol/l was enough to prevent the formation of the beaded fibres. Reasons as to why the addition of ionic surfactant prevents the formation of beaded fibres can be found in the effect of ionic surfactant on polymer properties. When the polymer concentration is the same, with the increase in surfactant concentration, the solution viscosity has a slight increase and the surface tension has a minor reduction; however, the solution conductivity increases considerably. Increasing the solution conductivity leads to an increase in the solution charge density. With increased charges within the filaments, their interaction with the external electric field and their repulsion to each other are enhanced, which improves the stretch of the filaments. Under the stronger stretching forces, the liquid filaments are extended in a more rapid and uniform way, which prevents the filaments from forming beaded sections. When the electrospun fibres become uniform, further addition of the ionic surfactant into the polymer solution reduces the fibre diameter. As shown in Fig. 5.5, as the surfactant concentration increases, the average fibre diameter decreases gradually, and the fibre distribution becomes narrow as well (Lin et al., 2004). Both the anionic and the cationic surfactants have been found to improve the solution charge density that enhances the fibre stretching process. It appears that there is no difference between adding an anionic and a cationic surfactant into the polystyrene solution, because they both lead to bead-free and uniform fibres. However, when a surfactant molecule is able to associate with the polymer linkage in the solution, it gives rise to an improvement in fibre evenness, because this polymer–surfactant interaction tends to make

© 2007, Woodhead Publishing Limited

98

Nanofibers and nanotechnology in textiles 1400

1200

Diameter (nm)

1000

800

600

400

200 0

5

10

15 20 [DTAB] (mmol/l)

25

30

35

5.5 Average fibre diameter vs surfactant concentration. The surfactant used is dodecyltrimethylammonium bromide (DTAB).

the polymer–surfactant become polyions. Other additives, such as NaCl, pyridinium formiate and polyelectrolyte, have also been reported to improve fibre evenness (Fong et al., 1999, Jun et al., 2003, Won et al., 2004). By comparison with the ionic surfactants, a non-ionic surfactant, Triton X-405, was added to the polystyrene solution, and the solution was electrospun in the same operating conditions. It was observed that the beaded fibres were still produced, but the fibre and bead morphologies are different from those that were electrospun from a polystyrene solution not containing surfactants. The addition of surfactant into the solution would affect the formation of ‘Taylor corn’, as observed by other researchers (Yao et al., 2003), who also found that a non-ionic surfactant made the electrospinning process stable. The feasibility of stopping the formation of fibre beads by adding ionic surfactants to a dilute polymer solution has also been studied (Lin et al., 2005a). As shown in Fig. 5.6, when a small amount of ionic surfactant was added to the dilute polymer solution, beads changed to a beads-on-string structure. Also, when the ionic surfactant was added to a semi-dilute polymer solution the fibres became quite uniform.

5.4.1

Surface morphology

The surface morphology of electrospun fibres is affected by the polymer used, the voltage applied and the solvent, as well as the environmental conditions (e.g. humidity). The fibres may have either a smooth or a porous/

© 2007, Woodhead Publishing Limited

Controlling the morphologies of electrospun nanofibres

99

(b)

(a)

5.6 SEM images of fibres electrospun from 2% PAN/DMF solution: (a) the solution does not contain any additives; (b) the solution contains 0.1% (w/v) dodecyltrimethylammonium bromide (DTAB). (a)

(b)

5.7 SEM images of electrospun PLA fibres: (a) chloroform as the solvent; (b) chloroform–DMF mixture (50:50 v/v) as the solution.

rough surface. A higher applied voltage leads to a rougher fibre surface (Deitzel et al., 2001). Porous fibres containing lots of ellipse-like holes on the fibre surface layer have been electrospun from a polylactic acid (PLA)–chloroform solution (Fig. 5.7a). However, when the solution used chloroform–DMF mixture as the solvent, the same operating condition gave a finer nanofibre with a smooth fibre surface (Fig. 5.7b). Porous fibres have also been electrospun from polycarbonate, poly (methyl methacrylate) and polystyrene (Megelski et al., 2002). Reasons for the formation of the porous surface have been explained as a phase separation occurring during the cooling of fibres. The rapid solvent evaporation and subsequent condensation of moisture into water particles result in the formation of nano- or micropores on the fibre surface (Casper et al., 2004). When the environment humidity increases, the pore size becomes larger. However, this result was observed only when the solution used a highly volatile organic solvent, such as chloroform, tetrahydrofuran and acetone.

© 2007, Woodhead Publishing Limited

100

5.5

Nanofibers and nanotechnology in textiles

Controlling fibre alignment and web morphologies

The driving forces for the deposition of electrospun fibres come from an electric field between the charged spinneret and the grounded collector. Owing to the existence of electric charges, the as-spun fibres increase the local electric potential of the collector. The deposition of the as-spun fibres on the collector is affected by the local electric potential. The fibres or fibre sections later deposited on the collector are electrically repulsed by the previously deposited fibres. Thus they adjust their direction automatically towards an area that has a lower electric potential. A dynamic change in the electric potential profile leads to random deposition of fibres, which results in a nonwoven fibrous web, as usually observed in an ordinary electrospinning process. However, orientated or aligned nanofibres can be formed if the fibres are deposited in a controlled way.

5.5.1

Fibre orientation

Earlier works on controlling the fibre orientation used a moving roller (Pedicini and Farris, 2003) (Fig. 5.8a) or a vibrating metal plate (Sundaray et al., 2004, Teo et al., 2005) (Fig. 5.8b) as the collector. When the roller rotated at a high speed (surface velocity 9–10 m/s), the fibre sections deposited first on the roller would move with the collector at the same speed. This additional movement drew other fibre sections which had not deposited on the collector to move with the movement of the collector. As a result, the fibres were deposited with their direction parallel to the direction of the movement, thus giving a partially orientated fibre mat. When the moving speed was high enough, the collector was able to further elongate the as-spun fibres, thus leading to a higher molecular orientation within the fibres (Fennessey and Farris, 2004). A sharp-edged metal disc has also been used to control the fibre alignment (Fig. 5.8c). Although the roller rotates at a relatively lower speed, fibres with very good alignment can be produced (Theron et al., 2001, Xu et al., 2004). Using a motionless metal frame (2 cm × 6 cm) as the electrode (Fig. 5.8d), the aligned nanofibres can be easily collected (Dersch et al., 2003). During the fibre deposition, a fibre first attaching to one side of the frame makes the local electric potential increase instantly. The electrical repulsion leads fibre sections arriving later to be deposited at the opposite side of the frame. This results in the fibres being deposited vertically to the frame side in order to avoid the repulsing force. Fibres deposited later on the frame are repulsed electrically by the fibres deposited previously, which therefore results in a parallel arrangement. The electrospinning process finally gives an aligned nanofibre array. When the collector uses an inter-parallel grounded electrode

© 2007, Woodhead Publishing Limited

Controlling the morphologies of electrospun nanofibres

101

(a) (b)

(c)

(d)

(f)

(e)

(g)

5.8 Collectors used for regulating the fibre collection process: (a) rotating roller; (b) vibrating plate; (c) sharp edged rotating disc; (d) frame; (e) fork; (f) point collector; (g) drum.

couple (the distance between micrometres to millimetres) (Fig. 5.8e), the nanofibres are aligned in the same way (Li et al., 2003). Based on the above concept, a silicon wafer, containing four couples of inter-paralleled Au stripes, has been used as the collector. The selective connection of one electrode couple to the ground electrode each time in

© 2007, Woodhead Publishing Limited

102

Nanofibers and nanotechnology in textiles

sequence in the electrospinning process enables the deposition of fibres layer-by-layer, with fibres parallel to each other within each layer, yet having a different orientation among the layers (Li et al., 2004b). A wire drum collector (Fig. 5.8g), consisting of a group of metal frames in the form of a cylindrical collector, has been used to produce the aligned nanofibres continuously (Katta et al., 2004). With rotating the drum slowly, a wire with the closest distance to the spinneret will be deposited by a fibre section firstly, followed by the deposition of other fibre sections on the next wire, due to the electric field attraction. With the movement of the collector, the fibres are deposited from one wire to the next. Because of the electrostatic forces, the fibres stretch across the shortest distance between the wires and thus cause the fibres to align. In addition, a well-aligned nanofibre bundle has been prepared using two fixed points as the collector (Teo and Ramakrishna, 2005) (Fig. 5.8f).

5.5.2

Nanofibre yarns

Yarns are continuous fibre bundles with the fibres partially oriented. The preparation of yarn has been based on either an in situ or a postspinning twisting process. The preparation of a continuous nanofibre yarn consists of a special deposition system which collect the fibres continuously and twists in sequence so as to form a continuous yarn. A metal cylinder (Ko et al., 2003) and a funnel-like collector (Kim et al., 2003) have been employed for this purpose, and when the nanofibres are electrospun to the inner side of the collector, they are drawn mechanically, or with the aid of vacuum, into a continuous thread, and twisted by rotating the collector or by using an additional twisting system, to form a continuous yarn (Fig. 5.9). Water has also been used as the collecting medium. When the fibres are directly electrospun in to water, they are usually floated on the water surface to form a thin fibre membrane. Drawing and twisting this nanofibre membrane results in a nanofibre yarn (Smit et al., 2005) (Fig. 5.9). In addition, simply twisting a thin nanofibre stripe cut from a regular electrospinning nanofibre mat also gives a nanofibre yarn (Fennessey and Farris, 2004). The twisting treatment leads to a partially orientated nanofibre bundle and the twisting angle affects the yarn’s mechanical strength.

5.5.3

Web structure

Although a random deposition of electrospun fibres gives a nonwoven fibre structure, the web morphology varies depending on the polymer properties and the operating conditions. Usually, the nanofibres are accumulated by a simple physical interaction and the fibres do not bond with each other. However, when the electrospinning distance is very short, the solvent has insufficient

© 2007, Woodhead Publishing Limited

Controlling the morphologies of electrospun nanofibres

103

Ventilation conditioning HV Orientation HV Attenuation Twisting Take-up Spin Vacuum

Winder

Twist HV

5.9 Process for the preparation of nanofibre yarns.

time to evaporate from the filaments because of the short flying time. Wet or semidry nanofibres are stuck and bonded together, forming an interconnected fibre web (Buchko et al., 1999). Even with dry electrospun fibres, if they come from an elastomeric polymer, an interconnected web structure could also be formed, because the polymer has a low glass transition temperature. As shown in Fig. 5.10, the elastomeric nanofibres merge at their ‘crossover points’ to form an interconnected web, or combine into larger fibres.

5.6

Bicomponent cross-sectional nanofibres

As with conventional bicomponent fibres, a bicomponent nanofibre consists of two or more polymer components within the same filament, with each component existing separately (Khatwani and Yardi, 2003). According to the cross-sectional morphology, the bicomponent fibres can be classified into

© 2007, Woodhead Publishing Limited

104

Nanofibers and nanotechnology in textiles

5.10 An interconnected nanofibre web electrospun from elastomeric polyurethane.

Core–sheath

Side-by-side

Pie-wedge

Islands in the sea

5.11 Cross-sectional morphologies of the bicomponent fibres.

four main types: ‘core–sheath’, ‘side-by-side’, ‘pie-wedge’ and ‘islands in the sea’. Their cross-sectional morphologies are illustrated in Fig. 5.11. Although the existing fibre-making technique is able to produce a bicomponent fibre of many cross-sectional structures, the production of bicomponent nanofibres has been limited to two basic types of cross-sectional structures, the ‘core–sheath’ and the ‘side-by-side’. These bicomponent nanofibres are electrospun via special spinnerets. Two polymer solutions flow within the spinneret as the sheath and core, or side-by-side, to the tip of the nozzle and then are subjected to a co-electrospinning process. The formation of bicomponent nanofibres is determined by the laminar bicomponent jet.

© 2007, Woodhead Publishing Limited

Controlling the morphologies of electrospun nanofibres

5.6.1

105

‘Core–sheath’ nanofibres and hollow nanofibres

The core–sheath nanofibres are prepared by using a co-axial spinneret as the nozzle (Sun et al., 2003, Yu et al., 2004). In the electrospinning process, the two polymer solutions are delivered to the tip of the nozzle via the small tube and the interlayer between this small tube and the larger co-axial tube, separately, and then co-electrospun into ‘core–sheath’ nanofibres (Fig. 5.12). The dimensions of the core and the sheath can be adjusted via the solution concentration and their relative flow rates. The two polymer solutions can use either the same or different types of solvents. Such a co-electrospinning process is also suitable for a polymer couple which generates a precipitate while mixing (Li et al., 2004a). Even a pure liquid or a solution of a small molecule which is not able to be electrospun alone can also be used as the core solution, and the co-axial bicomponent electrospinning process still works very well (Li and Xia, 2004a). ‘Hollow nanofibres’ (also called ‘nanotubes’) can be prepared easily from this core–sheath nanofibre by selectively removing the ‘core’ from the bicomponent nanofibres. To facilitate removal, a heavy mineral oil is used as the core material, which can therefore be removed by a simple dissolving/ extracting process (Li and Xia, 2004a). SEM and TEM images can clearly confirm the tubulous morphology. Another approach to prepare the hollow nanofibres used electrospun nanofibres as the template (‘tubes-by-fibre-template’ technique, TUFT). The nanofibres are firstly coated with a layer of shell material and then the fibre material is removed by dissolving or calcination, giving tubulous fibres. Electrode Polymer A Polymer B

HV

5.12 Apparatus for electrospinning ‘core–sheath’ nanofibres.

© 2007, Woodhead Publishing Limited

106

Nanofibers and nanotechnology in textiles

This technique has been used to prepare metal (Pinto et al., 2004), polymer (Bognitzki et al., 2000, Hou et al., 2002) and inorganic oxide (Caruso et al., 2001, Zhang et al., 2005) nanotubes.

5.6.2

‘Side-by-side’ nanofibres and sharp-edged crosssectional nanofibres

Early attempts to prepare the ‘side-by-side’ bicomponent nanofibres were also based on a co-electrospinning process (Gupta and Wilkes, 2003). In the experiment, two charged polymer solutions were delivered via different syringes to the tip of a single nozzle and then electrospun into nanofibres. Although energy dispersive X-ray spectrometry (EDS) analysis confirmed that the asspun nanofibre mat had both polymer components, there is no clear proof to indicate whether the fibre has a ‘side-by-side’ bicomponent cross-sectional morphology. Using a microfluidic device as the nozzle and putting two polymer solutions into the same microfluidic channel side-by-side, Lin et al. (2005b) have electrospun the side-by-side bicomponent nanofibres successfully (Fig. 5.13). The ‘side-by-side’ fibre morphology can be clearly proved by selective removal of one of the polymer components from the nanofibres. As shown in Fig. 5.14b, when the polyurethane moiety is removed from a polyacrylonitrile– polyurethane (PAN-PU) bicomponent nanofibre, the residual PAN moiety retains its fibrous structure, except that one side is removed from the fibre. Further proof can be found in the cross-sectional view of the nanofibres. As shown in Fig. 5.14c and d, the TEM image indicates that all the PU-PAN

Microfluidic device Polymer A

Polymer B

HV

5.13 Schematic for electrospinning the ‘side-by-side’ bicomponent nanofibres.

© 2007, Woodhead Publishing Limited

Controlling the morphologies of electrospun nanofibres (a)

(c)

(b)

(d)

107

5.14 SEM image of PU-PAN side-by-side nanofibres (a) and the same side-by-side nanofibre after dissolving the PU moiety (b). TEM image of the cross-sectional morphology of PU-PAN side-by-side nanofibres (c, d). Scale bars are 1 µm (c) and 100 nm (d).

fibres have both a light and a dark area in the fibre cross-sections, which suggests that the fibre has two side-by-side components. The light area has been confirmed to be the polyurethane, while the dark area is the PAN moiety. The PAN side has a ‘U’-shaped cross-section. Selective removal of the PU ‘side’ from the nanofibres leaves a sharp-edged cross-sectional nanofibre (Fig. 5.14b). It is very interesting to observe that when the ‘side-by-side’ bicomponent nanofibres are electrospun from an elastomeric polymer and a thermal plastic, the as-spun nanofibres have curly or helically crimped fibre morphologies (Fig. 5.14a). The diameter of the helix can be as small as 500 nm. Therefore, the side-by-side co-electrospinning process could be an effective route for the production of helical or crimped polymer nanofibres.

5.7

Future trends

Although significant progress has been made in electrospinning in recent times, theoretical research on electrospinning nanofibres in a controlled way is still at an early stage. There is considerable potential for the improvement of the electrospinning process and for the modification of fibre morphology. One of the challenges to the electrospinning technique is for the production of uniform and fine nanofibres (with fibre diameter less than 100 nm) from

© 2007, Woodhead Publishing Limited

108

Nanofibers and nanotechnology in textiles

different types of polymers. Other challenges include controlling the fibre morphology in a precise way, such as preparing the nanofibres with specific dimension, morphology and web structure, or with specific component arrangement. The ‘core–sheath’ and ‘side-by-side’ bicomponent fibres are the basic structures of bicomponent fibres. Methods for the preparation of more complex bicomponent nanofibres such as ‘island-in-the-sea’ and ‘piewedges’ have yet to be developed. In addition, electrospinning of polymer melt is also interesting as no solvent is involved in the process. Further research is warranted to examine the effects of both operating conditions and material properties on the properties and morphologies of nanofibres electrospun from polymer melts.

5.8

Acknowledgements

We thank the Nanostructural Analysis Network Organisation (NANO) for its kind assistance with the TEM observation of side-by-side bicomponent nanofibres. We also thank Deakin University for financial support through a Central Research Grant (G004169).

5.9

References

Bercea, M., Morariu, S., Ioan, C., Ioan, S. and Simionescu, B. C. (1999) European Polymer Journal, 35, 2019–2024. Bognitzki, M., Hou, H., Ishaque, M., Frese, T., Hellwig, M., Schwarte, C., Schaper, A., Wendorff, J. H. and Greiner, A. (2000) Advanced Materials (Weinheim, Germany), 12, 637–640. Bognitzki, M., Czado, W., Frese, T., Schaper, A., Hellwig, M., Steinhart, M., Greiner, A. and Wendroff, J. H. (2001) Advanced Materials (Weinheim, Germany), 13, 70–72. Buchko, C. J., Chen, L. C., Shen, Y. and Martin, D. C. (1999) Polymer, 40, 7397–7407. Buer, A., Ugbolue, S. C. and Warner, S. B. (2001) Textile Research Journal, 71, 323–328. Caruso, R. A., Schattka, J. H. and Greiner, A. (2001) Advanced Materials (Weinheim, Germany), 13, 1577–1579. Casper, C. L., Stephens, J. S., Tassi, N. G., Chase, D. B. and Rabolt, J. F. (2004) Macromolecules, 37, 573–578. Deitzel, J. M., Kleinmeyer, J., Harris, D. and Beck Tan, N. C. (2001) Polymer, 42, 261– 272. Dersch, R., Liu, T., Schaper, A. K., Greiner, A. and Wendorff, J. H. (2003) Journal of Polymer Science, Part A: Polymer Chemistry, 41, 545–553. Dong, H. and Jones, W. E. (2004) In Abstracts of Papers, 228th ACS National Meeting, Philadelphia, PA, United States, August 22–26, 2004, pp. INOR-317. Dong, H., Megalamane, U. and Jones, W., Jr (2003) Polymer Preprints (American Chemical Society, Division of Polymer Chemistry), 44, 124–125. El-Aufy, A. K., Naber, B. and Ko, F. K. (2003) Polymer Preprints (American Chemical Society, Division of Polymer Chemistry), 44, 134–135. Fennessey, S. F. and Farris, R. J. (2004) Polymer, 45, 4217–4225. Fong, H., Chun, I. and Reneker, D. H. (1999) Polymer, 40, 4585–4592.

© 2007, Woodhead Publishing Limited

Controlling the morphologies of electrospun nanofibres

109

Formhals, A. (1934) US Patent 1975504 (Schreiber-Gastell, Richard). Frenot, A. and Chronakis, I. S. (2003) Current Opinion in Colloid & Interface Science, 8, 64–75. Gibson, P. and Schreuder-Gibson, H. (2000) Materials Division (American Society of Mechanical Engineers), 91, 45–61. Gibson, P., Schreuder-Gibson, H. and Pentheny, C. (1998) Journal of Coated Fabrics, 28, 63–72. Gibson, P., Schreuder-Gibson, H. and Rivin, D. (2001) Colloids and Surfaces, A: Physicochemical and Engineering Aspects, 187–188, 469–481. Gupta, P. and Wilkes, G. (2003) Polymer Preprints (American Chemical Society, Division of Polymer Chemistry), 44, 82–83. Hou, H., Jun, Z., Reuning, A., Schaper, A., Wendorff, J. H. and Greiner, A. (2002) Macromolecules, 35, 2429–2431. Huang, Z.-M., Zhang, Y.-Z., Kotaki, M. and Ramakrishna, S. (2003) Composites Science and Technology, 63, 2223–2253. Jayaraman, K., Kotaki, M., Zhang, Y., Mo, X. and Ramakrishna, S. (2004) Journal of Nanoscience and Nanotechnology, 4, 52–65. Jia, H., Zhu, G., Vugrinovich, B., Kataphinan, W., Reneker, D. H. and Wang, P. (2002) Biotechnology Progress, 18, 1027–1032. Jun, Z., Hou, H., Schaper, A., Wendorff, J. H. and Greiner, A. (2003) e-Polymers, Paper No. 9. Kameoka, J., Czaplewski, D., Liu, H. and Craighead, H. G. (2004) Journal of Materials Chemistry, 14, 1503–1505. Katta, P., Alessandro, M., Ramsier, R. D. and Chase, G. G. (2004) Nano Letters, 4(11), 2215–2218. Katti, D. S., Robinson, K. W., Ko, F. K. and Laurencin, C. T. (2004) Journal of Biomedical Materials Research, Part B: Applied Biomaterials, 70B, 286–296. Kebarle, P. and Peschke, M. (2000) Analytica Chimica Acta, 406, 11–35. Khatwani, P. A. and Yardi, S. S. (2003) Man-Made Textiles in India, 46, 19–23. Khil, M. S., Cha, D. I., Kim, H. Y., Kim, I. S. and Bhattarai, N. (2003) Journal of Biomedical Materials Research, Part B: Applied Biomaterials, 67B, 675–679. Kim, H. Y., Khil, M. S., Kim, H. J., Jung, Y. H. and Lee, D. R. (2003) In IEEE-Nano 2003, San Francisco, CA, USA, pp. 801–803. Kim, H. Y., Lee, B. M., Kim, I. S., Jin, T. H., Ko, K. H. and Ryu, Y. J. (2004) Polymeric Materials: Science and Engineering, 91, 712–713. Ko, F., Gogotsi, Y., Ali, A., Naguib, N., Ye, H., Yang, G., Li, C. and Willis, P. (2003) Advanced Materials (Weinheim, Germany), 15, 1161–1165. Koski, A., Yim, K. and Shivkumar, S. (2003) Materials Letters, 58, 493–497. Krishnan, J., Kotaki, M., Yanzhong, Z., Xiumei, M. and Ramakrishna, S. (2004) Journal of Nanoscience and Nanotechnology, 4, 52–65. Lee, J. S., Choi, K. H., Ghim, H. D., Kim, S. S., Chun, D. H., Kim, H. Y. and Lyoo, W. S. (2004) Journal of Applied Polymer Science, 93, 1638–1646. Lee, K. H., Kim, H. Y., Bang, H. J., Jung, Y. H. and Lee, S. G. (2003) Polymer, 44, 4029– 4034. Li, D. and Xia, Y. (2004a) Nano Letters, 4, 933–938. Li, D. and Xia, Y. (2004b) Advanced Materials (Weinheim, Germany), 16, 1151–1170. Li, D., Wang, Y. and Xia, Y. (2003) Nano Letters, 3, 1167–1171. Li, D., Babel, A., Jenekhe, S. and Xia, Y. (2004a) Advanced Materials (Weinheim, Germany), 16 (22) 2062–2066.

© 2007, Woodhead Publishing Limited

110

Nanofibers and nanotechnology in textiles

Li, D., Wang, Y. and Xia, Y. (2004b) Advanced Materials (Weinheim, Germany), 16, 361– 366. Li, W. J., Laurencin, C. T., Caterson, E. J., Tuan, R. J. and Ko, F. K. (2002) Journal of Biomedical Materials Research, 60, 613–621. Lin, T., Wang, H., Wang, H. and Wang, X. (2004) Nanotechnology, 15, 1375–1381. Lin, T., Wang, H., Wang, H. and Wang, X. (2005a) Journal of Material Science and Technology, 21, 9–12. Lin, T., Wang, H. and Wang, X. (2005b) Advanced Materials (Weinheim, Germany), 17, 2699–2703. Matthews, J. A., Wnek, G. E., Simpson, D. G. and Bowlin, G. L. (2002) Biomacromolecules, 3, 232–238. Megelski, S., Stephens, J. S., Chase, D. B. and Rabolt, J. F. (2002) Macromolecules, 35, 8456–8466. Pedicini, A. and Farris, R. J. (2003) Polymer, 44, 6857–6862. Pinto, N. J., Carrion, P. and Quinones, J. X. (2004) Materials Science & Engineering, A: Structural Materials: Properties, Microstructure and Processing, 366, 1–5. Reneker, D. H. and Chun, I. (1996) Nanotechnology, 7, 216–223. Shin, Y. M., Hohman, M. M., Brenner, M. P. and Rutledge, G. C. (2001a) Applied Physics Letters, 78, 1149–1151. Shin, Y. M., Hohman, M. M., Brenner, M. P. and Rutledge, G. C. (2001b) Polymer, 42, 9955–9967. Smit, E., Buttner, U. and Sanderson, R. D. (2005) Polymer, 46(8), 2419–2423. Subramanian, A., Lin, H. Y., Vu, D. and Larsen, G. (2004) Biomedical Sciences Instrumentation, 40, 117–122. Sun, Z., Zussman, E., Yarin, A. L., Wendorff, J. H. and Greiner, A. (2003) Advanced Materials (Weinheim, Germany), 15, 1929–1932. Sundaray, B., Subramanian, V., Natarajan, T. S., Xiang, R.-Z., Chang, C.-C. and Fann, W.-S. (2004) Applied Physics Letters, 84, 1222–1224. Suthar, A. and Chase, G. (2001) The Chemical Engineer, 726, 26–28. Suthar, A. and Chase, G. (2002) Fluid/Particle Separation Journal, 14, 58–64. Teo, W. E. and Ramakrishna, S. (2005) Nanotechnology, 16(9), 1878–1884. Teo, W. E., Kotaki, M., Mo, X. M. and Ramakrishna, S. (2005) Nanotechnology, 16(6), 918–924. Theron, A., Zussman, E. and Yarin, A. L. (2001) Nanotechnology, 12, 384–390. Tsai, P. and Schreuder-Gibson, H. L. (2003) Advances in Filtration and Separation Technology, 16, 340–353. Won, K. S., Ji, H. Y., Taek, S. L. and Won, H. P. (2004) Polymer, 45, 2959–2966. Xu, C. Y., Inai, R., Kotaki, M. and Ramakrishna, S. (2004) Biomaterials, 25, 877–886. Yao, L., Haas, T. W., Guiseppi-Elie, A., Bowlin, G. L., Simpson, D. G. and Wnek, G. E. (2003) Chemistry of Materials, 15, 1860–1864. Yarin, A. L. (1993) Free Liquid Jets and Films: Hydrodynamics and Rheology, John Wiley & Sons Inc., New York. Yoshimoto, H., Shin, Y. M., Terai, H. and Vacanti, J. P. (2003) Biomaterials, 24, 2077– 2082. Yu, J. H., Fridrikh, S. V. and Rutledge, G. C. (2004) Advanced Materials (Weinheim, Germany), 16(17), 1562–1566. Zhang, G., Kataphinan, W., Teye-Mensah, R., Katta, P., Khatri, L., Evans, E. A., Chase, G. G., Ramsier, R. D. and Reneker, D. H. (2005) Materials Science & Engineering, B: Solid-State Materials for Advanced Technology, B116(3), 353–358.

© 2007, Woodhead Publishing Limited

Part II Carbon nanotubes and nanocomposites

111

© 2007, Woodhead Publishing Limited

6 Synthesis, characterization and application of carbon nanotubes: the case of aerospace engineering M. R E G I, University of Rome ‘La Sapienza’, Italy

6.1

Introduction

Nanotechnology is the design, production (synthesis) and application of nanomaterials and nanostructures in advanced macro- and micro-systems, through an understanding of the fundamental relationships between physical properties and materials. This innovative science deals with materials and structures at the nanometer scale, typically from subnanometer to several micrometer (hundred nanometers). Similar to quantum mechanics, in this case, the properties and characterization of a material or structure can significantly differ from the macro-scale conditions. Some nanomaterial properties, for example, tuning semiconductor band gaps by varying the material dimensions, are already known. Ultra-strong and ultra-light multifunctional materials (e.g. carbon nanotubes) can be produced through enhanced nanotechnology. Furthermore, super-conductivity and superior magnetic properties provided by nanomaterials can be valuable in improving the properties of electromagnetic devices (biomedical sensors, thermal management, health monitoring, etc.). An important advantage is component miniaturization made possible by nanotechnology, which has a strong impact on systems performance. To date, it is possible to develop very powerful computers able to simulate some functions of the human brain, smart molecular biosensors (nuclear, bacteriological and chemical (NBC) applications and environment monitoring), micro- and nano-electromechanical systems (MENS/ NEMS) or nanorobots that can repair internal damage and remove toxins from human bodies. Nanotechnology has a very broad range of potential applications in many scientific fields. Therefore, its development requires multidisciplinary teams (engineers, chemists, materials scientists, biologists and others) working together on: • nanomaterial and nanostructure fabrication; • nanomaterial and nanostructure characterization, evaluating their properties and possible applications; 113

© 2007, Woodhead Publishing Limited

114

Nanofibers and nanotechnology in textiles

• understanding the relationship between physical properties and nanoscale structure; • designing the nano-device and its integration into a macro-system; • evaluating the human impact of the nanotechnology. For each step it is necessary to develop specific and innovative procedures and tools able to satisfy a range of requirements. The synthesis of nanomaterials and nanostructures is an essential aspect of nanotechnology. A real application is possible only when the nanostructured materials are available with the specific properties desired (size, chemical composition, morphology, physical behaviour). The fabrication of nanomaterials started a long time ago, but it is only in the past ten years that nanotechnology has been a specific scientific sector. Owing to its rapid expansion, it is very difficult to cover all sectors of this innovative science. However, it is important to observe that in nanotechnology many scientific sectors, such as engineering and biology, typically very ‘distant’ from each other, can work together in the development of nanosystems and devices. The final target of nanotechnology is the control of materials and apparatus on nanoscale dimensions. A ‘nano’ material or device can improve the properties and characteristics of many systems. In several sectors, current technologies (e.g. the use of silica in a computer’s microchip) have reached their physical limit. Nonetheless, using the bottom-up or the top-down approach, further improvement is possible. Time and cost are two other important aspects in nanotechnology. As an innovative scientific sector, nanotechnology requires meaningful schedules and costs to be established. Nanosystems and devices must be economically competitive with respect to other traditional and established methodologies. This is a critical issue, since industry requires greater economic competitiveness for nanotechnology products, with tighter schedules and costs compared with those requested by R&D scientific programs. Consequently, there is a need for a double approach (industrial and scientific) to realize competitive nanotechnology systems and devices, which requires the following: • synthesis of nanomaterials; • purification (post-synthesis, useful to obtain a high degree of purification of the nanomaterial); • characterization (using electronic microscopy, and other techniques of observation); • functionalization (necessary to ‘prepare’ the nanomaterial for the next step); • integration of the nanomaterial, or device, in a macro-advanced system. In the above phases, control of parameters and reproducibility of results are essential. Therefore, it is clear that nanotechnology activities are very complex and characterized by different specialized scientific sectors, requiring:

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

• • • •

115

significant time and economic investments; ‘multidisciplinarity’; interaction between technicians and researchers of different scientific sectors; use of complex and advanced instruments and apparatus.

The possible applications of the nanotechnologies to aerospace engineering are: • composites (metallic and/or polymeric) with nanoparticles (e.g. carbon nanotubes) embedded in the matrix, for structural applications; • flat panels for aerospace electronics devices; • special nanostructured coatings for high-temperature conditions (re-entry space mission phase); • nanosensors (thermal, electromagnetic, biological, etc.); • MEMS/NEMS; • solid nanostructured propellants; • frequency selective surfaces (FSS) with electromagnetic compatibility; • thermal management; • nanodevices (nano-cantilever, -diode, -memories for personal computer, -transistor, etc.); • astronaut health monitoring; • biological and biomedical applications; • development of nanosystems in zero gravity conditions. These aspects will be illustrated in the following sections.

6.2

The development and structure of carbon nanotubes1–64

The first carbon filament possessing a very small diameter (less than 10 nm) was produced in the 1970s by the decomposition of hydrocarbon at high temperature using transition metal catalysts. The 1985 discovery of the third allotropy form of ordered carbon (after the graphite and diamond forms, respectively hybridized sp2 and sp3), commonly known as fullerene, spurred the subsequent discovery of a number of related forms of carbon. Among these the most famous are nanotubes (Fig. 6.1), observed for the first time in 1991 by Iijima of the NEC Laboratory in Tsukuba, Japan, using HRTEM (High-Resolution Transmission Electron Microscopy). During the same period, Russian researchers also reported the observation of carbon nanotubes and other forms of carbon nanostructures. The discovery of these nano-elements and nanostructures offered the opportunity to understand how carbon atoms bond with other carbon atoms, and how carbon reacts with other elements under specific conditions of temperature and pressure. With the benefit of theoretical models and

© 2007, Woodhead Publishing Limited

116

Nanofibers and nanotechnology in textiles

6.1 SEM micrograph of carbon nanotube bundles deposited on the cathode electrode surface.56

experimental data, it is now possible to use fullerenes to significantly improve the properties and characteristics of many advanced technologies and systems. The studies of Kroto, Smalley and coworkers (at Rice University) showed that carbon nanotubes constitute a particular case of the fullerene family. The most famous and most stable of the fullerene molecules is C60, with a computer-simulated image very similar in appearance to a soccer ball. This molecule consists of 20 hexagonal and 12 pentagonal faces, with the carbon atoms at each corner of the individual polygons. C60 stability has been attributed to the ‘pentagon rule’ and the satisfaction of all valences when the pentagon faces lead to the closure of individual C60 molecules. The connection between carbon nanotubes and other fullerenes has been defined by the observation that the nanotubes were closed by fullerene-like caps or hemispheres. It is interesting to observe that the smallest reported carbon nanotube diameter is the same as the diameter of C60. This is important in evaluating the minimum dimension of carbon nanostructures. It is necessary to identify all types of nanoparticles and nanostructures of the fullerene family (multiwall and/or single-wall nanotubes, carbon-encapsulated metal nanoparticles, fullerene black and soot, carbon onion, nanowhiskers, etc.). For each nanostructure it is possible to define a set of physical and chemical properties and subsequent applications. It is also interesting to explore the interrelationships between the various nanostructured carbon forms, as well as their relation to the traditional forms of ordered carbon atoms such as diamond and graphite. Carbon is a unique material and can be a good metallic

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

117

conductor in the form of graphite, a wide band gap semiconductor in the form of diamond, and a polymer when reacted with hydrogen. As discussed above, carbon provides a significant example of a material showing the entire range of intrinsic nanometer-scaled structures ranging from fullerene, one-dimensional nanoparticles, carbon nanotubes, one-dimensional nanowires, graphite, two-dimensional layered anisotropic materials, fullerene solids, and three-dimensional bulk materials with the fullerene molecules as the fundamental building block of the crystalline phase. Iijima’s discovery of multiwalled carbon nanotubes enabled many scientists to explore the field of nanotechnology, and in particular the world of carbon nanostructures, stimulated initially by the significant one-dimensional quantum effects predicted for their electronic properties, and subsequently by the possibility that the remarkable structure and properties of carbon nanotubes might provide significant improvements in many scientific and technological sectors. Two years after the HRTEM observation of the multiwalled carbon nanotubes, Iijima and his group, in collaboration with the IBM Almaden Laboratory, discovered single-wall carbon nanotubes. This was a very important discovery, because the single-wall configuration represents the fundamental form of carbon nanotubes. It has proved possible to study the fundamental structure by both numerical simulations and experimental tests. The goal is to correlate carbon nanotube properties (mechanical, thermal, electromagnetic, chemical and physical behaviour) to the geometrical characteristics (diameter, length, chirality, hexagon orientation in respect to the nanotubes axis, defects, etc.). It has proved possible to successfully synthesize bundles of carbon nanotubes with very good alignment and a high purity. Control of the parameters used (pressure, temperature, power supply, inert environment condition) is crucial in obtaining a consistent product. Theoretical and experimental work has focused on the correlation between carbon nanotube morphology and properties. It gives the opportunity to investigate the potential applications in macro-, micro- and nanoscience and technologies. Nanotubes can be utilized, individually or as an assembly, to build functional device prototypes in many scientific sectors, with improved properties and characteristics. The full potential of carbon nanotubes will be reached when growth and synthesis mechanisms are well defined and controlled. Its real application requires the availability of large quantities of material of high quality at low cost. In hydrogen storage, for example, it is necessary to obtain high-quality carbon nanotubes in kilogram quantities using a simple, efficient and cheaper synthesis method. In electronics applications the quantities needed are significantly smaller. The challenge is to produce carbon nanotubes efficiently, obtaining defect-free nanostructures with high length, in large scale with complete dimensional control (length, diameter, chiralities, etc.) provided by a well-defined synthesis process.

© 2007, Woodhead Publishing Limited

118

6.2.1

Nanofibers and nanotechnology in textiles

The structure of carbon nanotubes65–91

The structure of carbon nanotubes is conveniently explained in terms of a seamless array of one or more coaxial cylindrical sheets of graphite with an aspect ratio typically greater than 100 and with outer diameter measuring tens of nanometers, and closed at the end with two semi-domes. The creation of carbon nanotubes can be traced back to the discovery of the fullerene structure C60 (buckyball) in 1985 by Richard Smalley and Harold Kroto. When the buckyball is elongated to form a long and narrow tube with a diameter of approximately 1 nm (10–9 m), it provides the basic form of a carbon nanotube (Fig. 6.2). The basic element is graphite, constituted by a series of planes one above the other, held together by van der Waals forces. Each plane has a bidimensional covalent structure. Through a series of processes to fold up these planes of graphite, it is possible to create a tubular seamless structure that does not exist in nature. This structure takes the name of carbon nanotube. Essentially, two families of carbon nanotubes exist (Fig. 6.3): • SWNT or single-wall nanotubes, that are made up of only one straight tubular unit; • MWNT or multiwall nanotubes, that are made up of a series of coaxial tubes about 0.34 nm apart (same distance among the various planes of the graphite). Carbon nanotubes can thus be visualized as a sheet of graphite that has been rolled into a tube. Unlike the diamond (sp2 hybridization), where the 3D diamond cubic crystal structure is formed with each carbon atom having four

Diamond

Buckyball C60

Graphic

Armchair nanotube (10,10)

6.2 The carbon family: graphite, diamond, fullerenes and carbon nanotubes.92

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

119

6.3 TEM micrographs of the two principal carbon nanotube typologies: SWNT – single-wall nanotubes (in the center), and MWNT –multiwall nanotubes (on the left and right).92

nearest neighbours arranged in a tetrahedron, graphite (sp3 hybridization) is formed as a 2D sheet of carbon atoms arranged in a hexagonal array. In this case, each carbon atom has three nearest neighbours. ‘Rolling’ the sheet of graphite into cylinders forms carbon nanotubes. The properties of carbon nanotubes depend on atomic arrangement (how the sheet of graphite is ‘rolled’), the diameter and length of the tubes, and the morphology or nanostructure. Using different synthesis methods and specific process parameters, it is possible to obtain different carbon nanotube morphologies and properties, with the potential for multipurpose applications in many scientific fields. It is possible to define the geometric parameters of carbon nanotubes. The diameter is expressed in terms of the chiral vector (Ch) which connects two crystallographically equivalent sites on a 2D graphite sheet (Fig. 6.4): Ch = na1 + ma 2

[6.1]

where

and

a1 =

a 3 x+ ay 2 2

[6.2]

a2 =

a 3 x– ay 2 2

[6.3]

© 2007, Woodhead Publishing Limited

120

Nanofibers and nanotechnology in textiles

y

B′

x

B θ

T A τ

a1

O

Ch a2 Ψ (a)

Zigzag (0, 0)

(1, 0)

(2, 0)

(1, 1)

(3, 0)

(2, 1)

(4, 0)

(3, 1)

(2, 2)

(5, 0)

(4, 1)

(3, 2)

(5, 1)

(4, 2)

(3, 3)

(6, 0)

(6, 1)

(5, 2)

(4, 3)

(7, 0)

(8, 0)

(7, 1)

(6, 2)

(5, 3)

(10, 0)

1

3

(8, 1)

(7, 2)

(6, 3)

(9, 0)

(9, 1)

(8, 2)

(7, 3)

(4, 4)

a2

(5, 4)

(6, 4)

24

(11, 1)

3

20

(9, 2)

(10, 2)

(11, 2)

10

19

56

(8, 3)

(9, 3)

7

18

48

(7, 4)

(8, 4)

(9, 4)

(10, 4)

5

17

43

92

(5, 5)

(6, 5)

(7, 5)

(8, 5)

(9, 5)

1

1

13

37

80

x Metal

17 (10, 1)

(8, 3)

a1 y

(11, 0) (12, 0)

Semiconductor

(6, 6)

(6, 7)

(7, 7)

15

32

87

Armchair

(b)

6.4 Geometry construction rule for different carbon nanotube typologies (θ = chiral angle, ψ = length of vector a1, τ indicates direction (dotted line) given by the vector a1).65

with a = 2.46 ⇒ Å in an (x, y) coordinate system. The chiral angle (θ) is defined as (Fig. 6.4): cos(θ ) =

© 2007, Woodhead Publishing Limited

2n + m 2 n + m 2 + nm 2

[6.4]

Carbon nanotubes in aerospace engineering

121

where n and m are two integer numbers. The chiral vector is perpendicular to the tube axis, while the chiral angle is the angle between Ch and the socalled zigzag direction (θ = 0°). The carbon nanotube diameter (d) is calculated by the following formula: d=

| Ch | a cc 3( n 2 + m 2 + nm ) = π π

[6.5]

with: 1.41Å ≤ acc ≤ 1.44 Å

[6.6]

where the inferior extreme corresponds to the graphite, while the superior extreme corresponds to the buckyball C60. Since Ch, d and θ are expressed as a function of the integers n and m, there is enough to represent any carbon nanotube typology using the following notation: (n, m)

(n, m) = (5,5)

(n, m) = (9,0)

(n, m) = (10,5)

6.5 Carbon nanotube typologies: armchair, zigzag and chiral.92

© 2007, Woodhead Publishing Limited

[6.7]

122

Nanofibers and nanotechnology in textiles

Then, changing the chiral angle (θ), three carbon nanotube families can be distinguished (Fig. 6.5): 1. θ = 0° → zigzag carbon nanotubes with (n, 0) or (0, m). 2. θ = 30° → armchair carbon nanotubes with n = m. 3. 0° < θ < 30° → chiral carbon nanotubes with n ≠ m. If n ≠ m ≠ 0, there is chiral symmetry, while if n = 0 or n = m there is achirality. Each carbon nanotube typology can be characterized in various ways: • • • • •

morphological; mechanical; thermal; electromagnetic; chemical and physical stability and/or reactivity, etc.

Figure 6.4 indicates the carbon nanotubes that are semiconducting and those that are metallic. It shows the number of distinct fullerene caps that can be used to close the end of (n, m) the nanotubes. The electrical behavior of carbon nanotubes can be determined, using the following simple rule:  n – m = 3 q = metallic   n – m ≠ 3 q = semiconducting  q = integer 

[6.8]

Table 6.1 summarizes the principal carbon nanotube parameters and Table 6.2 shows their principal properties (mechanics, electrics, thermal, etc.). Since 1991, great progress has been made in understanding the morphology and properties of carbon nanotubes. There has been a constant fruitful interplay between theoretical models and experimental activity which has enhanced researchers’ knowledge in this field. Although there are still many fundamental studies to be done, it is possible to define applications of carbon nanotubes in many sectors. These are mostly related to three unique features of the nanotubes: small dimensions, electronic and mechanical properties. They can be used for example as nanomolds and templates for making small structures of other materials; they can be modified for catalytic purposes and used for gas storage. Conductive carbon nanotubes are excellent for field emissions owing to the high curvature at the tips, while in bulk form they might make good wires. In this case the technological problem is to be able to manipulate the carbon nanotubes individually in a practical way to find use in nanoelectronics devices. In general, bulk applications are most promising at present. Among these, nanotube reinforced materials (polymeric, ceramic, metallic) are excellent candidates. Their exceptionally high strength combined with their light weight makes them ideal for this structural purpose. Nanotubes are the ultimate carbon fibers in terms of strength to weight ratio so it would

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

123

Table 6.1 Characteristic parameters of carbon nanotubes Symbol

Description

Formula

Value

aC–C

Carbon–carbon atom distance

–

1.42 ⇒ Å

a

Length of unit vector

a1, a2

Unit vectors

 3 1  3 1  2 , 2  a,  2 , – 2  a    

in (x, y) coordinate system

Ch

Chiral vector

Ch = na1 + ma2 ≡ (n, m)

(n, m): integer

L

Circumference of nanotube

L = | Ch | = a n 2 + m 2 + nm

0° ≤ | m | ≤ 30°

d

Nanotube diameters

d= L = π

θ

Chiral angle

sin ϑ =

2.46 ⇒ Å

3a C–C

cos ϑ =

tan ϑ =

n 2 + m 2 + nm

π 3m 2

2 n + m 2 + nm

a

0° ≤ | ϑ | ≤ 30°

2n + m 2

2 n + m 2 + nm 3m 2n + m

Table 6.2 Principal properties of carbon nanotubes Parameter

Value

Diameter

O (nanometers)

Length

Several micrometers

Density

1.33–1.40 g/cm3

Tensile strength

45 GPa

Young’s modulus

∼1–4 TPa

Electrical properties

Metal or semiconductor

Current density

1 × 109 A/cm2 (estimated)

Field emission

Activation of phosphorus compounds at ∼1–3 V with 1 µm spacing electrodes

Thermal conductivity

6000 W/mK

Thermal stability

Stable up to 2800 °C in vacuum and 750 °C in atmosphere

© 2007, Woodhead Publishing Limited

124

Nanofibers and nanotechnology in textiles

be surprising if they could not find a niche in this engineering sector. Other possible applications are: health monitoring, biomedical sensors, propulsion and thermal managements. The critical aspect is their production costs. Until this is brought down to a level competitive with existing fibers, large-scale use of carbon nanotubes will not take place. But, with the right theoretical models and with experimental confirmation of their unique qualities and properties, there will be a greater incentive to develop industrial scale production. In the meantime, the increased know-how brought about by the study of nanotubes is having a strong impact on traditional carbon research and development and on many other fields.

6.3

Synthesis of carbon nanotubes

The first step in the development of advanced nanotechnology systems is the synthesis of the nanomaterial. Regarding carbon nanotubes, three principal methods are available: • arc discharge (on inert environment, water immersed, plasma arc); • laser ablation (CO2, Nd-Yag); • chemical vapor deposition (CVD; thermal, hot filament, plasma enhanced). The target is to produce a large quantity of carbon nanotubes (or nanomaterials in general) with a high degree of purity and alignment levels, uniform property distributions and low costs. Producing carbon nanotubes with the above qualities is necessary for current potential applications to become marketable. This requires solving some scientific and technological problems that can be more or less complex depending on each specific synthesis method. One example is the chirality control of carbon nanotubes, with production capability adapted to each specific application (composite material, electronics devices, heat management, etc.). One requirement would be to understand perfectly the mechanism of nanotube nucleation and growth, which remains a key area of research. This problem is partially explained by the complex physical aspects of the synthesis process such as the control of parameters and the reproducibility of results. Thanks to the large number of experimental parameters involved in the synthesis process, and considering the large range of conditions influencing each process, it is legitimate to suppose that many physical and chemical mechanisms affect the synthesis process. Using the above methods, typical synthesis products are: • nanoparticles, carbon nanocapsules, nanofibers and whiskers; • graphite structures; • carbon nanotubes. These products use graphite as the base material. In fact, using different precursors (e.g. aluminium, ceramics), with the same technological methods,

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

125

it is possible to produce other kinds of nanomaterials (e.g. aluminium nanotubes, ceramic nanoparticles). The principal parameters employed during the synthesis process are: • • • • •

temperature; pressure; gas (inert or not); time of synthesis; voltage and amperage (and the consequent power provided to the synthesis apparatus); • base materials (graphite); • catalysts (yttrium, cobalt, nickel, molybdenum, etc.), used for optimizing the synthesis capability thanks to the improvement of chemical and physical phenomena occurring during the process. Each parameter is fundamental in order to obtain the specific carbon nanotube typology. In particular, temperature and electrical parameters (voltage, amperage) represent the energy provided to the graphite necessary for the growth of carbon nanotubes. One of the advantages of these different synthesis techniques is the possibility of varying a large number of parameters (described above), allowing the characterization of the optimal conditions for the control of carbon nanotube formation. The major drawback of these techniques is that the carbon nanotubes are never pure, i.e. they are associated with other carbon phases and catalysts, therefore requiring a post-synthesis purification step. The full application potential of carbon nanotubes will not be realized until the growth mechanisms can be optimized and well controlled. Realworld applications of this innovative material (sometimes called the material of the 21st century) require either large quantities of bulk materials or device integration in scaled-up form. For applications such as structural composites and hydrogen storage, it is necessary to obtain high-quality nanotubes at the kilogram or tonne level (industrialization of the process) using growth methods that are simple, efficient and inexpensive. For devices such as nanotubebased electronics, scale-up will unavoidably rely on self-assembly or controlled growth strategies on surfaces combined with microfabrication techniques. Significant work has been carried out in recent years to tackle these issues. Nevertheless, many challenges remain in the nanotube growth area. First, an efficient growth approach to structurally perfect nanotubes on a large scale is still lacking. Second, growing defect-free nanotubes continuously to macroscopic lengths has been difficult. Third, control is needed over nanotube growth on the surface in order to obtain large-scale ordered nanowire and nanofiber structures. Finally, there is the seemingly formidable task of controlling nanotube chirality by any existing growth method. Before describing the synthesis methods, it is useful to outline their characteristics.

© 2007, Woodhead Publishing Limited

126

Nanofibers and nanotechnology in textiles

• Arc discharge in inert environment: • The simplest and oldest technique. • Gives high quantities of carbon nanotubes. • Requires a specific control of the environment using inert gases (helium and/or argon). • It is necessary to provide an electrode cooling system. • The purity and alignment level cannot be optimized without specific parameter control (use of the plasma arc discharge makes it possible to solve these problems). • It is very fast. • Arc discharge water immersed method: • The use of inert gases is not needed. • Thanks to the deionized water, cooling systems are not necessary. • The partial evaporation of deionized water may provide arc instability during the synthesis. • CVD: • Provides an intrinsic good alignment level of the carbon nanotubes produced. • Only small quantities of carbon nanotubes can be produced. • High level of purity. • Many crystallographic defects in carbon nanotubes produced may be found. • It is slow. • Laser ablation: • High quantity of carbon nanotubes produced. • Better control and repeatability of the process parameters compared to the arc methods. • Cost reduction. • Complex facilities (lasers, ovens) required. For realistic use of the carbon nanotubes in advanced applications using the above methods, it is necessary to achieve: • • • •

improved synthesized quantity; better parameter control and consistency of product; significant cost reduction; an industrial scale process, i.e. it is necessary to transform the prototype synthesis facilities into industrial apparatus.

6.3.1

Arc discharge1–14

In 1991 Iijima discovered carbon nanotubes in the cathode deposit of an electrode used in the arc discharge. This method is the oldest technique used in nanomaterial synthesis processes. To understand the phenomenon of

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

127

nanotube growth in the arc discharge method it is relevant to consider the optimization of other techniques employed (laser ablation, CVD) and, that in general, arc technology is widely employed in several industrial applications. It is also necessary to evaluate the macroscopic phenomena, as per electrode spot, material flow and atomics behavior during the arc process. Typical experimental apparatus used to perform synthesis by the arc method is shown in Fig. 6.6. The arc was generated between two electrodes with the following parameters: • Voltage: 20–30 V. • Current: 60–120 A (there are two possible arc discharge configurations: DC (direct current) and AC (alternating current). • Pure or doped graphite electrodes. • Distance between the electrode surface exposed to the arc: 1–3 mm. • Inert gas: helium and/or argon (the pressure in the synthesis chamber is controlled, and after the process a vacuum is obtained to avoid oxidation of electrodes). • Discharge time: 10–60 s. The electrode configuration is an important aspect of this method. Typically, a homo-electrode configuration is employed in which the cathode electrode is made of graphite (pure or doped, diameter: 10–15 mm), and the anode electrode is also made of graphite (pure or doped, diameter: 3–5 mm). Another configuration is the so-called hetero-electrode in which over the graphitic

6.6 Arc discharge apparatus in inert environment conditions, consisting of a synthesis chamber, electrode alignment and cooling system, observation window, inert gas ignition and exhausting points, electrode gag control system.53

© 2007, Woodhead Publishing Limited

128

Nanofibers and nanotechnology in textiles

cathode, a metallic anode is used (e.g. molybdenum). The catalysts used to dope the electrodes are: yttrium, cobalt, nickel. They improve the quantity and quality of carbon nanotubes synthesized. During the process, arc stability is needed (using the methodology of the arc plasma discharge it is possible to obtain a self-sustained arc with an improvement in the synthesis of carbon nanotubes). This can be obtained by accurate control of the electrode gaps (usually an electronic controller is used), and also by using a specific shape of the electrode surfaces exposed to the arc (typically flat or cupped for the cathode, and conical for the anode). After the process the cathode electrode surface shows several consecutive craters created by the random movement of the generated arc. The areas surrounding the craters appeared shiny gray or silver on visual observation. As shown in Fig. 6.7, on the cathode electrode surface it is possible to distinguish four typical regions, A, B, C and D. • In A, the crater spot generated by the arc is visible, and in this region no carbon nanotubes are present, only micro-spheres deposited during the process. • In B, many carbon nanotubes are observed (Fig. 6.8) with a good purity level (some residual catalysts are present), and the alignment depends on

6.7 Cathode electrode surface morphology after the exposure to the arc discharge. Four characteristic regions (A, B, C and D) are present: A no carbon nanotubes, B high quantities of carbon nanotubes (purity and alignment level depends on the parameters employed and by process control), C small carbon nanotube quantities and more impurities deposited on the surface, D without nanostructured materials (this region is not affected by arc phenomena).51

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

129

6.8 SEM micrograph with carbon nanotubes produced by the arc discharge method, in the region B of Figure 6.7.55

6.9 SEM micrograph with carbon nanotubes produced by the arc discharge method, in the region C of Figure 6.7 with residual catalysts and amorphous micro- and nano-carbon particles.55

the parameter control used (an intrinsic alignment of the nanotubes is typically provided by the CVD method). • C (Fig. 6.9) is characterized by the presence of carbon nanotubes with major quantities of impurities (catalysts, amorphous carbon, other carbon

© 2007, Woodhead Publishing Limited

130

Nanofibers and nanotechnology in textiles

micro- and nano-structures) compared with B. Usually after the arc process a purification step for the produced nanomaterials is required. • In D, no change caused by the arcing process was observed in the original graphite surface. The aim of the arc method is to improve the quantity of carbon nanotubes produced (industrialization), maximize region B and reduce catalysts and amorphous residuals. Varying the parameters changes the synthesis results. For example, as shown in Fig. 6.10, changing the electrode gap and increasing the inert gas flow make it possible to modify the geometry and the dimensions of the above four regions. In particular, it is possible to observe an extension of the crater. Naturally, if other parameters are modified (voltage, amperage, discharge time, etc.), further different results are obtainable. The configuration most often used, the homo-electrode system, consists of both graphitic anode and cathode (C-anode and C-cathode). The cathode deposit is cylindrically composed of two macroscopic parts (in which the above four regions are contained): the outer glossy gray hard shell, and inner dark black soft core. In the homo-electrode system the anode spot is larger than the cathode spot and the mass erosion of the anode is much greater than that of the cathode. This explains the large quantity of carbon nanotubes present in the C-cathode in respect to the C-anode in the DC arc discharge. In the AC arc discharge method, theoretically, it is possible to find the same

6.10 Using different synthesis parameters (gas flux, pressure, temperature, etc.) makes it possible to change the morphology of the cathode electrode surface.56

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

131

carbon nanotube quantities in both electrodes, but in this case it is more difficult to control the arc stability, due to the electrical polarity inversion. In the homo-electrode systems (DC arc discharge), during the arc, carbon atoms, evaporated, due to the high temperature developed from the anode, are deposited on the cathode surface (B and C). These are then reheated and re-evaporated by the arc, with consequent carbon nanotube growth. In fact, in the central part of the cathode surface involved in the process, nanotubes are synthesized by the random movement of the arc, while on the external region (D), materials transferred from the anode are deposited without the reheating and re-evaporation phase. This explains the absence of carbon nanotubes in D. This explanation of the phenomena occurring during arc discharge is just macroscopic; in fact, several theories describe carbon nanotube growth during arc discharge with atomic models. It is necessary to distinguish between models (mainly semi-empirical) useful for an engineering implementation of carbon nanotube synthesis, and theoretical models indispensable for acquiring the know-how on physical and chemical phenomena characterizing carbon nanotube formation. From a technological point of view, the synthesis results depend strongly on a large number of parameters. Complex parametric analysis and scientific investigations are required for a microscopic understanding of carbon nanotube growth mechanism. Optimizing the arc process produces very interesting results. Figure 6.11a shows the cathode electrode surface in which the four regions (A, B, C, D) are well controlled, i.e. with a clean transition from one region to another, with the maximum extension of B, thanks to the reduction of the other three regions (A, C and D). Figure 6.11b shows an SEM micrograph of the carbon nanotubes present in B, with a very high purity level. The alignment is not excellent, but this is an intrinsic characteristic of the arc discharge method. Another method is the so-called arc discharge water immersed (using deionized water) method. As shown in Fig. 6.12, vacuum and inert gas systems are not required. In this case the carbon nanotubes on the electrode surfaces, are present in the water suspension. This ‘water method’ presents some problems: • the control of the arc stability is very complex; • the quantity of carbon nanotubes produced is less than that obtained by ‘inert environment arc discharge’; • the industrialization of the method seems more complex. Owing to the many possible configurations applicable and several modifiable parameters, the arc method, which can be considered the traditional carbon nanotube synthesis method, has a greater possibility of being improved.

© 2007, Woodhead Publishing Limited

132

Nanofibers and nanotechnology in textiles (a)

(b)

6.11 Optimization of the arc discharge process parameters makes it possible to improve the carbon nanotubes produced by specifically controlling the four characteristic regions and the purity level (as shown by the SEM micrograph in panel b) of the nanomaterials.56

6.3.2

Laser ablation15–24

This technique is an upgrade of the arc discharge method. Two typical laser methodologies are employed: • Nd-Yag: laser ablation of a carbon rod (graphite pure or doped) at a temperature of 1200 °C in an argon inert flow. A two pulse sequence at 10 Hz from an Nd-Yag laser (wavelength: green at 532 nm; energy: ∼50 mJ

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

133

6.12 Water immersed arc discharge prototype apparatus, used to perform a base experimental test of innovative synthesis methods.55

per pulse, 6–7 ns) followed by 50 ns delayed second laser pulse at a wavelength of 1064 nm (IR, energy: ∼50 mJ per pulse, 4–6 ns) was directed to the above carbon target, with the growth of carbon nanotubes. • CO2: laser ablation of a carbon rod (graphite pure or doped) at room temperature (no oven is needed) in an argon inert flow. A single continuous CO2 laser (wavelength: 1064 nm; power: 400–900 W) was directed to the above carbon target with the growth of carbon nanotubes. The typical laser experimental set-up used to produce carbon nanotubes is shown in Fig. 6.13. It consists of a quartz tube inside a furnace (not needed in the case of the CO2 ablation). The tube is sealed and connected to a vacuum system and an inert gas reservoir. The laser beam enters the quartz tube through a special window mounted in a vacuum flange. The carbon target is placed in the center of the quartz tube and is aligned to the laser beam. The distance between the laser and the carbon target is changeable due to a small quartz tube coaxial to the water-cooled metallic collector mounted at the other end of the external tube. During the synthesis process the temperature profile is measured by thermocouples collocated into the tube. Briefly, the laser ablation process can be summarized as follows: • The laser beam (double pulsed or continuous) shoots at the carbon target. • An atomic carbon plume is produced and carried, by the inert gas flow, to the water-cooled metallic (typically aluminum or copper) collector.

© 2007, Woodhead Publishing Limited

134

Nanofibers and nanotechnology in textiles

6.13 Laser ablation CO2 apparatus for carbon nanotube synthesis.55

• The nanostructures (carbon nanotubes, fullerenes and other carbon elements) are deposited to the collector surface. The laser ablation method has been proven to be the most efficient technique for high-purity carbon nanotube production. Many papers focus on studying the effects of processing parameters, i.e. the synthesis optimization (as a function of the catalyst’s concentration, furnace temperature, gas flow, pressure, energy provided, etc.). For instance, the experimental tests demonstrate that the bimetallic catalyst mixture (e.g. Ni/Co) is more efficient than the use of a single metal. Moreover, the furnace temperature and gas conditions (flow and pressure) directly influence the production yield of the carbon nanotubes and, in particular, their diameter distribution. The increase in the laser intensity favors the growth of large nanotube diameters. Also the temperature has an important role in the synthesis. In particular, for values less than 800–900 °C the yield of the produced nanostructures decreases, improving the amorphous carbon element deposited on the collector. It is important to observe that the temperature value is proportional to the energy provided to the carbon target. In general, it is found that the tendency to favor the growth of high carbon nanotube yield is a function of the temperature and the laser intensity. Thus, each optimized laser intensity corresponds with an ‘optimal heating’ of the carbon target that gives better synthesis results. This is also consistent with the fact that the double pulse configuration, with respect to the continuous laser spot, provides a maximization of the energy absorption by the target surface, leading to a higher carbon nanotube yield.

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

135

As for the arc discharge, the phenomena occurring during the laser ablation process are very complex. By in situ imaging and spectroscopic diagnostic investigation, it is possible to analyse the carbon nanotube growth mechanism, corresponding to a single laser pulse. Initially the laser pulse produces atomic molecular vapor (plume) containing ∼5 × 1016 carbon atoms and ∼1014 catalyst atoms (this value has been estimated by weighing the target before and after ablation). The evaporated materials remain in the vapor phase until approximately 100 µs after the ablation. Then, laser plasma, which is initially very hot, cools rapidly, increasing the population of the atomic and molecular species. The analysis shows that carbon atoms condense and form clusters after 200 µs from the initial ablation, while the metallic catalyst atoms (typically Ni/Co) condense much later. The atomic catalyst population is maximum at t = 0.8 ms and then condenses by t = 2 ms. At this time all atoms and molecules have converted into clusters and nanoparticles, as evidenced by the vortex ring structures of the plume; with a temperature of ∼1400 °C, just above the Ni/C and Co/C eutectic temperatures. By t = 4 ms, the plume has reached the furnace temperature, with the initial formation of carbon nanotubes. Continuing the process, the length of the nanostructures increases. In fact, if for example the process is stopped at ∼t = 25 ms, only short carbon nanotubes are obtained, indicating that the majority of the growth takes place after the ablation provided by the laser pulse. In any case, the laser ablation samples, observed with SEM and HRTEM, exhibit numerous bundles containing a high quantity of carbon nanotubes of excellent purity (carbon nanoparticles and residual catalysts are minimized in respect to the non-optimized arc discharge results). Both the carbon nanotubes and the bundle diameters depend strongly on the process parameters (temperature, pressure, gas flow) and in particular on the laser configuration (single or pulsed, intensity, etc.). For each method (Nd-Yd, CO2), owing to the experimental activities, it is possible to determine the set of parameters that optimize carbon nanotube synthesis (growth rate, quantity, purity, chirality, single wall or multiwall, metallic or semiconductor).

6.3.3

Chemical vapor deposition25–39

CVD is the third method employed in the synthesis of carbon nanotubes. The growth process involves heating a catalyst material to high temperature in a tube furnace and flowing a hydrocarbon gas through the tube reactor for a specific period of time. The key parameters in carbon nanotube CVD growth are: • hydrocarbons (gas flow and rate); • catalysts; • temperature (thermal cycles, described below).

© 2007, Woodhead Publishing Limited

136

Nanofibers and nanotechnology in textiles

The typical active catalyst species used are transition metal nanoparticles formed on the support (substrate) such as silica (Si) or other materials. Using different CVD typologies (e.g. thermal CVD (TCVD) or plasma enhanced CVD (PHCVD)) the substrate and the catalyst employed can be changed: Si substrate for TCVD, plastic materials for PHCVD. General nanotube growth in the CVD method is characterized by the dissociation of the hydrocarbon molecules, forming a region rich in carbon atoms reacting with the metallic particles of the substrate, with subsequent carbon nanotube formation (base and tip growth criteria). The CVD method is characterized by four fundamental aspects: 1. The possibility of performing a free-standing or well-aligned carbon nanotube growth (in the first case the nanostructures on the substrate surface are randomly distributed, while in the second case very good alignment is obtained. This aspect is relevant for the electronics application as nanodevices or MEMS/NEMS). 2. The carbon nanotubes produced have high defect densities (the nature of these defects remains to be thoroughly understood, but is most probably due to the relatively low growth temperature, which does not provide sufficient thermal energy to anneal the nanotubes into perfect crystalline structures. Growing perfect carbon nanotubes with the CVD process remains a challenge because the presence of the above defects strongly influences the carbon nanotube behavior in many applications such as composite nanostructure materials, electronics, thermal management). 3. It is possible to choose a specific pattern of carbon nanotube growth, creating specific distribution of the nanostructures on the substrate thanks to chemical inhibition and/or activation. (This provides the possibility of realizing nanoelectronics circuits or specific paths useful to the controlled thermal dissipation of advanced apparatus and devices). 4. The time scale is not fast (each synthesis requires at least one hour and the quantities of carbon nanotubes produced are much less than by the laser ablation and arc discharge processes, which are, in addition, faster than CVD. In this case the industrialization of the method involves different aspects regarding, principally, the alignment and density defects). Many CVD configurations have been developed: • • • •

TCVD: thermal CVD; HFCVD: hot filament CVD; PHCVD: plasma enhanced CVD; CCVD: catalytic CVD.

In particular, through using PHCVD it is possible to deposit nanostructures in a low-temperature synthesis condition. This is particularly important in the case of a substrate with a low melting point (plastic materials, for example).

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

137

Other methods (TCVD, HFCVD) are normally used with high melting point materials (e.g. Si(100) doped substrate). The substrate is usually made up of an Si(100) film thermally oxidized and coated with a thin layer (100–200 nm) of metal catalyst. A typical TCVD facility is shown in Fig. 6.14. The apparatus consists of a quartz tube inside a furnace, with the gas (inert and hydrocarbons) and the vacuum systems. Figure 6.15 illustrates the typical thermal cycles employed in TCVD carbon nanotube synthesis. In particular: • phase 1: heating the system using an inert gas (argon); • phase 2: chemical etching (by hydrogen) of the metallic catalyst deposited

6.14 Thermal CVD apparatus.64 T [°C] 3

T2 2

T1

2 bis

4 1

Tamb

t1 0

t2

t2b

t3

t4

t [sec]

6.15 Typical thermal CVD cycle used in carbon nanotube synthesis: 1, heating phase (inert condition); 2, chemical substrate etching for producing on the surface catalytic nanoparticles useful for the growth of carbon nanotubes; 2 bis, heating phase (inert condition); 3, carbon nanotube growth; 4, final cooling phase (inert condition).64

© 2007, Woodhead Publishing Limited

138

Nanofibers and nanotechnology in textiles

on the Si substrate (T1 = 500–700 °C, t2 = 10–20 min). The aim is to create, on the substrate surface, a deposit of high-density metallic nanoparticles (derived from the metal coating) that act as nucleation seeds for carbon nanotube growth (base and tip growth model); • phase 3: second heating of the system using inert gas (argon); • phase 4: carbon nanotube growth (T2 = 800–1200 °C, t3 = 10–60 min) using hydrocarbon gases (hydrogen, methane, etc.) • phase 5: cooling the system (t4 = 30–60 min) with inert gas (argon), to avoid oxidation of the nanostructures synthesized in the previous phase. Figure 6.16 shows the typical Si(100) doped substrate employed. With CVD it is possible to produce nanostructures and also to realize special nanocoatings, i.e. to deposit materials (nanostructures or simple nanoparticles) on different surfaces (carbon–carbon composites, metallic plates, fibres, multilayer composites, etc). Figure 6.17 shows the nacelle (inserted in the quartz tube) with a carbon–carbon, metallic plate, carbon fiber sample used to perform a TCVD nanocoating test. The results are reported in Figs 6.18–6.21 respectively: • SEM micrograph of the free-standing carbon nanotubes produced (Fig. 6.18). • Nanoparticles deposited on the metallic plates (this is a possible alternative

6.16 Silicon, with relative Miller index of 100, doped substrate coated with (ex. nickel) catalyst.64

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

Si

139

Carbon fibers

Metal C-C

6.17 Different materials used in TCVD.63

6.18 Bundles of free-standing carbon nanotubes produced by TCVD on the Si(100) doped substrate.63

method to plasma spray and traditional technologies used for the coating) (Fig. 6.19). • Nanoparticles on the carbon fibers (this can represent a fundamental step for chemical bonding between the fibers and the polymeric, or other typology, matrix and for the subsequent mechanical behavior of composite materials) (Fig 6.20). • Carbon–carbon coating (depositing ceramic materials is possible in order to manufacture a nanostructured thermal barrier useful for aerospace vehicles during the atmosphere re-entry phase of a mission) (Fig. 6.21).

© 2007, Woodhead Publishing Limited

140

Nanofibers and nanotechnology in textiles

6.19 Carbon coating deposited on the metal plate surface using the TCVD method.63

6.20 Carbon coating deposited on the carbon fibers using the TCVD method.63

CVD provides the possibility of developing innovative and advanced technological methods useful to improve traditional techniques employed in many scientific sectors.

6.4

Characterization techniques1, 50–64, 68–73

In each step in the development of an advanced nanotechnology system/ apparatus the characterization of the nanomaterials and nanostructures

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

141

6.21 Carbon coating deposited on the carbon–carbon composite using the TCVD method. Using ceramic materials, with the same technology, it is possible to produce a coating with high mechanical, thermal and chemical characteristics useful for hybrid composite materials employed in space vehicles (for the re-entry phase mission in which the thermal and oxidation conditions are extreme).63

employed (synthesized, purified, etc.) is a critical phase. In fact, it is necessary to perform a specific evaluation of the fundamental properties and characteristics of the ‘nano-elements’ integrated in the above advanced system. To observe and to analyse structures of micro- and nanometric dimensions is not simple but it is necessary to develop a specific standard protocol for every single step, i.e.: • • • • •

sample preparation methods; analysis criteria; evaluation and interpretation of results; analysis reliability and repeatability; storage procedure for samples.

Several analytical characterization tools have been used successfully in past to determine the principal properties of nanostructures and nanomaterials, but a lack of standard methodologies makes it difficult to compare these measurements. The development of a protocol in which standardized analysis methods and procedures are defined is needed. Typical characterization analyses used in the nanotechnology science are: • optical microscopy; • optical laser microscopy; • SEM (scanning electron microscopy);

© 2007, Woodhead Publishing Limited

142

• • • • •

Nanofibers and nanotechnology in textiles

TEM (transmission electron microscopy); EDX (energy dispersion X-ray); AFM (atomic force microscopy); STM (scanning tunneling microscopy); Raman spectroscopy.

It is clear that the numerous nanomaterials/structures, available due to the different nanotechnologies developed and modified by different processes (synthesis, purification, integration, etc.), demand close examination when they are used for each application. It is fundamental to have a well-characterized material in order to access the variability of the numerous steps required in the design and development of the previously mentioned applications. For example, the nanocomposite material’s preparation and evaluation require measurements that can follow the matrix before and after the addition of nanostructures (e.g. carbon nanotubes). There is a strong need for standard methods to characterize the nanomaterials in order to improve the capability of comparing different samples employed as raw materials. The requested protocol must be characterized by a standard procedure useful for performing a short ‘nano-elements’ characterization, with high reliability levels. This section illustrates the techniques available and the typical uses: • Optical microscopy: used to characterize the micromorphologies of the sample surface (e.g. electrode cathode surface after the arc discharge) or micro-powders. • Optical laser microscopy: used for the same purpose as optical microscopy, but with better resolution and the ability to perform 3D surface morphology digital reconstructions. • SEM: used to determine the nature of the nanomaterials obtained (e.g. to define whether the synthesis products are carbon nanotubes or simple amorphous graphite) and to acquire an idea about the material’s quality and morphology. • TEM: provides specific characterization. In high-resolution modality, for example, it is possible to determine if the carbon nanotubes analysed are single or multi wall. • EDX: gives a spectra from which the chemical elements present in the studied nanomaterials can be separated and identified. • AFM: allows 3D nanotopography and morphology profiling of the microand nanomaterial/structures. In addition, with the cantilever tip of this instrument, it is possible to determine the principal mechanical (Young’s modulus) and electrical (V–I characteristic) nanostructure properties. • STM: provides 3D real images with subatomic spatial resolution of electrically conductive samples. • Raman spectroscopy: for the specific applications in carbon nanotube science, provides an express, non-destructive and preparation-free estimation of the carbon content in a sample.

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

143

The techniques listed enable the following characterization of nanomaterials: • Morphology: evaluation of nanometric geometry and characteristics of the observed nanostructures. For example, to determine the typology (single or multiwall), the chiral angle, the twisted angle, etc. of carbon nanotubes. • Homogeneity: to determine the statistical distribution of the various nanomaterials/structures present in a sample or, for example, dispersed into a matrix or in any device. • Dispersability: to determine the capability of nanostructures to form a stable suspension at specific concentration values in bundles or single elements; or to evaluate the dispersion level of the same materials, into a matrix or in a micro-/macro-system. In particular, three different levels have been defined: macro-, micro- and nano-dispersion. • Reactivity: for each nanomaterial and nanostructure determining the chemical, physical and thermal reactivity is necessary. These are important parameters in each step (synthesis, purification, integration and operative conditions of the ‘nano-elements’). • Purity: in each phase of their development (synthesis, purification, integration, etc.) the nanomaterials are constituted by nanostructures and amorphous residuals. It is always necessary to evaluate the exact percentage of the each element’s typology contained in a sample (powders, massive elements, apparatus, etc.). Clearly, the target is the greatest reduction of impurities. Each analysis typically requires a specific sample preparation procedure. In the case of massive samples, if requested, it is only necessary to deposit a carbon (or gold) coating on the observed surface to improve the electron microscopy contrast and resolution. In contrast, for the micro- and nanopowders, the preparation procedure is more complex. In fact, generally, the following steps are needed: • The mechanical removal of powder from the sample surface (for example from the cathode electrodes after the arc discharge synthesis). • To sonicate the powders in order to separate the single nanostructures, avoiding formation of agglomerates. • To deposit powder in a holder (characteristic for each microscopy typology: aluminum holders with carbon tape for the SEM, metallic grid with a specific mesh size in the TEM analysis, etc.). • To perform, when requested, the above-mentioned carbon or gold coating of the sample surface (this is specific for SEM, TEM and EDX). Some analysis typologies require vacuum conditions (SEM, TEM, EDX), or the electrical conductivity of the sample. Other techniques are more simple and rapid (no sample preparation, no vacuum) and the analysis quality is not reduced, as in AFM.

© 2007, Woodhead Publishing Limited

144

Nanofibers and nanotechnology in textiles

For each technique, the calibration phase and the comparison of preliminary results with standard measurements are fundamental. They ensure that subsequent analyses can be performed with a high reliability level. The evaluation of results is very difficult in nanotechnological microscopic analysis. In fact, the reduced size of samples, together with the complexity of the instruments and physical and chemical phenomena occurring in nanoscale dimensions (e.g. relativistic effects), introduces numerous stochastic variables, making measurement and relative analysis very complex, with complicated evaluation and interpretation of results.

6.4.1

Optical microscopy

With optical microscopy it is possible to perform an accurate macro-morphology analysis of various sample typologies. Obviously, it is not possible to observe nanostructures, but, for example, one can see the surface of the cathode electrodes before and after the arc discharge, identifying the characteristic four regions (A, B, C and D) previously described. In addition, it is possible to characterize the surface (with and without electrical conductivity) and determine the morphologies of the fracture section of composite samples after mechanical (static and dynamic), thermal and corrosion tests. Figures 6.7 and 6.11a show optical images of the cathode electrode surfaces after arc discharge corresponding to different synthesis parameters. As shown, the four regions (A–D) are immediately observable, leading to preliminary studies of process results. The target is to realize an SEM analysis and associate to each region a specific characterization of the nanostructures contained on it. Using the same synthesis parameters it is possible to determine, for comparison, on numerous electrodes, the nanodeposit characterization using only optical microscopy – a cheap and easy analysis method. Figure 6.22 illustrates the fracture section of polymeric nanostructured composite materials after dynamic testing. Much information can be gained from studying these images.

6.4.2

Optical laser microscopy

Optical laser microscopy represents an innovative analysis typology. In fact, in the co-focal no contact mode (using laser light with wavelength (λ) in the order of nanometers – O (nm)), it is possible to build a micrometric 3D topography of the surface. As shown in Fig. 6.23, the cathode electrode surface (treated by arc discharge) is analyzed, permitting the detection of the four characteristic zones with a very high detailed reconstruction of their morphologies. The image acquisition is extremely rapid, and no vacuum and electrical conductivities are required.

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

145

6.22 Fracture section of nanostructured polymeric composite sample used to perform a mechanical test.53

Data name : Sample_135.ols Comment :

460.000 230.000 0.000 um0.00 384.000 768.000 1152.000 1536.000 1920.000 um 2560.00 Z X

Y

0.00

1280.00 Ob. Zoom Acq. Info.

: : : :

5x 1.0x XYZ–M–C CF–H

6.23 Optical laser micrographs showing the 3D morphologies of the four characteristic regions present on the cathode surface after the arc discharge process.60

© 2007, Woodhead Publishing Limited

146

6.4.3

Nanofibers and nanotechnology in textiles

Scanning Electron Microscopy (SEM)

SEM micrographs are used to determine the nature of the deposit produced by the synthesis and to acquire a rough idea of the material quality (e.g. purity level). As illustrated, for example, in Figs 6.1 and 6.8, the information acquired by SEM is preliminary; in fact, it is not possible to say whether the observed filaments are bundles of carbon nanotubes or micro-carbon fibers. In addition, in the case of carbon nanotubes data relative to the presence of single walls or multiwalls are not provided. The same applies for other parameters such as chirality, twisted angle, etc. The micro-mechanics studies of the nanostructured composite are realizable. Figure 6.24(a) shows the micro-fracture line’s deviation and interruption due to cavities inside the tested sample. Figure 6.24(b) illustrates the crosssection of a polymeric thin film with carbon nanotubes. By integrating specific software in the SEM it is possible to perform a statistical analysis of nanostructure size. Figure 6.25 shows an SEM picture representation of the carbon particles with different dimensions. From this micrograph the software calculates: area, diameter (max, min and mean) and perimeter of each particle. This is a very important characterization, because in many nanotechnology applications the size of the ‘nano-elements’ influences strongly the behaviour of apparatus and devices.

6.4.4

Transmission Electron Microscopy (TEM)

Scanning or high-resolution TEM configurations allow the real characteristics of the observed nano-elements to be determined. As mentioned above with SEM, the characterization is incomplete. Only with nanoscale observations is it possible to determine, for example, if a carbon nanotube is single or multiwall. Figure 6.26 gives example of multiwall carbon nanotubes, in which the morphologies and, in particular, their bamboo-like cap end are clearly visible. Typically nanotube extremities contain the catalyst particles useful to the growth process.

6.4.5

Energy Dispersion X-ray (EDX)

Many catalysts are employed to improve the kinetics of nanostructure growth, and quantities and quality (purity level) of synthesized materials, such as cobalt, nickel and yttrium. It is necessary to determine the presence of these elements in order to establish, for each synthesis result, the influence of a specific typology and quantity of catalyst, or a particular mixture. Moreover, an evaluation of the after-purification residual elements (catalysts, amorphous elements) is fundamental for understanding phenomena and subsequent behavior of devices.

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

147

(a)

(b)

6.24 (a) SEM analysis of the fracture line on the transverse section of a nanostructured composite material after dynamic testing.50 (b) Cross-section SEM micrograph of polymeric nanostructured thin film.50

These data are obtained by EXD, as illustrated by Fig. 6.27, in which it is possible to evaluate the catalysts used and the relative percentage (in this case for nano-powders containing carbon nanotubes and produced by laser

© 2007, Woodhead Publishing Limited

148

Nanofibers and nanotechnology in textiles

6.25 SEM showing statistical analysis of micro-carbon particles. The software is integrated in the SEM, and through the use of different colors, determines the dimensional ranges of the particles.64

6.26 High-resolution TEM micrograph of a multiwall carbon nanotube synthesized by TCVD.64

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

149

Au Ni

Ni C

Au 1.00

2.00

3.00

4.00 5.00 6.00 7.00 Energy of X-ray [keV]

8.00

9.00

10.00

6.27 A typical EDX spectrum used to determine the chemical composition of the nanomaterials produced and purified using different methodologies.64

ablation). EDX systems are integrated both in SEM and TEM microscopes, but as described in many protocols reported in the literature, the EDX realized by SEM is more accurate and reliable than that performed by TEM, in which the minor dimension of the analyzed zone can provide unreliable results (with reference to the effective chemical composition of samples, principally due to the interaction between the holders and the measured region).

6.4.6

Atomic Force Microscopy (AFM)

With atomic force microscopy it is possible to realize a 3D nanoscale topography of nanostructures, and define ‘nanowrinkledness’ profiles. The principal parameters (e.g. Young’s modulus) and properties (e.g. V–I characteristic) of carbon nanotubes have been determined by the tip of the AFM cantilever. Figure 6.28 illustrates an example of carbon nanotubes deposited on a Si substrate and analysed by AFM. The lines marked as red and green show nanometric sample profiles. The relative statistical characterization is reported in the table.

6.4.7

Raman spectroscopy

Raman spectroscopy is used to corroborate SEM and TEM examinations. All carbon allotropic species (fullerene, graphite, carbon nanotubes and diamonds) are Raman active. All carbon forms contribute to the Raman spectra in the range of 1000–1700 cm–1 with two characteristic peaks at 1300 cm–1 (D-band) and 1600 cm–1 (G-band). The position, width and relative

© 2007, Woodhead Publishing Limited

150

nm Line profile: Red 0.102

0.4

µm

30

0.8 µm

1.2

1.6

Line profile: Green

20

0.041 0

0 0

0.4

0.8 µm

1.2

30 nm

40

1.6

6.869

0.4 10

4.238

0.585nm

0.095 +8 +6 +4 +2 +0 0.4

0.8 µm

1.2

1.6

400 pxl/div

Histogram

0

10

20

50

Statistics Line Min(nm)Max(nm) Mid(nm) Mean(nm) Rpv(nm) Rq(nm) Ra(nm) Rz(nm) Rsk(pm) Rku(pm) Red 1.900 14.322 8.111 5.722 12.422 2.623 1.901 4.923 –192.199 657.531 Green 0.585 10.522 5.553 4.058 9.938 1.978 1.490 3.225 –160.973 605.737

6.28 AFM analysis of carbon nanotubes. The graphics and table report the nanotopography analysis useful to characterize nanostructures. In addition, AFM microscopy is used to perform mechanical and electrical characterization of carbon nanotubes.64

© 2007, Woodhead Publishing Limited

Nanofibers and nanotechnology in textiles

0.054 0

0.8

8.33

1.2

+10 +7.5 +5 +2.5 +0

6.138

40

1.9nm

1.6

50

Carbon nanotubes in aerospace engineering

151

intensity of peaks vary with carbon allotropes and thus can be used to evaluate the typologies of the ‘nano-elements’ analyzed. In particular: • G-Band (∼1590 cm–1) is assigned to the tangential radial mode of the graphite. The Breit–Wigner–Fano (BWF) line shape in this band indicates the metallic catalysts in the carbon nanotubes; • D-Band (∼1370 cm–1) is associated with the Raman mode of the amorphous carbon – the G/D ratio provides the purity level of the analyzed carbon nanotubes; • RBM (radial breathing mode at ∼130–300 cm–1) represents, in the lowfrequency region of the Raman spectra, the radial breathing mode of carbon nanotubes. The peak position of the RBM (ω cm–1) is related to the single-wall carbon nanotube diameter (d) by the following relation: d = 248 ω

[6.9]

in the range of 0.8–1.5 nm. Raman spectroscopy is very useful for a correct carbon nanotube characterization, when SEM and TEM analyses do not provide univocal results. Figure 6.29 shows an example of the carbon nanotube Raman analysis.

Intensity [a.u.]

G-band

BWF

D-band RBM

200

400

600 800 1000 1200 1400 1600 Wave number [cm–1]

6.29 Raman spectroscopy of nano-powders containing carbon nanotubes. For each band it is possible to associate a specific nanomaterial property.26

© 2007, Woodhead Publishing Limited

152

6.5

Nanofibers and nanotechnology in textiles

Purification techniques40–49

After a synthesis process, the produced nanostructures contain some residuals. In particular: • graphite = base materials not transformed in nanostructures (e.g. carbon nanotubes); • metallic micro- and nanoparticles = employed as catalysts to improve the chemical kinetics of the synthesis process and then increasing the quantities of the produced nanostructures; • fullerene = besides carbon nanotubes, the produced deposit can contain other nanostructure typologies (e.g. nanowhiskers, ‘buckyballs’, buckyonions, etc). Using different synthesis methods, the quantities of the unwanted residuals can vary. Using laser ablation and CVD facilities the purification level of produced nanomaterials is typically high. Arc discharge, if it is not optimized, may be characterized by a significant presence of residuals, but, as mentioned with specific control of the process parameters this method can provide high quantities of nanomaterials with a significant purification level. The availability of carbon nanotubes, or in general of nanostructures, of high purification level enables the optimization of the exceptional theoretical properties of these innovative ‘nano-elements’. In fact, when developing a nanotechnology system, the effective properties of nanostructures will be inferior to the theoretical predicted values. This depends, principally, on the insufficient purification level of nanostructures, giving a ‘nanosystem’ that embraces less effective properties than those predicted theoretically. It is important to underline the economic aspect. The high cost of a poorly optimized nanotechnology system (or apparatus) is not justified in respect to a traditional apparatus that provides the requested performance with minor costs. Realistic large-scale use of nanotechnologies will depend, strongly, on a real economic cost reduction with respect to the actual competitive economic level characteristic of nano prototypes and demonstrators. With devoted purification methodologies it is possible to obtain nanostructures with high purification levels (>80% with respect to the post-synthesis materials employed). The principal purification techniques are: • • • •

sonication; chemical etching; selective oxidation; electrophoresis.

Each method is characteristic of a different synthesis process.

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

6.5.1

153

Sonication

Sonication is employed in the sample preparation procedures for microscopy characterization (SEM, TEX, EDX, etc.). It is useful to separate nanostructures from residuals using the following procedures: • mechanical removal of the nano-powders from the synthesis deposit (e.g. in the case of the arc discharge from the cathode electrode surface); • dispersion of the removed materials in ethylic alcohol; • sonication of the liquid solution; • drying of the above solution and SEM analysis of the just-treated nanomaterials. By this method it is possible to: • separate nano-elements from the residual; • open the edges of carbon nanotubes, removing the caps (this is a typical functionalization, useful for preparing nanotubes for their integration in composite materials or in electronic devices. In fact, the external caps are substituted by other chemical functional groups devoted to the chemical and physical interactions between the nanostructures and the system in which they are dispersed); • cut the nanostructures, etc. The principal parameters of this method are time and ultrasound frequency. During the SEM and TEM sample preparations, the materials are sonicated for 3–5 min. In the case of purification (removing of the residual and/or of the cap ends), the time increases. This is a critical aspect. In fact, as shown by Fig. 6.30, after an ultrasound treatment of over 60 min, the carbon nanotube morphologies are very degraded (Fig. 6.30b), with respect to the pre-treatment conditions (Fig. 6.30a). Decreasing the time (Fig. 6.31) reduces post-processing carbon nanotube morphology degradation.

6.5.2

Chemical etching

With chemical etching, using a specific acid solution, it is possible to reduce residuals as a result of the different chemical reactivity of the nanostructures, amorphous particles and catalysts. An example is oxidation with potassium permanganate (KMnO4) in an acidic solution (sulfuric acid H2SO4). The chemical reaction useful to remove the undesired residual containing carbon nanotubes proceeds as follows: 3C + 4KMnO4 + 4H+ → 4MnO2 + 3CO2 + 4K+ + 2H2O After reaction, the mixture is washed in water and dried at ∼150 °C for 12 h. Figure 6.32a shows carbon nanotube morphologies before (a) and after (b) chemical treatment.

© 2007, Woodhead Publishing Limited

154

Nanofibers and nanotechnology in textiles (a)

(b)

6.30 (a) SEM micrograph of carbon nanotube deposition before the ultrasound purification treatment.54 (b) SEM micrograph of the same deposit after the treatment. It is not possible to observe good results due to an evident degradation of the carbon nanotube morphology.54

Naturally, it is possible to use a different chemical reaction process depending on the specific nanomaterials employed. The chemical reaction with potassium permanganate, for example, is employed for carbon nanotubes produced by laser ablation rather than for those obtained by arc discharge, owing to the different chemical reactivity caused by the different morphologies.

6.5.3

Selective oxidation

Selective oxidation represents the third methodology exploitable to purify nanomaterials and in particular carbon nanotubes. It consists of a selective chemical reaction among carbon nanotubes and residual elements (graphite,

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

155

6.31 Reducing the ultrasound treatment time makes it possible to obtain an improvement in the carbon nanotube morphology, as shown by SEM.54

amorphous structures, catalysts, etc.) in the presence of an oxidating hot environment. In theory, nanostructures (carbon nanotubes or, more generally, fullerenes) possess a chemical activity less than that of other elements (e.g. amorphous carbon). Using DTA (Differential Thermal Analysis) and TGA (Thermogrametric Analysis), under particular operative conditions, it is possible to obtain a deposit containing carbon nanotubes of high purification. Figure 6.33 shows a typical DTA–TGA profile employed to purify carbon nanotubes. The line marked ‘Temp’ indicates the temperature profile characterized by two steps: • heating phase; • static thermal operative condition (the optimized temperature value is obtained from repeated testing, useful for determining the condition at which the highest purification level is obtained). Both phases are carried out in an oxidative environment. Usually, the gases employed are nitrogen (N2) and oxygen (O2) with percentages defined by the oxidation ‘intensity’ required (e.g. 90% N2 and 10% O2). The aim is to obtain powders containing the largest quantity of carbon nanotubes, thus minimizing residuals. The ‘TG’ line in Fig. 6.33 indicates the mass (m) loss percentage (∆m/∆t). This curve indicates that, during the process, the materials are oxidized. It is necessary to evaluate whether these consist only of residuals or also of

© 2007, Woodhead Publishing Limited

156

Nanofibers and nanotechnology in textiles (a)

(b)

6.32 (a) SEM micrograph of carbon nanotubes before the purification by oxidation with potassium permanganate (KMnO4) in acid solution.56 (b) Carbon nanotube morphology after the oxidation treatment.56

carbon nanotubes. The second hypothesis is more realistic. In fact, by this purification methodology, the real target is to find two simultaneous conditions: the maximum residuals oxidation and minimum carbon nanotube ‘destruction’. It is fundamental to observe that all characterizations (SEM, TEM, etc.) and purification methods can modify, in some cases strongly, the morphology and the properties of nanostructures. The ‘DTA’ curre in Fig. 6.33 represents the DTA analysis, which indicates the carbon nanotubes mass reduction with respect to an inert mass (e.g. alumina Al2O3) used as reference. Figure 6.34 shows a deposit, containing carbon nanotubes, before (a) and after (b) selective oxidation experiments. The reduction of residuals with a deposit prevalently consisting of bundles of carbon nanotubes is shown.

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

157

NETZSCH Simultaneous thermal analyzer STA 409 10

800

10

700

0

600

–10

500

–20

–10

DTA 400

–20

300

∆m

Temp [°C]

Mass [%]

0

–30 –40

200

–50

100

–60

∆t

–30

TG –40 0

20

Identity No. Date Operator Laboratory

40

M77 H554001 04 Aug 2003 Severi Terminal Lsi

60

80 100 Time (min)

Sample Reference Atmosphere Crucible

120

CN 1 Al203 N2 90 02 1/I/h AL203

140

0 160

Temperature difference [µV]

Temp

–70

250 mg : — 50.70 mg TG 50 µV : — 51.50 mg DTA 12 ⇒ segment: 1…2

6.33 Typical DTA–TG profile used in the purification of carbon nanotubes using the selective oxidation method.52

6.5.4

Electrophoresis

Good carbon nanotube alignment and purification levels are obtainable with electrophoresis. The electrophoresis (DC or AC) technique consists of dispersing nanomaterials (carbon nanotubes and residuals) in isopropyl alcohol dropped onto a coplanar metallic electrode (e.g. Al) with a gap of glass materials. An AC electric field is applied with a specific frequency at room temperature The results consist of well-aligned purified carbon nanotubes. However, carbon particles, contained as impurities, become harder to move (on the electrode surfaces) with increasing frequency, and the degree of nanotube orientation is higher when the frequency is higher and nanotubes are longer. Also in this case parameter optimization is required to obtain carbon nanotubes that are usable in advanced nanotechnology systems.

6.6

The use of carbon nanotubes in aerospace engineering50–66, 92, 93

The properties of nanomaterials (mechanical, electrical, thermal, etc.), and in particular those of carbon nanotubes, are considered as key factors for future improvement of technical characteristics of many engineering macro-

© 2007, Woodhead Publishing Limited

158

Nanofibers and nanotechnology in textiles (a)

(b)

6.34 (a) SEM micrograph of carbon nanotubes before selective oxidation in nitrogen and oxygen.53 (b) SEM micrographs of the same deposit after the selective oxidation. A significant purification level of the nanomaterials has been obtained.53

and nanosystems. The synthesis, purification and characterization of nanomaterials are primary requirements for their realistic use in many engineering sectors. Characterization and functionalization are required steps to prepare nanomaterials for the next phase of the composite manufacturing process or other nanotechnology systems and devices. Control of the synthesis parameters (voltage, current, laser power, raw materials typology, catalysts used, gas pressure, etc.) is the major hurdlle to obtaining the following requirements: • producing nanomaterials in high quantity; • obtaining nanomaterials of high purity and degree of alignment; • controlling the typology of nanomaterials produced and the relative morphology. The purified materials (e.g. carbon nanotubes with high or lower degree of

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

159

alignment) must be functionalized. This means preparing (for example, with a specific chemical treatment) carbon nanotubes for integration in other systems (a matrix for the composite applications, into a MEMS/NEMS, into a field emission system, sensors, etc). This is a critical phase: in fact, with non-optimized functionalization it is not possible to utilize the properties of carbon nanotubes. To date, the integration of carbon nanotubes in a polymeric matrix with poor chemical functionalization gives poor transmission loads between the matrix and nanomaterials. In addition, the integration of nanomaterials in the matrix (polymeric, metallic and ceramic) is an important step in the realization of innovative nanostructured materials (carbon nanotubes reinforced). Possible fields of application of these innovative advanced materials are represented by the following aerospace applications: • • • • • • • •

nanocomposite materials for structural applications; special coatings; frequency selective surfaces (FSS); thermal barrier for multilayer composite materials (unmanned space vehicles); thermal management; health monitoring; MEMS/NEMS; nanosensors and nanodevices.

This new generation of materials has real and interesting applications in aerospace technologies, and the typical research activities are: • theoretical study of the nanostructured material (e.g. carbon nanotubes reinforced); • study of nanomaterial synthesis, characterization (optical, SEM, TEM, Xray and chemical) and its relative functionalization; • manufacturing process definition; • sample manufacturing; • the characterization of materials for specific application (nanocomposite polymeric materials for structural applications, special coatings, FSS, thermal barrier for multilayer composite materials, thermal management). The properties of these new advanced materials will be calculated with specific tests (considering the probability of the presence of uncertainty within results, a large number of tests will be performed: minimum 10 specimens for each test). Nanocomposite materials with revolutionary new capabilities constitute an essential element in the design of advanced systems. In particular, carbon nanotube composite materials were identified as those with the highest expectation in terms of performance benefit for many applications and in

© 2007, Woodhead Publishing Limited

160

Nanofibers and nanotechnology in textiles

particular in the aerospace sector. The general field of nanoscience and technology offers potential as the next great technological revolution. In the field of materials science, we may see a paradigm shift away from the traditional materials role of developing metals, polymers, ceramics and composites to a revolutionary role of developing nanostructured, functionalized, self-assembling materials. Looking to the future, the theoretical potential of these revolutionary new materials will enable technological developments that are barely imaginable today. Material systems based on carbon nanotubes are a particularly attractive new class of materials. On the basis of computer simulations and limited experimental data, some specific forms of carbon nanotubes appear to possess extraordinary mechanical, thermal and electrical properties. If the properties of carbon nanotubes observed at the molecular level can be translated into useful macro-scale materials, the potential benefits to the aerospace industry include applications to vehicle structures, propulsion systems, thermal management, energy storage, electronic and computing, sensors and devices, and biological and medical applications. The computer simulation results and limited experimental studies have shown that small diameter, single-wall carbon nanotubes may possess elastic modulus in excess of 1 TPa, and strengths approaching 200 GPa. For example, if small diameter, single-wall tubes can be produced in large quantities, and embedded into a supporting polymeric matrix to form structural materials, the resulting structures could be considerably lighter and stronger than current aluminum alloys and carbon fiber reinforced polymer composites used in conventional aerospace structures. The mechanical properties of carbon nanotubes give opportunities to develop new advanced materials. It is possible to use different kinds of matrix such as: • metallic; • ceramic; • polymeric. The composite manufacture requires the following steps: • selection of materials (matrix, curing agent, nanometric particles); • studying the sample production methodologies; • obtaining a uniform distribution of the nanometric particles (with carbon nanotubes of opportune percentage) in the matrix; • achieving mechanical testing (dynamic and static); • performing morphological testing; • studying the mechanical fracture of the composite; • improving the characteristics of the composite; • applying the advanced composites in the aerospace systems;

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

161

• performing theoretical and numerical analysis of the composite behavior. With polymeric, metallic and ceramic matrices, it is possible to develop new applications in aerospace engineering: • • • • •

lighter structures; thermal barriers; thermal cooling systems; pins; special nanostructured coatings, etc.

Sample manufacture is necessary to evaluate the properties of these new materials. In particular: • studying the theoretical and numerical models; • synthesizing carbon nanotubes with high quality (enough to produce a sufficient number of samples), high purification and degree of alignment with a perfect control of the typology of nanomaterials synthesized. Price is important in project development; • performing the characterization and functionalization of carbon nanotubes for their successful integration into a nanosystem/device; • determining the sample’s manufacture procedures and the correlated methodologies of analysis; • acquiring all the capabilities essential to develop prototypes. Considering the possible aerospace applications of nanotechnology, the International Space Agencies (e.g. NASA, ESA) have identified the following long-term goals: • providing safe and affordable orbital transfer and interplanetary transportation capabilities to enable scientific research; • human and robotic exploration; • the commercial development of space; • cost reduction and high reliability. Numerous scientific and engineering breakthroughs will be required to develop the technology needed to achieve these targets. Critical technologies include advanced vehicle primary and secondary structure, radiation protection, propulsion and power systems, fuel storage, electronics and devices, sensors and science instruments, and medical diagnostics and treatment. Advanced materials with revolutionary new capabilities are an essential element of each of these technologies. Based on a survey of emerging materials with applications to aerospace systems (e.g. vehicle structures and propulsion systems), nanostructured materials have been identified as those with the highest expectation in terms of performance benefit for many applications. The principal aerospace applications of the nanotechnologies are described below.

© 2007, Woodhead Publishing Limited

162

6.7

Nanofibers and nanotechnology in textiles

Nanostructured composite materials for aerospace applications50–64, 92–127

In aerospace applications, the use of nanostructured materials is finalized to a significant improvement of the mechanical properties, Young’s modulus, ultimate strength, Poisson coefficient, etc. and reliability. In other sectors (e.g. electronics) the principal requirements are concentrated on the morphology characteristics of employed nanomaterials. In fact, the required quantities are always limited, and concurrent with their availability. Instead, for composites and structures, considering the large size of aerospace systems, the quantity and the relative costs represent a critical step. For realistic use of nanostructured composite materials it is fundamental to reduce the costs of developing industrial methodologies (synthesis, purification, integration, etc.) to provide the required quantities while respecting the imposed requirements relative to morphology, typology and purification level. The manufacturing of nanostructured composite samples requires the following specific steps: • definition of the matrix typology: polymeric (thermoplastic or thermosetting), metallic, ceramic; • selection of the specific curing agent, curing process and conditions (temperature, pressure, vacuum, etc.); • determination of the typology necessary to obtain the properties required; • definition of the manufacturing and test procedures – for traditional materials numerous standard procedures are available, but in the case of nanotechnologies, often, the development of specific non-standard procedures is necessary; • Analysis of results using: mathematical/theoretical and numerical models, electronic microscopy, fracture mechanics and non-destructive testing. Comparing theoretical and experimental results allows us to understand the realistic behavior of these innovative materials. In particular, considering the cost of these materials, traditional composite materials are more competitive in terms of economic value and properties. This is a critical aspect with reference to the large dimensions of aerospace components and structures. During sample manufacturing processes, particular attention is required in the following steps: • Matrix de-gassing period, this is needed to reduce the micro-void embedded in the composite. This represents an important condition not only in the composite sector, but in nanotechnology application areas. • Mixing procedure. During this phase it is essential to obtain uniform distribution of the nanomaterials in the matrix, without void and avoiding random chemical concentrations.

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

163

• Use of the correct curing agent quantities. In fact, including nanoparticles in the matrix means the ratio of the matrix/curing agent concentrations can vary. • Sample characterization and storage. In this case it is necessary to adapt specific standard procedures of traditional composite materials, to nanostructured composites. Considering the mechanical aspect, the load transfer properties are the main problem in ensuring high performance of the nanocomposite. This depends strongly on nanostructured functionalization and the interfacial stress between the matrix and embedded nanoparticles. There are three typical load transfer mechanisms: • micro-mechanical interlock; • chemical bonding; • van der Waals interaction forces. The first mechanism is not influenced by the nanocomposite, since the nanostructured surface typically appears atomically smooth. Some studies demonstrate that the interfacial chemical bonding, between the matrix and nano-elements, could be very high, and the sliding friction between carbon nanotubes and matrix is much greater than that among graphite sheets. With HRTEM analysis of carbon nanotube composite materials, after mechanical testing no fracture has been observed on the carbon nanotube surfaces. Another fundamental result is that if the carbon nanotubes can restore their original undeformed shape when the matrix has been heated, the compressive stress due to the shrinkage (produced by the curing) could be released. It has been observed that using a small percentage of carbon nanotubes embedded in the matrix results in a significant improvement in mechanical properties. This is very interesting carbon nanotube behavior, in respect to problems of bridging, pull out and delaminating (in this case on a nanoscale dimension). In addition, by a random carbon nanotube distribution into a polymeric matrix it is possible to obtain specific electrical properties in order to avoid electrostatic charge, provide sufficient matrix conductivity, etc. An example of the manufacturing and characterization of a nanostructured composite material is now reported. A complex step is the definition of a procedure aimed at the realization of a homogeneous dispersion of a nanometric powder in epoxy resin. Moreover, adhesion problems related to the interfacial activity of the resin, the powder and the nanotubes must be solved. The materials employed to manufacture the samples were: • epoxy resin; • commercial curing agent; • nanometric graphite powder (granulometria: 20 µm) with a carbon nanotube addition.

© 2007, Woodhead Publishing Limited

164

Nanofibers and nanotechnology in textiles

The total concentration of the dispersed powder was 10% and 20% in wt. The sample was 10 mm × 10 mm × 120 mm3 in dimension (Fig. 6.35). The curing process adopted was: • room temperature for 24 h; • furnace curing 80 °C for 3 h. Impact tests were performed, during which the following considerations were relevant: • the reduction of powder granulometria increased the impact resistance properties; • good surface finishing improved the mechanical properties. Figure 6.22 shows the fracture surface appearance of a sample containing 20% of powder. The pre-crack length was 2 mm and the brittle behavior of crack propagation was seen. To understand the fracture-mechanic behavior of the composite, SEM characterization of the fracture surface is necessary. Figure 6.36 shows an SEM image of the sample containing 10% nanocarbon powders. In areas A (crack initiation) and B (propagation) there is no presence of preferential directions for crack propagation. In contrast, in the sample containing 20% powder, preferential directions of crack propagation are observed (Fig. 6.37 areas A, B, C and D). The presence of preferential direction is due to the non-uniformity of powder dispersion in the matrix. A further observation made was that the fracture lines change direction corresponding to cavities (or voids). In Fig. 6.38 two fracture lines (A and B)

6.35 Nanostructured polymeric composite specimen for mechanical tests.58

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

165

are diverted by the presence of a void (see points C and D), and are stopped at point E. The static test shows that using nanometric particles (graphite and carbon nanotubes) the Young’s modulus is 12% more than the sample with only resin and curing agent. In theory using only carbon nanotubes (with theoretical Young’s modulus: 1 TPa) the mechanical properties of the composite become very interesting. This represents only a preliminary example of nanostructured composite materials. In addition, using the same procedure, it is possible to produce a thin film (Fig. 6.39) that has important applications in electronics, providing devices with specific electromagnetic properties, but also with high mechanical performance. Typically, with traditional technologies, it is very difficult to integrate, in the same material or element, mechanical and electrical properties (or other specific requirements). Nanostructured composite materials are employed to design and manufacture aerospace structures with high performance, reliability and lightness. An important example is that of the multigrid lattice structures (Fig. 6.40) used for launchers, fuselages and structural elements. Traditional composite technologies can be used to produce such structures, with different shapes and dimensions. The design is carried out using the Vasiliev model, which allows the dimensions of the resistance elements to be determined with respect to the following three conditions: • minimum mass;

6.36 SEM analysis of the fracture surface of nanostructured composite specimen with 10 wt% carbon nano-powder addition.55

© 2007, Woodhead Publishing Limited

166

Nanofibers and nanotechnology in textiles

6.37 SEM analysis of fracture surface of nanostructured composite specimen with 20 wt% carbon nano-powder addition, indicating the preferential direction of the fracture line propagation.55

E D

A C B

6.38 SEM analysis of fracture surface of nanostructured composite specimen with 20 wt% carbon nano-powders addition, showing the fracture line behavior corresponding to the micro-voids embedded in the matrix.55

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

167

6.39 Nanostructured thin film (polymeric matrix reinforced with carbon nanotubes).64

6.40 A 3D CAD (computer-aided design) drawing of a conic anisogrid lattice structure for aerospace application.59

© 2007, Woodhead Publishing Limited

168

Nanofibers and nanotechnology in textiles

• static resistance (the maximum applied load is equal to the ultimate strength of the material employed); • stability (unitary eigenvalue). By introducing nanotechnologies, and, in particular, the polymeric composite reinforced with carbon nanotubes, significant improvements in mechanical properties can be obtained. Using the Vasiliev model it is possible to observe that, using a low percentage of carbon nanotubes embedded in a traditional composite, the structures mass reduction, corresponding to the same mechanical behavior, is significant (∼20% using only 5 wt% of carbon nanotubes). The composites are: • polymeric or metallic matrix reinforced with fibers or particles; • polymeric or metallic matrix reinforced with nanoparticles or nanostructures; • hybrid configuration. Embedding nanoparticles in a matrix changes the manufacturing procedure. In this case, for nanostructured composite materials manufacturing, RTM (Resin Transfer Molding) is the better technology. The following steps for a future industrial and automatic production are used: • structure design; • 3D CAD of the positive simulacra of the structure is useful to produce, by rapid prototyping, the positive mold; • silicone negative mold is produced by using the positive mold. The use of silicone reduces the cost, allows the manufacturing of the complex shape and is reusable. A better dimensional stability is also obtainable. An example of a flat anisogrid structure is shown in Fig. 6.41. When using only nanoparticles, RTM technology is the best method. Figure 6.42a illustrates a typical RTM facility used to manufacture a plate of a polymeric composite reinforced with carbon nanoparticles (Fig. 6.42b). The samples have a very smooth surface without macro- and microscopic defects. This structural typology offers, simultaneously, different properties: mechanical, thermal, electrical and magnetic. Using nondestructive testing (NDT) techniques (X-ray and ultrasound systems) it is possible to control the internal morphology. Each specific application aims to obtain an isotropic, continuous and homogeneous material (on macro- and micro-scale): the NDT allows the sample characteristics to be evaluated and then the manufacturing technologies and processes adopted to be validated. Figure 6.43 shows the NDT tests of the plate reported in Fig. 6.42b. The features of the nanostructured polymeric composite materials have been broadly investigated. Some observations are useful to the metallic and ceramic matrices. Metal matrix nanostructured composites have been little investigated. These materials are generally produced using metallurgical methodologies

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

169

6.41 Flat anisogrid lattice structures produced in a silicone mold with a polymeric composite material reinforced with nanoparticles of graphite and carbon nanotubes.51

but the nanoparticles, and in particular carbon nanotubes, are not optimized (quantities, morphologies, etc.). SEM and TEM observations show that the carbon nanotubes embedded in the matrix are not damaged after manufacturing and are well distributed (the alignment degree depends on the manufacturing method adopted). The mechanical test shows that the use of the carbon nanotubes (dispersed for example in a Ti matrix) provide significant improvement of the Young’s modulus, hardness, and wear loss and friction coefficient. The nanostructured metallic composites are a significant development of MMC (metal matrix composites). The ceramic nanoreinforced composites are typically prepared by a mechanical mixing technique using specific precursors and a hot pressing sintering procedure (a sol gel or CVD is also employed). A homogeneous dispersion is observed without no significant damage to the carbon nanotubes embedded in the matrix. The use of carbon nanotubes allows refinement of composite microstructures, a strong reduction of the relative density and an improvement of the toughness and the friction coefficient. In addition, their presence when well-dispersed confers an electrical conductivity to the otherwise insulating ceramic matrix composites. Besides, thermal management using partially stabilized zirconia (PSZ) mixed with carbon nanotubes is important in aerospace applications.

© 2007, Woodhead Publishing Limited

170

Nanofibers and nanotechnology in textiles (a)

(b)

6.42 (a) The RTM facility used to manufacture a plate constituted by nanostructured polymeric composite material.64 (b) An example of a prototype obtained by RTM.64

6.8

Nanostructured solid propellants for rockets61, 76–79

Propulsion by means of solid propellant is one of the advanced aerospace and missile fields of research. The remarkable performance, with a technologically more simple system than the liquid propellant jet, put this type of propellant at the heart of many advanced applications. Solid propellant engines find their application in all the ‘propelled mission phases’ where the reference parameters (Trust Vectoring Control (TVC), time of ignition,

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

171

(a)

(b)

6.43 (a) Ultrasound spectra produced to analyze the morphology of the nanostructured plate produced with the RTM facility (see Fig. 6.42(b)).64 (b) The analysis of the negative image (obtained from X-ray) of the demonstrator similar to that reported in Fig. 6.42(b). In this case internal defects are observed. This indicates that the RTM process has not been performed correctly.64

© 2007, Woodhead Publishing Limited

172

Nanofibers and nanotechnology in textiles

predefined throw/trajectory and orbital corrections) are defined. Applications such as the orbital corrections of a satellite to avoid ambiguous gravitational perturbations, need liquid propulsion systems that provide a more flexible system. However, such systems are more complex, and have a different intrinsic reliability than a solid propulsion system. The main reference parameters for the propellant performances are: • combustion velocity; • specific impulse; • stability of the flame front. The propellant is characterized by an appropriate mixture of a fuel and a comburent (more additives and eventual catalysts). Apart from the physical effects related to the geometrical/dimensional characterization of the propellant reservoir, all the intrinsic performances of the solid grain depend on the physical /mechanical properties of the fuel– comburent mixture. The rocket thrust (T) is defined as: ˙ e T = mV

[6.10]

where: T = thrust m˙ = mass flow Ve = exhaust gas velocity and, in particular:

m˙ = ρrA

[6.11]

where: ρ = propellant density A = the nozzle exhaust section r = burning rate the burning rate (r) is a fundamental parameter and it is defined by the Vielle law: r = αPn

[6.12]

where: α and n = ballistic coefficients P = combustion chamber pressure. A solution typically used to increase the performance (r and then T) consists of charging the propellant with metallic micro-particles that are extremely energetic, such as aluminium. During combustion this allows further energy to be supplied that improves the specific impulse. There are substantial problems associated with this technique such as: • homogeneous distribution of the powder is necessary in the solid mixture of the grain, in order to avoid localised conglomerates;

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

173

• incomplete combustion of the micro-charge, that gives: 䊊 mechanical degradation of the inner exhaust nozzles, 䊊 toxicity of the exhausted gas, 䊊 reduced improvement of the grain performances, 䊊 thermo-fluid dynamics problems associated with the burned gas, due to the presence of bi-phase flow, 䊊 in the case of missiles, there is greater intercept opportunity, for the detection of a hot trace due to unburnt solid particulate; • environmental compatibility of the exhausted gases; • difficulties in reducing the micro-charge dimensions (to increase the thermal exchange surface) on a nanometric scale; • energetic limits of the charge itself; • high costs. The utilization of different techniques for powder production (Electrical Explosion Wire (EEW), plasma condensation, mechanical shattering) and their coating may lead to significant variations in grain performance. In addition, formulations including not only one, but two, oxides, ammonium perclorate and ammonium nitrate, criteria are under study, with the aim of reducing costs without heavy impact on the performance. Grain characterization, is necessary for any chemical formulation through the following: • Determination of the combustion velocity as a function of the pressure, the measurement of the ignition delay, the study of the combustion of the propellant under laser radiation (for a stability analysis), the analysis of the combustion residuals. For applications where the engine will be subject to high acceleration, it will be necessary to analyze the eventual effects of the acceleration towards the combustion velocity. • Dimensional, morphological and chemical characterization of the ingredients, with particular attention to the metallic nano-powders and the combustion residuals, by means of electron microscopy (SEM, TEM, HRTEM), XPS (X-ray photoelectron spectroscopy) and XRD (X-ray diffraction). • Study of the powder agglomeration at the combustion surface level, by means of high-speed cameras. • Theoretical–numerical study of the simulation of the combustion processes, needed for a basic understanding of the complex phenomena involved, the development of parametric studies and to address experimental research activities. With the nanotechnology, using nanoparticles or nanostructures, it is possible to obtain further improvement of the rocket performance. In particular, the combustion of nano-sized elements gives a significant increase of the thermal spatial gradient in the combustion chamber, with a consequent increase of

© 2007, Woodhead Publishing Limited

174

Nanofibers and nanotechnology in textiles

the burning rate (r) and then of the thrust (T). The possible use of carbon nanotubes has entered the frame of innovative formulation studies, in addition to other materials with nanometric granulometria. Thanks to their covalent unidirectional configuration, the carbon nanotubes may: • improve the mechanical characteristics of the propellant; • improve the thermal energy released during the combustion, so improving the specific impulse and the relevant exhaust; • reduce the environmental pollution problems; • reduce the mass fraction of the unburnt residuals. The use of simple graphite with nanometric granulometria, whose plane crystal characterization sp2 is the same as that of the carbon nanotubes, may give the same performances at macroscopic levels, with a significant cost reduction and a remarkable simplification of the technological processes. To increase rocket performance (burning rate, thrust, combustion stability, etc.) it is possible to decrease the propellant quantities required, with a consequent improvement of the payloads transportable. The principal objectives of the research programs in nanotechnology propulsion are: • to implement the manufacturing technologies of the propellants with nanometric metallic particles; • to test bi-oxidizing solutions (ammonium perclorate and nitrate); • to study (theoretically and experimentally) carbon-based (graphite, carbon nanotubes) nanometric charges (nanostructured and non-nanostructured); • to characterize the new solid propellant and evaluate its performances; • to foresee the eventual use of a metallic charge, associated to a carbon base. Experimental activity is relevant for the manufacturing of solid propellant that, typically, includes the use of the following base materials: • commercial aluminum powder charge (micrometric granulometria, typically 30–50 µm); • nanometric aluminum powder charge (granulometria less than 1 µm); • micrometric graphite powder charge (<20 µm); • micrometric graphite powder charge (<20 µm) with the addition of aluminum; • carbon nanotubes charge; • carbon nanotubes charge and the addition of aluminum; • aluminum nitrure nanotubes charge (with and without nanometric aluminum and carbon-based powders). Figure 6.44 shows a combustion example of the nanostructured solid propellant sample.

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

175

6.44 Experimental combustion test of a nanostructured solid propellant sample for rockets employed in aerospace systems.61

The new formulation propellants must be compared with those actually in use. The final goal is to lead to new, innovative formulations to reach the following targets: • • • • •

the development of lighter propulsion systems, with improved performances; an increased payload mass; a cost reduction; an environmental pollution reduction; improved thermo-mechanical characteristics.

6.9

Frequency selective surfaces for aerospace applications64, 128–137

Frequency selective surfaces (FSS) are an important application in many engineering sectors. In particular, it is useful for radomes, filters and radar communications. Traditional FSSs are constituted by two possible configurations: periodically perforated metallic screens, or arrays of metallic patches printed on dielectric substrates. Aeronautic, military and naval applications are the typical technological fields involved in the FSS developments. The requirements of an FSS are: • pass-band filter behavior; • to hide the surface from the external observer.

© 2007, Woodhead Publishing Limited

176

Nanofibers and nanotechnology in textiles

With a specific design of the traditional FSS, as a result of the Floquet theorem, it is possible to select a unique frequency value for the filter. By solving Maxwell equations, it is possible to determine the theoretical electromagnetic FSS behavior. However, considering the complexity of the phenomena involved, it is only possible to perform a realistic characterization of the FSS with an experimental test. New kinds of FSS have been introduced that incorporate nanotechnology. In fact, using particular matrices (thermoplastic and thermosetting polymeric resins, silicones, etc.) with homogeneously dispersed, nanoparticles (structured and non-structured) it is possible to produce a material necessary for the electromagnetic, structural and mechanical requirements. In contrast, traditional FSSs provide only electromagnetic properties. For a nanostructured FSS a particular technological manufacturing process is required to obtain the following micro- and macro-properties: homogeneity, continuity and isotropic characterization (necessary to the imposed requirements). Moreover, these innovative FSSs provide a continuous monolithic structure with a significant improvement in the reliability of the systems in which they are integrated. Traditional FSSs are periodic structures with filtering properties, traditionally manufactured either as periodically perforated metallic screens or as arrays of metallic patches printed on dielectric substrates. A novel approach is proposed, whereby narrowband filtering properties are created from random composite structures based on the physical resonant properties of the constituents and the geometry of micro- and nano-inclusions. In this manner, a bulk continuous material rather than a lattice formation is used to manipulate and shape the electromagnetic propagation. The novel artificial dielectrics constitute conformal FSSs to be applied by means of a uniform coating process to simple planar or complex curvilinear shapes. The approach is guided by a theoretical design for a random mixture with frequencyselective properties, characterized by a concentric geometry for the inclusions. The frequency dispersion of the proposed composite is driven by the use of a Lorentzian resonant dielectric as one of the constituent media. The novel complex medium is an amorphous ensemble of micro- and nanospheres composed of a lossy core, coated with a highly resonant dielectric layer and embedded in a dielectric host (a polymeric matrix). This is an innovative application of nanotechnologies in the electronic field. With these innovative nanostructured FSSs, it is necessary to develop appropriate software useful for evaluating the electromagnetic and mechanical behavior. The nanostructured FSS consists of a dielectric matrix (polymeric or silicone resin) with micro- and nanoparticles embedded into it. For this composite nanostructured material, it is required to have isotropy, homogeneity and continuity properties. In fact, when the nano-FSS works at high frequency,

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

177

all possible material defects can significantly modify the electromagnetic behavior of the element. This is the reason why it is necessary to define a specific sample-manufacturing procedure required to obtain samples with the above characteristics. Another problem is relevant to the choice of materials. First, a matrix with a specific electromagnetic property is required. It is necessary to study different aspects such as the thermal, mechanical, physical and chemical characterization of the matrix. In fact, the design of an innovative nano-FSS is not limited only to the electromagnetic analysis, but also to its integration in a real operative system with specific requirements, relevant to all the operative aspects and conditions. For example, the resistance to flames and salty water are only two conditions of the possible specific required characterizations to certificate the nano-FSS produced. The surface of the sample must be very smooth to guarantee well-defined control of the electromagnetic behavior. The static and dynamic resistance of the materials are important parameters concerning the mechanical properties. Using a polymeric matrix, it is possible to ensure enough mechanical resistance both for specific electromagnetic instruments (advanced devices) and for naval/ aeronautics applications (e.g. nano-FSS panel integrated on a boat and aircraft as an electromagnetic shield). It is possible to think also of a flexible nanoFSS to produce devices with very complex geometries. In this case it is necessary to choose a specific matrix (e.g. silicone materials). A further aspect is relevant to the nanoparticles employed. The morphological characterization gives the opportunity to evaluate the possible use of a specific kind of micro- and/or nanoparticle that has particular properties (electromagnetic, mechanical, chemical, etc). In this case it is possible to use simple particles (non-nanostructured) or particular elements (carbon nanotubes, for example). The manufacturing procedures for the sample preparation can be very different, with significant changes in the relevant behavior of the nano-FSS. For example, embedding carbon nanotubes on the matrix with different alignments and degrees of purification can provide very different behaviors of the FSS. This gives an idea of the technological problems involved in manufacturing a nano-FSS with characteristics defined by the theoretical and numerical models. The manufacturing procedure is much simpler with the use of simple nanoparticles. Cost and availability of a suitable quantity of nanostructured materials are two fundamental parameters in the development on these innovative elements. The nano-elements embedded in the matrix can provide other properties. For example, carbon nanotubes are studied for the development of structural composites for aerospace applications, thermal management, electrical systems, etc. In this case it is possible to think of advanced materials with manifold properties (mechanical, thermal, chemical, and mainly electromagnetic). The materials typically employed in nano-FSS manufacture are:

© 2007, Woodhead Publishing Limited

178

Nanofibers and nanotechnology in textiles

• matrix: polymeric and silicones; • commercial or specific curing agents; • particles: micro-powder of non-nanostructured graphite, carbon nanotubes and metal oxides. It is possible to produce two kinds of nanostructured sample: continuous and multilayer. In each case the uniformity of the sample thickness is fundamental to provide a homogeneous electromagnetic behavior in the studied band (for example the X-band). The preliminary activities in the nano-FSS design are devoted to defining the sample manufacturing procedures, useful for obtaining materials with specific parameters (dimensions, geometry, porosity, roughness, etc.). With these samples (Fig. 6.35) it is possible to perform a static and dynamic mechanical tests that provide important results on the structural behaviour of this material that will be employed for electromagnetic applications. Using SEM, a fracture-section characterization is also possible (Fig. 6.38). This allows the internal morphology of materials and the micromechanics composite behavior to be investigated. Also in this case by using specific software integrated in the SEM, a statistical analysis of the particle dimension can be performed. It is necessary to identify a region in which numerous micro- and/or nanoparticles are present, and then to define the range dimensions to study. In this case, the software, using different colors, determines particle groups, each with a specific dimension range. Figure 6.25 shows an SEM micrograph of microparticles with different dimensions and shapes. It is interesting to produce nanostructured thin films that can be used in hybrid multilayer composites dedicated to mechanical and FSS applications (Fig. 6.39). After the preliminary base material characterizations it is possible to produce nanostructured FSS samples. For example, the materials employed are: • graphite (0%, 50%, 65% in wt with respect to the resin + curing agent); • epoxy and polyester resin and silicone; • curing agent. For each material a specific curing cycle has been adopted, as specified by the material’s datasheet and by the curing test previously performed. It is important to note that the curing process of the matrices can vary in significant ways when the particles are embedded in it. Figure 6.45 shows the various methodologies and phases of the sample manufacturing. To evaluate the selectivity of the developed nanostructured materials it is necessary to measure (in magnitude and phase) the scattering parameters (Sij). An accurate calibration of the instrument is required, so as to eliminate the effects of the various transitions on the performances of the Device Under Test (DUT).

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

179

Considering an analyzer experimental facility (Fig. 6.46) with two ports (1 and 2), the magnitude of the scattering parameter S21 (that is, the response at port 2 when port 1 is excited), is defined as: S21 (dB) = 101g 10

Pout Pin

[6.13]

where: Pout = electromagnetic power measured at port 2 Pin = electromagnetic power measured at port 1 In particular, Pin indicates the power acting on the nanostructured materials (deposited in the waveguide; see Fig. 6.45), and Pout the transmitted power through the nanomaterials themselves. In a simplified way: if Pout = Pin → dB = 0 if Pout < Pin → dB < 0 (in this case the material absorbs the electromagnetic power acting on it). In the study of FSSs, two other important physical parameters are: • electrical permittivity (ε): providing the electrical behavior of the material = Re(ε) + j Im(ε) ⇒ εc = ε – j (σ/ω), with σ = power loss due to Joule effect, ω = 2πf (f = frequency) with Re(ε) always > 0 (capacitive effect = electric energy absorption); by Poynting Theorem: Im(ε) > 0 (active material); Im(ε) < 0 (electrical power loss).

6.45 Nanostructured FSS samples (before the final mechanical removal of the residual material on the waveguide surfaces) and the relative manufacturing procedures.64

© 2007, Woodhead Publishing Limited

180

Nanofibers and nanotechnology in textiles

6.46 Agilent portable network analyzer (PNA) E8363B – experimental set-up.64

• magnetic permeability (µ): providing the magnetic behavior of the material = Re(µ) + j Im(µ) with Re(µ) always > 0 (inductive effect = magnetic energy absorption); by Poynting theorem: Im(µ) > 0 (active material); Im(µ) < 0 (magnetic power loss). In the case of FSSs, it is necessary to obtain the following general results: [Im(ε), Im(µ)] < 0 (the specific values will depend on the materials employed). In the case of a ‘metamaterial’ it may be: [Re(ε), Re(µ)] < 0. Figure 6.47 illustrates an example of the S21 measure in the X-band. It is possible to evaluate (in dB): • mean electromagnetic power loss; • minimum electromagnetic power loss; • maximum electromagnetic power loss. In addition, it is possible to observe that the electromagnetic power absorbed by materials (dielectric + graphite) is increased with respect to the host (only dielectric). This indicates that the micro- and nano-inclusions, embedded on the polymeric matrix, allow the transmitted power to be reduced. With experimental measures, it is possible to observe that by increasing the quantity of carbon nanoparticles embedded in the polymeric matrix, the electromagnetic power absorption increases and Pout decreases (with respect to the same value of Pin). In addition, analogous results are obtainable measuring the electrical permittivity (ε) and the magnetic permeability (µ). In fact, if the percentage of the micro- and carbon nanoparticles embedded in the polymeric matrix increases, the following results are obtained:

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

0.00

181

dB-S21

–0.50 –1.00 –1.50 –2.00 –2.50 –3.00 –3.50 –4.00 –4.50 –5.00 Ch9: Start 8.00000 GHz

Stop 12.0000 GHz

6.47 Frequency response (magnitude in dB) of the sample with nanoparticles of graphite embedded in the matrix.64

| Im(ε) | ↑ with Im(ε) < 0 and | Re(ε) | ↑ with Re(ε) > 0 | Im(µ) | ↑ with Im(µ) < 0 and | Re(µ) | ↑ with Re(µ) > 0 The use of the silicone materials is indicated due to the possibility of producing flexible FSSs, with the important implication of making FSS elements with a complex shape. Then, the principal characterizations of these innovative materials are: • to obtain a specific electromagnetic behavior; • to provide mechanical and thermal properties; • to select a definite frequency in which S21 ∼ 100% and S12 ∼ 0% (expressed in terms of the percentage of the external power excitation acting on the FSS surface). After experimental investigations, it is necessary to analyze, for each sample typology, the selectivity behavior. In particular, for each percentage of the particles embedded in the matrix, it is necessary to determine the specific frequency value at which materials provide the requested selectivity. This is an important goal, with a broad relevance to many scientific sectors and engineering applications. The principal activities are focused on the technological methods necessary to manufacture the nano-FSS. Particular attention is dedicated to characterization of the base materials (polymeric matrix, curing agent,

© 2007, Woodhead Publishing Limited

182

Nanofibers and nanotechnology in textiles

nanoparticles) using electron microscopy (SEM, HRTEM and EDX). The degassing phase (used also for the nanostructured composite) is essential to obtain the three following micro- and macro-properties: homogeneity, continuity and isotropic characteristics. Also, the curing phase (temperature, pressure, inert conditions) characterizes the final properties of the produced FSS. Non-destructive testing allows us to verify if the above properties are obtained and then to qualify the technological process employed. The experimental electromagnetic test provides interesting results. In particular, using different quantities of the nanoparticles embedded on the polymeric matrix it is possible to obtain various electromagnetic behaviors. In fact, by increasing the percentage of the particles embedded in the matrix, the sample presents a considerable improvement of the electromagnetic radiation absorbed by the material. For each percentage it is necessary to determine the specific frequency of selectivity. The future applications of these innovative nanomaterials include the development of multifunctional hybrid nanostructured composite materials able to provide, simultaneously, mechanical, thermal and electromagnetic specific behaviours for aerospace, military, navy and communication applications.

6.10

Other aerospace applications of carbon nanotubes65, 66, 138

Nanotechnology can be applied in many other scientific sectors of aerospace engineering. • 3D NonoTopography: the good mechanical properties of carbon nanotubes make them an ideal force sensor in scanning probe microscopy with high durability, reliability and capability to reconstruct 3D Images surface with high resolution, resolving the typical limitations of conventional force sensors (typically ceramics) • Chemical force microscopy: nanostructures and in a special way functionalized carbon nanotubes allows selective micrographs to be created based on chemical discrimination (chemical force microscopy). With functionalized nanostructures it is possible to improve the special resolution of the micrographs produced by increasing the chemical reactivity of the samples. The chemical interaction between nanotubes and chemical species present on the observed surface, allows a chemical mapping of the analyzed sample to be performed. • Field emission: carbon nanotubes have been demonstrated to be very efficient field emitters. Using this fundamental property new innovative electronic devices may be developed, including flat panel, computers, Xray facilities, etc. For example, replacing the traditional glass support of

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

•

•

•

•

•

• •

183

TV and PC screens with nanostructured polymeric materials would make it possible to produce a flexible screen. In addition, the use of carbon nanotubes gives the possibility of developing devices with high electronic emission stability, mechanical resistance and longer lifetimes. Chemical sensors: the electrical properties of carbon nanotubes (single wall or multiwall) were recently demonstrated to be very sensitive to chemical composition variations of the surrounding atmosphere at room temperature. This very interesting property would allow an innovative chemical sensor to be designed with high sensibility and capability to detect various chemical substances. Catalyst support: due to their ability to be tailored to specific needs, carbon nanotubes are candidate supports in heterogeneous catalytic processes. Carbon nanotubes are also employed as catalytic support due to their high surface area, chemical and thermal stability (in a non oxidative environments). Carbon nanofibres, soot and graphite are used in these applications. Adsorption: significant adsorption and interaction phenomena occur between carbon nanotubes and gases. There are two possible applications for this sector. The first is characterized by the molecular adsorption, which is related to carbon nanotube electrical properties and, then, with the possibility of developing chemical sensors. The second includes gas storage and separation due to the high surface area of carbon nanotubes. Storage: It is fundamental to develop lightweight and safe hydrogen storage systems in the automotive industry. Due to results obtained with hydrogen storage studies several researchers, using carbon nanotubes, have tried to develop innovative gas storage facilities (oxygen, nitrogen, inert and noble gases, hydrocarbon, etc). Gas separation: this aspect is a consequence of the above storage application. Due to nanostructure properties, it is possible to control the sorption phenomena with specific conditions (pressure, temperature, nanostructure morphology) Absorbtion: experimental activities demonstrate that carbon nanotubes are able to absorb toxic gases (e.g. dioxin, fluoride or alcohols). This application is fundamental for military applications (NBC). Biosensors: attaching molecules to a carbon nanotube surface is a very interesting way to realize nano-biosensors. The use of the internal cavity of carbon nanotubes for drug delivery is another amazing possible application for many sectors (propulsion, health monitoring, medicine, etc.). For example, introducing carbon nanotubes holding specific sensors or chemical elements into the human body, would make it possible to monitor heath. This is a possible application for long-term aerospace missions. Two basic aspects must be investigated: the toxicity and reliability of these systems in respect to traditional medical apparatus.

© 2007, Woodhead Publishing Limited

184

Nanofibers and nanotechnology in textiles

In each application described above, the functionalization of carbon nanotubes, or of nanostructures in general, represents the fundamental step.

6.11

Conclusions

As described in this chapter nanotechnology offers numerous opportunities for significant improvement in many sectors of the aerospace engineering (structures, composite radar shield, propulsion, etc.). In fact, by introducing nanostructures and nano-elements, it is possible to design devices and apparatus with properties and characteristics that cannot be obtained by traditional technology. Development cost will be the key to realistic use of nanotechnology in aerospace engineering. The possibility of translating base research activities to the industrial scale, with competitive costs, represents the principal challenge of this new branch of science. Nanotechnologies are characterized by a strong interdisciplinary character, which allows scientific sectors, typically very different from one another, to work jointly to design advanced systems. However, it should be considered that, for each application, a specific set of requirements must be satisfied, employing different specific ‘nano-methodologies’. Aerospace engineering is characterized by large dimension systems (launchers, for example) with high performance. For this reason nanotechnology should overcome the limitations of traditional technologies. Some researchers consider ‘nano-elements’ to be the materials and structures of the 21st century. The economic, scientific and technological efforts necessary are significant, but with strong collaboration among universities, industries, researchers and scientists the application of nanotechnology science to human life will become a reality.

6.12

Acknowledgments

I would like to thank the following professors and researchers for the important contribution they made to the realization of this chapter: Prof. Mario Marchetti: Full Professor of Aerospace Structures at the High School of Aerospace Engineering of the University of Rome ‘La Sapienza’ Dr Franco Mancia: Researcher at C.S.M. Centro Sviluppo Materiali S.p.A., Rome, Italy Prof. Gilberto Rinaldi: Professor of Chemistry at the Faculty of Engineering of the University of Rome ‘La Sapienza’ Prof. Luciano Galfetti: Full Professor of Aerospace Propulsion at the Politecnico of Milano, Italy Dr Stefano Bellucci, Dr Giorgio Giannini: Researchers at INFN (Istituto Nazionale di Fisica Nucleare) dei Laboratori Nazionali di Frascati (LNF) – Laboratory of Nanotechnology

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

185

Dr Luigi Paoletti, Dr Biagio Bruni: Researchers at ISS Istituto Superiore di Sanità – Health and Technology Department, Rome, Italy Mr Mario Brecciaroli: Researcher at ENEA of Frascati, Italy Dr Raffaele Mucciato, Dr Gilberto Gaggiotti: Researchers at 2M Strumenti S.p.A., Italy Mr Giancarlo Spera: Alitalia Maintenance Division, Fiumicino Airport, Italy Special thanks go to Dr Vincenzina Fusco, PhD in Food Science and Technology at the University of Naples Federico II, Department of Food Science, Division of Microbiology, for reading this chapter and providing helpful suggestions. The photos (6.1 and from 6.6 to 6.47) showed in this Chapter are also included in the following Degree Thesis: 1. Marco Regi – “Strutture Multigrid Realizzate Con Compositi Polimerici Rinforzati Connanotubi in Carbonio Per Applicazioni Aerospaziali”, Università of Roma “La Sapienza”, 2006. 2. Marco Regi – “Studio E Realizzasione Di Materiali Nanostrutturati Per Applicazioni Aerospaziali”, Università of Roma “La Sapienza”, 2006.

6.13

References

1. Takikama, H., Kusano, O., Sakakibara, T. – ‘Graphite cathode spot produces carbon nanotubes in arc discharge’, Appl. Phys. 32 (1999) 2433–2437 2. Zeng, H., Zhu, L., Hao, G., Sheng R. – ‘Synthesis of various forms of carbon nanotubes by ac arc discharge’, Carbon 36 (1998) 259–261 3. Takikama, H., Ikeda, M., Hirahara, K., Hibi, Y., Tao, Y., Ruiz jr., P. A., Sakakibara, T., Itoh, S., Iijima S. – ‘Fabrication of single walled carbon nanotubes and nanohorns by means of a torch arc in open air’, Physica B 323 (2002) 277–279 4. Bae, J. C., Yoon, Y. J., Lee, S., Baik, H. K. – ‘Field emission of carbon nanotubes deposited by electrophoresis’, Physica B 323 (2002) 168–170 5. Zhen-Hua, L., Miao, W., Xin-Qing, W., Hai-Bin, Z., Huan-Ming, L., Ando Y. – ‘Synthesis of large single walled carbon nanotubes by arc discharge’, Chin. Phys. Lett. 19 (2002) 91–93 6. Tang, D. S., Xie, S. S., Chang, B. H., Sun, L. F., Liu, Z. Q., Zou, X. P., Li, Y. B., Ci, L. J., Liu, W., Zhou, W. Y., Wang, G. – ‘Effect of acetylene in buffer gas on the microstructures of carbon nanotubes in arc discharge’, Nanotechnology 13 (2002) L1–L4 7. Takikama, H., Tao, Y., Miyano, R., Sakakibara, T., Ando, Y., Zhao, X., Hirahara, K., Iijima, S. – ‘Carbon nanotubes on electrodes in short-time heteroelectrode arc’, Mater. Sci. Eng. C 16 (2001) 11–16 8. Dong-Sheng, T., Wei-Ya, Z., Li-Jie, C., Xiao-Qin, Y., Hua-Jun, Y., Zhen-Ping, Z., Ying-Xin, L., Dong-Fang, L., Wei L. – ‘Morphologies and microstructures of carbon nanotubes prepared by self-sustained arc discharging’, Chin. Phys. 11(5) (2002) 496–501 9. Lange, H., Sioda, M., Huczko, A., Zhu, Y. Q., Kroto, H. W., Walton, D. R. M. –

© 2007, Woodhead Publishing Limited

186

10.

11. 12.

13. 14. 15.

16.

17.

18.

19. 20.

21.

22.

23. 24.

25.

Nanofibers and nanotechnology in textiles ‘Nanocarbon production by arc discharge in water’, Carbon 41 (2003) 1617– 1623 Zhang, H., Xue, X., Wang, D., He, Y., Peng S. – ‘The effect of different kinds of inert gases and their pressures on the preparation of carbon nanotubes by carbon arc method’, Mater. Chem. Phys. 58 (1999) 1–5 Li, Y., Xie, S., Zhou, W., Tang, D., Liu, Z., Zou, X., Wang, G. – ‘Small diameter carbon nanotubes synthesized in an arc discharge’, Carbon 39 (2001) 1421–1446 Lee, S. J., Baik, H. K., Yoo, J., Han, J. H. – ‘Large scale synthesis of carbon nanotubes by plasma rotating arc discharge technique’, Diamond Related Mater. 11 (2002) 914–917 Huang, H., Kajiura, H., Murakami, Y., Ata, M. – ‘Metal sulphide catalyzed growth of carbon nanofibers and nanotubes’, Carbon 41 (2003) 579–625 Takikama, H., Kusano, O., Sakakibara, T. – ‘Graphite cathode spot produces carbon nanotubes in arc discharge’, Appl. Phys. 32 (1999) 2433–2437 Maser, W. K., Munoz, E., Benito, A. M., Martinez, M. T., de la Fuente, G. F., Maniette, Y., Anglaret, E., Sauvajol, J. L. – ‘Production of high density single walled nanotube material by a simple laser ablation method’, Chem. Phys. Lett. 292 (1998) 587–593 Yudasaka, M., Yamada, R., Sensui, N., Wilkins, T., Ichihashi, T., Iijima, S. – ‘Mechanism of the effect of nico, Ni and Co catalysts on the yeld of single wall carbon nanotubes formed by pulsed Nd:Yag laser ablation’, J. Phys. Chem. B 103 (1999) 6224–6229 Yudasaka, M., Sensui, N., Takizawa, M., Bandow, S., Ichihashi, T., Iijima, S. – ‘Formation of single wall carbon nanotubes catalyzed by Ni separation from Y in laser ablation or in arc discharge using a C target containing a NiY catalyst’, Chem. Phys. Lett. 312 (1999) 155–160 Zhang, M., Yudasaka, M., Iijima, S. – ‘Single wall carbon nanotubes: a high yield of tubes through laser ablation of a crude-tube target’, Chem. Phys. Lett. 336 (2001) 196–200 Guo, T., Nikolaev, P., Thess, A., Colbert, D. T., Smalley, R. E. – ‘Catalytic growth of single walled nanotubes by laser ablation’, Chem. Phys. Lett. 243 (1995) 49–54 Zhang, H., Ding, Y., Wu, C., Chen, Y., Zhu, Y., He, Y., Zhong, S. – ‘The effect of laser power on the formation of carbon nanotubes prepared in CO2 continuous wave laser ablation at room temperature’, Physica B 325 (2003) 224–229 Burakov, V. S., Bokhonov, A. F., Nerdel’ko, M. I., Savastenko, N. A., Tarasenko, N. V. – Optical Emission Characteristics of Carbon Plasma Produced by Single and Double Pulse Laser Ablation, Institute of Molecular and Atomic Physics National Academy of Science of Belarus Yudasaka, M., Komatsu, T., Ichihashi, T., Iijima, S. – ‘Single wall carbon nanotube formation by laser ablation using double-targets of carbon and metal’, Chem. Phys. Lett. 278 (1997) 102–106 Arepalli, S., Nikolaev, P., Holmes, W., Scott, C. D. – ‘Diagnostic of laser-produced plume under carbon nanotube growth conditions’, Appl. Phys. A 69 (1999) 1–9 Scott, C. D., Arepalli, S., Nikolaev, P., Smalley, R. E. – ‘Growth mechanisms for single wall carbon nanotubes in a laser ablation process’, Appl. Phys. A 72 (2001) 573–580 Yokomichi, H., Sakai, F., Ichihara, M., Kishimoto, N. – ‘Attempt to synthesize carbon nanotubes by hot wire chemical vapor deposition’, Thin Solid Films 395 (2001) 253–256

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

187

26. Okazaki, T., Shinohara, H. – ‘Synthesis and characterization of single wall carbon nanotubes by hot filament assisted chemical vapor deposition’, Chem. Phys. Lett. 376 (2003) 606–611 27. Obraztsov, A., Zolotukhin, A., Ustinov, A., Volkov, A., Svirko, Y. – ‘Chemical vapor deposition of carbon films: in situ plasma diagnostics’, Carbon 41 (2003) 836–839 28. Chen, M., Chen, C., Chen, C. F. – ‘Preparation of high yield multiwalled carbon nanotubes by microwave plasma chemical vapor deposition at low temperature’, J. Mater. Sci. 37 (2002) 3561–3567 29. Wang, X., Hu, Z., Chen, X., Chen, Y. – ‘Preparation of carbon nanotubes and nanoparticles by microwave plasma enhanced chemical vapor deposition’, Scripta Mater. 44 (2001) 1567–1570 30. Jung, K., Boo, J., Hong, B. – ‘Synthesis of carbon nanotubes grown by hot filament plasma enhanced chemical vapor deposition method’, Diamond Related Mater. 13 (2004) 299–304 31. Dillon, A., Mahan, A., Alleman, J., Heben, M., Parilla, P., Jones, K. – ‘Hot wire chemical vapor deposition of carbon nanotubes’, Thin Solid Films 430 (2003) 292–295 32. Park, D., Kim, Y., Lee, J. – ‘Synthesis of carbon nanotubes on metallic substrates by a sequential combination of PECVD and thermal CVD’, Carbon 41 (2003) 1025–1029 33. Kong, J., Cassell, A., Dai, H. – ‘Chemical vapor deposition of methane for single walled carbon nanotubes’, Chem. Phys. Lett. 292 (1998) 567–574 34. Pradhan, D., Sharon, M. – ‘Carbon nanotubes, nanofilaments and nanobeads by thermal chemical vapor deposition process’, Mater. Sci. Eng. B 96 (2002) 24–28 35. Mauron, P., Emmenegger, C., Zuttel, A., Nutzenadel, C., Sudan, P., Schlapbach, L. – ‘Synthesis of oriented nanotube films by chemical vapor deposition’, Carbon 40 (2002) 1339–1344 36. Han, J., Choi, S., Lee, T., Yoo, J., Park, C., Jung, T., Yu, S., Yi, W., Han, I., Kim, J. – ‘Growth characteristics of carbon nanotubes using platinum catalyst by plasma enhanced chemical vapor deposition’, Diamond Related Mater. 12 (2003) 878–883 37. Allouche, H., Monthioux, M., Jacobsen, R. – ‘Chemical vapor deposition of pyrolytic carbon on carbon nanotubes. Part 1: synthesis and morphology’, Carbon 41 (2003) 2897–2912 38. Lee, C., Park, J. – ‘Growth and structure of carbon nanotubes produced by thermal chemical vapor deposition’, Carbon 39 (2001) 1891–1896 39. Grujicic, M., Cao, G., Gersten, B. – ‘Optimization of chemical vapor deposition process for carbon nanotubes fabrication’, Appl. Surface Sci. 191 (2002) 223–239 40. Bonard, J., Stora, T., Salvetat, J., Maier, F., Stockli, T., Duschl, C., Forrò, L., Heer, W., Chatelain, A. – ‘Purification and size selection of carbon nanotubes’, Adv. Mater. 9(10) (1997) 827 41. Colomer, J., Piedigrosso, P., Willems, I., Journet, C., Bernier, P., van Tendeloo, G., Fonseca, A., Nagy, J. – ‘Purification of catalytically produced multi wall nanotubes’, J. Chem. Soc., Faraday Trans., 94 (1998) 3753–3758 42. Bandow, S., Rao, A., Williams, K., Thess, A., Smalley, R., Eklund, P. – ‘Purification of single wall carbon nanotubes by microfiltration’, J. Phys. Chem B 101 (1997) 8839–8842 43. Yamamoto, K., Akiya, S., Nakayama, Y. – ‘Orientation and purification of carbon nanotubes using AC electrophoresis’, Appl. Phys. 31 (1999) L34–L36

© 2007, Woodhead Publishing Limited

188

Nanofibers and nanotechnology in textiles

44. Shelimov, K., Esenaliev, R., Rinzler, A., Huffman, C., Smalley, R. – ‘Purification of single wall carbon nanotubes by ultrasonically assisted filtration’, Chem. Phys. Lett. 282 (1998) 429–434 45. Hafner, J., Bronikowski, M., Azamian, B., Nikolaev, P., Rinzler, A., Colbert, D., Smith, K., Smalley, R. – ‘Catalytic growth of single wall carbon nanotubes from metal particles’, Chem. Phys. Lett. 296 (1998) 195–202 46. Dai, H., Rinzler, A., Nikolaev, P., Thess, A., Colbert, D., Smalley R. – ‘Single wall nanotubes produced by metal catalyzed disproportionation of carbon monoxide”, Chem. Phys. Lett. 260 (1996) 471–475 47. Flahaut, E., Govindaraj, A., Peigney, A., Laurent, C., Rousset, A., Rao, C. – ‘Synthesis of single walled carbon nanotubes using binary (Fe, Co, Ni) alloy nanoparticles prepared in situ by the reduction of oxide solid solution’, Chem. Phys. Lett. 300 (1999) 236–242 48. Yamamoto, K., Akiya, S., Nakayama, Y. – ‘Orientation and purification of carbon nanotubes using AC electrophoresis’, Appl. Phys. 31 (1999) L34–L36 49. Morishita, K., Takarada, T. – ‘Scanning electron microscope observation of the purification behaviour of carbon nanotubes’, J. Mater. Sci. 34 (1999) 1169–1174 50. Regi, M., Mancia, F., Marchetti, M., Amantini, L. – ‘Study of carbon nanotubes process for their application in the aerospace engineering’, IX Workshop on Microwave Engineering, Metamaterials and Special Materials for Electromagnetic Applications and TLC, Aula Magna University of ‘ROMA TRE’, 5 April 2004, Proceeding of Ronchi Foundation 1–2 (2005) 267 51. Regi, M., Mancia, F., Marchetti, M. – ‘Design and characterization of anisogrid lattice structures with carbon nanotubes’, Proceedings International Workshop on MEMS and Nanotechnology Integration (MNI): Application, 10–11 May 2004, Montreux (Svizzera) 52. Regi, M., Marchetti, M., Mancia, F., Allegri, G. – ‘Synthesis of carbon nanotubes and their application in “anisogrid lattice structures” ’, Proceedings SEM X International Congress & Exposition on Experimental and Applied Mechanics, 5th International Symposium on MEMS and Nanotechnology, 7–10 June 2004, Costa Mesa, California, USA 53. Regi, M., Mancia, F. – ‘Synthesis and characterization of carbon-nanotubes and their application in the aerospace engineering’, Proceedings Polymer Fibres 2004, 14–16 July 2004, UMIST Conference Centre, Manchester, UK 54. Regi, M., Mancia, F., Marchetti, M. – ‘Synthesis of carbon nanotubes and their application in “anisogrid lattice structures” ’, Proceedings International Conference on Experimental Mechanics ICEM12, 29 August–2 September 2004, Bari, Italy 55. Regi, M., Mancia, F., Marchetti, M., Totaro, G., De Nicola, F., Vasiliev, V. V., Rasin, A. F. – ‘Nanostructured composite materials and anisogrid lattice structures for aerospace applications, Part A: synthesis of carbon nanotubes and their application in “anisogrid lattice structures”, Part B: optimized design of isogrid and anisogrid lattice structures’, Proceedings 55th IAC International Astronautical Congress IAF, 4–8 October 2004, Vancouver, Canada 56. Mancia, F., Regi, M., Marchetti, M. – ‘Optimization of the synthesis of carbon nanotubes using the arc discharge method’, Proceedings of “School and Workshop on Nanotubes and Nanostructures” NN2004, LNF – INFN, Frascati, 14–20 October 2004 57. Bellucci, S., Balasubramanian, C., Mancia, F., Marchetti, M., Regi, M., Tombolini, F. – ‘Electrical and mechanical properties of composite materials based on carbon nanotubes for aerospace applications’, Proceedings (Poster + Paper) Canues 2004,

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

58.

59.

60.

61.

62.

63.

64.

65. 66. 67. 68.

69. 70.

71. 72. 73.

74.

189

Conference on Micro-Nano-Technologies for Aerospace Applications: From Concepts to Systems, 1–5 November 2004, Monterey, CA Bellucci, S., Balasubramanian, C., Mancia, F., Marchetti, M., Regi, M., Tombolini, F. – ‘Composite materials based on carbon nanotubes for aerospace applications’, Proceedings ICEM04 International Conference on Experimental Mechanics, 29 November–1 December 2004, Singapore Marchetti, M., Regi, M., Mancia, F. – ‘Damage response of anisogrid lattice aerospace structures’, Proceedings dell’ICOSAAR2005 9th International Conference on Structural Safety and Reliability, Rome, Italy, 19–23 June 2005 Regi, M., Marchetti, M., Mancia, F. – ‘Ricostruzione laser 3D della morfologia superficiale di elettrodi di grafite contenenti nanotubi in carbonio’, Tutto Misure 4 (2005) Galfetti, L., De Luca, L. T., Meda, L., Marra, G., Marchetti, M., Regi, M., Bellucci, S. – ‘Nanoparticles for solid rocket propellants’, Nanoscience and Nanotechnology NN2005, INFN–LNF, Villa Mondragone Monteporzio Catone, Frascati, Rome, 14–16 November 2005 Regi, M., Mucciato, R., Bellucci, S., Marchetti, M., Gaggiotti, G., Borin, P., Giannini, G. – ‘Study and characterization of carbon nanotubes with the atomic force microscopy (AFM)’, Nanoscience and Nanotechnology NN2005, INFN–LNF, Villa Mondragone Monteporzio Catone, Frascati, Rome, 14–16 November 2005 Bellucci, S., Balasubramanian, C., Giannini, G., Borin, P., Regi, M., Micciulla, F. – ‘Field emission studies from multiwalled carbon nanotubes’, Nanoscience and Nanotechnology NN2005, INFN–LNF, Villa Mondragone Monteporzio Catone, Frascati, Rome, 14–16 November 2005 Marchetti, M., Frezza, F., Regi, M., Mazza, F., Carnà, E. – ‘Development and characterization of nanostructured frequency selective surfaces (FSS)’, Proceeding 46th IACAS – Israel Annual Conference on Aerospace Structures, 1–2 March 2006 Dresslhaus, M. S., Dresslhaus, G., – Carbon Nanotubes: Synthesis, Structure, Properties and Application, Springer Verlag Berlin Heidelberg New York, 2001 Guozhong, C. – Nanostructures and Nanomaterials, Imperial College Press, 2004 Nan, C., Shi, Z., Lin, Y. – ‘A simple model for thermal conductivity of carbon nanotube based composites’, Chem. Phys. Lett. 375 (2003) 666–669 Arepalli, S., Nikolaev, P., Gorelik, O., Hadjiev, V., Holmes, W., Files, B., Yowell, L. – ‘Protocol for the characterization of single walled carbon nanotube material quality’, Carbon 42 (2004) 1783–1791 Morishita, K., Takarada, T. – ‘Scanning electron microscope observation of the purification behaviour of carbon nanotubes’, J. Mater. Sci. 34 (1999) 1169–1174 Feng, S. Q., Yu, D. P., Hu, G., Zhang, X. F., Zhang, Z. – ‘The HRTEM observation of cross-sectional structure of carbon nanotubes’, J. Phys. Chem. Solids 58(11) (1997) 1887–1892 Hirsch, P., Howie, A., Nocholson, R., Pashley, D. W., Whelan, M. J. – Electron Microscopy of Thin Crystals, Krieger Publishing Company, 1997 Thomas, G., Goringe, M. J. – Transmission Electron Microscopy of Materials, John Wiley & Son, 1979 Valentini, L., Biagiotti, J., Kenny, J., Santucci, S. – ‘Morphological characterization of single walled carbon nanotubes-pp composites’, Composites Sci. Technol. 63 (2003) 1149–1153 Nan, C., Shi, Z., Lin, Y. – ‘A simple model for thermal conductivity of carbon nanotube based composites’, Chem. Phys. Lett. 375 (2003) 666–669.

© 2007, Woodhead Publishing Limited

190

Nanofibers and nanotechnology in textiles

75. Chambers, A., Park, C., Baker, R., Rodriguez, N. – ‘Hydrogen storage in graphite nanofibres’, J. Phys. Chem. B 102(22) (1998) 4253–4256 76. Dokhan, A., Price, E. W., Sigman, R. K., Seitman, J. M. – ‘The effect of the Al particle size on the burning rate and residual oxide in aluminized propellants’, Proceedings of 37th AIAA/ASME/SAE/ASEE Joint Propulsion Conference and Exhibition ‘Fundamentals of Solid Propellant Combustion’, AIAA–2001–3581, American Institute of Aeronautics and Astronautics (2001) 1–11 77. Désilets, S., Brousseau, P., Coté, S. – ‘Ignition of energetic materials containing carbon nanotubes’, 34th International Annual Conference of ICT, 24–27 June 2003 78. Ramaswamy, A., Kaste, P., Trevino, S. – Studies in Nanopropulsion for Environmentally Benign Propellants, University of MD, College Park, MD, Army Research Laboratory 79. Dokhan, A., Price, E., Sigman, R., Seitzman, J. – ‘The effects of al particle size on the burning rate and residual oxide in aluminized propellants’, AIAA 2001-3581 37th AIAA/ASME/SAE/ASEE Joint Propulsion Conference, 2001 80. Peigney, A., Laurent, C., Flahaut, E., Rousset, A. – ‘Carbon nanotubes in novel ceramic matrix nanocomposites’, Ceramics Int. 26 (2000) 677–683 81. Roche, S. – ‘Carbon nanotubes: exceptional mechanical and electronic properties’, Ann. Chim. Sci. Mat. 25, (2000) 529–532 82. Hammel, E., Tang, X., Trampert, M., Schmitt, T., Mauthner, K., Eder, A., Potschke, P. – ‘Carbon nanofibers for composite applications’, Carbon 42 (2004) 1153–1158 83. Sax, N. I., Feiner, B. – ‘Dangerous Properties of Industrial Materials’, Nostrand Reinold, 1984 84 Garg, A., Sinnott, S. – ‘Effect of chemical functionalization on the mechanical properties of carbon nanotubes’, Chem. Phys. Lett. 295 (1998) 273–278 85. Little, R. – ‘Mechanistic aspects of carbon nanotubes nucleation and growth’, J. Cluster Sci. 14 (2) (2003) 86. Shyu, Y., Hong, F. – ‘Low-temperature growth and field emission of aligned carbon nanotubes by chemical vapor deposition’, Mater. Chem. Phys. 72 (2001) 223–227 87. Lee, N., Park, C., Lee, S., Kang, J., Kim, C., Yun, M., Park, N., Han, I., Kim, J., Jung, J., Kim, J. – ‘New emitter techniques for field emitter displays’, Sixth Applied Diamonds Conference/Second Frontier Carbon Technology, Auburn University, July 2001, NASA/CP-2001-210948 88. Dean, K., Chalamala, B., Coll, B., Talin, A., Trujillo, J., Wei, Y., Jaskie, J. – ‘Fundamental properties of nanotube field emitters for large area electron sources’, Sixth Applied Diamonds Conference/Second Frontier Carbon Technology, Auburn University, July 2001, NASA/CP-2001-210948 89. Jeong, S., Hwang, H., Lee, K. – ‘Multiwall carbon nanotubes growth in nanotemplate and their application to a field emission device’, Sixth Applied Diamonds Conference/Second Frontier Carbon Technology, Auburn University, July 2001, NASA/CP-2001-210948 90. Wong, W., Au, F., Bello, L., Lee, C., Lee, S. – ‘Field emission from carbon nanotubes grown on plasma treated nickel/silicon substrate’, Sixth Applied Diamonds Conference/Second Frontier Carbon Technology, Auburn University, July 2001, NASA/CP-2001-210948 91. Fu, W., Xiao, B., Albin, S. – ‘Field emission properties of ultra fine multi wall carbon nanotubes grown by microwave plasma CVD’, Sixth Applied Diamonds Conference/Second Frontier Carbon Technology, Auburn University, July 2001, NASA/CP-2001-210948

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

191

92. Lau, K., Hui, D. – ‘The revolutionary creation of new advanced materials-carbon nanotubes composites’, Composite: Part B 33 (2002) 263–277 93. Thostenson, E. T., Ren, Z., Chou, T. – ‘Advances in the science and technology of carbon nanotubes and their composites’, Composites Sci. Technol. 61 (2001) 1899– 1912 94. Andrews, R., Rantell, T., Haddon, R., Dickey, E., Bergosh, R., Hu, H., Landis, C., Meier, M. – ‘Composite materials from modified carbon nanotubes’, Sixth Applied Diamonds Conference/Second Frontier Carbon Technology, Auburn University, July 2001, NASA/CP-2001-210948 95. Ruoff, R. – ‘Mechanics of carbon nanotubes’, Sixth Applied Diamonds Conference/ Second Frontier Carbon Technology, Auburn University, July 2001, NASA/CP2001-210948 96. Kimura, A., Mizutani, A., Toriyama, T., Oguri, K., Tonegawa, A., Nishi, Y. – ‘Fracture stress enhancement by EB treatment of carbon fibres’, Sixth Applied Diamonds Conference/Second Frontier Carbon Technology, Auburn University, July 2001, NASA/CP-2001-210948 97. Cagin, T., Che, J., Deng, W., Xu, X., Goddard, W. – ‘Computational studies on formation and properties of carbon nanotubes’, Sixth Applied Diamonds Conference/ Second Frontier Carbon Technology, Auburn University, July 2001, NASA/CP2001-210948 98. Kalamamkarov, A. – Composite and Reinforced Elements of Construction, John Wiley & Sons Ltd, 1992 99. Fortscue, P., Stark, J., Swinerd G. – Spacecraft Systems Engineering, John Wiley & Sons Ltd, third edition 2003 100. Gou, J., Miniaie, B., Wang, B., Liang, Z., Zhang, C. – Computational and Experimental Study of Interfacial Bonding of Single Walled Nanotubes Reinforced Composites, Department of Mechanical Engineering, University of South Alabama, USA 101. Cooper, C., Cohen, S., Barber, A., Wagner, H. – ‘Detachment of nanotubes from a polymeric matrix’, Appl. Phys. Lett. 81(20) (2002) 3873–3875 102. Lourie, O., Cox, D., Wagner, H. – ‘Buckling and collapse of embedded carbon nanotubes’, Phys. Rev. Lett. 81(8) (1998) 1638–1641 103. Wei, C., Cho, K. – ‘Chemical bonding of polymer on carbon nanotube’, MRS 2001 Spring Meeting Proceeding Paper 104. Barber, A., Cohen, S., Wagner, H. – ‘Measurement of carbon nanotubes polymer interfacial strength’, Appl. Phys. Lett. 82(23) (2003) 4140–4142 105. Ruoff, R., Qian, D., Liu, W. – ‘Mechanical properties of carbon nanotubes: theoretical prediction and experimental measurements’, C. R. Physique 4 (2003) 993–1008 106. Salvetat, J., Bonard, J., Thomson, N., Kulik, A., Forrò, L., Beniot, W., Zuppiroli, L. – ‘Mechanical properties of carbon nanotubes’, Appl. Phys. A 69 (1999) 255– 260 107. Cooper, C., Ravich, D., Lips, D., Mayer, J., Wagner, H. – ‘Distribution and alignment of carbon nanotubes and nanofibrils in a polymer matrix’, Composites Sci. Technol. 62 (2002) 1105–1112 108. Alloui, A., Bai, S., Cheng, H., Bai, J. – ‘Mechanical and electrical properties of a MWNT/epoxy composite’, Composites Sci. Technol. 62 (2002) 1993–1998 109. Ruoff, R., Lorents, D. – ‘Mechanical and thermal properties of carbon nanotubes’, Carbon 33 (7) (1995) 925–930

© 2007, Woodhead Publishing Limited

192

Nanofibers and nanotechnology in textiles

110. Frankland, S., Harik, V., Odegard, G., Brenner, D., Gates, T. – ‘The stress–strain behaviour of polymer nanotube composites from molecular dynamics simulation’, Composites Sci. Technol. 63 (2003) 1655–1661 111. Valentini, L., Biagiotti, J., Kenny, J., Santucci, S. – ‘Morphological characterization of single walled carbon nanotubes-PP composites’, Composites Sci. Technol. 63 (2003) 1149–1153 112. Lau, K., Hui, D. – ‘The revolutionary creation of new advanced materials–carbon nanotube composites’, Composites: Part B 33 (2002) 263–277 113. Wagner, H., Lourie, O., Zhou, X. – ‘Macrofragmentation and microfragmentation phenomena in composite materials’, Composites: Part A 30 (1999) 59–66 114. Xiao, Q., Zhou, X. – ‘The study of multi walled carbon nanotube deposited with conducting polymer for super capacitor’, Electrochimica Acta 48 (2003) 575–580 115. Peigney, A., Laurent, C., Flahaut, E., Rousset, A. – ‘Carbon nanotubes in novel ceramic matrix nanocomposites’, Ceramics Int. 26 (2000) 677–683. 116. Koratkar, N., Wei, B., Ajayan, P. – ‘Multifunctional structural reinforcement featuring carbon nanotube film’, Composites Sci. Technol. 63 (2003) 1525–1531 117. Binnott, S., Shenderova, O., White, C., Brenner, D. – ‘Mechanical properties of nanotubule fibres and composites determined from theoretical calculations and simulations’, PII: S0008-6223(97)00144-9 118. Stephan, C., Nguyen, T., Chapelle, M., Lefrant, S., Journet, C., Bernier, P. – ‘Characterization of single walled carbon nanotubes-PMMA composites,’ Synthetic Metals 108 (2000) 139–149 119. Garg, A., Sinnott, S. – ‘Effect of chemical functionalization on the mechanical properties of carbon nanotubes’, Chem. Phys. Lett. 295 (1998) 273–278 120. Andler, J., Shaffer, M., Prasse, T., Bauhofer, W., Schulte, K., Windle, A. – ‘Development of dispersion process for carbon nanotubes in a epoxy matrix and the resulting electrical properties’, Polymer 40 (1999) 5967–5971 121. Currao, S., Davey, A., Coleman, J., Dalton, A., McCarthy, B., Maier, S., Drury, A., Gray, D., Brennan, M., Ryder, K., de la Chapelle, M., Journet, C., Bernier, P., Byrne, H., Carroll, D., Ajayan, P., Lefrant, S., Blau W. – ‘Evolution and evaluation of polymer/nanotube composite’, Synthetic Metals 103 (1999) 2559–2562 122. Odegard, G., Gates, T., Nicholson, L., Wise, K. – ‘Equivalent continuum modelling of nano structured materials’, Composites Sci. Technol. 62 (2002) 1869–1880 123. Hammel, E., Tang, X., Trampert, M., Schmitt, T., Mauthner, K., Eder, A., Potschke, P. – ‘Carbon nanofibers for composite applications’, Carbon 42 (2004) 1153–1158 124. Vasiliev, V., Barynin, V., Rasin, A. – ‘Anisogrid lattice structures – survey of development and application’, Composite Structures 54 (2001) 361–370 125. Slinchenko, D., Verijenko, V. – ‘Structural analysis of composite lattice shells of revolution on the basis of smearing stiffness’, Composite Structures 54 (2001) 341–348 126. Vasiliev, V. V., Jones, R. M. – Mechanics of Composite Structures, Taylor & Francis, 1993 127. Vasiliev, V. V., Morozov, E. V. – Mechanics and Analysis of Composite Materials, Elsevier Science Ltd, 2001 128. Kyriazidou, C., Diaz, R., Alexopoulos, N. – ‘Novel material with narrow band transparency window in the bulk’, IEEE Trans. Antenna Propagation 48(1) (2000) 107–116. 129. Campos, F., D’Assuncao, A., Neto, G. – ‘Scattering characteristics of FSS on two anisotropic layers for incident co-polarized plane waves’, Microwave Optical Technol. Lett. 33(1) (2002) 57–61

© 2007, Woodhead Publishing Limited

Carbon nanotubes in aerospace engineering

193

130. Lu, C., Chern, R., Chang, C. – Analysis of Frequency Selective Surfaces by Spectral Galerkin Method, Institute of Applied Mechanics, National Taiwan University 131. Cui, S., Weile, D. – Efficient Analysis of Scattering for Periodic Structures Composed of Arbitrary Inhomogeneous and Anisotropic Materials using FE-BI Method Accelerated by FFT, Dept. of Electrical & Computer Engineering, University of Delaware, Newark, DE 132. Chakravarty, S., Mittra, R. – ‘Application of micro genetic algorithm to the design of spatial filters with frequency selective surfaces embedded in dielectric media’, IEEE Trans. Electromagnetic Compatibility 44(2) (2002) 338–346. 133. Chakravarty, S., Mittra, R. – ‘Design of a frequency selective surface (FSS) with very low cross-polarization discrimination via parallel micro genetic algorithm’, IEEE Trans. Antennas Propagation 51(7) (2003) 1664–1668 134. Lynch, J., Colburn, J. – ‘Modelling polarization mode coupling in frequency selective surfaces’, IEEE Trans. Microwave Theory Techniques 52(4) (2004) 1328–1338 135. Kristensson, G. – On the Generation of Surface Waves in Frequency Selective Structures, Department of Electroscience Electromagnetic Theory, Lund Institute of Technologies, Sweden, 2003 136. Li, Z., Papalambros, P., Volakis, J. – ‘Frequency selective surface design by integrating optimisation algorithms with fast full wave numerical methods’, IEE Proc. Microwave Antennas Propagation 149(3) (2002) 175–180 137. Munk, B. – Frequency Selective Surfaces, John Wiley & Sons Inc., 2000 138. Bhushan, B. – Handbook of Nanotechnology, Springer, 2004

© 2007, Woodhead Publishing Limited

7 Carbon nanotube and nanofibre reinforced polymer fibres M. S. P. S H A F F E R, Imperial College London, UK and J. K. W. S A N D L E R, University of Bayreuth, Germany

7.1

Introduction

Carbon nanotubes (CNTs) can be seen as a bridge between traditional carbon fibres and the fullerene family;1 this intermediate position between the molecular and continuum domains is the classic signature of a nanomaterial. Research on these structures blossomed only recently, following the electricarc synthesis of multiwalled nanotubes by Iijima, in 1991;2 since then, in excess of 10 000 papers have appeared discussing the science of CNTs, including a large fraction on polymer composites. This interest was initially stimulated by the recognition of the relationship with the closed, curved, carbon shells of the fullerene family that had been discovered a few years previously, in 1985.3 However, although Iijima is often credited with the discovery of CNTs, there are earlier reports in the literature, notably by Endo in 1976, of the synthesis of tubular carbon structures using hydrocarbon decomposition,4 as well as earlier in the catalysis literature of the 1950s, and possibly even the late 19th century.5 In fact, nanotubes are now known to occur naturally, having been observed in 10 000-year-old ice cores6 and metallic swords.7 The investigation of nanotubes today is driven by their elegant and diverse structures, and their emerging remarkable intrinsic properties. CNTs have typical diameters in the range of ~1–50 nm and lengths of many micrometres (even centimetres in special cases).8 They can consist of one or more concentric graphitic cylinders. In contrast, commercial (polyacrylonitrile (PAN) and pitch) carbon fibres are typically in the 7–20 µm diameter range, while vapour-grown carbon fibres (VGCFs) have a broad range of possible diameters (see Fig. 7.1). Compared with carbon fibres, the best nanotubes can have almost atomistically perfect structures; indeed, there is a general question as to whether the smallest CNTs should be regarded as very small fibres or heavy molecules, especially as the diameters of the smallest nanotubes are similar to those of common polymer molecules. 194

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

195

Diameter 10 µm

Carbon fibres Graphite whiskers

1µm Vapour-grown carbon fibres 0.1µm

Carbon nanofibres

10 nm Carbon nanotubes 1nm

Fullerenes (C60, C70, etc.)

7.1 Comparison of diameters of various fibrous carbon-based materials.

Consequently, it is not yet clear to what extent conventional fibre composite understanding can be extended to CNT composites. Structurally, CNTs are very diverse depending on their origins. Fundamentally, a single wall carbon nanotube (SWCNT) consists of a single layer of graphite rolled into a seamless cylinder; the orientation of the graphite lattice to the cylinder axis defines the chirality or helicity of the nanotube.9 Nanotube shells have large surface areas and prefer to gain van der Waals stabilisation10 either by forming parallel bundles of SWCNTs11–14 or by nesting concentrically to form a multiwalled carbon nanotube (MWCNT). The outer diameter of such MWCNTs can vary between 2 and a somewhat arbitrary upper limit of about 50 nm; the inner diameter is often (though not necessarily) quite large, about half that of the whole tube. As-grown, each nanotube is closed by a hemispherical, fullerenic cap associated with pentagonal rather than hexagonal rings in the graphitic structure. In addition, a wide range of defects can exist including vacancies, extraneous non-hexagonal rings, edge dislocations, local sp3 hybridisation, and non-carbon functional group defects which can give rise to longer range morphological changes in the structure, such as kinks/bends, and changes in diameter. The formation of kinks during synthesis is particularly significant as it encourages the development of an entangled network of nanotubes that is difficult to disperse.15 Straight nanotubes, on the other hand, are less likely to be entangled, can be aligned more easily, and are likely to have better performance in composites.16 Figure 7.2 shows a comparison of aligned and entangled MWCNTs. In addition to variations in the structure of the nanotubes, other contaminating materials can be present, including, for example, amorphous carbon, graphitic nanoparticles and catalyst metals, which can be difficult to remove.

© 2007, Woodhead Publishing Limited

196

Nanofibers and nanotechnology in textiles

(a)

100 nm (b)

7.2 Scanning electron micrographs of (a) aligned and (b) commercial entangled MWCNTs produced by chemical vapour deposition (CVD) methods.

Carbon nanofibres (CNFs) are mainly differentiated from nanotubes by the orientation of the graphene planes: whereas the graphitic layers are parallel to the axis in nanotubes, nanofibres can show a wide range of orientations of the graphitic layers with respect to the fibre axis. They can be visualised as

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

197

stacked graphitic discs or (truncated) cones, and are intrinsically less perfect as they have graphitic edge terminations on their surface. Nevertheless, these nanostructures can be in the form of hollow tubes with an outer diameter as small as ~5 nm, although 50–100 nm is more typical. The stacked cone geometry is often called a ‘herringbone fibre’ due to the appearance of the longitudinal cross-section. Slightly larger (100–200 nm) fibres are also often called CNFs, even if the graphitic orientation is approximately parallel to the axis. An example of the complicated structure of a commercial CNF material is shown in Fig. 7.3.

7.2

Synthesis and properties of carbon nanotubes

Both MWCNTs and SWCNTs can be produced by a variety of different processes which can broadly be divided into two categories: high-temperature evaporation using arc-discharge11, 12, 17–19 or laser ablation,13, 20 and various chemical vapour deposition (CVD) or catalytic growth processes.14, 21–23 In the high-temperature methods, MWCNTs can be produced from the evaporation of pure carbon, but the synthesis of SWCNTs requires the presence of a metallic catalyst. The CVD approach requires a catalyst for both types of CNTs but also allows the production of CNFs. The products of the hightemperature routes tend to be highly crystalline, with low defect concentrations, but are relatively impure, containing other, unwanted carbonaceous impurities. These methods usually work on the gram scale and are, therefore, relatively expensive. For the use of nanotubes in composites, large quantities of nanotubes are required at low cost, ideally without the requirement for complicated +/-15° Parallel segments 5 nm

250 nm

Inner core

7.3 Representative transmission electron micrographs of commercial carbon nanofibres, highlighting structural variations both in overall morphology and in the orientation of the graphitic planes. The leftmost image shows a ‘bamboo’ and a ‘cylindrical’ CNF, while the rightmost image shows a high magnification image of one wall of the cylindrical fibre which reveals the graphitic arrangement sketched in the central panel.

© 2007, Woodhead Publishing Limited

198

Nanofibers and nanotechnology in textiles

purification. At present, only CVD-grown nanotubes satisfy these requirements and, as such, tend to be the materials of choice for composite work, both in academia and in industry;24 a number of companies have scaled up such processes to 100 tonnes per year or more. The main contaminants in CVD materials are residual catalyst particles which are mostly incorporated into the nanotubes. On the other hand, these gas-phase processes operate at lower temperatures and lead to structurally more imperfect nanotubes, as shown by the deviation from the ideal cylindrical structure in Fig. 7.4.

7.2.1

Mechanical properties

The interest in carbon nanotubes, particularly their application in composites, has been driven by their remarkable intrinsic properties; however, these properties depend critically on the structural characteristics mentioned above, with crystalline quality and orientation being especially important. The fundamental mechanical properties of nanotubes are quite difficult to determine, but a number of attempts have been made, based on transmission electron microscopy (TEM) studies of thermal vibrations,25, 26 bending measurements using atomic force microscopy (AFM),27, 28 direct nano-tensile tests,29–31 fragmentation/deformation under load in a composite,32–34 and experiments on macroscopic aligned bundles,35, 36 as well as various computational approaches.37–41 The errors in such measurements are often large because of sample variability and the challenging experimental environment; in addition, the calculations generally assume a classical mechanical behaviour, and usually use the van der Waals thickness of the graphitic sheet(s) to calculate the appropriate normalising cross-sectional area (in fact, the presence of the hollow core reduces the real effective modulus, which, in any case, is hard

4 nm

2.5 nm

7.4 Transmission electron micrographs of commercial MWCNTs grown by CVD methods, with the beam perpendicular (left) and parallel (right) to the axis.

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

199

to define in bending).42 Despite these difficulties, there is a reasonably clear picture emerging that highly crystalline SWCNTs or individual shells of MWCNTs, grown at high temperatures, have axial stiffnesses similar to the in-plane properties of graphite (1.1 TPa), and strengths of around 50 GPa, implying high strains to failure (~10%). Under bending deformation, nanotubes are highly resilient, undergoing reversible kinking.43–46 These values match or exceed all other materials (specific properties are even more impressive given the low density) and have stimulated significant efforts to exploit the performance at the macroscopic scale,47 as discussed in a later section. One problem is that CVD-grown material, which is otherwise highly suitable for composites applications, has a high defect concentration. These defects dramatically degrade the intrinsic properties of the nanotubes. Although point defects can be significant,48 deviations from a perfectly parallel alignment of the graphitic layers to the axis is particularly detrimental owing to the high anisotropy of graphite. Reductions in strength and stiffness of up to one to two orders of magnitude have been measured,49–51 with typical values around 2 GPa, and 50–100 GPa, respectively. The high anisotropy of graphite has a further consequence; namely, that individual MWCNTs and SWCNT bundles can suffer internal failures due to the low shear strength parallel to the graphene layers. The result is that only the outer layer(s) of an MWCNT or SWCNT bundle actually carries the tensile load when embedded in a polymer matrix. Whether end effects, high aspect ratios or modest defect concentrations can alleviate this problem remains to be seen. Shear failures have been observed in nano-tensile tests29–31 and scanning tunneling microscope (STM) observations52 of neat nanotubes and in some composite experiments.33, 53 For these reasons, some have suggested that individually dispersed SWCNTs should be the ideal reinforcement; however, other problems with flexibility, processing and large interfacial areas may be worse.

7.2.2

Transport properties

Theoretical studies of the electronic properties of SWCNTs indicate that nanotube shells are one-dimensional conductors with characteristic Van Hove singularities in their density of states. Depending critically on helicity, they can be either metallic or semiconducting,54–56 with a small or moderate band gap (for semiconducting tubes) inversely proportional to the tube radius.57, 58 On average, approximately one-third of SWCNTs are metallic and twothirds semiconductors.54 Since MWCNTs have larger diameters, confinement effects disappear, and the transport properties approach those of turbostratic graphite.59 Interlayer interactions which might be important in small diameter MWCNTs are thought to be weak.60, 61 However, structural defects as well as bends or twists again have a strong effect on the transport properties.62

© 2007, Woodhead Publishing Limited

200

Nanofibers and nanotechnology in textiles

Experimental electronic transport measurements on individual CNTs using bottom63–65 or top66 contacts, or scanning tunnelling spectroscopy (STS),67, 68 have broadly confirmed these theoretical predictions. The predicted Van Hove singularities have been observed in STS experiments and a wide range of visible/near IR, fluorescence and Raman spectroscopy studies.69, 70 Typical room temperature conductivity is in the range 105–106 S/m and 10 S/m for metallic and semiconducting nanotubes, respectively. The conductivity of SWCNT bundles, which is influenced by the significant semiconductor content,71 has been found to be between 1 × 104 and 3 × 106 S/m72, 73, 74 at room temperature, depending on sample type. Metallic nanotubes thus approach the in-plane conductivity of graphite (2.5 × 106 S/m).75 Conductivities of individual MWCNTs have been reported to range66 between 20 and 2 × 107 S/m, depending on the helicities of the outermost shells76 or the presence of defects.77 The electronic properties of larger diameter MWCNTs approach those of graphite. Smaller MWCNTs exhibit ballistic conductivity over micrometre lengths, with scattering occurring only at the contacts.78 Lastly, the axial thermal conductivity of individual, perfect CNTs is expected to be very high,79 greater than that of diamond. Experimental values for individual MWCNTs have reached 3300 W/m K,80 and aligned SWCNT arrays have been shown to be highly anisotropic.81

7.2.3

Physical properties

Carbon nanotubes burn in air at temperatures in the range of 450–750 °C depending on the crystallinity, size, purity and surface chemistry of the sample. In inert atmospheres, MWCNTs are essentially stable up to 3000 °C or more, although graphitisation may occur;82 SWCNTs begin to reform into MWCNTs from around 1500 °C. The density of CNTs relates to that of graphite (2200 kg/m3), but is effectively reduced, as long as the central hole remains empty, by an amount that depends on the ratio of the internal and external diameters. Typical values for MWCNTs and CNFs are in the range of 1600–2000 kg/m3. The density of SWCNTs varies systematically with diameter in the approximate range 1300–1600 kg/m3 and has been used to separate them by centrifugation.83 The surface areas generally follow the classical geometric expectation, although it can be increased by surface activation of MWCNTs.84 Typical values for MWCNTs are hundreds of m2/g, with closed SWCNTs reaching 1340 m2/g in theory, although the experimental value is generally reduced by bundling.85 The value for SWCNTs is particularly high as every atom lies on the surface; if the inner surface is accessible to the probe, the total area is even higher, with every atom present on two surfaces.

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

7.3

201

Developing nanotube/nanofibre–polymer composites

There are at least three general experimental methods to produce bulk polymer nanocomposites: mixing in the liquid state, solution-mediated processes and in-situ polymerisation techniques. The direct melt-blending approach is much more commercially attractive than the latter two methods, as both solvent processing and in-situ polymerisation are less versatile and more environmentally contentious. Although various mixing methodologies are commonly used, final sample fabrication usually is by injection-moulding or hot pressing. The literature on processing and evaluating macroscopic nanotube/ nanofibre–polymer composites is still in its infancy but developing rapidly. This situation is not surprising, given that initial attempts to produce such nanocomposites were hindered by the small quantities of nanotubes available. The focus on CVD synthesis techniques has opened the door to the manufacture of large-scale polymer nanocomposites.

7.3.1

Nanotube dispersion

A high-quality nanotube dispersion in any polymer matrix is a crucial prerequisite for good mechanical composite performance, and is often difficult to achieve. Each nanotube should be loaded separately and equally; if the filler is agglomerated, some nanotubes will be shielded from the mechanical load. In addition, the agglomerates will act as defects, leading to stress concentration and premature failure. Good dispersion becomes harder to achieve as the particle size shrinks into the nanoscale; as surface areas increase, particles become more mobile, the distances between them decrease and shear forces become harder to apply. Nanoparticles have a strong tendency to agglomerate, and the high aspect ratio of nanotubes only makes matters worse as they sweep out large hydrodynamic volumes and can easily become entangled (see the rheology discussion in Section 7.5). High loading fractions favour agglomeration not only because the particles come into contact more often, but also because there can be a shortage of polymer matrix to ‘wet out’ the surface of the filler. The problem of debundling SWCNTs is particularly acute, as the van der Waals forces between the deformable nanotubes are strong, and the amount of polymer required to wet the surface is large. A trivial estimate suggests that in a 1 vol% composite containing individual SWCNTs, all the polymer chains are touching a nanotube surface (assuming a 5 nm radius of gyration); it is then easy to understand why it might be hard to add additional SWCNTs. Indeed, it is quite a common result for nanocomposites in general that properties increase at low loading fractions but cannot be increased further due to agglomeration above a few volume

© 2007, Woodhead Publishing Limited

202

Nanofibers and nanotechnology in textiles

percent. The situation is more ambiguous when addressing transport properties, especially electrical conductivity, as a network of touching nanotubes is desired. However, even in this case, best results may be obtained by generating a good dispersion initially, and then allowing the network to form.16 A particular practical problem is that dispersion is very hard to quantify. In fact, no good objective measures are available. Characterisation typically consists of a qualitative assessment of a freeze–fracture surface studied under an SEM. This approach is quite successful for discovering dense aggregates (typical of CNTs synthesised in the electric arc) or looser agglomerates in low volume fraction systems. However, it is less useful at high loading fractions where the filler is necessarily densely packed; here it is hard to distinguish between closely packed but isolated CNTs and an agglomerated network, since any contacts may not lie in the fracture plane. Optical microscopy or even optical clarity is another useful guide, again chiefly for low loading fractions. Since nanotubes are usually at least micrometres long, agglomerates are generally visible. Optical clarity is, therefore, a necessary though not sufficient condition for well-isolated CNTs, although at higher loadings, samples simply become very dark. Rheological measurements can also provide an indication of dispersion, particularly as data can be collected during the dispersion process; as discussed further in Section 7.5. The primary method of dispersion is usually physical. Substantial shear forces appear to be necessary during the initial composite processing steps, in order to disperse either CNTs or CNFs in the (pre)polymer melt or polymer solution, especially when using the filler in the as-produced state or at highvolume fractions. For high-viscosity systems, particularly thermoplastic melts, twin-screw extrusion has been frequently applied and found to be effective for up to 60 wt% CNFs86 and about 30 wt% MWCNTs.87 The high intrinsic viscosity of thermoplastic matrices in general has the dual advantage of increasing the shear applied to the aggregates (even breaking the individual CNTs/CNFs) and minimising the opportunity for reaggregation. In the case of nanotubes especially, the degree of dispersion depends strongly on the entanglement state of the as-received material, since even extensive twinscrew extrusion does not lead to a complete break-up of the entanglements in commercially available catalytically grown MWCNTs,88–90 although some materials seem to be more easily processed,87–91 particularly those based on aligned CNTs.92 CNFs tend to break more easily during processing, leading to the removal of entanglements;93 SWCNT bundles are not generally broken up by thermoplastic processing. Other high-shear processes can be applied, for example, the use of ball milling of the raw filler material prior to processing;94 however, this approach degrades the aspect ratio of individual particles significantly more than shear-intensive melt processing.95 Solvent processing is popular as an alternative to melt processing, or as a preliminary step. Dispersion in low-viscosity solutions is typically achieved

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

203

using ultrasound; the high intensity used separates CNT aggregates and even SWCNT bundles, but can also cause damage.96 Reagglomeration may be prevented by the presence of the polymer in solution which can adsorb onto the exposed CNT surfaces or even become grafted due to chain scission.97 Amphiphilic polymers dissolved in water,98 such as poly(hydroxyaminoether) (PHAE), 99, 100 poly(vinyl alcohol) (PVA),101, 102 and PVA/poly(vinyl pyrrolidone) (PVP)103 have proved particularly effective, although organic systems have also been explored, based on polystyrene (PS),104–106 ultrahigh molecular weight polyethylene (UHMWPE)107 and polypropylene (PP).108 In addition, surfactants,103, 109 polymer-functionalised nanotubes110, 111 and other chemical treatments of the constituents112–114 are often employed. Some of the most successful nanotube–polymer composites have been created by in-situ polymerisation, such that the nanotubes become grafted to or ensnared in the growing polymer.115–119 In all low-viscosity systems, the individual nanotubes should form an inherently electrostatically or sterically stabilised dispersion, with a long lifetime relative to the subsequent processing. Once the solvent is removed, thermoplastic processing can be applied without necessarily losing the good dispersion.87, 91, 95, 114, 120, 121 The chemical methods of providing stabilisation, particularly those that encourage a strong, or even covalent interaction with the polymer, tend to have a positive impact on subsequent composite properties through improved load transfer.

7.3.2

Mechanical properties of nanotube/nanofibre– polymer composites

Although many nanotube/nanofibre composite systems have been prepared, the mechanical enhancements have been significant but somewhat limited compared with theoretical predictions. The main challenges lie in obtaining a good dispersion, optimising the interface between polymer and filler, and obtaining high-quality structures in sufficient quantities (as discussed above, CVD-grown nanotubes are used commonly in composites but have relatively poor properties). As in any fibre reinforced composite system, the interfacial shear strength is an important parameter, although difficult to assess. A growing body of computational work122–124 as well as some initial experimental work33, 125–127 has addressed this issue. Both approaches have shown significant spread in the data, yet the overlap seems to suggest values in the range of 50–100 MPa for non-covalently bonded composites47 although much higher values are anticipated for covalently bonded nanotubes.122 The published mechanical data show that the tensile modulus of nanotube/ nanofibre–thermoplastic composites is generally improved, although a detailed comparison of the data is difficult due to the different types of fillers, surface treatments, matrices, processing techniques and test methods that have been used. In general, the stiffening effect of nanotubes and nanofibres appears to

© 2007, Woodhead Publishing Limited

204

Nanofibers and nanotechnology in textiles

be more prominent in semicrystalline rather than amorphous thermoplastics, possibly due to nucleation effects commonly observed in such nanocomposites.108, 128–130 Even when a homogeneous dispersion of the nanoscale filler is claimed for all concentrations, the stiffness enhancement is usually most prominent for low filler weight fractions, with the critical concentration depending on the specific materials and processing conditions used. This observation might relate to decreasing nanofiller alignment or increasing void content with increasing weight fraction,95, 106, 131 although alignment variations of the polymer matrix often have not properly been taken into account. In semicrystalline matrices the (often unanalysed) variations in crystallinity can also be a source of non-linearity. Furthermore, in most cases, there are likely to be changes in dispersion, as the larger surface areas associated with high loading fractions become increasingly difficult to accommodate within the polymer; similar effects have been seen in nanoclay-filled polymers.132 However, even in the presence of nanotube clusters, enhancements in composite stiffness can be observed.133 A concise overview of reported mechanical nanocomposite properties is presented by Coleman et al.47 Composites based on chemically-treated nanotubes show the best results on average. This conclusion is not surprising, given that nanotube functionalisation should significantly improve both the dispersion as well as the stress transfer. In comparison to solution-based composites, the average reinforcement effect seen in melt-processed systems is somewhat disappointing, most likely reflecting alignment issues in bulk samples. Yet, the best individual results reported134, 135 have shown stiffness improvements approaching the maximum predicted by conventional composite theories, while strength improvements are generally low. Lastly, it is interesting to note that MWCNTs appear to outperform SWCNTs in many cases, most likely reflecting the difficulties in obtaining straight, debundled SWCNTs. Compared to stiffening, enhancements in composite yield stress, strength and toughness generally appear more difficult to achieve, especially for filler loading fractions exceeding about 10 wt%. These properties depend more on the homogeneity of specimens achieved during processing, as well as on interfacial issues relating to the specific filler types and matrices. For example, the impact properties of nanofibre– polycarbonate (PC) composites are significantly decreased, even at low nanofibre contents, most likely as a result of aromatic hydrocarbons on the nanofibre surface enhancing chemical stress cracking of the polycarbonate.136 The most prominent strength enhancements of bulk nanocomposites have been achieved for well-dispersed and aligned nanofibres in poly(ether ether ketone) (PEEK),137 up to filler loading fractions of 10 vol%, and in PP,95 at nanofibre loading fractions as low as 5 vol%. CNTs and CNFs are also interesting additives for tribological applications;138, 139 they can significantly reduce the wear rate of polymers,

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

205

apparently without much dependence on the degree of dispersion achieved during compounding. Benefits are obtained even when compounded with established tribological aids, but are potentially accessible in components (such as fibres) in which conventional additives cannot be incorporated.

7.3.3

Physical properties of nanotube/nanofibre–polymer composites

As well as mechanical reinforcement, there is considerable interest in the exploitation of other physical properties of nanotubes, particularly the high thermal and electrical conductivities. Electrically conductive polymer composites, for example, are used in antistatic packaging applications, as well as in highly specialised components in the electronics, automotive and aerospace sectors. The incorporation of conductive filler particles into an insulating polymer host can lead to sufficient bulk conductivities to exceed the antistatic limit of 10–6 S/m. Common conductive fillers are metallic or graphitic particles in any shape (spherical, platelet-like or fibrous) and size. However, the incorporation of CNTs allows for a low percolation threshold, a high-quality surface finish, a robust network and good mechanical properties – a combination not obtained with any other filler. The use of CNTs/CNFs as a conductive filler is their biggest current application, and is widespread across the automotive and electronic sectors. The electrical properties of nanofibre–thermoplastic composites exhibit a characteristic percolation behaviour,140–142 with a rapid increase in conductivity as the loading fraction passes through the critical threshold. In the case of untreated CNFs, the threshold is around 5–10 vol% and depends on aspect ratio, surface chemistry, dispersion and alignment. Similarly, the electrical percolation threshold of thin MWCNT–thermoplastic films also depends on the type of nanotube and the processing route. Threshold values range from around 5 wt% for oxidised catalytic MWCNTs in PVA101 to around 0.06 wt% and 0.5 wt% for arc-discharge MWCNTs in PVA143 and PMMA144, respectively. In bulk, thermosetting systems,145–150 CNT-based composites tend to have higher conductivities and lower percolation thresholds than either carbon black or CNF-based ones. Indeed, a CNT-based epoxy system currently shows the lowest percolation threshold observed in any system, at around 0.0025 wt%.146 It is tempting to attribute this and other low thresholds simply to the high aspect ratio of the conductive filler. However, these low values can be explained only in the light of a complicated dispersion and reaggregation behaviour during processing; in essence, well-dispersed nanotubes are destabilised and trapped just as a network forms.151 This network formation behaviour can be manipulated not only by temperature and shear rate, but also by the application of external electrical fields.152 Interestingly, low loading fractions of nanotubes in thin

© 2007, Woodhead Publishing Limited

206

Nanofibers and nanotechnology in textiles

films may allow a useful level of conductivity while retaining optical transparency.118 Effects to increase thermal conductivity have been less successful. Increases have been modest and linear; percolation does not play a major role due to the less dramatic difference in conductivity between filler and matrix.95, 131, 141 Results are similar to those observed for short carbon fibres in similar systems.141, 153

7.4

Adding nanotubes and nanofibres to polymer fibres

The fundamental motivation for adding nanotubes to polymer fibres is similar to those outlined above for other composite systems: namely, that there is an opportunity to improve the mechanical and functional properties of the matrix by drawing on the unique properties of the nanotubes. However, there are some additional attractions. At the current stage of development, bulk nanotube composites appear to offer only comparable properties to systems reinforced with conventional chopped fibres. However, unlike the nanotubes, such conventional fillers cannot be accommodated within fine polymer fibres. Thus, CNTs/CNFs offer a new route to nano-reinforced polymer fibres, and can provide unique improvements in performance. These improvements can be significant even at the current level of development of nanotube composites, using existing commercial CVD materials, because no other established filler can compete in the confined environment of the polymer fibre. In addition to acting as a reinforcing agent, such nanoscale fillers can also act as critical processing aids by modifying the polymer rheology; in particular, the elongational flow properties that are relevant at the high extension rates typically encountered during fibre-spinning operations.154 Last but not least, initial work on such fibre systems requires only small quantities of nanotubes, allowing a wide range of scarce, experimental nanotube materials or modifications to be tested in composite form.92 Polymer fibre systems thus also provide a useful test-bed for developing our understanding of nanotube composites in general and moving towards the full exploitation of their fundamental properties. This type of basic development is additionally helped by the fact that the nanotubes tend to become aligned to the fibre axis during processing, naturally providing the optimal orientation. Perhaps most significant of all, this uniaxial orientation is almost certainly the most plausible arrangement for fully exploiting the anisotropic, intrinsic properties of the nanotubes. The aligned arrangement not only provides the optimal loading geometry, but also maximises the volume content of nanotubes, since rods pack more efficiently when aligned;155 the ultimate (possibly unobtainable) goal of many researchers is a perfectly aligned infinite crystal of nanotubes.156

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

207

There are, therefore, two strands to nanotube–polymer fibre composites. On the one hand, there is interest in improving the properties of existing polymer fibre systems either by improving critical properties of highperformance fibres (see, for example, the discussion of PEEK below), or by upgrading the performance of commodity polymers to an ‘engineering’ standard (e.g. PP). On the other hand, there are ongoing attempts to generate fibres with an absolutely higher level of mechanical performance than any existing system, based either on improving the current top performers (such as poly(pphenylenebenzobisoxazole), PBO) or by developing new, high CNT content fibres. Finally, there are interesting attempts to exploit the potential of nanotubes to produce functional fibres that can, for example, act as electronic devices. Processing is, of course, crucially important in determining nanotube dispersion and orientation, as well as the more traditional but equally important factors associated with polymer morphology. Polymer nanocomposite fibres may be spun from solution or from the melt, often following an initial dispersion step, as discussed in Section 7.3.1 above. Much of the initial work has been exploratory in nature, and there remains considerable scope for applying the wider understanding of fibre spinning to these systems.

7.4.1

Solution spinning

Solution spinning is a widespread and attractive route for the production of polymer fibres, and has been applied to a variety of CNT systems. A particularly interesting possibility is the use of a lyotropic nematic nanotube solution as a route to a highly aligned fibre. Much of this work is directed at high loadings of nanotubes and examples include the use of surfactant-stabilised dispersions of CNTs injected into a PVA157, 158 or poly(ether imide) (PEI)159 bath, to form a fibre that can be handled and drawn, and the use of pure SWCNT dispersions in ‘superacid’.160, 161 A number of other, lower-volume fraction systems have also been spun from nanotube/polymer solutions, although the ‘spinning’ process is often a relatively rudimentary small-scale demonstration. Solvent routes are particularly attractive for SWCNTs as pure melt-mixing does not lead to adequate exfoliation of the bundles; matrix/solvent (maximum CNT content) examples include polyaniline/dichloroacetic acid (0.3%),162 chitosan/water (4.8%),163 polyacrylonitrile/N, N-dimethylformamide (DMF) (1%)164 and PVOH/DMSO/ H2O (polyvinyl alcohol/dimethylsulphoxide/water) (3%),165 although the highest content system163 was poorly dispersed. An interesting variant is based on the use of caprolactam as both solvent and monomer for an in-situ ring opening polymerisation of nylon-6 to produce nanocomposite fibres (1.5%).166 More straightforwardly, the in situ polymerisation can be carried out in the presence of the solvent before spinning, as in the case of SWCNTs added to the rigid rod polymer PBO.116 Lastly, both MWCNTs and CNFs

© 2007, Woodhead Publishing Limited

208

Nanofibers and nanotechnology in textiles

have also been processed into polyetherketone (PEK) fibres following an in situ polymer grafting reaction (10%).167 It is also possible to produce nano-reinforced polymer fibres by electrospinning nanotube polymer solutions;168 however, this topic will not be discussed further as it is covered at length elsewhere in this book (see Chapters 1–5)

7.4.2

Melt spinning

The majority of nanotube/nanofibre-filled thermoplastic fibres are made by variants of melt spinning, or approximations based on sample collection from rheometers and the like, representing an optimisation of previous alignment approaches based on die designs95 and simple melt strand or film stretching.91 While conventional melt processing of the nanocomposite dope is desirable for economical reasons, initial solution-blending is often performed in the case of SWCNT-filled polymers,120, 121 yet the effectiveness of this approach remains debatable. Although the amount of nanocomposite material is often limited, the principal applicability of traditional fibre processing technology for the manufacture of nanotube/nanofibre-filled systems has been demonstrated. As such, most studies have focused on low to medium draw ratios and a whole range of fibre diameters ranging from a few to hundreds of micrometres. In addition, experimental efforts aimed at maximising alignment of both filler and matrix are still at early stages; initial cold- and hot-drawing approaches following spinning have not yet been optimised. It can be anticipated that production techniques will improve as a more fundamental understanding of nanocomposite rheology and thermal properties emerges and more nanocomposite material is prepared.

7.5

Analysing the rheological properties of nanotube/nanofibre–polymer composites

The addition of a nanofiller to a polymer melt can significantly affect the rheological properties; in certain cases, both processing and final composite properties can be enhanced with the same filler. To date, rheological studies of nanotube/nanofibre-filled systems have focused mainly on the shear behaviour;88, 136, 140, 169–171 however, shear tests alone cannot characterise the melt elongation properties that are important for foaming, film-blowing or fibre-spinning processes. There are as yet limited data available on the elongational viscosity of nanocomposite melts at low extension rates related to foaming processes172, 173 and very little174 at the high extension rates typically encountered during fibre-spinning operations. In addition, interactions between filler and melt occurring during processing can also

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

209

have pronounced impacts on the resulting matrix morphology129, 154 (see Section 7.6.1).

7.5.1

Shear properties

For particle-filled thermoplastic melt systems, oscillatory shear experiments are common and avoid issues with time-dependent alignment effects. As expected, the shear viscosity of such nanocomposites generally increases with loading fraction, to an extent that depends on the nature of the filler and matrix. Percolation effects have been observed, generally at low shear, although even low nanotube concentrations can lead to significant viscosity increases;170 at high shear rates, thinning associated with the polymer dominates, particularly if the filler can fragment. For example, CNF loadings up to 10 wt% had no significant influence on the shear viscosity of PP140 and PEEK154 composites in the shear rate regime typically encountered during thermoplastic processing. In case of a PC matrix, the shear viscosity was even reduced with increasing nanofibre content up to 10 wt%,136 most likely as a result of pronounced shear alignment of the filler, a well-known behaviour for short fibre-filled polymers.175 With increasing strain amplitude, nanotubes and nanofibres gradually align parallel to the flow direction, thus reducing the tube–tube or fibre–fibre interactions, as it has been observed in nanoclay systems.176 The formation or presence of filler network structures and/or aggregates at elevated contents is reflected by a pronounced increase in shear viscosity, especially at low shear rates (Fig. 7.5 shows the behaviour of nanofibre reinforced PEEK as an example.154 Such behaviour is typical for highly filled nanoparticulate systems, independent of filler type and geometry140, 170, 171 and is associated with a transition to an elastic pseudo solid-like response. This rheological percolation threshold depends on the nanotube/nanofibre type and treatment as well as on the polymer matrix and generally indicates the onset of interactions between individual filler particles or clusters. Often (for example the PEEK system shown in Fig. 7.5), strong shear thinning is observed;154 the effect can be attributed to the alignment of the nanofibres which, given their large size and rigidity, will experience a larger torque and have longer relaxation times than the polymer molecules. It is also interesting to note that meltcompounded nanocomposites tend to have similar rheological and electrical thresholds,154 although nanocomposites prepared by the coagulation method177, 178 have shown relatively lower rheological thresholds; an effect that is attributed to the fact that denser nanotube networks are required to achieve electrical conductivity than to restrict polymer motion. The comparatively high rheological percolation threshold (between 10 and 15 wt%) observed for this particular PEEK/CNF system (Fig. 7.5) is in agreement with studies on similar systems140, 169 but is much higher than that

© 2007, Woodhead Publishing Limited

210

Nanofibers and nanotechnology in textiles

15 wt% CNF 10 wt% CNF 5 wt% CNF Pure PEEK 104

103

102 10–2

Shear thinning exponent ‘n’

Complex shear viscosity η* [Pa s]

105

0.60

0.55

Slope ‘n’ 0.50 0

5 10 15 CNF content (wt%)

10–1

20

100 101 Frequency ω [rad/s]

102

103

7.5 Complex shear viscosity of carbon nanofibre reinforced PEEK composites as a function of frequency, at a temperature of 360 °C, in the linear viscoelastic regime. The insert shows the resulting shear thinning exponent as a function of nanofibre content.

observed in many MWCNT-filled thermoplastics.170, 171, 179 The most likely explanation is that the CNF aspect ratio is much more significantly degraded during processing owing to their larger absolute dimensions and lower strength. Comparison with the classic Guth equation180 suggests an aspect ratio of 15, much lower than that of the as-grown CNFs. Yet, degradation of aspect ratio during processing has also been reported by Kuriger et al.95 and Kharchenko et al.,179 investigating both CNF- and MWCNT-filled PP, respectively, as well as by Ma et al.181 evaluating melt-spun CNF/polyethylene terephthalate (PET) fibres.

7.5.2

Elongational properties

Although-the shear rheology data do not indicate any particular degradation of melt processability when adding such carbon nanofillers, they do not explain variations in flow behaviour or microstructure arising from elongation of the melt as it occurs during foaming,182 fibre-spinning,183 and film blowing.184 Uniaxial extension at high strain rates relevant for fibre-spinning can be explored using the Rheotens test:185 the force required to elongate a melt strand is measured as a function of a linearly accelerated draw-down velocity, under quasi-isothermal conditions.186 The force required to rupture

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

211

the strand is commonly defined as the ‘melt strength’ and the maximum draw-down velocity or draw-down ratio is defined as the ‘elongation to failure’ or ‘drawability’. The experimental data shown in Fig. 7.6 clearly highlight a substantial increase in PEEK melt strength and drawability with increasing nanofibre content, effects that can be attributed to a ‘reinforcement’ of the melt.154 A direct conversion of such Rheotens data into an elongational viscosity is complex187, 188 but it is possible to compare the drawing behaviour of different polymer nanocomposites at defined experimental conditions.189 For the PEEK– CNF system, the calculated elongational viscosity (Fig. 7.7) indicates an increase in initial elongational viscosity and an earlier onset of elongational thinning with increasing CNF content. Thus, the processing window relevant for operations such as foaming and fibre-spinning can be usefully adjusted. Both the increasing melt strength as well as the higher drawability of the PEEK nanocomposites explain the relatively good melt-spinnability of this system.190 In general, such effects may also allow significantly finer filaments to be spun, in the future, without early rupture of the melt strand. For example, fine nanofibre-reinforced polyester PET fibres with diameters as low as 25 µm have been produced under stable conditions.181 160

140

Tensile force [mN]

120 100

80

60

40

15 wt% PEEK 10 wt% PEEK 5 wt% PEEK Pure PEEK

20 0 0

20

40

60 80 100 Draw-down ratio

120

140

7.6 Experimental average Rheotens curves for PEEK–CNF nanocomposites at 360 °C. The lines represent a Levenberg– Marquardt fit of the data according to Wagner et al.187

© 2007, Woodhead Publishing Limited

160

212

Nanofibers and nanotechnology in textiles

Apparent elongational viscosity [Pa s]

106

105

104

103

15 wt% CNF 10 wt% CNF 5 wt% CNF Pure PEEK

102 10–3

10–2

10–1 100 101 Elongational rate [1/s]

102

103

7.7 Apparent elongational viscosity as a function of elongation rate for carbon nanofibre reinforced PEEK at 360 °C, calculated using the Wagner model.187

7.6

Analysing the microstructure of nanotube/ nanofibre–polymer composites

As outlined in Section 7.4, most experimental approaches towards nanocomposite fibre spinning are still at early stages, operating on smallscale material quantities. These initial attempts have, nevertheless, demonstrated a significant potential of producing good-quality nanocomposite fibres. Obviously, as in the case of bulk nanocomposites, both the general melt or solution spinnability as well as the fibre surface finish are strongly related to the quality of nanofiller dispersion achieved during processing prior to spinning. Both the use of highly entangled raw nanotube materials, as well as insufficient shear mixing or excessive post-shear reagglomeration, lead to clusters which can prohibit melt-spinning completely191 or induce rather low-quality fibre surfaces. 92 Rather than reinforcing the melt and even enhancing its drawability154 as highlighted in Section 7.5.2, nanofibre or nanotube agglomerates induce significant stress concentrations leading to a severe reduction of melt elongation. Intermediate levels of dispersion or the presence of very small agglomerates induce voids, rough surface finishes and/or diameter variations,92 as illustrated in Fig. 7.8.

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

(a)

(b)

(c)

(d)

213

7.8 Comparative SEM micrographs of melt-spun polyamide-12 nanocomposite fibres containing (a) 10 wt% CNFs, (b) 10 wt% entangled catalytic MWCNTs, (c) 5 wt% aligned CVD MWCNTs, and (d) 5 wt% arc-grown MWCNTs. Reproduced from Ref. 92.

The variations in fibre quality can be directly related to the nanotube/ nanofibre dispersion which, in turn, reflects the condition of the as-produced materials. Arc-grown material often is fused together by contaminating graphite; catalytically-grown MWCNTs separate well given sufficient shear forces but often retain some entanglements, while straighter and unentangled CVDgrown MWCNTs and most nanofibre types disperse most successfully. Efforts aimed at spinning SWCNT-reinforced composite fibres face related difficulties. At high loading fractions, exceeding a few weight percent, variations in the rheological behaviour prevent uniform and stable melt-spinning conditions.192 Furthermore, debundling of the SWCNTs remains a challenge; even solvent-blending prior to melt-compounding and spinning did not lead to individually dispersed nanotubes in polycarbonate fibres.121 Nevertheless, some successful nanotube-filled fibres have been spun at lower loading fractions.120, 193

© 2007, Woodhead Publishing Limited

214

7.6.1

Nanofibers and nanotechnology in textiles

Matrix microstructure

As mentioned before, the addition of carbon nanotubes and nanofibres to a polymer matrix can have significant implications on the resulting matrix microstructure. In the case of semicrystalline thermoplastics especially, crystallisation kinetics are often strongly affected by high surface area fillers, altering the degree of crystallinity, the crystalline orientation, and even the preferred crystal structure. The impact of CNTs and CNFs in highly oriented systems such as nanocomposite fibres has not yet been fully established, although related behaviour is known for conventional additives such as colouring pigments.194 Such effects are critical to the understanding of nanocomposite performance in general, but are too often ignored when evaluating the observed improvements in fibre properties with regard to the intrinsic properties of the filler. In the case of polyvinyl alcohal (PVOH), it has been suggested that ordering of PVOH around CNTs is the dominant determinant of the mechanical performance, rather than direct reinforcement.195 As for conventional polymer fibres, spectroscopy, diffraction196 and thermal analysis are commonly used to evaluate the final matrix microstructure of the nanocomposite fibre. As in the bulk case, significant variations in these microstructural features have been observed in semicrystalline fibre systems, depending on the filler type, treatment and content, as well as on the specific processing conditions. An example for melt-spun PEEK–CNF fibres is shown in Fig. 7.9; the differential scanning calorimetry (DSC) data indicate a significant increase in matrix crystallinity with the addition of nanofibres, although apparently independently of filler content. The two-chain orthorhombic crystal I form of PEEK usually seen for melt-spun and annealed monofilaments197 is maintained for the nanocomposites,190 yet the increased glass transition temperature of the neat polymer fibre (Fig. 7.9) strongly implies a significant variation in molecular alignment prior to quenching, in agreement with the elongational rheology behaviour of this particular system discussed in Section 7.5.2. In polyamide92 and polypropylene193 nanocomposite fibres, nucleation effects of the additives were observed, although no changes in the favoured crystal structure were reported. Laser Raman spectroscopy as well as wide-angle X-ray diffraction are commonly used for the characterisation of SWCNT and MWCNT/CNF alignment, respectively; although quantitative interpretation of the data can be challenging. For example, in the case of carbon nanofibres,92, 181, 190 an apparent lower degree of alignment is often determined from 2D wide-angle X-ray scattering (WAXS) patterns due to misalignment of the graphitic planes relative to the nanofibre axis (Section 7.1). In polypropylene fibres for example,198, 199 evaluation is further complicated by the superposition of graphitic and polymer peaks. Raman spectroscopy yields a much stronger signal when polarised parallel to the nanotube axis, and, once calibrated, can

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

215

10 wt% Endotherm

10 wt%

5 wt%

5 wt%

Pure PEEK

Pure PEEK

(a) As-spun 100

200

(b) Heat-treated 300

400 100 Temperature [°C]

200

300

400

7.9 Representative DSC traces of melt-spun PEEK–CNF fibres, (a) after melt-spinning and (b) following subsequent annealing at 220 °C, obtained on composite fibre bundles heated at 10 °C/min. Curves have been shifted for clarity.

indicate degree of alignment; an example of the resulting data for SWCNTs121 is shown in Fig. 7.10. As expected, overall, the data suggest a greater degree of both nanotube/ nanofibre and polymer alignment in fibre systems than in bulk nanocomposites. Some discrepancies remain, reflecting the significant variations in materials, experimental conditions and data analysis techniques. Filler alignment in general appears to critically depend on the quality of the initial dispersion, independently of processing technology; the lack of additional alignment of multiwalled structures during cold-drawing92 suggests that the nanofillers are fully aligned by the relatively low draw ratios applied during meltspinning; the remaining misorientation is likely to reflect both the misorientation of the graphitic layers, and the intrinsic ‘waviness’ of the CNTs/CNFs. That the nanofillers align easily is not suprising given their large size, long relaxation times and rigidity compared to the matrix polymer, and is consistent with the rheology data discussed above. The intrinsic ‘waviness’ that is characteristic of CVD-grown material and that cannot apparently be removed, at least for multiwalled structures, is likely to have a major, limiting effect on the maximum performance that can be achieved.200 In principle, more flexible SWCNT bundles may suffer from this effect to a lesser extent, but data so far seem to suggest a similar behaviour both in solution and in the melt.

© 2007, Woodhead Publishing Limited

216

Nanofibers and nanotechnology in textiles 1.2 Normalised intensity at 1588 cm–1

Matrix = PC 1.0

SWNT-Soln SWNT-Dry

0.8

Theory

0.6 0.4 Perfect alignment 0.2 0.0 0

10

20

30

40 50 60 ψ [degrees]

70

80

90

7.10 Raman intensity of the tangential mode at 1588 cm–1 of SWCNTs in melt-spun PC fibres, containing 5 wt% of SWCNTs, as a function of the fibre angle. Reproduced from Ref121 The dashed line represents the relationship between relative intensity and fibre angle for perfect or unidirectional alignment of SWCNTs along the fibre axis. The label ‘SWNT-Soln’ refers to an initially solution-blended SWCNT/PC system that was subsequently extruded and melt-spun in an identical manner to the ‘SWCNT-Dry’ sample.

7.7

Mechanical, electrical and other properties of nanocomposite fibres

As outlined above, well-dispersed nanotubes and nanofibres tend to align with the molecular flow and can experience a significant load transfer in the molten state during fibre spinning.154 Such reinforced melt systems show a good spinnability, leading to high-quality fibres even at elevated draw ratios. Similarly, the drawability and strain to failure of both as-prepared and treated solid-state nanocomposite fibres also critically depend on the quality of the filler dispersion, alignment and interfacial bonding. As an example, the stress– strain behaviour of both as-prepared and annealed melt-spun PEEK–CNF composite fibres is shown in Fig. 7.11.190 In most fibre systems, stiffness and strength increase, but strain to failure decreases with increasing filler entanglement and content. However, for some high loading fraction, solutionspun fibres and significant improvements in fibre toughness have been reported.157, 158 Overall, there is considerable variability in the published mechanical data due to the largely different systems and experimental approaches. Nevertheless, a tentative comparison of data for melt-spun fibres to the typical data for bulk systems processed in the melt (see Section 7.5.2)47, 201 indicates a more

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

217

200 (a) As-spun

150

100

Stress [MPa]

50

Pure PEEK 5 wt% 10 wt%

0 0

50

100

150

200

200 (b) Heat-treated

150

100

50

Pure PEEK 5 wt% 10 wt%

0 0

50

100 Strain [%]

150

200

7.11 Representative engineering stress–strain diagrams of (a) as-spun and (b) heat-treated PEEK–CNF nanocomposite fibres as a function of nanofibre content. (Tensile tests were based on a constant force ramp.190)

pronounced reinforcement in fibrous nanocomposites. This observation can be attributed to the improved alignment of the filler to the loading direction; a conventional short fibre interpretation based on the Krenchel model suggests that orientation efficiency of nanofibres may behave similarly137 although issues of fibre waviness are also important.200 As in the case of bulk nanotube composites, absolute values of the mechanical properties are mostly disappointing relative to the expectations generated by the mechanical properties of perfect nanotubes; the reasons are

© 2007, Woodhead Publishing Limited

218

Nanofibers and nanotechnology in textiles

likely to be similar: the relative imperfection of the CVD nanotubes used, difficulties with dispersion, and the need to optimise interfacial stress transfer without chemically damaging the nanotubes. A detailed analysis of the literature is complicated, as variations in the polymer microstructure, particularly of semicrystalline matrices, have not always been considered. Yet there is evidence that, all else being equal, the crystalline quality of the filler is important92 with dispersed MWCNTs outperforming CNFs; Fig. 7.12, for example, summarises the fibre modulus and yield stress for melt-spun polyamide-12 2.0 eMWCNT

PA fibre modulus [GPa]

1.8 1.6 1.4

CNF

aMWCNT

1.2 1.0 AMWCNT

0.8 0.6 0.0

2.5

5.0

7.5 (a)

10.0

12.5

15.0

PA12 fibre yield stress [MPa]

55 50 eMWCNT 45 40 CNF 35

aMWCNT

30 25 AMWCNT 20 15 0.0

2.5

5.0 7.5 10.0 Filler content [wt%] (b)

12.5

15.0

7.12 Plots summarising the relationship between (a) tensile modulus and (b) yield stress of melt-spun polyamide-12 nanocomposites and nanoscale filler weight fraction for CNFs, arc-grown (aMWCNT) entangled (eMWCNT) and aligned (aMWCNT) multiwall structures. (Tensile tests were based on a constant force ramp.92)

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

219

nanocomposite fibres containing different multiwalled fillers. It is not yet clear what size of nanotube will prove to be ideal; small diameter MWCNTs may be the best choice, as higher loading fractions of individual tubes can be realised as compared to most SWCNT systems, while retaining a relatively high degree of perfection and surface area. The deformation micro-/nano-mechanics of such fibrous nanocomposites are only slowly emerging; the picture is complicated by the quite pronounced variations in matrix microstructure following nanocomposite spinning. Some attempts have been made both to determine interfacial phenomena and to define the laws that govern their properties, for instance, the nanoscale scaling of flaw violations.202 However, the main focus remains on experimental methodologies to study the evolution of the microstructure during deformation. Laser Raman spectroscopy, for example, not only delivers static information regarding the local dispersion and alignment of nanotubes within a fibre, but also allows mapping of the local deformation in a composite by following the molecular deformation of the reinforcement.203 These molecular deformations are evidenced by a shift in Raman band position as a function of stress or strain applied to the composite during in-situ deformation experiments and sensitively highlight nanotube/matrix interactions and the degree of load transfer.204, 205 Alternatively, synchrotron X-ray sources can be used to study more explicitly the relationship of polymer and filler deformation and orientation, using a combination of small- and wide-angle studies.206, 207 For example, functionalised SWCNTs have been shown to suppress the reconstruction of lamellae stacks in poly(ethylene-propylene) fibres.114 Although individual polymer systems will need to be studied in detail, such techniques will shed light on the filler/matrix interactions in nanotube/nanofibre composite fibres and may lead to new micromechanical concepts for polymer fibre deformation in general.

7.7.1

Electrical properties

A range of interesting properties beyond straightforward mechanical reinforcement are expected.208 One interesting possibility is that nanotubefilled fibres may provide convenient, robust, electrically conductive textiles for a range of ‘intelligent’ applications, although the simplest advantage may be in electrostatic dissipation. Static build-up in textiles is not only be uncomfortable but can be dangerous in many industrial environments. Existing alternative approaches include post-spinning coating with nanoparticles209 or intrinsically conductive polymers,210 but often lack cohesion and longterm stability. As in the case of bulk nanocomposites, carbon nanotubes/ nanofibres are well suited to the formation of electrically conductive percolating networks (see Section 7.3.3 above). In principle, a good dispersion of high aspect ratio nanotubes can percolate at modest loadings, although the alignment

© 2007, Woodhead Publishing Limited

220

Nanofibers and nanotechnology in textiles

of straight nanotubes is expected to increase the percolation threshold.155 In reality, the intrinsic curvature and waviness of dispersed nanotubes, although detrimental for mechanical reinforcement, is beneficial for percolation. Nevertheless, the alignment induced by spinning has been shown to reduce or remove the conductivity observed in the bulk nanocomposites for both PC211 PP213. An early study in 1999 highlighted a significant decrease in volume resistivity of an isotropic pitch fibre containing 5 wt% of SWCNTs.192 Yet, SWCNT loading fractions of up to 10 wt% did not induce a measurable improvement in conductivity of rigid-rod PBO polymer fibres prepared by solution-spinning, again taken as evidence for a high degree of filler alignment.116 For comparison, solution-spun polyamide-11 fibres containing intrinsically conductive polyaniline showed a percolation threshold at around 5 wt% due to phase separation and fibrillation.212 In this case, the fibre conductivity even increased with increasing draw ratio, whereas nanofibrefilled PP fibres showed the reverse trend,213 as highlighted by the normalised current–voltage characteristics of such PP–CNF nanocomposite fibres shown in Fig. 7.13. Yet, a promising compromise between electrical and mechanical properties was attained at relatively low filler contents, verifying the potential of using such carbon nanostructures to accomplish electrically conductive thermoplastic fibres.

7.7.2

Other properties

An important feature of nanocomposite fibres is the reduced axial shrinkage116, 214 during both spinning and post-treatments, although these

Current, I (V )/Imax

1.0 0.8 0.6 0.4 10% VGCF DR = 1.0 DR = 1.5 DR = 3.5

0.2 0.0 0

20

40 60 80 Applied voltage [V]

15% VGCF DR = 1.0 DR = 3.5 DR = 8.0 100

7.13 Normalised current–voltage characteristics of melt-spun PP fibres containing 10 and 15 vol% of carbon nanofibres drawn to different draw ratios (DR). Reproduced from Ref. 213.

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

221

0.16

PBO PBO / SWNT (90/10)

Change in length [%]

0.12

0.08

0.04

0.00 0

2

4

6

8 10 Time [min]

12

14

16

7.14 Creep behaviour of neat PBO and PBO nanocomposite fibres containing 10 wt% of SWCNTs at 400 °C and a stress of 250 MPa. Reproduced from Ref. 116.

effects may be particularly pronounced due to the preliminary nature of the spinning studies. The ability of the nanofillers to constrain the polymer matrix will strongly depend on the dispersion and polymer microstructure,215 as well as on the effective surface area and chemical interactions. One positive example is the significantly improved axial creep behaviour of solution-spun high-performance PBO fibres containing 10 wt% of SWCNTs at elevated temperatures, shown in Fig. 7.14.116 Enhancements in other properties such as flame retardance,216 wear resistance,138 and thermal conductivity95, 131, 141 have been observed in bulk nanocomposite systems and are likely to be relevant to polymer fibres, although they are, as yet, much less thoroughly investigated. Indeed, the wear improvement observed in PEEK138 is already in a system that can be spun into fibres.190

7.8

Future trends

The introduction of nanotubes and nanofibres into polymer fibres is an appealing prospect, and a natural embodiment of this nanocomposite system. As discussed in the motivation section (Section 7.4) above, there is scope to provide immediate mechanical enhancements, using existing nanotube and nanofibre materials, that cannot be obtained using other fillers. Nanocomposite fibres may represent a useful and early mechanical application of carbon

© 2007, Woodhead Publishing Limited

222

Nanofibers and nanotechnology in textiles

nanomaterials. In addition, the presence of the nanofiller, if well dispersed, may aid the spinning process by increasing the melt strength and/or drawability of the polymer. While developing these potentially useful systems, a number of challenges associated with nanotube composites in general can be addressed. The long-term goal must be to access the properties of individual perfect nanotubes on a macroscopic scale. As in the case of bulk applications, the performance of such fibrous nanocomposites critically depends on the quality of nanostructure dispersion achieved prior to spinning and drawing. The manufacture of nanocomposite fibres with uniform diameter and good surface finish relies on well-dispersed filler particles, even at low draw ratios. Methods to improve both dispersion and to optimise the interface between polymer and filler are emerging, based particularly on chemical functionalisation of the surfaces. Probably the most central remaining difficulty is to identify the nature of the ideal nanotube for use in composites and then to synthesise sufficiently large quantities at low enough cost for practical usage. Although a variety of synthesis methods now exist to produce carbon nanotubes and nanofibres, the products differ greatly in terms of diameter, aspect ratio, crystallinity, crystalline orientation, purity, entanglement, surface chemistry and straightness. These structural variations dramatically affect intrinsic properties, processing and behaviour in composite systems. However, it is not yet clear which type of nanotube material is most suitable for composite applications, nor is there much theoretical basis for rational design. Ultimately, the selection will depend on the matrix material, processing technology and the property enhancement required. Further comparative studies are needed, together with theoretical understanding of the nanomechanical mechanisms at work and their interaction with the polymer microstructure. Many questions remain outstanding but there are rich rewards to be gained, if the true potential of carbon nanotubes can be harnessed.

7.9

References

1. Ebbesen, T., ‘Carbon Nanotubes, Preparation and Properties’, CRC Press, Boca Raton (1997) 2. Krätschmer, W., Lamb, L.D., Fostiropoulos, K., Huffmann, D.R. ‘Solid C60: A new form of carbon’, Nature 347(6291) (1990), 354–358 3. Kroto, H.W., Heath, J.R., O’Brian, S.C., Curl, R.F., Smalley, R.E. ‘C 60 : Buckminsterfullerene’, Nature 318(6042) (1985), 162–163 4. Oberlin, A., Endo, M., Koyama, T. ‘Filamentous growth of carbon through benzene decomposition’, J. Cryst. Growth 32(3) (1976), 335–349 5. Monthioux, M., Kuznetsov, V.L. ‘Who should be given the credit for the discovery of carbon nanotubes?’, Carbon 44(9) (2006), 1621–1623 6. Esquivel, E.V., Murr, L.E. ‘A TEM analysis of nanoparticulates in a Polar ice core’, Mater. Charact. 52(1) (2004), 15–25 7. Reibold, M., Paufler, P., Levin, A.A., Kochmann, W., Pätzke, N., Meyer, D.C. ‘Carbon nanotubes in an ancient Damascus sabre’, Nature 444(7117) (2006), 286

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

223

8. Zheng, L.X., O’Connell, M.J., Doorn, S.K., Liao, X.Z., Zhao, Y.H., Akhadov, E.A., Hoffbauer, M.A., Roop, B.J., Jia, Q.X., Dye, R.C., Peterson, D.E., Huang, S.M., Liu, J., Zhu, Y.T. ‘Ultralong single-wall carbon nanotubes’, Nature Mater. 3(10) (2004), 673–676 9. Meyyappan, M., ‘Carbon Nanotubes: Science and Applications’, CRC Press, Boca Raton (2005) 10. Tersoff, J., Ruoff, R.S. ‘Structural properties of a carbon-nanotube crystal’, Phys. Rev. Lett. 73(5) (1994), 676–679 11. Bethune, D.S., Kiang, C.H., de Vries, M.S., Gorman, G., Savoy, R., Vazquez, J., Beyers, R. ‘Cobalt-catalysed growth of carbon nanotubes with single-atomic-layer walls’, Nature 363(6430) (1993), 605–607 12. Journet, C., Master, W.K., Bernier, P., Loiseau, A., Lamy de la Chapelle, M., Lefrant, S., Deniard, P., Lee, R., Fischer, J.E. ‘Large-scale production of singlewalled carbon nanotubes by the electric-arc technique’, Nature 388(6644) (1997), 756–758 13. Thess, A., Lee, R., Nikolaev, P., Dai, H., Petit, P., Robert, J., Xu, C., Lee, Y.H., Kim, S.G., Rinzler, A.G., Colbert, D.T., Scuseria, G.E., Tomanek, D., Fischer, J.E., Smalley, R.E. ‘Crystalline ropes of metallic carbon nanotubes’, Science 273(5274) (1996), 483–487 14. Cheng, H.M., Li, F., Su, G., Pan, H.Y., He, L.L., Sun, X., Dresselhaus, M.S. ‘Large-scale and low-cost synthesis of single-walled carbon nanotubes by the catalytic pyrolysis of hydrocarbons’, Appl. Phys. Lett. 72(25) (1998), 3282–3284 15. Endo, M., Takeuchi, K., Igarashi, S., Kobori, K., Shiraishi, M., Kroto, H.W. ‘The production and structure of pyrolytic carbon nanotubes (PCNTs)’, J. Phys. Chem. Solids 54(12) (1993), 1841–1848 16. Sandler, J.K.W., Kirk, J.E., Kinloch, I.A., Shaffer, M.S.P., Windle, A.H. ‘Ultra-low electrical percolation threshold in carbon-nanotube–epoxy composites’, Polymer 44(19) (2003), 5893–5899 17. Ebbesen, T.W., Ajayan, P.M. ‘Large-scale synthesis of carbon nanotubes’, Nature 358(6383) (1992), 220–222 18. Colbert, D.T., Zhang, J., McClure, S.M., Nikolaev, P., Chen, Z., Hafner, J.H., Owens, D.W., Kotula, P.G., Carter, C.B., Weaver, J.H., Rinzler, A.G., Smalley, R.E. ‘Growth and sintering of fullerene nanotubes’, Science 266(5188) (1994), 1218–1222 19. Cadek, M., Murphy, R., McCarthy, B., Drury, A., Lahr, B., Barklie, R.C., in het Panhuis, M., Coleman, J.N., Blau, W.J. ‘Optimisation of the arc-discharge production of multi-walled carbon nanotubes’, Carbon 40(6) (2002), 923–928 20. Rinzler, A.G., Liu, J., Dai, H., Nikolaev, P., Huffman, C.B., Rodriguez-Macias, F.J., Boul, P.J., Lu, A.H., Heymann, D., Colbert, D.T., Lee, R.S., Fischer, J.E., Rao, A.M., Eklund, P.C., Smalley, R.E. ‘Large-scale purification of single-wall carbon nanotubes: process, product, and characterization’, Appl. Phys. A 67(1) (1998), 29–37 21. Endo, M., Takeuchi, K., Igarashi, S., Kobori, K., Shiraishi, M., Kroto, H.W. ‘The production and structure of pyrolytic carbon nanotubes (PCNTs)’, J. Phys. Chem. Solids 54(12) (1993), 1841–1848 22. Rao, C.N.R., Govindaraj, A., Sen, R., Satishkumar, B.C. ‘Synthesis of multi-walled and single-walled nanotubes, aligned-nanotube bundles and nanorods by employing organometallic precursors’, Mater. Res. Innovat. 2(3) (1998), 128–141. 23. Andrews, R., Jacques, D., Rao, A.M., Derbyshire, F., Qian, D., Fan, X., Dickey,

© 2007, Woodhead Publishing Limited

224

24.

25. 26. 27.

28. 29.

30.

31. 32. 33.

34.

35.

36.

37. 38. 39. 40.

41. 42.

Nanofibers and nanotechnology in textiles E.C., Chen, J. ‘Continuous production of aligned carbon nanotubes: A step closer to commercial realization’, Chem. Phys. Lett. 303(5–6) (1999), 467–474 Tibbetts, G.G., Gorkiewicz, D.W., Alig, R.L. ‘A new reactor for growing carbonfibers from liquid-phase and vapor-phase hydrocarbons’, Carbon 31(5) (1993), 809–814. Treacy, M.M.J., Ebbesen, T.W., Gibson, J.M. ‘Exceptionally high Young’s modulus observed for individual carbon nanotubes’, Nature 381(6584) (1996), 678–680 Krishnan, A., Dujardin, E., Ebbesen, T.W., Yianilos, P.N., Treacy, M.M.J. ‘Young’s modulus of single-walled nanotubes’, Phys. Rev. B 58(20) (1998), 14013–14019 Falvo, M.R., Clary, G.J., Taylor, R.M., Chi, V., Brooks Jr., F.P., Washburn, S., Superfine, R. ‘Bending and buckling of carbon nanotubes under large strain’, Nature 389(6651) (1997), 582–584 Wong, E.W., Shehan, P.E., Lieber, C.M. ‘Nanobeam mechanics: elasticity, strength and toughness of nanorods and nanotubes’, Science 277(5334) (1997), 1971–1975 Yu, M.-F., Lourie, O., Dyer, M.J., Moloni, K., Kelly, T.F., Ruoff, R.S. ‘Strength and breaking mechanism of multiwalled carbon nanotubes under tensile load’, Science 287(5453) (2000), 637–640 Yu, M.-F., Files, B.S., Arepalli, A., Ruoff, R.S. ‘Tensile loading of ropes of single wall carbon nanotubes and their mechanical properties’, Phys. Rev. Lett. 84(24) (2000), 5552–5555 Nardelli, M.B., Yakobson, B.I., Bernholc, J. ‘Brittle and ductile behavior in carbon nanotubes’, Phys. Rev. Lett. 81(21) (1998), 4656–4659 Lourie, O., Cox, D.M., Wagner, H.D. ‘Buckling and collapse of embedded carbon nanotubes’, Phys. Rev. Lett. 81(8) (1998), 1638–1641 Wagner, H.D., Lourie, O., Feldmann, Y., Tenne, R. ‘Stress-induced fragmentation of multiwall carbon nanotubes in a polymer matrix’, Appl. Phys. Lett. 72(2) (1998), 188–190 Li, F., Chai, H.M., Bai, S., Su, G., Dresselhaus, M.S. ‘Tensile strength of singlewalled carbon nanotubes directly measured from their macroscopic ropes’, Appl. Phys. Lett. 77(20) (2000), 3161–3163 Walters, D.A., Ericson, L.M., Casavant, M.J., Liu, J., Colbert, D.T., Smith, K.A., Smalley, R.E. ‘Elastic strain of freely suspended single-wall carbon nanotube ropes’, Appl. Phys. Lett. 74(25) (1999), 3803–3805 Pan, Z.W., Xie, S.S., Lu, L., Chang, B.H., Sun, L.F., Zhou, W.Y., Wang, G., Zhang, D.L. ‘Tensile tests of ropes of very long aligned multiwall carbon nanotubes’, Appl. Phys. Lett. 74(21) (1999), 3152–3154 Liu, J.Z., Zheng, Q., Jiang, Q. ‘Effect of a rippling mode on resonances of carbon nanotubes’, Phys. Rev. Lett. 86(21) (2001), 4843–4846 Yakobson, B.I., Brabec, C.J., Bernholc, J. ‘Nanomechanics of carbon tubes: Instabilities beyond linear response’, Phys. Rev. Lett. 76(14) (1996), 2511–2514 Yakobson, B.I., Smalley, R.E. ‘Fullerene nanotubes: C1,000,000 and beyond’, Am. Sci. 85(6) (1997), 324–337 Bernholc, J., Brabec, C.J., Nardelli, M.B., Maiti, A., Roland, C., Yakobson, B.I. ‘Theory of growth and mechanical properties of nanotubes’, Appl. Phys. A 67(1) (1998), 39–46 Belytschko, T., Xiao, S.P., Schatz, G.C., Ruoff, R.S. ‘Atomistic simulation of nanotube fracture’, Phys. Rev. B 65(235430) (2002) Govindjee, S., Sackman, J.L. ‘On the use of continuum mechanics to estimate the properties of nanotubes’, Solid State Commun. 110(4) (1999), 227–230

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

225

43. Pantano, A., Parks, D.M., Boyce, M.C. ‘Mechanics of deformation of single- and multi-wall carbon nanotubes’, J. Mech. Phys. Solids 52(4) (2003), 789–821 44. Despres, J.F., Daguerre, E., Lafdi, K. ‘Flexibility of graphene layers in carbon nanotubes’, Carbon 33(1) (1995), 87–92 45. Ruoff, R.S., Lorents, D.C. ‘Mechanical and thermal properties of carbon nanotubes’, Carbon 33(7) (1995), 925–930 46. Iijima, S., Brabec, C., Maiti, A., Bernholc, J. ‘Structural flexibility of carbon nanotubes’, J. Chem. Phys. 104(5) (1996), 2089–2092 47. Coleman, J.N., Khan, U., Blau, W.J., Gun’ko, Y.K. ‘Small but strong: a review of the mechanical properties of carbon nanotube–polymer composites’, Carbon 44(9) (2006), 1624–1652 48. Chandra, N., Namilae, S. ‘Tensile and compressive behavior of carbon nanotubes: effect of functionalization and topological defects’, Mech. Adv. Mater. Struct. 13(2) (2006), 115–127 49. Salvetat, J.-P., Kulik, A.J., Bonard, J.-M., Briggs, G.A.D., Stökli, T., Metenier, K., Bonnamy, S., Beguin, F., Burnham, N.A., Forro, L. ‘Elastic modulus of ordered and disordered multiwalled carbon nanotubes’, Adv. Mater. 11(2) (1999), 161–165 50. Enomoto, K., Kitakata, S., Yasuhara, T., Ohtake, N., Kuzumaki, T., Mitsuda, Y. ‘Measurement of Young’s modulus of carbon nanotubes by nanoprobe manipulation in a transmission electron microscope’, Appl. Phys. Lett. 88(15) (2006), 153115 51. Pan, Z.W., Xie, S.S., Lu, L., Chang, B.H., Sun, L.F., Zhou, W.Y., Wang, G., Zhang, D.L. ‘Tensile tests of ropes of very long aligned multiwall carbon nanotubes’, Appl. Phys. Lett. 74(21) (1999), 3152–3154 52. Ge, M.H., Sattler, K. ‘Bundles of carbon nanotubes generated by vapor-phase growth’, Appl. Phys. Lett. 64(6) (1994), 710–711 53. Schadler, L.S., Giannaris, S.C., Ajayan, P.M. ‘Load transfer in carbon nanotubes epoxy composites’, Appl. Phys. Lett. 73(26) (1998), 3842–3844 54. Saito, R., Fujita, M., Dresselhaus, G., Dresselhaus, M.S. ‘Electronic structure of chiral graphene tubules’, Appl. Phys. Lett. 60(18) (1992), 2204–2206 55. Hamada, N., Sawada, S.-I., Oshiyama, A. ‘New one-dimensional conductors: graphitic microtubules’, Phys. Rev. Lett. 68(10) (1992), 1579–1581 56. Mintmire, J.W., Dunlop, B.I., White, C.T. ‘Are fullerene tubules metallic?’, Phys. Rev. Lett. 68(5) (1992), 631–634 57. White, C.T., Robertson, D.H., Mintmire, J.W. ‘Helical and rotational symmetries of nanoscale graphitic tubules’, Phys. Rev. B 47(9) (1993), 5485–5488 58. Mintmire, J.W., White, C.T. ‘Electronic and structural properties of carbon nanotubes’, Carbon 33(7) (1995), 893–902 59. Benedict, L.X., Crespi, V.H., Louie, S.G., Cohen, M.L. ‘Static conductivity and superconductivity of carbon nanotubes: relations between tubes and sheets’, Phys. Rev. B 52(20) (1995), 14935–14940 60. Tanaka, K., Aoki, H., Ago, H., Yamabe, T., Okahara, K. ‘Interlayer interaction of two graphene sheets as a model of double-layer carbon nanotubes’, Carbon 35(1) (1997), 121–125 61. Lambin, P., Meunier, V., Rubio, A. ‘Electronic structure of polychiral carbon nanotubes’, Phys. Rev. B 62(8) (2000), 5129–5135 62. Kane, C.L., Mele, E.J. ‘Size, shape, and low energy electronic structure of carbon nanotubes’, Phys. Rev. Lett. 78(10) (1997), 1932–1935 63. Langer, L., Stockman, L., Heremans, J.P., Bayot, V., Olk, C.H., van Haesendonck, C., Bruynseraede, Y., Issi, J.P. ‘Electrical resistance of a carbon nanotube bundle’, J. Mater. Res. 9(4) (1994), 927–932

© 2007, Woodhead Publishing Limited

226

Nanofibers and nanotechnology in textiles

64. Heremans, J.P., Olk, C.H., Morelli, D.T. ‘Magnetic susceptibility of carbon structures’, Phys. Rev. B 49(21) (1994), 15122–15125 65. Tans, S.J., Devoret, M.H., Dai, H., Thess, A., Smalley, R.E., Geerligs, L.J., Dekker, C. ‘Individual single-wall carbon nanotubes as quantum wires’, Nature 386(6624) (1997), 474–477 66. Ebbesen, T.W., Lezec, H.J., Hiura, H., Bennett, J.W., Ghaemie, H.F., Thio, T. ‘Electrical conductivity of individual carbon nanotubes’, Nature 382(6586) (1996), 54–56. 67. Odom, T.W., Huang, J.-L., Kim, P., Lieber, C.M. ‘Atomic structure and electronic properties of single-walled carbon nanotubes’, Nature 391(6662) (1998), 62–64 68. Wildöer, J.W.G., Venema, L.C., Rinzler, A.G., Smalley, R.E., Dekker, C. ‘Electronic structure of atomically resolved carbon nanotubes’, Nature 391(6662) (1998), 59– 62 69. Bachilo, S.M., Strano, M.S., Kittrell, C., Hauge, R.H., Smalley, R.E., Weisman, R.B. ‘Structure-assigned optical spectra of single-walled carbon nanotubes’, Science 298(5602) (2002), 2361–2366 70. Telg, H., Maultzsch, J., Reich, S., Hennrich, F., Thomsen, C. ‘Chirality distribution and transition energies of carbon nanotubes’, Phys. Rev. Lett. 93(17) (2004), 177401 71. Bockrath, M., Cobden, D.H., McEuen, P.L., Chopra, N.G., Zettl, A., Thess, A., Smalley, R.E. ‘Single-electron transport in ropes of carbon nanotubes’, Science 275(5308) (1997), 1922–1925 72. Kim, G.T., Choi, E.S., Kim, D.C., Suh, D.S., Park, Y.W., Liu, K., Duesberg, G., Roth, S. ‘Magnetoresistance of an entangled single-wall carbon-nanotube network’, Phys. Rev. B 58(24) (1998), 16064–16069 73. Fischer, J.E., Dai, H., Thess, A., Lee, R., Hanjani, N.M., Dehaas, D.L., Smalley, R.E. ‘Metallic resistivity in crystalline ropes of single-wall carbon nanotubes’, Phys. Rev. B 55(8) (1997), 4921–4924 74. Bozhko, A.D., Sklovsky, D.E., Nalimova, V.A., Rinzler, A.G., Smalley, R.E., Fischer, J.E. ‘Resistance vs. pressure of single-wall carbon nanotubes’, Appl. Phys. A 67(1) (1998), 75–77 75. Charlier, J.R., Issi, J.-P. ‘Electrical conductivity of novel forms of carbon’, J. Phys. Chem. Solids 57(6–8) (1996), 957–965 76. Lee, J.-O., Park, C., Kim, J.-J., Kim, J., Park, J.W., Yoo, K.-H. ‘Formation of lowresistance ohmic contact between carbon nanotubes and metal electrodes by a rapid thermal annealing method’, J. Phys. D: Appl. Phys. 33(16) (2000), 1953– 1956 77. Dai, H., Wong, E.W., Lieber, C.M. ‘Probing electrical transport in nanomaterials: conductivity of individual carbon nanotubes’, Science 272(5261) (1996), 523–526 78. Poncharal, P., Berger, C., Yi, Y., Wang, Z.L., de Heer, W.A. ‘Room temperature ballistic conduction in carbon nanotubes’, J. Phys. Chem. B 106(47) (2002), 12104– 12118 79. Hone, J., Whitny, M., Zettl, A. ‘Thermal conductivity of single-walled carbon nanotubes”, Synth. Metals 103(1–3) (1999), 2498–2499 80. Kim, P., Shi, L., Majumdar, A., McEuen, P.L. ‘Thermal transport measurements of individual multiwalled nanotubes’, Phys. Rev. Lett. 87(21) (2001), 215502 81. Hone, J., Llaguno, M.C., Nemes, N.M., Johnson, A.T., Fischer, J.E., Walters, D.A., Casavant, M.J., Schmidt, J., Smalley, R.E. ‘Electrical and thermal transport properties of magnetically aligned single walled carbon nanotube films’, Appl. Phys. Lett. 77(5) (2000), 666–668

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

227

82. Andrews, R., Jacques, D., Qian, D., Dickey, E.C. ‘Purification and structural annealing of multi-walled carbon nanotubes at graphitization temperatures’, Carbon 39(11) (2001), 1681–1687 83. Arnold, M.S., Green, A.A., Hulvat, J.F., Stupp, S.I., Hersam, M.C. ‘Sorting carbon nanotubes by electronic structure using density differentiation’, Nature Nanotechnol. 1 (2006), 60–65 84. Niu, C.M., Sichel, E.K., Hoch, R., Moy, D., Tennet, H. ‘High power electrochemical capacitors based on carbon nanotube electrodes’, Appl. Phys. Lett. 70(11) (1997), 1480–1482 85. Peigney, A., Laurent, C., Flahaut, E., Bacsa, R.R., Rousset, A. ‘Specific surface area of carbon nanotubes and bundles of carbon nanotubes’, Carbon 39(4) (2001), 507–514 86. Lozano, K., Barrera, E.V. ‘Nanofiber-reinforced thermoplastic composites I: Thermoanalytical and mechanical analyses’, J. Appl. Polym. Sci. 79(1) (2001), 125–133 87. Jin, Z., Pramoda, K.P., Xu, G., Goh, S.H. ‘Dynamic mechanical behavior of meltprocessed multi-walled carbon nanotube/poly(methyl methacrylate) composites’, Chem. Phys. Lett. 337(1–3) (2001), 43–47 88. Pötschke, P., Fornes, T.D., Paul, D.R. ‘Rheological behavior of multiwalled carbon nanotube/polycarbonate composites’, Polymer 43(11) (2002) 3247–3255 89. Pötschke, P., Dudkin, S.M., Alig, I. ‘Dielectric spectroscopy on melt processed polycarbonate-multiwalled carbon nanotube composites’, Polymer 44(17) (2003), 5023–5030 90. Meincke, O., Kaempfer, D., Weickmann, H., Friedrich, C., Vathauer, M., Warth, H. ‘Mechanical properties and electrical conductivity of carbon-nanotube filled polyamide-6 and its blends with acrylonitrile/butadiene/styrene’ Polymer 45(3) (2004), 739–748 91. Thostenson, E.T., Chou, T.-W. ‘Aligned multi-walled carbon nanotube-reinforced composites: processing and mechanical characterization’, J. Phys. D: Appl. Phys. 35(16) (2002), 77–80 92. Sandler, J.K.W., Pegel, S., Cadek, M., Gojny, F., van Es, M., Lohmar, J., Blau, W.J., Schulte, K., Windle, A.H., Shaffer, M.S.P. ‘A comparative study of melt-spun polyamide-12 fibres reinforced with carbon nanotubes and nanofibres’, Polymer 45(6) (2004), 2001–2015 93. Sato, E., Takahas, T., Koyama, K. ‘Comparison of the length of vapor-grown carbon fiber before and after mixing process’, Kobunshi Ronbunshu 61(2) (2004), 144–148 94. Tibbetts, G.G., McHugh, J.J. ‘Mechanical properties of vapor-grown carbon fiber composites with thermoplastic matrices’, J. Mater. Res. 14(7) (1999), 2871–2880 95. Kuriger, R.J., Alam, M.K., Anderson, D.P., Jacobsen, R.L. ‘Processing and characterization of aligned vapor grown carbon fiber reinforced polypropylene’, Comp. Part A 33(1) (2002), 53–62 96. Lu, K.L., Lago, R.M., Chen, Y.K., Green, M.L.H., Harris, P.F., Tsang, S.C. ‘Mechanical damage of carbon nanotubes by ultrasound’, Carbon 34(6) (1996), 814–816 97. Koshio, A., Yudasaka, M., Zhang, M., Iijima, S. ‘A simple way to chemically react single-wall carbon nanotubes with organic materials using ultrasonication’, Nano Lett. 1(7) (2001), 361–363 98. O’Connell, M.J., Boul, P., Ericson, L.M., Huffman, C., Wang, Y., Haroz, E., Kuper,

© 2007, Woodhead Publishing Limited

228

99. 100. 101. 102.

103.

104.

105.

106.

107.

108.

109.

110.

111.

112.

113. 114.

115.

Nanofibers and nanotechnology in textiles C., Tour, J., Ausman, K.D., Smalley, R.E. ‘Reversible water-solubilization of singlewalled carbon nanotubes by polymer wrapping’, Chem. Phys. Lett. 342(3–4) (2001), 265–271 Jin, L., Bower, C., Zhou, O. ‘Alignment of carbon nanotubes in a polymer matrix by mechanical stretching’, Appl. Phys. Lett. 73(9) (1998), 1197–1199 Bower, C., Rosen, R., Han, J., Zhou, O. ‘Deformation of carbon nanotubes in nanotube–polymer composites’, Appl. Phys. Lett. 74(22) (1999), 3317–3319 Shaffer, M.S.P., Windle, A.H. ‘Fabrication and characterization of carbon nanotube/ poly(vinyl alcohol) composites’, Adv. Mater. 11(11) (1999), 937–941 Cadek, M., Coleman, J.N., Barron, V., Hedicke, K., Blau, W.J. ‘Morphological and mechanical properties of carbon-nanotube-reinforced semicrystalline and amorphous polymer composites’, Appl. Phys. Lett. 81(27) (2002), 5123–5125 Zhang, X., Liu, T., Sreekumar, T.V., Kumar, S., Moore, V.C., Hauge, R.H., Smalley, R.E. ‘Poly(vinyl alcohol)/SWCNT composite film’, Nano Lett. 3(9) (2003), 1285– 1288 Qian, D., Dickey, E.C., Andrews, R., Rantell, T. ‘Load transfer and deformation mechanisms in carbon nanotube–polystyrene composites’, Appl. Phys. Lett. 76(20) (2000), 2868–2870 Watts, P.C., Hsu, W.K., Chen, G.Z., Fray, D.J., Kroto, H.W., Walton, D.R.M. ‘A low resistance boron-doped carbon nanotube-polystyrene composite’, J. Mater. Chem. 11(10) (2001), 2482–2488 Safadi, B., Andrews, R., Grulke, E.A. ‘Multiwalled carbon nanotube polymer composites: synthesis and characterization of thin films’, J. Appl. Polym. Sci. 84(14) (2002), 2660–2669 Ruan, S.L., Gao, P., Yang, X.G., Yu, T.X. ‘Toughening high performance ultrahigh molecular weight polyethylene using multiwalled carbon nanotubes’, Polymer 44(19) (2003), 5643–5654 Assouline, E., Lustiger, A., Barber, A.H., Cooper, C.A., Klein, E., Wachtel, E., Wagner, H.D. ‘Nucleation ability of multiwall carbon nanotubes in polypropylene composites’, J. Polym. Sci. B: Polym. Phys. 41(5) (2003), 520–527 Dufresne, A., Paillet, M., Putaux, J.L., Canet, R., Carmona, F., Delhaes, P., Cui, S. ‘Processing and characterization of carbon nanotube/poly(styrene-co-butyl acrylate) nanocomposites’, J. Mater. Sci. 37(18) (2002), 3915–3923 Paiva, M.C., Zhou, B., Fernando, K.A.S., Lin, Y., Kennedy, J.M., Sun, Y.-P. ‘Mechanical and morphological characterization of polymer–carbon nanocomposites from functionalized carbon nanotubes’, Carbon 42(14) (2004), 2849–2854 Lin, Y., Zhou, B., Fernando, K.A.S., Liu, P., Allard, L.A., Sun, Y.-P. ‘Polymeric carbon nanocomposites from carbon nanotubes functionalised with matrix polymer’, Macromolecules 36(19) (2003), 7199–7204 Mitchell, C.A., Bahr, J.L., Arepalli, S., Tour, J.M., Krishnamoorti, R. ‘Dispersion of functionalized carbon nanotubes in polystyrene’, Macromolecules 35(23) (2002), 8825–8830 Shaffer, M.S.P., Fan, X., Windle, A.H. ‘Dispersion and packing of carbon nanotubes’, Carbon 36(11) (1998), 1603–1612 Chen, X.M., Burger, C., Fang, D.F., Sics, I., Wang, X.F., He, W.D., Somani, R.H., Yoon, K., Hsiao, B.S., Chu, B. ‘In-situ X-ray deformation study of fluorinated multiwalled carbon nanotube and fluorinated ethylene-propylene nanocomposite fibers’, Macromolecules 39(16) (2006), 5427–5437 Jia, Z.J., Wang, Z.Y., Xu, C.L., Liang, J., Wei, B.Q., Wu, D.H., Zhu, S.W. ‘Study

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

116.

117.

118.

119.

120. 121.

122.

123. 124. 125. 126. 127.

128.

129.

130.

131.

132. 133.

229

on poly(methyl methacrylate)/carbon nanotube composites’, Mater. Sci. Eng. A 271(1–2) (1999), 395–400 Kumar, S., Dang, T.D., Arnold, F.E., Bhattacharyya, A.R., Min, B.G., Zhang, X., Vaia, R.A., Park, C., Adams, W.W., Hauge, R.H., Smalley, R.E., Ramesh, S., Willis, P.A. ‘Synthesis, structure, and properties of PBO/SWNT composites’, Macromolecules 35(24) (2002), 9039–9043 Barraza, H.J., Pompeo, F., O’Rear, E.A., Resasco, D.E. ‘SWNT-filled thermoplastic and elastomeric composites prepared by miniemulsion polymerization’, Nano Lett. 2(8) (2002), 797–802 Ounaies, Z., Park, C., Wise, K.E., Siochi, E.J., Harrison, J.S. ‘Electrical properties of single wall carbon nanotube reinforced polyimide composites’, Comp. Sci. Tech. 63(11) (2003), 1637–1646 Roslaniec, Z., Broza, G., Schulte, K. ‘Nanocomposites based on multiblock polyester elastomers (PEE) and carbon nanotubes (CNT)’, Comp. Interfaces 10(1) (2003), 95–102 Kearns, J.C., Shambaugh, R.L. ‘Polypropylene fibers reinforced with carbon nanotubes’, J. Appl. Polym. Sci. 86(8) (2002), 2079–2084 Fornes, T.D., Baur, J.W., Sabba, Y., Thomas, E.L. ‘Morphology and properties of melt-spun polycarbonate fibers containing single- and multi-wall carbon nanotubes’, Polymer 47(5) (2006), 1704–1714 Frankland, S.J.V., Caglar, A., Brenner, D.W., Griebel, M. ‘Molecular simulation of the influence of chemical cross-links on the shear strength of carbon nanotubepolymer interfaces’, J. Phys. Chem. B 106(12) (2002), 3046–3048 Liao, K., Li, S. ‘Interfacial characteristics of a carbon nanotube–polystyrene composite system’, Appl. Phys. Lett. 79(25) (2001), 4225–4227 Wong, M., Paramsothy, M., Xu, X.J., Ren, Y., Li, S., Liao, K. ‘Physical interactions at carbon nanotube-polymer interfaces’, Polymer 44(25) (2003), 7757–7764 Cooper, C.A., Cohen, S.R., Barber, A.H., Wagner, H.D. ‘Detachment of nanotubes from a polymer matrix’, Appl. Phys. Lett. 81(20) (2002), 3873–3875 Barber, A.H., Cohen, S.R., Wagner, H.D. ‘Measurement of carbon nanotube-polymer interfacial strength’, Appl. Phys. Lett. 82(23) (2003), 4140–4142 Barber, A.H., Cohen, S.R., Kenig, S., Wagner, H.D. ‘Interfacial fracture energy measurements for multi-walled carbon nanotubes pulled from a matrix’, Comp. Sci. Tech. 64(15) (2004), 2283–2289 Xie, X.L., Aloys, K., Zhou, X.P., Zeng, F.D. ‘Ultrahigh molecular mass polyethylene/ carbon nanotube composites – crystallization and melting properties’, J. Therm. Anal. Calorim. 74(1) (2003), 317–323 Sandler, J., Broza, G., Nolte, M., Schulte, K., Lam, Y.-M., Shaffer, M.S.P. ‘Crystallization of carbon nanotube and nanofiber polypropylene composites’, J. Macromol. Sci. – Phys. B 42(3–4) (2003), 479–488 Grady, B.P., Pompeo, F., Shambaugh, R.L., Resasco, D.E. ‘Nucleation of polypropylene crystallization by single-walled carbon nanotubes’, J. Phys. Chem. B 106(23) (2002), 5852–5858 Patton, R.D., Pittman Jr., C.U., Wang, L., Hill, J.R. ‘Vapor grown carbon fiber composites with epoxy and poly(phenylene sulfide) matrices’, Comp. Part A 30(9) (1999), 1081–1091 Ray, S.S., Okamoto, M. ‘Polymer/layered silicate nanocomposites: a review from preparation to processing’, Prog. Polym. Sci. 28(11) (2003), 1539–1641 Tang, W., Santare, M.H., Advani, S.G. ‘Melt processing and mechanical property

© 2007, Woodhead Publishing Limited

230

134.

135.

136.

137.

138.

139. 140.

141.

142. 143.

144.

145.

146.

147. 148.

149.

Nanofibers and nanotechnology in textiles characterization of multi-walled carbon nanotube/high density polyethylene (MWNT/ HDPE) composite films’, Carbon 41(14) (2003), 2779–2785 Coleman, J.N., Cadek, M., Blake, R., Nicolosi, V., Ryan, K.P., Belton, C., Fonseca, A., Nagy, J.B., Gun’ko, Y.K., Blau, W.J. ‘High-performance nanotube-reinforced plastics: understanding the mechanism of strength increase’, Adv. Funct. Mater. 14(8) (2004), 791–798 Putz, K.W., Mitchell, C.A., Krishnamoorti, R., Green, P.F. ‘Elastic modulus of single-walled carbon nanotube/poly(methyl methacrylate) nanocomposites’, J. Polym. Sci. Part B – Polym. Phys. 42(12) (2004), 2286–2293 Carneiro, O.S., Covas, J.A., Bernado, C.A., Caldeira, G., Van Hattum, F.W.J., Ting, J.-M., Alig, R.L., Lake, M.L. ‘Production and assessment of polycarbonate composites reinforced with vapour-grown carbon fibres’, Comp. Sci. Technol. 58(3– 4) (1998), 401–407 Sandler, J., Werner, P., Shaffer, M.S.P., Demchuk, V., Altstädt, V., Windle, A.H. ‘Carbon-nanofibre-reinforced poly(ether ether ketone) composites’, Comp. Part A 33(8) (2002), 1033–1039 Werner, P., Altstädt, V., Jaskulka, R., Jacobs, O., Sandler, J.K.W., Shaffer, M.S.P., Windle, A.H. ‘Tribological behaviour of carbon-nanofibre-reinforced poly(ether ether ketone)’, Wear 257(9–10) (2004), 1006–1014 Zoo, Y.S., An, J.W., Lim, D.P., Lim, D.S. ‘Effect of carbon nanotube addition on tribological behavior of UHMWPE’, Tribol. Lett. 16(4) (2004), 305–309 Lozano, K., Bonilla-Rios, J., Barrera, E.V. ‘A study on nanofiber-reinforced thermoplastic composites II: Investigation of the mixing rheology and conduction properties’, J. Appl. Polym. Sci. 80(8) (2001), 1162–1172 Gordeyev, S.A., Macedo, F.J., Ferreira, J.A., Van Hattum, F.W.J., Bernado, C.A. ‘Transport properties of polymer-vapour-grown carbon fibre composites’, Physica B 279(1–3) (2000), 33–36 Finegan, I.C., Tibbetts, G.G. ‘Electrical conductivity of vapor-grown carbon fiber/ thermoplastic composites’, J. Mater. Res. 16(6) (2001), 1668–1674 Kilbride, B.E., Coleman, J.N., Fraysse, J., Fournet, P., Cadek, M., Drury, A., Hutzler, S., Roth, S., Blau, W.J. ‘Experimental observation of scaling laws for alternating and direct current conductivity in polymer–carbon nanotube composite thin films’, J. Appl. Phys. 92(7) (2002), 4024–4030 Stephan, C., Nguyen, T.P., Lahr, B., Blau, W.J., Lefrant, S., Chauvet, O. ‘Raman spectroscopy and conductivity measurements on polymer-multiwalled carbon nanotubes composites’, J. Mater. Res. 17(2) (2002), 396–400 Bai, J.B., Allaoui, A. ‘Effect of the length and the aggregate size of MWCNTs on the improvement efficiency of the mechanical and electrical properties of nanocomposites – experimental investigation’, Comp. Part A 34(8) (2003), 689– 694 Sandler, J., Shaffer, M.S.P., Prasse, T., Bauhofer, W., Schulte, K., Windle, A.H. ‘Development of a dispersion process for carbon nanotubes in an epoxy matrix and the resulting electrical properties’, Polymer 40(21) (1999), 5967–5971 Allaoui, A., Bai, S., Cheng, H.M., Bai, J.B. ‘Mechanical and electrical properties of a MWNT/epoxy composite’, Comp. Sci. Technol. 62(15) (2002), 1993–1998 Barrau, S., Demont, P., Peigny, A., Laurent, C., Lacabanne, C. ‘DC and AC conductivity of carbon nanotubes–polyepoxy composites’, Macromolecules 36(14) (2003), 5187–5194 Prasse, T., Cavaille, J.-Y., Bauhofer, W. ‘Electric anisotropy of carbon nanofibre/

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

150.

151.

152.

153. 154.

155.

156.

157.

158.

159.

160.

161.

162.

231

epoxy resin composites due to electric field induced alignment’, Comp. Sci. Technol. 63(13) (2003), 1835–1841 Valentini, L., Puglia, D., Frulloni, E., Armentano, I., Kenny, J.M., Santucci, S. ‘Dielectric behavior of epoxy matrix/single-walled carbon nanotube composites’, Comp. Sci. Technol. 64(1) (2004), 23–33 Martin, C.A., Sandler, J.K.W., Shaffer, M.S.P., Schwarz, M.-K., Bauhofer, W., Schulte, K., Windle, A.H. ‘Formation of percolating networks in multi-wall carbonnanotube-epoxy composites’, Comp. Sci. Technol. 64(15) (2004), 2309–2316 Martin, C.A., Sandler, J.K.W., Windle, A.H., Schwarz, M.-K., Bauhofer, W., Schulte, K., Shaffer, M.S.P. ‘Electric field-induced aligned multi-wall carbon nanotube networks in epoxy composites’, Polymer 46(3) (2005), 877–886 Fu, S.-Y., Mai, Y.-W. ‘Thermal conductivity of misaligned short-fiber-reinforced polymer composites’, J. Appl. Polym. Sci. 88(6) (2003), 1497–1505 Bangarusampath, D.S., Ruckdäschel, H., Sandler, J.K.W., Altstädt, V., Shaffer, M.S.P. ‘Melt rheology of carbon nanofibre-reinforced poly(ether ether ketone) under shear and elongational flow’, J. Rheol (2007), submitted to J. Rheol Rahatekar, S.S., Koziol, K.K.K., Butler, S.A., Elliott, J.A., Shaffer, M.S.P., Mackley, M.R., Windle, A.H. ‘Optical microstructure and viscosity enhancement for an epoxy resin matrix containing multiwall carbon nanotubes’, J. Rheol. 50(5) (2006), 599–610 Pugno, N.M. ‘On the strength of the carbon nanotube-based space elevator cable: from nanomechanics to megamechanics’, J. Phys.: Condens. Matter 18(33) (2006), S1971–S1990 Vigolo, B., Penicaud, A., Coulon, C., Sauder, C., Pailler, R., Journet, C., Bernier, P., Poulin, P. ‘Macroscopic fibers and ribbons or oriented carbon nanotubes’, Science 290(5495) (2000), 1331–1334 Dalton, A.B., Collins, S., Munoz, E., Razal, J.M., Ebron, V.H., Ferraris, J.P., Coleman, J.N., Kim, B.G., Baughman, R.H. ‘Super-tough carbon-nanotube fibres – these extraordinary composite fibres can be woven into electronic textiles”, Nature 423(6941) (2003), 703 Munoz, E., Suh, D.S., Collins, S., Selvidge, M., Dalton, A.B., Kim, B.G., Razal, J.M., Ussery, G., Rinzler, A.G., Martinez, M.T., Baughman, R.H. ‘Highly conducting carbon nanotube/polyethyleneimine composite fibers’, Adv. Mater. 17(8) (2005), 1064 Eriscon, L.M., Fan, H., Peng, H.Q., Davis, V.A., Zhou, W., Sulpizio, J., Wang, Y.H., Booker, R., Vavro, J., Guthy, C., Parra-Vasquez, A.N.G., Kim, M.J., Ramesh, S., Saini, R.K., Kittrell, C., Lavin, G., Schmidt, H., Adams, W.W., Billups, W.E., Pasquali, M., Hwang, W.F., Hauge, R.H., Fischer, J.E., Smalley, R.E. ‘Macroscopic, neat, single-walled carbon nanotube fibers’, Science 305(5689) (2004), 1447– 1450 Zhou, W., Vavro, J., Guthy, C., Winey, K.I., Fischer, J.E., Ericson, L.M., Ramesh, S., Saini, R., Davis, V.A., Kittrell, C., Pasquali, M., Hauge, R.H., Smalley, R.E. ‘Single wall carbon nanotube fibers extruded from super-acid suspensions: preferred orientation, electrical, and thermal transport’, J. Appl. Phys. 95(2) (2004), 649– 655 Mottaghitalab, V., Spinks, G.M., Wallace, G.G. ‘The development and characterisation of polyaniline–single walled carbon nanotube composite fibres using 2-acrylamido2 methyl-1-propane sulfonic acid (AMPSA) through one step wet spinning process’, Polymer 47(14) (2006), 4996–5002

© 2007, Woodhead Publishing Limited

232

Nanofibers and nanotechnology in textiles

163. Spinks, G.M., Shin, S.R., Wallace, G.G., Whitten, P.G., Kim, S.I., Kim, S.J. ‘Mechanical properties of chitosan/CNT microfibers obtained with improved dispersion’, Sens. Actuat. B – Chem. 115(2) (2006), 678–684 164. Chae, H.G., Minus, M.L., Kumar, S. ‘Oriented and exfoliated single wall carbon nanotubes in polyacylonitrile’, Polymer 47(10) (2006), 3494–3504 165. Zhang, X.F., Liu, T., Sreekumar, T.V., Kumar, S., Hu, X., Smith, K. ‘Gel spinning of PVA/SWNT composite fiber’, Polymer 45(26) (2004), 8801–8807 166. Gao, J.B., Itkis, M.E., Yu, A.P., Bekyarova, E., Zhao, B., Haddon, R.C. ‘Continuous spinning of a single-walled carbon nanotube-nylon composite fiber’, J. Am. Chem. Soc. 127(11) (2005), 3847–3854 167. Oh, S.-J., Lee, H.-J., Keum, D.-K., Lee, S.-W., Wang, D.H., Park, S.-Y., Tan, L.S., Baek, J.-B. ‘Multi-walled carbon nanotubes and nanofibres grafted with polyetherketones in mild and viscous polymeric acids’, Polymer 47(4) (2006), 1132–1140 168. Ko, F., Gogotsi, Y., Ali, A., Naguib, N., Ye, H.H., Yang, G.L., Li, C., Willis, P. ‘Electrospinning of continuous carbon nanotube-filled nanofiber yarns’, Adv. Mater. 15(14) (2003), 1161 169. Lozano, K., Yang, S.Y., Zeng, Q. ‘Rheological analysis of vapor-grown carbon nanofiber-reinforced polyethylene composites’, J. Appl. Polym. Sci. 93(1) (2004), 155–162 170. Pötschke, P., Abdel-Goad, M., Alig, I., Dudkin, S., Lellinger, D. ‘Rheological and dielectrical characterization of melt mixed polycarbonate-multiwalled carbon nanotube composites’, Polymer 45(26) (2004), 8863–8870 171. McNally, T., Pötschke, P., Halley, P., Murphy, M., Martin, D., Bell, S.E.J., Brennan, G.P., Bein, D., Lemoine, P., Quinn, J.P. ‘Polyethylene multiwalled carbon nanotube composites’, Polymer 46(19) (2005), 8222–8232 172. Ramesh, N.S., Lee, S.T. ‘Blowing agent effect on extensional viscosity calculated from fiber spinning method for foam processing’, SPE Foams ’99 (1999), 85–96 173. Werner, P., Verdejo, R., Wöllecke, F., Altstädt, V., Sandler, J.K.W., Shaffer, M.S.P. ‘Carbon nanofibres allow foaming of semicrystalline poly(ether ether ketone)’, Adv. Mater. 17(23) (2005), 2864–2869 174. Joseph, R., Martyn, M.T., Tanner, K.E., Coates, P.D., Bonfield, W. ‘Rheological characterisation of hydroxyapatite filled polyethylene composites Part 1–Shear and extensional behaviour’, Plast. Rubber Comp. 30(5) (2001), 197–204 175. Ramazani, A., Ait-Kadi, A., Grmela, M. ‘Rheological modelling of short fiber thermoplastic composites’, J. Non-Newton. Fluid Mech. 73(3) (1997), 241–260 176. Krishnamoorti, R., Ren, J.X., Silva, A.S. ‘Shear response of layered silicate nanocomposites’, J. Chem. Phys. 114(11) (2001), 4968–4973 177. Hu, G., Zhao, C., Zhang, S., Yang, M., Wang, Z. ‘Low percolation thresholds of electrical conductivity and rheology in poly(ethylene terephthalate) through the networks of multi-walled carbon nanotubes’, Polymer 47(1) (2006), 480–488 178. Du, F., Scogna, R.C., Zhou, W., Brand, S., Fischer, J.E., Winey, K.I. ‘Nanotube networks in polymer nanocomposites: rheology and electrical conductivity’, Macromolecules 37(24) (2004), 9048–9055 179. Kharchenko, S.B., Douglas, J.F., Obrzut, J., Grulke, E.A., Migler, K.B. ‘Flowinduced properties of nanotube-filled polymer materials’, Nature Mater. 3(8) (2004), 564–568 180. Guth, E. ‘Theory of filler reinforcement’, J. Appl. Phys. 16(1) (1945), 20–25 181. Ma, H., Zeng, J., Realff, M.L., Kumar, S., Schiraldi, D.A. ‘Processing, structure,

© 2007, Woodhead Publishing Limited

Carbon nanotube and nanofibre reinforced polymer fibres

182.

183. 184.

185. 186.

187. 188. 189.

190.

191.

192.

193.

194. 195.

196. 197.

198. 199.

233

and properties of fibers from polyester/carbon nanofiber composites’, Comp. Sci. Technol. 63(11) (2003), 1617–1628 Stange, J., Münstedt, H. ‘Rheological properties and foaming behaviour of polypropylenes with different molecular structures’, J. Rheol. 50(6) (2006), 907– 923 Han, C.D., Kim, Y.W. ‘Studies on melt spinning. 5. Elongational viscosity and spinnability of two-phase systems’, J. Appl. Polym. Sci. 18(9) (1974), 2589–2603 Münstedt, H., Steffl, T., Malmberg, A. ‘Correlation between rheological behaviour in uniaxial elongation and film blowing properties of various polyethylenes’, Rheol. Acta 45(1) (2005), 14–22 Meissner, J. ‘Dehnungsverhalten von Polyäthylen-Schmelzen’, Rheol. Acta 10 (1971), 230–242 Muke, S., Ivanov, I., Kao, N., Bhattacharya, S.N. ‘Extensional rheology of polypropylene melts from the Rheotens test’, J. Non-Newton. Fluid Mech. 101(1– 3) (2001), 77–93 Wagner, M.H., Bernnat, A., Schulze, V. ‘The rheology of the rheotens test’, J. Rheol. 42(4) (1998), 917–928 Laun, H.M., Schuch, H. ‘Transient elongational viscosities and drawability of polymer melts’, J. Rheol. 33(1) (1989), 119–175 Prasad, R., Pasanovic-Zujo, V., Gupta, R.K., Cser, F., Bhattacharya, S.N. ‘Morphology of EVA based nanocomposites under shear and extensional flow’, Polym. Eng. Sci. 44(7) (2004), 1220–1230 Sandler, J., Windle, A.H., Werner, P., Altstädt, V., van Es, M., Shaffer, M.S.P. ‘Carbon-nanofibre-reinforced poly(ether ether ketone) fibres’, J. Mater. Sci. 38(10) (2003), 2135–2141 Sennett, M., Welsh, E., Wright, J.B., Li, W.Z., Wen, J.G., Ren, Z.F. ‘Dispersion and alignment of carbon nanotubes in polycarbonate’, Mat. Res. Soc. Symp. Proc. 706 (2002), Z3.31.1–6 Andrews, R., Jacques, D., Rao, A.M., Rantell, T., Derbyshire, F., Chen, Y., Chen, J., Haddon, R.C. ‘Nanotube composite carbon fibers’, Appl. Phys. Lett. 75(9) (1999), 1329–1331 Bhattacharyya, A.R., Sreekumar, T.V., Liu, T., Kumar, S., Ericson, L.M., Hauge, R.H., Smalley, R.E. ‘Crystallization and orientation studies in polypropylene/single wall carbon nanotube composite’, Polymer 44(8) (2003), 2373–2377 Marcincin, A. ‘Modification of fiber-forming polymers by additives’, Prog. Polym. Sci. 27(5) (2002), 853–913 Coleman, J.N., Cadek, M., Ryan, K.P., Fonseca, A., Nagy, J.B., Blau, W.J., Ferreira, M.S. ‘Reinforcement of polymers with carbon nanotubes. The role of an ordered polymer interfacial region. Experiment and modelling’, Polymer 47(26) (2006), 8556–8561 Young, R.J., Eichhorn, S.J. ‘Deformation mechanisms in polymer fibres and nanocomposites’, Polymer 48(1) (2007), 2–18 Liu, T.X., Wang, S.E., Mo, Z.S., Zhang, H.F. ‘Crystal structure and drawinginduced polymorphism in poly(aryl ether ether ketone). IV’, J. Appl. Polym. Sci. 73(2) (1999), 237–243 Kumar, S., Doshi, H., Srinivasarao, M., Park, J.O., Schiraldi, D.A. ‘Fibers from polypropylene/nano carbon fiber composites’, Polymer 43(5) (2002), 1701–1703 Chatterjee, A., Deopura, B.L. ‘High modulus and high strength PP nanocomposite filament’, Comp. Part A 37(5) (2006), 813–817

© 2007, Woodhead Publishing Limited

234

Nanofibers and nanotechnology in textiles

200. Fisher, F.T., Bradshaw, R.D., Brinson, L.C. ‘Effects of nanotube waviness on the modulus of nanotube-reinforced polymers’, Appl. Phys. Lett. 80(24) (2002), 4647– 4649 201. Coleman, J.N., Khan, U., Gun’ko, Y.K. ‘Mechanical reinforcement of polymers using carbon nanotubes’, Adv. Mater. 18(6) (2006), 689–706 202. Gao, H.J., Ji, B.H., Jäger, I.L., Arzt, E., Fratzl, P. ‘Materials become insensitive to flaws at nanoscale: lessons from nature’, Proc. Natl Acad. Sci. USA 100(10) (2003), 5997–5600 203. Zhao, Q., Wagner, H.D. ‘Raman spectroscopy of carbon-nanotube-based composites’, Proc. Trans. Royal Soc. London Ser. A 362(1824) (2004), 2407–2424 204. Cooper, C.A., Young, R.J., Halsall, M. ‘Investigation into the deformation of carbon nanotubes and their composites through the use of Raman spectroscopy’, Comp. Part A 32(3–4) (2001), 401–411 205. Zhao, Q., Wood, J.R., Wagner, H.D. ‘Stress fields around defects and fibers in a polymer using carbon nanotubes as a sensor’, Appl. Phys. Lett. 78(12) (2001), 1748–1750 206. Chen, X.M., Yoon, K.W., Burger, C., Sics, I., Fang, D.F., Hsiao, B.S., Chu, B. ‘Insitu X-ray scattering studies of a unique toughening mechanism in surface-modified carbon nanofiber/UHMWPE nanocomposite films’, Macromolecules 38(9) (2005), 3883–3893 207. Kelarakis, A., Yoon, K., Sics, I., Somani, R.H., Hsiao, B.S., Chu, B. ‘Uniaxial deformation of an elastomer nanocomposite containing modified carbon nanofibers by in situ synchrotron X-ray diffraction’, Polymer 46(14) (2005), 5103–5117 208. Karst, D., Yang, Y.Q. ‘Potential advantages and risks of nanotechnology for textiles’, AATCC Rev. 6(3) (2006), 44–48 209. Huang, C.-Y., Wu, C.-C. ‘The EMI shielding effectiveness of PC/ABS/nickelcoated-carbon-fibre composites’, Europ. Polym. J. 36(12) (2000), 2729–2737 210. Hirase, R., Hasegawa, M., Shirai, M. ‘Conductive fibers based on poly(ethylene terephthalate)–polyaniline composites manufactured by electrochemical polymerization’, J. Appl. Polym. Sci. 87(7) (2003), 1073–1078 211. Pötschke, P., Brünig, H., Janke, A., Fischer, D., Jehnichen, D., ‘Orientation of multi-walled carbon nanotubes in composites with polycarbonate by melt spinning’, Polymer 46(23) (2005), 10355–10363 212. Zhang, Q.H., Jin, H.F., Wang, X.H., Jing, X.B. ‘Morphology of conductive blend fibers of polyaniline and polyamide-11’, Synth. Metals 123(3) (2001), 481–485 213. Gordeyev, S.A., Ferreira, J.A., Bernardo, C.A., Ward, I.M. ‘A promising conductive material: highly oriented polypropylene filled with short vapour-grown carbon fibres’, Mat. Lett. 51(1) (2001), 32–36 214. Zeng, J., Saltysiak, B., Johnson, W.S., Schiraldi, D.A., Kumar, S. ‘Processing and properties of poly(methyl methacrylate)/carbon nano fiber composites’, Comp. Part B 35(2) (2004), 173–178 215. Chae, H.G., Sreekumar, T.V., Uchida, T., Kumar, S. ‘A comparison of reinforcement efficiency of various types of carbon nanotubes in polyacrylonitrile fiber’, Polymer 46(21) (2005), 10925–10935 216. Schartel, B., Pötschke, P., Knoll, U., Abdel-Goad, M. ‘Fire behaviour of polyamide6/ multiwall carbon nanotube nanocomposites’, Europ. Polym. J. 41(5) (2005), 1061– 1070

© 2007, Woodhead Publishing Limited

8 Structure and properties of carbon nanotube-polymer fibers using melt spinning R. E. G O R G A, North Carolina State University, USA

8.1

Introduction

The fabrication of advanced fibers for protective applications continues to be of utmost interest. This chapter focuses on the mechanical and morphological properties of polymer/carbon nanotubes composites via melt processing. In particular, this chapter will focus on (1) melt processing optimization for improved nanotube dispersion; (2) the effect of fiber draw ratio on morphology and mechanical properties; (3) the effect of nanotube type and geometry on morphology and mechanical properties; and (4) details on future trends for the technology and potential applications as well as a section outlining further resources for the interested reader. Carbon nanotubes are graphitic sheets rolled into seamless tubes (i.e. arrangements of carbon hexagons into tube-like fullerenes) and have diameters ranging from about a nanometer to tens of nanometers with lengths up to centimeters. Nanotubes have received much attention due to their interesting properties (high modulus and electrical/thermal conductivity) since their discovery by Iijima in 1991.1, 2 For example, theoretical calculations and preliminary experimentation have shown that carbon nanotubes have excellent mechanical properties, electrical conductivity (5.1 × 10–6 to 5.8 Ω cm),3 and thermal conductivity (1750–5800 W/m K).4 Treacy et al.5 found the Young’s modulus of individual nanotubes to be in the range of 1 TPa using intrinsic thermal vibrations. Although carbon nanotubes show exceptional properties on the nanoscale, the difficulty lies in creating a material that exhibits carbon nanotube properties on the macro-scale. Incorporating the nanotubes as filler into polymer matrices is the most common method currently explored.6–15 Similar to other composites made from chopped fiber in a polymer matrix, filler dispersion and orientation are essential to achieve optimal property improvements. Researchers have used many different techniques to attempt to disperse nanotubes in polymer matrices including solution chemistry to functionalize the nanotube surface,16–21 the use of polymers to coat the nanotube 235

© 2007, Woodhead Publishing Limited

236

Nanofibers and nanotechnology in textiles

surface,22 in situ polymerization of the nanocomposite,23, 24 ultrasonic dispersion in solution, 14, 25 melt processing, 10, 26–30 the use of surfactants, 21, 31 electrospinning,32 gelation/crystallization33 and electrode chemistry.34 Furthermore, research has shown improved mechanical properties via nanotube orientation by melt drawing after melt compounding in a poly(methyl methacrylate) (PMMA) matrix with low levels (~1 wt%) of multiwall carbon nanotubes (MWNTs).27 Polypropylene (PP) has been a common matrix used for carbon nanotube composites. Both single wall26, 35–38 and multiwall11, 39–41 nanotubes have been used to study crystallization behavior in PP and polymer morphology as well as mechanical properties. Results have been mixed, especially for mechanical properties, where one study26 showed no significant improvement in mechanical properties, and others have shown moderate improvements in tensile strength, but decreased toughness.38

8.2

Producing carbon nanotube-polymer fibers

Exxon Mobil Type 3155 Polypropylene with a melt flow rate of 36 g/10 min, density of 0.9 g/cm3, and molecular weight distribution of 2.8, was utilized as the matrix. MWNTs were supplied by Nano-Lab (Newton, MA). These nanotubes (with purity > 95%) were produced via plasma-enhanced chemical vapor deposition (PE-CVD) using acetylene and ammonia with iron catalyst particles on a mesoporous silica substrate.42 The diameter is specified as 20–50 nm with lengths ranging between 5 and 20 µm. The PP pellets were ground into a powder using a SPEX® CertiPrep Freezer Mill (Metuchen, NJ). The pellets were precooled for 5 min followed by three 2 min grinding cycles at 10 Hz. Between each cycle the sample was cooled for a 1 min interval. Nanocomposites were created by dry blending the PP powder with a given ratio of MWNTs (0.25, 0.50, 1 and 3 wt%). The preblend was then fed into a Haake Mini-Lab twin screw extruder and processed for 10 min at 100 rpm and 200 °C (conditions previously optimized via a controlled experiment). After 10 min, the composite was extruded through a 1.75 mm cylindrical die. Oriented samples were created using a specially designed winding apparatus with melt and solid drawing capabilities. The winding apparatus consists of three PP rollers each driven by variable speed motors. The center to center distance for the first and second rollers is 15 in (380 mm), and for the second and third rollers is 24.75 in (630 mm). Each motor is capable of linear speeds on the outer roller surface of 45–450 in/min (1.1–11 m/min). The rollers can all be controlled independently. For melt drawing, the material is drawn using the first roller and the sample is collected using the third roller. Using this procedure, meltdrawn samples had a nominal draw ratio of 12:1. In coupled melt/solid

© 2007, Woodhead Publishing Limited

Carbon nanotube-polymer fibers using melt spinning

237

drawing, the material is wound around the first roller at one speed, and then wound around the second roller at a faster speed, and is collected by the third roller at the same speed as that for the second roller. Using this procedure, coupled melt/solid-drawn samples had nominal draw ratios of 23:1 and 24:1. After extruding both oriented and unoriented samples (i.e. no drawing via the winding apparatus), the mechanical and morphological properties were characterized using the techniques described in the following sections.

8.3

Thermal characterization

A Perkin Elmer Pyris 1 Thermal Gravimetric Analyzer (TGA) was used to determine the weight percent concentration of nanotubes in the composite samples. Approximately 20 mg of each sample was heated from 25 to 950 °C at a rate of 20 °C/min in a nitrogen environment. As the sample is heated, the mass is measured as function of temperature. The mass retained is calculated by dividing the mass at the temperature of interest by the initial mass. Once the polymer has degraded, the remaining mass is assumed to be the mass of the MWNTs since the MWNTs are thermally stable in nitrogen to temperatures above 1000 °C (from discussion with David Carnahan, President of NanoLab). Owing to the high thermal stability of carbon nanotubes, the PP matrix will degrade several hundred degrees before nanotubes in a nitrogen environment (from discussion with David Carnahan, President of NanoLab). Figure 8.1 shows a comparative plot of the 100% PP sample with nanocomposites ranging from 0.25 to 3 wt% MWNTs. Owing to the thermal

100

Weight [%]

80

60

40

20

0 100

0% MWNT 0.25% MWNT 0.5% MWNT 1% MWNT 3% MWNT 150

200

250

300 350 400 450 Temperature [°C]

500

550

600

650

8.1 TGA plot of PP nanocomposites as a function of MWNT loading.

© 2007, Woodhead Publishing Limited

238

Nanofibers and nanotechnology in textiles

stability of the nanotubes, the 3 wt% composite sample shows thermal stability (i.e. no mass loss) to 192 °C (80 °C higher than that for pure PP).

8.3.1

Differential scanning calorimetry

The crystallinity of the composite samples was studied using a Perkin Elmer Diamond differential scanning calorimeter. Each sample was heated from 25 to 200 °C at a rate of 10 °C/min. The thermograms were used to determine the onset melting temperature, peak melting temperature, peak area and enthalpy of melting (∆H). To achieve more accurate values for these temperatures, a straight baseline was drawn connecting each flat side of the melting peak. The shapes of the curves at different loading levels were evaluated qualitatively to determine changes in crystal structure. Finally, the overall percent crystallinity was calculated by dividing the enthalpy of melting for the sample (∆H) by the enthalpy of melting for 100% crystalline PP ( ∆H fo = 207.1 J/g).43 Figure 8.2 shows the melting endotherms from differential scanning calorimetry (DSC) for 0, 1 and 3 wt% drawn samples (to a 12:1 draw ratio), respectively. The crystallinity was calculated using a ∆H fo value of 207.1 J/g for 100% crystalline PP43 and Equation 8.1, shown below (where ∆H is the enthalpy measured from the experiment). The percent crystallinity, onset temperature and melting peak temperature are tabulated in Table 8.1. % Crystallinity = ∆Ho × 100 ∆H f

[8.1]

Figure 8.2 shows the melting peak has a similar shape at the different loading levels. Table 8.1 shows little quantitative change in the overall percent crystallinity and peak melting temperature. The addition of up to 3 wt% MWNTs in PP did not significantly alter the overall crystallinity as observed by DSC, which is consistent with results from the literature.26, 39, 40

8.4

Fiber morphology

The morphology of the PP/nanotube composite samples was observed both qualitatively and quantitatively. The dispersion and orientation of the nanotubes was verified through transmission electron microscopy (TEM) and scanning electron microscopy (SEM). Finally, the polymer crystal structure and orientation was investigated quantitatively through wide-angle X-ray diffraction (WAXD).

8.4.1

Transmission electron microscopy

A Jeol Inc. model 100S transmission electron microscope operating at 100 kV was utilized to view the orientation of the nanotubes. Both drawn and undrawn

© 2007, Woodhead Publishing Limited

Carbon nanotube-polymer fibers using melt spinning

239

–54.0 Peak = 167.43 °C –54.5

Heat flow [mW]

–55.0 –55.5 –56.0

Onset = 154.5 °C

–56.5

Area = 225.688 mJ Delta H = 78.364 J/g

–57.0 –57.3 110.3 120

130

140 150 160 170

180 190 206.3

Temperature °C (a) –52.39 –53.0 Peak = 165.93 °C

–53.5 Heat flow [mW]

–54.0 –54.5

Area = 368.589 mJ Delta H = 79.954 J/g

–55.0 –55.5 –56.0

Onset = 155.73 °C

–56.5 –57.0 –57.5 –57.78 120.2

130

140

150 160 170 Temperature °C (b)

180

190

199

–52.66 –53.0 Peak = 166.63 °C

Heat flow [mW]

–53.5 –54.0

Area = 270.243 mJ Delta H = 83.408 J/g

–54.5 –55.0 –55.5 –56.0

Onset = 137.73 °C

–56.5 –56.89 120.3

8.2 DSC melting 130

© 2007, Woodhead Publishing Limited

140

150 160 Temperature °C (c)

170

180

188.9 peaks for (a) 0 wt%,

(b) 1 wt%, and (c) 3 wt% MWNTs in PP.

240

Nanofibers and nanotechnology in textiles Table 8.1 Percentage crystallinity and peak melting temperature Figure 8.2

MWNT (wt%)

Draw ratio

Crystallinity (%)

Peak melting temperature (°C)

(a) (b) (c)

0 1 3

12:1 12:1 12:1

38 39 40

167 166 167

100 nm

100 nm

(a)

(b)

8.3 TEM images of 3 wt % MWNT in PP: (a) unoriented, (b) oriented, 12:1 draw ratio.

samples were observed to determine the extent of orientation for the nanotubes. Ultrathin cross-sections of the fiber sample were microtomed and placed on standard TEM grids. Images were captured at magnifications of 25 000× and 50 000×. Figure 8.3 shows TEM images of 3 wt% MWNTs in PP both unoriented (no draw) and oriented at 12:1 draw ratio. These images depict the crosssection of the fiber parallel to the drawing direction. Comparing the two images reveals how melt drawing the nanocomposite orients the nanotubes. In the unoriented image (Fig. 8.3a) whole nanotubes lying at various oblique angles are observed; whereas the image of the drawn sample (Fig. 8.3b) shows nanotubes that run parallel to the fiber axis, therefore indicating flowinduced orientation along the fiber axis.

8.4.2

Scanning electron microscopy

A Jeol Inc. Field Emission model J600F scanning electron microscope, operating at 5 kV, was utilized to further examine the orientation of the nanotubes. Both drawn and undrawn samples were imaged to determine the effect of melt drawing on nanotube orientation. The samples were cleaved

© 2007, Woodhead Publishing Limited

Carbon nanotube-polymer fibers using melt spinning

241

with a sharp razor normal to the fiber axis. Next, the samples were coated with an Au/Pd film approximately 100 Å thick to reduce charging. After coating, they were mounted with the microscope looking down the fiber axis, depicting the cross-section of the fiber normal to the drawing direction. Images were captured at magnifications of 20 000×, 30 000× and 40 000×. Figure 8.4 shows SEM images of 3 wt% MWNTs in PP both unoriented (no draw) and oriented at 12:1 draw ratio. These images depict the crosssection of the fiber normal to the drawing direction. Comparing the two SEM images further reveals how drawing the nanocomposite orients the nanotubes. Since the samples are cleaved, the fracture plane will propagate along the weakest section of the composite, namely the interface between the nanotube and the matrix. In the case where the nanotubes are unoriented, the fracture plane will contain nanotubes lying at various oblique angles. As is shown in Fig. 8.4a, whole nanotubes are observed indicating the random orientation of the nanotubes, whereas, the image of the drawn sample (Fig. 8.4b) shows only the tips of the nanotubes (which run parallel to the fiber axis), therefore indicating flow-induced orientation along the fiber axis. Qualitative evidence of nanotube alignment is observed in both the transmission and scanning electron microscopy images. Therefore as the material is being extruded and melt-drawn, the extensional flow causes nanotube orientation along the fiber axis. As a result, the maximum load transfer can be achieved, leading to the property improvements discussed in Section 8.5.

8.4.3

Wide-angle X-ray diffraction

A Bruker AXS wide-angle X-ray diffractometer with a Cu Kα average source (λ = 1.5418 Å) was utilized to look at the crystalline structure of the overall composite sample and to calculate the Herman orientation factor. The samples were run in transmission and reflection modes. In the reflection mode, the

(a)

(b)

8.4 SEM images of 3 wt% MWNT in PP: (a) unoriented, (b) oriented, 12:1 draw ratio.

© 2007, Woodhead Publishing Limited

242

Nanofibers and nanotechnology in textiles

fibers were mounted on a 1 cm × 1 cm silicon wafer. When mounting the sample in the unit, the fiber axis was oriented in the vertical direction. In this mode, the beam hits the sample at a predetermined angle, and the X-rays diffract into the detector creating the diffraction pattern. The instrument was set to run the scan with ω set at 15°. The scan was run for 900 seconds. From the diffraction pattern, the intensity is measured as a function of 2Θ, and plots of intensity versus 2Θ are created. In the transmission mode, the fiber axis is oriented in the horizontal direction. The beam passes through the sample, and the X-rays diffract off the sample to the detector, creating cones seen as rings on the diffraction pattern. From these diffraction patterns, the intensity around each ring is measured and plotted against χ, the angle around the ring starting at 6 o’clock moving counter-clockwise. The data were then corrected for background scattering. Using ImageJ software, the transmission images were transformed to plots of intensity versus scattering vector, q. Figure 8.5 shows the 2D wide-angle X-ray diffraction patterns and corresponding scattering vector versus intensity plots for drawn samples (12:1) at 0 and 3 wt% MWNT loading. The plots both reveal the four peaks corresponding to the four-ring pattern common to PP fiber X-ray diffraction. However, all four peaks for the 3 wt% sample (Fig. 8.5b) are more distinct and intense than the peaks for the 0 wt% sample (Fig. 8.5a). Similar transitions for intensity were observed by Broda44 in pure PP fiber extruded with increasing take-up speeds. The study concluded the decrease in intensity resulted from the presence of a mesophase coupled with the PP α phase. Based on comparison with Broda’s X-ray diffraction plots, the 0 wt% sample contains mesophase coupled with α phase PP crystals. However, the 3 wt% sample contains only α phase PP crystals indicating a more ordered crystal structure as a result of the aligned MWNTs. Although different PP crystal phases can be responsible for changes in the mechanical properties,43 the strengthening mechanism proposed here is primarily due to load transfer from the polymer matrix to the nanotube. Using the intensity as a function of angle, χ, around each diffraction ring, Herman’s orientation factor can be determined. Herman’s orientation factor (P2) is defined in Equation 8.2:

3 < cos 2 χ > – 1 [8.2] 2 In Equation 8.2, is the average cosine squared value for the diffraction ring and is calculated using the following equation: P2 =

90

< cos 2 χ > =

Σ I i cos 2 χ i sin χ i i =0 90

Σ I i sin χ i

i =0

© 2007, Woodhead Publishing Limited

[8.3]

Carbon nanotube-polymer fibers using melt spinning

243

40 35 30

Intensity

25 20 15 10 5 0 0.4

0.5

0.6

0.7

0.8 q/q*

0.9

1.0

1.1

1.2

1.0

1.1

(a) 40 35 30

Intensity

25 20 15 10 5 0 0.4

0.5

0.6

0.7

0.8

0.9

q/q* (b)

8.5 WAXD patterns for (a) 0 wt% and (b) 3 wt% MWNTs in PP with 12:1 draw ratio.

In Equation 8.3, Ii and χi are the intensity and angle at the ith (0.5° step) position along the diffraction ring. The diffraction data were corrected to shift the background intensity to a value of zero. After correcting the background, Herman’s orientation factors for each crystallographic ring of a

© 2007, Woodhead Publishing Limited

244

Nanofibers and nanotechnology in textiles

Table 8.2 Herman’s orientation factors Sample

110

040

130

13 1

0% 0% 3% 3%

0.42 0.76 –0.10 0.43

–0.22 –0.37 –0.08 –0.34

–0.08 –0.33 –0.11 –0.26

–0.18 –0.24 –0.27 –0.22

Unoriented Oriented Unoriented Oriented

Table 8.3 ∆P2 values Ring

0 wt% MWNT

3 wt% MWNT

110 040 130 13 1

0.34 –0.15 –0.25 –0.06

0.53 –0.26 –0.15 0.05

virgin PP sample and a 3 wt% MWNT loaded sample were calculated using Equations 8.2 and 8.3. The results are tabulated in Table 8.2. Herman’s orientation factor can vary between –0.5 to 1. If the factor is –0.5, the crystal plane is oriented perpendicular to the reference direction of χ = 0, the direction of draw, whereas a factor of 1 denotes orientation parallel to the reference direction. Random orientation is observed when the factor equals 0. Herman’s orientation analysis can be used to quantify crystal orientation as a function of nanotube concentration and polymer and nanotube orientation. In the as-extruded (or undrawn) samples, the nanotube addition promotes isotropization of the composite as demonstrated by the P2 coefficients being slightly negative but close to zero (for nanotube filled samples). Further examination of the change in the P2 coefficients (∆P2 = P2,oriented – P2,unoriented) from the unoriented to the oriented state demonstrates the effects of drawing both pure PP and filled PP (with 3 wt% MWNT). The ∆P2 factors are tabulated in Table 8.3. As shown by Table 8.3, drawing the fiber samples results in improved orientation overall. The 110 crystal planes become more oriented along the fiber axis, whereas the 040 and 130 planes become more oriented perpendicular to the fiber axis. However, the 13 1 plane remains randomly oriented. It is germane to note that the orientation along the fiber axis for the 110 plane is more highly oriented for the nanotube loaded sample. Likewise, the orientation perpendicular to the fiber direction for the 040 plane is more highly oriented in the nanotube loaded sample. This indicates that the crystal structure becomes more oriented with nanotube orientation, whereas the opposite is true of the 130 plane.

© 2007, Woodhead Publishing Limited

Carbon nanotube-polymer fibers using melt spinning

8.5

245

Mechanical properties of fibers

Tensile tests were preformed to measure the tensile toughness (MJ/m3), modulus (MPa) and yield strength (MPa) of the oriented and unoriented samples at each loading level. The tests were performed on an Instron® Model 5544 using the Bluehill® Version 1.00 software package. Ten replicates of each sample were run to obtain average, standard deviation and standard error values. Each experiment was conducted with a gauge length of 25.4 mm, crosshead speed of 25.4 mm/min and data acquisition rate of 50 points/ second. To obtain the elastic modulus, a linear regression technique was utilized to define the slope of the stress–strain curve in the initial region before yield. The toughness was calculated using the product of the energy at break at 98% maximum load and the sample volume. The energy at break was defined as the area under the force–elongation curve up to the break point, which was defined as 98% of the maximum load. The volume was defined for a cylinder with initial diameter, d, and initial length, l, as follows: V = π∗(d2/4)*l

[8.4]

Therefore, the defined calculation for toughness equals the energy at break divided by the volume of the sample. The yield strength was calculated using the slope threshold algorithm from the BlueHill® software package. In the algorithm, the yield is calculated as the point where the slope of the stress/ strain curve decreases to a user-selected percentage of the modulus slope. After experimenting with several different percentages, the value chosen was 2%, because the marker placed on the stress–strain curve most closely approximated the appropriate location for yield in a polymer system. Figure 8.6 shows a representative set of stress–strain curves for the 12:1 melt-drawn samples at each nanotube concentration. The addition of nanotubes significantly alters the stress–strain behavior of the fibers. The ultimate stress, yield stress and modulus increase with the addition of nanotubes. In contrast, the ultimate elongation slightly decreases with the addition of nanotubes. The significant increases in ultimate and yield stress combined with a small decrease in ultimate elongation lead to the observed increases in toughness. Figure 8.7 shows the results for the tensile toughness (Fig. 8.7a), modulus (Fig. 8.7b) and yield stress (Fig. 8.7c) as a function of nanotube concentration (from 0 to 3 wt%) and orientation (no draw, or unoriented, and oriented with a draw ratio of 12:1, 23:1 and 24:1). The white bars represent undrawn samples, the black bars represent fibers melt-drawn to a 12:1 ratio, the vertical striped bars represent fibers that are melt and then cold drawn to a 23:1 ratio, and the horizontal striped bars represent fibers coupled melt/solid drawn to a 24:1 ratio. The error bars are based on the standard error of ten samples tested at each concentration level. Because of the difficulty of drawing the PP/3 wt% MWNT samples to the 24:1 draw ratio, homogeneous samples could not be collected. In both the tensile toughness and modulus, optimal

© 2007, Woodhead Publishing Limited

246

Nanofibers and nanotechnology in textiles

140 120

Stress [MPa]

100 80 60 40 20 0 0

200

400

600

800 1000 Strain [%]

1200

1400

1600

1800

8.6 Stress–strain curves for 12:1 melt-drawn fibers (䉫 PP, 䊐: PP/0.25, 䉭: PP/0.5, × PP/1, 䊊 PP/3).

loading was achieved at 0.25 wt%. For the yield strength, the maximum occurred at about 1 wt%. As observed in Fig. 8.7a, the toughness of the unoriented (no draw) material is so low at all concentrations that it does not show up on the scale of the figure. Therefore, the orientation of PP alone improves tensile toughness. However, with the addition of nanotubes, the oriented samples show a statistically significant improvement at concentrations of 0.25–0.5 wt% MWNTs. As proposed by Gorga and Cohen27 the toughness increase in amorphous materials results from the crack bridging ability of the nanotubes. As microcracks form, the nanotubes bridge the gap formed by the crack, slowing down the crack propagation and increasing the overall material toughness. Synonymously, in semicrystalline materials, such as polypropylene, the nanotubes act as tie molecules between crystalline regions rather than crack bridges, thereby also acting as tougheners. At concentrations greater than 0.5 wt%, toughness decreases. The decrease results as the nanotubes begin to aggregate, forming stress concentrations similar to those caused by voids in composite systems as discussed by Wilbrink et al.45 Similarly, Fig. 8.7b shows an increasing modulus with concentrations up to 0.25 wt% followed by a decrease for both undrawn and drawn samples. Figure 8.7b shows orientation of PP alone increases the modulus but not to the level attained with the addition of nanotubes. Adding low levels of nanotubes to PP improves the mechanical properties up to an optimal concentration (0.25 wt%) above which nanotube aggregation leads to property decreases via stress concentrations (as discussed above). Although modulus decreases above 0.25 wt%, the filled samples with orientation exhibit improved modulus

© 2007, Woodhead Publishing Limited

Carbon nanotube-polymer fibers using melt spinning

247

800 No draw 12:1 23:1 24:1

Tensile toughness [MJ/m3]

700 600 500 400 300 200 100 0 PP

PP/0.25

PP/0.5

PP/1

PP/3

(a) 1600

No draw 12:1 23:1 24:1

1400

Modulus [MPa]

1200 1000 800 600 400 200 0 PP

PP/0.25

PP/0.5 (b)

PP/1

No draw 12:1 23:1 24:1

50 45 40 Yield stress [MPa]

PP/3

35 30 25 20 15 10 5 0 PP

PP/0.25

PP/0.5 (c)

PP/1

PP/3

8.7 (a) Tensile toughness (MJ/m3), (b) modulus (MPa) and (c) yield stress (MPa) as a function of nanotube loading in PP for several draw ratios.

© 2007, Woodhead Publishing Limited

248

Nanofibers and nanotechnology in textiles

over that of the drawn PP for all nanotube concentrations studied. The modulus increase can either be attributed to load transfer (from matrix to particle) or crystallinity changes in PP due to the addition of MWNTs (although none was detected via DSC). According to the rule of mixtures, the maximum modulus should reach 2.2 GPa for a 0.25 wt% sample, assuming nanotube modulus of 1 TPa and PP modulus of 550 MPa. However, owing to less than optimal load transfer, the maximum modulus achieved at a loading of 0.25 wt% is 1.3 GPa. As shown by electron microscopy (Sections 8.4.1 and 8.4.2), the nanotubes are not perfectly aligned. Poor adhesion to the matrix material and imperfections and defects in the nanotube structure will also result in a reduced composite modulus. Likewise, Fig. 8.7c shows statistically significant increases in the yield stress with the addition of nanotubes to PP. Furthermore, the increase to a maximum value followed by a decrease is also observed. However, the maximum occurs at a higher concentration range of about 0.5–1 wt% MWNTs. Additionally, the increase is not as significant and pronounced as in the case of the toughness and modulus, indicating that the nanotubes are not as effective at resisting plastic deformation. Additionally, little to no statistical difference between the melt-drawn (12:1) samples and the coupled melt/solid-drawn (23:1, 24:1) samples is observed (as shown by the error bars). However, the toughness for the 0.5 wt% sample is significantly lower for reasons we cannot explain, but appears to be an anomaly. Therefore, the addition of fibers that are melt and then soliddrawn does not affect the values and trends achieved by melt drawing alone. The ability to melt-draw to higher levels could possibly lead to further property improvements. Table 8.4 gives a summary of the average mechanical properties for each loading level and draw ratio.

8.5.1

Dynamic mechanical analysis

Dynamic mechanical analysis (DMA) was used to investigate the viscoelastic properties of the sample. The Q800 Dynamic Mechanical Analyzer from TA Instruments was used for these experiments. The fiber clamp was used for the oriented samples, and the film clamp was used for the unoriented samples. The samples were run using a DMA multi-frequency–strain procedure with a temperature step and frequency sweep. All samples were cylindrical with length l and diameter d. The temperature was stepped from 30 to 150 °C with 20 °C increments. The frequency sweep was 0.01–100 Hz on a logarithmic scale with 3 points per decade. Time–temperature superposition master curves were created by selecting a reference temperature in the middle of the experimental temperature range, 90 °C, and shifting the other curves in relation to the reference curve. The storage modulus, loss modulus and tan δ master curves were analyzed for trends as a function of nanotube loading.

© 2007, Woodhead Publishing Limited

Carbon nanotube-polymer fibers using melt spinning

249

Table 8.4 Summary of mechanical properties for PP and composite samples as a function of nanotube concentration and fiber draw ratio Sample name

MWNT (wt%)

Draw ratio

Tensile toughness (MJ/m3)

Modulus (MPa)

Yield stress (MPa)

PP PP PP PP PP/0.25 PP/0.25 PP/0.25 PP/0.25 PP/0.5 PP/0.5 PP/0.5 PP/0.5 PP/1 PP/1 PP/1 PP/1 PP/3 PP/3 PP/3

0 0 0 0 0.25 0.25 0.25 0.25 0.5 0.5 0.5 0.5 1 1 1 1 3 3 3

0 12:1 23:1 24:1 0 12:1 23:1 24:1 0 12:1 23:1 24:1 0 12:1 23:1 24:1 0 12:1 23:1

0.8 457 397 523 1.1 608 549 512 1.7 620 474 237 0.7 500 355 511 0.9 254 286

257 551 585 611 528 1312 1167 1250 254 852 877 967 266 782 1064 949 238 957 662

15.8 31.8 29.9 33.3 21.2 34.6 33.4 40.3 19.4 38.0 37.7 31.7 15.8 38.8 38.8 40.0 16.4 34.2 32.6

Additionally, the activation energy of the composite fibers was calculated using the Arrhenius equation and was analyzed for trends as a function of nanotube loading. The experiments were performed at several temperature intervals and shifted to a reference temperature (90 °C). The storage modulus was measured as a function of frequency for seven temperatures (ranging from 30 to 150 °C) for the 0.25 wt% MWNT sample. At each temperature the frequency ranged from 0.01 to 100 Hz. Horizontally shifting the curves around a reference temperature, i.e. 90 °C, produced the master curve. The x-axis shift factor, aT, for each temperature was plotted as a function of the temperature. The data can be fit using the Arrhenius equation (Equation 8.5) as a mathematical model (as shown in Fig. 8.8): ln a T =

– Ea  1 – 1 R  T0 T

[8.5]

In Equation 8.5, Ea is the activation energy of the polymer system, aT is the time-based shift factor, R is the gas constant, T is the measurement temperature, and T0 is the reference temperature (in absolute units). As the polymer is strained, it must overcome an energy barrier to allow the chains to move and rotate about the main chain bonds. The energy barrier is known as the activation energy of the polymer system.46 Figure 8.9 shows a plot of this activation

© 2007, Woodhead Publishing Limited

250

Nanofibers and nanotechnology in textiles 10

7.5 5

aT

2.5 0 –2.5 –5 –7.5 –10 290

310

330

350 370 Temperature [K]

390

410

430

8.8 Shift factor as a function of temperature for 0.25 wt% MWNTs in PP (12:1 draw ratio) with Arrhenius model. 450 400 350

Ea [kJ/mole]

300 250 200 150 100 50 0 0

0.25

0.5 MWNT [wt%]

1

3

8.9 Activation energy of fibers as a function of nanotube loading with 12:1 draw ratio.

energy as a function of nanotube loading with errors bars representing the standard error of the Arrhenius model fit for the 0.25 wt% sample at the 12:1 draw ratio.

© 2007, Woodhead Publishing Limited

Carbon nanotube-polymer fibers using melt spinning

251

Similar to the tensile data, a maximum, with the addition of nanotubes, is seen in the activation energy. However, the trend is only marginally statistically significant as shown by the error bars. With the addition of high modulus nanotubes, the resistance to strain should increase similar to that for the toughness (resistance to rupture) and modulus (resistance to deformation). However, as the concentration of nanotubes increases to concentrations greater than 0.5 wt%, aggregates of the nanotubes form, therefore decreasing the reinforcing effect.

8.6

Conclusions and future trends

Dispersion and orientation of MWNTs in PP simultaneously created a tougher and stiffer material. The tensile properties increased as a function of loading to a maximum value at a loading level of 0.25 wt% MWNTs. Toughness increases up to 32% over pure PP were due to the nanotubes acting as tie molecules between the crystalline regions. Modulus increases of up to 138% were due to load transfer from polymer matrix to nanotube. The activation energy, calculated from dynamic mechanical experiments, also revealed a maximum as a function of nanotube loading. Therefore, the mechanical properties (modulus, toughness, yield and activation energy) all exhibited a maximum at a low nanotube loadings (<1 wt% MWNTs). The DSC results showed no change in overall crystallinity. Wide-angle X-ray diffraction indicates a crystal structure change from α phase coupled with mesophase to α phase only with nanotube loading in oriented samples; however, the trend in mechanical properties as a function of nanotube loading supports stiffening as a function of load transfer. In drawn samples, the nanotubes were shown by transmission electron microscopy and scanning electron microscopy to be highly aligned. Although nanotube orientation was revealed through these techniques, isotropization (for randomly dispersed nanotubes) and increased orientation for the melt-drawn (or oriented) samples was quantified by Herman’s orientation factor analysis of the wide-angle X-ray diffraction patterns. Through this work, the processing parameters and morphological and mechanical properties of a promising nanocomposite material were developed for potential use in advanced fiber and material applications. Future work will focus on the thermal and electrical conductivity of these nanocomposite fibers. The current results reinforce the need to design the nanotube/polymer interface so that sufficient load transfer occurs. Improved adhesion and stress transfer across the interface would enable the exceedingly high modulus value of the nanotubes to contribute in the expected way to the nanocomposite modulus. At the same time, it is important to tailor the surface adhesion properly to capitalize on the role of nanotubes in the toughening of the nanocomposites. If the interfaces of the bridging nanotubes are exceedingly well bonded, fracture would occur at very high stresses, but without significant

© 2007, Woodhead Publishing Limited

252

Nanofibers and nanotechnology in textiles

development of strain, thereby producing strong but brittle materials. A properly designed polymer/nanotube interface would adhere at low stresses to provide modulus enhancement while providing frictional slip modes at higher stresses for energy dissipation. Therefore future work should focus on the ability to engineer the interface to promote both load transfer and provide frictional energy dissipation. Additionally, researchers are also making nanotube ropes (without a polymer matrix). With this methodology, one can imagine that the strength of the ropes should approach that of the individual nanotubes. Issues here will include ductility, abrasion resistance and processability.

8.7

Sources of further information and advice

In addition to the references cited throughout this chapter, there are some good review sources for polymer nanotube composites. These include: • Thostenson, E.T., Ren, Z. and Chou, T.W., Advances in the science and technology of carbon nanotubes and their composites: a review. Composites Science and Technology, 2001. 61(13): p. 1899–1912. • Harris, P.J.F., Carbon nanotube composites. International Materials Review, 2004. 49(1): p. 31–43. • Lau, K.T. and Hui, D., The revolutionary creation of new advanced materials – carbon nanotube composites. Composites Part B – Engineering, 2002. 33(4): p. 263–277. Another source which discusses polymer nanocomposites (not specific to carbon nanotubes) is: • Krishnamoorti, R. and Vaia, R.A., Polymer Nanocomposites: Synthesis, Characterization, and Modeling. ACS Symposium Series, 2002. 804. For detailed information on carbon nanotubes, the following will provide the reader with a solid foundation: • Dresselhaus, M.S. and Dai, H., Advances in carbon nanotubes. MRS Bulletin, 2004. 29(4). • Dresselhaus, M.S., Dresselhaus, G., Charlier, J.C. and Hernandez, E., Electronic, thermal and mechanical properties of carbon nanotubes. Philosophical Transactions of the Royal Society of London Series A – Mathematical Physical and Engineering Sciences, 2004. 363(1823): p. 2065–2098.

8.8

Acknowledgments

The author would like to thank William Dondero for running the experiments, collecting the data, and performing the analyses. The author gratefully acknowledges Professor Richard Spontak for useful discussions regarding

© 2007, Woodhead Publishing Limited

Carbon nanotube-polymer fibers using melt spinning

253

the analysis of the X-ray diffraction data. In addition, the author thanks Professor Jon Paul Maria for use of the X-ray facility and training. This work was supported by the North Carolina State University Faculty Research and Professional Development Program and the NCSU Nanotechnology Initiative.

8.9

References

1. Iijima, S., Helical microtubules of graphitic carbon. Nature, 1991. 354(6348): p. 56– 58. 2. Iijima, S. and T. Ichihashi, Single-shell carbon nanotubes of 1-nm diameter. Nature, 1993. 363(6430): p. 603–605. 3. Ebbesen, T.W., et al., Electrical conductivity of individual carbon nanotubes. Nature, 1996. 382(6586): p. 54–56. 4. Hone, J., et al., Thermal conductivity of single-walled carbon nanotubes. Physical Review B, 1999. 59(4): p. R2514–R2516. 5. Treacy, M.M.J., T.W. Ebbesen, and J.M. Gibson, Exceptionally high Young’s modulus observed for individual carbon nanotubes. Nature, 1996. 381(6584): p. 678–680. 6. Bower, C., et al., Deformation of carbon nanotubes in nanotube–polymer composites. Applied Physics Letters, 1999. 74(22): p. 3317–3319. 7. Cooper, C.A., et al., Distribution and alignment of carbon nanotubes and nanofibrils in a polymer matrix. Composites Science and Technology, 2002. 62(7–8): p. 1105– 1112. 8. Haggenmueller, R., et al., Aligned single-wall carbon nanotubes in composites by melt processing methods. Chemical Physics Letters, 2000. 330(3,4): p. 219–225. 9. Jin, L., Bower, C. and Zhou, O. Alignment of carbon nanotubes in a polymer matrix by mechanical stretching. Applied Physics Letters, 1998. 73(9): p. 1197–1199. 10. Jin, Z., et al., Dynamic mechanical behavior of melt-processed multi-walled carbon nanotube/poly(methyl methacrylate) composites. Chemical Physics Letters, 2001. 337(1,2,3): p. 43–47. 11. Kearns, J.C. and R.L. Shambaugh, Polypropylene fibers reinforced with carbon nanotubes. Journal of Applied Polymer Science, 2002. 86(8): p. 2079–2084. 12. Lozano, K. and E.V. Barrera, Nanofiber-reinforced thermoplastic composites. I. Thermoanalytical and mechanical analyses. Journal of Applied Polymer Science, 2001. 79(1): p. 125–133. 13. Potschke, P., Fornes, T.D. and Paul, D.R. Rheological behavior of multiwalled carbon nanotube/polycarbonate composites. Polymer, 2002. 43(11): p. 3247–3255. 14. Safadi, B., Andrews, R. and Grulke, E.A. Multiwalled carbon nanotube polymer composites: synthesis and characterization of thin films. Journal of Applied Polymer Science, 2002. 84(14): p. 2660–2669. 15. Schadler, L.S., Giannaris, S.C. and Ajayan, P.M. Load transfer in carbon nanotube epoxy composites. Applied Physics Letters, 1998. 73(26): p. 3842–3844. 16. Chen, J., et al., Solution properties of single-walled carbon nanotubes. Science 1998. 282(5386): p. 95–98. 17. Mitchell, C.A., et al., Dispersion of functionalized carbon nanotubes in polystyrene. Macromolecules, 2002. 35(23): p. 8825–8830. 18. Bubert, H., et al., Characterization of the uppermost layer of plasma-treated carbon nanotubes. Diamond and Related Materials, 2003. 12(3–7): p. 811–815. 19. Eitan, A., et al., Surface modification of multiwalled carbon nanotubes: toward the

© 2007, Woodhead Publishing Limited

254

20.

21. 22.

23. 24. 25. 26. 27.

28.

29. 30.

31. 32. 33.

34. 35. 36.

37.

38. 39.

Nanofibers and nanotechnology in textiles tailoring of the interface in polymer composites. Chemistry of Materials, 2003. 15(16): p. 3198–3201. Jang, J., Bae, J. and Yoon, S.H. A study on the effect of surface treatment of carbon nanotubes for liquid crystalline epoxide–carbon nanotube composites. Journal of Materials Chemistry, 2003. 13(4): p. 676–681. Shaffer, M.S.P., Fan, X. and Windle, A.H. Dispersion and packing of carbon nanotubes. Carbon, 1998. 36(11): p. 1603–1612. Star, A., et al., Preparation and properties of polymer-wrapped single-walled carbon nanotubes. Angewandte Chemie, International Edition, 2001. 40(9): p. 1721– 1725. Jia, Z., et al., Study on poly(methyl methacrylate)/carbon nanotube composites. Materials Science and Engineering A, 1999. A271: p. 395–400. Deng, J., et al., Carbon nanotube–polyaniline hybrid materials. European Polymer Journal, 2002. 38(12): p. 2497–2501. Qian, D., et al., Load transfer and deformation mechanisms in carbon nanotube– polystyrene composites. Applied Physics Letters, 2000. 76(20): p. 2868–2870. Bhattacharyya, A.R., et al., Crystallization and orientation studies in polypropylene/ single wall carbon nanotube composite. Polymer, 2003. 44(8): p. 2373–2377. Gorga, R.E. and R.E. Cohen, Toughness enhancements in poly(methyl methacrylate) by addition of oriented multiwall carbon nanotubes. Journal of Polymer Science Part B – Polymer Physics, 2004. 42(14): p. 2690–2702. Potschke, P., Bhattacharyya, A.R. and Janke, A. Melt mixing of polycarbonate with multiwalled carbon nanotubes: microscopic studies on the state of dispersion. European Polymer Journal, 2004. 40(1): p. 137–148. Siochi, E.J., et al., Melt processing of SWCNT – polyimide nanocomposite fibers. Composites Part B – Engineering, 2004. 35(5): p. 439–446. Tang, W.Z., Santare, M.H. and Advani, S.G. Melt processing and mechanical property characterization of multi-walled carbon nanotube/high density polyethylene (MWNT/ HDPE) composite films. Carbon, 2003. 41(14): p. 2779–2785. Gong, X., et al., Surfactant-assisted processing of carbon nanotube/polymer composites. Chemistry of Materials, 2000. 12(4): p. 1049–1052. Dror, Y., et al., Carbon nanotubes embedded in oriented polymer nanofibers by electrospinning. Langmuir, 2003. 19(17): p. 7012–7020. Bin, Y.Z., et al., Development of highly oriented polyethylene filled with aligned carbon nanotubes by gelation/crystallization from solutions. Macromolecules, 2003. 36(16): p. 6213–6219. Chen, G.Z., et al., Carbon nanotube and polypyrrole composites: coating and doping. Advanced Materials (Weinheim, Germany), 2000. 12(7): p. 522–526. Grady, B.P., et al., Nucleation of polypropylene crystallization by single-walled carbon nanotubes. Journal of Physical Chemistry B, 2002. 106(23): p. 5852–5858. Valentini, L., et al., Effects of single-walled carbon nanotubes on the crystallization behavior of polypropylene. Journal of Applied Polymer Science, 2003. 87(4): p. 708–713. Valentini, L., et al., Morphological characterization of single-walled carbon nanotubes-PP composites. Composites Science And Technology, 2003. 63(8): p. 1149– 1153. Moore, E.M., et al., Enhancing the strength of polypropylene fibers with carbon nanotubes. Journal of Applied Polymer Science, 2004. 93(6): p. 2926–2933. Assouline, E., et al., Nucleation ability of multiwall carbon nanotubes in polypropylene

© 2007, Woodhead Publishing Limited

Carbon nanotube-polymer fibers using melt spinning

40.

41.

42. 43. 44. 45. 46.

255

composites. Journal of Polymer Science Part B – Polymer Physics, 2003. 41(5): p. 520–527. Sandler, J., et al., Crystallization of carbon nanotube and nanofiber polypropylene composites. Journal of Macromolecular Science – Physics, 2003. B42(3–4): p. 479– 488. Xia, H.S., et al., Preparation of polypropylene/carbon nanotube composite powder with a solid-state mechanochemical pulverization process. Journal of Applied Polymer Science, 2004. 93(1): p. 378–386. Ren, Z.F., et al., Synthesis of large arrays of well-aligned carbon nanotubes on glass. Science, 1998. 282(5391): p. 1105–1107. Karger-Kocsis, J., ed. Polypropylene: An A-Z Reference. 1999, Kluwer Academic Publishers: Boston. Broda, J., Polymorphism in polypropylene fibers. Journal of Applied Polymer Science, 2003. 89(12): p. 3364–3370. Wilbrink, M.W.L., et al., Toughenability of nylon-6 with CaCO3 filler particles: new findings and general principles. Polymer, 2001. 42(26): p. 10155–10180. McCrum, N.G., Buckley, C.P. and Bucknall, C.B. Principles of Polymer Engineering. Second ed. 1997, Oxford University Press: New York.

© 2007, Woodhead Publishing Limited

9 Multifunctional polymer nanocomposites for industrial applications S. J. B U L L, University of Newcastle, UK

9.1

Introduction

The development of composite materials with nanostructured reinforcements is well documented; such composites may have metallic (Michalski et al., 2003), ceramic (Ohji, 1999) or polymeric (Merinska et al., 2004) matrices and be developed in bulk or thin film form (Pavoor et al., 2004). Reinforcements are often chosen to improve the structural performance of the composite without compromising other properties though in some cases it is possible to enhance both structural and functional properties considerably. This has led to emerging applications in a number of industries, particularly the automotive and sports goods industries. For textile applications polymer matrix composites are the most relevant. In polymer nanocomposites the reinforcements are typically selected to increase the strength of the material and many workers have demonstrated the benefits of adding exfoliated clays, nanoparticles and even carbon nanotubes. The major problems arise in ensuring reliability of processing and achieving a uniform dispersion and distribution of the reinforcement. However, the benefits in structural strength are not always much greater than can be achieved with traditional reinforcements, and to get the greatest benefits from such nanocomposites, additional functionality must be considered. Since a given improvement in mechanical properties can be achieved by a smaller volume fraction of nanoscale reinforcement, there is volume available for the addition of particles which can deliver other properties rather than structural strength and stiffness. Thus there is considerable potential for multifunctional nanocomposite materials, both as bulk composite materials and fibres. In this chapter the development of nanocomposites for mechanical, tribological and fire-protection applications will be introduced and the potential for multifunctional nanocomposite fibres will be discussed.

256

© 2007, Woodhead Publishing Limited

Multifunctional polymer nanocomposites

9.2

257

The development of functional polymer nanocomposites

Making good samples of polymer nanocomposites is a challenge and a range of processing techniques are being actively developed at present including melt mixing (Chan et al., 2002), in situ polymerisation (Yang et al., 1998, Reynaud et al., 2001) and other approaches. A full discussion of the technologies available is beyond this chapter but the interested reader is referred to the review of Jordan et al. (2005). A single processing approach is unlikely to deliver viable composites in every system and the different processing techniques often do not give the same results (Park et al., 2001). One of the main issues in preparing good polymer matrix nanocomposites is the good dispersion of the nanoparticles in the polymer matrix which is a strong function of the preparation technique. This is a particular problem as the volume fraction of the particles increases. An important group of nanocomposites is those based on clay reinforcement where the processing critically depends on the final morphology required for the reinforcement within the composite, i.e. particulate, exfoliated or intercalated (Komarneni, 1992). In the intercalated form the matrix polymer molecules are introduced between ordered layers of clay resulting in an increase in the interlayer spacing. However, in the exfoliated form the clay layers are separated and distributed randomly within the matrix. It is possible that some portion of a particular composite will form the intercalation morphology whereas another part will form the exfoliated structure – this is determined by balancing the interaction between the polymer matrix and silicate platelets against interactions between the silicate platelets themselves and may depend on attractive and steric interactions (Koo et al., 2003). Exfoliation of layered materials is often hampered by the fact that the materials exhibit a strong tendency to agglomerate due to their large surface areas. However, in general, exfoliated materials show better properties than intercalated materials with the same nanoplatelet concentration and this has driven the development of exfoliated composite systems. Intercalated nanocomposites are usually formed by mixing in the melt or in situ polymerisation whereas exfoliation may require more complex processing depending on the properties of the clay (Usuki et al., 1993). However, such layered silicate-based polymer nanocomposites have attracted considerable recent interest after the commercialisation of polypropyleneand nylon-6-based materials (Krishnamoorti and Yurekli, 2001, Kiersnowski and Piglowski, 2004). The major barrier to commercialisation has been developing techniques to ensure a reliable and reproducible product which has now been addressed for clay-based composites some thirty or so years after they were first developed. The use of carbon nanotubes (CNT) in polymer composites has attracted

© 2007, Woodhead Publishing Limited

258

Nanofibers and nanotechnology in textiles

considerable attention (Wagner et al., 1997, Shaffer and Kinloch, 2004). The high aspect ratio of the carbon nanotubes and their exceptional mechanical properties could provide the ultimate reinforcement for structural composite materials and improved electrical and thermal conductivity are also possible due to the low percolation limit of CNT. However, the planar graphitic surface of these nanotubes provides relatively few sites for easy chemical modification and bonding to a matrix material and hence processing viable composites can be difficult. Furthermore, the properties of the nanocomposites made with carbon nanotubes vary considerably; in part this is due to the fact that mixtures of single and multiwall nanotubes are produced by most manufacturing techniques and this will vary from process to process and even from batch to batch. Another source of variation is the different defect densities which are observed for nanotubes from different sources. Considerable improvements in reliability and reproducibility will be required if CNT are to be used in composites in the same way that clay systems are used today. To date much of the development of polymer nanocomposites has been for structural applications with current commercial applications such as the step assist for the Chevrolet Astro van introduced by General Motors in 2002. However, there are other composite functions, such as tribological resistance, low friction or fire retardancy which are important in other applications and with nanoscale reinforcements it is possible to mix several different types of reinforcement to generate improvements in a range of properties. The following sections discuss how such properties are improved in nanocomposites.

9.3

Improving the mechanical properties of polymer nanocomposites

Many workers have reported that the elastic modulus of nanocomposites increases as the size of the reinforcement is reduced, provided that there was a good interaction between the filler and the polymer matrix (Vollenberg and Heikens, 1989, Chan et al., 2002). For some nanocomposites with very small reinforcement particles, Young’s modulus is greater than might be expected from the volume law of mixtures of the constituents and this is attributed to the modification to the structure of the matrix surrounding the nanoparticles due to their high surface area (Akita and Hattori, 1999). For polymer systems where a high degree of crystallinity is possible, the increase in modulus with a reduction in particle size is found to be even greater unless there is poor interaction between filler and matrix. The modulus increases with the volume fraction of the reinforcement as expected until aggregation of the particles occurs when the modulus can be reduced in some systems (Akita and Hattori, 1999).

© 2007, Woodhead Publishing Limited

Multifunctional polymer nanocomposites

259

The yield strength of a polymer nanocomposite is critically dependent on the interaction between the reinforcement and the matrix. Where this interaction is strong the yield stress tends to increase with increasing volume fraction of reinforcement and decreasing particle size. This changes when there is poor interaction where an increase in volume fraction leads to a yield stress decrease from the value for the unreinforced matrix, regardless of the filler concentration or size. The ultimate tensile strength follows a similar pattern to the yield stress. Thus, to achieve nanocomposites with good mechanical properties it is critical that the matrix–reinforcement interfacial chemistry is controlled to give strong adhesion. As the stiffness of the nanocomposite increases, so, in general, does its strength and this is accompanied by a reduction in its strain to failure. The strain to failure for a nanocomposite material is often higher than when the reinforcement is micrometre-sized (Petrovic et al., 2000). However, there are exceptions to this and, as with strength, it critically depends on the bonding between the reinforcement and the matrix – if this is poor then both the strength and strain to failure are reduced (Chan et al., 2002). For clay-reinforced nanocomposites, increases in modulus compared with the unfilled polymer matrix have been observed in many systems with the effect increasing with filler content as expected but the properties are highly sensitive to microstructure (Luo and Daniel, 2003). In general, to maximise stiffness (and thermal properties) it is necessary to achieve full exfoliation and dispersion which is not readily achieved (Vu et al., 2001, Zhang et al., 2004). Filler particles are added to composite materials to improve their viscoelastic properties. The heat distortion temperature for nanocomposites tends to be higher than that of the unfilled matrix or microreinforced materials because of an increase in viscosity during composite manufacture and a reduction in viscous deformation afterwards. For nanocomposites the storage modulus increases with an increase in the reinforcement content for both clay reinforced (Shelley et al., 2001) and silica nanoparticle reinforced nylon (Reynaud et al., 2001). Similar results have been observed for silica–polyurethane composites (Petrovic et al., 2000). Morphological details such as exfoliation, intercalation or matrix cross-linking also have a significant effect on the viscoelastic response. In epoxy–clay nanocomposites an increase in the amount of exfoliation is associated with an increase in viscosity and a relatively large increase in storage modulus (Park and Jana, 2003). In general, the resistance to viscoelastic deformation tends to be higher in polymer nanocomposites than in pure polymer systems. Where there are strong matrix– reinforcement interactions the storage modulus increases with increasing volume fraction of filler and increases as the reinforcement size decreases. The increase in tensile strength or Young’s modulus for nanocomposites compared with microcomposites can partly be explained by the details of the interaction between the filler and the matrix. Good adhesion between matrix

© 2007, Woodhead Publishing Limited

260

Nanofibers and nanotechnology in textiles

and particle results in less debonding when a stress is applied and consequently the strength is improved. The polymer in the proximity of a particle to which there is good adhesion tends to have higher density which also results in an increase in stiffness (Vollenberg and Heikens, 1989). The polymer adjacent to this high-density region will have a lower density as the chains are moved towards the particle to make the high-density region. For large particles this lower density region will be relatively large and the contribution of the high modulus filler will be diminished compared with the case when the filler particle is small. For nanocomposites, the number of particles for a given volume fraction is much larger and the particles will be much closer to each other. If the reinforcement particles are very densely packed then the highdensity boundary layer will make up a larger proportion of the matrix and the modulus will increase. Thus, the major improvements to the mechanical properties of nanocomposites arise from the small particle spacing, rather than directly from the size or volume fraction of the reinforcement particles used. The use of CNT in polymer composites has also received considerable attention (Kuriger et al., 2002, Shaffer and Kinloch, 2004). The combination of the high aspect ratio and extreme mechanical properties (strength and stiffness) of CNT provides the ultimate reinforcement for composite materials. Even with only 1% by weight of CNT added to a polymer matrix increases in strength and stiffness in excess of 40% have been measured (Lozano and Barrera, 2001). However, the properties of most CNT are very variable, since single and multiwall nanotubes with differing numbers of defects may be used for composite manufacture. For this reason the properties of nanotube composites can be very variable and considerable work is necessary to develop a reliable supply of material to generate CNT reinforced composites with controlled properties.

9.4

Improving the fire-retardant properties of polymer nanocomposites

Polymer–clay nanocomposites have been shown to greatly improve the barrier properties and fire retardancy of polymers as might be expected from their highly distributed inorganic material content (Gilman et al., 1997). This has been reviewed by several authors for both polymer–clay and polymer– nanoparticle composites (Porter et al., 2000, Koo and Pilato, 2005, Yang et al., 2005). Intercalated and exfoliated complexes show increased thermal stability compared with the unreinforced polymer. The fire properties of materials are evaluated in many different ways including by cone calorimetry (ASTM E1354), radiative gasification and limiting oxygen index measurement (ASTM D2863, ISO4589). Cone calorimetry is the most widely used laboratory method to assess nanocomposites – this technique generates information on

© 2007, Woodhead Publishing Limited

Multifunctional polymer nanocomposites

261

the heat which is released during combustion of the material and the reduction in peak heat release rate (PHRR) is often used to characterise potential fireretardant materials (Schartel et al., 2005). For instance, in polystyrene–clay nanocomposites, Zhu and Wilkie (2000) reported that the peak heat release rate may be decreased by up to 58% depending on the composite structure and filler content. Improvements in fire performance have been reported for polymer–clay nanocomposites with a range of polymer matricies including polyethylene (Bergaya et al., 2005), polypropylene (Tidjani, 2005), polyamide (Lei et al., 2004), epoxy (Camino et al., 2005), polystyrene (Chigwada et al., 2006a), acrylonitrile butadiene styrene (ABS, Chigwada et al., 2006a) and ethylene– vinyl acetate copolymer (EVA, Szep et al., 2006). In most cases the fire performance of a polymer–clay nanocomposite is dictated by its composition, rather than by the details of the microstructure, in contrast to the mechanical properties of the composite. The benefits increase as the clay content increases and further improvements can be achieved by the use of compatibilisers (Modesti et al., 2006). Good dispersion of the clay is essential for the most reliable materials (Duquesne et al., 2003). Modification of the clay with different salts can improve performance, e.g. by the addition of quinolinium and pyridinium surfactants to polystyrene (Chigwada et al., 2006b). It has also been shown that the use of synthetic clays to improve fire performance is more effective than the use of natural clays in the composite (Morgan et al., 2005). In many cases the clay catalyses the formation of a clay reinforced carbonaceous char which is responsible for the lower flammability of the nanocomposites (Hull et al., 2003, Preston et al., 2004, Zanetti et al., 2004). The fire retardancy of polymer/clay nanocomposites is thus controlled by a change in the degradation pathway of a polymer by incorporation of the clay (Jang et al., 2005). Since the clay layers act as a barrier to mass transport and lead to superheated conditions in the matrix, extensive random scission of the products formed by radical recombination is an additional degradation pathway of polymers in the presence of clay. The polymers that show good fire retardancy upon nanocomposite formation exhibit significant intermolecular reactions, such as inter-chain aminolysis/acidolysis, radical recombination and hydrogen abstraction. This increases as the clay loading increases (Jang and Wilkie, 2005). In the case of the polymers that degrade through a radical pathway, the relative stability of the radical is the most important factor for the prediction of the effect that nanocomposite formation has on the reduction in the peak heat release rate. The more stable is the radical produced by the polymer, the better is the fire retardancy, as measured by the reduction in the peak heat release rate, of the polymer/clay nanocomposite. Improvements in fire performance have also been reported for nanocomposites reinforced with nanoparticles. For instance, PMMA–silica

© 2007, Woodhead Publishing Limited

262

Nanofibers and nanotechnology in textiles

nanocomposites show improved thermal stability compared with the unreinforced polymer (Yang and Nelson, 2004) and polyimide–silica nanocomposites show improved fire retardancy (Liu et al., 2000). Oxide nanoparticles generate a decrease in PHRR with a range of fillers and the effect increases with filler content. Synergistic effects have also been reported with the combination of TiO2 and organoclays (Laachachi et al., 2005). The best fire performance is usually achieved by a combination of traditional fire-retardant additives with the nano-reinforcement, whether clay-based or nanoparticles (Zhang and Horrocks, 2003). For instance the combination of magnesium hydroxide and montmorillonites with an EVA matrix results in improved performance (Szep et al., 2006). Similarly an improvement in fireretardancy of clay–polystyrene nanocomposites has been achieved by the addition of an oligomeric material consisting of vinyl benzyl chloride, styrene and dibromostyrene (Chigwada et al., 2005). Different clay materials also have differing effects on fire retardancy (Jash and Wilkie, 2005) and combinations of different clays can also generate improved properties (Marosfosi et al., 2006). The combination of clay–nanocomposites with added phosphorus-based fire retardants has also been tested and a 92% decrease in PHRR was recorded in the best case (Zhu and Wilkie, 2000). Synergistic effects have been reported in a number of systems, and this, combined with the ability to reduce the fire-retardant content and replace the halogen and phosphorus-containing systems with more environmentally friendly materials, is a major advantage for the nanocomposites currently being developed.

9.5

Improving the tribological properties of polymer nanocomposites

There are a number of ways in which the addition of a particle to a polymer can improve its friction and wear properties and all of these have been exploited in the production of nanocomposites. One approach is to reduce the adhesion of the composite to its sliding counterface by adding frictionmodifying reinforcements which may also reduce the heat generated in a sliding contact. These friction modifiers may be in the form of low-friction particles which form low-friction transfer layers at the interface between the composite and its sliding counterface when the surface layers are disrupted by sliding. The fillers may also modify the properties of the transfer layer which forms, reacting chemically with the matrix and counterface materials to produce a low shear material. Another approach is to increase the hardness, stiffness and strength of the material using reinforcement particles as discussed previously. Yet another approach is to use fillers to control the heat transfer through the material in order to dissipate the heat generated by friction. Not all fillers increase the wear resistance of the composite; in the case where the filler decomposes during sliding and the reaction products strongly adhere to

© 2007, Woodhead Publishing Limited

Multifunctional polymer nanocomposites

263

the counterface (Briscoe, 1993) or where the adhesion between the filler and matrix is poor the wear rate will increase (Dasari et al., 2005). To modify friction and adhesion, solid lubricants such as graphite or polytetrafluoroethylene (PTFE) are used. These are either layered structures (e.g. graphite or MoS2) with low shear strength planes within their structure or smooth molecular species (PTFE) which can become aligned in the sliding direction in a similar manner to combed hair. In both cases these form a modified transfer film on the surface during wear, e.g. a PTFE film (Hager and Davies, 1993). One of the problems with the layered lubricants is the fact that chemical bonding at the edge of the low shear planes can lead to increases in the shear stress required to maintain sliding and a consequent increase in friction. A recent development which addresses this problem is the production of fullerene-like MoS2 nanoparticles where the low-friction basal planes are wrapped into an approximately spherical shell with a much reduced number of chemical bonding sites, usually at kinks in the planes (Rapoport et al., 2005a). This type of material shows low friction and improved life and load bearing capacity compared with conventional flake particles both in composites and as an addition to grease (Rapoport et al., 2005b). Another approach is to soak lubricants into a nanoporous composite (Ahn et al., 2003). The reinforcement material composition affects the composition and properties of the transfer layer that forms during sliding which controls friction and wear behaviour in most nanocomposite systems. A uniform, tenacious transfer film is produced in clay–nylon-6 nanocomposites, which leads to a reduction in friction and a lower wear rate (Srinath and Gnanamoorthy, 2005). Similar results are observed for silica nanoparticle reinforcement (Garcia et al., 2004). The use of reinforcement particles that improve the mechanical properties of the composite in order to improve its wear performance is well documented for a range of polymer matrix materials. As the particle size is reduced the improvement in tribological properties tends to increase as smaller regions of matrix material are exposed to the sliding counterface during wear. For instance the wear performance of silica reinforced epoxy composites increases as the particle size is reduced from 500 to 120 nm (Xing and Li, 2004). However, opposite trends were observed for glass reinforced nylon in abrasive wear tests (Friedrich, 1986b). In fact the behaviour is often a complex function of particle size with an increase in wear rate at intermediate size and a reduction at the smallest sizes (Fig. 9.1). For composites with nanoscale particle reinforcement (i.e. less than 100 nm diameter) there is generally an improvement in wear performance as the particle size is reduced (Wang et al., 1996, Bahadur and Sunkara, 2005). Optimum wear performance is usually obtained with a fixed volume fraction of filler (Cai et al., 2003). For a particular particle size the wear resistance increases with volume fraction up

© 2007, Woodhead Publishing Limited

Nanofibers and nanotechnology in textiles

Wear rate [m3/N m]

264

10–14

10–15

10–16 10

100

1000 Particle size [nm]

104

9.1 Effect of reinforcement diameter on wear performance for silica– polyurethane nanocomposites (12 vol%).

to a maximum, usually when the filler particles start to interact with each other. The maximum in performance thus represents a limit imposed by the dispersion of the particles in the matrix. In cases where particle interactions occur, the wear can be very severe and apparently abrasive in nature whereas at lower reinforcement volume fractions the wear mechanism is different (e.g. adhesive and fatigue wear). A strong adhesion between reinforcement and matrix is necessary to prevent pull-out and high wear in nanocomposites designed for tribological applications. This has driven the development of coupling agents and other chemical modifications to improve performance (Zhang et al., 2002a,b). An added advantage of nanocomposites manufactured with small diameter reinforcements is the reduction in surface roughness of the sliding surface. This generally leads to a reduction in the contact stresses at asperity contacts and a resultant reduction in damage to the composite system. The roughness of the counterface is also critical because this dictates the nature of the initial contact with the nanocomposite surface and controls the mechanics of formation of the transfer film (Friedrich, 1986b) – when the roughness of the counterface is less than the particle size, the presence of the nanoparticles tends to increase the wear rate because detached particles get trapped between the sliding surfaces and cannot be taken out of the contact by moving into the roughness valleys. In this case three body abrasion occurs before a protective transfer film forms and the wear rate increases. Improvements in friction and wear have also been reported with carbon nanotube reinforced polymer composites (Igarashi et al., 2005). This is due to both increases in strength of the material and the modification of the transfer layer by fragments of nanotube which can reduce friction. There is an optimum nanotube composition (Cai et al., 2004) typically about 10% by weight, for the best wear performance (Werner et al., 2004).

© 2007, Woodhead Publishing Limited

Multifunctional polymer nanocomposites

9.6

265

Case-study: development of a nanocomposite sliding seal ring

The problems with developing successful functional applications of nanocomposites can be illustrated with the following example. In this casestudy nanocomposite materials were developed for an application in which a composite ring was sliding against a flat steel component in the absence of lubrication. The mean contact pressure at the seal face was 1 MPa. The wear rate and coefficient of friction of the sliding composite part were to be minimised and the sliding speed maximised. A traditional approach to achieve a highly wear-resistant polymer composite is to combine fillers with different functions (e.g. Friedrich, 1986a); this approach has been adopted here. A range of conventional composite and nanocomposite materials has been investigated to determine which is most suitable. Benchmarking newly developed nanocomposites against currently available composite systems is essential to justify their use in any application and in most cases there is a considerable cost implication when selecting a nanocomposite material.

9.6.1

Materials investigated

A range of structural composite materials were produced with different reinforcements for performance assessment. The systems were based on a conventional woven roving/thermosetting resin design but with added nanoscale reinforcement which was mixed with the resin prior to lamination. The systems investigated were: 1. glass reinforced phenolic; 2. glass reinforced polyester; 3. glass reinforced vinyl ester (conventional and scrimp (higher glass content)); 4. carbon fibre reinforced epoxy; 5. glass reinforced vinyl ester with added PTFE (100 nm); 6. glass reinforced vinyl ester with added MoS2 (100 nm); 7. glass reinforced vinyl ester with added MoS2 (100 nm) + 10% flake graphite (>10 µm); 8. glass reinforced vinyl ester with added flake graphite (>10 µm); 9. glass reinforced vinyl ester with added 3.5–5 µm silica particles; 10. glass reinforced vinyl ester with added 50–70 nm diameter silica particles; 11. glass reinforced vinyl ester with added silica particles; 50% at 50–70 nm and 50% at 3.5–5 µm. Sample 1 was chosen because the thermal degradation products of phenolic materials are known to be less sticky that the other matrix materials (Vishwanath et al., 1993). Sample 2 was chosen because it is cheap and easy to process

© 2007, Woodhead Publishing Limited

266

Nanofibers and nanotechnology in textiles

but it was expected to have poor wear performance. Silica particles were treated with a titanate coupling agent to aid wetting by the resins; particles were easy to mix with the resin but tended to form clumps in the composite after processing. Sample 4 was chosen because carbon fibre additions to materials such as nylon have been shown to reduce friction in seals by modifying the chemistry of the transfer layer formed. All the additives in samples 6 to 8 are low friction materials that are known to reduce friction in some circumstances as an additive to polymers, composites (Yu et al., 1996), lubricating oils and greases. The main difference with previous studies was the size of the reinforcing particles. The samples were made by laminating against glass using a release agent; this produced a very smooth surface which is suitable for wear testing. The additives were mixed with the vinyl ester resin prior to hand lamination according to the proportions in Table 9.1. Four samples were made at a time. Some 400 g of the vinyl ester resin was used with the various fillers to lay up the first four layers of the composite and then another four layers were laid up using just the resin. This produces samples with a smooth side (next to the glass) which can be tested and a rough side at the rear. The total sample thickness is sufficiently large that no substrate or backing effects were observed during testing. For the phenolic and polyester matrix materials both the smooth and rough sides were available for testing and both were assessed to determine if roughness has a big effect on the friction and wear results. The results were compared with a standard steel sample which was tested either in the as-received condition or coated with a layer of lubricating grease. In addition to the modification of structural composites a number of polymer nanocomposites were made by mixing nanoscale reinforcements with a polyurethane resin. In this case the resin was again cast against a glass Table 9.1 Fillers added to 400 g vinyl ester resin to make the structural composites used in this study Sample

Glass mat

Filler (1)

Weight (1) [g]

Micro silica (3.5–5 µm)

Woven roving 0/90

Silica

300

Nano silica (50–70 nm)

Woven roving 0/90

Silica

200

Mixed silica

Woven roving 0/90

Microsilica

100

PTFE (100 nm)

Woven roving 0/90

PTFE

50

Graphite

Woven roving 0/90

Graphite

30

MoS2 (100 nm)

Woven roving 0/90

MoS2

100

Mixed

Woven roving 0/90

MoS2

90

© 2007, Woodhead Publishing Limited

Filler (2)

Weight (2) [g]

Nanosilica

100

Graphite

10

Multifunctional polymer nanocomposites

267

block and the smooth side was tested. Silica nanoparticles of a range of different sizes were added to the resin at 5% and multiwall carbon nanotubes were added at 0.5% by weight.

9.6.2

Wear tests

Tribological tests were performed on a modified pin on disc tester fitted with a rotation speed tachometer and friction and wear transducers which were all calibrated prior to use. In this tester a flat-ended pin is loaded against a disc by a dead weight on the end of the loading arm. In each test 60 mm diameter composites discs of approximately 10 mm thickness were used sliding against a 10 mm diameter, 17 mm long steel pin. The disc is rotated by a motor whose speed is controlled by the tachometer. The rotation of the disc forces the loading arm against a load transducer to measure the frictional force and a displacement transducer is placed against the loading arm to measure the total arm displacement during the test. Great care was taken to ensure that the end of the pin was flat and parallel to the disc surface in the test to minimise errors due to misalignment. This was achieved by running in the pin on an unimportant region of disc prior to moving to the test diameter for friction and wear evaluation. All tests were carried out unlubricated. Each test was performed three times and the results averaged; test results were quite consistent with errors of less than 10% except in the case of the nanoscale silica particles where considerable scatter was observed and the extremes of the behaviour observed are shown in the following figures. Two different types of test were carried out: • Variation of friction coefficient with sliding speed. In this case tests were carried out on the same track. The sample was rotated at a low speed and left for a few minutes for the friction coefficient to stabilise. A friction measurement was then taken averaged over 60 s of sliding and then the speed was increased and the process repeated. The speed was increased in this manner until seizure occurred and the rotation stopped. This defines the limit in sliding velocity for the composite/pin pair at the normal load chosen. • Fixed speed test at 6 m/s for a total of 10 km sliding distance to determine wear behaviour. In all tests a load of 75 N was applied to the pin giving a contact pressure of ~1 MPa. Stylus profilometry was used to assess the depth of the wear scar produced at four places on each track and the results averaged to determine the wear. Friction coefficients were determined by dividing the measured frictional force by the applied normal load.

© 2007, Woodhead Publishing Limited

268

Nanofibers and nanotechnology in textiles

9.6.3

Friction coefficient measurements

For all materials the friction coefficient was approximately constant as a function of sliding speed until close to seizure when friction increases rapidly. Typical plots are shown in Fig. 9.2 and 9.3. Two important parameters can thus be extracted from the test data: • the average friction coefficient at all sliding speeds; • the speed at which seizure occurs under the contact conditions in the test. The average friction coefficient data for all the samples are collected in Fig. 9.4. It can clearly be seen that most of the composite samples have lower friction coefficients than ungreased steel but all have higher friction than greased steel. As expected the phenolic reinforced with glass shows low friction in dry sliding as does the polyester reinforced with glass and epoxy reinforced with carbon. Only the nanoscale PTFE and MoS2 additives have been successful at reducing the friction of the glass reinforced vinyl ester (VE). This is consistent with the observation that the scrimp glass reinforced vinyl ester, which has a higher glass content than the other material tested, also shows a higher friction coefficient. It appears that increasing the amount of filler in the material is thus counterproductive if a low friction coefficient

Steel pin, 1MPa contact pressure, dry 0.6

Coefficient of friction

0.5

0.4 Steel Greased steel Phenolic/glass smooth Phenolic/glass rough Polyester/glass smooth Polyester/glass rough

0.3

0.2

0.1

0

0

1

2

3 4 Speed [m/s]

5

6

7

9.2 Variation of coefficient of friction with speed for a range of materials tested against steel at 1 MPa contact pressure in dry sliding.

© 2007, Woodhead Publishing Limited

Multifunctional polymer nanocomposites

269

Steel pin, 1MPa contact pressure, dry 0.6 Scrimp vinyl ester/glass Vinyl ester/glass Epoxy/glass Epoxy/Kevlar Epoxy/glass-Kevlar hybrid Epoxy/carbon

Coefficient of friction

0.5

0.4

0.3

0.2

0.1 0

1

2

3

4

5

Speed [m/s]

9.3 Variation of friction coefficient against steel with sliding speed for a range of vinyl ester and epoxy matrix composites tested at 1 MPa contact pressure in dry sliding.

Dry friction VE/nano silica (2) VE/mixed silica VE/nano silica (1) VE/micro silica VE/graphite VE/MoS2 + graphite VE/MoS2 VE/PTFE VE/glass Scrimp VE/glass Epoxy/Kevlar Epoxy/glass + Kevlar Epoxy/glass Epoxy/carbon Polyester/glass rough Polyester/glass smooth Phenolic/glass rough Phenolic/glass smooth Greased steel Steel 0

0.1

0.2 0.3 Average friction coefficient

0.4

0.5

9.4 Average friction coefficient against steel (1 MPa contact pressure) as a function of material tested in this programme.

© 2007, Woodhead Publishing Limited

270

Nanofibers and nanotechnology in textiles

is required unless the filler itself can deliver low friction. It was initially thought that this increase in wear was due to interactions between the reinforcement materials at high loadings. However, temperature measurements showed that increasing the filler content has reduced the thermal conductivity of the composite and thus increased the thermal degradation and stickiness of the contact zone. The highly filled composites show a higher wear rate (see later) and more surface damage is evident. Damage dissipates energy and therefore increases friction. Increasing surface roughness slightly also increases friction for the glass reinforced polyester tested here due to the excessive damage created. This is much less significant for the phenolic matrix material which shows no change in friction with roughness. The glass reinforced vinyl ester reinforced with nanoscale reinforcement was very different to process and clumps of glass were visible in the composite. Repeat tests on this material show very variable friction and wear performance depending on the location of the glass clumps. This highlights the importance of developing processing methods that generate a uniform dispersion of any reinforcement. The maximum speed before seizure for all the composite materials is very low (Fig. 9.5) but can be increased by the addition of nanoscale MoS2 reinforcements.

Speed limit (m/s) VE/nano silica (2) VE/mixed silica VE nanosilica (1) VE/micro silica VE/graphite VE/MoS2 + graphite VE/MoS2 VE/PTFE VE/glass Scrimp VE/glass Epoxy/Kevlar Epoxy/glass + Kevlar Epoxy/glass Epoxy/carbon Polyester/glass rough Polyester/glass smooth Phenolic/glass rough Phenolic/glass smooth Greased steel Steel 0

1

2

3 Speed limit [m/s]

4

5

6

9.5 Maximum speed before seizure for samples tested in this study (1 MPa contact pressure).

© 2007, Woodhead Publishing Limited

Multifunctional polymer nanocomposites

9.6.4

271

Wear

During the wear test experiments it was found that not all samples could be run under the same conditions as originally planned since the maximum speed limit of most of the composites was very low. Wear measurements were thus made on each disc at the highest sliding speed achievable for a total sliding distance of at least 1 km and normalised to 10 km (assuming wear depth is linearly proportional to sliding distance) to compare with the best samples. Some of the samples show very good wear behaviour (Fig. 9.6). In particular the glass reinforced phenolic shows very low wear rates. The carbon fibre reinforced epoxy has a much higher wear rate than any of the other epoxy matrix composites – the fibres are easily damaged during testing and any advantage that carbon might impart in reducing friction is lost as the roughened damage surface wears. Most of the nano-additions to vinyl ester reduce its wear rate and the silica reinforcement can give major improvements in the best samples. However, as the glass content increases wear is increased due to poor interaction between the particles and the matrix leading to easy pullout; in the case of the scrimp vinyl ester glass pull-out is very apparent in the wear scar. Of all the additives the nanoscale silica has the greatest effect but the nanoscale MoS2 offers some advantages. The glass reinforced polyester has a very high wear rate as expected. The improved wear resistance of nanosilica reinforced vinyl ester prompted a more detailed investigation of the effects of particle size on the wear 10 km Sliding distance, 1 MPa contact pressure, dry VE/nano silica (1) VE/mixed silica VE/nano silica (2) VE/micro silica VE/graphite VE/MoS2 + graphite VE/MoS2 VE/PTFE VE/glass Scrimp Vinyl ester/glass Epoxy/Kevlar Epoxy/glass + Kevlar Epoxy/glass Epoxy/carbon Polyester/glass rough Polyester/glass smooth Phenolic /glass rough Phenolic/glass smooth Greased steel Steel 0

200

400

600 800 Wear depth [µm]

1000

1200

9.6 Maximum wear scar depth of all materials tested in this study after 10 km dry sliding against steel (1 MPa contact pressure).

© 2007, Woodhead Publishing Limited

272

Nanofibers and nanotechnology in textiles

resistance of a polymer nanocomposite. In this case a polyurethane resin was used, rather than a structural composite, to avoid the complicating effects of macroscale reinforcement. Figure 9.1 shows that the wear rate is critically dependent on particle size, initially increasing as the reinforcement diameter is reduced but dramatically falling as the reinforcement size becomes nanoscaled. The results presented so far indicate that no single nanoscale reinforcement generates ideal results, either alone or when used in conjunction with more conventional reinforcement materials. To control wear, a hard reinforcement such as silica is preferred whereas to reduce friction a low shear stress reinforcement such as MoS2 is more effective. Thus a combination of nanoscale reinforcements in vinyl ester resin has been investigated in further tests and the benefits of the mixed compositions on friction (Fig. 9.7) and speed at seizure (Fig. 9.8) have been demonstrated without compromising wear performance. Visual examination of the worn surfaces of the tested composites shows some evidence of char formation. Thus the high-speed sliding wear performance of these polymer nanocomposite materials is controlled by heat generation and dissipation. The approach of adding particles which generates a reduction in friction has been successful at improving tribological performance by limiting heat generation but an alternative approach would be to add material to the composite which can improve its thermal conductivity so that any heat generated can be more easily dissipated. Potential nanoscale reinforcements which could achieve this include metals and carbon nanotubes since the conductivity along the tube is very high. The best results are achieved with single wall nanotubes but it is currently difficult to obtain this material in Friction coefficient Vinyl ester/50nm silica Vinyl ester/200nm + 100nm silica Vinyl ester/200nm silica Vinyl ester/7000nm silica Vinyl ester/graphite Vinyl ester/200nm MoS2 + graphite Vinyl ester/100nmMoS2/100nm silica Vinyl ester/100nm PTFE/100nm silica Vinyl ester 0

0.1

0.2

0.3 0.4 Friction coefficient

0.5

9.7 Friction coefficient of vinyl ester nanocomposites with different reinforcements.

© 2007, Woodhead Publishing Limited

0.6

Multifunctional polymer nanocomposites Vinyl ester/50nm silica

273

Speed limit (m/s)

Vinyl ester/200nm + 100nm silica Vinyl ester/200 nm silica Vinyl ester/7000nm silica Vinyl ester/graphite Vinyl ester/200nm MoS2 + graphite Vinyl ester/100nmMoS2/100nm silica Vinyl ester/100nm PTFE/100nm silica Vinyl ester

0

0.2

0.4

0.6 0.8 1.0 Speed limit [m/s]

1.2

1.4

1.6

9.8 Speed limit of vinyl ester nanocomposites with different reinforcements.

sufficient quantity for tribological composites. Thus unprocessed multiwall nanotube mats have been incorporated in the polyurethane resin used previously to assess the viability of the approach. At 0.5% loading the tubes have little or no effect on the friction and wear rate of the polyurethane at low sliding speeds as might be expected but the speed limit for seizure for the composite is increased from 1 to 3.5 m/s, indicating that the dissipation of the frictional heat has been improved (Fig. 9.9). More work is necessary to assess this encouraging result further. The results presented here highlight the need to understand the mechanisms by which a composite fails in a given application if the best material is to be developed. As the functional requirements become more complex this will be more difficult to achieve but the benefits of doing so will be greater as more synergisms can be exploited.

9.7

Enhancing the functionality of polymer nanocomposites

The results presented here demonstrate the benefit of nanoscale reinforcement to improve wear resistance of the composite material but to improve the overall performance of the materials for a sliding seal application no single reinforcement material is suitable. However, a combination of reinforcements can be developed which impart multiple functionality to the composite. For instance, both silica and MoS2 nanoparticles added to a polyurethane resin matrix can generate improvements in both friction and wear behaviour. The work here has been carried out with thermosetting resin matrices but there is no reason to suspect that similar results will not be achieved using thermoplastic resins which can be processed into fibres, e.g. by melt electrospinning. Enhanced functionality can be achieved at relatively small

© 2007, Woodhead Publishing Limited

274

Nanofibers and nanotechnology in textiles 0.6 Vinyl ester Vinyl ester/MWNT

Friction coefficient

0.5

0.4

0.3

0.2

0.1 0

1

2 3 Speed [m/s]

4

5

9.9 Variation of coefficient of friction with sliding speed for vinyl ester resin with and without 0.5% multiwall nanotube reinforcement.

particle loadings where the viscosity of the material is not increased too much to process. In many cases the nanocomposite composition range does not need to be tightly controlled as good properties are achieved with a range of reinforcement contents and the processing is quite forgiving in this respect. However, achieving a uniform dispersion of the reinforcement is much more important and this is a major goal of process development. It has been shown that nanotubes at similar loadings to those investigated in this study can be added to thermoplastic matrices and processed by injection moulding (Yalcin and Cakmak, 2004) or spun into fibres (Dalton et al., 2003) and fibres containing nano-reinforcements have been produced by electrospinning (Sen et al., 2004). The potential for the manufacture of nanocompsite fibres is therefore excellent but more work is necessary to develop and optimise appropriate processing conditions. A reliable source of defect-free nanotubes with reproducible properties will need to be developed if the benefits are to be exploited commercially to any great extent. It is clear that multifunctional materials may be developed by using several reinforcements in the resin matrix and the choice of these should be based on the required functionality. Many possibilities exist, including: • high-strength impact management systems; • fire-retardant systems; • friction and touch/feel modifiers;

© 2007, Woodhead Publishing Limited

Multifunctional polymer nanocomposites

• • • •

275

colourants; electrical conductivity modifiers/capacitors/batteries; wettability modifiers; self-cleaning fibres.

In some cases these functionalities may be achieved by other methods, e.g. by coating or surface texturing after fibre manufacture and the relative merits of the competing technologies will need to be assessed. One key concern is the cost and availability of the reinforcements used. In this study synthetic nanoparticles and nanotubes have been used but these would not be economic for large-scale composite manufacture at present. Careful control of particle size distribution and the processing of particles is essential and novel processing methods, such as precipitation in the spinning disc reactor (Brechtelsbauer et al., 2001), offer great promise in this respect. A range of natural nanoparticles exist which can be obtained from clay deposits which could lead to improvements in strength and fire protection but these need to be investigated in more detail. The characterisation of mineral deposits and the particles that come from them is a key part of this and much progress has been made in recent years.

9.8

Conclusions

Multifunctional nanocomposite materials can be produced by combining a range of nanoscale reinforcement materials with a polymer matrix. Given a knowledge of the performance requirements of a given application it is possible to design a composite with a range of functionalities necessary for successful operation. However, processing of the composite remains a key factor in success and in particular ensuring a uniform dispersion of reinforcement is critical to reliable performance. There are many potential textile applications of multifunctional composite materials ranging from self-cleaning, water-resistant and fire-retardant fibres, to novel colouration schemes and fabrics with improved toughness and feel. However, much work is necessary to make such applications a reality.

9.9

Acknowledgements

The author would like to thank Katarina Horvathova, Rob McNaught and Thomas Malkow who produced and tested the composite materials.

9.10

References

Ahn, H. S., Julthongpiput, D., Kim, D. I. and Tsukruk, V. V. (2003) Dramatic enhancement of wear stability in oil-enriched polymer gel nanolayers, Wear, 255, 801–807.

© 2007, Woodhead Publishing Limited

276

Nanofibers and nanotechnology in textiles

Akita, H. and Hattori, T. (1999) Studies on molecular composite. I. Processing of molecular composites using a precursor polymer for poly(p-phenylene benzobisthiazole), Journal of Polymer Science Part B – Polymer Physics, 37 (3), 189–197. Bahadur, S. and Sunkara, C. (2005) Effect of transfer film structure, composition and bonding on the tribological behavior of polyphenylene sulfide filled with nano particles of TiO2, ZnO, CuO and SiC, Wear, 258 (9), 1411–1421. Bergaya, F., Mandalia, T. and Amigouet, P. (2005) A brief survey on CLAYPEN and nanocomposites based on unmodified PE and organo-pillared clays, Colloid and Polymer Science, 283 (7), 773–782. Brechtelsbauer, C., Lewis, N., Oxley, P., Ricard, F. and Ramshaw, C. (2001) Evaluation of a spinning disc reactor for continuous processing, Organic Process Research and Development, 5, 65–68. Briscoe, B. J. (1993) ‘The tribology of composite materials: a preface’, in Friedrich, K. (Ed), Advances in Composite Tribology. Elsevier, Amsterdam, pp. 3–15. Cai, H., Yan, F. Y., Xue, Q. J. and Liu, W. M. (2003) Investigation of tribological properties of Al2O3-polyimide nanocomposites, Polymer Testing, 22 (8), 875–882. Cai, H., Yan, F. Y. and Xue, Q. J. (2004) Investigation of tribological properties of polyimide/carbon nanotube nanocomposites, Materials Science and Engineering A – Structural Materials Properties Microstructure and Processing, 364 (1–2), 94– 100. Camino, G., Tartaglione, G., Frache, A., Manferti, C. and Costa, G. (2005) Thermal and combustion behaviour of layered silicate-epoxy nanocomposites, Polymer Degradation and Stability, 90 (2), 354–362. Chan, C. M., Wu, J., Li, J.-X. and Cheung, Y.-K. (2002) Polypropylene/calcium carbonate nanocompsites, Polymer, 43, 2981–2992. Chigwada, G., Jash, P., Jiang, D. D. and Wilkie, C. A. (2005) Synergy between nanocomposite formation and low levels of bromine on fire retardancy in polystyrenes, Polymer Degradation and Stability, 88 (3), 382–393. Chigwada, G., Wang, D. Y., Jiang, D. D. and Wilkie, C. A. (2006a) Styrenic nanocomposites prepared using a novel biphenyl-containing modified clay, Polymer Degradation and Stability, 91 (4), 755–762. Chigwada, G., Wang, D. Y. and Wilkie, C. A. (2006b) Polystyrene nanocomposites based on quinolinium and pyridinium surfactants, Polymer Degradation and Stability, 91 (4), 848–855. Dalton, A. B., Collins, S., Munoz, E., Razal, J. M., Ebron, V. H., Ferraris, J. P., Coleman, J. N., Kim, B. G. and Baughman, R. H. (2003) Super-tough carbon nanotube fibres – these extraordinary composite fibres can be woven into electronic textiles, Nature, 423, 703. Dasari, A., Yu, Z. Z., Mai, Y. W., Hu, G. H. and Varlet, J. L. (2005) Clay exfoliation and organic modification on wear of nylon 6 nanocomposites processed by different routes, Composites Science and Technology, 65 (15–16), 2314–2328. Duquesne, S., Jama, C., Le Bras, M., Delobel, R., Recourt, P. and Gloaguen, J. M. (2003) Elaboration of EVA–nanoclay systems – characterization, thermal behaviour and fire performance, Composites Science and Technology, 63 (8), 1141–1148. Friedrich, K. (1986a) Friction and Wear of Polymer Composites. Elsevier, Amsterdam. Friedrich, K. (1986b) ‘Wear of reinforced polymers by different abrasive counterparts’, in Friedrich, K. (Ed), Friction and Wear of Polymer Composites. Elsevier, Amsterdam, pp. 233–287. Garcia, M., de Rooij, M., Winnubst, L., van Zyl, W. E. and Verweij, H. (2004) Friction

© 2007, Woodhead Publishing Limited

Multifunctional polymer nanocomposites

277

and wear studies on nylon-6/SiO2 nanocomposites, Journal of Applied Polymer Science, 92 (3), 1855–1862. Gilman, J. W., Kashiwagi, T. and Lichtenhan, J. D. (1997) Nanocomposites: A revolutionary new flame retardant approach, Sampe Journal, 33 (4), 40–46. Hager, A. M. and Davies, M. (1993) ‘Short fibre reinforced, high temperature resistant polymers for a wide field of tribological applications’, in Friedrich, K. (Ed), Advances in Composite Tribology. Elsevier, Amsterdam, pp. 107–157. Hull, T. R., Price, D., Liu, Y., Wills, C. L. and Brady, J. (2003) An investigation into the decomposition and burning behaviour of ethylene–vinyl acetate copolymer nanocomposite materials, Polymer Degradation and Stability, 82 (2), 365–371. Igarashi, A., Terasawa, T., Kanie, M., Yamanobe, T. and Komoto, T. (2005) A morphological study of the effect of carbon nanotube filler on tribology of phenol/formaldehyde resin-based composites, Polymer Journal, 37 (7), 522–528. Jang, B. N. and Wilkie, C. A. (2005) The thermal degradation of polystyrene nanocomposite, Polymer, 46 (9), 2933–2942. Jang, B. N., Costache, M. and Wilkie, C. A. (2005) The relationship between thermal degradation behavior of polymer and the fire retardancy of polymer/clay nanocomposites, Polymer, 46 (24), 10678–10687. Jash, P. and Wilkie, C. A. (2005) Effects of surfactants on the thermal and fire properties of poly(methyl methacrylate)/clay nanocomposites, Polymer Degradation and Stability, 88 (3), 401–406. Jordan, J., Jacob, K. I., Tannenbaum, R., Sharaf, M. A. and Jasiuk, I. (2005) Experimental trends in polymer nanocomposites – a review, Materials Science Engineering A, 393, 1–11. Kiersnowski, A. and Piglowski, J. (2004) Polymer-layered silicate nanocomposites based on poly(e-caprolactone), European Polymer Journal, 40, 1199–1207. Komarneni, S. (1992) Nanocomposites, Journal of Materials Chemistry, 2, 1219–1230. Koo, C. M., Kim, S. O. and Chung, I. E. (2003) Study on morphology evolution, orientational behaviour, and anisotropic phase formation of highly-filled polymer-layered silicate nanocomposites, Macromolecules, 36, 2748–2757. Koo, J. H. and Pilato, L. A. (2005) Polymer nanostructured materials for high temperature applications, Sampe Journal, 41 (2), 7–19. Krishnamoorti, R. and Yurekli, K. (2001) Rheology of polymer layered silicate nanocomposites, Current Opinion on Colloidal Interface Science, 6, 464–470. Kuriger, R. J., Alam, M. K., Anderson, D. P. and Jacobsen, R. L. (2002) Processing and characterisation of aligned vapour grown carbon fiber reinforced polypropylene, Composites A, 33, 53–62. Laachachi, A., Leroy, E., Cochez, M., Ferriol, M. and Cuesta, J. M. L. (2005) Use of oxide nanoparticles and organoclays to improve thermal stability and fire retardancy of poly(methyl methacrylate), Polymer Degradation and Stability, 89 (2), 344–352. Lei, S., Yuan, H., Lin, Z. H., Xuan, S. Y., Wang, S. F., Chen, Z. Y. and Fan, W. C. (2004) Preparation and properties of halogen-free flame-retarded polyamide 6/organoclay nanocomposite, Polymer Degradation and Stability, 86 (3), 535–540. Liu, J., Gao, Y., Wang, F. D. and Wu, M. (2000) Preparation and characteristics of nonflammable polyimide materials, Journal of Applied Polymer Science, 75 (3), 384– 389. Lozano, E. and Barrera, V. (2001) Nanofiber-reinforced thermoplastic composites I: Thermoanalytical and mechanical analyses, Journal of Applied Polymer Science, 79, 125–133.

© 2007, Woodhead Publishing Limited

278

Nanofibers and nanotechnology in textiles

Luo, J.-J. and Daniel, I. M. (2003) Characterisation and modelling of mechanical behaviour of polymer–clay nanocomposites, Composite Science Technology, 63, 1607– 1616. Marosfosi, B., Matko, S. and Marosi, P. A. G. (2006) Fire retarded polymer nanocomposites, Current Applied Physics, 6 (2), 259–261. Merinska, D., Kovarova, L., Kalendova, A., Vaculik, J., Weiss, Z., Chimelova, M., Malac, J. and Simonak, J. (2004) Journal of Polymer Engineering, 23, 241. Michalski, J., Konopka, K. and Trszaska, M. (2003) Description of Al2O3 powders coated by Ni-P particles obtained through an electroless chemical reaction and possibilities to obtain an Al2O3/Ni-P composite, Materials Chemistry and Physics, 81, 407–410. Modesti, M., Lorenzetti, A., Bon, D. and Besco, S. (2006) Thermal behaviour of compatibilised polypropylene nanocomposite: effect of processing conditions, Polymer Degradation and Stability, 91, (4) 672–680. Morgan, A. B., Chu, L. L. and Harris, J. D. (2005) A flammability performance comparison between synthetic and natural clays in polystyrene nanocomposites, Fire and Materials, 29 (4), 213–229. Ohji, T. (1999) Fracture behavior of ceramic based nanocomposites, Key Engineering Materials, 2, 391–396. Park, C. I., Park, O. O., Lim, J. G. and Kim, H. J. (2001) The fabrication of syndiotactic polystyrene/organophilic clay nanocomposites and their properties, Polymer, 42, 7465– 7475. Park, J. H. and Jana, S. C. (2003) Mechanism of exfoliation of nanoclay particles in epoxy–clay nanocomposites, Macromolecules, 36, 2758–2768. Pavoor, P. V., Gearing, B. P., Gorga, R. E., Bellare, A. and Cohen, R. E. (2004) Engineering the friction-and-wear behavior of polyelectrolyte multilayer nanoassemblies through block copolymer surface capping, metallic nanoparticles, and multiwall carbon nanotubes, Journal of Applied Polymer Science, 92 (1), 439–448. Petrovic, Z. S., Javni, I., Waddon, A. and Banhegyi, G. (2000) Structure and properties of polyurethane–silica nanocomposites, Journal of Applied Polymer Science, 76, 133– 151. Porter, D., Metcalfe, E. and Thomas, M. J. K. (2000) Nanocomposite fire retardants – a review, Fire and Materials, 24 (1), 45–52. Preston, C. M. L., Amarasinghe, G., Hopewell, J. L., Shanks, R. A. and Mathys, Z. (2004) Evaluation of polar ethylene copolymers as fire retardant nanocomposite matrices, Polymer Degradation and Stability, 84 (3), 533–544. Rapoport, L., Fleischer, N. and Tenne, R. (2005a) Applications of WS2 (MoS2) inorganic nanotubes and fullerene-like nanoparticles for solid lubrication and for structural nanocomposites, Journal of Materials Chemistry, 15 (18), 1782–1788. Rapoport, L., Nepomnyashchy, O., Lapsker, I., Verdyan, A., Moshkovich, A., Feldman, Y. and Tenne, R. (2005b) Behavior of fullerene-like WS2 nanoparticles under severe contact conditions, Wear, 259 (1–6), 703–707. Reynaud, E., Jouen, T., Gautheir, C. and Vigier, G. (2001) Nanofillers in polymeric matrix: a study on silica reinforced PA6, Polymer, 42, 8759–8768. Schartel, B., Bartholmai, M. and Knoll, U. (2005) Some comments on the use of cone calorimeter data, Polymer Degradation and Stability, 88, (3) 540–547. Sen, R., Zhao, B., Perea, D., Itkis, M. E., Hu, H., Love, J., Bekgarova, E. and Haddon, R. C. (2004) Preparation of single-walled carbon nanotube reinforced polystyrene and polyurethane nanofibers and membranes by electrospinning, Nano Letters, 4, 459– 464.

© 2007, Woodhead Publishing Limited

Multifunctional polymer nanocomposites

279

Shaffer, M. and Kinloch, I. A. (2004) Prospects for nanotube and nanofibre composites, Composites Science and Technology, 64 (15), 2281–2282. Shelley, J. S., Mather, P. T. and de Vries, K. L. (2001) Reinforcement and environmental degradation of nylon-6/clay nanocomposites, Polymer, 42, 5849–5858. Srinath, G. and Gnanamoorthy, R. (2005) Effect of nanoclay reinforcement on tensile and tribo behaviour of nylon 6, Journal of Materials Science, 40 (11), 2897–2901. Szep, A., Szabo, A., Toth, N., Anna, P. and Marosi, G. (2006) Role of montmorillonite in flame retardancy of ethylene–vinyl acetate copolymer, Polymer Degradation and Stability, 91 (3), 593–599. Tidjani, A. (2005) Polypropylene-graft-malefic anhydride-nanocomposites: II – fire behaviour of nanocomposites produced under nitrogen and in air, Polymer Degradation and Stability, 87 (1), 43–49. Usuki, A., Kojima, Y., Kawasumi, M., Okada, A., Fukushima, Y., Kurauchi, T. and Kamigaito, O. (1993) Synthesis of nylon-6 clay hybrid, Journal of Materials Research, 8, 1179– 1184. Vishwanath, B., Verma, A. P., Rao, C. V. S. K. and Gupta, R. K. (1993) Effect of different matrices on wear characteristics of glass woven roving polymer composites, Journal of Composites, 24, 347–353. Vollenberg, P. H. T. and Heikens, D. (1989) Particle size dependence of the Young’s modulus of filled polymers. 1 Preliminary experiments, Polymer, 30, 1656–1662. Vu, Y. T., Mark, J. E., Pham, L. H. and Englehardt, M. (2001) Clay nanolayer reinforcement of cis-1,4-polyisoprene and epoxidized natural rubber, Journal of Applied Polymer Science, 85, 1392–1403. Wagner, H. D., Lourie, O., Feldman, Y. and Tenne, R. (1997) Stress-induced fragmentation of multiwall carbon nanofiber in a polymer matrix, Applied Physics Letters, 72, 188. Wang, Q., Xu, J., Shen, Q. and Liu, W. (1996) An investigation of the friction and wear properties of nanometer Si3N4 filled PEEK, Wear, 196, 82–86. Werner, P., Altstadt, V., Jaskulka, R., Jacobs, O., Sandler, J. K. W., Shaffer, M. S. P. and Windle, A. H. (2004) Tribological behaviour of carbon-nanofibre-reinforced poly(ether ether ketone), Wear, 257 (9–10), 1006–1014. Xing, X. S. and Li, R. K. Y. (2004) Wear behaviour of epoxy matrix composites filled with uniform-sized sub-micron spherical silica particles, Wear, 256, 21–26. Yalcin, B. and Cakmak, M. (2004) The role of plasticizer on the exfoliation and dispersion and fracture behavior of clay particles in PVC matrix: a comprehensive morphological study, Polymer, 45, 6623–6638. Yang, F. and Nelson, G. L. (2004) PMMA/silica nanocomposite studies: synthesis and properties, Journal of Applied Polymer Science, 91 (6), 3844–3850. Yang, F., Ou, Y. and Yu, Z. (1998) Polyamide 6 silica nanocomposites prepared by in situ polymerization, Journal of Applied Polymer Science, 69, 355–361. Yang, F., Yngard, R. and Nelson, G. L. (2005) Flammability of polymer–clay and polymer– silica nanocomposites, Journal of Fire Sciences, 23 (3), 209–226. Yu, L. G., Nie, M. D. and Lian, Y. F. (1996) The tribological behaviour and application of rare earth lubricants, Wear, 197, 206–210. Zanetti, M., Bracco, P. and Costa, L. (2004) Thermal degradation behaviour of PE/clay nanocomposites, Polymer Degradation and Stability, 85 (1), 657–665. Zhang, M. Q., Rong, M. Z., Yu, S. L., Wetzel, B. and Friedrich, K. (2002a) Effect of particle surface treatment on the tribological performance of epoxy-based nanocomposites, Wear, 253, 1088–1095. Zhang, M. Q., Rong, M. Z., Yu, S. L., Wetzel, B. and Friedrich, K. (2002b) Improvement

© 2007, Woodhead Publishing Limited

280

Nanofibers and nanotechnology in textiles

of tribological performance of epoxy by the addition of irradiation grafted nanoinorganic particles, Macromolecular Materials and Engineering, 287 (2), 111–115. Zhang, S. and Horrocks, A. R. (2003) A review of flame retardant polypropylene fibres, Progress in Polymer Science, 28 (11), 1517–1538. Zhang, Y.-Q., Lee, L.-H., Jang, H.-J. and Nah, C.-W. (2004) Preparing PP/clay nanocomposites using a swelling agent, Composite Engineering B, 35, 133–138. Zhu, J. and Wilkie, C. A. (2000) Thermal and fire studies on polystyrene-clay nanocomposites, Polymer International, 49 (10), 1158–1163.

© 2007, Woodhead Publishing Limited

10 Nanofilled polypropylene fibres M. S F I L I G O J S M O L E and K. S T A N A K L E I N S C H E K, University of Maribor, Slovenia

10.1

Introduction

Polypropylene (PP) is, besides polyesters, one of the most widely used polymers for producing synthetic fibres, especially for technical applications. PP fibres are mostly used in different technical fields due to their excellent mechanical properties, high chemical stability and processability. However, because of low surface energy, lack of reactive sites and sensitivity to photo- or thermal oxidation the polymer properties are insufficient for some applications. Therefore, several techniques for fibre modification have been reported, e.g. plasma treatment, chemical modification and nanomodification, i.e. production of nanocoated and nanofilled materials. Polypropylene (PP) fibres have good mechanical properties and can withstand temperatures up to 140 °C (softening point 140–160 °C) before melting at about 170 °C.1 The low polymer density (0.90 g/cm3) offers several specific application possibilities.2 Low costs, good chemical resistance to acid and alkaline environments have greatly influenced the high production quantity of this polymer type.1, 3 Modifications are needed for some purposes due to PP’s high hydrophobicity (moisture regain < 0.1%) and chemical unreactivity and to obtain functional materials with superior physical and mechanical properties for different applications. There is a wide variety of both synthetic and natural crystalline fillers that are able, under specific conditions, to influence the properties of PP. In PP nanocomposites, particles are dispersed on the nano-scale.2, 4 The incorporation of one-, two- and three-dimensional nanoparticles, e.g. layered clays,5 nanotubes,6, 7 nanofibres,8, 9 metal-containing nanoparticles,10 carbon black,11, 12 etc. is used to prepare nanocomposite fibres. However, the preparation of nanofilled fibres offers several possibilities, such as the creation of nanocomposite fibres by dispersing of nanoparticles into polymer solutions, the polymer melt blending of nanoparticles, in situ prepared nanoparticles within a polymeric substrate (e.g. PP/silica nanocomposites prepared in situ via sol–gel reaction),13, 14 the intercalative polymerization of the monomer, 281

© 2007, Woodhead Publishing Limited

282

Nanofibers and nanotechnology in textiles

and the introduction of nanoparticles from dispersion into a porous polymer. Nanomodification creates improved fibre characteristics, e.g. mechanical strength, thermal stability, the enhancement of barrier properties, fire resistance, ion exchange capability, etc., for use in different application fields. In order to follow modification efficiency, various characterization techniques can be used, e.g. X-ray analysis to study composite structure, morphological observations by electron microscopy, mechanical tests, determination of electrokinetic properties, calorimetric measurements. The dispersion of particles within the hybrid system is of fundamental importance, and thus to observe particles a method based on selective etching of the polymer using a plasma or chemical etching can be used.15–17 The dimensions of spherulites and nanoparticles can be determined by image analysis of the micrographs .17

10.2

Polymer layered silicate nanocomposites

Recently, in order to design materials with the desired properties, nanotechnologies have become of great importance since organic–inorganic nano-scale composites frequently exhibit unexpected hybrid properties synergistically derived from the two components.18, 19 In addition, there are also some property improvements caused by nano-scale material modification that could not be realized by conventional fillers as, for example, a general flame-retardant character and a dramatic improvement in barrier properties. The properties of the particles themselves (size, shape, distribution) can profoundly change the characteristics of a polymer system. Therefore understanding the structure/property relations in polymer/nanoparticle nanocomposites is of major significance.19, 20 The idea of mixing polymers with appropriately modified clay minerals and synthetic clays is not new. Polymer layered silicate nanocomposites (PLNC) were reported by Carter and coworkers in patent literature, as early as 1950.21 However, two major findings have led to its revival. Firstly, researchers from Toyota reported a polyamide 6/montmorillonite (MMT) composite with a remarkable enhancement of thermal and mechanical properties due to very moderate inorganic loading. In addition, it was discovered that it is possible to melt-mix polymers with clays without the use of organic solvents.19, 22 Since then, the technology for incorporating nanoparticles into a PP matrix has offered several challenges for research, and for industrial applications. Several different nanoparticles for nanofilled composites, e.g. layered silicates,2, 19, 23–27 silica nanoparticles,11, 28, 29 carbon black,11, 12 carbon nanotubes,6, 7 metal containing nanoparticles,10 elastomeric nanoparticles30 and TiO2,31 have been reported. The most common PP nanocomposites are composed of organically modified silicates, e.g. MMT and polymeric matrix.

© 2007, Woodhead Publishing Limited

Nanofilled polypropylene fibres

283

Generally, polymer–clay nanocomposites use smectic-type clays as fillers, such as hectorite, MMT, kaolin or synthetic mica, all minerals with a layered structure.2, 23, 32, 33 They are of great industrial value because of their high aspect ratio, plate morphology, intercalative capacity, natural abundance and low costs.23 The layers are characterized by a thickness of about 1 nm, and the other dimensions vary from 30 nm to several micrometres or more. Several layers are stacked in clay particles, kept together by weak van der Waals forces. The performance of polymer–clay nanocomposites strongly depends on the breaking-up of clay particles in the polymer matrix.2 However, the most frequently used layered silicate, MMT, is a naturally occurring 2:1 phyllosilicate, which has the same layered and crystalline structure as talc and mica but a different layer charge. A central octahedral sheet of alumina fused between two external silica tetrahedral sheets (the oxygens from the octahedral sheet also belong to the silica tetrahedral) forms the layers of the crystal lattice. Isomorphic substitution within the layers (e.g. Mg2+ or Fe+2 substitutes Al3+) causes a negative charge, which is defined through the charge exchange capacity (CEC): MMT is typically 90–120 meq/100 g, depending on the mineral’s origin.19, 33 Interlayers or galleries are parallel layers forming stacks with a regular van der Waals gap between them. The negative charge of the pristine MMT is balanced by cations (Na+, Li+, Ca++) from the interlayer.19, 25 There is a considerable difference in polarity between polymer and clay,2 the pre-treatment of both polymer and clay is, therefore, necessary. Clays are usually modified by cationic surfactants such as organic ammonium salts (e.g. stearyl ammonium18, 34) or alkyl phosphonium, that make the clay surface more organophilic.2 PP’s compatibility with organoclays is obtained by grafting polar functional groups such as maleic anhydride, diethyl maleate (DEM),35 methyl methylacrylate and butylacrylate,36 and mono-ethanol stearamide and others. Another approach to enhance the compatibility between PP and modified clay is suggested by Kim et al. 37 They prepared PP layered silicate nanocomposites via melt mixing of three components: PP, layered silicates modified with octadecylamine (C18-MMT) and antioxidant.37

10.2.1 Preparation of layered silicate polypropylene nanocomposites Three different techniques can be used to prepare nanocomposites:25, 38 1. Dispersing the layered silicate in a solution of the polymer in an organic solvent, followed by either solvent evaporation or polymer precipitation. The limitations of this method are shown by the large amounts of organic solvents, polymer solubility and poor filler dispersion.25

© 2007, Woodhead Publishing Limited

284

Nanofibers and nanotechnology in textiles

2. Melt intercalation of the polymer into previously organo-modified silicates. It is a very effective method when polymers, such as polyamide-6, polysiloxane and even polystyrene are intercalated. For the formation of PP-based nanocomposites, a compatibilizer such as maleic anhydride grafted polypropylene (PP-MA) or acrylic acid grafted PP is involved, which improves the polyolefin–filler interactions.25 3. Intercalative polymerization of the monomer. The monomer together with the polymerization initiator or catalyst is intercalated within the silicate layers and polymerization is initiated either thermally or chemically. The macromolecule chains exfoliate in silicate layers and make them disperse in the polymer matrix evenly.36 This method allows the formation of nanocomposites based on non-polar polymers such as polyolefins.25 The dispersion of clay particles in polymers can result in the following: • Clay sheets may remain stacked in structures called tactoids. Original mineral structure does not contribute to any improvement compared with the usual microcomposites with a low filler loading.2, 23 • Intercalated nanocomposite, where some polymer molecules are inserted between individual silicate layers.23 • Delaminated or exfoliated nanocomposite in which the layered structure of the clay is disrupted by the silicate layers no longer being close enough to interact with each other, and the nanometric particles are fully dispersed in the matrix.2, 23, 24 The coherent order of stacked layers strongly depends on the clay content. The grafted polar groups (e.g. maleic anhydride) in the PP-MA chains promote interaction with the clay particles by the diffusion of the PP chain into the space between the silicate galleries (intercalating sites); therefore, the decrement of intercalating sites leads to exfoliation towards the individual silicate layers.34 In order to favour these morphologies, long chain alkylammonium cations are exchanged for the constitutive cations (Na+, Li+) of the layered silicates, making the silicate surface more lipophilic, and appropriate for interacting with the organic polymer.25 Above all, the reinforcement effect is observed by exfoliated nanocomposites, since the nanometric dispersion of clay platelets creates a maximal interfacial surface between the filler and the polymer matrix.25

10.3

The structure and properties of layered silicate polypropylene nanocomposites

Generally isotactic polypropylene (iPP) is used for nanofilled materials, although nanocomposites prepared from syndiotactic polypropylene (sPP) have also been reported.39, 40 The degree of crystallinity in PP depends on the

© 2007, Woodhead Publishing Limited

Nanofilled polypropylene fibres

285

processing conditions, but is usually between 50 and 60%.1 The common crystalline PP form is α monoclinic, but other forms (β and γ) have been observed. Polymorphism is important for technological reasons, because each phase has different physical and mechanical characteristics. The crystals are chain-folded lamellae and are aggregated into spherulites or row-nucleated structures, depending on the processing conditions.1 Some additives in polymers become nucleation centres, leading to an increase of crystal growth in the crystallization process of the polymer. A foreign surface reduces the nucleus size needed for crystal growth through the creation of the interface between polymer crystal and substrate. The important effect of such nucleation is a modification of polymer morphology, which can result in change of crystallographic form. The nucleating efficiency of various organic and inorganic fillers of iPP – talc, chalk, wood flour, nano clay particles, carbon black, chitosan – was studied by Mucha and Królikowski.41 The best nucleating agents in this research were talc and carbon black. The organic filler as a chitosan powder forms amorphous inclusions in the composites on which iPP molecules cannot be adsorbed. Their presence disturbs a macromolecular diffusion and delays the crystallization process of iPP.41 A mixed α and β phase was observed in nanocomposites.2 The relative amount of β phase never exceeded 30%, denoting a rather uniform crystallization in the bulk and on the samples’ surface. The authors presume that the high β content could be attributed to stearic-acid derivatives which were added as processing aid agent and could act as nucleating agent for β phase. Stearic-acid derivatives are known as promoters of β crystallographic phase.2 X-ray diffraction (XRD), is commonly used to probe the nanocomposite structure. However, XRD can only detect the periodically stacked clay layers, disordered or exfoliated layers are undetected.19 In general, in layered silicatefilled polymers a coexistence of exfoliated, intercalated and disordered layers is observed. The monoclinic crystal structure of neat PP and nanofilled PP shows reflections assigned to planes (100), (040), (130) and others.23 The peaks recorded between 13° and 26° (14°, 17°, 18.7°, 20°, 21.5°, 25.5°) are associated with the α crystalline structure of the PP corresponding to the basal reflections on the (110), (040), (130) and (111) crystallographic planes.42 The formation of an intercalated nanocomposite structure can be defined by analysing the interlayer spacing (d001) of clay, although the diffraction maximum originating from the (002) crystallographic plane can be observed on the scattering curves also.23 Only when d001 in the composite is higher than in the pure clay then the polymer molecules were positioned between clay layers and, hence, an intercalated nanocomposite is produced. If the peak corresponding to d001 is not observed in a polymer/clay diffractogram, this implies that an exfoliated nanocomposite structure was most likely obtained.23

© 2007, Woodhead Publishing Limited

286

Nanofibers and nanotechnology in textiles

Pristine MMT shows a characteristic diffraction peak between 6.94° 34 and 7.3° 17 2θ corresponding to the (001) plane diffraction. It corresponds to an interlayer spacing of 1.2 nm.17 For modified MMT the interlayer spacing of 1.9 nm obtained from the characteristic peak (001) at 2θ 4.7° was reported.43 Pure organoclay cloisite 15-A shows a (001) basal signal at 2θ 2.8°, the corresponding d001 are 3.15 nm distant from one another.42 In nanocomposites a shift of this peak towards small angles would be associated with intercalation while its disappearance would be a sign that exfoliation has occurred.34 However, the disappearance of a basal peak in a wide-angle X-ray scattering (WAXS) diffractogram alone should not be intended as a clear sign of exfoliation, unless it is supported by small-angle X-ray scattering (SAXS) which is sensitive to the crystalline regions organized in lamellar stacks.42 The SAXS profiles of pure PP samples and nanocomposites registered in the lower angular region show the presence of a maximum, which is associated with the long period resulting from the presence of a macro-lattice formed by centres of adjacent lamellae of the polymer.43 Evidence that the nanocomposite structure was not formed is shown by the diffraction peaks due to the presence of MMT in PP polymer which occur in the range of 2θ 2–8°. For PP/unmodified bentonite d001 = 1.51 nm (2θ = 5.82°), which is similar to the value observed for the clay (d001 = 1.50 nm; 2θ = 5.70°). On the other hand, the system containing PP/modified bentonite forms a diffraction peak at 2θ = 2.18° which is attributed to d001 = 4.21 nm, and is much higher than the one for the modified bentonite clay (d001 = 2.05 nm, 2θ = 4.32°).23 This indicates that the organophilic bentonite forms a nanocomposite with PP, as deduced from the increase in the distance between the clay lamellae (d001) caused by the intercalation of polymer molecules.23 The characteristics of the coupling agent (molecular weight and grafting content) influence the process of the clay intercalation.17 The use of a coupling agent with low molecular weight and high grafting content leads to high and uniform intercalation, while exfoliation does not arise. In contrast, high molecular weight and low grafting content leads to more heterogeneous intercalation accompanied with some exfoliation. Disordered and more distanced layer structure is observed.17 The presence of clay affects the PP morphology, i.e. spherulite size, and brings about crystalline orientation, causes the crystallization at lower temperature, but does not change the PP crystallinity.17

10.3.1 Preparation techniques and nanocomposite structure There are two possibilities for preparing nanocomposites: 1. Master batches containing a high content of clay (e.g. 50% wt of clay) in PP are initially prepared. The master batch is then added to neat PP in

© 2007, Woodhead Publishing Limited

Nanofilled polypropylene fibres

287

appropriate amounts to obtain nominal contents of 0.5–5 wt% clay in composites. The extrusion temperature usually used is 150–210°C. The extruded composites are cooled, generally pelletized and then dried23, 42 and re-extruded using a higher screw speed. 2. Preparation of intercalated PPCNs (PP/clay nanocomposite) using a modified PP and organophilic clay via melt extrusion processing.18, 34, 44 PP–clay hybrids are prepared by using melt blending technology, i.e. simple melt mixing of three components – i.e. PP, modified PP oligomers and intercalated clays.18, 45 There are two important factors to achieve the exfoliated and homogeneous dispersion of the layers in the hybrids: the intercalation capability of the oligomers in the layers and the miscibility of the oligomers with PP. Almost completed hybrids were obtained in the case where the PP-MA has both intercalation capability and miscibility.18 Mostly nanofilled PP is produced by injection molding; however, composites prepared from iPP and sPP, respectively, and organic layered silicate by the spinning procedure have been reported.39, 42, 46–48 The XRD spectra show that both the spun fibres and injection moulded specimens contain a mixture of intercalated and exfoliated organosilicate layered (OSL) structures.42 On the diffraction curves of nanocomposite fibres no OSL peaks over the full 2θ range from 2° to 30° were observed. In contrast the injection moulded specimens exhibit six distinct OSL peaks in the 2θ region between 2° and 12° and two additional peaks between 26° and 30°. It can be concluded that under melt spinning conditions, the polymer matrix can exfoliate the OSL gallery more effectively than injection moulding. Spinning velocity influences the structure formed, i.e. OSL structures exhibit more improved exfoliation in the fibres spun at higher velocity. The exfoliation of pre-intercalated OSL structures was much more significant under an extensional velocity gradient during melt spinning rather than shear flow in injection molding.42 Joshi et al. reported that nanoclay reinforced PP nanocomposites could be spun and drawn successfully for 0.5, 1.0 and 1.5 wt% of the modified clay loading. Beyond 1.5 wt% the spinnability is poor.49 The modified clay content in the range of 0.5-1.0 wt% gives optimum properties.49

10.3.2 Properties of layered silicate polypropylene nanocomposites Nanoparticles are able to provide PP with stiffening, reinforcing and toughening effects at rather low filler concentration. The influence of processing conditions on the nanocomposite structure, i.e. intercalated or exfoliated, and on the enhancement of mechanical properties of PP nanocomposites was studied by different researchers.44 Most polymer/clay nanocomposites studies report

© 2007, Woodhead Publishing Limited

288

Nanofibers and nanotechnology in textiles

tensile properties as a function of MMT content.19 The enhancements are strictly related to the processing conditions, the filler content and the presence of compatibilizer.44 When comparing properties of neat PP and nanocomposites, there is a sharp increase of the Young’s modulus for very small inorganic loadings followed by much slower increase beyond approximately 5 wt%. With increasing MMT content the yield stress does not change markedly compared to the neat polymer value. Considering the same processing conditions, the elastic modulus is higher in the presence of compatibilizer for different filler contents due to the polymer–inorganic adhesion improvement. This implies that the stress is much more efficiently transferred from the polymer matrix to the inorganic filler, resulting in a higher increase in the Young’s modulus.19, 42 Similar improvements in mechanical properties can also be achieved by other layered fillers; however, much higher filler loadings are required (e.g. by loading 30–60 wt% of talc or mica).19 Silicate clay can increase the modulus, decomposition temperature, yield strength and fatigue strength, and has no effect on glass transition temperature and melt temperature.50 Polymer/silicate nanocomposites are characterized by very strong reduction of gas and liquid permeability and at the same time the solvent uptake decreases accordingly.19 When single layers are dispersed in a polymer matrix the resulting nanocomposite is optically clear in the visible region, as clays are just 1 nm thick, whereas there is a loss of intensity in the UV region mostly due to scattering by the MMT particles.19 Five principal types of generic flame-retardant systems for inclusion in PP fibres have been identified as phosphorus-containing, halogen-containing, silicon-containing, metal hydrate and the more recently developed nanocomposite flame-retardant formulations.51, 52 The most effective are halogen– antimony and phosphorus–bromine combinations, however their application is limited by ecological criteria.51 Several kinds of nanocompounds can be used to enhance the flame retardancy of PP, such as modified MMT, TiO2, Sb2O3 and boroxosiloxanes. Of these, the sodium cation exchanged MMT is the most common because of its low price.19, 51, 53 The general view of the flame-retardant mechanism is that a high-performance carbonaceous silicate char builds up on the surface during combustion: this insulates the underlying material and slows the mass loss rate of decomposition products.19, 53 All MMT-based composite systems nowadays reported show that MMT must be nanodispersed for it to affect the flammability.53 Investigations have shown that the largest increase in mechanical properties are obtained in exfoliated nanocomposite forms while the intercalated materials show the best fire performance.51 The results for thermal ageing showed that PP compounds based on organically modified bentonite had higher thermal stability than those with natural clay23 because of the formation of a nanocomposite structure with

© 2007, Woodhead Publishing Limited

Nanofilled polypropylene fibres

289

reduced oxygen diffusion into the material. However, the thermal degradation of PP with the modified clay is higher than the pure polymer. The phenomenon is attributed to the acidic nature of the clay, the interaction between the clay and PP stabilizers and the decomposition of the organic salt during processing. When the degradation was done in the melt state the thermal stability of the composites may be higher than the pure polymer.23 Nanostructured materials, i.e. PP filled with an extra-pure synthetic fluorohectorite modified by means of interlayer exchange of sodium cations for protonated octadecylamine NH 3+ (ODA) in a weight concentration of maximum 6%, may find new and upgraded application in the electrical and electronic industry, replacing conventional insulation.54

10.4

Nanosilica filled polypropylene nanocomposites

The incorporation of inorganic fillers in polymers contributes effectively to the improvement of the mechanical properties and in particular the toughness; however, the procedure is often limited owing to the high filler contents that are needed for desired composite properties. For this reason nanocomposites with unique nano-scale microstructure, obtained by low particle loading represent a particular advantage. However, it is very difficult to disperse nanoparticles in a polymeric matrix homogeneously owing to a strong tendency to agglomerate. An irradiation grafting method for modification of nanoparticles, e.g. nanosilica, presents an appropriate solution.55 Through irradiation grafting polymerization, nanoparticle agglomerates transform into a nanocomposite microstructure. Grafted monomers of low molecular weight can penetrate into the agglomerated nanoparticles easily and react with the activated sites of the particles inside and outside the agglomerates. Strong interfacial interactions with the polymeric matrix during the subsequent mixing procedure result. Different monomers are appropriate for the process, e.g. styrene, methyl methacrylate, butyl acrylate, ethyl acrylate, methyl acrylic acid, vinyl acetate and others.56 Reinforcing and toughening effects were obtained by mixing of modified nanosilica into PP, precisely tensile properties – such as strength, modulus and elongation at break – were increased.56 Low nanosilica-loaded PP composites were produced by Rong et al.57 by a conventional compounding technique in which the nanoparticles were grafted by polystyrene using irradiation beforehand. The significantly increased hydrophobility of the nanoparticles obtained by the grafted polymers assures better interfacial interaction between the nanoparticles and the polymer. PP-based nanocomposites filled with polystyrene-grafted nano-SiO2 (SiO2-g-PS) show an increase of strength with a rise in filler content only to the filler content critical value and afterwards the strength remains constant irrespective of filler content.57

© 2007, Woodhead Publishing Limited

290

Nanofibers and nanotechnology in textiles

Embedded precipitated nanosilica enhances stiffening, reinforcing and toughening of PP nanocomposites at lower concentration as fumed nanosilica. By grafting with polymers such as styrene, methyl methacrylate, butyl acrylate, and ethyl acrylate, the efficiency in the composites’ mechanical properties is increased depending on the grafting polymer used.58 Nanosilica particles can be prepared in situ via a sol–gel reaction. Jain et al. reported on PP–silica nanocomposites preparation by the sol–gel procedure13, 16, 59 from tetraethoxy orthosilicate (TEOS), which was added to PP powder. The effect of in situ formed silica nanoparticles on the nonisothermal crystallization kinetics of PP–silica nanocomposites was studied in addition. Silica nanoparticles act as a nucleating agent, thereby a twostage crystallization process was observed. Within the primary stage, nucleation and spherulitic growth occurred, while the secondary stage includes the perfectioning of crystals; thereby more perfect crystals are formed. The nucleation effect of silica results in a more narrow lamellar thickness distribution.11 Silica nanoparticles promote β crystallographic phase.11 When nanosilica–PP nanocomposites are prepared from isotactic PP and SiO2 nanoparticles by melt-mixing, a compatibilizer should be added to break off the hydrogen bonds created between the nanoparticles’ surface hydroxyl groups. By the addition of PP copolymer with maleic anhydride groups (PP-g-MA), the maleic anhydride groups can react with the surface hydroxyl groups of SiO 2 nanoparticles and, through this, reduce agglomeration.60 Mechanical properties such as tensile strength at break and Young’s modulus are influenced by the content of silica particles and copolymer content.60 After the critical concentration of nanoparticles (2.5 wt%) has been exceeded, the tensile and impact strength are decreased. This behaviour is attributed to appearance of large silica particle agglomerates.60 Furthermore, recently research work has focused on the possibility of using nanofiller as a compatibilizer for immiscible polymer blends. Quin Zhang with co-authors reports an improvement of the compatibility of PP/ polystyrene blends with the addition of nano-SiO2 particles by changing the phase morphology and properties of the two polymers.61 A drastic reduction of polystyrene phase size and a very homogeneous size distribution were observed by introducing nano-SiO2 particles at short mixing time, but at longer mixing time an increase of polystyrene size was observed, indicating a kinetics-controlled compatibilization of nano-SiO2 particles by changing the phase morphology and properties of the two polymers.61 Low nanoparticle-loaded polymer composites with improved mechanical performance can be prepared by a conventional melt blending technique in which the nanoparticles are chemically pregrafted by diglycidyl ether of bisphenol-A (DGEBA).62 Reddy et al.62 prepared a composite by melt blending propylene–ethylene copolymer (EP) with DGEBA-grafted nanosilica. The addition of epoxy resin grafted nanosilica to the polymer matrix produced a

© 2007, Woodhead Publishing Limited

Nanofilled polypropylene fibres

291

homogeneous dispersion of particles in the form of micro-domains and thereby provides EP with stiffening, strengthening and toughening effects.

10.5

Calcium carbonate and other additives

Calcium carbonate (CaCO3) is one of the most commonly used inorganic fillers in PP.63 Several authors reported that mechanical properties of PP filled with nano-sized calcium carbonate particles are essentially improved.64 Chan et al. prepared nanofilled PP fibres by adding 9.2 vol% surfactant treated CaCO3. The impact strength of the modified PP was more than twice as high when compared with the neat PP. The intrinsic toughness of the PP matrix influences the toughening effect of the nanoparticles. The highest increase in toughness was found in the case of the moderate matrix toughness.63 In attempts to prevent the agglomeration of nanoparticles several surface modifications of the nanoparticles have have reported, e.g. nano-CaCO3 treated with titanate coupling agent, silane coupling agent, etc.63 The particle size (0.07–1.9 µm) has no influence on the thermal composite’s properties, and also the particle content does not affect the melting temperature or the crystallinity of PP. Surface treatment of the particles by the stearic acid coating showed an important positive effect on the impact strength due to the improved dispersion of the CaCO3 particles. In addition PP molecular weight should be considered when designing a CaCO3–PP nanocomposite due to its profound impact on the toughening properties.65

10.5.1 Carbon black-filled polypropylene composites Carbon additives are used in polymer composites as fillers, reinforcing agents and pigments. In addition, carbon black is used to enhance UV stability, electrical conductivity and weather resistance.12 PP geotextiles containing carbon black are applied for soil reinforcement, filtration and other construction purposes.11, 12 In addition, carbon black can be used for preparation of magnetostrictive materials which are defined as materials that undergo a change in shape due to the change in the magnetization state of the material.66 Carbon black, which has a polyaromatic structure containing various oxygen functional groups, is produced by partial combustion of liquid or gaseous hydrocarbons. Its antioxidant activity is due to the catalytic decomposition of peroxides and free radical scavenging, which is more effective at low temperature.11 The effect of carbon black on thermal and photo-oxidative degradation of oriented PP geotextile tapes was studied by Horrocks et al.11 The influence of particle size (16–60 nm), structure, aggregate shape, specific surface volatile content and concentration was determined. Significant increases in thermal stabilities were observed when increasing carbon black content in PP composite.

© 2007, Woodhead Publishing Limited

292

Nanofibers and nanotechnology in textiles

However, the opposite effect of the carbon black with highest volatile content on thermal stability was observed.12 This phenomenon is attributed to the adsorption of antioxidant by the carbon black surface or to the sensitization of thermal oxidative reactions by the surface oxygenated groups present.11 It has been proved that different behaviour is related to the volatile content, due to the presence of carboxylic and sulphonic acid groups.12 Not only are the temperature and rate of decomposition influenced by embedded carbon black particles, but also the decomposition products. The presence of 1alkene oligomers with 3n C atoms is reduced, while 2-alkenes and 1-alkenes with 3n+1 C atoms are increased. Carbon black promotes chain scission and participates in the radical transfer reactions.12 Incorporated particles of carbon black, especially those with small particle size, improve UV durability.11

10.5.2 Alumina nanoparticle-filled polypropylene Nanocomposites containing 1.5–5.0 wt% of spherical alumina (Al2O3) nanoparticles, which were pre-treated with silane coupling agent, were prepared by Zhao and Li.67 Tensile tests show that Young’s modulus and the yield strength of the nanocomposite increase with the particle content increasing, suggesting that the interfacial interaction between the nanoparticles and the matrix is relatively strong. Structural investigation, i.e. X-ray analysis, differential scanning calorimetry (DSC) and optical microscopy measurements, show that a small amount of the β-crystal form results after adding the Al2O3 nanoparticles. The Al2O3 nanoparticles reduce the size of PP spherulites and enhance the crystallization temperature of PP, by acting as an effective nucleating agent.67

10.5.3 Polypropylene-polyhedral oligomeric silsesquioxane nanocomposites Polyhedral oligomeric silsesquioxanes (POSS) were first synthesized in 1946. They belong to the group of silsesquioxanes characterized by the general formula (RSiO1,5)n, where R is hydrogen or an organic group, such as alkyl, aryl or any of their derivatives. POSS can be easily functionalized by chemically altering the R substituent group, thus having the potential to undergo copolymerization or grafting reactions. The addition of the thermally robust POSS moiety drastically modifies the polymer thermal properties supplying greater thermal stability to the polymer matrix, also allowing the tailoring of the polymer glass transition temperature by varying the POSS concentration.68 The addition of POSS also improves mechanical properties and reduces polymer composite flammability. Fu et al.69, 70 first prepared octamethyl-POSS/PP composites by melt blending and studied their crystallisation behaviour.69 Fina et al.68 investigated

© 2007, Woodhead Publishing Limited

Nanofilled polypropylene fibres

293

the influence of the POSS substituent groups on the morphological and thermal characteristics of melt-blended POSS/PP composites by varying the aliphatic chain length. Composites with octamethyl-, octaisobutyl- and octaisooctyl-POSS were prepared. Increasing alkyl chain length provoked substantial differences in morphology of the composites, i.e. the compatibility between POSS and PP, nucleating ability, spherulitic morphology formation, etc.68 Recently the same authors reported about thermal and thermo-oxidative degradation of PP-based composites, using different metal containing POSS. Metal POSS derivatives (Al and Zn) were prepared by deprotonation of incompletely condensed POSS trisilanol (i-C4H9)7Si7O9(OH)3 with either triethylaluminium or diethylzinc.71

10.6

Conclusion

In recent years, nanostructured materials have attracted much attention because of their potential for large gains in mechanical and physical properties as compared with standard structural materials. Although numerous different nanofilled products were developed, there are still unlimited challenges for researchers and technology.

10.7

References

1. Lewin M., Pearce E.M., Handbook of Fiber Chemistry, New York, Marcel Dekker, 1998. 2. Benetti E.M., Causin V., Marega C., Marigo A., Ferrara G., Ferraro A., Consalvi M., Fantinel F., ‘Morphological and structural characterization of polypropylene based nanocomposites’, Polymer, 2005 46 8275–8285. 3. Cook J.G., Handbook of Textile Fibres, II Man-made Fibres, Durham, Merrow Publishing, 1984. 4. Poole C.P., Owens F.J., Introduction to Nanotechnology, Hoboken, New Jersey, Wiley Interscience, 2003. 5. Manias E., Touny A., Wu L., Lu B., Strawhecker K., Gilman J.W., Chung T.C., ‘Polypropylene/silicate nanocomposites, synthetic routes and materials properties’, Polymeric Materials: Science & Engineering, 2000 82 282–283. 6. Chatterjee A., Deopura B.L., ‘High modulus and high strength PP nanocomposite filament’, Composites Part A – Applied Science and Manufacturing, 2006 37/5 813– 817. 7. Kearns J.C., Shambaugh R.L., ‘Polypropylene fibers reinforced with carbon nanotubes’, Journal of Applied Polymer Science, 2002 86/8 2079–2084. 8. Hine P., Broome V., Ward I., ‘The incorporation of carbon nanofibres to enhance the properties of self reinforced, single polymer composites’, Polymer, 2005 46 10936– 10944. 9. Li J., Wang Q., Chan C., Wu J., ‘Effect of molding temperature on crystalline and phase morphologies of HDPE composites containing PP nano-fibers’, Polymer, 2004 45 5719–5727.

© 2007, Woodhead Publishing Limited

294

Nanofibers and nanotechnology in textiles

10. Gubin S.P., ‘Metal-containing nano-particles within polymeric matrices: preparation, structure, and properties’, Colloids and Surfaces, A: Physicochemical and Engineering Aspects, 2002 202 155–163. 11. Horrocks A.R., Mwila J., Miraftab M., Liu M., Chohan S.S., ‘The influence of carbon black on properties of oriented polypropylene, 2. Thermal and photodegradation’, Polymer Degradation and Stability, 1999 65 25–36. 12. Jakab E., Omastova M., ‘Thermal decomposition of polyolefin/carbon black composites’, Journal of Analytical Applied Pyrolysis, 2005 74 204–214. 13. Jain S., Goossens H., van Duin M., Lemstra P., ‘Effect of in situ prepared silica nano-particles on non-isothermal crystallization of polypropylene’, Polymer, 2005 46 8805–8818. 14. Hribernik, S., Sfiligoj Smole, M., Stana-Kleinschek K., Strnad S., Bele M., Ribitsch V., ‘Preparation of nanocomposite cellulose fibres’ 4th Symposium of science and technology of nanomaterials in Slovenia, Slonano 2005. 15. Mozetič M., Zalar A., Panjan P., Bele M., Pejovnik S., Grmek R., ‘A method of studying carbon particle distribution in paint films’, The Solid Films, 2000 376 5–8. 16. Stakne K., Dissertation, University of Maribor, 2004. 17. Perrin–Sarazin F., Ton-That M.T., Bureau M.N., Denault J., ‘Micro- and nano-structure in polypropylene/clay nanocomposites’, Polymer, 2005 46 11624–11634. 18. Kawasumi M., Hasegawa N., Kato M., Usuki A., Okada A., ‘Preparation and mechanical properties of polypropylene-clay hybrids’, Macromolecules 1997 30 6333–6338. 19. Manias E., Touny A., Wu L., Strawhecker K., Lu B., Chung T.C., ‘Polypropylene/ montmorillonite nanocomposites. Review of the synthetic routes and material properties’, Chemistry of Materials, 2001 13 3516–3523. 20. Privalko V.P., Shumsky V.F., Privalko E.G., Karaman V.M., Walter R., Friedrich K., Zhang M.Q., Rong M.Z., ‘Viscoelasticity and flow behaviour of irradiation grafted nano-inorganic particle filled polypropylene composites in the melt state’, Journal of Materials Processing Technology, 2003 137 208–213. 21. Carter L., Hendricks J.G., Bolley D.S., US Patent 2,531,396, 1950 (assigned to National Lead Co.), 31. 22. Vaia R.A., Ishii H.‚ Giannelis E.P., ‘Synthesis and properties of 2-dimensional nanostructures by direct intercalation of polymer melts in layered silicates’, Chemistry of Materials, 1993 5 1694–1696. 23. Filho F.G.R., Mélo T.J.A., Rabello M.S., Silva S.M.L., ‘Thermal stability of nanocomposites based on polypropylene and bentonite’, Polymer Degradation and Stability, 2005 89 383–392. 24. Ray S.S., Okamoto M., ‘Polymer/layered silicate nanocomposites: a review from preparation to processing’, Progressin Polymer Science, 2003 28 1539–1641. 25. Alexandre M., Dubois P., Sun T., Garces J.M., Jérôme R., ‘Polyethylene-layered silicate nanocomposites prepared by the polymerization-filling technique: synthesis and mechanical properties’, Polymer, 2002 43 2123–2132. 26. Han B., Ji G., Wu S., Shen J., ‘Preparation and characterisation of nylon 66/ montmorillonite nanocomposites with co-treated montmorillonites’, European Polymer Journal, 2003 39 1641–1646. 27. Tang Y., Hu Y., Song L., Zong R., Gui Z., Chen Z., Fan W., ‘Preparation and thermal stability of polypropylene/montmorillonite nanocomposites’, Polymer Degradation and Stability, 2003 82 127–131. 28. Rong M.Z., Zhang M.Q., Zheng Y.X., Zengh H.M., Friedrich K., ‘Improvement of

© 2007, Woodhead Publishing Limited

Nanofilled polypropylene fibres

29.

30.

31.

32.

33.

34.

35.

36. 37.

38.

39.

40.

41.

42.

43.

44.

295

tensile properties of nano-SiO2/PP composites in relation to percolation mechanism’, Polymer, 2001 42 3301–3304. Rong M.Z., Zhang M.Q., Zheng Y.X., Zengh H.M., Walter R., Friedrich K., ‘Structure– property relationships of irradiation grafted nano-inorganic particle filled polypropylene composites’, Polymer, 2001 42 167–183. Zhang M., Liu Y., Zhang X., Gao J., Huang F., Song Z., Wei G., Qiao J., ‘The effect of elastomeric nano-particles on the mechanical properties and crystallization behaviour of polypropylene’, Polymer, 2002 43 5133–5138. Wei Q., Gao W., Wang X., ‘Preparation and characterisation of functionally nanostructured fibres’ 2nd International Textile Clothing & Design Conference, Magic World of Textiles, Dubrovnik, 2004, 151–156. Mareri P., Bastide S., Binda N., Crespy A., ‘Mechanical behaviour of polypropylene composites containing fine mineral filler: effect of filler surface treatment’, Composites Science and Technology, 1998 58 747–752. Lagaly G., Ziesmer S., ‘Colloid chemistry of clay minerals: the coagulation of montmorillonite dispersions’, Advances in Colloid and Interface Science, 2003 100– 102 105–128. Nam P.H., Maiti P., Okamoto M., Kotaka T., Hasegawa N., Usuki A., ‘A hierarchical structure and properties of intercalated polypropylene/clay nanocomposites’, Polymer 2001 42 9633–9640. García-López D., Picazo O., Merino J.C., Pastor J.M., ‘Polypropylene–clay nanocomposites: effect of compatibilizing agents on clay dispersion’, European Polymer Journal, 2003 39 945–950. Ding C., Jia D., He H., Guo B., Hong H., ‘How organo-montmorillonite truly affects the structure and properties of polypropylene’, Polymer Testing 2005 24 94–100. Kim J.H., Koo C.M., Choi Y.S., Wang K.H., Chung I.J., ‘Preparation and characterization of polypropylene/layered silicate nanocomposites using an antioxidant’, Polymer, 2004 45 7719–7727. Lee J.H., Jung D., Hong C.E., Rhee K.Y., Advani S.G., ‘Properties of polyethylenelayered silicate nanocomposites prepared by melt intercalation with a PP-g-MA compatibilizer’, Composites Science and Technology, 2005 65 1996–2002. Mlynarčiková Z., Kaempfer D., Thomann R., Mülhaupt R., Borsig E., Marcinčin A, ‘Syndiotactic polypropylene/organoclay nanocomposite fibres: influence of the filler and compatibilizer on the fibre properties’, 2nd International Textile Clothing & Design Conference, Magic World of Textiles, Dubrovnik, 2004. Gorrasi G., Tortora M., Vittoria V., Kaempfer D., Mülhaupt R., ‘Transport properties of organic vapors in nanocomposites of organophilic layered silicate and syndiotactic polypropylene’, Polymer, 2003 44 3679–3685. Mucha M., Królikowski Z., ‘Application of DSC to study crystallization kinetics of polypropylene containing fillers’, Journal of Thermal Analysis and Calorimetry, 2003 74 549–557. Lew C.Y., Murphy W.R., McNally G.M., Abe K., Yanai S., ‘Metallocene–catalyzed polypropylene nanocomposite fibers: the effect of melt spinning on the exfoliation of layered-silicates’, Polymer Fibers 2004, Manchester 2004. Bertini F., Canetti M., Audisio G., Costa G., Falqui L., ‘Characterization and thermal degradation of polypropylene–montmorillonite nanocomposites’, Polymer Degradation and Stability, 2006 91 600–605. Modesti M., Lorenzetti A., Bon D., Besco S., ‘Effect of processing conditions on

© 2007, Woodhead Publishing Limited

296

45. 46.

47.

48.

49. 50.

51. 52.

53.

54.

55.

56.

57.

58.

59. 60.

Nanofibers and nanotechnology in textiles morphology and mechanical properties of compatibilized polypropylene nanocomposites’, Polymer, 2005 46 10237–10245. Utracki L.A., Sepehr M., Li J., ‘Melt compounding of polymeric nanocomposites’, International Polymer Processing, 2006 21 (1) 3–16. Marcinčin A., Jurenkova M., Kubičkova M., Hricova M., Borsig E., Brejka O., ‘Nanocomposite fibres for advanced textiles’, 2nd International Textile, Clothing and Design Conference, Magic World of Textiles, Dubrovnik 2004 Sfiligoj Smole M., Stakne K., Stana Kleinschek K., Gregor Svetec D., Ribitsch V., ‘Properties of nanocomposite PP fibres’ SPIE Opto-Ireland 2005 Nanotechnology and Nanophotonics, Dublin 2005, 306–313. Sfiligoj Smole M., Stakne K., Gregor Svetec, D., Stana Kleinschek K., Bele M., Ribitsch V., Musil V., ‘Nanofilled PP fibres’, Polymer Fibres 2004, Manchester 2004. Joshi M., Shaw M., Butola B.S., ‘Studies of composite filaments from nanoclay reinforced polypropylene’, Fibres and Polymers, 2004 5(1) 59–67. Zhou Y., Rangari V., Mahfuz H., Jeelani S., Mallick P.K., ‘Experimental study on thermal and mechanical behaviour of polypropylene, talc/polypropylene and polypropylene/clay nanocomposites’, Materials Science and Engineering, 2005 402 109–117. Zhang S., Horrocks A.R., ‘A review of flame retardant polypropylene fibres’, Progressin Polymer Science, 2003 28 1517–1538. Marosi G., Márton A., Szép A., Csontos I., Keszei S., Zimonyi E., Toth A., Almeras X., Le Bras M., ‘Fire retardancy effect of migration in polypropylene nanocomposites induced by modified interlayer’, Polymer Degradation and Stability, 2003 82 379– 385. Qin H., Zhang S., Zhao C., Feng M., Yang M., Shu Z., Yang S., ‘Thermal stability and flammability of polypropylene/montmorillonite composites’, Polymer Degradation and Stability, 2004 85 807–813. Montanari G.C., Fabiani D., Palmieri F., ‘Modification of electrical properties and performance of EVA and PP insulation through nanostructure by organophilic silicates’, IEEE Transactions on Dielectrics and Electrical Insulation 2004 11 (5) 754–762. Wu C.L., Zhang M.Q., Rong M.Z., Friedrich K., ‘Tensile performance improvement of low nanoparticles filled-polypropylene composites’, Composites Science and Technology, 2002 62 1327–1340. Rong M.Z., Zhang M.Q., Zheng Y.X., Zengh H.M., Walter R., Friedrich K., ‘Structure– property relationships of irradiation grafted nano-inorganic particle filled polypropylene composites’, Polymer, 2001 42 167–183. Rong N.Z., Zhang M.Q., Zheng Y.X., Zengh H.M., Friedrich K., ‘Improvement of tensile properties of nano-SiO2/PP composites in relation to percolation mechanism’, Polymer, 2001 42 3301–3304. Wu C.L., Zhang M.Q., Rong M.Z., Friedrich K., ‘Silica nanoparticles filled polypropylene: effect of particle surface treatment, matrix ductility and particle species on mechanical performance of the composites’, Composites Science and Technology, 2005 65 635–645. Stöber W., Fink A., ‘Controlled growth of monodisperse silica spheres in the micron size range’, Journal of Colloids and Interface Science, 1968 26 62–69. Bikiaris D.N., Vassiliou A., Pavlidou E., Karayannidis G.P., ‘Compatibilisation effect of PP-g-MA copolymer on iPP/SiO2 nanocomposites prepared by melt mixing’, European Polymer Journal, 2005 41 1965–1978.

© 2007, Woodhead Publishing Limited

Nanofilled polypropylene fibres

297

61. Zhang Q., Yang H., Fu Q., ‘Kinetics-controlled compatibilization of immiscible polypropylene/polystyrene blends using nano-SiO2 particles’, Polymer, 2004 45 1913– 1922. 62. Reddy C.S., Das C.K., Narkis M., ‘Propylene–ethylene copolymer nanocomposites: Epoxy resin grafted nanosilica as a reinforcing filler’, Polymer Composites 2005 26/6 806–812. 63. Yang K., Yang Q., Li G., Sun Y., Feng D., ‘Morphology and mechanical properties of polypropylene/calcium carbonate nanocomposites’, Materials Letters, 2006 60 (6) 805–809. 64. Weon J.I., Gam K.T., Boo W.J., Sue H.J., Chan C.M., ‘Impact-toughening mechanisms of calcium carbonate-reinforced polypropylene nanocomposite’, Journal of Applied Polymer Science, 2006 99/6 3070–3076. 65. Zuiderduin W.C.J., Westzaan C., Huétink J., Gaymans R.J., ‘Toughening of polypropylene with calcium carbonate particles’, Polymer 2003 44 261–275. 66. Nai-Xiu D., Mao-Sheng Z., ‘Magnetostrictive properties of carbon black filled polypropylene composites’, Polymer Testing, 2004 23 523–526. 67. Zhao H.X., Li R.K.Y., ‘Crystallization, mechanical and fracture behaviours of spherical alumina-filled polypropylene nanocomposites’, Journal of Polymer Science B – Polymer Physics, 2005 43 (24) 3652–3664. 68. Fina A., Tabuani D., Frache A., Camino G., ‘Polypropylene–polyhedral oligomeric silsesquioxanes (POSS) nanocomposites’, Polymer, 2005 46 7855–7866. 69. Fu B.X., Yang L., Somani R.H., Zong S.X., Hsiao B.S., Phillips S., Blanski R., Ruth P., ‘Crystallization studies of isotactic polypropylene containing nanostructured polyhedral oligomeric silsesquioxane molecules under quiescent and shear conditions’, Journal of Polymer Science Part B – Polymer Physics, 2001 39 (22) 2727–2739. 70. Fu B.X., Gelfer M.Y., Hsiao B.S., Phillips S., Viers B., Blanski R., Ruth P., ‘Physical gelation in ethylene-propylene copolymer melts induced by polyhedral oligomeric silsesquioxane (POSS) molecules’, Polymer, 2003 44 (5) 1499–1506. 71. Fina A., Abbenhuis H.C.L., Tabuani D., Frache A., Camino G., ‘Polypropylene metal functionalised POSS nanocomposites: a study by thermogravimetric analysis’, Polymer Degradation and Stability, 2006 91 (5) 1064–1070.

© 2007, Woodhead Publishing Limited

298

Nanofibers and nanotechnology in textiles

© 2007, Woodhead Publishing Limited

Part III Improving polymer functionality

299

© 2007, Woodhead Publishing Limited

11 Nanostructuring polymers with cyclodextrins A. E. T O N E L L I, North Carolina State University, USA

11.1

Introduction

For over a decade my colleagues and I, and several other research groups, have been forming crystalline inclusion compounds (ICs) between host cyclodextrins (CDs) and various guest polymers.1–73 When guest polymers are threaded by host cyclodextrins (CDs) to form crystalline polymer-CDICs, the included polymer chains are highly extended and isolated from neighboring chains. This is a consequence of the stacking of the cyclic oligosaccharides, α-, β- or γ-CD containing six, seven, or eight glucose units, respectively, which produces continuous narrow channels with ~ 0.5– 1.0 nm diameters, where the guest polymers are included and confined (see Fig. 11.1).6

0.95 nm

0.57 nm

β-CD

α-CD

0.78 nm

0.78 nm

γ -CD

(b)

(a)

(d) (c)

(e)

(f)

11.1 (a) γ-CD chemical structure; (b) approximate dimensions of α-, β-, and γ-CDs; schematic representation of packing structures of (c) cage-type, (d) layer-type and (e) head-to-tail channel-type CD crystals; and (f) CD-IC channels containing included polymer guests.

301

© 2007, Woodhead Publishing Limited

302

Nanofibers and nanotechnology in textiles

My laboratory alone has formed polymer-CD-ICs with more than three dozen high molecular weight polymer guests,1–52 covering a wide range of different chemical structures, including the fibroin protein from the Bombyx mori silkworm.47 As a consequence, we have been attempting to understand why randomly coiling polymer chains in solution or the melt become threaded or thread into the nano-pores of dissolved or solid CDs, where they are highly extended and segregated from other polymer chains. The nano-threading of guest polymers into CDs to form ICs has added importance, because they can serve as model systems to probe various aspects of molecular recognition and supramolecular chemistry, which are so critical to life processes. This chapter brings together and summarizes observations previously made in our laboratory that illuminate several important aspects of the nano-threading of polymers to form polymer-CD-ICs.74 These include: (i) competitive CD threading of polymers with different chemical structures and molecular weights from their solutions containing suspended solid or dissolved CDs, (ii) the threading and insertion of undiluted liquid polymers into solid CDs, and (iii) suspension of polymer A or B-CD-IC crystals in a solution of polymer B or A and observation of the transfer of polymer B or A from solution to displace polymer A or B and form polymer B or A-CD-ICs, without dissolution of the polymer-CD-ICs. Comparison of these observations has enabled an assessment of the relative importance of several factors that have been previously suggested as being crucial in the formation of CD-ICs with both polymer and smallmolecule guests and to the nano-threading of polymers in general. The formation of and coalescence from polymer-CD-ICs are briefly described along with the methods used to characterize both the ICs and the coalesced guest polymers. Here the focus is on the structural organization of solid polymer samples coalesced from their CD-ICs, which are found to be quite distinct from solid polymer samples formed from their randomly coiling and entangled solutions and melts. The alteration and control of properties displayed by solid polymers coalesced from their CD-ICs are also discussed. Polymers treated with and containing CDs, but that are only partially or not included at all by CDs, or that are covalently bound to CDs, are also considered, with particular emphasis placed on how their properties may be modified and controlled by their constituent CDs.

11.2

Formation and characterization of polymer– cyclodextrin–inclusion compounds

Formation of polymer-CD-ICs (see Fig. 11.1f and 11.2) begins with either the dissolution or suspension of CD cage-structure crystals (see Fig. 11.1c) or their conversion to channel- or columnar-structure crystals (see Fig. 11.1e) containing only water of hydration as a guest.24 The formation of crystalline polymer-CD-ICs from solutions requires the combination of solutions

© 2007, Woodhead Publishing Limited

Nanostructuring polymers with cyclodextrins

303

Polymer Time

Filter precipitate

RT ∆

∆ Inclusion complexation

Organic solvent CD

Coalescing process ∆

Removal of CD with water and amylase enzyme

Water

Coalesced polymer chains in extended conformations with less entangled state

CD-IC powder

11.2 Schematic representation of polymer-CD-IC formation, the coalescence process and the coalesced polymer.

containing the host CDs and the guest polymers, generally conducted with stirring at elevated temperatures,48 followed by quiescent cooling, precipitation and filtration. On the other hand, suspension of either as-received cagestructure or columnar-structure CD crystals in polymer solutions40 or in neat liquid polymers,44, 45 can also produce polymer-CD-IC crystals. By employing solutions containing two or more chemically distinct polymers during CD-IC formation, we may produce common CD-ICs that contain two or more guest polymers. Coalescence of the common guest polymers results in well-mixed polymer blends formed of two or more polymers that are generally observed to be immiscible.17, 33, 39, 46 X-ray diffraction and/or solidstate 13C-NMR (nuclear magnetic resonance) observations of the resulting products can confirm the formation of channel-structure CDs and the presence of both the host CD and the guest polymer in the case of the latter technique. In addition, Fourier transform infrared (FTIR) and differential scanning calorimetry (DSC) observations enable the observation of the guest polymers and their inclusion in the channels of the host CDs, respectively. Dissolution of polymer-CD-ICs and observation by 1 H-NMR can reveal their stoichiometries.

11.2.1 Coalescence of guest polymers from their cyclodextrin–inclusion compounds As noted in Fig. 11.2, guest polymers may be recovered from their CD-ICs in the form of bulk solid samples by treating the polymer-CD-IC crystals with warm water or another CD solvent, which is a non-solvent for the guest

© 2007, Woodhead Publishing Limited

304

Nanofibers and nanotechnology in textiles

polymer, and or treatment with an amylase enzyme.20 This process is termed coalescence and is usually conducted at temperatures below which the resultant consolidated guest polymers are immobile, i.e. below their glass-transition temperatures, Tg, to retain, as much as possible, their CD-processed and reorganized structures indicated in Fig. 11.2. A wide variety of analytical techniques, such as wide-angle X-ray scattering (WAXS), FTIR, DSC, thermogravimetric analysis (TGA), solid-state NMR and mass spectrometry, are used to characterize the structures of polymer samples coalesced from their CD-IC crystals. The results of these observations are always compared with those obtained for as-received or as-synthesized, or melt or solution processed samples of the same polymers. In this manner, we attempt to assess process-dependent differences in their organization, such as morphologies, crystallinities and even the conformations adopted by their constituent polymer chains.

11.3

Properties of polymer–cyclodextrin–inclusion compounds

Liu and Guo75 have summarized the interactions/driving forces that are often cited as playing significant roles in the formation of soluble small-molecule guest/host CD-ICs, and attempted to prioritize them in order of importance. Even though the formation is being discussed of solid crystalline columnarstructure CD-ICs containing polymer guests (see Fig. 11.1f and 11.2), where either both components are initially in solution or with solid CDs suspended in solutions of or in neat polymer guests, it is also useful to discuss and evaluate these same potential interactions/driving forces in connection with their formation. At the same time, because soluble and crystalline solid CDICs are most significantly distinguished by the regular packing of host CDs in the IC crystals, this distinction must also be considered in the present discussion.

11.3.1 Electrostatic interactions The dipolar interactions between polar host CDs and included guests with permanent dipole moments are considered to affect at least the conformations and structures of soluble small-molecule guest/host CD-ICs. 75 Our observations29 of the preference for inclusion of poly (ε-caprolactone) (PCL) over poly(L-lactic acid) (PLLA) chains in both dissolved and crystalline suspended α-CD (α-CDCS), as well as the displacement of PLLA chains by PCL from PLLA-α-CD-ICs when suspended in PCL solution, indicate that dipole–dipole electrostatic interactions may not be critical to the formation of polymer-CD-ICs. If they were, then we might expect PLLA to be preferentially included in α-CD compared to PCL, because two PLLA repeat units and two ester group dipoles occupy each α-CD, while only single PCL

© 2007, Woodhead Publishing Limited

Nanostructuring polymers with cyclodextrins

305

repeat units with a single identical ester group dipole are included in each αCD. Because both aliphatic polyesters likely adopt nearly fully extended alltrans conformations when included in their α-CD-ICs,76, 77 the dipole moments in neighboring repeat units point in approximately opposite directions. This might cause partial cancellation of the net PLLA dipole moment in each αCD, because two PLLA repeat units occupy each host α-CD, while only a single PCL ester group is included. Aside from this potential caveat, and because purely non-polar hydrocarbon polymers may be included in CDs,42 it is likely that dipolar electrostatic interactions do not play a major role in the nano-threading and subsequent formation of polymer-CD-ICs.

11.3.2 van der Waals interactions Liu and Guo75 concluded that van der Waals interactions are a major driving force for the formation of soluble small-molecule guest/host CD-ICs. Because van der Waals interactions depend on molecular polarizabilities, and the two PLLA repeat units included in each α-CD are more polarizable than a single included PCL repeat unit, we tentatively suggest that van der Waals interactions may not be important in the formation of polymer-CD-ICs, because inclusion of PCL was observed to be preferred over that of PLLA.29

11.3.3 Hydrophobic interactions Because PCL is less polar than PLLA, with an increased potential for hydrophobic interactions, we conclude that hydrophobic interactions may be important in the formation of polymer-CD-ICs, because, compared with PLLA, PCL is preferentially included by α-CD.29 In addition, the observed33, 46 competitive preference for the solution inclusion of bisphenol-A polycarbonate (PC) in γ-CD in the presence of poly(methyl methacrylate) and/or poly(vinyl acetate) further strengthens this conclusion, because PC is more hydrophobic than the other two polymers.

11.3.4 Hydrogen bonding Once again the preference of PCL over PLLA inclusion29 and the fact that all hydrocarbon polyolefins can form CD-ICs42 imply that hydrogen-bonding between included guest polymers and host CDs is not likely crucial in the formation of polymer-CD-ICs.

11.3.5 Relief of conformational strain in cyclodextrins Relief of conformational strain found in pure cage-structure CDs, which adopt asymmetric conformations,75 cannot occur during polymer-CD-IC

© 2007, Woodhead Publishing Limited

306

Nanofibers and nanotechnology in textiles

formation from solutions, where CDs are dissolved, may play a minor role in the case of the formation of polymer-CD-ICs through suspension of solid cage-structure CDs in polymer solutions40 or in neat polymers.45

11.3.6 Exclusion of cavity-bound, high-energy water This is not likely to be an important factor for polymer-CD-ICs formed in solution. When forming polymer-CD-ICs by suspension of CDs in neat liquid polymers,44, 45 or in their solutions,40 however, some of the water bound in CD cavities must be displaced by polymer chains as they thread and are included in the suspended host CDs. In the case of poly (N-acylethylenimine) (PNAI),40 inclusion from acetone solutions into both suspended cage and columnar structure γ-CDs, was prevented when using chloroform solutions. This strongly implies that the cavity water in CDs must have a suitable place to go when displaced by the inclusion of polymer guests, and likely is important in the formation of polymer-CD-ICs by suspension of CDs in their solutions and possibly also in neat polymer liquids. Drying cage structure α-CD before suspending into neat poly(ethylene glycol) (PEG) did not affect PEG inclusion,44, 45 but this may have been the result of the compatibility between water and PEG. To further assess whether or not exclusion of cavity-bound, high-energy water is an important factor when forming polymer-CD-ICs by suspension of CDs in neat liquid polymers or in their solutions, α-CDCS should be utilized. Vacuum drying of α-CDCS removes water of hydration residing in the α-CD channels, but does not result in changes in the columnar crystalline packing of α-CDs.24 As a consequence, if air-dried and vacuum-dried αCDCS are suspended in neat polymers or their solutions, we would expect the inclusion of polymers to be faster in the vacuum-dried α-CDCS if exclusion of cavity-bound, high-energy water is an important factor. Such experiments are currently in progress.

11.3.7 Crystalline packing of host cyclodextrins in solid cyclodextrin–inclusion compounds When the α-CD-IC with guest propionic acid (PA) is formed either from solution or by suspending as-precipitated, air-dried α-CDCS in neat PA, a cage-structure PA-α-CD-IC results. Vacuum-drying α-CDCS, which removes nearly two-thirds of the water contained by air-dried α-CDCS, apparently stabilizes the columnar packing structure sufficiently to force PA to be included, without structural reversion to the generally preferred cage-structure PA-αCD-IC.24 Thus, interactions (presumably hydrogen-bonding) between neighboring α-CDs are increased by removal of interstitial hydration water upon vacuum-drying.

© 2007, Woodhead Publishing Limited

Nanostructuring polymers with cyclodextrins

307

Columnar-structure CD-ICs are always formed with polymer guests, because of their long-chain nature, which requires their threading and inclusion by many CDs. This is consistent with the observation,44, 45 that neat PEG oligomers are included in both as-received and vacuum-dried cage-structure α-CDs at the same rate and to the same quantity even though vacuum-dried cagestructure α-CDs have lost ~one-third of their hydration water,78 presumably from their cavities, where they do not affect or stabilize the cage packing of α-CDs. However, when comparing the inclusion of both PEG oligomers,44, 45 neat and in solution, and PNAI chains40 in solution, into suspended cageand columnar-structure α-CDs, the inclusion into α-CDCS is observed to be more facile. In the former/latter inclusion the packing structure of α-CDs is altered/unaltered, thereby illustrating the importance of the crystalline packing of host CDs in the formation of polymer-CD-ICs with solid CDs.

11.3.8 Nano-threading of polymers into solid cyclodextrins Some all-hydrocarbon polyolefins in certain solutions have been observed to thread into γ-CDs and form polyolefin-γ-CD-ICs.42 In concert with the large number of CD-ICs that have been formed with guest polymers having a wide variety of chemical structures, this strongly suggests that the nano-threading of polymers is a general phenomenon characteristic of their long-chain natures, or polymer physics, and not their detailed chemical structures, or polymer chemistry.74 We have observed: (i) the competitive CD threading of polymers with different chemical structures and molecular weights from their solutions containing suspended solid or dissolved CDs, (ii) the threading and insertion of undiluted liquid polymers into solid CDs, and (iii) suspension of polymer A or B-CD-IC crystals in a solution of polymer B or A and observation of the transfer of polymer B or A from solution to displace polymer A or B and form polymer B or A-CD-ICs, without dissolution of the polymer-CD-ICs. From these observations we have been able to illuminate several important aspects of the nano-threading of polymers. In particular, the value in observations of the inclusion of guest polymers, as well as small-molecule guests, into solid CDs suspended in their solutions and in neat guest liquids was made apparent, because interactions between host CDs, between CDs and solvents, and between guests and solvents, which complicate and make understanding the formation of polymer-CD-ICs difficult, are either eliminated or can be independently controlled in these experiments.74 Extension of the investigations summarized here should eventually permit a more complete answer to the question: ‘Why do randomly coiling polymer chains in solution or the neat melt become threaded or thread into the nano-pores of dissolved or solid CDs

© 2007, Woodhead Publishing Limited

308

Nanofibers and nanotechnology in textiles

and zeolites, where they are highly extended and segregated from other polymer chains?’ However, we can currently conclude that electrostatic, van der Waals and hydrogen-bonding interactions and relief of conformational strain in CDs are not important, while hydrophobic interactions, exclusion of high energy, cavity-bound water and crystalline packing of host CDs are important in the formation of polymer-CD-ICs. Also the nano-threading of polymers appears to be a general phenomenon characteristic of their long-chain natures, or polymer physics, and not their detailed chemical structures, or polymer chemistry.74

11.4

Homo- and block copolymers coalesced from their cyclodextrin–inclusion compounds

When poly(ethylene terephthalate) (PET) was coalesced from its IC formed with γ-CD, it was observed to be significantly reorganized with respect to the as-received and solution and melt processed (normal) samples.21, 32 For example, coalesced PET was found to have an FTIR spectrum that was distinct from those of normal PET samples, with much improved resolution. DSC observations of coalesced PET repeatedly evidence high-temperature crystallization, with resulting high crystalline contents, upon cooling rapidly from the melt, which is uncharacteristic of PETs that are generally slow to crystallize and can easily be quenched into a totally amorphous material. Furthermore, repeated DSC heating scans of coalesced PET fail to reveal a macroscopic glass-transition, and this was confirmed on a microscopic scale by measurement of temperature-dependent solid-state 13C-NMR-observed 1 H spin-lattice relaxation times, T1ρ(1H).32 Each of the behaviors unique to coalesced PET can be attributed to the g ± tg ⫿ kink conformations assumed by its ethylene glycol fragments when included in the narrow channels of its γ-CD-IC, which are compared schematically in Fig. 11.3 (bottom) to the all-trans crystalline conformation (top) of PET. The kink conformations are nearly as extended as the crystalline, all-trans conformation, but they have a smaller cross-section, which explains their preferential inclusion in the narrow channels of its γ-CD-IC (see Fig. 11.1).79 When coalesced from its γ-CD-IC the PET chains can readily and rapidly convert to the all-trans conformation through simple counter rotations about the —CH2—O— bonds and therefore crystallize. Normal melts of PET consist of chains with predominantly g ± —CH2—CH2— bonds, which must be rotated to the trans conformation during crystallization. This conformational transition is difficult to accomplish without long-range movements of chain segments, consistent with the normally slow crystallization rate of PET. Furthermore, both solid-state FTIR and 13C-NMR observations

© 2007, Woodhead Publishing Limited

Nanostructuring polymers with cyclodextrins

309

O O O

O

O

O

O

O

O

O O

H

O H

O O O

O C

C H

O

H

O O O

H H

O O

C

C

O

H H

O

11.3 Schematic of the crystalline all-trans (top) and γ-CD-included g ± tg ⫿ (gauche ± trans-gauche kink ⫿) (bottom) conformations of PET.

make clear that the kink conformations are largely retained by PET chains in the noncrystalline regions of the coalesced sample, and are also apparently substantially retained for a considerable period of time in the melt. This explains the ability of coalesced PET to be repeatedly and rapidly crystallized from its melt to high levels of crystallinity without application of hightemperature (T > Tg) annealing or solvent-induced crystallization. The nanostructuring of PET by processing with γ-CD has important consequences for other macroscopic properties.21, 32, 80 For example, compared with PET annealed at high temperature to achieve the same level of crystallinity as coalesced PET, the coalesced sample has a higher density and exhibits a reduced permeability to CO2.80 This implies a tighter packing of chains in the noncrystalline regions of coalesced PET.

11.4.1 PCL-b-PLLA di-block copolymer When coalesced from its α-CD-IC, the structure and properties of the biodegradable-bioabsorbable di-block copolymer (PCL-b-PLLA) are found to be significantly different from the as-synthesized sample,18, 20 as can be seen in Table 11.1. Note, for example, that both the levels of block phase segregation and crystallinity are substantially reduced in the coalesced di-block. This results in a much more rapid biodegradation of the coalesced di-block,20 compared with the as-synthesized sample, which is critical to several of its potential applications. The phase segregation and crystallinity of block copolymers may not be reduced only via the formation of and coalescence from their CD-ICs, as mentioned above for PCL-b-PLLA, but may actually be controlled through the formation of ICs using CDs which incorporate only some or all of the constituent blocks. For example, the phase segregation and crystallinity of

© 2007, Woodhead Publishing Limited

310

Nanofibers and nanotechnology in textiles

Table 11.1 Thermal properties of as-synthesized and coalesced PCL-b-PLLA di-block copolymer (melting and crystallization temperatures Tm, Tcc, melting enthalpy ∆Hm, and % crystallinity) Identity

Tm PCL +Tcc PCL (°C) (°C)

∆Hm PCL Xc PCL (J/g) (%)

Tm PLLA (°C)

+ Tcc PLLA (°C)

∆Hm PLLA χc PLLA (J/g) (%)

Assynthesized di-block

56.1

17.6

66.9

48

160

77.9

61.4

66

Coalesced di-block

63.7

29.3

35.0

25

164

93.9

12.5

13

Table 11.2 Thermal properties (DSC) of various PCL-PPG-PCL tri-block copolymer samples25 Identity

Tm PCL (°C)

∆HPCL (J/g)

χc PCL (%)

As-synthesized copolymer

57.3

58.6

56.5

Sample coalesced from α-CD-copolymer IC

63.8

76.8

74.1

Sample coalesced from γ-CD-copolymer

63.0

51.3

49.5

PCL blocks, χc-PCL, in the tri-block copolymer PCL-poly(propylene glycol)PCL (PCL-PPG-PCL) can be increased and decreased, respectively, compared to the as-synthesized sample, by employing α- and γ-CDs.25 In the IC formed between PCL-PPG-PCL and α-CD only the PCL blocks are included in the α-CD channels, while the PPG blocks are excluded, which serves to segregate the PCL and PPG blocks. On the other hand, both PCL and PPG blocks are included in the IC channels formed with γ-CD, leading to an expected decrease in block segregation upon coalescence. As can be observed in Table 11.2, both expectations are met. Phase segregation, as indicated by PCL block crystallinity, is enhanced compared with the as-synthesized sample when PCL-PPG-PCL is processed with αCD, while processing with γ-CD enhances the mixing of crystallizable PCL and noncrystallizable PPG blocks.

11.5

Constrained polymerization in monomer– cyclodextrin–inclusion compounds

When the IC formed between γ-CD and styrene was suspended in water, which in this case did not result in dissolution of γ-CD, and a water-soluble free-radical initiator was added, the styrene was observed to polymerize to

© 2007, Woodhead Publishing Limited

Nanostructuring polymers with cyclodextrins

311

polystyrene (PS),37 without disruption of the channel structure of the host γCDs. Based on modeling the configurations and conformations of PS tetramers in cylindrical channels mimicking γ-CD,9, 81 it was concluded that isotacticPS (i-PS) could and syndiotactic-PS (s-PS) could not be accommodated in the γ-CD channels. Dissolution of the PSs obtained by the constrained polymerization of styrene in γ-CD channels and by the unconstrained homogeneous polymerization of styrene in an organic solvent and observation of their 13C-NMR spectra, did in fact reveal that the isotactic content of the PS obtained from styrene-γ-CD-IC was significantly greater than that of the homogeneously polymerized PS. Thus it appears that the microstructures of polymers may be controlled/altered by the spatially constrained polymerization of monomer-CD-ICs.

11.6

Coalescence of common polymer–cyclodextrin– inclusion compounds to achieve fine polymer blends

Though most polymers are found to be immiscible, it is possible to obtain well-mixed blends of any two or more polymers that can be dissolved in common solvents by processing with CDs.39 This is achieved by first making a common CD-IC containing the polymers to be blended and coalescing them subsequently into a solid blend sample. Blending polymers by processing their common CD-ICs will be illustrated with the immiscible pair of biodegradable/bioabsorbable polyesters PCL and PLLA.10 Figure 11.4 compares the X-ray diffractograms of pure PCL and PLLA and their blends obtained by casting from their dioxane solution and by the hot water coalescence from their common IC with α-CD. Note that diffraction peaks from both PCL and PLLA crystals are prominent in the solution-cast blend (Fig. 11.4c), indicating a highly phase-separated morphology. On the other hand, in Fig. 11.4d the diffractogram of the PCL/PLLA blend coalesced from their common α-CD-IC, no diffraction peaks are observed for PCl crystals and only very weak peaks are observable for PLLA crystals. In fact, from DSC observation of the coalesced PCL/PLLA blend, it is estimated that no PCL crystals are present, and that only 5% of the PLLA is crystalline. Clearly then, the coalesced PCL/PLLA blend appears to be nearly totally amorphous and presumably well-mixed. Two-dimensional solid-state spin-diffusion Hetcor NMR experiments revealed82 that the coalesced PCL/PLLA blend was intimately mixed. The average minimum dimension for the amorphous coalesced PCL/PLLA blend was 4.9 nm, versus 7.4 nm for the amorphous regions of the semicrystalline PCL/PLLA blend prepared by dissolution in dioxane. The former dimension is comparable to the 4–5 nm radii of gyration for the particular PLLA and PCL polymers used, indicative of molecular level mixing in the coalesced

© 2007, Woodhead Publishing Limited

312

Nanofibers and nanotechnology in textiles

Intensity

6000 (a)

4000 (b)

(c) 2000

(d) 0 5

10

15

20

25

30

35

40

2θ

11.4 X-ray diffractograms of pure PCL (a) and PLLA (b) and PCL/PLLA blends obtained by casting from dioxane solution (c) and hot water coalescence from PCL/PLLA-α-CD-IC (d).10

blend. Individual spin-diffusion coefficients, D, for PLLA and PCL chains in the blends were calculated based on direct experimental measurement of the polymers in the blend. The magnitude of D was found to decrease by a factor of two for PLLA chains in the coalesced blend compared to pure PLLA, corroborating further the molecular mixing of PCL and PLLA chains in their coalesced blend. It is important to stress that blending polymers by coalescing their common CD-ICs is the only currently known method to achieve intimate solid-state mixing of normally incompatible polymers.

11.7

Temporal and thermal stabilities of polymers nanostructured with cyclodextrins

In general we have observed that homopolymer, block copolymers and polymer blends nanostructured by processing with CDs retain their unique solid-state organizations for considerable periods of time even at temperatures exceeding their Tg and Tm values, where their chains are potentially mobile. For example, PET coalesced from its IC with γ-CD continues to be rapidly crystallizable upon cooling from its melt, even after spending 2 h at ~300 °C.21, 32 The PCL-b-PLLA di-block copolymer coalesced from its IC with α-CD remains more homogeneous, amorphous, and less phase segregated than an as-

© 2007, Woodhead Publishing Limited

Nanostructuring polymers with cyclodextrins

313

synthesized sample even after it is heated above the Tm of PLLA and then cooled from the melt.18 In fact the intimately mixed PCL/PLLA blend made by coalescence of PCL and PLLA from their common IC with α-CD remains well mixed and does not phase separate even after being heated for several hours at 200 °C, well above the melting temperatures of both polyesters.10 The thermal and temporal stabilities of polymers nanostructured with CDs suggests that they may be melt-processed under normal conditions and still substantially retain their unique solid-state organizations, which should result in fibers, films and molded articles with distinct and possibly improved properties

11.8

Cyclodextrin-modified polymers

The behaviors and properties of polymers may also be modified with CDs and additive-CD-ICs. Because CD-ICs formed with small-molecule additives, such as antibacterials and flame retardants, are stable to temperatures beyond 250 °C, they may be compounded intact into many molten polymers. In this manner antibacterial83 and flame-retardant84 polymer fibers and films have been achieved. Uncomplexed CDs may also be used to modify polymers. For example, when aqueous solutions of poly(vinyl alcohol) (PVA) are subjected to freezing/thawing cycles they produce hydrogels. When small quantities of γ-CD are added to the aqueous PVA before freezing and thawing, the resultant hydrogels are softer (lower modulus) and are more swellable then when prepared without γ-CD.43 This behavior has been attributed to the partial threading of γ-CD by PVA chains, thereby reducing the number of microcrystalline cross-links subsequently formed between PVA chains during cyclic freezing/thawing. The solution viscosity of the hydrophobically modified, alkali-soluble emulsion (HASE) associative polymer described in Fig. 11.5 can be controlled by the addition of α- and β-CDs.30 CDs with their hydrophobic inner cores interact with the pendant macromonomeric segments of the associative HASE copolymer containing hydrophobic end groups and reduce their interchain hydrophobic interactions, leading to a reduction in solution viscosity and dynamic moduli by several orders of magnitude. The effect of CDs on the solution rheology of HASE copolymers is clearly demonstrated in Fig. 11.6, where the steady shear viscosities of aqueous HASE copolymer solutions are observed to be reduced ~a thousand-fold upon the introduction of α- and β-CDs. Furthermore, the reduction in polymer viscoelasticity in the presence of CD is reversibly recovered upon subsequent addition of different nonionic surfactants that have a higher propensity to complex with the CDs than the hydrophobic macromonomer segments of the HASE copolymer.

© 2007, Woodhead Publishing Limited

314

Nanofibers and nanotechnology in textiles Polyelectrolyte backbone

Hydrophobic groups

CH3 CH2

C

O

C OH

PEO spacers

CH3

H H2C

C C

x

Methacrylic acid

CH2

C

H3C

C

O

O CH2 CH3 Ethyl acrylate

CH3

NH y

C

O

O CH2 H2C O p

R Macromonomer

z

11.5 Schematic representation of an HASE associative polymer and the molecular constitution of the HASE polymer used in this study.30 R refers to the C22H45 hydrophobe, p = 40, and x/y/z = 43.6/56.2/0.20 by mole.

11.9

Polymers with covalently bonded cyclodextrins

Recently several reports of covalently bonding CDs to pre-synthesized polymers85 and the synthesis of polymers incorporating CDs86 have been published. The reactive Cl atom on monochlorotriazinyl-β-CD (CDMCT) can react with nucleophilic groups, such as the –OH groups on cellulose to permit grafting of CDMCT on cotton.85 When cotton fabric grafted with CDMCT was immersed in solutions containing insecticides or insect repellents,

© 2007, Woodhead Publishing Limited

Nanostructuring polymers with cyclodextrins 103

315

0 3 5

η (Pa ·s)

102 7 10 101 12 100 15 20

25

10–1

(a)

103 0

5

102 η (Pa· s)

3

10 101

15 25 20

100

10–1

10–2

10–1

γ˙ (s –1)

100 (b)

101

102

11.6 Effects of addition of (a) α-CD and (b) β-CD on the steady shear viscosity of HASE associative polymer solutions.30 Numbers correspond to the moles of cyclodextrin per moles of hydrophobe.

these active additives were complexed with the grafted CDMCTs, thus rendering the cotton fabric insecticidal or insect repellant. Nylon-6,10 containing covalently bonded methyl-β-CDs (MBCD) was produced via interfacial polymerization of hexamethylene diamine in water and sebacoyl chloride plus MBCD pre-reacted with sebacoyl chloride in xylenes.86 The MBCD–nylon-6,10 fibers drawn from the bi-layer interface were generally cross-linked due to the large number of –OH groups on MBCD, though cross-linking could be controlled by adjusting the ratio of MBCD pre-reacted with sebacoyl chloride to unreacted sebacoyl chloride. When placed in a dye bath containing Acid Blue 29, the MBCD–nylon6,10 fibers were observed to dye faster and absorb more dye than pure nylon-6,10 fibers. Apparently a portion of Acid Blue 29 dye enters the MBCD cavity and forms an Acid Blue 29-MBCD inclusion complex.

© 2007, Woodhead Publishing Limited

316

Nanofibers and nanotechnology in textiles

11.10 Conclusions By way of the many examples discussed above, it should be clear that solid polymers may be effectively nanostructured by processing with CDs. Polymer solutions and gels may also be modified/controlled with CDs, and various functionalities may be delivered to them in the form of additive-CD-ICs. Finally, CDs may be covalently incorporated into polymers to provide them with permanent capabilities to bind and/or release a host of active agents.

11.11 References 1. Huang, L., Vasanthan, N., Tonelli, A. E., J. Appl. Polym. Sci., 64, 281, 1997. 2. Tonelli, A. E., Polymer Int., 43, 295, 1997. 3. Huang, L., Allen, E. J., Tonelli, A. E., Recent Research Developments in Macromolecular Research, S. G. Pandalai, Ed., Research Signpost, Trivandrum, India, 1997, Vol. 2, p. 175. 4. Huang, L., Allen, E., Tonelli, A. E., Polymer, 39, 4857, 1998. 5. Huang, L., Tonelli, A. E., ACS Symposium Series # 728, Intelligent Materials for Controlled Release Technologies, S. M. Dinh, J. D. DeNuzzio and A. R. Comfort, Eds., ACS, Washington, DC, 1999, Chapter 10. 6. Huang, L., Tonelli, A. E., J. Macromol. Science, Revs. Macromol. Chem. and Phys.,C 38(4), 781, 1998. 7. Huang, L., Allen, E., Tonelli, A. E., Polymer, 40, 3211, 1999. 8. Lu, J., Shin, I. D., Nojima, S., Tonelli, A. E., Polymer, 41, 5871, 2000. 9. Rusa, C. C., Tonelli, A. E., Macromolecules, 33, 1813, 2000. 10. Rusa, C. C., Tonelli, A. E., Macromolecules, 33, 5321, 2000. 11. Huang, L., Gerber, M., Taylor, H., Lu, J., Tapaszi, E., Wutkowski, M., Hill, M., Nunalee, F. N., Harvey, A., Rusa, C. C., Porbeni, F., Edeki, E., Tonelli, A. E., ACS Symposium Series # 790, Film Formation in Coatings: Mechanisms, Properties, and Morphology, T. Provder and M. Urban, Eds., Am. Chem. Soc., Washington, DC, 2001, Chapter. 14. 12. Lu, J., Mirau, P. A., Rusa, C. C., Tonelli, A. E., Cyclodextrin: From Basic Research to Market, Proceedings of the 10th International Cyclodextrin Symposium (CD2000), May 21–24, 2000, Ann Arbor, MI., J. Szejtli, Ed., Mira Digital Publ. (314776-6666), 2001. 13. Rusa, C. C., Lu., J., Huang, L., Tonelli, A. E., Cyclodextrin: From Basic Research to Market, Proceedings of the 10th International Cyclodextrin Symposium (CD2000), May 21–24, 2000, Ann Arbor, MI., J. Szejtli, Ed., Mira Digital Publ. (314776-6666), 2001. 14. Porbeni, F. E., Edeki, E. M., Shin, I. D., Tonelli, A. E., Polymer, 42(16), 6907, 2001. 15. Rusa, C. C., Luca, C., Tonelli, A. E., Macromolecules, 34, 1318, 2001. 16. Lu, J., Mirau, P. A., Shin, I. D., Nojima, S., Tonelli, A. E., Macromol. Chem. Phys., 203, 71, 2001. 17. Wei, M., Tonelli, A. E., Macromolecules, 34, 4061, 2001. 18. Shuai, X. Porbeni, F. E. Wei, M. Shin, I. D. Tonelli, A. E. Macromolecules, 34, 7355, 2001. 19. Huang, L., Gerber, M., Taylor, H., Lu, J., Tapazsi, E., Wutkowski, M., Hill, M.,

© 2007, Woodhead Publishing Limited

Nanostructuring polymers with cyclodextrins

20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38.

39. 40. 41. 42. 43. 44. 45.

317

Harvey, A., Rusa, C. C., Wei, M., Porbeni, F. E., Lewis, C. S., Tonelli, A. E., Macromol.Chem., Macromol. Sympos., 176, 129, 2001. Shuai, X., Wei, M., Porbeni, F. E., Bullions, T. A., Tonelli, A. E., Biomacromolecules, 3, 201, 2002. Bullions, T. A., Wei, M., Porbeni, F. E., Gerber, M. J., Peet, J., Balik, M., White, J. L., Tonelli, A. E., J. Polym. Sci., Polym. Phys. Ed., 40, 992, 2002. Shuai, X., Porbeni, F. E., Wei, M., Bullions, T. A., Tonelli, A. E., Macromolecules, 35, 3126, 2002. Wei, M., Davis, W., Urban, B., Song, Y., Porbeni, F. E., Wang, X., White, J. L., Balik, C. M., Rusa, C. C., Fox, J., Tonelli, A. E., Macromolecules, 35, 8039, 2002. Rusa, C. C., Bullions, T A., Fox, J., Porbeni, F. E., Wang, X. Tonelli, A. E. Langmuir, 18, 10016, 2002. Shuai, X., Porbeni, F. E., Wei, M., Bullions, T. A., Tonelli, A. E., Macromolecules, 35, 2401, 2002. Shuai, X., Porbeni, F. E., Wei, M., Bullions, T. A., Tonelli, A. E., Macromolecules, 35, 3778, 2002. Wei, M., Shuai, X., Tonelli, A. E., Biomacromolecules, 4, 783, 2003. Bullions, T. A., Edeki, E. M., Porbeni, F. E., Wei, M., Shuai, X., Tonelli, A. E., J. Polym. Sci., Polym. Phys. Ed., 41, 139, 2003. Rusa, C. C., Fox, J., Tonelli, A. E. Macromolecules, 36, 2742, 2003. Abdala, A. A., Khan, S., Tonelli, A. E., Macromolecules, 36, 7833, 2003. Tonelli, A. E., Macromol.Chem., Macromol. Sympos., 203, 71, 2003. Wei, M., Bullions, T. A., Rusa, C. C., Wang, X., Tonelli, A. E., J. Polym. Sci. Part B, Polym. Phys. Ed., 42, 386, 2003. Rusa, C. C., Uyar, T., Rusa, M., Hunt, M. A., Wang, X., Tonelli, A. E., J. Polym. Sci. Part B, Polym. Phys. Ed., 42, 4182, 2004. Wei, M., Shin, I. D., Urban, B., Tonelli, A. E., J. Polym. Sci. Part B, Polym. Phys. Ed., 42, 1369, 2004. Rusa, C. C., Shuai, X., Bullions, T. A., Wei, M., Porbeni, F. E., Lu, J., Huang, L., Fox, J., Tonelli, A. E., J. Polym. Environ., 12(3), 157, 2004. Pang, K., Schmidt, B., Kotek, R., Tonelli, A. E., J. Appl. Polym., Sci., 102(6), 6049– 6053, 2005. Uyar, T., Rusa, M., Tonelli, A. E., Macromol. Chem. Rapid Commun., 25, 1382, 2004. Rusa, C. C., Wei, M., Bullions, T. A., Rusa, M., Gomez, M. A., Porbeni, F. E., Wang, X., Shin, I. D., Balik, C. M., White, J. L., Tonelli, A. E., Crystal Growth Design, 4, 1431, 2004. Rusa, C. C., Wei, M., Shuai, X., Bullions, T. A., Wang, X., Rusa, M., Uyar, T., Tonelli, A. E., J. Polym. Sci. Part B, Polym. Phys. Ed., 42, 4207, 2004. Rusa, M., Wang, X., Tonelli, A. E., Macromolecules, 37, 6898, 2004. Rusa, C. C., Wei, M., Bullions, T. A., Shuai, X., Uyar, T. A., Tonelli, A. E., Polym. Adv. Technol., 16, 269, 2005. Rusa, C. C., Rusa, M., Gomez, M. A., Shin, I. D., Fox, J. D., Tonelli, A. E., Macromolecules, 37, 7992, 2004. Hernandez, R., Rusa, M., Rusa, C. C., Lopez, D., Mijanos, C., Tonelli, A. E., Macromolecules, 37, 9620, 2004. Peet, J., Rusa, C. C., Hunt, M., Tonelli, A. E., Balik, C. M., Macromolecules, 38, 537, 2005. Harada, A., Okada, M., Kawaguchi, Y., Chem. Letts., 34, 542, 2005.

© 2007, Woodhead Publishing Limited

318

Nanofibers and nanotechnology in textiles

46. Uyar, T., Rusa, C. C., Wang, X., Rusa, M., Hacaloglu, J., Tonelli, A. E., J. Polym. Sci. Part B, Polym. Phys. Ed., 43, 2578, 2005. 47. Rusa, C. C., Bridges, C., Ha, S.-W., Tonelli, A. E., Macromolecues, 38, 5640, 2005. 48. Uyar, T., Rusa, C. C., Hunt, M. A., Aslan, E., Hacaloglu, J., Tonelli, A. E., Polymer, 46, 4762, 2005. 49. Porbeni, F. E., Shin, I. D., Shuai, X., Wang, X., White, J. L., Jia, X., Tonelli, A. E., J. Polym. Sci. Part B, Polym. Phys. Ed., 43, 2086, 2005. 50. Uyar, T., Aslan, E., Tonelli, A. E., Hacaloglu, J., Polym. Degrad. Stabil., 9(1), 1–11, 2006. 51. Uyar, T., El-Shafei, A., Hacaloglu, J., Tonelli, A. E., J. Inclus. Phenom. Macrocyclic Chem., 55, 109, 2006. 52. Uyar, T., Hunt, M. A., Gracz, H. S., Tonelli, A. E., Cryst. Growth Design, 6, 113, 2006. 53. Harada, A., Kamachi, M., Macromolecules, 23, 2821, 1990. 54. Harada, A., Li, J., Kamachi, M., Nature, 356, 325, 1992. 55. Harada, A., Li, J., Kamachi, M., Nature, 364, 516, 1993. 56. Harada, A., Li, J., Kamachi, M., J. Am. Chem. Soc., 116, 3192, 1994. 57. Harada, A., Li, J., Kamachi, M., Macromolecules, 28, 8406, 1995. 58. Harada, A., Suzuki, S., Okada, M., Kamachi, M., Macromolecules, 29, 5611, 1996. 59. Harada, A. Adv. Polym. Sci., 133, 141, 1997. 60. Harada, A., Kawaguchi, Y., Nishiyama, T., Kamachi, M., Macromol. Rapid Commun., 30, 7115, 1997. 61. Harada, A., Nishiyama, T., Kawaguchi, Y., Okada, M., Kamachi, M., Macromolecules, 30, 7115, 1997. 62. Kawaguchi, Y., Nishiyama, T., Okada, M., Kamachi, M., Harada, A., Macromolecules, 33, 4472, 2000. 63. Harada, A., Acc. Chem. Res., 34, 456, 2001. 64. Huh, K. M., Ooya, T., Sasaki, S., Yui, N., Macromolecules, 34, 2402, 2001. 65. Li, J., Ni, X., Leong, K., Angew. Chem. Int. Ed., 42, 69, 2003. 66. Choi, H. S., Ooya, T., Sasaki, S., Kwon, I. C., Jeong, S. Y., Yui, N., Macromolecules, 36, 9313, 2003. 67. Li, J., Ni, P., Zhou, Z., Leong, K. W., J. Am. Chem. Soc., 125, 1788, 2003. 68. Choi, H. S., Takahashi, A., Ooya, T., Yui, N., Macromolecules, 37, 10036, 2004. 69. Shin, K., Dong, T., He, Y., Taguchi, Y., Oishi, A., Nishida, H., Inoue, Y., Macromol. Biosci., 4, 1075, 2004. 70. Dong, T., He, Y., Shim, K., Inoue, Y., Macromol. Biosci., 4, 1084, 2004. 71. Dong, T., He, Y., Zhu, B., Shin, K., Inoue, Y., Macromolecules, 38, 7736, 2005. 72. Nepal, D., Samal, S., Geckeler, K. E., Macromolecules, 36, 3800, 2003. 73. Girardeau, T. E., Zhao, T., Leisen, J., Beckham, H. W., Bucknall, D. G., Macromolecules, 38, 2261, 2005. 74. Rusa, C. C., Rusa, M., Peet, J., Uyar, T., Fox, J., Hunt, M. A., Wang, X., Balik, C. M., Tonelli, A. E., J. Inclus. Phenom. Macrocyclic Chem., 55, 185, 2005. 75. Liu, L., Guo, Q-X., J. Inclus. Phenom. Macrocyclic Chem., 42, 1, 2002. 76. Tonelli, A. E., Macromolecules, 24, 1275, 1991. 77. Tonelli, A. E., Macromolecules, 25, 3581, 1992. 78. Hunt, M. A., Rusa, C. C., Tonelli, A. E., Balik, C. M., Carbohydr. Res., 339, 2805, 2004. 79. Tonelli, A. E., Computat. Theor. Polym. Sci., 2, 80, 1992. 80. Vedula, J., Tonelli, A. E., J. Polym. Sci. Part B, Polym. Phys. Ed., 45, 735, 2007.

© 2007, Woodhead Publishing Limited

Nanostructuring polymers with cyclodextrins

319

81. Hunt, M. A., Uyar, T., Shamsheer, R., Tonelli, A. E., Polymer, 45, 1345, 2004. 82. Jia, X., Wang, X., Tonelli, A. E., White, J. L., Macromolecules, 38, 2775, 2005. 83. Lu, J., Hill, M., Hood, M., Greeson, Jr., D. F., Horton, J. R., Orndorff, P. E., Herndon, S. A., Tonelli, A. E., J. Appl. Polym. Sci., 82, 300, 2001. 84. Huang, L., Gerber, M., Lu, J., Tonelli, A. E., Polym. Degrad. Stabil., 7(2), 279, 2001. 85. Romi, R., Lo Nostro, P., Bocci, E., Ridi, F., Baglioni, P., Biotech. Prog., 21(6), 2005. 86. Busche, B. J., Rusa, M., Tonelli, A. E., Balik, C. M., Macromolecules, submitted.

© 2007, Woodhead Publishing Limited

13 Polyolefin/clay nanocomposites R. A. K A L G A O N K A R and J. P. J O G, National Chemical Laboratory, India

13.1

Introduction

Composites are widely used in such diverse areas as transportation, construction, electronics and consumer products. They offer unusual combinations of stiffness, strength and weight that are difficult to attain separately from the individual components. The advent of nanoscience and nanotechnologies has continuously provided the impetus pushing for the development of materials with fascinating properties and a rich variety of applications. Polymer–clay (PC) nanocomposites represent a new class of materials based on reinforcement of polymeric materials by dispersion of nano-scale clay particles at molecular level in the polymer matrix. Because of their nanometer size features, nanocomposites possess unique properties typically not shared by their more conventional microcomposite counterparts and therefore offer new technology and business opportunities. PC nanocomposites exhibit exceptional improvement in mechanical properties including stiffness, strength, dimensional stability as well as exhibiting barrier properties far better than conventionally filled polymers. In addition, because of the length scale involved that minimizes scattering, nanocomposites are usually transparent. Furthermore, PC nanocomposites exhibit a significant increase in thermal stability as well as self-extinguishing characteristics and enhanced flame retardancy. Since PC nanocomposites achieve composite properties at much lower volume fraction of reinforcement, they avoid many of the costly and cumbersome fabrication techniques common to conventional fiber- or mineralreinforced polymers. Instead they can be processed by techniques such as extrusion, injection molding and casting normally reserved for unfilled polymers. Furthermore, they are adaptable to films and fibers as well as monoliths. The combination of enhanced properties and weight reduction has already led to a few commercial applications. Toyota Motor Company has successfully introduced an automotive timing-belt cover made from nylonlayered silicate nanocomposite.1 Ube Industries in Japan in collaboration 351

© 2007, Woodhead Publishing Limited

352

Nanofibers and nanotechnology in textiles

with Toyota is also developing nanocomposite barrier films for food packaging and other applications. Similar R&D efforts focusing on silicate nanocomposites are already underway in several US companies. Potential applications include airplane interiors, fuel tanks and components in electrical or electronic parts, under-the-hood structural parts, brakes and tires. Polyolefins (POs) represent one of most widely used polymeric materials. However, fabrication of PO-based clay nanocomposites still remains a serious challenge for researchers. This is mainly due to the nonpolar nature of POs and the highly hydrophilic nature of clays. In the past efforts have been made to fabricate PO/clay nanocomposites using a variety of strategies. In this chapter we present a systematic review of the ongoing and past research concerning a variety PO/clay nanocomposites. The developments in this area with respect to the type of PO and clay, preparation technique of the nanocomposite, compatibilization and organic modification strategies used, and their effect on the various properties of these nanocomposites are reviewed critically.

13.1.1 Layered silicate clay minerals Natural clays are most commonly formed either by the in situ alteration of volcanic ash or by the hydrothermal alteration of volcanic rocks. Bentonite is a rock that consists of commonly known smectite or montmorillonite group clays as its major component and auxiliary minerals such as kaolin, quartz, gypsum and iron ore. Sodium or calcium bentonite is not itself a mineral name: more correctly, it is smectite clay composed primarily of the mineral montmorillonite having sodium or calcium ions as its predominant cations. Smectite or montmorillonite group clays are further classified into two subgroups, trioctahedral and dioctahedral smectite clays. Montmorillonite, beidilite and nontronite are trioctahedral smectite clays whereas hectorite and saponite are dioctahedral smectite clays. Montmorillonite is a threelayer mineral, formed by stacking of several layers of tetrahedron and octahedron sheets, electrostatically held together by isomorphic interlayer cations. As the electrostatic attraction is low, exposure to polar fluids such as water will cause the formation of a monomolecular layer of water between the silicate layers. The basis behind bentonite swelling is that the several layers of water dipoles can form into weak ‘stacked’ tetrahedral structures, causing the silicate layers to separate – this is termed intercrystalline swelling. Purity and quality of bentonite will vary, as per the depositional environment and subsequent weathering processes and also differs by region and deposit. The lattice structure of trioctahedral smectite clays comprises an octahedral alumina sheet, sandwiched between two tetrahedral silica sheets. The structural details of trioctahedral smectite clay are given in Fig. 13.1. A single crystal lattice of montmorillonite is negatively charged owing to the isomorphism

© 2007, Woodhead Publishing Limited

Polyolefin/clay nanocomposites

353

Exchangeable cations, n H2O

13.1 Structure of smectite clay.

substitution mainly in the octahedral layer (e.g. Mg2+ for Al3+). The negative charge of the lattice is balanced by the exchangeable cations, which are held in the interlayer space of the clay. The exchangeable cations can be sodium and/or calcium ions, which can be replaced by suitable organic or inorganic cations. The idealized unit cell formula of the sodium form of montmorillonite is Na0.67 [Al3.33Mg0.67] [Si8] O20 (OH)4 in which one of every six octahedral Al3+ has been replaced by Mg2+. The charge deficit is 0.67 electrostatic units (esu) per unit cell, or 0.67 equivalent charge per 734 g (formula weight) of clay. This gives a value of 91.3 milliequivalents (meq) charge deficit in a layer per 100 g of clay. This deficiency of charge must be balanced by equal quantity of cation charge at its surface for electrical neutrality. The maximum amount of any one cation that can be taken up by a particular clay is constant and is known as the cation exchange capacity (CEC) or base exchange capacity (BEC) of that clay. The amount is expressed in meq per 100 g of dry clay. CEC varies from 80 to 100 depending up on the substituted cation. Owing to the isomorphous substitution for M3+ by M2+ and M2+ by M+ in its structure, bentonite clay has a unique nature of cation exchange and adsorption capacity. The adsorption efficiency of bentonite is greatly enhanced after purification and/or modification. These properties have been used for a long time to decolorize edible oils, clarify alcoholic beverages and remove grease from raw wool.

© 2007, Woodhead Publishing Limited

354

13.2

Nanofibers and nanotechnology in textiles

Organomodification of clays

The smectite clay possesses a large surface area (~750 m2/g). In order to utilize their high surface area efficiently, the chemical modification of such clays is necessary. During modification the exchangeable interlamellar cations are replaced by the other cations, inorganic or organic, present in the external environment. This can be achieved by several methods and the modified products obtained can be classified into various categories including acidtreated clays, pillared clays and nanoclays. Of particular interest in polymer/ clay nanocomposites are nanoclays. Nanoclays are complexes derived from clay–organic reactions. The clays used are smectite group clays, generally montmorillonite or hectorite. These clays are hydrophilic and accessible to intercalation. Therefore, they are liable to intercalate organics through ion exchange and/or adsorption when reacted with organic compounds and to be transformed to hydrophobic or organophilic nature. The organophilicity is ascribed to the interlamellar surface coverage with organics. Owing to the organophilic nature, the modified clays are named organophilic clays or organoclays. Despite ignorance of the scientific basis of such interaction process, the uptake of certain organic compounds by clays has been known for a very long time, being the basis of a wide use of clays. Such clays are generally named ‘bleaching’ or ‘Fuller’s earth’. Complex-forming compounds may be classified into the following groups: • uncharged polar organic compounds (neutral organic compounds); • negatively charged organic compounds (acidic organic compounds); and • positively charged organic compounds (basic organic compounds). The nanoclays of particular interest with respect to usage of clay as a nanofiller in polymer composites are prepared by modification of clay using positively charged organic compounds. This class includes the organic compounds such as quaternary ammonium compounds (quaternaries), amines, alkaloids, purines, nucleosides, proteins, etc. The properties and end use of clays modified by such organic bases are largely dependent on the organic compound adsorbed by the clays.

13.2.1 Structure and properties of organomodified clays The organomodified clays are prepared by reacting more voluminous organic onium cations with the montmorillonite clay. The reaction results in the exchange of relatively small sodium ions with organic counter-ions. This ion exchange results in the increase of the clay interlayer space, enabling organic cation chains to move in between the layers. In addition, the surface properties of each clay sheet are changed from being hydrophilic to hydrophobic.

© 2007, Woodhead Publishing Limited

Polyolefin/clay nanocomposites

355

During the cation exchange reaction the organic cations could either replace the inorganic cations or neutralize the hydrogen ions at clay surfaces.2–4 Hendricks5 concluded from his work that the adsorption of organic cations by clays is influenced both by electrostatic (columbic) and van der Waals attractive forces. Grim et al.6 found that in the case of organic cations of relatively small size the maximum amount adsorbed did not go beyond the CEC of the clays, even when a large excess of the organic ion was present in the solution. Under similar conditions larger organic ions were taken up in the excess of the CEC because of the influence of van der Waals forces on the adsorption of large organic cations by clays. Similar findings were reported by Cowan and White7 using isotherms for the adsorption of a homologous series of primary n-alkyl ammonium ions from ethyl (C 2 ) to tetradecylammonium (C 14) by sodium montmorillonite. Vansant and Uytterhoeven8 concluded that apart from van der Waals forces of organic cations, the combination of variation in hydration status of the cations and electrostatic interactions between cation and clay surface played an important role in adsorption of n-alkyl ammonium ions on Na-montmorillonite. On interaction with clay particles, the cationic portion of positively charged organic salts are accommodated in the interlayer space of clay crystals by replacing inorganic cations initially present. The interaction between the clay platelets and organic cations results in the replacement of Na+ cations by organic cation moieties, which leads to an increase in the basal spacing. Barrer and Brummer found this increment to be directly related to the amount of organic cations present after the ion exchange.9 Uncharged organics having strongly polar groups adopted an alpha-II orientation.10 Jordan11 observed a similar orientation in montmorillonite complexes with primary n-alkyl ammonium ions of varying chain length (from C3 to C18). He further concluded that the basal spacing values were sensibly constant at 1.36 nm for C3 to C10 and 1.76 nm for C10 onwards. Further, when the cation area was smaller than half the area per exchange site, the amine cation adsorbed on one surface was found to be fitted into the gaps between the cations lying on the opposing surface to form monolayer complexes. When the area of the cation was greater than half the area per exchange position, interpenetration of this kind could not be realized and double layer complexes formed (e.g. for n-hexadecyl ammonium). Jordan et al.12, using X-ray diffraction (XRD), observed the following results for a long chain cation such as n-octadecyl ammonium: (a) the formation of double layer complexes when the amounts of cations adsorbed were equivalent to the exchange capacity, (b) the alkyl chain in the doublelayer complex stands at an angle to the silicate layer at amounts adsorbed higher than the CEC, and (c) at still higher amounts adsorbed, the chains were found reoriented into near-vertical positions with respect to the surface giving rise to a close packed arrangement which allows extensive van der Waals interactions between adjacent alkyl chains to be established. Greene-

© 2007, Woodhead Publishing Limited

356

Nanofibers and nanotechnology in textiles

Kelly13 reported similarly for pyridinium and its derivatives in montmorillonite. Thus, besides being dependent on the alkyl chain length of the cation and the charge on the mineral layer, the arrangement of intercalated n-alkyl ammonium/ pyridinium ions and/or molecules is also influenced by the amount of the cations and/or molecules adsorbed which ultimately depends on the amount of the compound reacted. The orientation of alkyl ammonium ions between the clay aggregates has been investigated by Lagaly.14 He has reported that besides the formation of paraffin-type structures of chains in all-trans conformation, aggregates of chains containing gauche conformations are commonly observed. More recently, molecular dynamics simulations have provided insights into the packing orientation of alkyl ammonium surfactant chains in the clay interlayers.15 It was observed that a disordered liquid-like arrangement of the alkyl chains was preferred in the clay gallery. In this disordered state the alkyl chains do not remain flat, but instead overlap and co-mingle with the adjacent onium ions in the opposing layers within the galleries.

13.3

Polymer/clay nanocomposites

PC nanocomposites are an emerging class of organic–inorganic hybrids that contain a relatively low wt% of nanometer-sized clay. These were first developed in the late 1980s. The dispersion of the nanometer-sized clay in the polymer matrix significantly improves the mechanical, thermal, barrier properties and flame retardancy of the base polymer. Three main types of nanocomposites can be obtained when clay is dispersed in a polymer matrix. This depends on the nature of the components used, including polymer matrix, clay and organic cation. If the polymer cannot intercalate between the silicate sheets, a microcomposite is obtained. The phase-separated composite that is obtained has the same properties as traditional microcomposites. Beyond this traditional class of polymer–filler composites, two types of nanocomposites can be obtained. Intercalated structures are formed when a single (or sometimes more) extended polymer chain is intercalated (sandwiched) between the silicate layers. The result is a well-ordered multilayer structure of alternating polymeric and inorganic layers. Exfoliated or delaminated structures are obtained when the silicates are completely and uniformly dispersed in the continuous polymer matrix. The delamination configuration is of particular interest because it maximizes the polymer–clay interactions, making the entire surface of the layers available for the polymer. This should lead to the most significant changes in mechanical and physical properties. A schematic of the types of possible structure formations in PC composites is depicted in Fig. 13.2.

© 2007, Woodhead Publishing Limited

Polyolefin/clay nanocomposites

Clay

Microcomposite

Intercalated nanocomposite

357

Polymer

Exfoliated nanocomposite

13.2 Schematic representation of possible structure formation in PC composites.

13.3.1 Preparation techniques for polyolefin/clay nanocomposites Different paths have been proposed to prepare PC nanocomposites, such as solution intercalation, in situ polymerization and melt intercalation. Solution intercalation This is one of the most common routes for preparation of PC nanocomposites. In this technique polar solvents are used to swell the organoclay followed by dissolving the polymer in the same solvent and mixing the two so that the polymer chains intercalate between the clay layers. The compensation of the decrease in the conformational entropy of the intercalated polymer chains by the entropy gained due to desorption of the solvent molecules is the driving force for polymer intercalation to take place using the solution intercalation technique. Although this is an advantageous method on the laboratory scale it is difficult to apply the solution intercalation method in industry owing to the problems associated with the use of large quantities of solvent. In situ polymerization This is a conventional process of synthesizing PC nanocomposites, especially those based on thermoset polymers. The organoclay is first swollen in the monomer. This requires a certain amount of time, which is governed by the polarity of the monomer molecules, surfactant molecules in the organoclay,

© 2007, Woodhead Publishing Limited

358

Nanofibers and nanotechnology in textiles

and the swelling temperature.16 This is followed by the initiation of the reaction. The driving force for this mechanism is linked to the polarity of the monomer molecules: e.g., during the swelling phase, the high surface polarity of the clay attracts the polar monomer molecules so that they diffuse between the silicate layers. After reaching a certain equilibrium the diffusion stops and the layered silicate is swollen in the monomer to the extent equivalent to the perpendicular orientation of the alkyl ammonium surfactant.17 When the polymerization is initiated, either by heat or radiation, by diffusion of a suitable initiator or by a catalyst fixed through cation exchange prior to the swelling step, the overall polarity of the intercalated molecules is lowered. This displaces the thermodynamic equilibrium so that more polar molecules are driven between the clay layers. This continuum eventually results in the organic molecules separating the clay layers.18 Melt intercalation Melt intercalation consists of blending a molten thermoplastic with an organically modified clay in order to optimize the polymer–clay interactions. It is a promising approach for forming nanocomposites that would greatly expand the commercial opportunities of this technology. The polymer chains have a significant loss of conformational entropy during the intercalation. The proposed driving force for this mechanism is the important enthalpic contribution of the polymer/organically modified clay interactions during the blending and annealing steps. In the melt intercalation process, rheological and thermodynamic characters of the materials are important parameters that affect the degree of intercalation and properties of the final nanocomposites. Generally the degree of intercalation of the polymer in the organically modified clay is governed by matrix viscosity, average shear rate and mean residence time in the mixing process. If technically possible, melt compounding could be significantly more economical and simple than in situ polymerization processes. This approach would allow nanocomposites to be formulated directly using conventional extruders and mixers as needed without the necessary involvement of resin producers. Indeed, PC nanocomposites have been successfully produced by extrusion. A wide range of thermoplastics, including strongly polar polyamide-6, ethylene vinyl acetate and polystyrene, has been intercalated between clay layers. However POs, which represent the biggest volume of polymers produced, have so far only been successfully intercalated to a very limited extent. As there are very few studies on formation of nanocomposites by direct melt intercalation, the corresponding knowledge of this process and its accomplishments is still far from complete.

© 2007, Woodhead Publishing Limited

Polyolefin/clay nanocomposites

359

13.3.2 Compatibilization issues in polyolefin/clay nanocomposites Mere organophilization of the clay is often not sufficient to obtain intercalated or exfoliated PC nanocomposites. This is mostly true in the case of PO. This is mainly due to the low polarity of these polymers, which makes it difficult to get the exfoliation and homogeneous dispersion of the silicate layers at the nanometer level in the polymer. Silicate layers have polar hydroxyl groups, which are compatible only with polymers containing polar functional groups. This necessitates the use of a compatibilizer to facilitate the formation of an intercalated or exfoliated polymer/clay nanocomposite. Another problem in the development of nanocomposites based on POs is the hydrophilic nature of the clay surfaces and the hydrophobic nature of PO. Thus the clay needs to be modified using suitable organic surfactants. The organically modified clay is adequate for most of the polar polymers, while for nonpolar PO, use of a compatibilizer is often required additionally to facilitate the intercalation of the polymer. The structure and properties of PC nanocomposites are governed by the choice of compatibilizer, its content and the nature of organic modifiers. A lot of work has been reported in polypropylene (PP)/clay nanocomposites using maleic anhydride (MA)-grafted PP as the compatibilizing agent.19–33 The effectiveness of the compatibilizer depends on the molecular weight and the MA content. Higher MA content may facilitate melt intercalation; however, it may lead to immiscibility between the PP matrix and the compatibilizer. Most of the literature reports that the use of oligomers enhances the melt intercalation process, thereby resulting in better improvement in the mechanical properties. Generally the ratio of MA PP to clay used varies from 1:1 to 3:1. However, higher content of MA PP may not be cost effective and may also lead to inferior mechanical properties. In spite of lots of studies this truly exfoliated PP/clay hybrid cannot be prepared by melt mixing techniques. Ding et al.34 prepared PP/clay nanocomposites using a highly effective PP solid-phase graft as a compatibilizer (solid-phase graft contained MA and butyl acrylate as grafting monomers and had a grafting percentage of 11.8%). This technique facilitated increase in mechanical properties, thermal stability and crystallization temperature of the nanocomposites. Wang et al.32 studied melt processed PP/clay nanocomposites modified with maleated PP compatibilizers. The organoclay was first blended with PP-g-MA. They used different grades of PP-g-MA with a wide range of MA content and molecular weight. PP/organically modified clay nanocomposites were then modified with different levels of PP-g-MA compatibilizers on a twin-screw extruder. They reported that although PP-g-MA with lower molecular weight and higher MA content could lead to good clay dispersion in PP/clay composites, it caused deterioration in both the mechanical and thermal properties of the

© 2007, Woodhead Publishing Limited

360

Nanofibers and nanotechnology in textiles

composites. They concluded that there were two important factors contributing to exfoliation and homogeneous dispersion of the clay layers: (1) intercalation capability of the compatibilizers in the clay layers and (2) the composition of the compatibilizer in PP/clay composite. Garcia-Lopez et al.33 have studied the effect of compatibilizing agents on clay dispersion in PP/clay nanocomposites. For this purpose they prepared PP/clay nanocomposites using two different coupling agents, diethyl maleate (DEM) and MA, and two different clays. They observed that the differences in mechanical properties when using different clays are smaller if DEM is used instead of MA. They concluded that clay dispersion and interfacial adhesion are greatly affected by the kind of matrix modification.

13.4

Polypropylene/clay nanocomposites

PP is one of the most widely used PO polymers and a very attractive candidate for the matrix phase of polymer nanocomposites. However, as PP does not include any polar groups in its backbone, it was thought that the homogeneous dispersion of silicate layers in PP would not be possible. Earlier experimental approaches to intercalate PP in to the silicate layers involved using clays modified with nonpolar organic molecules.35 In this approach a PO oligomer with telechelic OH groups, which was used as a compatibilizer, was first intercalated in the organically modified silicate layers through strong hydrogen bonding. This results in weakening of interlayer interaction through expansion of the silicate layers. A PP/clay hybrid was obtained by mixing PP with the organoclay modified using the functional oligomer. Although improved dispersion of the clay layers in PP was obtained through this process, some aggregates of the clay minerals were still observed in the hybrid. In a subsequent paper, Kawasumi et al. reported the use of MA-modified PP oligomers as compatibilizers.20 They postulated that there are two important factors, that govern the homogeneous dispersion of clay layers and formation of the exfoliated PP/clay hybrid, viz. (1) the intercalation capability of the oligomers in the clay interlayers and (2) the miscibility of the oligomers with PP. In a following paper Hasegawa et al. reported the preparation of PP/clay nanocomposites using PP modified with MA by the melt intercalation technique.36 On comparison of the dispersibility of the clay in the hybrid containing PP-g-MA with that containing homo-PP it was observed that the former did not show the presence of clay stacks while an apparent peak corresponding to the d001 plane reflection of the clays was observed in the latter. Further, the homo-PP/clay hybrid showed the same d-spacing as that of the organoclay used in this study. This research served as the basis for the use of MA-grafted PP as the compatibilizer, which was widely applied as the compatibilizing agent in the fabrication of PP/clay nanocomposites over the following years.

© 2007, Woodhead Publishing Limited

Polyolefin/clay nanocomposites

361

Few approaches other than the usage of MA-g-PP have been reported for the preparation of PP/clay nanocomposites. Thermodynamically it is not possible to produce miscible nanocomposites of PP and alkyl ammonium modified clays without the assistance of solvents or extensive shear.37 Along the same lines, Manias et al. developed two approaches. In the first approach, nanocomposites were formed by attaching minute amounts of polar functional groups to PP.38 These functionalized PP derivatives were either attached to PP in blocks or randomly grafted so that the polymers still resembled the neat PP closely (only 0.5% functional groups) but at the same time had adequate polar character to become miscible with alkyl ammonium modified clays ether by direct melt intercalation or co-extrusion of PP with the organically modified PP. In the second approach, the formation of PP/clay nanocomposites was realized by using neat/unmodified PP and a semifluorinated surfactant modification for the clay. Here also nanocomposite formation was achieved by direct melt interaction unassisted by shearing or solvents suggesting sufficiently favorable thermodynamic interactions between the PP and the clay. In both the cases microstructural characterization of PP/clay nanocomposites revealed a coexistence of intercalated and exfoliated clay layers throughout the polymer matrix. These nanocomposites exhibited enhanced mechanical properties, including higher moduli and strength, and at the same time were found to be thermally more stable than neat PP as well as exhibiting enhanced flammability characteristics. Moreover they were found to be more resistant to solvents and to have improved starch resistance compared with neat PP. In a subsequent paper, Manias and coworkers proposed another novel approach for the preparation of PP/clay nanocomposites.39 They used ammonium terminated PP as the organic modification for montmorillonite to obtain exfoliated PP/clay nanocomposites. They also show the advantage of chain-end-functionalized PP over side-chain-functionalized or block copolymer PP to obtain exfoliated structures rather than intercalated ones. Sun and Garces reported a novel synthesis approach to make high-performance PP/ clay nanocomposites by in situ polymerization with metallocene/clay catalysts under mild polymerization conditions, which resulted in great improvements in mechanical performance and processability of the polymer.40 This approach is free from the usage of any external activators needed for initiating the olefin polymerization and also high pressure and temperature process conditions are not required. Kato et al. have reported a new production method for PP/clay nanocomposites.41 The highlight of this method was a lack of prior organic modification of clay. This method focuses on the nature of the clay mineral, which was exfoliated in water. By controlling the pressure of the water vapor, the exfoliation of the clay mineral was achieved in a twin-screw extruder. Two compatibilizers, MA-modified PP and octadecyl trimethyl

© 2007, Woodhead Publishing Limited

362

Nanofibers and nanotechnology in textiles

ammonium chloride, were added to the mixture of the clay mineral and PP to prevent aggregation of the clay in the twin-screw extruder. The PP/clay nanocomposite thus formed consisted of exfoliated clay layers, which were dispersed uniformly in the polymer matrix as evidenced by XRD and transmission electron microscopy (TEM) studies. The PP/clay nanocomposite had almost the same excellent properties as a conventional PP/clay nanocomposite. The obvious advantage of this method is the elimination of the organic modification of the clay, which has to be carried out separately from the melt compounding in an extruder. The process of organic modification of clay, which consists of various industrial processes of agitation, filtration, drying and milling, can be cumbersome and is also costly. Thus, the elimination of organic modification will simplify the overall process of PC nanocomposite production as well as reduce the cost of the final product as compared with PC nanocomposites prepared using the conventional processes. However, the successful implementation of this process for other polymers is yet to be demonstrated. For the successful production and application development of PP/clay nanocomposites a better understanding of the structure–property relationships in these multicomponent systems from the point of view of a wide spectrum of variations in the nature of ingredients and processing environments is necessary. Taking this into consideration one of the major criteria governing the performance of PP/clay nanocomposites is the inherent structure of the organically modified clay. The specific variations imposed on the organically modified clays involve the initial interlayer spacing and the packing density of the organic chains within the clay gallery space.30, 42 During the process of intercalation, the polymer chains penetrating the clay layers have to overcome a large entropy barrier, which is a result of the closely packed clay layers. The entropy barrier can be lowered by favorable enthalpic contributions that are possible by modification of the clay using suitable organic moiety such that the polymer–clay interactions are more favorable than the clay–surfactant interactions. Increasing the modifier concentration of the surfactant leads to a more efficient packing and increased interlayer contact between the clay layers, resulting in more cohesive van der Waals interactions between the clay layers and, thus, a greater interlayer solid-like character. Formation of intercalated hybrid will be favored in systems with favorable enthalpic contributions, which are possible when polymer–clay interactions are more favorable than the clay–intercalant interactions. Thus, systems that contain organically modified clays with intercalant concentration, which is thermodynamically favorable to compensate for the loss of conformational entropy by enthalpic gains when the clay comes in contact with polymer chains, are more likely to have a higher amount of polymer intercalated in the clay interlayer. Hence, higher intercalant concentration in the clay interlayer makes it more dense and reduces the polymer–clay interactions, which in

© 2007, Woodhead Publishing Limited

Polyolefin/clay nanocomposites

363

turn makes it harder for the free polymer chains to infiltrate the clay gallery space. To understand the mechanisms underlining the property enhancements observed in PC nanocomposites it is necessary to understand how the nanomorphologies of these advanced materials influence their properties. Gilman et al. have studied the flammability properties of PP/clay nanocomposites to understand the use of these nanocomposites as flame retardants.43 The type of clay, level of dispersion in polymer matrix and processing degradation influence the magnitude of the flammability reduction in PP/clay nanocomposites. It is concluded that a high-performance carbonaceous-silicate char builds up on the surface during burning, which insulates the underlying material and slows down the mass loss rate of decomposition products. Zanetti and coworkers reported the flame-retardant properties based on PP–graft–MA and organically modified clays using oxygen consumption cone calorimetry.44 The effect of addition of traditional flame retardants, viz. decabromodiphenyl oxide (DB) and antimony trioxide (AO), to the nanocomposites was investigated by cone calorimetry and conventional flame-retardant evaluation techniques such as limiting oxygen index and vertical burning test. For comparison, the flammability properties of the PP/ clay nanocomposites were studied both in the presence and absence of DB and AO. The nanocomposites showed better flame-retardant properties than the pure polymer. The peak heat release rate reduced still further with the use of AO or BD. When both the additives were added to the nanocomposite a synergistic effect results, which did not occur under identical testing conditions, when AO and BD were added to the pure polymer. Thermal characteristics of PP/clay nanocomposites and the organically modified clays used as nano fillers were investigated by Lee et al. using XRD, TEM and Fourier transform infrared (FTIR) spectroscopy as characterization tools.45 Two organically modified clays containing different alkylammonium surfactants were used in this study. The PP/clay nanocomposites were prepared by melt blending of the polymer with the organically modified clays and MA-grafted PP oligomer as a compatibilizer. The modified clays showed decrease in the interlayer spacing at the processing temperature due to thermal decomposition of the organic cation. It was observed that the thermal characteristics of the modified clays are dependent not only on the type of alkylammonium surfactant but also on the interlayer structure. Partial exfoliation of PP composite was observed during processing at high temperatures, using the alkylammonium-modified clay containing ethoxy groups (A), which initially had a small interlayer spacing and polar character. This behavior was attributed to the decrease in the interlayer spacing and less surface organophilicity due to thermal decomposition of this clay during processing. The PP/clay nanocomposite prepared using the modified clay containing the alkylammonium surfactant with purely aliphatic chains (B)

© 2007, Woodhead Publishing Limited

364

Nanofibers and nanotechnology in textiles

exhibited the formation of intercalated or exfoliated structure. This difference in behavior was because the clay B was more organophilic initially compared with clay A, which led to melt intercalation in spite of thermal decomposition, resulting in intensive intercalation or exfoliation aided by shear. Crystallization can be used effectively to control the amount of polymer intercalating the clay gallery space, which in turn will control the morphology, mechanical and other properties of PC nanocomposites. Jog and coworkers were first to report the crystallization behavior and the morphology of PP/ clay nanocomposites prepared by melt processing and using PP–graft–MA as the compatibilizer.27, 28 They found that the PP/clay nanocomposites crystallize at a higher temperature than pure PP. Also the spherulitic morphology of PP is completely altered in PP/clay nanocomposites, which crystallize in the form of peculiar birefringent structures, which grow with time. This indicates that the surface of the exfoliated clay layers acts as a nucleating agent that promotes crystallization in PP. Isothermal crystallization studies on PP/clay nanocomposites revealed that the nanocomposites exhibit a narrower isothermal crystallization peak than pure PP. The PP/clay nanocomposites exhibited lower values of total crystallization time, indicating that the crystallization process of PP is accelerated in the presence of clay. In a subsequent paper Jog and coworkers reported similar observations from the isothermal crystallization studies carried out on PP/clay nanocomposites prepared using three different grades of PP, two grades of MA-grafted PP containing different percentages of MA and two clays.29 Okamoto and coworkers reported the crystallization behavior of three different PP/clay nanocomposites.46 They found that the linear growth rate of the spherulites and the overall rate of crystallization is not influenced by the presence of clay. XRD studies revealed that the intergallery spacing increases with increasing crystallization temperature (Tc) and at constant Tc the extent of intercalation increased with decreasing clay content. TEM observations revealed that clay particles are well dispersed at low Tc, while segregation of clay layers occurs at high Tc. Dynamic mechanical analysis (DMA) shows that the storage modulus (G′) increases with increasing Tc. This increase is limited to 30% for nanocomposites with low clay content (4 wt% clay) and it decreases to lower values with increasing clay content. Thus, by controlling the intercalation through crystallization at a suitable temperature the microstructure, morphology and mechanical properties of crystalline PC nanocomposites can be controlled. DMA is an important tool for studying the structure–property relationships in polymer composites. DMA essentially probes the relaxations in polymers, thereby providing a method to understand the mechanical behavior and the molecular structure of these materials under various conditions of stress and temperature. Hasegawa et al. reported the mechanical properties of PP/clay nanocomposites prepared using MA-modified PP oligomer as a com-

© 2007, Woodhead Publishing Limited

Polyolefin/clay nanocomposites

365

patibilizer.22 The storage moduli values observed for PP/clay nanocomposites were found to be higher than those of pure PP up to 130 °C. The relative storage moduli of PP/clay nanocomposite to those of pure PP were studied to understand the effect of clay hybridization. These values were found to be relatively small below the glass transition temperature (Tg) of PP; however, above Tg they drastically increased and then decreased to melt. This tendency was emphasized more strongly as clay layers were more uniformly dispersed in the PP matrix. The relative storage moduli of poorly dispersed clay layers were found to be relatively small and almost independent of temperature. Jog and coworkers have reported that the PP/clay nanocomposites exhibit higher storage moduli (increase of about 56%) than neat PP over the entire temperature range studied (–40 to 120 °C).27, 28 The loss modulus that corresponds to the dissipation of the energy shows a peak at about 11 °C for PP, which corresponds to the glass transition temperature of PP. In PP/clay nanocomposites, this peak is not well defined. It was thus concluded that the cooperative relaxation of PP in the nanocomposites becomes weak owing to the restricted mobility of the chains in the presence of the clay layers. There are many reports in the literature that show remarkable improvements in the tensile properties of the polymeric materials when organically modified clays are used as nano-fillers. A combination of improved stiffness and strength is only possible when either intercalated or exfoliated nanocomposites having uniformly dispersed clay platelets are formed. Better stress transfer between the filler and the polymer will result in improved mechanical properties of the polymer composite. This is possible with the formation of nanostructures. Reichert et al. have systematically evaluated the influence of organic modification of clays and compatibilizer functionality on the structure and properties of PP/clay nanocomposites using a series of organophilic clays modified using various protonated alkyl amines and two types of PP–graft– MA oligomers with different anhydride functionalities.24 The results showed that tensile property enhancement was possible only when a specific compatibilizer was used in combination with appropriate organically modified clays. Jog et al. reported an enhancement in the flexural properties of PP/ clay nanocomposites prepared by melt blending.27 The flexural modulus increased by 30% while the flexural strength increased by 25% compared with neat PP. An increase in the tensile modulus and tensile strength of the nanocomposites was also reported. The improved mechanical properties depended on the type of organic modifier and the inorganic content of the clay. It was observed that the clay containing more inorganic matter showed better improvement than that containing less. A rheological study of polymer nanocomposites is of great importance considering their fabrication and final applications. Rheology is a powerful tool, which provides important information on the internal microstructure of the nanocomposites, the state of dispersion of the clay, its orientation and

© 2007, Woodhead Publishing Limited

366

Nanofibers and nanotechnology in textiles

aspect ratio, as well as the effects of particle–particle interactions in clay and particle–polymer interactions on the viscoelastic behavior of the nanocomposites. Therefore, an understanding of the viscoelastic properties of polymer nanocomposites, both in solid and molten state, is essential from a processability and structure–property point of view. Dynamic oscillatory shear experiments are representative of the durability of the sample under various conditions of vibrations and/or external stress. Galgali et al. have studied the rheological response of isotactic PP/clay nanocomposites in the presence and absence of the compatibilizer, viz. PP– graft–MA.47 They show that the nanocomposites show a solid-like rheological response at low frequencies, which is completely dependent on the amount of clay loading in the nanocomposites, and is independent of the finer structure of the nanocomposites, i.e. whether they are end-tethered or melt intercalated. They also show that the typical rheological response of the PC nanocomposites does not arise due to restricted mobility of the confined polymer chains between the clay galleries, but is an effect of the frictional interactions between the clay layers. Further they have shown that beyond the apparent yield stress the zero shear viscosity of the PP/clay nanocomposite containing 7 wt% clay and the compatibilizer drops dramatically by more than three orders of magnitude, suggesting that the solid-like behavior of the PP/clay nanocomposites after annealing is due to the formation of a percolating network that strongly resists deformation. Recently, Abranyi et al. reported through rheological investigation that a percolated clay network formation is possible only under certain conditions.48 Accordingly, at low clay content a higher degree of exfoliation is needed to produce the number of silicate layers to form a percolated structure. When the clay content is higher, which increases the number of individual platelets, the amount of compatibilizer required for the formation of the network is less. To summarize a threshold concentration of MAPP exists in PP/clay nanocomposites for the formation of the clay network, which depends on the silicate content. Lele et al. studied the rheology–microstructure links in syndiotactic PP/ clay nanocomposites using in situ rheo-X-ray measurements of the nanocomposites during shear.49 The microstructure of the nanocomposite at low strain and shear rates consisted of a percolating three-dimensional network of dispersed clay tactoids. Rheology indirectly indicated that the observed yielding of the material at high stresses might be linked to the orientation of the clay tactoids, which was evidenced using the rheo-X-ray diffraction measurements. The transient experiments showed that the clay tactoids relax their orientation incompletely after cessation of shear. It was observed that the stress relaxation of the matrix chains, which were compatibilized with clay platelets, drives the orientation relaxation in about 10–3 s. After this time scale the clay tactoids are jammed and produce a percolating network having a residual orientation and a residual relaxation modulus. In steady

© 2007, Woodhead Publishing Limited

Polyolefin/clay nanocomposites

367

shear experiments at shear rates below 10–3 s–1 the response corresponding to that of a percolating network would be observed while at shear rates above 10–3 s–1, a yield-like response would be observed. Okamoto et al. investigated the elongational flow-induced structure formation of PP/clay nanocomposites by carrying out uniaxial elongational tests at constant Hencky strain in molten state using elongational flow optorheometry.50 The PP/clay nanocomposites exhibited high viscosity and a tendency of strong strain-induced hardening. This was attributed to the perpendicular alignment of the clay layers to the stretching direction. They observed two features from the shear viscosity curve. First, the extended Trouton rule is invalid for PP/clay melt, in contrast to the melt of ordinary homopolymers. Second, the shear viscosity of the PP/clay melt increases continuously with time without reaching the steady state within the time spans studied. This time-dependent thickening behavior is termed as rheopexy. These differences reflect the difference in the shear flow-induced versus elongational flow-induced internal structure fomation in PP/clay nanocomposite melts. Rheopexy was not observed for PP homopolymer modified with PP– graft–MA melts. TEM results were presented for the nanocomposites along the x- and yaxes of the elongated specimens, which represented the directions parallel and perpendicular to the stretching direction. A converging flow is applied to the thickness direction (y- and z-axis) with stretching if the assumption of affine deformation without volume change is valid. For the specimen elongated with high strain rate, perpendicular alignment of the clay layers (edges) was observed along the stretching direction (x-axis) in the x–y plane. For the x– z plane, the clay layers (edges) disperse into the PP matrix along the z-axis direction rather than randomly, but these faces cannot be observed in this plane. From the experimental results and the two-directional TEM results the authors conclude the formation of a house of cards-like structure under slow elongational flow.

13.5

Polyethylene/clay nanocomposites

Polyethylene can be used for a wide range of applications because, depending on its structure, it can be produced in many different forms. Low-density polyethylene (LDPE) was the first type of polyethylene to be commercially exploited. It can be characterized by a large degree of branching, forcing the molecules to be packed rather loosely, forming a low-density material. LDPE is soft and pliable and has a variety of applications including plastic bags, containers, textile and electrical insulation, as well as coatings for packaging materials. There have been few attempts to develop nanocomposites based on LDPE and clays for improvement in materials properties. Morawiec et al. reported the preparation of nanocomposites based on compatibilized LDPE

© 2007, Woodhead Publishing Limited

368

Nanofibers and nanotechnology in textiles

and organically modified clays.51 The nanocomposites were prepared by melt-mixing MA-grafted LDPE with octadecyl amine-treated clay in an internal mixer. Improved thermal stability in the nanocomposites was observed which was attributed to the exfoliation of the clay platelets in the polymer matrix. Further, it was observed that the uncompatibilized LDPE/clay composite fractures early during drawing due to decohesion of the polymer from nonexfoliated clay platelets and did not exhibit improved thermal properties. The nanocomposite containing more compatibilizer could be deformed to a higher elongation. The authors argued that the mechanical performance of these nanocomposites is not only governed by clay exfoliation and clay content but also by the presence of a significant amount of compatibilizer, which is reflected in the above result. They concluded that MA-grafted polyethylene not only promotes the exfoliation of the clay and its good adhesion as compared with pristine LDPE but also toughens the polymer matrix. In an earlier study Antipov et al. reported the structure and deformation behavior of LDPE/clay nanocomposites.52 The effects on the type and concentration of a filler on the structure and deformation behavior of LDPE/ clay nanocomposites was studied. The clay was organically modified using various substrates. From the crystallization study it was observed that a part of the polymer crystallizes on the surface of layered silicate particles as heterogeneous nuclei, giving rise to a finely crystallite fraction and a bimodal crystallite size distribution. The nanocomposites showed an improved Young’s modulus, while the tensile strength and elongation at break decreased compared with the pristine polymer. They also argued that the cold drawing of composite films decreases the dimensions of the crystallites by more than two orders, although the melting temperature of LDPE in the nanocomposite with oriented clay platelets remained almost the same as in the isotropic pure polymer film.

13.5.1 Linear low-density polyethylene/clay nanocomposites Earlier research on linear low-density polyethylene (LLDPE)/clay nanocomposites focused on the preparation of LLDPE-based clay nanocomposites via in situ ethene homo- and copolymerization as well as by the melt compounding technique.53 It was observed that, compared with melt compounding, in situ polymerization, catalyzed with methylaluminoxaneactivated zirconocene, nickel and palladium catalysts, was more effective in the formation of LLDPE/clay nanocomposites as evidenced by larger interlayer spacing and formation of exfoliated anisotropic nanoclays with high aspect ratio. Another method for the preparation of LLDPE/clay nanocomposites was reported by Wang et al.54 They used MA-grafted polyethylene to prepare the

© 2007, Woodhead Publishing Limited

Polyolefin/clay nanocomposites

369

nanocomposites via the melt compounding technique. It was observed that the morphology of the LLDPE/clay nanocomposite (intercalated or exfoliated structure) was governed by the hydrophobicity of the organically modified clay and the hydrophilicity of the maleated polyethylene. It was also observed that in case of hybrids prepared with neat LLDPE, the intercalation capability of the polymer depended upon the thermodynamical equilibrium state at the clay surface rather than the initial interlayer spacing of the organophilic clay. Recently, Sanchez-Valdes et al. reported the preparation of LLDPE/clay nanocomposites using a zinc-neutralized carboxylate ionomer as a compatibilizer. 55 The results so obtained were compared with the nanocomposites prepared using LLDPE–graft–MA as compatibilizer. The nanocomposites prepared with ionomer showed good mechanical performance only slightly below that of the nanocomposites prepared using MA. This was attributed to the interactions between the ionomer functional groups and the polar groups in the clay surfactant and the nano-size of the clay, as well as the uniform dispersion of the clay in the polymer matrix. The oxygen permeability of the nanocomposites prepared using the ionomer was decreased by the addition of the clay. It was concluded that ionomeric compatibilization not only promotes the exfoliation of clay but also improves adhesion to LLDPE, and increases the mechanical performance of the polymer. Lew et al. reported the preparation of LLDPE/clay nanocomposites from metallocene-catalyzed and conventional Ziegler–Natta-catalyzed LLDPE by melt compounding.56 Owing to higher extrusion shear stress, attributed to a narrower molecular weight distribution, metallocene-catalyzed LLDPE was found to be more effective in exfoliating the clay layers than that synthesized from Ziegler–Natta catalyst. It was also shown that the polymer chainbranching, clay exfoliation, and PE–graft–MA compatibilizer concentration have significant effects on the three α, β and γ relaxation temperatures. Further it was observed that the degree of crystallinity increased with the addition of the compatibilizer PE–graft–MA and the presence of clay platelets appeared to have suppressed the overall crystalline lamellar formation as a result of restricted chain mobility. Hotta and Paul studied the effect of the use of organic modification (they have used clays modified with different alkyl ammonium modifiers, viz. having one alkyl tail and having two alkyl tails) on the morphology and mechanical properties of LLDPE/clay nanocomposites.57 The effect of use of LLDPE–graft–MA as a compatibilizer was also investigated. Nanocomposites derived from organic clay modified with two alkyl tails exhibited better clay dispersion and improvement in the mechanical properties as compared with that derived from the organically modified clay having only one alkyl tail. This behavior was attributed to the better affinity of LLDPE for the alkyl tails; thus increasing their number resulted in better

© 2007, Woodhead Publishing Limited

370

Nanofibers and nanotechnology in textiles

dispersion of clay in the LLDPE matrix. The rheology and gas permeability of the nanocomposites derived from the organically modified clay having two alkyl tails were also investigated and the results showed that both melt viscosity and melt strength increased with increased clay content and LLDPE– graft–MA content, while the gas permeability decreased by the addition of the clay.

13.5.2 High-density polyethylene/clay nanocomposites High-density polyethylene (HDPE) is a version of polyethylene that is defined by its higher density (greater than or equal to 0.941 g/cm3). It is harder, stronger and a little heavier than LDPE, but less ductile. It has a low degree of branching, thus stronger intermolecular forces. HDPE has an opaque wax-like appearance and its properties include good flexibility, good low temperature toughness, good resistance to chemicals and weathering, good processability by most of the processing techniques, and it is less expensive. HDPE is known to have poor barrier properties for gases, solvents and hydrocarbons. Thus, enhancement of barrier, mechanical and thermal properties of HDPE through incorporation of clay and formation of nanocomposites can open new avenues of applications for HDPE. Bafna et al. prepared HDPE/clay nanocomposites using a melt extrusion technique and maleated LDPE as a compatibilizer.58 They reported a study on the three-dimensional (3D) hierarchical orientation of six different structural features of these nanocomposites using two-dimensional (2D) small-angle X-ray scattering and 2D wide-angle X-ray scattering. The structural features studied included: clay clusters/tactoids (120 nm), modified clay (002) (2.4– 3.1 nm), unmodified clay (002) (1.3 nm), clay (110) and (020) planes normal to the previous two, polymer crystalline lamellae (001) (19.0–26.0 nm), and polymer unit cell (110) and (200) planes. The orientation data showed two surprising results. First, the clay tactoids are associated with the intercalated clay and not with the unmodified clay. Secondly, the unmodified clay is associated with the polymer lamellae and may modify the lamellar surface energy. By tuning the clay platelet orientation, the permeation and strength of the PO films could be controlled. They concluded that it is the lamellae and not the chains that govern the final orientation of the polymer crystals, since chain tilt leads to randomization of the unit cells and not the lamellar normals. Osman and Atallah have reported a comparative study on the effect of filler particle shape, size and surface treatment on polymer crystallinity and gas permeability using spherical and plate-like inclusion in HDPE.59 Platelike inclusions strongly reduce the polymer permeability coefficient while the spherical particles do not have any effect on it. The reduction in gas permeability depends on the average aspect ratio of the fillers, which in turn

© 2007, Woodhead Publishing Limited

Polyolefin/clay nanocomposites

371

depends on the exfoliation of the clay and consequently on its surface tension. Neither the spherical particles (calcium carbonate modified and unmodified) nor the plate-like particles (organically modified clay) were found to nucleate crystallization of HDPE under slow cooling conditions. In a following paper Osman et al. reported the gas permeability properties of HDPE/clay nanocomposites of organically modified clays having different surface coverage and alkyl chain packing density.60 An appreciable decrease in the permeation coefficient of the composites was observed for partially exfoliated clays in the polymer matrix. Xu et al. reported that on grafting with acrylic acid, the dispersion and intercalation of bentonite clay in HDPE/clay nanocomposite increased.61 Consequently, with increasing clay content in the nanocomposite formed with acrylic acid grafted HDPE, the tensile strength and Young’s modulus increased, while that of neat HDPE/clay nanocomposites decreased. Also the addition of clay to the grafted polymer was found to affect the melting temperature and the degree of crystallization of the matrix; by decreasing these values, however, no change was observed in the crystallization behavior of the neat HDPE/clay composite. Tanniru et al. studied the mechanical response of HDPE/clay nanocomposite. The micromechanism of plastic deformation during impact loading is studied with scanning electron microscopy (SEM), and the impact strength of the composites is linked to the structural studies performed using differential scanning calorimetry (DSC), DMA, and TEM and SEM.62 With the addition of clay the impact strength of HDPE decreases; however, the toughness continued to be higher even at very low temperatures. It was observed that the fracture of HDPE initiates with crazing, while the propagation zone involves a combination of different processes, viz. fast propagation of crack and shear process. The fracture initiation and propagation of HDPE/clay nanocomposites is characterized by stretching of fibrils (fibrillation) interdispersed with microvoids. The low toughness of the nanocomposites in relation to neat HDPE is related to the crystal structure and the interfacial interaction between the filler and the polymer matrix. The primary mechanism of deformation in the nanocomposites is altered from a combination of craze and drawing of fibrils in HDPE to a microvoid coalescence-fibrillated process.

13.5.3 Ultra-high molecular weight polyethylene/clay nanocomposites Ultra-high molecular weight polyethylene (UHMWPE) has extremely high abrasion resistance when compared with other thermoplastics. It also has exceptional impact resistance, even at cryogenic temperatures, and is superior to stainless steel. Other advantages of UHMWPE include: low moisture absorption, good electrical and thermal insulation, self-lubrication and chemical

© 2007, Woodhead Publishing Limited

372

Nanofibers and nanotechnology in textiles

inertness (except in some acids). Some formulations in which UHMWPE can be used are sheet, rod and tube. Some of its excellent uses are as wear strips, chain guides, chute and hopper linings, bushings, and boat and truck bed liners. Recently Park et al. reported the preparation and characterization of UHMWPE/clay nanocomposites by using the melt intercalation technique.63 XRD analysis of the nanocomposites indicated that the nanocomposites were formed upon exfoliation of the clay layers in the polymer matrix. The thermal stability of the nanocomposites increased with increasing clay content compared with the pristine polymer as evidenced from the thermogravimetric analysis (TGA) results. This was attributed to the tortuous path encountered by the diffusing volatiles within the nanocomposites. The mechanical property studies of the nanocomposites showed that they exhibited higher tearing energies and tensile strengths than those of the pristine polymer. They concluded that the interfacial and mechanical properties of UHMWPE were improved upon addition of the clay and formation of the nanocomposite.

13.6

Higher polyolefin/clay nanocomposites

13.6.1 Poly(4-methyl-1-pentene)/clay nanocomposites Poly(4-methyl-1-pentene) (PMP) is an important member of the PO family. It has wide-ranging application in various industries, including automotive components, light covers, as well as special medical applications such as pacemaker parts, blood collection and transfusion devices, medical and laboratory apparatus. It is also used for commodity applications such as microwave components, cookware and electronic components. Of particular interest are properties such as low density, higher thermal stability, chemical resistance, optical transparency and high permeability. However, there is a limitation on the applications of PMP as its mechanical properties deteriorate even at temperatures just above the ambient temperature. It is well known that incorporation of clay enhances the thermomechanical properties of the polymers. Thus, a study that focuses on this aspect through the formation of PMP/clay nanocomposites is of academic as well as industrial interest. However, very few studies have reported the preparation and properties of PMP/clay nanocomposites. Wanjale and Jog have fabricated PMP/clay nanocomposites using the melt intercalation technique.64 They studied the effects of clay modification and compatibilizer on the formation and properties of the nanocomposites. Two commercially available organically modified clays were used for the preparation of PMP/clay nanocomposites. X-ray diffraction studies showed the formation of intercalated, disordered and/or partially exfoliated hybrids. The reinforcing effect of the clays was studied using DMA. The nanocomposite formed using the clay modified with dimethyl hydrogenated tallow 2-ethylhexyl-quaternary ammonium methyl-sulfate

© 2007, Woodhead Publishing Limited

Polyolefin/clay nanocomposites

373

showed better mechanical properties than that formed using the clay modified with dimethyl, dihydrogenated tallow quaternary-ammonium chloride. Although the addition of compatibilizer improved the d-spacing of the nanocomposites along with their thermal stability it failed to bring about much improvement in the storage moduli values in the rubbery regime. This was attributed to the low melting point of the compatibilizer. In a following paper, Wanjale and Jog studied the effect of the organic treatment of the clays on the properties of PMP/clay nanocomposites.65 Three different organically treated clays were used for this study. The microstucture of the nanocomposites was elucidated using XRD. It was observed that the octadecyl amine (ODA)-treated clay showed better thermomechanical properties than the tallow-modified clays. This was attributed to the better dispersion of the clay layers in the ODA-treated clay compared with the tallow-treated clays.

13.6.2 Poly(1-butene)/clay nanocomposites Poly(1-butene) (PB) is a semicrystalline PO with some outstanding mechanical properties, viz. excellent resistance to creep and environmental cracking. It also exhibits time-dependent polymorphism. Wanjale and Jog were first to report the formation and properties of PB/clay nanocomposites.66, 67 The nanocomposites were characterized for structural and thermomechanical properties using XRD, DMA and TGA. The effect of clay modification on the crystallization behavior of PB and on its solid state transformations was also investigated. The nanocomposites showed an intercalated structure as evidenced from XRD studies. They exhibited improved thermal stability, increased storage modulus and a notable decrease in the coefficient of thermal expansion. The nanocomposites showed enhanced rate of isothermal crystallization compared with the pristine polymer. The most interesting observation made by the authors was the transformation of the metastable tetragonal form to a stable hexagonal form of the nanocomposite at a faster rate than the pristine polymer. They concluded that the observed changes in the properties of PB in PB/clay nanocomposites could be due to the higher aspect ratio of the clay particles and the extent of percolation as a result of intercalation of the polymer in the clay interlayers.

13.6.3 Other polyolefin/clay nanocomposites This section primarily deals with PC nanocomposites prepared from PO copolymers and blends. The commercial importance of polymer blends and copolymers is well known. A few reports involving PC nanocomposites based on these categories of polymers are summarized. Arroyo et al. studied the effect of pristine and organically modified clay

© 2007, Woodhead Publishing Limited

374

Nanofibers and nanotechnology in textiles

filler on the morphology and tensile properties of PP/LDPE blend matrices using a statistical design.68 The tensile behavior of the nanocomposites is shown to be more dependent on the matrix composition than on the clay content. A more resistant nanocomposite material is formed from organic modification of clay. A better compatibility between the polymer blend matrix and the clay can be observed from the SEM studies. Thermal and flammability properties of ethylene–vinyl acetate copolymer (EVA)/clay nanocomposites by blending with LLDPE were reported by Chuang et al.69 Morphological studies revealed the formation of an intercalated nanocomposite structure. An increase in the tensile strength was observed for the addition of 5 parts per hundred resin (phr) of the clay to the EVA/ LLDPE blend. The nanocomposite exhibited better protection and stabilization towards thermo-oxidation that is attributed to the barrier diffusion of the volatile decomposition products as well as oxygen from the gas phase to the polymer. The nanocomposite exhibits lower peak heat release rate than the neat polymer blend during combustion. In the presence of conventional fire retardants such as aluminum trihydroxide and antimony trioxide the peak release rate of the nanocomposite is lowered further. The addition of conventional flame retardants to the nanocomposite shows a synergistic effect on the flame retardancy and smoke suppression. Mishra et al. reported the preparation and properties of a ternary nanocomposite consisting of LLDPE, millable polyurethane (PU) and organically modified clay.70 Heat shrink behavior and the mechanical response of this nanocomposite was reported. A decrease in the heat shrinkability as the amount of the clay increased was observed with the entanglement points serving as the memory points during shrinkage. This was attributed to the reduction of the entanglement points in millable PU owing to the presence of the clay that disturbs the entanglement network of the millable PU. The tensile modulus of the nanocomposite increased substantially with increasing amount of clay. Shah et al. reported a study on the structure–property relationships for nanocomposites based on a sodium ionomer of poly(ethylene-co-methacrylic acid) and a series of organically modified clays.71 The presence of surfactant structural aspects that result in more shielding of the silicate surface and the increased alkyl material lead to improved exfoliation in the ionomer.

13.7

Conclusions

The results of work on PO/clay nanocomposites are summarized in Table 13.1. The preparation and properties of and issues underlining the structure– property relationships in nanocomposites based on various POs and clays have been systematically reviewed. It was observed that the non-polar nature of POs necessitates the usage of a compatibilizer such as an anhydride or the presence of acidic or ionic groups to improve the matrix polarity. Incorporation

© 2007, Woodhead Publishing Limited

Table 13.1 Summary of reported work on PO/clay nanocomposites Clay modifier

Compatibilizer

Method of preparation

Key observations/conclusions

Reference

PP

Alkyl ammonium cation; octadecyl amine

MA-g-PP

Melt intercalation

Better dispersion of clay. Enhanced crystallization rate. Significant changes in the morphology. Clay layers disturb spherulitic morphology. Enhance mechanical behavior.

22, 27, 28, 29, 36

Dioctadecyl-dimethyl ammonium bromide; octadecyl-ammonium bromide

Functionalized PP derivatives

Melt intercalation

Minute amounts of compatibilizer required, thus resemble neat PP closely.

38

Melt intercalation

Better mechanical, thermal and solvent resistance.

38

Physical mixing and melt blending

Exfoliated nanocomposite.

39

In situ polymerization

Mild polymerization conditions. Improved mechanical performance and processability without usage of external activators. Uniform dispersion of clay. Elimination of organic modification of clay hence cost effective.

40

Semi-fluorinated surfactants Octadecyl and dioctadecyl ammonium salts

Ammonium terminated PP

Amine complexes

© 2007, Woodhead Publishing Limited

Melt extrusion

41

375

MA-g-PP and octadecyl terimethyl ammonium chloride

Polyolefin/clay nanocomposites

Polymer

376

Table 13.1 Continued Polymer

Compatibilizer

Method of preparation

Key observations/conclusions

Reference

Dimethyl bis(hydrogenated) tallow ammonium chloride; bis(2-hydroxyethyl) methyl tallow ammonium chloride

MA

Melt blending

Increased flammability resistance.

43, 44

Dimethyl bis(hydrogenated tallow) ammonium chloride; bis(2-hydroxy-ethyl) methyl tallow ammonium chloride

MA-g-PP oligomer

Melt blending

Thermal properties are dependent not only on type of alkyl ammonium surfactant but also on clay interlayer structure.

45

Dimethyl bis(hydrogenated) tallow ammonium chloride; octadecyl amine

MA-g-PP

Melt blending

Solid-like rheological response arises due to frictional interactions between clay layers. Threshold concentration of MAPP governs the formation of a percolated network. First direct evidence of flow-induced orientation of clay tactoids. Strain-induced hardening and rheopexy originate from the perpendicular alignment of clay layers to the stretching direction.

47, 48, 49 50

© 2007, Woodhead Publishing Limited

Nanofibers and nanotechnology in textiles

Clay modifier

LDPE

Octadecyl amine

LLDPE

Dimethyldistearyl ammonium; dimethylstearylbenzyl ammonium salts

Ultra-low density PE-g-MA

Melt mixing

Improved thermal stability. Higher levels of compatibilizer lead to increased elongation of the nanocomposite. Mechanical performance is governed by clay exfoliation and content as well as by amount of compatibilizer. MA-g-LDPE toughens the polymer matrix.

51

In situ polymerization; melt compounding

Metallocene catalyzed in situ polymerization proved more effective than melt compounding in nanocomposite preparation.

53

MA modified PE

Melt compounding

Hydrophobicity of the modified clay and hydrophilicity of maleated PE govern the morphology of the nanocomposite.

54

Dimethyl dihydrogenated tallow ammonium chloride

Zincneutralized carboxalate ionomer. MA

Melt blending

Mechanical performance of nanocomposites with ionomer is only slightly below that of nanocomposites with MA.

55

Octadecyl amine

MA-g-LLDPE oligomer

Melt compounding

Nanocomposites were prepared from metallocene and Ziegler–Natta catalyzed LLDPE. The former was found to be more effective in exfoliating the clay layers than the latter.

56

Polyolefin/clay nanocomposites

Dimethyl dihydrogenated tallow ammonium chloride; octadecyl amine

377

© 2007, Woodhead Publishing Limited

378

Table 13.1 Continued Polymer

Compatibilizer

Method of preparation

Key observations/conclusions

Reference

Dimethyl bis(hydrogenated) tallow ammonium chloride; trimethyl hydrogenated tallow ammonium salt

LLDPE-g-MA

Melt compounding

Modifier with two alkyl tails found to be better than that with one tail.

57

Quaternary ammonium salt Octadecyltrimethyl ammonium; dioctadecyldimethyl ammonium; benzylhexadecyltrimethyl ammonium; tributyloctadecyl phosphonium; triphenylhexadecyl phosphonium

Maleated PE

Melt compounding Melt blending

Lamellae not chains govern the final orientation of polymer crystals. Reduces the polymer permeability coefficient. Gas permeability strongly dependent on average aspect ratio of the fillers.

58

Octadecyl trimethyl ammonium salt

Acrylic acid

Melt compounding

Enhanced mechanical performance. Decrease in melting temperature and degree of crystallization of the matrix.

61

Melt processing

Primary mechanism of deformation in nanocomposites is altered from combination of craze and drawing

62

Dimethyl dialkyl ammonium salt

© 2007, Woodhead Publishing Limited

59, 60

Nanofibers and nanotechnology in textiles

HDPE

Clay modifier

of fibrils of HDPE to microvoid– coalescence–fibrillation process. UHMWPE PMP

Better thermomechanical properties. Compatibilizer with high softening/ melting point governs the properties of nanocomposites.

63 64

Octadecyl amine; dimethyl hydrogenated 2-ethylhexyl quaternary ammonium methyl sulfate; quaternary ammonium methyl sulfate;

Melt blending

Enhanced mechanical performance.

65

PB

Octadecyl amine

Melt blending

Enhanced thermomechanical properties. Faster transformation of the metastable tetragonal form to a stable hexagonal form.

66, 67

PP/LLDPE

Alkylbenzene dimethyl ammonium chloride

Melt blending

Tensile behavior depended on matrix composition rather than clay content.

68

PE-MA-acrylic ester terpolymer

Polyolefin/clay nanocomposites

Melt intercalation Melt blending

Dimethyl hydrogenated tallow 2-ethylhexylquaternary ammonium methyl sulfate; dimethyl dihydrogenated tallow ammonium chloride

379

© 2007, Woodhead Publishing Limited

380

Table 13.1 Continued Clay modifier

EVA/LLDPE

LLDPE/PU

Compatibilizer

Method of preparation

Key observations/conclusions

Reference

Organic modifier contained an allyl group and a lauryl group and two hydroxyethyl groups

Melt blending

Nanocomposites exhibit lower peak heat release rate. Synergistic effect on fire retardancy and smoke suppressing after addition of conventional fire retardants.

69

Bis(2-hydroxy-ethyl) methyl tallow ammonium chloride

Melt blending

Decrease in heat shrinkability with increasing clay content. Enhanced mechanical properties.

70

© 2007, Woodhead Publishing Limited

Nanofibers and nanotechnology in textiles

Polymer

Polyolefin/clay nanocomposites

381

of suitable functional groups is also found to be beneficial for PO/clay nanocomposite production. Thus by using a suitable combination of compatibilizer and organic modification of clay, PO/clay nanocomposites having superior properties can be fabricated. These nanocomposites possess several advantageous properties, including the following: • Superior mechanical properties in both solid and melt state compared with the pristine PO as well as conventional filler reinforced POs due to the reinforcing effect of clay at the nano-scale. • Microstructural properties such as morphology, rate of crystallization and the crystallinity of the PO are greatly affected by the presence of clay layers. Transformation to commercially useful stable polymorphic forms in the presence of clay has been observed in POs that exhibit polymorphism. • Dramatic improvement is observed in the barrier properties of the PO/ clay nanocomposites due to the formation of tortuous paths in the presence of the clay. • Thermal properties and flame retardancy of the POs in the presence of clays improve due to such phenomena as building up of high-performance carbonaceous–silicate char on the surface during burning, which insulates the underlying material and slows down the mass loss rate of decomposition products. Intercalation of the polymer chains in the clay interlayer and dispersion level of the clay platelets in the polymer matrix are some of the aspects that govern the properties of these PO/clay nanocomposites. The remarkable property improvement at low filler loadings makes these nanocomposites an attractive choice for the development of materials with fascinating properties and a rich variety of applications. Although much has been reported on the various aspects of preparation and property enhancements in PO/clay nanocomposites, a more in-depth understanding of the structure–property relationships in these nanocomposites is essential for the total utilization of their complete commercial potential. Only then can PO/clay nanocomposites truly be termed modern ‘high-performance commodity plastics’.

13.8

References

1. Kurauchi T., Okada A., Nomura T., Nishio T., Saegusa S. and Deguchi R., ‘Nylon 6– clay hybrid – synthesis, properties and application to automotive timing belt cover’, SAE Technical Paper Series 910584, International Congress and Exposition, Detroit, 1991. 2. Bailey S. W., ‘The status of clay mineral structures’, Clay and Clays Minerals Proc 14th National Conference, Berkley, California, Pergamon Press, 1966. 3. Greenland D. J., Laby R. H. and Quirk J. P., ‘Adsorption of amino-acids and peptides by montmorillonite and illite. Part 1. Cation exchange and proton transfer’, Trans Faraday Soc, 1965, 61, 2013–2023.

© 2007, Woodhead Publishing Limited

382

Nanofibers and nanotechnology in textiles

4. Van Olphen H., An Introduction to Clay Colloids Chemistry, New York, John Wiley and Sons Inc, 1977. 5. Hendricks S. B., ‘Base exchange of the clay mineral montmorillonite for organic cations and its dependence upon adsorption due to Van der Waals forces’, J Phys Chem, 1941, 45, 65–81. 6. Grim R. E., Allaway W. H. and Cuthbert F. L., ‘Reaction of different clay minerals with some organic cations’, J Am Ceram Soc, 1947, 30, 137–142. 7. Cowan C. T. and White D., ‘The mechanism of exchange reactions occurring between sodium montmorillonite and various n-primary aliphatic amine salts’, Trans Faraday Soc, 1958, 54, 691–697. 8. Vansant E. F. and Uytterhoeven J. B., ‘Thermodynamics of the exchange of nalkylammonium ions on sodium–montmorillonite’, Clay Clay Miner, 1972, 20, 47– 54. 9. Barrer R. M. and Brummer K., ‘Relations between partial ion exchange and interlamellar sorption in alkylammonium montmorillonite’, Trans Faraday Soc, 1963, 59, 959–968. 10. Greenland D. J., ‘Interactions between clays and organic compounds in soils’, Soils Fertilizers, 1965, 28, 415–425. 11. Jordan J. W., ‘Organophilic bentonites I. Swelling in organic liquids’, J Phys Colloid Chem, 1949, 53, 294–306. 12. Jordan J. W., Hook B. J. and Finlayson C. M., ‘Organophilic bentonites II’ J Phys Colloid Chem, 1950, 54, 1196–1208. 13. Greene-Kelly R., ‘Sorption of aromatic organic compounds by montmorillonite. Part 1. – orientation studies’, Trans Faraday Soc, 1955, 51, 412–424. 14. Lagaly G., ‘Interaction of alkylamines with different types of layered compounds’, Solid State Ionics, 1986, 22, 43–51. 15. Hackett E., Manias E. and Giannelis E. P., ‘Molecular dynamics simulations of organically modified layered silicates’, J Chem Phys, 1998, 108, 7410–7415. 16. Kornmann X., Lindberg H. and Berglund L. A., ‘Synthesis of epoxy–clay nanocomposites: influence of the nature of the clay on the structure’, Polymer, 2001, 42, 1303–1310. 17. Messersmith P. B. and Giannelis E. P., ‘Synthesis and characterization of layered silicate–epoxy nanocomposites’, Chem Mater, 1994, 6, 1719–1725. 18. Kornmann X., Lindberg H. and Berglund L. A., ‘Synthesis of epoxy–clay nanocomposites: influence of the nature of the curing agent on the structure’, Polymer, 2001, 42, 4493–4499. 19. Kurokawa Y., Yasuda H., Kashiwagi M. and Oya A., ‘Structure and properties of a montmorillonite/polypropylene nanocomposite’, J Mater Sci Lett, 1997, 16, 1670– 1672. 20. Kawasumi M., Hasegawa N., Kato M., Usuki A. and Okada A., ‘Preparation and mechanical properties of polypropylene–clay hybrids’, Macromolecules, 1997, 30, 6333–6338. 21. Kato M., Usuki A. and Okada A., ‘Synthesis of polypropylene oligomer–clay intercalation compounds’, J Appl Polym Sci, 1997, 66, 1781–1785. 22. Hasegawa N., Kawasumi M., Kato M., Usuki A. and Okada A., ‘Preparation and mechanical properties of polypropylene–clay hybrids using a maleic anhydride– modified polypropylene oligomer’, J Appl Polym Sci, 1998, 67, 87–92. 23. Oya A., Kurokawa Y. and Yasuda H., ‘Factors controlling mechanical properties of clay mineral/polypropylene nanocomposites’, J Mater Sci, 2000, 35, 1045–1050.

© 2007, Woodhead Publishing Limited

Polyolefin/clay nanocomposites

383

24. Reichert P., Nitz H., Klinke S., Brandsch R., Thomann R. and Mulhaupt R., ‘Poly(propylene)/organoclay nanocomposite formation: influence of compatibilizer functionality and organoclay modification’, Macromol Mater Eng, 2000, 275, 8–17. 25. Ishida H., Campbell S. and Blackwell J., ‘General approach to nanocomposite preparation’, Chem Mater, 2000, 12, 1260–1267. 26. Wang H., Zeng C., Svoboda P. and James Lee L., ‘Preparation and properties of polypropylene nanocomposites’, 59th ANTEC Proceedings, Dallas, 2001, 2203– 2207. 27. Kodgire P., Kalgaonkar R., Hambir S., Bulakh N. and Jog J. P., ‘PP/clay nanocomposites: effect of clay treatment on morphology and dynamic mechanical properties’, J Appl Polym Sci, 2001, 81, 1786–1792. 28. Hambir S., Bulakh N., Kodgire P., Kalgaonkar R. and Jog J. P., ‘PP/clay nanocomposites: a study of crystallization and dynamic mechanical behavior’, J Polym Sci Part B: Polym Phys, 2001, 39, 446–450. 29. Hambir S., Bulakh N. and Jog J. P., Polypropylene/clay nanocomposites: effect of compatibilizer on thermal, crystallization and dynamic mechanical behavior’, Polym Eng Sci, 2002, 42, 1800–1807. 30. Kim K.-N., Kim H. and Lee J.-W., ‘Effect of interlayer structure, matrix viscosity, and composition of a functionalized polymer on the phase structure of polypropylene– montmorillonite nanocomposites’, Polym Eng Sci, 2001, 41, 1963–1969. 31. Szazdi L., Pukanszky Jr B., Foldes E. and Pukanszky B., ‘Possible mechanism of interaction among the components in MAPP modified layered silicate PP nanocomposites’, Polymer, 2005, 46, 8001–8010. 32. Wang Y., Chen F.-B., Li Y.-C. and Wu K.-C., ‘Melt processing of polypropylene/ clay nanocomposites modified with maleated polypropylene compatibilizers’, Compos Part B – Eng, 2004, 35, 111–124. 33. Garcia-Lopez D., Picazo O., Merino J. C. and Pastor J. M., ‘Polypropylene–clay nanocomposites: effect of compatibilizing agents on clay dispersion’, Eur Polym J, 2003, 39, 945–950. 34. Ding C., Jia D., He H., Guo B. and Hong H., ‘How organo–montmorillonite truly affects the structure and properties of polypropylene’, Polym Test, 2005, 24, 94–100. 35. Usuki A., Kato M., Okada A. and Kurauchi T., ‘Synthesis of polypropylene–clay hybrid’, J Appl Polym Sci, 1997, 63, 137–139. 36. Hasegawa N., Okamoto H., Kato M. and Usuki A., ‘Preparation and mechanical properties of polypropylene–clay hybrids based on modified polypropylene and organophilic clay’, J Appl Polym Sci, 2000, 78, 1918–1922. 37. Vaia R. A. and Giannelis E. P., ‘Lattice model of polymer melt intercalation in organically-modified layered silicates’, Macromolecules, 1997, 30, 7990–7999. 38. Manias E., Touny A., Wu L., Lu B., Strawhecker K., Gilman J. W., and Chung T. C., ‘Polypropylene/montmorillonite nanocomposites. Review of the synthetic routes and material properties’, Chem Mater, 2001, 13, 3516–3523. 39. Wang, Z. M., Nakajima, H., Manias, E. and Chung T. C., ‘Exfoliated PP/clay nanocomposites using ammonium-terminated PP as the organic modification for montmorillonite’, Macromolecules, 2003, 36, 8919–8922. 40. Sun T. and Garces J. M., ‘High-performance polypropylene–clay nanocomposites by in-situ polymerization of metallocene/clay catalysts’, Adv Mater, 2002, 14, 128–130. 41. Kato M., Matsushita M. and Fukumori K., ‘Development of new production method for a poypropylene–clay nanocomposite’, Polymer Nanocomposites 2003, International

© 2007, Woodhead Publishing Limited

384

42. 43.

44.

45.

46.

47. 48.

49.

50.

51.

52.

53.

54.

55.

56. 57.

Nanofibers and nanotechnology in textiles Symposium on Polymer Nanocomposites Science and Technology, Quebec, Canada, 2003. Zhao Z., Tang T., Qin Y. and Huang B., ‘Effects of surfactant loadings on the dispersion of clays in maleated polypropylene’, Langmuir, 2003, 19, 7157–7159. Gilman J. W., Jackson C. L., Morgan A. B., Harris R., Manias E., Giannelis E. P., Wuthenow M., Hilton D. and Philips S. H., ‘Flammability properties of polymerlayered–silicate nanocomposites. Polypropylene and polystyrene nanocomposites’, Chem Mater, 2000, 12, 1866–1873. Zanetti M., Camino G., Canavese D., Morgan A. B., Lamelas F. J. and Wilkie C. A., ‘Fire retardant halogen–antimony–clay synergism in polypropylene layered silicate nanocomposites’, Chem Mater, 2002, 14, 189–193. Lee J. W., Lim Y. T. and Park O. O., ‘Thermal characteristics of organoclay and their effects upon the formation of polypropylene/organoclay nanocomposites’, Polym Bull, 2000, 45, 191–198. Maiti P., Nam P. H., Okamoto M., Hasegawa N. and Usuki A., ‘Influence of crystallization on intercalation, morphology, and mechanical properties of polypropylene/clay nanocomposites’, Macromolecules, 2002, 35, 2042–2049. Galgali G., Ramesh C. and Lele A., ‘Rheological study on the kinetics of hybrid formation in polypropylene nanocomposites’, Macromolecules, 2001, 34, 852–858. Abranyi A., Szazdi L., Pukanszky Jr B., Vancso G. J. and Pukanszky B., ‘Formation and detection of clay network structure in poly(propylene)/layered silicate nanocomposites’, Macromol Rapid Commun, 2006, 27, 132–135. Lele A., Mackley M., Galgali G. and Ramesh C., ‘In situ rheo-X-ray investigation of flow-induced orientation in layered silicate–syndiotactic polypropylene nanocomposite melt’, J Rheol, 2002, 46, 1091–1110. Okamoto M., Nam P. H., Maiti P., Kotaka T., Hasegawa N. and Usuki A., ‘A house of cards structure in polypropylene/clay nanocomposites under elongational flow’, Nano Lett, 2001, 1, 295–298. Morawiec J., Pawlak A., Slouf M., Galeski A., Piorkowska E. and Krasnikowa N., ‘Preparation and properties of LDPE/organo-modified montmorillonite nanocomposites’, Eur Polym J, 2005, 41, 1115–1122. Antipov E. M., Guseva M. A., Gerasin V. A., Korlev, Yu M., Rebrov A. V., Fischer H. R. and Razumovskaya I. V., ‘Structure and deformation behavior of nanocomposites based on LDPE and modified clays’, Polym Sci Ser A, 2003, 45, 1130–1139. Heinemann J., Reichert P., Thomann R. and Mulhaupt R., ‘Polyolefin nanocomposites formed by melt compounding and transition metal catalyzed ethane homo- and copolymerization in presence of layered silicates’, Macromol Rapid Commun, 1999, 20, 423–430. Wang K. H., Choi M. H., Koo C. M., Choi Y. S. and Chung I. J., ‘Synthesis and characterization of maleated polyethylene/clay nanocomposites’, Polymer, 2001, 42, 9819–9826. Sanchez-Valdes S., Lopez-Quintanilla M. L., Ramirez-Vargas E., Medellin-Rodriguez F. J. and Gutierrez-Rodriguez J. M., ‘Effect of ionomeric compatibilizer on clay dispersion in polyethylene/clay nanocomposites’, Macromol Mater Eng, 2006, 291, 128–136. Lew C. Y., Murphy W. R. and McNally G. M., ‘Preparation and properties of polyolefinclay nanocomposites’, Polym Eng Sci, 2004, 44, 1027–1035. Hotta S. and Paul D. R., ‘Nanocomposites formed from linear low density polyethylene and organoclays’, Polymer, 2004, 45, 7639–7654.

© 2007, Woodhead Publishing Limited

Polyolefin/clay nanocomposites

385

58. Bafna A., Beaucage G., Merabella F. and Mehta S., ‘3D hierarchical orientation in polymer–clay nanocomposite films’, Polymer, 2003, 44, 1103–1115. 59. Osman M. A. and Atallah A., ‘High-density polyethylene micro- and nanocomposites: effect of particle shape, size, and surface treatment on polymer crystallinity and gas permeability’, Macromol Rapid Commun, 2004, 25, 1540–1544. 60. Osman M. A., Rupp J. E. P. and Suter U. W., ‘Gas permeation properties of polyethylenelayered silicate nanocomposites’, J Mater Chem, 2005, 15, 1298–1304. 61. Xu Y., Fang Z. and Tong L., ‘On promoting intercalation and exfoliation of bentonite in high-density polyethylene by grafting acrylic acid’, J Appl Polym Sci, 2005, 96, 2429-2434. 62. Tanniru M., Yuan Q. and Misra R. D. K., ‘On significant retention of impact strength in clay-reinforced high-density polyethylene (HDPE) nanocomposites’, Polymer, 2006, 47, 2133–2146. 63. Park S.-J., Li K. and Hong S.-K., ‘Preparation and characterization of layered silicatemodified ultrahigh-molecular-weight polyethylene nanocomposites’, J Ind Eng Chem, 2005, 11, 561–566. 64. Wanjale S. D. and Jog J. P., ‘Effect of modified layered silicates and compatibilizer on properties of PMP/clay nanocomposites’, J Appl Polym Sci, 2003, 90, 3233– 3238. 65. Wanjale S. D. and Jog J. P., ‘Poly(4-methyl-1-pentene) nanocomposites: effect of organically modified layered silicates’, Polym Int, 2003, 53, 101–105. 66. Wanjale S. D. and Jog J. P., ‘Poly(1-butene)/clay nanocomposites: preparation and properties’, J Polym Sci Part B: Polym Phys, 2003, 41, 1014–1021. 67. Wanjale S. D. and Jog J. P., ‘Poly(1-butene)/clay nanocomposites: a crystallization study’, J Macromol Sci, 2003, B42, 1141–1152. 68. Arroyo M., Suarez R. V., Herrero B. and Lopez-Manchado M. A., ‘Optimization of nanocomposites based on polypropylene/polyethylene blends and organo-bentonite’, J Mater Chem, 2003, 13, 2915–2921. 69. Chuang T.-H., Guo W., Cheng K.-C., Chen S.-W., Wang H.-T. and Yen Y.-Y., ‘Thermal properties and flammability of ethylene–vinyl acetate copolymer/montmorillonite/ polyethylene nanocomposites with flame retardants’, J Polym Res, 2004, 11, 169– 174. 70. Mishra J. K., Kim I. and Ha C., ‘Heat shrinkable behavior and mechanical response of a low-density polyethylene/millable polyurethane/organoclay ternary nanocomposite’, Macromol Rapid Commun, 2004, 25, 1851–1855. 71. Shah R. K., Hunter D. L. and Paul D. R., ‘Nanocomposites from poly(ethylene-comethacrylic acid) ionomers: effect of surfactant structure on morphology and properties’, Polymer, 2005, 46, 2646–2662.

© 2007, Woodhead Publishing Limited

14 Multiwall carbon nanotube–nylon-6 nanocomposites from polymerization Y. K. K I M and P. K. P A T R A, University of Massachusetts Dartmouth, USA

14.1

Introduction

Since their discovery in the early 1990s1 carbon nanotubes (CNTs) have excited scientists and engineers with their wide range of unusual physical properties. Typical features of the carbon nanotubes are: extremely small size (diameter from 1 to 40 nm and length around few micrometers), excellent mechanical properties (elastic modulus around 1 TPa, strength around 30 GPa and fracture strains about 20%), novel electric properties i.e. from perfect electric conductors (resistivity around 10–4 Ω cm and current density as high as 109 A/cm2) to semiconductors and high thermal conductivity.2–10 To take advantage of this unique combination of size and properties, a wide variety of applications have been proposed for carbon nanotubes, including: chemical and genetic probes, batteries, fuel cells, fibers, cables, pharmaceutics and biomedical materials, field emission tips, mechanical memory, supersensitive sensors, hydrogen and ion storage, scanning probe microscope tips and structural materials.11 Owing to small production quantities in the past, research to date has concentrated on small-volume applications (electronics, sensors, etc.). However, worldwide development in synthesis of large-scale single wall carbon nanotubes (SWNTs) and multiwall carbon nanotubes (MWNTs), has encouraged research in the use of carbon nanotubes in composites. Carbon nanotube-based composites have the potential to revolutionize structural materials for aerospace, electrical and thermal conductors for energy applications, nano-biotechnology and other disciplines.3 Nanotubes are a very attractive reinforcing material for polymer composite structures. Their extremely high aspect ratio and their amazingly high modulus and strength are the main reasons for so many studies that are conducted in the polymer–nanotubes composites field. An extremely high elastic modulus and strain to failure coupled with a tensile strength an order of magnitude higher than conventional carbon fibers, qualify carbon nanotubes especially as the ultimate reinforcement in polymer composite materials. 386

© 2007, Woodhead Publishing Limited

Multiwall carbon nanotube–nylon-6 nanocomposites

387

There are however three main problems in successful preparation of carbon nanotube-based nanocomposites. These are dispersion, orientation of carbon nanotubes and adhesion between carbon nanotubes and matrix.12 In this research work, MWNT–nylon-6 nanocomposites were prepared using the in situ polymerization technique (ring opening polymerization of ε-caprolactam to form nylon-6 polymer in the presence of carbon nanotubes), assisted by a few minutes (4 min) of ultrasonication. Micrometer-sized fibers were then extruded from prepared nanocomposite using a single screw extruder. The prepared fibers were characterized for the dispersion and orientation of carbon nanotubes using scanning electron microscopy (SEM), non-isothermal crystallization studies using differential scanning calorimetry (DSC) and mechanical properties using Instron. The dilute solution viscometry and simple theoretical models of filled polymers were used to approximately determine the effect of the MWNTs on the molecular weight of the prepared nylon-6 polymer.

14.2

Nanocomposite synthesis and production

14.2.1 Materials • • • • •

ε-Caprolactam (Fisher Scientific). Sodium hydride (NaH), 60% dispersion in oil (Fisher Scientific). Polyoxyethylene (POE), MW = 6000 g/moles (Fisher Scientific). N-acetylcaprolactam ( Fisher Scientific). Multiwall carbon nanotubes (0.5% and 1% on weight of polymer) (Catalytic Materials, Holliston, MA).

14.2.2 Equipment • Glass beaker. • Balance. • Cole Parmer Ultrasonic Processor 750 Watts.

14.2.3 Synthesis procedure • ε-Caprolactam (40 g), polyoxyethylene (0.88 g) carbon nanotubes (0.5%, 1% on weight of polymer) and N-acetylcaprolactam (20 drops) were placed in a beaker. • The mixture in the beaker was heated slowly using a Bunsen burner. • The molten solution was subjected to 4 min ultrasonication (Cole Parmer Ultrasonic Processor 750 W. Parameters – 80% amplitude, 5 s on, 5 s off) to break the agglomerations of carbon nanotubes. • NaH (0.15 g) was added when the mixture in the tube melted.

© 2007, Woodhead Publishing Limited

388

Nanofibers and nanotechnology in textiles

• Heating continued for 4–6 min until the reaction mixture became much more viscous. • Thin films of prepared nanocomposite were made using a hot press (230 °C and 30 s pressing time). Thin films were cut into pieces and washed with hot water to remove any unreacted species present in the polymer.

14.2.4 Carbon nanotube/nylon-6 composite fiber production The cut pieces of the nylon polymer and composites were placed in a Cole Parmer vacuum oven at 70 °C for 24 h to remove water and moisture contents. Neat nylon-6 and nanocomposite fibers were prepared using a Brabender single screw extruder (Intelli-torque) and single hole fiber die (diameter = 0.016 inches (0.4 mm), L/D ratio = 4) with the following parameters: • • • •

Temperature zone 1 = 250 °C. Temperature zone 2 = 230 °C. Temperature zone 3 = 230 °C. Screw speed = 3.

The extruded fibers were allowed to fall freely without any tension and then wound on bobbins. The extruded fibers were stretched with draw ratios 3 and 4 using Instron fiber clamps at room temperature. Stretched samples were used for further characterization.

14.3

Characterization techniques

14.3.1 Scanning electron microscopy analysis SEM studies were done on the cross-sectional surfaces of the broken fibers to study the dispersion and orientation of the carbon nanotubes in the fibers. The fibers were clamped vertically using Scotch tape and adhesive onto SEM stubs. The stubs and the samples were coated with gold using a Denton Vacuum Desk-II sputtering machine. The cross-sectional surfaces of the broken fibers were observed under different magnification levels using a JEOL JSM-5610 high vacuum scanning electron microscope. Intrinsic viscosities of nylon-6 polymer and nanocomposites were determined using dilute solution viscometry (ASTM D2857-95). Efflux times for 40% sulfuric acid and different concentrations of nylon-6 polymer and nanocomposite solutions were determined using a Ubbelohde-type viscometer. Four concentrations were prepared for each sample. The efflux time values were used further to determine the various terms listed in Table. 14.1, which shows the definition of the terms used in dilute solution viscometry, where

© 2007, Woodhead Publishing Limited

Multiwall carbon nanotube–nylon-6 nanocomposites

389

Table 14.1 Definition of different terms used in dilute solution viscometry Common name

Name proposed by IUPAC

Symbol and definition

Relative viscosity

Viscosity ratio

η rel =

ηsp

Specific viscosity

η = t η0 t0 = ηrel – 1

ηred = ηsp /c

Reduced viscosity

Viscosity number

Inherent viscosity

Logarithmic viscosity number

ηinh = ln(ηrel)/c

Intrinsic viscosity

Limiting viscosity number

[η] = lim (ηred) c →0

t0 and t are the efflux times for solvent and polymer solution respectively and c is concentration of solution in g/dl or g/ml. Viscosity average molecular weight for nylon-6 polymer was determined using the intrinsic viscosity and the Mark–Houwink equation. [η] KMa

[14.1]

where [η] = intrinsic viscosity a and K = Mark–Houwink constants M = polymer molecular weight a = 0.69 and K = 59.2 × 10–03 ml/g for nylon-6 and 40% H2SO4.13 In order to determine the molecular weight of nylon-6 polymer in the nanocomposite, it was assumed that the viscosity of nanocomposite solution had been altered by the carbon nanotubes according to the theory of dilute Brownian rods.14, 15 The intrinsic viscosity of nylon-6 polymer in the nanocomposite was determined by excluding the effect of carbon nanotubes:14, 15 2( L / R ) 2 η – ηs = ηsφ 45[ln ( L / R )]

  1 + 0.64   ln ( L / R )  1.659 +  1.5  (ln ( L / R )) 2   1 – ln ( L / R )  

    

[14.2]

where η = viscosity of solution ηs = viscosity of suspension L = length of rods (fillers) R = radius of rods φ = volume fraction of rods Viscosity-average molecular weight for nylon-6 polymer and the nanocomposite was then determined using the Mark–Houwink equation.

© 2007, Woodhead Publishing Limited

390

Nanofibers and nanotechnology in textiles

14.3.2 Differential scanning calerimetry and nonisothermal crystallization kinetics studies DSC analysis was performed on the nylon-6 and the nanocomposite samples to understand the effect of the carbon nanotubes on the melting behavior and the percentage crystallinity. The values of melting point and heat of melting were measured during a second heating cycle. Percentage crystallinity was determined using Equation 14.3: % Crystallinity =

∆H s × 100 ∆H c

[14.3]

where ∆Hs = heat of fusion for sample (J/g) ∆Hc = heat of fusion for 100% crystalline nylon-6 polymer (J/g) = 230 J/g13 Non-isothermal crystallization kinetics for nylon-6 and nanocomposite samples were studied. The onset of the crystallization, crystallization temperature and the heat of the crystallization were noted during the cooling cycle. Using Universal Analysis software, running integration function was performed for crystallization peak in heat flow (J/g) vs. time (minutes) plot to get the values of relative crystallinity Xt for different values of time t. The time required for the sample to crystallize 50% (t1/2) was noted for the nylon6 and nanocomposite samples. Then the chart of ln [–ln (1 – Xt)] vs. ln t was plotted. The Avrami equation was used to determine values of n and k from the slopes and the interception of the best fit (Equation 14.4). ln [–ln (1 – Xt)] = n ln t + ln k

[14.4]

where k = growth rate parameter of Avrami equation n = nucleation parameter of Avrami equation The following heat–cool–heat sequence was used in DSC and non-isothermal crystallization studies: 1. 2. 3. 4. 5.

Equilibrate at 0 °C. Heat to 240 °C at 20 °C/min. Isothermal at 240 °C for 5 min. Cool to 0 °C at –40 °C/min. Heat to 240 °C at 20 °C/min.

14.3.3 Tensile testing Before mechanical testing, the diameter of each Instron sample was measured using a Leica DMRX optical microscope and micrometer scale under 200× magnification with transmission mode illumination. Three readings per sample (middle and at two ends) were taken and an average value was determined.

© 2007, Woodhead Publishing Limited

Multiwall carbon nanotube–nylon-6 nanocomposites

391

Mechanical properties were measured using an Instron tensile tester and ASTM standard test procedures – D3822-01. • Sample conditioning – samples were kept for 24 h in testing laboratory for moisture equilibrium. • Gauge length – 5 cm. • Sample preparation – single filament was secured to two cardboard pieces with Quicktite superglue (cynoacrylate-based adhesive by Locktite). • Extension rate – 60% of initial specimen length/min, or 3 cm/min. • Sample size – 15 per batch. The following properties were measured: • initial modulus (MPa); • breaking strength (MPa); • breaking strain (%). Fifteen samples from different portions of the extruded fibers were used for tensile testing. The linear curve fit equation was determined for the initial part of the stress–strain curve (strain 0–3%) for each sample using Microsoft Excel. The slope of the equation gives the initial modulus. Breaking strength and breaking strain were calculated at the point of maximum load.

14.4

Properties of multiwall carbon nanotube–nylon6 nanocomposite fibers

During in situ polymerization we observed that polymerization became much more difficult with increasing wt% of MWNTs. With increasing amount of wt% MWNTs, the time required to build sufficient viscosity increased. For example it took around 2–3 min to polymerize (i.e. attain sufficient viscosity) neat nylon-6, but with 0.5 wt% MWNTs, it took around 4–5 min to attain similar viscosity. This time was increased to about 5 min with 1 wt% MWNTs in the mixture. This shows that the rate of polymerization has been affected by the presence of carbon nanotubes. With the method used in this work, nanocomposites with more than 1 or 1.5 wt% MWNTs can not be prepared. Some possible reasons for this are as follows: • Because of the significantly reduced polymerization rate, the time required to build sufficient viscosity increases. As time goes on, the active species (the sodium salt of caprolactam) in the formulation react with moisture in the air, resulting in termination of the growing chains. This results in very low molecular weight nylon-6 polymer. • With increasing time, the temperature of the formulation goes very high, which results in degradation of the contents.

© 2007, Woodhead Publishing Limited

392

Nanofibers and nanotechnology in textiles

The very high specific area of the MWNTs plays a major role in reducing the polymerization rate. Assuming a density of about 1.5 g/cm3 and length around 4 µm, MWNTs having a diameter of 30 nm give a surface area of 78 m2/g. Figure 14.1 shows the SEM image of the cross-section of a broken nanocomposite fiber (0.5 wt% MWNTs). Well-separated MWNTs were observed throughout the cross-section. Carbon nanotubes appear as bright tiny dots in the SEM images due to charging. Figures 14.2 and 14.3 are some additional images of the cross-section and show the well-separated carbon nanotubes. As carbon nanotubes appear as tiny, near-round dots in the crosssection of the fiber it is likely that they are nearly oriented along the direction of the fiber.

14.4.1 Mechanism of dispersion Viscosity effect It is now known that ultrasonication energy can efficiently disperse the carbon nanotubes or nanoparticles in the solutions having low viscosity.16 In the experiments, the carbon nanotubes were added to a molten monomer solution (ε-caprolactam) having low viscosity, i.e. before polymerization. So a small amount of ultrasonication energy and time are enough to break the agglomerations and disperse carbon nanotubes effectively.

14.1 Dispersion of MWNTs, SEM image of cross-section of broken fibers (0.5 wt% MWNTs) magnification ×3500.

© 2007, Woodhead Publishing Limited

Multiwall carbon nanotube–nylon-6 nanocomposites

393

14.2 Dispersion of MWNTs, SEM image of cross-section of broken fibers (0.5 wt% MWNTs).

14.3 Dispersion of MWNTs, SEM image of cross-section of broken fibers (1 wt% MWNTs).

© 2007, Woodhead Publishing Limited

394

Nanofibers and nanotechnology in textiles

Time effect Re-agglomeration of the carbon nanotubes in the solution is a time-dependent phenomenon.17 Ultrasonically separated particles tend to agglomerate again after some time. This phenomenon depends upon solvent–nanoparticle interaction, chemical treatment given to nanoparticles (functionalization), presence of dispersing agent or surfactant, viscosity and concentration of particles in solvent. In these experiments, polymerization occurs within a few minutes (2–5 min) after ultrasonication, so polymer chains initiate and propagate between well-separated carbon nanotubes, which lowers the cohesive force of attraction between carbon nanotubes that binds the nanoparticles. This can lead to efficient separation and dispersion of carbon nanotubes. Once the polymerization occurs, because of high viscosity, there is a little chance of getting particles re-agglomerated in further processing such as during fiber extrusion. Chain formation effect In the in situ technique, owing to the random nature of chain propagation and growth of polymeric molecules, the chances are very high that chains will form between individual carbon nanotubes. Use of ultrasonication can also encourage monomer molecules to go very close to the surfaces of a carbon nanotube. It is possible to wrap carbon nanotubes by the polymeric chain, depending on the affinity between the carbon nanotubes and monomer solution. It is also quite possible that monomer ε-caprolactam can penetrate and polymerize inside MWNTs, as it has an inner diameter of about 3–4 nm,16 which is much larger than molecular size of ε-caprolactam (few angstroms). The mechanism of dispersion is explained in Fig. 14.4. Differential scanning calorimetry analysis Figure 14.5 shows the DSC curves for the neat nylon-6 and nanocomposite samples (replica 1). The results show that there is no significant effect of the carbon nanotubes on the melting temperature of the composite; however, Monomer molecules

Nanoparticle

Polymer molecule

14.4 Dispersion mechanism in in situ polymerization.

© 2007, Woodhead Publishing Limited

Multiwall carbon nanotube–nylon-6 nanocomposites

395

0.0

Heat flow [W/g]

–0.5

–1.0

–1.5

216.52 °C 213.24 °C

–2.0

0.5% CNT 1% CNT Neat nylon

–2.5 120

140

216.49 °C 160

Exo Up

180 Temperature [°C]

200

220

240

Universal V4.1D TA instruments

14.5 DSC melting curves for neat nylon-6 and nanocomposites. Table 14.2 Melting characteristics and crystallinity of neat nylon and nanocomposites MWNT (wt%)

Melting point (°C)

Heat of fusion (J/g)

Crystallinity (%)

0 0.5 1

216 217 214

52 60 69

23 26 30

nanocomposites show higher crystallinity than neat nylon-6. An increase in crystallinity of 15% was observed for 0.5 wt% MWNT and 32% for 1 wt% MWNT nanocomposite (Table 14.2). It is important to note that neat nylon6 shows a narrow melting peak, whereas nanocomposites show a broad melting peak. Figure 14.5 also shows a big difference in onset of melting. The broad peak and early onset of melting suggest the presence of smaller and defective crystals in the nanocomposites. It is quite possible that MWNTs obstruct or interfere with the growth of crystals around them and makes them smaller, defective or both. The increase in crystallinity can be attributed to the reduced molecular weight, which was observed in dilute solution viscometry experiments. Shorter chains can align more easily during the formation and growth of the crystals. Neat nylon-6 shows a small pre-melting peak which is indication of re-crystallization or reorganization of nylon-6 molecules. Such peaks were absent in the case of nanocomposite samples.

© 2007, Woodhead Publishing Limited

396

Nanofibers and nanotechnology in textiles

Table 14.2 shows the melting characteristics (average) of neat nylon-6 and nanocomposite samples.

14.4.2 Melting characteristics and crystallization The crystallization exotherms of neat nylon-6 and nanocomposite samples (replica 1) are presented in Fig. 14.6. All the samples were subjected to the same heating profile. Crystallization in nanocomposites starts earlier (about 1 min earlier) or at higher temperature (about 25 °C higher) compared with neat nylon-6 samples. This indicates that carbon nanotubes act as nucleating agents at the early stages of crystallization. Table 14.3 shows the average crystallization characteristics of neat nylon6 and nanocomposite samples. The results show that carbon nanotubes

5 0.5% CNT 1% CNT Neat nylon

171.26 °C 4

Heat flow [W/g]

185.10 °C 181.52 °C

3

15.75 min 214.44 °C

2

1

16.49 min 184.92 °C

0 100

120

140

160 180 Temperature [°C]

Exo Up

15.90 min 208.97 °C

200

220

240

Universal V4.1D TA instruments

14.6 DSC crystallization curves for neat nylon-6 and nanocomposites. Table 14.3 Crystallization behavior of neat nylon-6 and nanocomposite sample

Neat nylon 0.5 wt% MWNT 1 wt% MWNT

© 2007, Woodhead Publishing Limited

Crystallization start time (min)

Crystallization onset temperature (°C)

Heat of crystallization (J/g)

Crystallization peak temperature (°C)

16.49 15.91 15.79

185 208 213

60 56 62

170 182 185

Multiwall carbon nanotube–nylon-6 nanocomposites

397

significantly alter the crystallization behavior of the nylon-6 polymer. Nanocomposite samples show higher heat of crystallization values. The shift of around 12–15 °C for crystallization peak can be observed for the nanocomposite fibers. Integration of the exothermic peak the non-isothermal crystallization can give the relative degree of crystallinity as function time. Figure 14.7 shows the normalized graph of relative crystallinity % vs. time for nylon-6 and nanocomposite samples (replica 1). Though nucleation occurs earlier the crystals in nanocomposites grow very slowly compared with the neat nylon6 sample. The Avrami equation was used to analyze the non-isothermal crystallization process.18 Figure 14.8 shows the plot of ln [–ln(1 – Xt)] vs. ln t to describe the non-isothermal crystallization process. Non-isothermal crystallization parameters, Avrami exponent n and rate constant k, were determined from the slope and the interception of the best fit. Table 14.4 shows the values of half-crystallization time and Avrami exponent n and rate constant k. The values of n around 1.4–1.5 suggest the rod-shaped crystal geometry and thermal nucleation type.17 Results show that rate constant k has been significantly changed due to inclusion of the carbon nanotubes. This suggests that carbon nanotubes restrict or interfere with the growth of the crystals that are around the vicinity of the carbon nanotubes.

14.4.3 Molecular weight Neat nylon-6 and nanocomposite samples were characterized using dilute solution viscometry to understand the effect of carbon nanotubes on the

100

Realtive crystallinity [%]

90 80 70 60 50 40 30 Neat nylon 0.5 wt% MWNT 1 wt% MWNT

20 10 0 0

0.5

1 Time [min]

1.5

14.7 Effect of carbon nanotubes on crystal growth.

© 2007, Woodhead Publishing Limited

2

398

Nanofibers and nanotechnology in textiles 3 (Neat nylon) y = 1.5537x + 1.1045 (0.5% CNT) y = 1.5184x + 0.1458 (1% CNT) y = 1.4652x + 0.0306

2 1

ln [– ln (1 – Xt)]

0 –1 –2 –3 –4 –5 0.5% CNT Neat nylon 1% CNT

–6 –7 –8 –5

–4

–3

–2 ln t [min]

–1

0

1

14.8 Determination of Avrami parameters. Table 14.4 Avrami parameters for neat nylon-6 and nanocomposite

Neat nylon 0.5 wt% MWNT 1 wt% MWNT

t1/2 Crystallization half-time (min)

K (min–1)

n

0.420 0.760 0.825

1.054 0.176 0.039

1.548 1.497 1.454

Avrami parameters

molecular weight of the nylon-6 polymer synthesized. Figure 14.9 shows the plots of the reduced and inherent viscosities vs. concentrations. The observation of the graphs shows that nanocomposite samples, even though they were expected to show higher intrinsic viscosities because of the presence of carbon nanotubes, have lower intrinsic viscosities. Therefore it can be concluded that the nanocomposites have lower molecular weights than the neat nylon6 polymer. Equation 14.2 was used to determine the intrinsic viscosities for the nanocomposite samples excluding the effect of the carbon nanotubes on the viscosity. The length of the carbon nanotubes was assumed to be 4 µm. The volume fraction for the 0.5 wt% and 1 wt% MWNTs on weight of polymers was determined to be 0.004 and 0.008. The diameter of the used MWNTs was 30 nm. Table 14.5 shows the average intrinsic viscosities for neat nylon-6 and nanocomposite samples after excluding the effect of carbon nanotubes on viscosity. The molecular weights for neat nylon-6 polymer and nanocomposite samples were then calculated using the Mark–Houwink equation (Table 14.6). The

© 2007, Woodhead Publishing Limited

Neat nylon

NC 0.5 70

70

y = 1147.8x + 51.423

y = 886.7x + 54.786

65

Viscosity [ml/g]

60

Reduced viscosity Inherent viscosity

55 50

y = –537.55x + 54.593

45

60

Reduced viscosity Inherent viscosity

55 50 45

40

y = –306.26x + 51.816

40 0

0.002

0.004 0.006 0.008 Concentration [g/ml]

0.01

0.012

0

0.002

0.004 0.006 0.008 Concentration [g/ml]

NC 1 65

y = 1053.1x + 49.711 Viscosity [ml/g]

60 55

Reduced viscosity Inherent viscosity

50 45

y = –277.44x + 49.936

40 0

0.002

0.004 0.006 0.008 Concentration [g/ml]

0.01

0.012

0.012

399

14.9 Determination of the intrinsic viscosity for neat nylon and nanocomposite samples.

© 2007, Woodhead Publishing Limited

0.01

Multiwall carbon nanotube–nylon-6 nanocomposites

Viscosity [ml/g]

65

400

Nanofibers and nanotechnology in textiles Table 14.5 Intrinsic viscosities for the neat nylon and nanocomposite samples after application of Equation 14.2 Intrinsic viscosity (ml/g)

Neat nylon NC 0.5 NC 1

Replica 1

Replica 2

Average

54 51 48

55 52 50

54 51 49

Table 14.6 Determination of molecular weights using the Mark–Houwink equation

Neat nylon NC 0.5 NC 1

Intrinsic viscosity (ml/g)

Mv (g/mol)

54 49 45

19 700 17 100 15 100

25000

Mv [g/mol]

20000

15000

10000

5000

0 Neat nylon

NC 0.5

NC 1

14.10 Viscosity average molecular weights for neat nylon-6 polymer and nanocomposite samples.

viscosity average molecular weight (Mv) was significantly reduced by the addition of carbon nanotubes; Mv decreased by about 13% with 0.5 wt% addition of MWNTs and by about 23% with the 1 wt% addition of MWNTs. Figure 14.10 graphically represents the affect of carbon nanotubes on the molecular weight. Results of dilute solution viscometry thus confirm the conclusions made in Section 14.2. The carbon nanotubes lower the rate of

© 2007, Woodhead Publishing Limited

Multiwall carbon nanotube–nylon-6 nanocomposites

401

polymerization. This decreased polymerization rate results in activation of more and more initiator molecules resulting in lowering of the molecular weight.

14.4.4 Tensile properties Table 14.7 shows average mechanical properties for nylon-6 and nanocomposite fibers. It also shows the standard deviation for the modulus, strength and breaking strain values. All the samples showed high variability in properties. Variability was higher in the case of nanocomposite samples. The reasons behind the high variability can be attributed to the absence of deaeration, mixing of the purging compound and localized MWNT agglomerates in the case of nanocomposites. The initial moduli for neat nylon-6 and nanocomposite samples are compared in Fig. 14.11. The initial modulus increased with draw ratio for all the samples. Increase in initial modulus was observed for nanocomposite samples compared with neat nylon samples. The nanocomposite sample with 0.5 wt% MWNTs stretched four times shows the highest initial modulus. Nanocomposite samples with 1 wt% MWNTs show lower initial modulus values compared with samples with 0.5 wt% MWNT but they show higher values than those of the neat nylon-6 sample. This is a result of the lowered molecular weight in the case of 1 wt% MWNTs nanocomposite compared with 0.5 wt% MWNTs nanocomposite. The breaking strengths of neat nylon-6 and nanocomposite samples are compared in Fig. 14.12. Increase in the strength was observed for the nanocomposite samples. The highest strength value was observed for the

Table 14.7 Average mechanical properties for nylon-6 and nanocomposite fibers Neat nylon DR-3

Neat nylon DR-4

NC 0.5 DR-3

NC 0.5 DR-4

NC 1 DR-3

NC 1 DR-4

1149

1302

1553

1837

1353

1641

96

112

189

258

118

169

205

224

261

327

226

277

SD strength (MPa)

12

10

29

30

20

34

Breaking strain (%)

33

24

39

23

38

17

6

5

7

4

7

3

Modulus (MPa) SD modulus (MPa) Strength (MPa)

SD breaking strain (%)

DR draw ratio, NC 0.5 nanocomposite with 0.5 wt% MWNTs, NC 1 nanocomposite with 1 wt% MWNTs, SD standard deviation.

© 2007, Woodhead Publishing Limited

402

Nanofibers and nanotechnology in textiles

2000 1800

Modulus [MPa]

1600 1400 1200 1000 800 600 400 200 0 Neat nylon DR-3

Neat nylon NC 0.5 DR-3 NC 0.5 DR-4 DR-4

NC 1 DR-3

NC 1 DR-4

14.11 Comparison of moduli of nylon-6 and nanocomposite samples. 350

Strength [MPa]

300 250 200 150 100 50 0 Neat nylon DR-3

Neat nylon NC 0.5 DR-3 NC 0.5 DR-4 DR-4

NC 1 DR-3

NC 1 DR-4

14.12 Comparison of strengths of nylon-6 and nanocomposite samples.

nanocomposite sample with 0.5 wt% MWNTs stretched four times; however, nanocomposite samples with 1 wt% MWNTs show lower strength values compared with samples with 0.5 wt% MWNT but higher values than those of the neat nylon-6 sample. This is a result of the lowered molecular weight in the case of 1 wt% MWNTs nanocomposite compared with 0.5 wt% MWNTs nanocomposite. A somewhat unusual trend can be observed for the strain-atbreak values for neat nylon-6 and nanocomposite samples (Fig. 14.13). Higher breaking strain values were observed in the case of nanocomposite samples that were stretched three times. But after stretching four times, breaking

© 2007, Woodhead Publishing Limited

Multiwall carbon nanotube–nylon-6 nanocomposites

403

45 40 Strain at break [%]

35 30 25 20 15 10 5 0 Neat nylon DR-3

Neat nylon DR-4

NC 0.5 DR-3 NC 0.5 DR-4

NC 1 DR-3

NC 1 DR-4

14.13 Comparison of strains at break of nylon-6 and nanocomposite samples. Table 14.8 Percentage change in mechanical properties of nanocomposite samples compared with neat nylon-6 samples Change (%) NC 0.5 DR-3 Modulus

Neat nylon DR-3 Neat nylon DR-4

+35

Strength

Neat nylon DR-3 Neat nylon DR-4

+27

Strain % at break

Neat nylon DR-3 Neat nylon DR-4

+19

NC 0.5 DR-4

NC 1 DR-3

NC 1 DR-4

+18 +41

+26 +10

+46

+24 +14

–7

–31

strain values go down drastically and are lower than those of neat nylon-6 samples stretched four times. Table 14.8 shows the percentage change in mechanical properties due to reinforcement of the carbon nanotubes in neat nylon-6. The maximum increase of 41% and 46% in modulus and strength respectively was observed in the case of nanocomposites with 0.5 wt% MWNT stretched four times. Less improvement can be observed in the case of nanocomposites with 1 wt% MWNT, likely due to the lower molecular weight of the nylon. Table 14.9 shows the effect of drawing on the percentage change in mechanical properties of the neat nylon-6 and nanocomposite samples. Results show that percentage improvements are higher in nanocomposite samples compared with neat nylon-6 samples when samples were further stretched from three to four times. For example, when neat nylon-6 samples were

© 2007, Woodhead Publishing Limited

404

Nanofibers and nanotechnology in textiles

Table 14.9 Effect of draw ratio on percentage change in mechanical properties of nanocomposite and neat nylon-6 samples Change (%) Neat nylon DR-4 Modulus

Strength

Strain % at break

Neat nylon DR-3 NC 0.5 DR-3 NC 1 DR-3

+13

Neat nylon DR-3 NC 0.5 DR-3 NC 1 DR-3

+9

Neat nylon DR-3 NC 0.5 DR-3 NC 1 DR-3

–27

NC 0.5 DR-4

NC 1 DR-4

+18 +21 +25 +23 –43 –56

Table 14.10 Change in mechanical properties of nanocomposite samples (0.5 wt% and 1 wt% MWNT) Change (%) NC 1 DR-3 Modulus Strength Strain % at break

NC 0.5 DR-3 NC 0.5 DR-4

–13

NC 0.5 DR-3 NC 0.5 DR-4

–14

NC 0.5 DR-3 NC 0.5 DR-4

–4

NC 1 DR-4

–11 –15 –26

further stretched from three to four times, modulus increased by 13%. On the other hand, 18 and 21% improvements can be observed in the case of 0.5 wt% and 1 wt% MWNT respectively. Similar results can be observed in the case of strength values. This must be a result of orientation of carbon nanotubes along the fiber axis when samples were further stretched from three to four times. Table 14.10 shows the comparison of mechanical properties between nanocomposite samples with 0.5 wt% MWNT and 1 wt% MWNT. All the properties decreased with further addition of carbon nanotubes. This can be possibly be attributed to the reduction of molecular weight in the case of the 1 wt% MWNT nanocomposite sample compared with the 0.5 wt% MWNT sample.

14.5

Conclusions

Nylon-6/MWNT nanocomposites were prepared by an in situ polymerization technique. Different studies were done on the nylon-6/MWNT nanocomposites

© 2007, Woodhead Publishing Limited

Multiwall carbon nanotube–nylon-6 nanocomposites

405

and extruded nanocomposite fibers and the following conclusions were made: • Nanocomposites with more than 1 or 1.5 wt% MWNTs could not be prepared by the proposed technique due to reduced polymerization rate. If the reaction were carried out for a longer time then polymerization failed either due to the reaction of the monomer mixture with moisture in the air or degradation of the chemicals due to heating for the longer duration. • SEM analysis of the cross-section of the stretched and broken nanocomposite fibers showed that carbon nanotubes were well separated in the polymer matrix. Carbon nanotubes also appeared to be nearly oriented along the direction of the fiber axis due to the effect of the extrusion and stretching. • Crystallization kinetics study using DSC indicated that although carbon nanotubes acted as a nucleating agent at the initial stage, they significantly lowered the crystallization rate or crystal growth. • Dilute solution viscometry study showed that nylon-6 polymer in the nanocomposites has a lower molecular weight than that of neat nylon-6 polymer polymerized with same concentration of the initiator. Viscosity average molecular weight decreased from 19 700 to 17 100 g/mol for 0.5 wt% addition of the MWNTs and to 15 100 g/mol for 1 wt% addition of the MWNTs due to restricted chain growth. • Tensile testing of neat nylon-6 and nanocomposite fibers showed that reinforcement of nylon-6 fiber by carbon nanotubes increased the modulus and the strength of the fibers significantly (95% confidence level). Compared with neat nylon fibers, maximum improvements (41% in modulus and 46% in strength) were observed for the 0.5 wt% MWNT nanocomposite fibers that were stretched four times. But improvements were reduced to 26% in modulus and 23% in strength for the 1 wt% MWNT nanocomposite fibers, stretched four times, possibly due to the reduced molecular weight (23%). The results further showed that additional stretching of the fibers from three to four times, possibly further oriented the carbon nanotubes and resulted in additional improvements in modulus and strength.

14.6

Acknowledgments

The authors would like to express their gratitude to Mr Leo Barish for his help and discussions in microscopy. Thanks are given to Professor Steve Warner for his suggestions. The authors acknowledge Hyperion Catalysis for providing us with multiwall nanotubes and Honeywell for supplying the nylon-6. Finally authors are grateful to the National Textile Center (NTC) for funding this project under Department of Commerce grant no. 0207400.

© 2007, Woodhead Publishing Limited

406

14.7

Nanofibers and nanotechnology in textiles

References

1. Iijima, S., Nature, London, 56, pp 354, 1991 2. Satio, R., Dresselhaus, G., Dresselhaus, M., Physical Properties of the Carbon Nanotubes, Imperial College Press, London, 1998 3. Harris, P., Carbon Nanotubes and Related Structures: New Materials for the 21st Century, Cambridge University Press, New York, 1999 4. Pan, Z., Xie, S., Lu, L., Chang, B., Sun, L., Applied Physics Letters, 74, pp 3152– 3154, 1999 5. Qian, D., Dickey, E., Andrews, R., Applied Physics Letters, 76, pp 2868–2870, 2000 6. Thess, A., Lee, R., Nikolaev, P., Dia, H., Petit, P., Science, 273, pp 483–487, 1996 7. Journet, C., Maser, W., Bernier, P., Loiseau, A., De La Chapelle, M., Nature, 388, pp 756–758, 1997 8. Tans, S.J., et al., Nature, 393, p 49, 1998 9. Wei, B.Q., et al., Applied Physics Letters, 79, p 1172, 2001 10. Hone, J., Topics In Applied Physics, 80, p 273, 2001 11. Collins, P. Avouris, Ph., Scientific American, No. 12, p 62, 2000 12. Ajayan, P., Schadler, L., Giannaris, C., Rubio, A., Advanced Materials, 12(10), pp 750–753, 2000 13. Brandrup, J., Immergut, E.H., Polymer Handbook, 3rd edn, Wiley, New York, 1989 14. Kirkwood, J.G., Auer, P.L., Journal of Chemical Physics, 19, pp 281–283, 1951 15. Davis, V.A., et al., Macromolecules, 37, pp 154–160, 2004 16. Sandler J., Shaffer, M.S.P., Prasse, T., Bauhofer, W., Schulte, K., Polymer, 40, pp 5967–5971, 1999 17. Hiemenz, P.C., Polymer Chemistry: The Basic Concepts, Marcel Dekker, New York, 1984 18. Li, Y., Zhang, G., Zhu, X., Yan, D., Journal of Applied Polymer Science, 88, pp 1311–1319, 2000

© 2007, Woodhead Publishing Limited

Part IV Nanocoatings and surface modification techniques

407

© 2007, Woodhead Publishing Limited

15 Nanotechnologies for coating and structuring of textiles T. S T E G M A I E R, M. D A U N E R, V. V O N A R N I M, A. S C H E R R I E B L E, A. D I N K E L M A N N and H. P L A N C K, ITV Denkendorf, Germany

15.1

Introduction

Fiber-based materials in technical applications are an increasing worldwide market and cover a wide area. Interesting applications for fibers with nanoscaled dimensions or nanostructured surfaces are in mobility (vehicles, aircraft, fuel cells), gas and liquid filtration, fiber reinforced materials (composites), protection and professional protective clothing. Tailoring and controlling of structures on a nano-scale level are considered to be key factors for the development of advanced materials or structural components and multifunctional applications. Nanotechnology is considered to increase the number, variety and effectivity of physical properties (electrical conductivity, magnetic susceptibility, interaction with light, photonics, corrosion protection, friction control, abrasion resistance, water and oil repellence, soil release, biocompatibility) of existing products and therefore act as an innovative base for new products. The dimensional aspect for the use of the ‘nanotechnology’ term is not clearly defined. It is commonly the control over a structural range from a few nanometers up to 50 and 100 nanometers that defines ‘nanotechnology’. In some cases the upper limit of the considered scale is regarded to be a few hundred nanometers or even 1 µm. In our research on nano-structured textiles we usually consider structural elements smaller than 100 nm. However, because ‘nano’ fibers with diameters of 100–500 nm are of great technological interest and need innovative fiber spinning techniques, these fibers are often included in the investigations of nanofibers. There are many ways to implement nano-scale controlled properties into textiles. One is to give the fiber itself a nano-scale by fiber spinning. With novel fiber spinning technologies it is possible to spin fibers with diameters between 20 and 500 nm; 10–500 times thinner than fibers spinnable by traditional fiber spinning techniques. An aim is to make the production of fibers with diameters below 100 nm highly productive and state-of-the-art. In order to achieve effects from nanostructures in the fiber bulk either 409

© 2007, Woodhead Publishing Limited

410

Nanofibers and nanotechnology in textiles

nanoparticle-filled polymer melts (e.g. pigments, TiO2, ZnO, clay) or nanophase-separating polymeric systems (e.g. elastane) can be spun. Nanofillers such as clay have been known for decades, but the processing of those filled polymeric melts is challenging because agglomeration has to be avoided and the influence of nanoparticles on rheological properties can be great.

15.2

Production of nanofiber nonwovens using electrostatic spinning

Porosity and pore size are crucial properties of filter media, which determine efficiency as well as pressure drop and permeability. Small pore sizes at high porosity of a textile filter medium depend on the fiber size. A reduction of the pore size below the fiber diameter greatly reduces the porosity and diminishes the permeability, which means the filtration efficiency increases with a reduction in the fiber diameter. The demand for filter media with high filter efficiency in the sub-micrometer range is increasing. The demand is based on the need for the filtration of aerosols and of industrially more and more important nanoparticles as well as on the requests for effective barrier effects against bacteria (<0.3 µm), viruses and other microorganisms. The diameter of natural as well as of synthetic fibers usually ranges from 10 to 20 µm. Microfibers and bi-component split fibers allow 3–7 µm. Melt blow and flash spinning end up with 1 µm fiber diameters.1 Below that, in the sub-micrometer range, glass fibers are produced, but should not be used for many filtration applications. Figure 15.1 shows the relation between fiber diameter and the resulting fiber surface: a reduction from microfibers (10 µm) to nanofibers (100 nm) increases the fiber surface in a textile formation with a weight of 300 g/m2 from 100 m2 to 10 000 m2.

15.2.1 Electrostatic spinning The basic design of electrostatic spinning and its realization in most research laboratories is very simple (Fig. 15.2). A polymer solution is fed to a nozzle by a defined low pressure. An electrical field is applied between the nozzle and a collector (rotating mandrel or a conveyor belt). The application of the voltage on the nozzle or the carrier is not important initially. The application of positive or negative charges is mandatory. Independent of the polymer, solvent and the concentration of the solution, a voltage of 1 kV per centimeter between the nozzle and the collector of the spun fibers is useful. The nozzle diameter is usually about 100 µm. Based on video recordings the mechanism leading to nanofibers was described at first as a multiple splaying similar to lightning.2 Ten times

© 2007, Woodhead Publishing Limited

Fiber surface [m2/m2]

Fiber diameter [µm]

100 10 1 0.1 0.01 1E-05 0.0001 0.001 0.01 0.1 Fiber count [dtex]

1

100 000 10 000 1 000 100

10

15.1 Increase of fiber surface through reduction of fiber diameter.

10 1 1E-05 0.0001 0.001 0.01 0.1 Fiber count [dtex]

1

10

Nanotechnologies for coating and structuring of textiles

Influence on fiber count, PET, area weight = 300 g/m2

Influence of fiber count PET

411

© 2007, Woodhead Publishing Limited

412

Nanofibers and nanotechnology in textiles Collector Polymer-jet P

Polymer reservoir

High voltage 1kV/cm

15.2 Apparatus for the electrostatic spinning, ‘Whipping’ mechanism from Ref. 3.

splaying of a 1 µm fiber leads to fibrils of about 0.3 µm in diameter. It can be easily calculated that an original single polymer solution jet needs to split into some 100 distinct jets in order to achieve nanofibers of 100 nm diameter. (In addition, the solvent extraction has to be taken in account.) In 2000, Reneker and Chun published the scientific proof for another mechanism, so-called ‘whipping’, which was discovered thanks to an improved video-resolution: after leaving the nozzle the primary fiber stays stable on the way to the carrier as long as surface tension, electrical charging and external influences (such as friction in air) stay in equilibrium. Any perturbation leads to a deviation of the fiber until a new equilibrium is reached. At constant margins (feeding and winding speed) the fiber becomes stretched. Here again a draw ratio of 100 can be easily achieved, reducing the diameter from 1 to 0.1 µm.2

15.2.2 Polymers and solvents The requirements on the polymer are comparable to other fiber-forming processes from solutions. Amorphous polymers result in regular nonwoven structures by use of the electrospinning process. Yet, according to our experiments and verified by literature data, nano-fibers could be produced only from crystallite-forming polymers. A further requirement for electrostatic spinning is that the polymer should be polar. Regular structures in the micrometer scale were produced from different polyurethanes and from copolyesters (e.g. Fig. 15.3 and 15.4). Microporous surfaces have been produced by the electrospinning process using solutions from polylactides.4 Using solutions from polyethyleneoxide (PEO), polyvinylalcohol (PVA) and polyacrylonitrile (PAN), as well as from polyimide, fibers in the nanometer scale were produced. The fiber diameter can be measured only by scanning electron microscope (SEM).

© 2007, Woodhead Publishing Limited

Nanotechnologies for coating and structuring of textiles

413

15.3 PVA in water, 1 kV/cm.

15.4 Polyurethane, CHCl3/MeOH, 1 kV/cm.

As with the polymers, a high polarity of the solvent improves the electrostatic process. Further a high electrical conductivity of the solvent is required for the production of nano-fibers.5 PVA and PEO are favorites for experiments as they can be processed from aqueous solutions and no special safety considerations have to be made regarding toxicity or explosivity. In addition, water is highly polar and most suitable for electrostatic spinning. (Their use for filters is limited, of course.) These are the main requirements on the solvents. In practice these ideal conditions will rarely be found. The use of chlorinated organic solvents is accepted when only small amounts are processed (e.g. for medical applications). It is difficult to obtain an operation permit for

© 2007, Woodhead Publishing Limited

414

Nanofibers and nanotechnology in textiles

large-scale production. Using inflammable organic solvents such as hydrocarbons carries risks of explosion, and safety concerns are substantial. The vapor pressure of the solvent is a minor concern for processing by electrostatic spinning. Dimethylacetamide (DMAC) and dimethylformamide (DMF) with a low vapor pressure respectively can be very useful for the spinning of certain polymers. But the complete removal of the solvents may be a problem. On the other hand a high vapor pressure may cause problems due to the early evaporation of the respective solvent. The appropriate concentration of the solutions depends of course on the polymer, its molecular mass and the solvent. Fibers do not form at very low concentrations. High concentrations hinder the feeding of the solution and its filtration. Good results are obtained at 5–20 vol.% with the best results in between.

15.2.3 Production methods Investments of tens of millions euros per year are made worldwide for the development of electrostatic spinning. The United States, Korea and Japan are particularly heavy investors. Within Europe, Germany is one of the leading countries in R & D on electrospinning. At INDEX 2005 in Geneva, Switzerland, the Czech start-up company Elmarco, Liberec presented a technology for the productive spinning of nano-fiber webs. This technology, which avoids the use of needle tips or nozzles, has been filed for patents by other companies before, yet is without commercial use. The process is based on a rotating roll dipping in a bath of polymer solution. Roll and bath are electrically on ground. The high-voltage electrode is placed at some distance above the roll. Nano-fibers can be continuously collected on a web which runs in the space between the roll and the electrode.

15.2.4 Productivity A general problem of electrostatic spinning is the poor productivity. There are technologies that do not use a nozzle system – as reported in most of the publications from scientific laboratories – but convey the polymer solution by distribution on a surface like a rotating rod. Productivity is measured by the polymer volume delivered per time unit by a spinning pump. This works with the nozzle system only. Here the concentration of the solution must be considered, which is the proportion of the fiber-forming polymer and the delivered solution. It directly influences the viscosity of the solution. Measuring the throughput of the polymer solution is currently the only way to determine the productivity of an electrospinning process in-line. The production speed (m2/min) at constant area weight of the nanofiber web as a measure for

© 2007, Woodhead Publishing Limited

Nanotechnologies for coating and structuring of textiles

415

productivity is needed. However, because of the extremely low weight of the nanofiber webs the electrospun web is difficult to weigh in-line. The first suggestion derived from basic research to technical products was to multiply the nozzles. The practical distance of the nozzles at which the meta-stable rotating fibers do not touch each other has been determined in our trials to be 5 mm. The linear order technical production lines will have an array of nozzles. Clearly the production speed depends on the nozzle numbers. Yet the effort needed to deliver a homogeneous solution to the nozzles and to clean them is increased as well. The approach to improve the nano-fiber formation by using fine capillary holes (100 µm) was found to be impractical, because the solvent evaporated too fast, which resulted in obstruction of the capillaries. An additional problem was caused by the parallel order of the capillaries. If they are closed, they will not be opened by a low processing pressure; the polymer solutions only run through open holes. This results in an inhomogeneous fiber diameter distribution and an unstable fiber manufacturing process with large maintenance efforts. One solution may be the cyclic automatic cleaning of the capillaries, but this greatly increases the cost of the equipment. Thus a diameter of the capillaries of 500 µm has been chosen, which is easier to handle in the production and processing stages. An L/D ratio of 20:1 enhances the orientation of the molecules, yet here as well problems of production occur after about 20–30 min. Even with a 50 capillary nozzle a spin pump (0.6 cm3/min) could not be used because of the extremely low throughput at which fibers were formed. The throughput could be increased and the spinning pump could run at its lowest turns by supporting the fiber formation and the evaporation of the solvent by an air stream blown concentrically around each capillary. The delivery is best made by gear pumps, which guarantees a constant volume flow of the conveyed solution independent of the pressure. The volume flow acts as a process parameter to determine the area weight produced. The alternative state-of-the-art process to produce fine nonwovens is the melt-blow process. A comparison of melt-blow and electrospinning with respect to productivity is therefore of interest. In order to produce 2 µm thin fibers by the melt-blow process using 1250 capillaries per meter, a throughput of 3000 cm3/m h is typical. An electrospinning set-up with 500 mm width and 50 capillaries in one row delivers about 1 cm3/m h polymer. In order to achieve the same throughput as the melt-blow example, the electrospinning set-up needs to be scaled up to an array of over 10 000 capillaries in 12–13 rows, with 100 capillaries per meter working width.

15.2.5 Centrifuge spinning Like electrostatic spinning, centrifuge spinning per se is not a new technology. Glass to sub-micrometer fibers, pitch to carbon fibers, melamin to the Basofil

© 2007, Woodhead Publishing Limited

416

Nanofibers and nanotechnology in textiles

fibers are some examples in which the technology has been used or at least tested. Yet no literature is known reporting the production of polymeric fibers below 1 µm. In cooperation with Reiter, Winnenden, Germany, ITV Denkendorf has established equipment having as the core unit a high-speed rotor up to 50 000 rpm driven by pressurized air. The polymer solution is delivered via the hollow axis. By centrifugal forces the solution is accelerated and sprayed. An air stream helps to bundle the fiber cone. Here electrostatic charging is not used for fiber formation but for fiber collection to a web. Depending on the process parameters one rotor covers about 330 mm. Thus, three rotors are required at least per 1 m working width. Based upon the equipment in planning a polymer volume of 500 cm3/m h can be processed to fibers in the range of 0.3 µm. Taking into account a linear dependency of productivity to the fiber diameter, the productivity of centrifuge spinning meets that of melt-blown.

15.2.6 Comparing technologies For laboratory use, electrostatic spinning using manifolds is superior in its simplicity. However, scaling up for production use is impaired by its low productivity, which can be increased only through considerable mechanical effort. Only with a controlled mass transport per capillary can the formation of the desired fine fibers be ensured. Automatic cleaning means are required to guarantee open capillaries. The use of an air stream can increase the productivity, but is limited by fiber fineness and diameter distribution. The centrifuge spinning requires a highly developed technology with more than 40 000 rpm. Reiter developed this many years ago for varnishing. The productivity is comparably high. The delivery of the polymer solution by a spin pump as well as the distribution of the fibers on the width of the substrate are important for a homogeneous fiber web. Electrostatic spinning may use spin pumps with nozzle manifolds for technical applications, yet obstruction of the nozzles will disturb the homogeneity. Homogeneous fiber distribution over the length and the width, as well as the effective area weight, requires an extensive means of control. The common very low area weight of the nanofiber web of about 0.1 to 1 g/m2 should be considered. Centrifuge spinning allows the online determination of the area weight by one pump per rotor. The centrifugal technology by itself ensures the homogeneous distribution of the fibers. The area weight can be controlled by appropriate means. The question of how to control the laminate strength of the web with the substrate for filtration use is common to all technologies. In addition to engineering measures knowledge of the chemistry is required and must be elaborated for each substrate and each application. Also common to all these

© 2007, Woodhead Publishing Limited

Nanotechnologies for coating and structuring of textiles

417

technologies is the tendency already known from melt-blown: finer fibers can be achieved best by reduction of the polymer mass processed per time, i.e. to reduce the productivity. For polymer solutions the reduction of polymer mass can mean reduced throughput as well as reduction of the polymer concentration in the solution.

15.3

Anti-adhesive nanocoating of fibers and textiles

The textile finishing industry increasingly makes use of functional nanoparticles in order to achieve new or improved properties of textiles. Nanoparticles such as antimicrobial silver, photo-active TiO2, conductive or magnetic metals or metal oxides as well as UV-absorbing particles (ZnO) are utilized in textile functionalization.6 The interest in nanoparticles to functionalize textiles also results from the micro-structured nature of most textile products because the thickness of functional coatings on micro-scaled fibers should be in the region of submicrometers. Therefore, beside all new functions, the trend towards even thinner fiber creates a need for thin nano-scaled or nanostructured finishing coatings in order to maintain the small fiber diameter and to make use of its related properties. One often desired textile property of huge interest is liquid repellence. Although the finishing of textiles with oil- and water-repelling resinbased fluorocarbons (FC) is state-of-the art (Scotchguard® and others), there are still attempts to increase production efficiencies and to improve the product properties. Simplified, an FC finishing process in textile production can be considered as a nanocoating process. A typical thickness of an FC finishing coating layer on a microfiber (10 µm fiber diameter) is 50 nm. By decreasing the diameter of the fiber to 1 µm, but keeping textile weight and liquor pick-up in a padding machine/padder constant, the coating thickness on each fiber will decrease to 10 nm and lower. The theoretical models predict even an FC monolayer on sub-micrometer fibers. Figure 15.5 shows the calculated dependence between FC layer thickness and fiber diameter for a polyester textile, assuming typical finishing conditions. There are different approaches to improve the effects of an FC finishing of textiles. The approach discussed next, plasma technology, takes into account that it is a waste of energy to dry off 99% of the applied finishing fluid to leave an FC layer that is a few nanometers thin on the fiber. Another approach that will be discussed later is based on the roughness dependence of wetting and aims at the generation of ultra-water-repelling surfaces by nano-structured rough surfaces with low surface energy.

© 2007, Woodhead Publishing Limited

418

Nanofibers and nanotechnology in textiles Influence on fiber dimension, PET, FC finishing, concentration = 0.8%, wet pick up = 70% Layer thickness [nm]

100 10 1 0.1 0.01 0.00001 (30nm)

0.0001

0.001 (300 nm)

0.01

0.1 (3 µm)

1

10 (30 µm)

Fiber thickness [dtex (nm)]

15.5 Dependence between estimated fluorocarbon layer thickness and fiber thickness.

15.4

Water- and oil-repellent coatings by plasma treatment

Plasma-based modifications are dry processes and therefore an interesting alternative to the traditional wet textile finishing systems for economic reasons. They make use of gases whose reactivity has been raised by electrical discharges in strong electric fields between electrodes. Atmospheric pressure plasma systems can be integrated easily in continuously running textile production and finishing lines. If the discharge energy is sufficiently controlled and the gas temperature is kept in the range of room temperature, it is called cold or low-temperature plasma and the plasma treatment is generally applicable to nearly all kind of fibers. Further advantages of plasma treatments are modifications of surface properties without changing the properties of the fiber bulk. They are water-free processes with a minimum consumption of chemicals going along with elimination of energy-intensive drying processes. Therefore, plasma processes can be highly environmentally friendly processes. Plasma treatment changes properties such as friction coefficient, surface energy and antistatic behavior. The technological basis of the wide applicability of atmospheric pressure processes in the textile industry was the enhancement of the established corona technology by coating both electrodes by a dielectric material (dielectrical barrier discharge, DBD), using an intermittent (pulsed) voltage source as well as by feeding defined gas mixtures into the discharge (Fig. 15.6). If reactive gases that are able to polymerize after excitation in the discharge are fed into the plasma zone, thin coatings can be deposited on the substrate from a non-equilibrium plasma by, for example, radical polymerization. The morphology of the coating and the deposition rate are controlled by the reaction mechanism and reaction rate. Readily polymerizing systems

© 2007, Woodhead Publishing Limited

Nanotechnologies for coating and structuring of textiles

419

15.6 Dielectric barrier discharge (DBD) in air.

form particles within the discharge which can be deposited on the substrate. However, for textile treatments, these dust-forming plasmas are less relevant because the particles typically lie loosely on the surface. Of very great interest are polymerization processes that predominantly take place at the substrate surface by forming a permanent functional surface coating. For example, highly cross-linked layers with varying surface energies, depending on the chemical composition, can be deposited from non-equivalent plasmas. Plasma polymerization processes need, in general, an encapsulated plasma device to control the plasma atmosphere. However, a continuous roll to roll and air to air process is still possible if either the plasma process runs within ambient air or gas-locks avoids the entry of air into the reactor chamber. The generation of water- and oil-repellent functional layers on textiles by plasma polymerization of fluorocarbons at atmospheric pressure under continuous inline conditions has been a major focus of collaborative research projects at ITV Denkendorf.7, 8 The structures achieved with plasma chemical deposited FC layers are characterized by a nanometer-scaled thickness and relatively high degree of cross-linking. Plasma polymerized layers with FCs in DBD show surface energies of 11 mN/m on polymeric films. These values are significantly lower than the typical value of PTFE with 18 mN/m.9 In encapsulated continuously working plasma units (Fig. 15.7), oil repellency grades of 5–6 (according to AATCC 118-1992) have been obtained on PETMonofil fabric at a process speed of 0.5 m/min (Fig. 15.8). Better oil-repellent properties than those of PTFE were obtained, but the properties of waterbased FC finishing have not been completely achieved until now. Oil repellencies on treated fabrics increase with decreasing process speeds, increasing fluorocarbon layer thicknesses. The thickness of plasma polymerized

© 2007, Woodhead Publishing Limited

420

Nanofibers and nanotechnology in textiles

(a)

(b)

15.7 Encapsulated plasma unit for 1 m textile width.

FC layers on fibers of up to 14 nm was extrapolated from IR measurements on plasma-treated Si-wafers (Fig. 15.9). Deposition rates of up to 1 nm/s on fibers were achieved.

15.4.1 Aerosol and spraying applications The use of aerosols in plasma technology increases the application spectrum of suitable chemicals enormously. With the help of aerosols in atmospheric pressure liquid chemicals, solutions and, in a limited way, dispersions can be used in plasma for surface modification. The potential of combinations from aerosols and spraying application in the DBD for the surface treatment of textiles is in the first development stage. Examples for current and future

© 2007, Woodhead Publishing Limited

Nanotechnologies for coating and structuring of textiles

421

Oil repellence grade number

1

2

3

4

5

6

7

8

Oil repellence grade number

15.8 Demonstration of oil repellence of plasma-finished fabric. 8 7 6 5 4 3 2

PET

1

PA

0 0

2

4 6 8 10 12 Estimated thickness of FC layer [nm]

14

16

15.9 Dependence of oil repellence on plasma FC layer thickness.

applications are physical surface modification, e.g. generation of electret properties on filters,10 chemical functionalization, energy saving finishing, and chemical and topographical nanostructuring.

15.5

Self-cleaning superhydrophobic surfaces

New products and new properties of products can be developed by learning from principles and functions in nature. ITV Denkendorf is working in basic and applied science in networks of botanic institutes, chemical companies, textile producers and consumers in different fields on bionic ideas.11 One of the main focuses of current work with nanotechnology within these networks is the development of self-cleaning superhydrophobic surfaces on textiles.

15.5.1 Principles The characteristic property of self-cleaning or so-called Lotus-Effect® surfaces12 is the capacity of complete cleaning only by means of water, for example, in the form of rain. The attribute is often called the self-cleaning effect, as there

© 2007, Woodhead Publishing Limited

422

Nanofibers and nanotechnology in textiles

is no need for cleaning agents or additional mechanical support, beside the droplet momentum. This characteristic was discovered and investigated on natural surfaces of plants, such as leaf and blossom surfaces, but also on animal surfaces. The most famous and probably the most ideal representative from the plant world is the lotus plant that acts as the eponym. The main function of nano-structured superhydrophobic surfaces in nature is most likely protection against pathogenic organic contamination such as bacteria or spores.13 These contaminants are completely removed from the leaves by rainfall. The self-cleaning effect is based on low surface energy and the minimization of adhesion area to attaching agents by nano- and micro-scaled surface structures. SEM photographs show the superposed double structure of these organism surfaces. The results are extremely high contact angles of contacting water drops, rolling off at slight inclinations and removing attaching pollutions.

15.5.2 Transfer to fiber-based products There is a variety of applications for fiber-based surfaces with self-cleaning characteristics. This includes outdoor applications, such as textile roofs for airports and railways, sunscreen textiles, outdoor clothing, but also indoor applications, which come into contact with water or water-based solutions (Fig. 15.10).14 One of the specific features of textiles in this context is that they readily bring rough structures with at least two topological structure elements represented by the filament’s fiber arrangement within the yarn structure and the yarn arrangement within the fabric structure. Subsequent approaches to

15.10 Honey droplet on a fabric with self-cleaning surface characteristic.

© 2007, Woodhead Publishing Limited

Nanotechnologies for coating and structuring of textiles

423

implement the self-cleaning effect on textile-based surfaces also implement fiber surface modification to low surface energies, the optimization of fabric and yarn construction structures. The alteration of textile surface finishing chemicals to meet the above-mentioned requirements can, for example, consist of polymer-based dispersions with nanoparticle additives. Other products are organic–inorganic hybrid materials on the basis of sol–gel chemistry, eventually also with nano-filler additives. In order to be transferred to modern textile production lines, the finishing systems should be water-based. Processes to apply the chemicals consist of standard textile finishing processes such as padding, face padding or spraying. Another attempt that is being investigated at ITV Denkendorf is to modify the fiber surface coating by the yarn.15 This process is especially interesting if finishing processes after fabric construction are limited or for sewing thread. However, dyeing processes have to be applied before the yarn finishing, such as yarn dyeing or spin dyeing. The textile construction also plays a crucial role for the effects of selfcleaning. In two ITV studies that were financially supported by the German Federal Ministry of Research and Technology the influence of textile structure on superhydrophobicity and self-cleaning was investigated. In the first study, to analyze the general feasibility of the development of extremely self-cleaning textiles, the influence of construction parameters on woven fabrics made of filament yarn was investigated. Particularly low wettability is measured for woven fabrics with open yarn structure. These fabrics have distinct microstructured surfaces and show superhydrophobicity. High filament fineness supports hydrophobicity compared with yarns with thicker filaments, if the yarn is constructed with low compactness, so that the filaments lie side by side with an adequate distance between them. Aspect rates between 1 and 2 based on distance and height differences of adjacent filaments turn out to be especially favorable for high water repellence. In the study to analyze the influence of structure and arrangement of staple fibers and filaments in fabrics of different types of constructions (knitted goods, nonwovens, warp knit fabrics and woven fabrics), various textile parameters (fiber material, yarn spinning method, filament and fiber construction and surface modification by mechanical and chemical methods) were varied. The investigations show significant influence of hairiness of the sample surfaces on the water repellence. Long, distant fibers, especially if present in samples made of ring-spun yarn, hinder the small water droplets of approximately 2 mm in diameter from rolling off and therefore result in poor repellence compared with equivalent samples made of open-end yarn and Vortex-yarn. In contrast to that the self-cleaning behavior is not affected by long, distant fibers and therefore an influence of the spinning method is not detected. The repellence is proportional to the fiber density of short distant fibers as the contact area decreases in the same way as when the

© 2007, Woodhead Publishing Limited

424

Nanofibers and nanotechnology in textiles

drop sits on the fiber endings like on flock-coated textiles, as shown in Fig. 15.11. The same effect is observed in the opposite way with singeing or calendering, which results in smoother surfaces. On the other hand for such surfaces that have fewer undercutting structures the accessibility of dirt particles and therefore self-cleaning ability is enhanced. The implementation of nanodimensional structures on the fibre surfaces enhances superhydrophobicity and the self-cleaning effect. This is shown in Fig. 15.12.

(a)

(b)

15.11 Static wetting behavior of water drops on hydrophobic flock textiles with varied surface roughness roughness: (a) low surface roughness, (b) high surface roughness.

© 2007, Woodhead Publishing Limited

Nanotechnologies for coating and structuring of textiles

425

Residual contamination [%]

100 90

Product A (nanostructure + hydrophobic)

80

Product C (nanostructure + hydrophobic)

70

Product D (hydrophobic) Product F (hydrophobic)

60 50 40 30 20 10 0 0

1

2

3 4 5 6 7 Number of water drops

8

9

10

15.12 Residual contamination with soot particles after impact of water drops; multifilament fabric with nano-structured rough superhydrophobic coating (products A and C) and respectively smooth hydrophobic finishing (products D and F).

15.5.3 Testing methods Methods to test superhydrophobicity and self-cleaning have been newly developed or adapted for textile applications at ITV Denkendorf. These methods were developed to sensitively differentiate between conventional soil-repellent finished textile samples that have a smooth fiber surface on the one hand and textiles that are finished with products that impose nanodimensional structures on the fiber surface on the other hand. The rate of superhydrophobicity is measured by determining the so-called repellent power, which was invented by Dr Keller BASF, Ludwigshafen, Germany, via the determination of the dynamic roll-off angle. The static dynamic contact angle used for the characterization of even surfaces such as foils is applicable to textiles only in special cases. When dealing with micro-rough surfaces, especially where distant fibres dominate the surface structures, the contact angles cannot be measured satisfactorily with optical testing methods. The dynamic roll-off angle represents the boundary value at which a liquid droplet with a defined volume that is placed on the inclined sample surface from a defined height rolls off the sample. Correlations of roll-off angle and contact angles are given by Furmidge.16 Self-cleaning efficiency is measured via testing methods with dirt that is known from other textile testing standards and consists of mixtures of different components or of single component particles such as carbon black. Dirt mixtures that are used for testing consist exemplarily of silica, mineral oil, olive oil and carbon black. In the applied

© 2007, Woodhead Publishing Limited

426

Nanofibers and nanotechnology in textiles

test method the dirt is mechanically rubbed into the textiles’ surface to simulate strong impact. After contamination the sample is sprayed with water. Evaluation of remaining contamination is either done qualitatively by standardized rating in comparison to the gray scale according to the norm DIN EN 20105 A02/A03 or with quantitative methods in which the residual contaminants are detected and quantified at a microscope with image processing and subsequent particle detection software.

15.5.4 The Denkendorf quality mark In order to prove the superhydrophobic and self-cleaning effect of textile products, ITV issues the quality mark ‘self-cleaning – inspired by nature’ that makes use of the foregoing testing methods (Fig. 15.13). Additionally, in the testing procedure to this seal of approval the fiber’s filament surface is examined with an SEM to qualify the surface structures, underlining the prerequisite of nano-scaled structures for the self-cleaning effect. Recent research and development work in the field of nano-scaled surfaces on textiles at ITV Denkendorf aims strongly at the optimization of the mechanical abrasion resistance of such surfaces. The durability of these surfaces has to be measured depending on the application of the product. Awning fabric, tested in climate exposure test cabinets with high UV penetration for 1000 h and intermediate application-dependent mechanical load in the form of grinding, exhibited better self-cleaning behavior than awning fabric prepared with conventional finishing systems. The ability for recovery of impaired surfaces is an important feature of further development activities.

15.13 Quality mark for self-cleaning textiles.

© 2007, Woodhead Publishing Limited

Nanotechnologies for coating and structuring of textiles

15.6

427

Sources of further information and advice

www.lotus-effect.de www.selfcleaning.eu www.nanopartikel.info

15.7

References

1. Dauner, M., Production of nano fiber nonwovens using electrostatic spinning; 7th Symposium ‘Textile Filter’, Chemnitz, 2–3 March, 2004 2. Reneker, D. H., Chun, I., Nanometer diameter fibers of polymer, produced by electro spinning; Nanotechnology 7 (1995), 216–223 3. Reneker, D. H. et al., Bending instability of electrically charged liquid jets of polymer solutions in electro spinning; J. Appl. Phys. 87 (9), (2000), 4531–4547 4. Bognitzki, M. et al., Nanostructured fibers via electro spinning; Adv. Mater. 13 (1), (2001), 70–72 5. Böbel, J., Entwicklung von nanostrukturierten Oberflächen für das Tissue Engineering auf Basis des Elektrospinnens; Studienarbeit, DITF/Universität Stuttgart, 2003 6. Soane, D. S., et al., US Patent 6,607,994 B2 2003 7. Stegmaier, T., Arnim, V. V., Dinkelmann, A., Planck, H., Behandlung von laufenden Textilbahnen im Atmosphärendruckplasma, Melliand textilberichte 6 (2004), 476– 481 8. Arnim, V. V., Stegmaier, T., Praschak, D., Bahners, T., Lunk, A., et al., Continuous plasma treatment of textiles under atmospheric pressure; Proceedings of the 29th Aachen Textile Conference, 2002 9. Lunk, A., Vinogradov, I. P., Dinkelmann, A., Deposition of fluorocarbon polymer films in a dielectric barrier discharge(DBD); Surface Coatings Technol. 174–175 (2003), 509–514 10. Ernst, M., Stegmaier, T., Planck, H., Elektretcoatings for Nonwovens; 7th Symposium ‘Textile Filter’, Chemnitz, 2–3 March, 2004 11. Stegmaier, T., Milwich, M., Scherrieble, A., Geuer, M., Planck, H., Bionik developments based on textile materials for technical applications, Bionik 2004 – International Conference, Hannover Messe, 22–23 April, 2004 12. Barthlott, W., Self-cleaning surfaces of objects and process for producing same; Patent specification EP0772514B1, 1997 13. Barthlott, W., Neinhuis, C., Purity of the sacred lotus, or escape from contamination in biological surfaces; Planta 202 (1997), 1–8 14. Stegmaier, T., Dauner, M., Dinkelmann, A., Scherrieble, A., von Arnim, V., Schneider, P., Planck, H., Nanostructured fibers and coatings for technical textiles; Techn. Textiles 47 (2004), 142–146 15. Stegmaier, T., Abele, H., Ernst, M., Hager, T., Scherrieble, A., Schneider, P., Witt, M.-U., Wunderlich, W., Planck, H., Functionalisation of filaments and fibres by coatings; Techn. Textiles 48 (2005), 16–19 16. Furmidge, C. G. L., Studies at phase interfaces. I. The sliding of liquid drops on solid surfaces and a theory for spray retention, J. Colloid Sci. 17 (1962), 309–324

© 2007, Woodhead Publishing Limited

16 Electrostatic self-assembled nanolayer films for cotton fibers G. K. H Y D E and J. P. H I N E S T R O Z A, Cornell University, USA

16.1

Introduction

The use of multilayered polymeric films offers the possibility of creating multicomposite molecular assemblies with great levels of reproducibility and controlled molecular architectures. While these high levels of control have been reached in planar and homogeneous surfaces such as silicon, glass, gold and other synthetic materials, their realization in natural fibers has not been achieved. Natural fibers offer unique challenges as not only are their cross-sections irregular, but their surfaces are chemically and physically heterogeneous. However, having the ability to control the surface of a natural fiber offers great rewards that go far beyond pure economics as natural fibers are renewable and biodegradable resources.

16.2

Principles of electrostatic self-assembly for creating nanolayer films

The Langmuir–Blodget (LB) technique allows monolayers to be created by using a non-solvent as a substrate. Once the monolayers have formed over the surface of a non-solvent system, they can be transferred onto a solid support. This technique, pioneered in the 1960s, is believed to be the first synthetic nano-scale heterostructure. These LB experiments were the first true nanomanipulations and they offered unprecedented control of the deposition of individual molecular layers. The creation of LB films requires the use of specialized equipment that is often expensive and difficult to maintain, and it is not compatible with existing fiber manufacturing techniques. Furthermore, the technique is limited by the size and topology of the substrate, making it unfeasible for large-scale robust manufacturing.1 Starting in the early 1990s, Gero Decher’s group pioneered work on a robust method to create nanolayer structures using electrostatic self-assembly (ESA) principles.2–5 The use of electrostatic interactions was selected as it offered the least steric demand of all self-assembly methods. Decher’s initial 428

© 2007, Woodhead Publishing Limited

Electrostatic self-assembled nanolayer films for cotton fibers

429

work was focused on rod-like molecules containing ionic groups at both ends, polyelectrolytes and various other charged materials in aqueous solutions.1 The ESA process has greatly increased in popularity due to the method’s simplicity and the fact that not only polyelectrolytes but also almost any type of charged nanomoieties can be used to create the nanolayers in a controlled manner.6 Figure 16.1 illustrates the process. In this example, a positively charged substrate, cotton, adsorbs a polyanion. This adsorption step is followed by a rinsing or excess removal procedure. The coated substrate, which now possesses an outer layer of a polyanion, can now adsorb a polycation. A surface charge reversal occurs with each adsorption step, leading to the formation of a layered structure.1 The strong electrostatic attraction between charged surfaces and oppositely charged molecules in solution is believed to be the dominant factor in the adsorption of the polyelectrolytes.2, 7–10 In theory, the adsorption of molecules possessing more than one equal charge allows for charge reversal on the surface. This behavior implies that (1) equally charged molecules will be repulsed, allowing for adsorption selfregulation and restriction of the deposition to a single layer, and (2) an oppositely charged molecule can be adsorbed in a second step on top of the first one. Multilayer films may be composed of polyions, charged molecular objects and/or colloidal objects. In theory, there are no limitations with respect to

PSS

Cationized cotton

H2O

PAH

Cationized cotton + 1 layer PSS

H2O

Cationized cotton + 1 layer PSS, 1 layer PAH

NH3Cl SO3Na

16.1 Schematic detailing deposition of anionic poly(styrene sulfonate) (PSS) and cationic poly(allylamine hydrochloride) (PAH) on cationized cotton fabric.

© 2007, Woodhead Publishing Limited

430

Nanofibers and nanotechnology in textiles

substrate size and topology as most of the time the process involves adsorption from an aqueous solution. Nanolayer films have been created on objects of a variety of sizes.11–15 Although good adhesion of a layer to the base substrate requires a particular number of ionic bonds, the overcompensation of the surface charge by the incoming layer is found to be more dependent on the properties of the polymer than on those of the substrate. The use of polyelectrolytes allows for the formation of bridges over individual underlying defects. This unique feature of ESA enables the creation of self-healing structures as the conformation of the polymer over the surface of the substrate can be controlled by manipulating the adsorption operating conditions.1 Numerous studies have validated this observation by demonstrating a linear increase of film thickness with the number of deposited layers independent of the nature of the initial substrate.11, 16–20

16.2.1 Deposition conditions Nanolayer films are normally deposited using adsorbate concentrations of several milligrams per milliliter. While these concentrations are greater than those needed to reach a plateau in an adsorption isotherm, the excess prevents depletion of the solutions during the deposition of multilayered structures.1 Washing or excess removal steps are often used after each adsorption step. The rinsing step is aimed at avoiding cross-contamination with the next adsorption solution as well as to remove weakly adsorbed polymer layers, hence stabilizing the multilayer structure.18 Adsorption times per layer can range from minutes for polyelectrolytes to hours for certain colloids. The magnitude of the adsorption times is a process limited by mass transfer and hence depends on factors such as the molar mass and concentrations of the polyelectrolyte solutions as well as deposition conditions such as mixing.1, 18 Several factors influence the composition of the multilayer film as well as the characteristics of the individual layers. The thickness of each layer appears to be dependent on both the characteristics of the immediately underlying surface and the deposition conditions. The nature and density of charged groups, their local mobility, and the surface roughness also appear to have an influence. Operational factors such as concentration, adsorption time, ionic strength, temperature, rinsing time, dipping speed and drying time also influence the nanolayer thickness.21 For a given pair of strongly dissociated polycations and polyanions, the thickness of the nanolayers is proportional to the salt concentration in the solution. Self-assembly of charged nanoparticles to oppositely charged substrate surfaces is governed by adsorption and desorption equilibria. While the efficient adsorption of the layers is the main objective of each immersion step, preventing the desorption or the rearrangement of the deposited layers during the rinsing process is of equal importance. The optimization of the

© 2007, Woodhead Publishing Limited

Electrostatic self-assembled nanolayer films for cotton fibers

431

ESA process requires the selection of proper stabilizers and careful control of the deposition kinetics.6, 22, 23

16.3

Advantages and disadvantages of electrostatic self-assembly

The process of layer-by-layer adsorption exhibits several advantages over similar surface modification techniques aimed at the production of multilayered films. One of the main advantages is that film architecture is almost completely determined by the deposition conditions, making ESA a manufacturingamenable process. Another advantage of layer-by-layer adsorption is that a wide variety of different materials can be used to create multilayer thin films, hence creating multicomposite or hybrid films.1 Current examples of multicomposite films include structures that contain proteins, clay platelets, metals and gold colloids.1, 24–27 Despite the fact that ESA has become widely used in recent years, certain details of the process are still not clearly understood. For example, the existence of a minimum time required to complete the deposition process has not been fully explained from first principles. The dynamics of the intermediate washing and drying steps have not been well characterized either. A quantitative evaluation of the assembly process will be necessary to make ESA a practical commercial method.28 ESA is also influenced by a variety of factors that may be difficult to control, such as polymer entropy, charge transfer interactions and hydrogen bonding.29 No single theory has been developed to completely describe the deposition process. However, a variety of seminal studies have clarified many aspects of the ESA method.7, 30, 31 The ease of preparation and the high degree of versatility render selfassembled films useful to a large variety of applications. Self-assembled films can function as barriers, with controllable levels of permeability, for gases, liquids, covalent molecules, ions and electrons. These properties have been used for the construction of insulators, passivators, sensors and modified electrodes. Self-assembled nanolayers are also suitable for the construction of devices based on molecular recognition. Molecules or nanoparticles within a self-assembled layer can be aligned spontaneously, or by changing the temperature, pressure and pH, or by the application of external electric or magnetic fields. These characteristics allow for the formation of superlattices with a controllable architecture opening a new avenue for the development of a number of photonic, electronic, magnetic and non-linear optical devices. When insulators, conductors and magnetic, ferroelectric and semiconductor nanoparticulate films are deposited using the layer-by-layer process, hybrid heterostructures can be constructed with molecular precision. By manipulating the size and the interparticle distances of monodispersed nanoparticles within self-assembled films, novel optic devices may be

© 2007, Woodhead Publishing Limited

432

Nanofibers and nanotechnology in textiles

developed.32–35 Self-assembled nanolayers have been used to create polymer light-emitting diode devices with improved performance characteristics.36 Furthermore, controlling some of the solution parameters such as surface charges and pair combinations has opened the possibility of creating new light-sensitive materials and optical devices.37 The self-healing capability of ESA nanolayers provides an increased tolerance to defects. This self-healing ability sets the electrostatic method apart from other self-organization techniques.6 The electrostatic method can be used on substrates with non-uniform surfaces and compensates for defects caused during the adsorption process.28, 38

16.4

Substrates used for electrostatic self-assembly

Synthetic substrates such as glass, quartz, mica, gold, silver and a wide array of polymers have been extensively used as base substrates for ESA deposition.6, 19, 22, 39–42 Both hydrophilic (fluorine, glass and silicon) and hydrophobic (silanized glass) substrates have also been successfully used to support nanolayer thin films.43 The choice of substrates has often been determined by their convenience for different analysis techniques. Glass and quartz are used so the deposition can be monitored using UV-VIS spectroscopy and optical microscopy.6 Silicon wafers have been used for ellipsometric studies.25, 44 and the smooth surfaces of quartz, mica and glass are the preferred choices for X-ray reflectivity studies.19, 20, 45

16.4.1 Influence of substrate characteristics Owing to the characteristics of the layer-by-layer deposition technique, the adsorption of the polyelectrolytes is dependent on the surface charge of the substrate rather than its topology.46 However, during the initial deposition steps the topology may play a major role.4, 13, 47–50 The amount of adsorbed polymer and the chemical composition of the outermost layer normally exhibit larger variations during the initial deposition steps before reaching a plateau.13, 17, 33 The surface charge density of the substrate usually determines how many deposition steps are required to reach this steady state. However, constant growth is eventually reached despite the substrate characteristics as long as the polyelectrolytes complement each other, creating electrostatic equilibrium.12, 47, 51 Previous work has been able to elucidate the influence of surface charge on the deposition of ESA layers. For example, Fou and Rubner used microscopic glass slides with hydrophilic, hydrophobic, negatively charged and positively charged surfaces as substrates. The surface charge of the substrates was found to have a great influence on the deposition time, layer thickness and layer uniformity.52 However, the topology of the substrate itself was found to have a negligible effect on the adsorption of the individual layers.46

© 2007, Woodhead Publishing Limited

Electrostatic self-assembled nanolayer films for cotton fibers

433

16.4.2 Polymers as substrates for layer-by-layer deposition The majority of the early work involving the layer-by-layer process involved inorganic substrates such as quartz and silicone since it was thought that the process required flat clean surfaces.20, 26, 29 Later on polymer films and other organic materials were used as substrates.53, 54 Some of the most studied polymers include poly(propylene) (PP), poly(isobutylene) (PIB), poly(styrene) (PS), poly(methyl methacrylate) (PMMA), poly(ethylene terephthalate) (PET), poly(phenylene oxide) (PPO) and poly(ether imide) (PEI).13 Delcorte and others used surface analysis techniques to demonstrate that alternate polyelectrolyte thin films could be built up on polymeric substrates. PP, PIB, PS, PMMA, PET, PPO and PEI were evaluated as substrates. Semicrystalline PET as well as polymers containing carbonyl groups and/or benzene rings were identified as the most promising substrate choices.13 PET is of particular interest to the textile industry. It can be surface modified using a variety of techniques including plasma, corona discharge, ion beam, laser treatment, photo-initiated graft polymerization, saponification, aminolysis, reduction and entrapment of poly(ethylene oxide). PET is a suitable substrate for several reasons. It contains carbonyl groups that are capable of hydrogen bonding. The surface can be readily hydrolyzed to introduce carboxylic acid, as well as alcohol, it is able to support negative charges (PET- CO 2– ) in a sufficiently basic solution and the PET surface can react with polyamines to incorporate amine functionality capable of introducing positive charges (PET- NH 3+ ) in a non-basic solution.11 Chen and McCarthy conducted a study involving the modification of PET with layer-by-layer deposition. Poly(sodium styrenesulfonate) and poly(allylamine hydrochloride) were used as polyelectrolytes for surface modification. Contact angle analysis and X-ray photoelectron spectroscopy (XPS) were used to illustrate the structure of the outermost layers as well as the thickness of each individual layer.11 Furthermore, XPS and contact angle data indicated that the layers were stratified and that the wettability of the multilayer assemblies could be controlled by the identity of the outer- most polyelectrolyte layer. The individual layers were found to be extremely thin (2–6 Å) and it was demonstrated that the thickness of each layer could be controlled by adjusting the ionic strength of the polyelectrolyte solutions. The stoichiometry of the deposition process (ammonium ion:sulfonate ion ratio) was also affected by the substrate chemistry and solution ionic strength. This observation was particularly appealing as it indicated that the layer-bylayer deposition process was quite forgiving and could be done under a variety of conditions. Peel tests further showed that the multilayer assemblies deposited over the PET films exhibited good mechanical properties as no failures were observed in the multilayers.11

© 2007, Woodhead Publishing Limited

434

Nanofibers and nanotechnology in textiles

16.4.3 Surface modification techniques Surface modification techniques used to charge the substrates on which the nanolayers are to be deposited can be categorized as physical or chemical. Chemical modification techniques include surface patterning, photobleaching or plasma treatment.5, 12, 55, 56 Methods of physical modification primarily use Langmuir–Blodgett films or layers of charged polyelectrolytes as primers for the deposition of multilayers.4, 57–59 While these techniques are fully demonstrated on synthetic substrates, there is little literature on their use to modify the surface of natural fibers. Several attempts have been made to add functionality to cotton fibers. One of the most viable methods includes the creation of cationic sites by using controlled epoxy-based chemical reactions. It has been previously reported that reacting cotton with 2,3-epoxypropyltrimethylammonium chloride forms cationic charges on the surface of the fibers. While this process was originally developed to improve the affinity of cotton for anionic dyes, it has been recently used to provide the cotton fabric with a positive surface charge aimed at supporting polyelectrolyte nanolayers.60

16.5

Polyelectrolytes used for electrostatic selfassembly

Aqueous solutions of polyelectrolytes are commonly preferred for depositing layer-by-layer assemblies. However, organic solvents have proven to be useful as well.6 Since the ESA method is mainly based on the attraction of opposite charges, it is necessary that the layer-forming compounds have at least a minimal number of charged groups. Below this minimum charge, the layer-by-layer deposition process no longer works. Past studies have led to the belief that large hydrophobic fragments found in polyelectrolytes could be detrimental to the layer-by-layer deposition technique as they reduce their charge density and they can interfere with ion–ion interactions. Experiments involving weak polyelectrolytes support this theory.18, 61 These observations illustrated that a minimum charge level is required before the polyelectrolytes will engage in self-assembly patterns. However, recent studies have successfully used polyelectrolytes with very low charge densities by manipulating other types of molecular interactions capable of reducing the minimum charge required for the layers to be adsorbed.16, 47, 62–66

16.5.1 Synthetic polyelectrolytes A large number of synthetic polyelectrolytes have been used to create a variety of nanostructured thin film coatings.1, 23, 67–69 Some of these polyelectrolytes include poly(ethyleneimine), poly(allylamine),

© 2007, Woodhead Publishing Limited

Electrostatic self-assembled nanolayer films for cotton fibers

435

poly(diallyldimethylammonium chloride), poly(styrene sulfonate), poly(vinyl sulfate), poly(acrylic acid) and poly[N-vinyl-(4-(39-carboxy-49hydroxyphenylazo) benzene sulfonamide)].6 One of the most studied and well-understood systems consists of poly(allylamine) (PAA) and poly(styrene sulfonate) (PSS).11, 19, 70–72 A number of more complex, functionalized polyelectrolytes have also been used, based on their ability to form structured coatings, and whether or not they can enable secondary chemical modifications. One of the greatest advantages of the layer-by-layer deposition technique is that almost any polyelectrolyte can be used as long as the appropriate oppositely charged partner polyelectrolyte is chosen.6 A large number of functional polymers have also been studied, including electrical and ionic conducting and light-emitting polymers.14, 73–81 Past experiments have also used non-conjugated redox-active polymers, reactive polymers and polymeric complexes.82–91 Standard polyelectrolytes modified with small numbers of functional groups have also been used for labeling purposes and for molecular recognition studies.12, 92–94 Polyelectrolytes labeled with dyes and fluorescent probes have been used in an effort to better understand the adsorption of the layers as well.13, 33, 95–100 Complementary chromophores have also been used to monitor multilayer adsorption on real time via UV-VIS spectroscopy and colorimetric methods.6

16.5.2 Modified and natural polyelectrolytes Charged nano-objects, usually referred to as rigid polyelectrolytes, such as stable colloidal dispersions of charged silica, metal oxides, microcrystallites, and metal colloids have been deposited using the layer-by-layer (LbL) technique.25, 95, 101–110 Most of the deposition work involves the use of fully charged polyelectrolytes. PSS and poly(allylamine hydrochloride) (PAH) are examples of polyelectrolytes that have often been deposited at pH values less than 7.0. Recent studies have aimed at depositing multilayers composed of weak polyelectrolytes. Weak polyelectrolytes are attractive as their charge density can be controlled by adjusting the pH of the solutions.23, 111 In addition, weak polyelectrolytes such as PAA and PAH allow for a more precise control over the physical characteristics of the multilayers. Weak polyelectrolytes can be deposited with a high percentage of the chains making loops and tails under pH conditions of incomplete charge. This is in contrast to strong polyelectrolytes which often deposit as molecularly thin layers (about 5 Å). Layer thicknesses greater than 80 Å have been achieved when using weak polyelectrolyte solutions of PAA/PAH.23, 111 Natural polyelectrolytes such as nucleic acids, proteins and polysaccharides have also been used for LbL ESA. 112–114 Studies involving natural polyelectrolytes have aimed at gaining deeper understanding of the biological

© 2007, Woodhead Publishing Limited

436

Nanofibers and nanotechnology in textiles

functions of film and their ability to simulate biological processes. LbL does not require chemical modification and should in theory maintain normal protein behavior.115–117 Cyclic compounds, dendrimers and hyperbranched polyelectrolytes such as poly(ethyleneimine) have all been used with success, indicating the robustness of the LbL method.44, 118–120

16.6

Analyzing self-assembled nanolayer films on cotton

Since the introduction of the ESA method, a variety of techniques have been adapted to assess self-assembled nanolayers. Recent work by Akin and collaborators illustrates the use of the most common analysis techniques namely XPS, X-ray reflectivity and atomic force microscopy (AFM).121 Self-assembled nanolayer films have also been characterized using infrared (IR) spectroscopy, UV-VIS spectroscopy, ellipsometry, planar optical wave guide systems and quartz crystal microgravimetry (QCM).22 Surface plasmon resonance (SPR) measurements have also been used to characterize multilayer thin films adsorbed onto gold and other noble metal substrates. SPR monitors the reflectivity of an incident light beam from a thin film that is attached to a glass prism as a function of the incident angle and it can be several orders of magnitude more sensitive than QCM measurements.122 Several other techniques used include gel permeation chromatography, nuclear magnetic resonance (NMR) spectroscopy, and end group titration. However, IR and XPS appear to be the most commonly used as they can easily identify the chemical functionality of the end groups in the outermost layer.123 Recent work by our research group, Hyde et al.,124 has demonstrated that the LbL deposition of oppositely charged polyelectrolytes can be used as a method of surface modification for cotton fibers and fabrics. The cotton fibers were functionalized by reacting 2,3-epoxypropyltrimethylammonium chloride with the hydroxyl groups of cellulose to create cationic charges. These cationic charges were used to deposit subsequent layers of PSS and PAH. XPS and transmission electron microscopy (TEM) were used to characterize the multilayer structures over the cotton fibers. XPS spectra provided an analysis of the chemical groups present on the outermost layer of the samples. Specifically the presence of nitrogen and sulfur was monitored as it related to the deposition of PSS and PAH respectively. The ratio of N/S on the outermost layer was found to be in quantitative agreement with previously published work that deposited PSS and PAH on synthetic substrates. Figure 16.2 compares the XPS survey spectra for samples of cationically charged woven cotton fabric and fabric supporting layers of PSS and PAH. Sharp peaks can be observed at 281.91 eV for carbon and 528.91 eV for oxygen. A small amount of nitrogen, believed to be generated during the cationization process, was detected at 398.91 eV for the cationized fabric.

© 2007, Woodhead Publishing Limited

Electrostatic self-assembled nanolayer films for cotton fibers

O1s

Counts per second (arb units)

Cationized cotton Cotton with 20 layers PSS-PAH

437

C1s

N1s S2p

1200

1000

800

600 Binding energy (eV)

400

200

0

16.2 XPS survey spectra of cationized cotton and cotton supporting 20 alternating layers of PSS and PAH.

The figure also shows a survey spectrum of a 20-layer PSS/PAH multilayer film deposited on a woven cationic cotton substrate. It can be seen that the peaks at 398.91 and 164.91 eV have increased in magnitude in comparison with those present in the cationic cotton sample, indicating the presence of a multilayer film. The peaks correspond to nitrogen and sulfur respectively. TEM was used to obtain direct evidence of the presence of the layers and their ability to provide a fully conformal coating over the cotton fibers. Figure 16.3 shows a TEM image of a cotton fiber supporting 20 layers. The cuticle of the cotton fiber can be seen on the right side of the image. The multilayer film provides a uniform, conformal coating to the surface of the fiber with a thickness between 325 and 375 nm. Since 20 layers were deposited, it can be speculated that each layer may be 16–19 nm thick. Figure 16.4 provides an enhanced image of the outermost layer of the multilayer structure, confirming that each layer had a thickness around 20 nm. Our research group has also investigated the use of LbL deposition as a method of surface modification for cotton fibers and fabrics. The wool fibers were functionalized by using 2,3-epoxypropyltrimethylammonium to create cationic charges. These cationic charges were used to deposit subsequent layers of PSS and PAH.

© 2007, Woodhead Publishing Limited

438

Nanofibers and nanotechnology in textiles

400 nm

16.3 TEM image of cotton fiber with 20-layer film of PSS/PAH.

20 nm

16.4 TEM image of outermost layer in 20-layer film of PSS/PAH.

© 2007, Woodhead Publishing Limited

Electrostatic self-assembled nanolayer films for cotton fibers

439

8.0 2.0

Counts per second (arb units)

7.0 1.5

6.0 1.0

5.0 0.5

4.0 0.0 2

3.0

4

6

8 10 12 14 16 18 20 22

2.0 1.0 0.0 0

2

4

6

8 10 12 14 16 18 20 22 24 Number of layers

16.5 N/S evolution ratio for wool substrate supporting a 20-layer film of PSS/PAH. Three different specimens are presented, illustrating robustness and reproducibility of the ESA procedure. The inset provides a closer look at the N/S ratio at the outer layers.

XPS analysis was used to analyze the chemical groups present on the outermost layers of the wool samples. Specifically the presence of nitrogen and sulfur was monitored as it related to the deposition of PSS and PAH respectively. Figure 16.5 provides the N/S ratio for a wool sample supporting a 20-layer film of PSS/PAH. The variance in initial layers occurs because of the uneven nature of the wool substrate. As the number of layers increases, the variance levels out appropriately. The alternating trend seen is similar to that for the cotton substrates.

16.7

Conclusions: functional textiles for protection, filtration and other applications

Any number of different textile fibers and fabrics could possibly be used as substrates for the electrostatic self-assembly of nanolayers as far as they could hold charges on their surfaces. Possible candidates include polyamides as well as hemp, silk and many others. Several papers have been published recently detailing new methods of LbL assembly, highlighting reduced deposition times and improved layer uniformity. For example, Kim et al. devised a new technique for LbL deposition called dynamic LbL assembly. This process makes use of the basic LbL deposition process and fluidic devices to create well-defined multilayer polyelectrolyte films that can be quickly and easily fabricated on a specific

© 2007, Woodhead Publishing Limited

440

Nanofibers and nanotechnology in textiles

area of a substrate. Using this procedure multilayer films can be fabricated in just 90 s of processing time. The resulting films are similar in terms of film thickness and roughness when compared with samples produced by the conventional LbL dipping process.125 Porcel et al. developed a novel method that simultaneously sprays polyanion and polycation solutions onto a vertically oriented charged surface. This process creates a uniform film that grows with time. The vertical position leads to continuous drainage and helps removing any material that is not fixed on the surface of the outermost layer. This deposition technique, much like the conventional technique, does not include a drying step allowing for a continuous removal of any excess materials and improving the uniformity of the film.126 The ESA deposition process has been used to deposit alternate nanolayers of PSS and PAH on substrates of cotton and wool fabric. Treatment of the samples with 2,3-epoxypropyltrimethylammonium chloride was proven to be an effective procedure to create a substrate able to support multilayer thin films. XPS and TEM provided direct and indirect evidence of the efficacy of the deposition process. In addition, quantitative agreement of the XPS data with previously published data using several synthetic substrates corroborates that the LbL deposition process can be used as a method for the modification of textile fibers and fabrics. The experimental results also show that ESA is more dependent on the nature of the polyelectrolytes than that of the original substrate. LbL deposition is a process that could be used to potentially develop functional textiles for protective clothing and selective filtration applications. Using nanolayer films as a method of textile modification will allow increases in the functionality of a variety of textile products. It is also possible that ESA could be easily integrated into existing textile manufacturing processes.

16.8

References

1. Decher, G., Fuzzy nanoassemblies: toward layered polymeric multicomposites. Science, 1997. 277: p. 1232–1237. 2. Lvov, Y., Decher, G. and Moehwald, H., Assembly, structural characterization, and thermal behavior of layer-by-layer deposited ultrathin films of poly(vinyl sulfate) and poly(allylamine). Langmuir, 1993. 9: p. 481–486. 3. Lvov, Y., Decher, G. and Sukhorukov, G., Assembly of thin films by means of successive deposition of alternate layers of DNA and poly(allylamine). Macromolecules, 1993. 26: p. 5396–5399. 4. Lvov, Y., Essler, F. and Decher, G., Combination of polycation/polyanion selfassembly and Langmuir–Blodgett transfer for the construction of superlattice films. J Phys Chem, 1993. 97: p. 13773–13777. 5. Lvov, Y., et al., Assembly of polyelectrolyte molecular films onto plasma-treated glass. J Phys Chem, 1993. 97: p. 12835–12841.

© 2007, Woodhead Publishing Limited

Electrostatic self-assembled nanolayer films for cotton fibers

441

6. Bertrand, P., et al., Ultrathin polymer coatings by complexation of polyelectrolytes at interfaces: suitable materials, structure and properties. Macromol Rapid Commun, 2000. 21: p. 319–348. 7. Hammond, P., Recent explorations in electrostatic multilayer thin film assembly. Curr Opin Coll Inter Sci, 2000. 4: p. 430–442. 8. Phuvanartnuruks, V. and McCarthy, T., Stepwise polymer surface modification: chemistry-layer-by-layer deposition. Macromolecules, 1998. 31: p. 1906–1914. 9. Tieke, B., et al., Ultrathin self-assembled polyelectrolyte multilayer membranes. Eur Phys J E, 2001. 5: p. 29–39. 10. Ulman, A., Formation and structure of self-assembled monolayers. Chem Rev, 1996. 96: p. 1533–1554. 11. Chen, W. and McCarthy, T., Layer-by-layer deposition: A tool for polymer surface modification. Macromolecules, 1997. 30: p. 78–86. 12. Decher, G., et al., New nanocomposite films for biosensors – layer-by-layer adsorbed films of polyelectrolytes, proteins or DNA. Biosensors Bioelectronics, 1994. 9(9– 10): p. 677–684. 13. Delcorte, A., et al., Adsorption of polyelectrolyte multilayers on polymer surfaces. Langmuir, 1997. 13(19): p. 5125–5136. 14. Liang, Z.Q. and Wang, Q., Multilayer assembly and patterning of poly(pphenylenevinylene)s via covalent coupling reactions. Langmuir, 2004. 20(22): p. 9600–9606. 15. Poptoshev, E., Schoeler, B. and Caruso, F., Influence of solvent quality on the growth of polyelectrolyte multilayers. Langmuir, 2004. 20: p. 829–834. 16. Cooper, T.M., Campbell, A.L. and Crane, R.L. Formation of polypeptide–dye multilayers by an electrostatic self-assembly technique. Langmuir, 1995. 11(7): p. 2713–2718. 17. Dubas, S. and Schlenoff, J., Factors controlling the growth of polyelectrolyte multilayers. Macromolecules, 1999. 32: p. 8153–8160. 18. Hoogeveen, N.G., et al., Formation and stability of multilayers of polyelectrolytes. Langmuir, 1996. 12(15): p. 3675–3681. 19. Losche, M., et al., Detailed structure of molecularly thin polyelectrolyte multilayer films on solid substrates as revealed by neutron reflectometry. Macromolecules, 1998. 31: p. 8893–8906. 20. Schmitt, J., et al., Internal structure of layer-by-layer adsorbed polyelectrolyte films: a neutron and X-ray reflectivity study. Macromolecules, 1993. 26: p. 7058– 7063. 21. Decher, G., ‘Polyelectrolyte multilayers, an overview’, in G. Decher and J. Schlenoff, Editors. Multilayer Thin Films, 2003, Wiley-VCH. p. 1–46. 22. Caruso, F., et al., Ultrathin multilayer polyelectrolyte films on gold: construction and thickness determination. Langmuir, 1997. 13: p. 3422–3426. 23. Yoo, D., Shiratori, S. and Rubner, M., Controlling bilayer composition and surface wettability of sequentially adsorbed multilayers of weak polyelectrolytes. Macromolecules, 1998. 31: p. 4309–4318. 24. Feldheim, D.L., et al., Electron transfer in self-assembled inorganic polyelectrolyte/ metal nanoparticle heterostructures. J Amer Chem Soc, 1996. 118: p. 7640– 7641. 25. Keller, S.W., Kim, H. and Mallouk, T.E., Layer-by-layer assembly of intercalation compounds and heterostructures on surfaces: toward molecular ‘beaker’ epitaxy. J Amer Chem Soc, 1994. 116: p. 8817–8818.

© 2007, Woodhead Publishing Limited

442

Nanofibers and nanotechnology in textiles

26. Lvov, Y., et al., Assembly of multicomponent protein films by means of electrostatic layer-by-layer adsorption. J Amer Chem Soc, 1995. 117: p. 6117–6123. 27. Lvov, Y., et al., Formation of ultrathin multilayer and hydrated gel from montmorillonite and linear polycations. Langmuir, 1996. 12(12): p. 3038– 3044. 28. Lvov, Y., et al., A careful examination of the adsorption step in the alternate layerby-layer assembly of linear polyanion and polycation. Coll Surf A, 1999. 146: p. 337–346. 29. Delcorte, A., et al., ToF-SIMS study of alternate polyelectrolyte thin films: chemical surface characterization and molecular secondary ions sampling depth. Surface Sci, 1996. 366: p. 149–165. 30. Fogden, A. and Ninham, B.W., Electrostatics of curved fluid membranes: the interplay of direct interactions and fluctuations in charged lamellar phases. Adv Coll Int Sci, 1999. 83: p. 85–110. 31. Ninham, B.W., Kurihara, K. and Vinogradova, O.I., Hydrophobicity, specific ion adsorption and reactivity. Coll Surf A, 1997. 123–124: p. 7–12. 32. Advincula, R., Fells, E. and Park, M., Molecularly ordered low molecular weight azobenzene dyes and polycation alternate multilayer films: aggregation, layer order, and photoalignment. Chem Mater, 2001. 13: p. 2870–2878. 33. Decher, G., Lvov, Y. and Schmitt, J., Proof of multilayer structural organization in self-assembled polycation–polyanion molecular films. Thin Solid Films, 1994. 244: p. 772–777. 34. Fendler, J., Preparation and utilization of self-assembled ultrathin films composed of polyelectrolytes, nanoparticles and nanoplatelets. Croatica Chim Acta, 1998. 71: p. 1127–1137. 35. Kato, N., et al., Thin multilayer films of weak polyelectrolytes on colloid particles. Macromolecules, 2002. 35: p. 9780–9787. 36. Advincula, R., Supramolecular strategies using the layer-by-layer sequential assembly technique: applications for PLED and LC display devices and biosensors. IEICE Trans Electron, 2000. E83-C: p. 1104–1111. 37. Advincula, R., et al., Photo-induced alignment of polymer ultrathin films fabricated by alternate self-assembly solution adsorption of polyelectrolytes and small azo dye chromophores. ACS Polymer Preprints, 1999. Spring: p. 1–2. 38. Lvov, Y., et al., Thin film nanofabrication via layer-by-layer adsorption of tubule halloysite, spherical silica, proteins, and polycations. Coll Surf A, 2002. 198–200: p. 375–382. 39. Cassagneau, T. and Fendler, J. Preparation and layer-by-layer self-assembly of silver nanoparticles capped by graphite oxide nanosheets. J Phys Chem B, 1999. 103: p. 1789–1793. 40. Artyukhin, A.B., et al., Layer-by-layer electrostatic self-assembly of polyelectrolyte nanoshells on individual carbon nanotube templates. Langmuir, 2004. 20(4): p. 1442–1448. 41. Caruso, F., et al., Quartz crystal microbalance and surface plasmon resonance study of surfactant adsorption onto gold and chromium oxide surfaces. Langmuir, 1995. 11: p. 1546–1552. 42. Lu, C.H., et al., Au nanoparticle micropatterns prepared from self-assembled films. Langmuir, 2004. 20(3): p. 974–977. 43. GirardEgrot, A.P., Morelis, R.M. and Coulet, P.R., Direct influence of the interaction between the first layer and a hydrophilic substrate on the transition from Y- to Z-

© 2007, Woodhead Publishing Limited

Electrostatic self-assembled nanolayer films for cotton fibers

44. 45.

46.

47.

48. 49. 50.

51. 52.

53. 54.

55.

56.

57.

58.

59.

60. 61.

443

type transfer during deposition of phospholipid Langmuir–Blodgett films. Langmuir, 1996. 12(3): p. 778–783. Watanabe, S. and Regen, S.L., Dendrimers as building-blocks for multilayer construction. J Amer Chem Soc, 1994. 116(19): p. 8855–8856. Tronin, A., Lvov, Y. and Nicolini, C., Ellipsometry and x-ray reflectometry characterization of self-assembly process of polystyrenesulfonate and polyallylamine. Colloid Polym Sci, 1994. 272(10): p. 1317–1321. Sano, M., Lvov, Y. and Kunitake, T., Formation of ultrathin polymer layers on solid substrates by means of polymerization-induced epitaxy and alternate adsorption. Ann Rev Mat Sci, 1996. 26: p. 153–187. Lutt, M., Fitzsimmons, M.R. and Li, D., X-ray reflectivity study of self-assembled thin films of macrocycles and macromolecules. J Phys Chem B, 1998. 102: p. 400– 405. Bergbreiter, D.E., Franchina, J.G. and Kabza, K., Hyperbranched grafting on oxidized polyethylene surfaces. Macromolecules, 1999. 32(15): p. 4993–4998. Schlenoff, J.B., Ly, H. and Li, M., Charge and mass balance in polyelectrolyte multilayers. J Amer Chem Soc, 1998. 120(30): p. 7626–7634. Ferreira, M., Cheung, J.H. and Rubner, M.F., Molecular self-assembly of conjugated polyions – a new process for fabricating multilayer thin-film heterostructures. Thin Solid Films, 1994. 244(1–2): p. 806–809. Kleinfield, E. and Ferguson, G., Healing of defects in the stepwise formation of polymer/silicate multilayer films. Chem Mater, 1996. 8: p. 1575–1578. Fou, A.C. and Rubner, M., Molecular-level processing of conjugated polymers. 2. Layer-by-layer manipulation of in situ polymerized p-type doped conducting polymers. Macromolecules, 1995. 28: p. 7115–7120. Brynda, E. and Houska, M., Multiple alternating molecular layers of albumin and heparin on solid surfaces. J Colloid Interface Sci, 1996. 183(1): p. 18–25. Godinez, L.A., et al., Multilayer self-assembly of amphiphilic cyclodextrin hosts on bare and modified gold substrates: controlling aggregation via surface modification. Langmuir, 1998. 14(1): p. 137–144. Riccardi, C., et al., Surface modification of poly(ethylene terephthalate) fibers induced by radio frequency air plasma treatment. Appl Surface Sci, 2003. 211: p. 386–397. Hammond, P.T. and Whitesides, G.M., Formation of polymer microstructures by selective deposition of polyion multilayers using patterned self-assembled monolayers as a template. Macromolecules, 1995. 28(22): p. 7569–7571. Advincula, R. and Knoll, W., Supramolecular thin films via the Langmuir–Blodgett– Kuhn (LBK) technique. Colloids Surfaces A–Physicochem Eng Aspects, 1997. 123: p. 443–455. Advincula, R., et al., In situ investigations of polymer self-assembly solution adsorption by surface plasmon spectroscopy. Langmuir, 1996. 12(15): p. 3536– 3540. Gao, M.Y., et al., Constructing Pbi2 nanoparticles into a multilayer structure using the molecular deposition (Md) method. J Chem Soc–Chem Commun, 1994. (24): p. 2777–2778. Hauser, P.J. and Tabba, A.R., Improving the environmental and economic aspects of cotton dyeing using a cationised cotton. Coloration Technol, 2001. 117: p. 282–288. Kolarik, L., et al., Building assemblies from high molecular weight polyelectrolytes. Langmuir, 1999. 15: p. 8265–8275.

© 2007, Woodhead Publishing Limited

444

Nanofibers and nanotechnology in textiles

62. Cochin, D., et al., Layered nanostructures with LC-polymers, polyelectrolytes, and inorganics. Macromolecules, 1997. 30: p. 4775–4779. 63. Linford, M.R., Auch, M. and Mohwald, H., Nonmonotonic effect of ionic strength on surface dye extraction during dye–polyelectrolyte multilayer formation. J Amer Chem Soc, 1998. 120: p. 178–182. 64. Ariga, K., Lvov, Y. and Kunitake, T., Assembling alternate dye–polyion molecular films by electrostatic layer-by-layer adsorption. J Amer Chem Soc, 1997. 119: p. 2224–2231. 65. Li, D., et al., Preparation, characterization and properties of mixed organic and polymeric self-assembled multilayers. J Amer Chem Soc, 1998. 120: p. 8797– 8804. 66. Kurth, D.G. and Osterhout, R., In situ analysis of metallosupramolecular coordination polyelectrolyte films by surface plasmon resonance spectroscopy. Langmuir, 1999. 15: p. 4842–4846. 67. Onda, M., et al., Sequential actions of glucose oxidase and peroxidase in molecular films assembled by layer-by-layer alternate adsorption. Biotechnol Bioeng, 1996. 51(2): p. 163–167. 68. Onitsuka, O., et al., Enhancement of light emitting diodes based on self-assembled heterostructures of poly(p-phenylene vinylene). J Appl Phys, 1996. 80(7): p. 4067– 4071. 69. Wu, A., et al., Solid-state light-emitting devices based on the tris-chelated ruthenium(II) complex: 3. High efficiency devices via a layer-by-layer molecularlevel blending approach. J Amer Chem Soc, 1999. 121(20): p. 4883–4891. 70. Arys, X., et al., Ordered polyelectrolyte multilayers. Rules governing layering in organic binary multilayers. J Amer Chem Soc, 2003. 125(7): p. 1859–1865. 71. Arys, X., Laschewsky, A. and Jonas, A.M. Ordered polyelectrolyte ‘multilayers’. 1. Mechanisms of growth and structure formation: a comparison with classical fuzzy ‘multilayers’. Macromolecules, 2001. 34(10): p. 3318–3330. 72. Cho, J. and Char, K., Effect of layer integrity of spin self-assembled multilayer films on surface wettability. Langmuir, 2004. 20(10): p. 4011–4016. 73. Dhanabalan, A., et al., A study of Langmuir and Langmuir–Blodgett films of polyaniline. Langmuir, 1997. 13(16): p. 4395–4400. 74. Facchetti, A., et al., Design and preparation of zwitterionic organic thin films: selfassembled siloxane-based, thiophene-spaced N-benzylpyridinium dicyanomethanides as nonlinear optical materials. Langmuir, 2001. 17(19): p. 5939–5942. 75. Hillebrandt, H., et al., High electric resistance polymer/lipid composite films on indium–tin-oxide electrodes. Langmuir, 1999. 15(24): p. 8451–8459. 76. Joly, S., et al., Multilayer nanoreactors for metallic and semiconducting particles. Langmuir, 2000. 16(3): p. 1354–1359. 77. Park, M.K., et al., Self-assembly and characterization of polyaniline and sulfonated polystyrene multilayer-coated colloidal particles and hollow shells. Langmuir, 2003. 19(20): p. 8550–8554. 78. Ram, M.K., et al., Physical properties of polyaniline films: assembled by the layerby-layer technique. Langmuir, 1999. 15(4): p. 1252–1259. 79. Raposo, M. and Oliveira, O.N., Adsorption of poly(o-methoxyaniline) in layer-bylayer films. Langmuir, 2002. 18(18): p. 6866–6874. 80. Wu, D.G., et al., Photosensitized electron injection from an ITO electrode to trichromophore dyes deposited on Langmuir–Blodgett films. Langmuir, 1999. 15(21): p. 7276–7281.

© 2007, Woodhead Publishing Limited

Electrostatic self-assembled nanolayer films for cotton fibers

445

81. Yuan, C.W., et al., Insertion of polypyrrole into arachidic acid Langmuir-Blodgettfilms. Langmuir, 1995. 11(1): p. 5–7. 82. Koetse, M., et al., Ultrathin coatings by multiple polyelectrolyte adsorption/surface activation (CoMPAS). Macromolecules, 1998. 31(26): p. 9316–9327. 83. Wu, A.P., Lee, J. and Rubner, M.F., Light emitting electrochemical devices from sequentially adsorbed multilayers of a polymeric ruthenium (II) complex and various polyanions. Thin Solid Films, 1998. 329: p. 663–667. 84. Arys, X., et al., Ultrathin multilayers made by alternate deposition of ionenes and polyvinylsulfate: from unstable to stable growth. Thin Solid Films, 1998. 329: p. 734–738. 85. Arys, X., et al., Structural studies on thin organic coatings built by repeated adsorption of polyelectrolytes. Prog Organic Coatings, 1998. 34(1–4): p. 108–118. 86. Lee, J.K., Yoo, D. and Rubner, M.F. Synthesis and characterization of an electroluminescent polyester containing the Ru(II) complex. Chem Mater, 1997. 9(8): p. 1710–1712. 87. Hodak, J., et al., Layer-by-layer self-assembly of glucose oxidase with a poly(allylamine)ferrocene redox mediator. Langmuir, 1997. 13(10): p. 2708–2716. 88. Yoo, D., et al., New electro-active self-assembled multilayer thin films based on alternately adsorbed layers of polyelectrolytes and functional dye molecules. Synthetic Metals, 1997. 85(1–3): p. 1425–1426. 89. Ninham, B.W. and Yaminsky, V., Ion binding and ion specificity: the Hofmeister effect and Onsager and Lifshitz theories. Langmuir, 1997. 13(7): p. 2097–2108. 90. Keller, S.W., et al., Photoinduced charge separation in multilayer thin films grown by sequential adsorption of polyelectrolytes. J Amer Chem Soc, 1995. 117(51): p. 12879–12880. 91. Keller, S.W. et al., Photochemically induced charge separation in electrostatically constructed organic–inorganic multilayer composites, in Nocera D.G. and Wishart, J.S. Editors, Photochemistry and Radiation Chemistry: Complementary Methods for the Study of Electron Transfer, 1998, Adv. Chem. Ser. 254 (Photochemistry and Radiation Chemistry), p. 359–379. 92. Anzai, J., et al., Layer-by-layer construction of multilayer thin films composed of avidin and biotin-labeled poly(amine)s. Langmuir, 1999. 15(1): p. 221–226. 93. Laschewsky, A., et al., Molecular recognition by hydrogen bonding in polyelectrolyte multilayers. Chem–A Europ J, 1997. 3(1): p. 34–38. 94. Decher, G., Hong, J.D. and Schmitt, J., Buildup of ultrathin multilayer films by a self-assembly process: III. Consecutively alternating adsorption of anionic and cationic polyelectrolytes on charged surfaces, Thin Solid Films, 1992. 210(2): p. 831–835. 95. Caruso, F., et al., Magnetic core–shell particles: preparation of magnetite multilayers on polymer latex microspheres. Adv Mater, 1999. 11(11): p. 950–953. 96. Dante, S., et al., Photoisomerization of polyionic layer-by-layer films containing azobenzene. Langmuir, 1999. 15(1): p. 193–201. 97. Lvov, Y., Yamada, S. and Kunitake, T., Non-linear optical effects in layer-by-layer alternate films of polycations and an azobenzene-containing polyanion. Thin Solid Films, 1997. 300(1–2): p. 107–112. 98. Ferreira, M. and Rubner, M.F., Molecular-level processing of conjugated polymers.1. layer-by-layer manipulation of conjugated polyions. Macromolecules, 1995. 28(21): p. 7107–7114. 99. Saremi, F., et al., Self-assembled alternating multilayers built-up from diacetylene

© 2007, Woodhead Publishing Limited

446

100.

101.

102.

103.

104.

105.

106.

107. 108.

109.

110.

111.

112. 113.

114. 115.

116.

Nanofibers and nanotechnology in textiles bolaamphiphiles and poly(allylamine hydrochloride) – polymerization properties, structure, and morphology. Langmuir, 1995. 11(4): p. 1068–1071. Lowack, K. and Helm, C.A., Polyelectrolyte monolayers at the mica/air interface – mechanically induced rearrangements and monolayer annealing. Macromolecules, 1995. 28(8): p. 2912–2921. Caruso, F. and Mohwald, H., Preparation and characterization of ordered nanoparticle and polymer composite multilayers on colloids. Langmuir, 1999. 15(23): p. 8276– 8281. He, J.A., et al., Electrostatic self-assembly of polydiacetylene nanocrystals: nonlinear optical properties and chain orientation. J Phys Chem B, 1999. 103(50): p. 11050– 11056. Voigt, A., et al., Membrane filtration for microencapsulation and microcapsules fabrication by layer-by-layer polyelectrolyte adsorption. Ind Eng Chem Res, 1999. 38(10): p. 4037–4043. Liu, Y.J., Rosidian, A. and Claus, R.O., Mechanical properties of electrostatically self-assembled Al2O3–ZrO2 nanocomposites prepared at room temperature. J Cluster Sci, 1999. 10(3): p. 421–428. Cheng, L., et al., Electrochemical growth and characterization of polyoxometalatecontaining monolayers and multilayers on alkanethiol monolayers self-assembled on gold electrodes. Chem Mater, 1999. 11(6): p. 1465–1475. Liu, Y.J. and Claus, R.O., Strong enhancement of optical absorbance from ionic self-assembled multilayer thin films of nanocluster Pt and polymer dye. J Appl Phys, 1999. 85(1): p. 419–424. Rosidian, A., Liu, Y.J. and Claus, R.O., Ionic self-assembly of ultrahard ZrO2/ polymer nanocomposite thin films. Adv Mater, 1998. 10(14): p. 1087–1091. Yonezawa, T., Onoue, S.Y. and Kunitake, T., Growth of closely packed layers of gold nanoparticles on an aligned ammonium surface. Adv Mater, 1998. 10(5): p. 414–416. Chen, T.Y. and Somasundaran, P., Preparation of novel core–shell nanocomposite particles by controlled polymer bridging. J Amer Ceramic Soc, 1998. 81(1): p. 140–144. Ingersoll, D., Kulesza, P.J. and Faulkner, L.R., Polyoxometalate-based layered composite films on electrodes – preparation through alternate immersions in modification solutions. J Electrochem Soc, 1994. 141(1): p. 140–147. Shiratori, S.S. and Rubner, M.F., pH-dependent thickness behavior of sequentially adsorbed layers of weak polyelectrolytes. Macromolecules, 2000. 33(11): p. 4213– 4219. He, J.A., et al., Bacteriorhodopsin thin film assemblies – immobilization, properties, and applications. Adv Mater, 1999. 11(6): p. 435–466. He, J.A., et al., Photoelectric properties of oriented bacteriorhodopsin/polycation multilayers by electrostatic layer-by-layer assembly. J Phys Chem B, 1998. 102(36): p. 7067–7072. He, J.A., et al., Oriented bacteriorhodopsin/polycation multilayers by electrostatic layer-by-layer assembly. Langmuir, 1998. 14(7): p. 1674–1679. Brynda, E., et al., The detection of human beta(2)-microglobulin by grating coupler immunosensor with three dimensional antibody networks. Biosensors Bioelectronics, 1999. 14(4): p. 363–368. Caruso, F., et al., Characterization of polyelectrolyte–protein multilayer films by atomic force microscopy, scanning electron microscopy, and Fourier transform

© 2007, Woodhead Publishing Limited

Electrostatic self-assembled nanolayer films for cotton fibers

117.

118. 119.

120. 121. 122.

123. 124. 125.

126.

447

infrared reflection-absorption spectroscopy. Langmuir, 1998. 14(16): p. 4559– 4565. Kong, W., et al., Immobilized bilayer glucose-isomerase in porous trimethylamine polystyrene based on molecular deposition. J Chem Soc Chem Commun, 1994. 11: p. 1297–1298. Fischer, P., et al., Polyelectrolytes bearing azobenzenes for the functionalization of multilayers. Macromolecular Symposia, 1999. 137: p. 1–24. Dermody, D.L., et al., Chemically grafted polymeric filters for chemical sensors: hyperbranched poly(acrylic acid) films incorporating beta-cyclodextrin receptors and amine-functionalized filter layers. Langmuir, 1999. 15(3): p. 885–890. Laschewsky, A., et al., Polyelectrolyte multilayers containing photoreactive groups. Macromol Chem Phys, 1997. 198(10): p. 3239–3253. Akin, F., et al., Nanostructure of fluorocarbon films deposited on polystyrene from C 3 F5+ ions. J Phys Chem B, 2004. 108: p. 9656–9664. Nelson, B., et al., Near-infrared surface plasmon resonance measurements of ultrathin films. 1. Angle shift and SPR imaging experiments. Anal Chem, 1999. 71: p. 3928– 3934. Hanton, S.D., Mass spectrometry of polymers and polymer surfaces. Chem Rev, 2001. 101: p. 527–569. Hyde, K., Rusa, M. and Hinestroza, J., Layer-by-layer deposition of polyelectrolyte nanolayers on natural fibres: cotton. Nanotechnology 2005. 16: S422–S428. Kim, H.J., et al., Dynamic sequential layer-by-layer deposition method for fast and region-selective multilayer thin film fabrication. Langmuir, 2005. 21(18): p. 8532– 8538. Porcel, C.H., et al., Ultrathin coatings and (poly(glutamic acid)/polyallylamine) films deposited by continuous and simultaneous spraying. Langmuir, 2005. 21(2): p. 800–802.

© 2007, Woodhead Publishing Limited

17 Nanofabrication of thin polymer films I. L U Z I N O V, Clemson University, USA

17.1

Introduction

Surface structure and behavior of fiber materials are of the utmost importance for the properties of fibers and textiles in processing and use, since friction, abrasion, wetting, adhesion, adsorption and penetration phenomena are involved. Further advances in industrial textiles impose rigorous requirements for the surface modification: a given textile material, depending on the conditions under which it is utilized, has to be hydrophobic or hydrophilic, acidic or basic, conductive or nonconductive, and deliver or adsorb some species. In order to obtain textile materials with the desired performance, the fiber surface is often modified with polymer layers before use. Numerous surface modifications involving oxidation, reduction, elimination, addition, cyclization and condensation, and grafting of macromolecules have been described in the literature.1 Among them, the grafting technique has several advantages over others,2 including easy and controllable introduction of new polymer chains with a high surface density, precise localization of the chain at the surface and long stability of the grafted layers. Moreover, covalent attachment of the macromolecules onto a polymer surface can avoid their delamination in liquid media. The polymer chains located at the interface can be anchored to the surface in several configurations. The macromolecules may form multiple connections with the substrate or be connected to the surface by one or both ends. Tethered polymer chains that are grafted to a solid substrate by one chain end may be definitely distinguished from other anchored polymer layers, since they form polymer brushes if relatively high grafting density is reached.3, 4 Brush-like layers are formed due to the excluded volume effect, when the substrate is completely covered with a relatively dense monolayer of grafted chains stretched normal to the support. There are several major parameters that control the grafted layer properties: grafting density, chain length, polydispersity and chemical composition of the chains.5, 6 This chapter focuses on synthesis and characterization of the nanothick, 448

© 2007, Woodhead Publishing Limited

Nanofabrication of thin polymer films

449

chemically grafted polymer films (polymer brushes) on inorganic and polymeric substrates including polymer fibers. The synthesis has been conducted employing the recently developed macromolecular anchoring layer approach.7 The chapter also offers examples of the application of the polymer grafting technique for the generation of hydrophobic, hydrophilic and switchable fibrous materials.

17.2

Macromolecular platform for nanofabrication

The chemical grafting of polymers can be accomplished by either ‘grafting to’ or ‘grafting from’ methods.5 According to the ‘grafting to’ technique, end-functionalized polymer molecules react with complementary functional groups located on the surface to form tethered chains. The ‘grafting from’ technique utilizes the polymerization initiated from the substrate surface by attached (usually by covalent bonds) initiating groups. It is necessary to highlight that most of the developed grafting (‘to’ and ‘from’) methods require attachment of end-functionalized polymers or low molecular weight substances (e.g. initiators) to the substrate for the polymer brush synthesis. There are two common approaches for the attachment of polymerization initiators or end-functionalized polymers for the brush fabrication. The first one relies on the reactions between end-functionalized initiator/polymer and native functional groups originally present on the substrate surface.8–10 A different approach involves the formation of a monolayer consisting of functional groups active towards terminally functionalized (e.g. epoxide, amine, anhydride or hydroxide) initiator/polymer.11, 12 Silane and thiol chemistries have proved to be suitable for the grafting in this case. Usually the coupling methods are relatively complex and specific for certain substrate/(macro)molecule combinations. An alternative method for the attachment involves primary polymer anchoring (mono)layer with activity towards both surface and functionalized (macro)molecule.13–16 The polymer is used for the initial surface modification as well as generation of the highly reactive primary anchoring layer. When deposited on a substrate, the primary layer first reacts with the surface through formation of covalent bonds (Fig. 17.1). The reactive units located in the ‘loops’ and ‘tails’ sections of the attached macromolecules are not connected to the surface.17 These free groups offer a synthetic potential for the further chemical modification reactions and serve as reactive sites for the subsequent attachment of the functionalized (macro)molecules. If the polymer used for building the primary layer contains functional groups highly active in various chemical reactions, the primary layer approach becomes virtually universal towards both surface and endfunctionalized species being used for the brush formation. For the majority of the initial experiments on the surface grafting a silicon wafer was used as a substrate, since it is now a ‘standard’ surface (along with

© 2007, Woodhead Publishing Limited

450

Nanofibers and nanotechnology in textiles 1.00

10.0

‘Free’ reactive groups in ‘loops’ and ‘tails’

(c)

(b)

(a) Reactive groups in ‘trains’ attach the polymer to surface

0

2.5

5.0 µm

7.5

7.5

0.75

5.0

0.50

2.5

0.25

0 10.0

0

0.25

0.50 µm

0.75

0 1.00

17.1 Schematic representation of reactive polymer attached to substrate (a). Atomic force microscopy (AFM) images showing topography of PGMA layer deposited by dip-coating on silicon wafer: (b) 10 × 10 µm2 and (c) 1 × 1 µm2.

mica) for the research in the field of thin polymer layers. A lot of brush investigations have been made using the wafer as a model surface, and thus we can compare our results with work of others. Poly(glycidyl methacrylate) (PGMA) was used as a primary anchoring polymer layer. A polymer with epoxy functionality was chosen, since the epoxy groups are quite reactive with carboxyl, hydroxyl, amino and anhydride functional groups. The versatile chemistry of the epoxy groups offered flexibility in selection of necessary initiators/macromolecules that are to be attached to the surface. The epoxy groups of the polymer chemically anchored PGMA to the surface.13 The glycidyl methacrylate units located in the ‘loops’ sections of the attached macromolecules were not connected to the surface. These free groups served as reactive sites for the subsequent attachment of polymerization initiators and/or polymer with functional groups, which exhibit an affinity for the epoxy modified surface. The attachment of PGMA to various surfaces was studied and it was found that the uniform and homogeneous epoxy containing polymer layer could be deposited on surfaces by adsorption or dip-coating.7, 18–22 The epoxy containing polymer layer could be deposited as a monolayer on polymeric poly(ethyleneterephthalate) (PET), polyethylene, polypropylene (PP), silicon resin, nylon) and inorganic (silica, glass, titanium, alumina, gold, silver) surfaces. It was possible to regulate the thickness of the layer and consequently the amount of epoxy groups on the surface by varying the solvent characteristics and concentration of solution being used for the deposition. Layers with a thickness from 1 to 10 nm were obtained. It was found that the layer could not be removed from the wafer using a vigorous solvent treatment, suggesting that PGMA was chemically bonded to the surface. The PGMA layer was smooth (atomic force microscopy, AFM, roughness 0.3 nm) and uniformly covered the surface on nano- and micro-levels (Fig. 17.1b and c). To check the activity of the epoxy groups in the loops and tails of the adsorbed polymer,

© 2007, Woodhead Publishing Limited

Nanofabrication of thin polymer films

451

a model reaction between the epoxy polymer layer and dodecyl amine (DA) was conducted.19 The PGMA layer retained its epoxy functionalities, which could be used for further chemical modification reactions.

17.3

‘Grafting from’ technique for synthesis of polymer films

In the ‘grafting from’ technique the polymerization is initiated from the substrate surface by attached (usually covalently bonded) initiating groups. When exceptionally high grafting density is needed, the ‘grafting from’ approach is the only method for brush formation. The polymer brushes grown from the surface by the technique indeed possess extremely high density of the attached chains. Molecules of a monomer penetrate through the already grafted polymer layer easily and significant grafted amounts can be reached. This technique was used for the preparation of thick grafted layers of high grafting density on the surface. Anionic,23, 24 cationic,25, 26 controlled/living27, 28 and conventional29, 30 free radical polymerizations have been successfully used to synthesize tethered polymer layers on solid substrate surfaces. By appropriate choice of initiating system, temperature, monomer and concentration, it is quite possible to synthesize layers possessing different morphology, thickness and composition.5 Thus, fine-tuning of the layer properties is possible. To realize a ‘grafting from’ approach employing the PGMA platform the primary layer was used to synthesize an effective macroinitiator for controlled/ ‘living’ atom transfer radical polymerization (ATRP).20, 21, 31 The controlled/ ‘living’ free radical polymerization has a number of advantages over traditional radical polymerization procedures. The main advantage of a ‘living’ process is that it provides reliable control over the polymer molecular weight and narrow polydispersities. Thus, the nature of the polymerization process permits structural characteristics of the grafted polymer brush to be readily varied and controlled. An added benefit is the frontal character of the chain growth on the surface. In this manner all chains have very similar history that may be translated in more predictable cooperative behavior of the chains and make the brush nearly ‘defect free’. Numerous effective approaches have been reported for the synthesis of polymer brushes by the ‘living’ free radical polymerization.5, 28, 32, 33

17.3.1 Synthesis of macroinitiator The synthesis of the ATRP macroinitiator included two major steps: deposition of the anchoring PGMA layer on the surface and attachment of bromoacetic acid (BAA) to the surface modified with the anchoring layer through the reaction between the carboxyl and epoxy functionalities. The reaction between

© 2007, Woodhead Publishing Limited

452

Nanofibers and nanotechnology in textiles

the epoxy groups and the carboxyl functionality of the BAA produces a bromoacetic ester derivative of the PGMA. Such α-bromoesters are known as effective initiators for ATRP of styrene, acrylic and some other vinyl monomers.34 Therefore, the ATRP macromolecular initiator, covalently anchored to the silicon surface, was synthesized. In order to prepare a thin layer of the PGMA macromolecular precursor, attached to the surface of silicon wafer, dip-coating from PGMA solution in methyl ethyl ketone (MEK) was employed. It was found that the thickness of PGMA layer is nearly proportional to the concentration of the solution, so the thickness of PGMA anchoring layers can be tuned easily by varying the concentration. After the dip coating, we observed not only smooth and uniform covering, but also stable reproducibility in the thickness of the reactive anchoring layer. Annealing of the adsorbed film for 20–40 min at 110 °C led to the permanent attachment of the deposited film to the substrate. The annealed PGMA layer was smooth and uniformly covered the substrate. The PGMA film treated with BAA vapor produced macroinitiator possessing smooth and uniform surface morphology (Fig. 17.2a and b). Variation of the PGMA layer thickness allowed control over the amount of BAA attached to the surface. There was a nearly linear correlation between the quantity of the epoxy polymer attached to the surface and the amount of the initiator anchored (Fig. 17.2c). The kinetics of the BAA deposition process was also suitable for regulation of the BAA amount reacted with the anchoring PGMA layer. The PGMA layer was contacted with BAA vapor at two different temperatures (30 and 90 °C) for different times. The thickness of the anchoring layer was kept constant (4 nm) in this series of experiments. At the beginning of the BAA deposition the surface concentration of the attached molecules increased with time and then leveled off after 4–6 h. At higher temperature the reaction rate proceeded more rapidly. The maximum initiator concentration at 30 °C was around 10 molecules/nm2, while at 90 °C it reached 28 molecules/nm2. 10.0

0

1.00

(b)

2.5

5.0 µm

7.5

7.5

0.75

5.0

0.50

2.5

0.25

0 10.0

0

0.25

0.50 µm

0.75

0 1.00

Initiator density [molecules/nm2]

(a)

(c) 30

20

10

1

2 3 4 Thickness of PGMA layer [nm]

5

17.2 AFM images of PGMA/BAA macroinitiator. Image size: (a) 10 × 10 µm2, (b) 1 × 1 µm2. Vertical scale – 10 nm. (c) Surface density of attached BAA versus thickness of the annealed (12 h at 90 °C) PGMA layer.

© 2007, Woodhead Publishing Limited

Nanofabrication of thin polymer films

453

17.3.2 Atom transfer radical polymerization from macroinitiator The fundamental idea of ATRP is the halogen exchange in the polymerizing system between the halogen terminated growing polymer chain/Cu(I) dNbP complex and macroradical/Cu(II) dNbP complex.34 Chain propagation is a first order process while termination is a second order reaction. For conventional bulk or solution ATRP the equilibrium is strongly shifted to the left. The free radical concentration is as low as 10–7–10–8 mole/l and Cu(II) concentration is approximately 5% of that for Cu(I).34 Consequently, the termination is diminished and all chains grow simultaneously during the polymerization without noticeable termination. An adequate concentration of Cu(II) is critical for effective reaction control. However, when the ATRP is initiated from the surface, the amount of the initiator located on the substrate is not sufficient to create the concentration of Cu(II) species required for polymerization control.28, 35 There are two methods developed for maintaining the Cu(II) concentration for ATRP initiated from a surface: (a) simultaneous initiation of ATRP from the surface and in solution, and (b) addition of the necessary amount of Cu(II) at the beginning of the process.5, 28 Both approaches were tested for the ATRP grafting from the PGMA/BAA macroinitiator adsorbed on the surface and successfully obtained the polymer brushes grafted to the silicon surface by the two methods. Specifically, polystyrene (PS) brushes of different thicknesses were synthesized on the PGMA/BAA modified silicon wafer by ATRP. At the beginning of the polymerization process, a linear increase of polystyrene layer thickness was observed. Later, the rate of the grafting decreased and the brush thickness practically leveled off. Different surface concentrations of BAA were used in the grafting experiments to acquire knowledge about the relationship between amount of the initiator anchored to the surface through PGMA and rate of the brush formation. The increase in the surface density of the initiating moieties led to the increase in the grafting rate. However, a cutoff initiator concentration beyond which no increase of the thickness of the grafted layer was observed. From comparison between the surface densities of the initiator and the attached polymer it was determined that the efficiency of the initiation from the surface was in the level of 5– 15%. It was also shown that other (than PS) polymer brushes could be successfully grown from the surface employing the macroinitiator based on the PGMA platform. N-isopropylacrylamide, poly(ethylene glycol) methyl ether methacrylate, butyl methacrylate and pentafluorostyrene were grafted to a substrate. To prove the controlled nature of the polymerization initiated from the surface by the PGMA/BAA macroinitiator, polymerization of styrene was conducted using the already grown PS brush of 19 nm as an initiator. It was

© 2007, Woodhead Publishing Limited

454

Nanofibers and nanotechnology in textiles

found that the thickness of the grafted layer increased from 19 to 31 nm, indicating successful chain extension. The experiment demonstrated that the PS brush obtained with the PGMA/BAA macroinitiator is carrying active Br ends. Following the polymerizations on the model silicon surface, the developed approach for the synthesis of the grafted layers on polymeric substrate was tested. The surface of the PET film was modified with a PGMA/BAA layer and ATRP of styrene initiated from the PET surface was carried out. As a result of the polymerization, a PS layer was firmly grafted to the PET surface. Figure 17.3 shows AFM images and values of water contact angles for virgin PET surface, PET surface covered with PGMA/BAA combination, and grafted PS layer. One can see that the surface morphology and wettability of the PET 1.00

1.00 (a)

0

(b) 0.75

0.75

0.50

0.50

0.25

0.25

0 1.00

0.25

0.50 0.75 µm Contact angle 72°

0

0.25

0.50 0.75 µm Contact angle 70°

0 1.00

1.00 (c) 0.75

0.50

0.25

0

0.25

0.50 0.75 µm Contact angle 92°

0 1.00

17.3 AFM phase images (1 × 1 µm2) of (a) virgin PET surface; (b) BAA layer attached to the PGMA layer; (c) PS layer obtained by ATRP initiated by attached BAA.

© 2007, Woodhead Publishing Limited

Nanofabrication of thin polymer films

455

film was changed after the polymerization. The obtained result suggested applicability of the developed synthetic approach to surface modification of fibers and textile materials.

17.4

‘Grafting to’ technique for synthesis of polymer films

Synthesis of the polymer brushes can be readily accomplished by a ‘grafting to’ method.36–38 In the ‘grafting to’ technique, end-functionalized polymer molecules react with complementary functional groups located on the surface to form tethered chains. End-functionalized polymers (with a narrow molecular weight distribution) can be synthesized by living, anionic, cationic, radical, group transfer and ring opening metathesis polymerizations. Thus the advantage of the method is that the well-defined end-functionalized polymers can be used for the grafting and, as a result, well-defined brushes can be readily obtained. The covalent bond formed between the polymer chain and the substrate makes the polymer brushes resistant to chemical environmental conditions. On the other hand, the technique has a constraint in terms of the maximum grafting that can be obtained, namely that the grafting is selflimiting.37, 39 Polymer chains must diffuse through the existing polymer film to reach the reactive sites on the surface. This barrier becomes more pronounced as the tethered polymer thickness increases. Thus, the polymer brush obtained typically has low grafting density and low thickness. The density of the brush obtained by the ‘grafting to’ method can be increased if the macromolecule attachment is conducted from a solution at Θ conditions36 or from melt.11, 37, 40, 41 Grafting from melt in particular offers potential advantages over grafting from solution, since the excluded volume interactions that make it difficult for chains to penetrate an initial grafted layer are screened out in the melt.37 If the grafting is carried out from the melt, many polymer chains are already in locations from which they do not need to pass the potential barrier, since they are already adjacent to the surface. The macromolecules only need to reorient themselves within the first monolayer in order to expose the terminal groups to the surface functionalities. An additional increase in grafting density for the attachment from the melt can be achieved when the PGMA anchoring layer is used for the introduction of reactive group on a substrate surface.18, 19, 22, 42 Attachment of end-functionalized PS from melt to a 1 nm thick layer of PGMA, attached to silicon wafers was studied.18 In fact, the free groups in the ‘tails’ and the ‘loops’ of the PGMA macromolecule (deposited on the substrate) served as reactive sites for the anchoring of the end-functionalized polymers. Comparison of the results for the grafting to the PGMA primary layer with published data11 obtained for the epoxysilane (ES) monolayer suggested that there are

© 2007, Woodhead Publishing Limited

456

Nanofibers and nanotechnology in textiles

many similarities between these grafting processes. The same major trends were observed. However, the grafting to the PGMA layer was much more effective. In Fig. 17.4 the surface coverage, Γ is plotted versus the degree of polymerization (N) of the grafting polymer. The surface coverage initially increases for the range 43 < N < 440, passes through a maximum at N = 440, and then decreases. The maximum is close to the critical entanglement molecular weight of PS, MC, which is 31 200 g/mol (NC = 300).43 However, it is necessary to stress that the surface coverage and hence the thickness of the PS grafted to the PGMA layer is two- to three-fold greater than that obtained for the ES monolayer. It appeared that the epoxy groups located in the loops/tails of the adsorbed PGMA macromolecule were more accessible to the end-functional groups of PS when compared with epoxysilane with terminal epoxy groups located mainly at the monolayer surface. It was concluded that the high efficiency of PGMA in the grafting reactions was related to the high mobility of the epoxy reactive groups and to the formation of an interpenetrating zone at the PS/PGMA interface. The grafting of end-functionalized polymers employing the macromolecular anchoring layer approach was further investigated considering the effect of the molecular weight and thickness of the PGMA layer.42 The obtained experimental data indicated that, when the higher molecular weight PS chains are considered, neither the thickness nor the molecular weight of the PGMA layer have any significant influence on the grafting of the end-functionalized PS to the macromolecular anchoring layer. Conversely, the lower molecular weight PS sample does not follow the trend, since significantly higher amounts of the polymer can be grafted to the thicker anchoring layers. The high grafting in this case may be a result of formation of an extremely extended

20

PGMA Epoxysilane

G [mg/m2]

15

10

5

0 100 1000 Degree of polymerization [N]

10000

17.4 Surface coverage, Γ (mg/m2) versus degree of polymerization (N) of grafted PS chains.

© 2007, Woodhead Publishing Limited

Nanofabrication of thin polymer films

457

interphase between loops of the adsorbed PGMA and the low molecular weight PS. The grafting of carboxylic acid end-functionalized poly(ethylene glycol) methyl ether (PEG) chains of different molecular weights from the melt onto a surface employing a PGMA anchoring layer was studied as well.19, 22 The grafting led to the synthesis of the complete PEG brushes possessing exceptionally high grafting density. The maximum thickness of the attached PEG films was strongly dependent on the length of the polymer chains being grafted. The maximum grafting efficiency was close to the critical entanglement molecular weight region for PEG. AFM imaging revealed that the grafting process led to complete PEG layers with surface smoothness on a nanometric scale. Practically all the samples were partly or fully covered with crystalline domains that disappeared under water. The PEG hydrophilic nature meant that the surface with the grafted layer exhibited a low (up to 21°) water contact angle. In general, it was shown that until a certain level of miscibility/compatibility between a polymer being grafted and the anchoring macromolecules (attached to a substrate) is reached, the effect of the layer thickness and molecular weight is not significantly pronounced. The same grafted layers were attached to relatively thin (1.0 nm) and thick (10 nm) PGMA films. If polymers to be grafted are miscible with PGMA, the grafted amount is strongly dependent on the anchoring layer thickness. In the case of such miscibility/compatibility the effect may become dramatic. Specifically, if end-functionalized polymer is miscible/compatible with the anchoring layer, the thickness and molecular weight of the anchoring layer must be seriously taken into account to avoid (or reach) the extensive interpenetration at the boundary. The developed grafting approach was also employed for attachment of polymers to polymeric films and fibers. Hydrophilic/polar (PEG, polyacrylamide, polyacrylic acid) and hydrophobic (PS and polypentafluorostyrene) polymers were attached to PET, polyethylene, cotton and nylon. Figure 17.5a and b shows the morphology and wettability of a PET surface modified with PS and PEG grafted layers. The AFM images revealed that the polymeric surface was completely covered with the grafted layers and the polymer grafted dictated the surface properties of the polymer film. The synthesized layers could not be removed by multiple rinsing in hot solvents, including such strong solvent as N, N-dimethylformamide (DMF). The obtained results suggested that polymers possessing functional groups could be indeed grafted to polymeric surfaces modified with the PGMA anchoring layer. Polypentafluorostyrene and/or PS were also successfully grafted to PET, cotton and nylon fiber/textile materials utilizing the grafting approach developed. Highly hydrophobic materials were obtained as a result of the surface modification (Fig. 17.5c).

© 2007, Woodhead Publishing Limited

458

Nanofibers and nanotechnology in textiles

(a)

(b)

(c)

17.5 (a) and (b) AFM topography images (1 × 1 µm ) and wettability measurements for the PET surface modified with grafted polystyrene (a) and poly(ethylene glycol) (b). (c) Droplet of water on the surface of polyester textile material modified with grafted polystyrene layer. The polymers were grafted through the PGMA layer. 2

17.5

Synthesis of smart switchable coatings

Polymer coatings that are responsive to environmental conditions are of great interest for various advanced applications including generation of ‘intelligent’ fibers and textiles. The advantage of these ‘smart’ materials is the ability to switch and/or tune the surface properties of the coatings by applying external stimuli to vary, for example, adhesion, wettability, friction, roughness, reactivity, biocompatibility and selectivity. In recent years, various efforts have been reported on the development of such smart materials that can act in response to environmental stimuli, such as changes in electric potential, temperature, pH, or the presence of a specific chemical substance.44–50 With the increasing demand for more sophisticated surfaces, one of the current targets is to fabricate and understand materials whose interfacial properties are capable of consistent reversible changes in their characteristics according to outside conditions or stimuli.45

17.5.1 Mixed polymer brushes Recently, the new class of interfacially active responsive materials, mixed polymer brushes, has been developed.50, 51 To generate the responsive brush, two or more different polymers have to be grafted to a surface (Fig. 17.6). Adaptive and responsive behavior of the thin film (chemically attached to a substrate) is based on phase segregation mechanism of two incompatible polymers constituting the mixed brushes. Upon outside stimuli (solvent quality, temperature, pH, ionic strength) the phase segregation results in switching of spatial distribution of functional groups within the ultrathin film and different chain fragments may be delivered to the brush exterior. The discovery of the

© 2007, Woodhead Publishing Limited

Nanofabrication of thin polymer films Polymer I

459

Polymer II Conditions selective for Polymer II

Conditions selective for Polymer I

17.6 Schematic of morphology variation of mixed polymer brush situated in non-selective (left) and selective (right) conditions.

brush switchable behavior has offered novel virtually endless possibilities to generate adaptive/responsive surfaces/interfaces expressing a variety of unique properties. Moreover, the combination of two or more polymers in the brush does not act as a simple addition of different functions, but it affects a specific morphology of the film driven by a subtle interplay between increased numbers of interactions. Therefore, the difference between a brush of a random copolymer (when different functions can be combined in the same polymer chain by random distribution of two or more different monomer units in the polymer molecule) and binary or multicomponent brushes with different homopolymers randomly grafted to the same substrate should be emphasized. The principal difference is introduced by the structure of the film when different end-attached polymer chains can segregate into nanoscopic domains affecting a unique morphology of the film. PS and poly(2-vinylpyridine) (PVP) switchable mixed brushes were synthesized employing the macromolecular anchoring layer used to activate the substrate boundary.52 The ‘grafting to’ approach was used to attach the polymer chains to the substrate. The PS and PVP were deposited on the wafers in a sequential fashion to chemically graft PS in a first step, and subsequently graft PVP. The investigation has demonstrated that it is possible to synthesize mixed polymer brushes of various compositions by grafting to a silicon surface using PGMA as an anchoring interlayer. The mixed brush synthesized in this manner exhibits changes in surface energy as measured by contact angles. Specifically, these contact angle measurements follow the change in concentration of PS content well, and indicate switching between PS and PVP with toluene and ethanol, respectively. Scanning probe microscopy (SPM) was used to visualize the sample surface morphology, observe microphase segregation within the brushes and sample roughness. Figure 17.7 shows how the surface morphology of the mixed brushes changes on content of PS and treatment with selective (ethanol and toluene) solvents. The surface topography of the mixed brushes showed a general trend toward smoother surfaces as the ratio of PS increases to the 95% composition. For samples with lower PS content, formation of clusters on the brush surface was observed, revealing pronounced phase segregation in the grafted film where grafted PS and PVP chains try to minimize their mutual contact.

© 2007, Woodhead Publishing Limited

460

Nanofibers and nanotechnology in textiles (a)

(b)

(c)

17.7 SPM topographical images of mixed PS-PVP brushes (grafted via PGMA anchoring layer) rinsed with ethanol (top row) and toluene (bottom row). Dimensions are 1 × 1 µm2 with a 5 nm vertical scale: (a) 31% PS, (b) 65% PS and (c) 94% PS.

In addition, an approach has been developed for fabrication of the stimulisensitive mixed brushes by ‘one-pot’ techniques, where all components of the smart nanolayers are simultaneously deposited on the surface and attached in one single step. Specifically, the one-step synthesis of PS-PVP mixed brushes was carried out. The end-functionalized PS and PVP were deposited on the surface (modified with PGMA) simultaneously from a joint solution in MEK. An initial study of the effect of depositing varying amounts of polymers indicated a preferential grafting of PVP at higher deposited amounts of polymer blend. However, the composition of the grafted nanolayer could be regulated by the ratio of the polymers in solution. Rinsing the mixed PS/ PVP polymer brush in selective solvents allowed observation of the change in water contact angle as a function of the nanolayer composition. Using toluene and ethanol as the selective solvents the hydrophilic/hydrophobic nature of the brush–air interface changes was probed. In toluene, PS dominates the interface with larger contact angle values. Conversely, ethanol results in a PVP dominated surface and lower contact angle values. Changes in the roughness and structure of the nanolayer surface were also observed (by AFM), corresponding to the solvent treatment and the layer composition (Fig. 17.8). The phase imagery indicated a degree of segregation between the PS and PVP rich phases.

© 2007, Woodhead Publishing Limited

Nanofabrication of thin polymer films THF

Ethanol

461

Toluene

17.8 AFM topographical images of PS-PVP mixed brushes of different thicknesses in their hydrophobic (toluene treatment), hydrophilic (ethanol treatment), and neutral (tetrahydrofuran (THF) treatment) states (1 × 1 µm2, vertical scale 5 nm). Thickness of the grafted layer: 9 nm (top row) and 5 nm (bottom row).

Switchable nanolayers were synthesized by combination of the grafting end-functionalized polymers (‘grafting to’) and polymerization initiated from the surface (‘grafting from’).53 The combination allows synthesis of responsive nanolayers consisting of polymers that can only be attached by a certain grafting method. The synthesis was conducted according to the following procedure. BAA was used as an initiator of ATRP polymerization. Attachment of the BAA molecules to the surface covered with the PGMA film was conducted from gaseous phase. Next, the synthesis of the poly(t-butyl acrylate) PTBA brush was carried out by the ‘grafting to’ method utilizing PTBA with an end carboxyl functional group. The PTBA melt grafting buried the ATRP initiator under the polymer brush with a significant thickness of 12–20 nm. To complete the fabrication of the mixed brush, ATRP of styrene initiated by the PGMA/BAA macroinitiator was carried out. As a result of the developed process, the mixed polymer brushes with PTBA brush thickness 12–20 nm and PS layer 1–100 nm were obtained. Hydrolysis of PTBA to polyacrylic acid (PAA) was utilized to synthesize polymer layers possessing hydrophobic/ hydrophilic properties. The brushes changed their surface morphology when they were exposed to solvent with different polarity (Fig. 17.9). For the best

© 2007, Woodhead Publishing Limited

462

Nanofibers and nanotechnology in textiles (a)

(b)

(c)

17.9 AFM topography images (1 × 1 µm2) of mixed PS–polyacrylic acid brushes treated with different solvent: (a) benzene (water contact angle 80°); (b) THF; (c) EtOH (water contact angle 40°). Thickness of PS/PAA 15/15 nm.

samples the contact angle was changed by 40° when the hybrid layer was exposed to the different environments.

17.5.2 Switchable unary polymer brush The mobile epoxy groups located in the loops/tails of the adsorbed PGMA macromolecule are shown to be accessible to the functional groups of an end-functionalized polymer and thus available for grafting. The mobility of the loops/tails of PGMA could be also effectively used to develop a novel system, which is robust, and possesses wettability on demand. In fact, the responsive unary polymer brush (UPB) system described below benefited from the mobility of the PGMA loops effectively to switch surface properties.54 UPB can be described as a binary system consisting of an end-functionalized polymer grafted to a macromolecular anchoring layer. The UPB system developed (Fig. 17.10) consists of end-grafted PS and a PGMA anchoring layer. It was anticipated that the mobile loops/trains of the macromolecule could be effectively used to tailor surface properties between the favorable state for PS (non-polar) and the favorable state for PGMA (relatively morepolar). AFM analysis of the substrates covered with UPB after solvent treatment (Fig. 17.10) showed a well-defined change in morphology for higher molecular weight PS. The surface changed from ‘smooth’ (toluene) to ‘ripple’ (MEK) after treatment with selective solvents. The phase segregation can be explained in terms of mobility of the free end of PS and restricted mobility of the ‘loops’ due to the anchored ‘train’ segments. The higher mobility of the PS chains indeed resulted in perpendicular segregation with one of the species enriched at the surface. This layered segregation resulted in the ‘smooth’ morphology after toluene treatment, while after MEK treatment, the restricted mobility of ‘loops’ prevented layer segregation of PGMA at the surface. This resulted in the two species self-assembling laterally into well-defined twodimensional structures corresponding to the ‘ripple’ morphology. Contact

© 2007, Woodhead Publishing Limited

Nanofabrication of thin polymer films

463

PS PGMA

MEK MEK

PGMA brought to boundary

Toluene

Toluene

PS brought to boundary

17.10 Left: Depiction of switchability of UPB system consisting of endgrafted PS and PGMA anchoring layer. Right: 1 × 1 µm2 AFM topography images of PS layers grafted to 3 nm PGMA after treatment with toluene and MEK. PS molecular weight: 672 000 g/mol. Vertical scale: 10 nm. The same PET fabric

17.11 Wettability of PET fabrics covered with switchable UPB after treatment with different solvents. Left: treated with toluene. Right: treated with MEK.

angle measurements after solvent treatments indicated that the highest molecular weight PS (Mn = 672 000 g/mol) showed maximum switching (approx. 20°). Using PGMA as an anchoring interlayer a switchable polymer nanolayer on the surface of PET textile material has been synthesized. The PET fabric changed the surface properties after being treated with different solvents (Fig. 17.11). When the fabric was exposed to toluene, it became hydrophobic and water did not penetrate the material. Conversely, water penetrated throughout the textile materials, if they were exposed to MEK. The wettability changes were reversible.

© 2007, Woodhead Publishing Limited

464

17.6

Nanofibers and nanotechnology in textiles

Synthesis of ultrahydrophobic materials

Water and soil repellency has been one of the major targets for fiber and textile scientists and manufacturers for centuries. Combinations of new materials for fiber production with a variety of surface treatments have been developed to reach the condition of limited wettability. Nevertheless, additional efforts has been needed to create fiber and textile materials with ideal repelling properties. Nature has already developed an elegant approach that combines chemistry and physics to create super-repellent surfaces.55, 56 Lotus leaves are unusually water repellent and keep themselves spotless, since countless miniature protrusions, coated with a water-repellent hydrophobic substance, cover their surface. Water cannot spread out on the leaves and it rolls around as droplets, removing grime and soil as it moves. The lotus effect is based on the surface roughness caused by different microstructures combined with hydrophobic properties of the wax covering the leaf surface.55 The surface roughness is the key prerequisite for the lotus effect. Owing to the rough surface the wettability of the lotus leaves is decreased and the contact area for dirt particles is reduced. A surface with both receding and advanced water contact angles above 150° is considered to be an ultrahydrophobic (or superhydrophobic) boundary.57–59 The common way for enhancing the hydrophobicity is to lower the surface energy. However, even materials with the lowest surface energy (6.7 mJ/m2 for a surface with regularly aligned closest-hexagonalpacked —CF3 groups) gives a water contact angle of only around 120°. In fact, surfaces with water contact angle of more than 150° may be developed only by introducing proper roughness on material boundaries having low surface energy. Classical works of Wenzel60 and Cassie and Baxter61 established that roughness as well as surface energy are the factors that determine wettability. Wenzel proposed a model describing the contact angle θ ′ at a rough surface: cosθ ′ = r cosθ

[17.1]

where r is a roughness factor, defined as the ratio of the actual area of a rough surface to the geometric projected area, and θ is the thermodynamic contact angle on a smooth surface of the material. Since r is always larger than unity, the surface roughness enhances both the hydrophilicity of hydrophilic surfaces and the hydrophobicity of hydrophobic ones. Cassie and Baxter proposed an equation describing the contact angle θ ′ at a surface composed of solid and air, assuming the water contact angle for air to be 180°: cos θ′ = f1 cos θ – f2

[17.2]

where f1 is the fraction of fluid area in contact with the material, and f2 is the fraction of the fluid area in contact with air. The equation can be used for

© 2007, Woodhead Publishing Limited

Nanofabrication of thin polymer films

465

hydrophobic surfaces that trap air in the hollows of the rough surface. It is possible to reach a transition from the Wenzel to the Cassie/Baxter regime (when the hysteresis is minute), if an optimum roughness is introduced to a hydrophobic substrate.62 These optimized surface structures can trap air and exhibit ultrahydrophobic properties necessary to achieve the lotus effect.

17.6.1 Fabrication of ultrahydrophobic textile materials For the fabrication of the lotus fibers two major requirements must be fulfilled: (1) the fibers need to have low surface energy and (2) the extended degree of roughness should be created. One of the methods to synthesize the rough and hydrophobic coating for fibers is utilization of a combination of a surfaceattached polymer layer and nanoparticles. The surface layer consisting of a low surface energy polymer will bring hydrophobicity to the fiber surface whereas nanoparticles will create necessary topography. The highly hydrophobic PET fabric was obtained by a combination of PS (low surface energy component) and silver nanoparticles (roughness initiation component). The desired ultrahydrophobic fabric was produced using a multistep grafting approach. To begin with, a polyester fabric was rinsed in multiple solvents to remove contaminants. After cleaning and drying at ambient conditions the textile material substrate was subjected to plasma discharge and rinsed in tetrahydrofuran (THF) to remove low molecular weight remnants formed due to the chain scission process. Next, the fabric was dip coated with a 70/30 PGMA/PVP mixed solution in MEK. The modified substrate was annealed at 110 °C for 10 min to aid self-cross-linking of the epoxy groups of PGMA. (Cross-linking of PGMA stabilizes the microstructure of the blend.) The annealed substrate was treated with ethanol (a good solvent for PVP) to ensure the presence of PVP on the surface. The fabric was then exposed to a suspension of silver nanoparticles (110–130 nm in diameter) in deionized water overnight. The silver nanoparticle adsorbed surface was dip coated with a second layer of PGMA. This second layer entraps silver particles in a cage between the first PGMA/PVP layer and the PGMA layer. This sandwich layer is robust, since its integrity is maintained by the cross-linked epoxy functionalities. Carboxy terminated PS was grafted to the unreacted epoxy functionalities of the top layer at 150 °C. Unreacted PS was removed by multiple rinsing in toluene. AFM study (images are not shown) demonstrated that PGMA/PVP blend (after treatment in ethanol) showed a typical dispersed morphology characteristic for the 70/30 immiscible polymer blend. The irregular PVP islands successfully immobilized particles (after overnight exposure to a suspension of silver nanoparticles in deionized water) due to the affinity of pyridyl groups to silver through the metal–ligand interactions of the nitrogen atoms. Varying the amount of PVP in the blend can regulate the density of

© 2007, Woodhead Publishing Limited

466

Nanofibers and nanotechnology in textiles

(a)

(b)

(c)

17.12 (a) Silver nanoparticles of size greater than 105 nm adsorbed on PET fiber surface. Magnification × 10 000. Static water contact angle on (b) control sample grafted with PS (no silver) and (c) ultrahydrophobic fabric.

silver nanoparticles adsorbed. The multilayered PS/PGMA/silver/PVP/PGMA system showed excellent mechanical integrity. The particles did not detach at high temperature (during PS grafting) or in toluene under ultrasonic treatment. Scanning electron microscope images of the polyester fabric surface after the modification are illustrated in Fig. 17.12a. A typical static contact angle analysis was performed on polyester fabric modified with the ‘silver/PS’ approach and on a control fabric modified with only PS (no silver). The contact angle of the fabric increased from 113° + 4° (Fig. 17.12b) for control surface to 157° + 3° for PS/silver multilayer system (Fig. 17.12c). The increase in contact angle was due to the partial contact of water with PS and entrapped air between silver nanoparticles. This synergistic effect of the hydrophobicity of PS and the roughness caused by silver nanoparticles indeed resulted in a contact angle beyond the superhydrophobic boundary.

17.7

Conclusions

In this chapter, a series of recent results in surface modification of various surfaces employing the macromolecular anchoring layer approach was overviewed. It was demonstrated that the approach could be used as a virtually universal method for grafting of functional polymer brushes. The properties of the brushes can be controlled by polymer nature, structural and morphological factors, and external stimuli. The polymer grafting technique developed can be readily applied to surface modification of fibers and textiles, leading to generation of hydrophobic, hydrophilic and switchable fibrous materials.

17.8

Acknowledgments

The research presented has been supported by the National Science Foundation CTS-0456550 and DMR-0602528 grants, ERC Program of National Science Foundation under Award Number EEC-9731680, and Grants M01-CL03,

© 2007, Woodhead Publishing Limited

Nanofabrication of thin polymer films

467

C05-CL01 and C04-CL06 from the Department of Commerce through the National Textile Center. The author would like to acknowledge present and former members of his research group: Viktor Klep, K. Swaminatha Iyer, Bogdan Zdyrko, Yong Liu, John Draper, Karthik Ramaratnam, and Oleksandr Burtovyy, who conducted most of the research presented. The author also acknowledges the participation of Professors P. J. Brown (Clemson University), G. Chumanov (Clemson University), S. Minko (Clarkson University) and M. Stamm (Institute for Polymer Research Dresden, Germany) and their research groups in the work outlined above.

17.9

References

1. Jagur-Grodzinski, J., Heterogeneous Modification of Polymers: Matrix and surface reactions, John Wiley & Sons, Chichester, 1997. 2. Uyama, Y., Kato, K., Ikada, Y., Adv. in Polym. Sci. 1998, 137, p. 1. 3. Alexander, S., J. Phys. (Paris) 1977, 38, 983. 4. de Gennes, P.-J., Macromolecules 1980, 13, 1069. 5. Zhao, B., Brittain W. J., Prog. Polym. Sci. 2000, 25, 677. 6. Laub, C. F., Koberstein, J. T., Macromolecules 1994, 27, 5016. 7. Luzinov, I., Iyer, K. L. S., Klep, V., and Zdyrko, B., US patent 7,026,014 B2, 2006. 8. Husseman, M., Malmstrom, E. E., McNamara, M., Mate, M., Mecerreyes, D., Benoit, D. G., Hedrick, J. L., Mansky, P., Huang, E., Russell, T. P., Hawker, C. J., Macromolecules 1999, 32, 1424. 9. Jones, D. M., Brown, A. A., Huck, W. T. S., Langmuir 2002, 18, 1265. 10. Jones, R. A. L., Lehnert, R. J., Schonerr, H., Vancso, J., Polymer 1999, 40, 525. 11. Luzinov, I., Julthongpiput, D., Malz, H., Pionteck, J., Tsukruk, V. V., Macromolecules 2000, 33, 1043. 12. Kong, X., Kawai T., Abe, J., Iyoda, T. Macromolecules 2001, 34, 1837. 13. Köthe, M., Müller, M., Simon, F., Komber, H., Jacobasch, H.-J., Adler, H.-J., Colloids and Surfaces 1999, 154, 75. 14. Klep, V., Luzinov, I., Polymer Preprints 2002, 43(2), 164. 15. Iyer, K. S., Klep, V., Luzinov, I., Polymer Preprints 2002, 43(1), 455. 16. Liu, Y., Klep, V., Luzinov, I., Polymer Preprints 2003, 44(1), 564. 17. Fleer, G. J., Cohen Stuart, M. A., Scheutjens J. M. H. M., Cosgrove, T., Vincent, B., Polymers at Interfaces, Chapman & Hall, New York, 1993. 18. Iyer, K. S., Zdyrko, B., Malz, H., Pionteck, J., Luzinov, I., Macromolecules 2003, 36, 6519. 19. Zdyrko, B., Klep, V., Luzinov, I., Langmuir 2003, 19, 10179. 20. Klep, V., Zdyrko, B., Liu, Y., Luzinov, I., in ‘Polymer Brushes’, Advincula, Brittain, Caster, Ruhe editors, Wiley-VCH Verlag GmbH& Co., Weinheim, 2004, p. 69. 21. Liu, Y., Klep, V., Zdyrko, B., Luzinov, I., Langmuir 2004, 20(16), 6710. 22. Zdyrko, B., Varshney, S. K., Luzinov, I., Langmuir 2004, 20, 6727. 23. Jordan, R., Ulman, A., Kang, J. F., Rafailovich, M. H., Sokolov, J., J. Am. Chem. Soc. 1999, 121, 1016. 24. Advincula, R., Zhou, Q., Park, M., Wang, S., Mays, J., Sakellariou, G., Pispas, S., Hadjichristidis, N., Langmuir 2002, 18, 8672. 25. Zhao, B., Brittain, W. J., Macromolecules 2000, 33, 342.

© 2007, Woodhead Publishing Limited

468

Nanofibers and nanotechnology in textiles

26. Jordan, R., West, N., Ulman, A., Chou, Y. M., Nuyken, O., Macromolecules 2001, 34 1606. 27. Husseman, M., Malmstrom, E. E., McNamara, M., Mate, M., Mecerreyes, D., Benoit, D. G., Hedrick, J. L., Mansky, P., Huang, E., Russell, T. P., Hawker, C. J., Macromolecules 1999, 32, 1424. 28. Matyjaszewski, K., Miller, P. J., Shukla, N., Immaraporn, B., Gelman, A., Luokala, B. B., Siclovan, T. M., Kickelbick, G., Vallant, T., Hoffmann, H., Pakula, T., Macromolecules 1999, 32, 8716. 29. Luzinov, I., Minko, S., Senkovsky, V., Voronov, A., Hild, S., Marti, O., Wilke, W., Macromolecules, 1998, 31, 3945. 30. Prucker, O., Rühe, J., Langmuir 1998, 14, 6893. 31. Liu, Y., Klep, V., Zdyrko, B., Luzinov, I., Langmuir 2005, 21, 11806. 32. Kim, J.-B., Bruening, M. L., Baker, G. L., J. Am. Chem. Soc. 2000, 122, 7616. 33. Shah, R. R., Merreceyes, D., Husemann, M., Rees, I., Abbott, N. L., Hawker, C. J., Hedrick, J. L., Macromolecules 2000, 33, 597. 34. Matyjasewski, K., in Controlled Radical Polymerization 1998; ACS Symposium Series No. 685, ACS, Washington, DC, Chapter 16, 258. 35. Jeyaprakash, J. D., Samuel, S., Dhamodharan, R., Rühe, J., Macromol. Rapid. Commun. 2002, 23, 277. 36. Karim, A., Tsukruk, V. V., Douglas, J. F., Satija, S. K., Fetters, L. J., Reneker, D. H., Foster, M. D., J. Phys. II France 1995, 5, 1441. 37. Jones, R. A. L., Lehnert, R. J., Schonerr, H., Vancso, J., Polymer 1999, 40, 525. 38. Auroy, P., Auvray, L., Leger, L., Macromolecules 1991, 24, 5158. 39. Jordan, R., Ulman, A., Kang, J. F., Rafailovich, M. H., Sokolov, J. S., J. Am. Chem. Soc. 1999, 121, 1016. 40. Clarke, C. J., Jones, R. A. L., Clough, A. S., Polymer 1996, 37, 3813. 41. Clarke, C. J., Polymer 1996, 37, 4747. 42. Iyer, K. S, Luzinov, I. Macromolecules 2004, 37, 9538. 43. Wool, R. P., Polymer Interfaces: Structure and strength, Hunser Publishers, Munich, 1995, p. 102. 44. Anastasidis, S., Retsos, H., Pispas, S., Hadjichristidis, N., Neophytides, S., Macromolecules 2003, 36, 1994. 45. Russel, T. P., Science 2002, 297, 964. 46. Lahann, J., Mitragotri, S., Tran, T.-N., Kaldo, H., Sundaram, J., Chol, I. S., Hoffer, S., Somorjai, G. A., Langer, R., Science 2003, 299, 371. 47. Zhao, B., Brittain, W. J., Macromolecules, 2000, 33, 8813. 48. Minko, S., Patil, S., Datsyuk, V., Simon, F., Eichhorn K.-J., Motornov, M., Usov, D., Tokarev, I., Stamm, M., Langmuir, 2002, 18, 289. 49. Minko, S., Usov, D., Goreshnik, E., Stamm, M., Macromol. Rapid Commun. 2001, 22, 206. 50. Sidorenko, A., Minko, S., Schenk-Meuser, K., Duschner, H., Stamm, M., Langmuir 1999, 15, 8349. 51. Luzinov, I., Minko, S., Tsukruk, V., Prog. Polym Sci., 2004, 29(7), 635. 52. Draper, J., Luzinov, I., Minko, S., Tokarev, I., Stamm, M., Langmuir 2004, 20, 4064. 53. Klep, V., Minko, S., Luzinov, I., Polymeric Mater. Sci. Eng. 2003, 89, 248. 54. Iyer, K. S, Luzinov, I., in ‘Responsive Polymer Materials: Designs and applications’, Sergiy Minko editor, Blackwell Publishing, Ames, 2006, p. 101. 55. Barthlott, W., Neinhuis, C., Planta 1997, 202, 1. 56. Von Baeyer, H. C., The Sciences 2000, January/February, 12.

© 2007, Woodhead Publishing Limited

Nanofabrication of thin polymer films

469

57. Miwa, M., Nakajima, A., Fujishima, A., Hashimoto, K., Watanabe, T., Langmuir 2000, 16, 5754. 58. Youngblood, J. P., McCarthy, T. J., Macromolecules 1999, 32, 6800. 59. Quere, D., Nature Mater. 2002, 1, 14. 60. Wenzel, R. N., Ind. Eng. Chem. 1936, 28, 988. 61. Cassie, A. B. D., Baxter, S., Trans. Faraday Soc. 1944, 3, 16. 62. Nakajima, A., Hashimoto, K., Watanabe, T., Chem. Monthly 2001, 132, 31.

© 2007, Woodhead Publishing Limited

18 Hybrid polymer nanolayers for surface modification of fibers S. M I N K O and M. M O T O R N O V, Clarkson University, USA

18.1

Introduction: smart textiles via thin hybrid films

A combination of several ingredients in the same hybrid material is useful in achieving a number of desirable properties, sometimes contra-properties. Very often we would like to switch between different properties of the material depending on the situation. Such behavior will be valuable for textiles, especially for cloth. Depending on environmental conditions or the state of our body the textile may change its properties to adapt to the situation in a desired way. These kinds of textiles are and cloth are known as ‘smart textiles/cloth’. The term ‘smart textile’ refers to two different major functions: (1) responsive textiles that change their physical/chemical properties upon external stimuli and (2) ‘information/communication textiles’ with built-in sensors that monitor the human body and environmental conditions and provide, send and exchange information. This chapter deals with the first type. However, it is noteworthy that the smart textile will combine both kinds of function and a suit of the future will use the information and responsive functions to demonstrate a smart response regulated by a computer chip. Switching of physical and chemical behavior of smart textiles is needed to regulate the temperature of the body and the humidity of the skin surface to evacuate excessive perspiration, to protect the skin from irradiation, aggressive chemicals and microbes (including chemical and biological threats), and to protect the cloth from becoming contaminated with dust and stains of foods, beverages and oil. At the same time the textile should be stable, resistant to washing/cleaning, demonstrate good feel and be attractive. That is a very complex task and doubtless such a textile will have a complex structure. In this case hybrid material is a key word. The use of an appropriate design of hierarchically arranged functional elements which are capable of rearranging upon external stimuli/signals is the way to the solution of this complex task. All levels of the organization in textile materials should be considered: structure of fibers, yarns, woven fabrics and their coatings. This chapter focuses on 470

© 2007, Woodhead Publishing Limited

Hybrid polymer nanolayers for surface modification of fibers

471

one of the structural elements of the textile: thin hybrid coatings. Although thin coatings cannot perform all the above functions of the smart textiles, the responsive coating is a very important component of complex smart materials to regulate wetting, adhesion and permeability.

18.2

Mechanisms of responsive behavior in thin polymer films

Materials respond to external stimuli via various mechanisms.1 For example if we set a match to a thin organic film the film will be burned out immediately. This kind of response appears uninteresting. However, if the film contains iodine isotopes which are captured in the smoke and reach a radioactivity detector, this mechanism can be used to activate a fire alarm. Thus, a definition of responsiveness depends on the type of response being produced. Here we consider responsiveness as an ability to reverse and adjust the changes of textile properties which help to regulate transport of aqueous solutions, dispersions and emulsions (of various origins) through the textile materials. We consider the application of thin polymer films (typically from 5 to 100 nm thick films) for such a regulation as a robust approach exploring interfacial interactions. Thin polymer films could switch surface composition upon external stimuli.2 Consequently, all relevant properties of the thin film such as wetting, adhesion, adsorption and reactivity will be switched at the same time. In this way a range of very important functions of textiles (water permeability, self-cleaning, stain resistance, antimicrobial properties) can be regulated and adjusted.

18.2.1 Responsiveness of polymer chains to their environment Responsiveness of thin polymer films is based on a phase segregation mechanism.2 It can be either intramolecular segregation of homopolymers and copolymers with a complex architecture when different segments of blocks are incompatible and segregate to avoid unfavorable interactions, or intermolecular segregation when two or more dissimilar polymers segregate to microscopic or macroscopic phases. The phase segregation in thin films can be regulated by their environment represented in terms of a solvent quality change and confinement effects of the interfaces. For example, two samples of a dry and a wet thin film can be considered as the materials introduced in two different environmental conditions created by two different solvents: air for the dry sample (a very poor solvent) and water for the wet sample. A change of pH or salt concentration in the aqueous solution which wets the film may affect solubility of polymers in water. Interfaces (polymer–substrate, polymer–air or polymer–water) may greatly alter the properties of the thin

© 2007, Woodhead Publishing Limited

472

Nanofibers and nanotechnology in textiles

polymer films. In a typical example one of the polymers (or polymer segments) preferentially adsorbs at the interface. Changes of the film environment may change polymer adsorption at the interface. Hydrophilic segments will adsorb at the polymer film–water interface, while hydrophobic segments tend to adsorb at the polymer film–dry air interface. Humidity may also affect the adsorption. The driving force for the change in the phase segregation is the change in balance between polymer–polymer and polymer–environment interactions due to the changes in the thin film environment. Polymer chains are very sensitive to their environment. This can be illustrated by conformational transition of single poly(2-vinylpyridine) (P2VP) chain vs. pH of aqueous solutions studied using in situ atomic force microscopy (AFM) (Fig. 18.1).3 P2VP is a weak polyelectrolyte, thus the degree of ionization depends on pH. At low pH values, (pH < 3) P2VP chains are highly protonated and possess an extended coil conformation (Fig. 18.1a) due to electrostatic repulsion of charged segments. An increase of pH value close to pH 4 results in the coil-to-globule phase transition when the P2VP chains form compact conformations (Fig. 18.1c). The ionization degree drops 2

4

(b) pH 4.04

1

0

pH 4.24

60 50 40 30

Chains

2

nm

100 nm

70

< r 2 >1/2 [nm]

1

(c)

(d)

80

3

100 nm

90

≥5

20 10

Fraction in loops [%]

(a) pH 3.89

15

3.0

3.5

4.0

4.5

pH 5.0

3.0

3.5

4.0

4.5

pH

(e)

10

5

100 nm 5.0

18.1 AFM-visualized conformations of adsorbed P2VP molecules: (a) pH 3.89, extended coils; (b) pH 4.04, intermediate state; (c) pH 4.24, compact coils. Z-scale bar shows a number of superposed chains assuming the height increment of 0.4 nm. (d) Directly measured values of root mean square (rms) end-to-end distance of P2VP single molecules adsorbed on mica surface versus pH. (e) Fraction of monomer units in loops versus pH. The gray zone in (d) and (e) is a pH range of the conformation transition. Reprinted with permission from Ref. 3. Copyright 2003 American Chemical Society.

© 2007, Woodhead Publishing Limited

Hybrid polymer nanolayers for surface modification of fibers

473

and the hydrophobic effect pushes the hydrophobic backbone to attain a compact conformation. The transition is very sharp. The characteristic size of the molecules (gyration radius) changes by a factor of two (Fig. 18.1d). In thin films polymer chains are highly concentrated. Assemblies of polymer chains respond cooperatively to their environment and form segregated domains. The sizes of the domains depend on interfacial tension and on the effects of constraints (for example, for chains grafted to a solid substrate, the polymer segments cannot move from the grafting point for a distance larger than the end-to-end distance of the polymer chain). Below we consider how the phase segregation mechanisms in thin polymer films can be explored for the fabrication of responsive materials.

18.2.2 Polymer brushes The systematic study of the surface restructuring of polymer materials began two decades ago.4–7 It was shown that the surfaces of polymer materials can rearrange and adapt their interfacial composition to approach a minimal interfacial tension. For example, polymers with specially tailored architecture (end functional groups, copolymers, segmented polymers, block– copolymers)8–13 expose to the interface their polar fragments in contact with water, and non-polar fragments in air. For bulk polymers the restructuring is a slow process (from minutes to hours). Much faster reorganization can be obtained for thin polymer films. Solvents soak into or evaporate from monolayers of polymer chains (of moderate molecular weight) very quickly (time scale of seconds). However, thin polymer films are not stable. They typically de-wet substrates and can be easily and rapidly dissolved. The solution to the problem was approached by grafting polymers to the substrate surface. The best result was found with polymer brushes. Polymer brushes are monolayers of end-tethered polymer chains (Fig. 18.2), in which the distance between grafting points is smaller than the chain end-to-end distance. That causes polymer chains to stretch away from the grafting surface due to the excluded volume effect. The stretching degree is balanced by entropic

Solvent quality

Good solvent (a)

Poor solvent (b)

18.2 Homopolymer brush constituted of end-tethered chains stretched in good solvent (a), and collapsed in poor solvent (b).

© 2007, Woodhead Publishing Limited

474

Nanofibers and nanotechnology in textiles

Swollen thickness [µm]

elasticity of the polymer coils.14 Polymer brushes are stable thin films because they are covalently (sometimes electrostatic or hydrophobic effects are used for the grafting) bound to the substrate. Brushes respond to their environment rapidly via conformational changes. Polymer chains of different architectures were used to design polymer brushes.15,16 Polymer brushes made up of homopolymer chains (homopolymer brushes) respond to their environment by expansion–contraction in good and poor solvents, respectively. At moderate grafting densities, homopolymer brushes segregate into spherical domains in a poor solvent, which allows them to minimize unfavorable interactions with the solvent. Thus, the transition between a good and a poor solvent is the driving force of the brush responsiveness. Polyelectrolyte homopolymer brushes respond to electrostatic interactions (Fig. 18.3)17 in a similar way to that of bulk polyelectrolyte chains. (This difference is caused by the tethered chain architecture and a high local concentration of ionizable groups in polyelectrolyte brushes which affects the local ionization degree.18) Although homopolymer brushes demonstrate very pronounced responsiveness in terms of switching their properties between good and poor interactions with their environment, the composition of the interface is not changed qualitatively. In other words, independent of the environment, the polymer exposed to the interface remains unchanged in its chemical nature. Thus, the switching of the thin film properties is limited by the constitution of the given polymer.

2.0

1.5

1.0

0.5 2

4

6 pH

8

10

18.3 Swollen thickness of a 20 nm (dry thickness) poly(methacrylic acid) (PMAA) brush on a La-prism as a function of the pH of solution (‘salt-free’). The brush was prepared by polymerization of methacrylic acid (bulk; t = 1.5 h; T = 60 °C). After the polymerization the PMAA brush was extracted with methanol and water for 15 h. The solid line is a guide to the eye. Reprinted with permission from Ref. 17. Copyright 2002, American Institute of Physics.

© 2007, Woodhead Publishing Limited

Hybrid polymer nanolayers for surface modification of fibers

475

The largest response can be obtained if two very different polymers or polymers and nanoparticles are combined in the same hybrid thin film. In this case, the difference in the response of two different species may result in a very strong alternation of the thin film properties because either one or another polymers could be exposed to the top of the thin film. We discuss the mechanisms of responsiveness of such systems below.

18.2.3 Mixed polymer brushes In mixed brushes (Fig. 18.4)19 two or more different polymers are grafted to the same substrate. The mechanism of responsiveness of the mixed brushes originates from the dependence of the incompatibility of two polymers on their environment. Polymers in the mixed brush segregate into nanoscopic phases. The size of the nanophase is related to the size of polymer molecules. In a nonselective solvent, the phase segregation is a lateral segregation when unlike polymers form spherical (dimples) or elongated (lamellar-like or ripples) clusters tethered to the substrate. Both of the polymers are exposed on the top of the brush (Fig. 18.4a). If selectivity of a solvent (solvent dissolves one polymer better than another polymer) increases, the mixed brush structure tends to form a layered structure (Fig. 18.4b) and the resulting morphology is a combination of lateral and layered segregation mechanisms.20 In selective

(a)

(b)

18.4 Schematic illustration of two possible morphologies of mixed brush irreversibly grafted to solid substrates (cross-section of the layer): lamellar morphology in a nonselective solvent (a); cluster morphology in a poor solvent for the black chains (b). Reprinted with permission from Ref.19. Copyright 2003 American Chemical Society.

© 2007, Woodhead Publishing Limited

476

Nanofibers and nanotechnology in textiles

solvents one polymer preferentially segregates to the top of the brush, while another polymer forms clusters segregated to the grafting surface. The most important difference of the mixed brush response compared with the homopolymer brush is that not only the transitions between stretched and collapsed conformations but also a change of the composition profile takes place. In other words, the surface composition of the brush is switched by changes in its environment (Fig. 18.5). The responsiveness of mixed polyelectrolyte brushes is modified by electrostatic interactions which can be used to regulate the mechanism of the phase segregation. Weak mixed polyelectrolyte brushes are of special interest since the electrostatic interactions can be affected by pH and ionic strength of aqueous environment and the surface composition of the mixed polyelectrolyte brush can be switched just by a change of external pH.21–23 Typically, two or more unlike polymers are grafted randomly to the substrate in mixed polymer brushes. A special case is represented by Y-shaped amphiphilic brushes which combine two dissimilar polymer chains attached to a single focal point or a leg.24, 25 Although the responsive behavior of 3.0

2.6

(φA – φB)/(φA + φB)

ζ = – 0.3 ζ = –0.2 ζ = –0.1 ζ = –0.03

0.0 –0.2 –0.4 0.0

2.4

0.5

1.0 z /R e

Dimple A

1.5

Ripple

Incompatibility χ

2.8

0.2

Dimple B

2.2 dis δ = 0.46 2.0

– 0.2

– 0.1 0 0.1 Solvent selectivity ζ

0.2

18.5 Phase diagram of a mixed brush as calculated in self-consistent field theory as a function of solvent selectivity. In addition to a laterally homogeneous (disordered) phase (below the line marked with dis), the phase diagram comprises a ripple phase (see Fig. 18.4a) and dimples (see Fig. 18.4b). The inset presents the laterally averaged composition profile expressed by fraction of monomers A(ψA) and B(ψb) vs. extension of the brush (Z) normalized by the Gaussion chains end-to-end distance (Re); δ-is the degree of chain stretching in the brush. Reprinted with permission from Ref. 20. Copyright 2002, American Physical Society.

© 2007, Woodhead Publishing Limited

Hybrid polymer nanolayers for surface modification of fibers

477

mixed and Y-shaped brushes can be explained by the same mechanism, the difference between randomly mixed polymer brushes and Y-shaped brushes is in the polymer distribution on the surface. The Y-shaped brushes could form more regular structures because the grafting of two different polymers is always coupled due to the Y-shaped architecture.

18.2.4 Block–copolymer brushes In block–copolymer brushes (Fig. 18.6)26 two or more chemically different polymers (typically two or three different blocks) constitute a polymer brush with block–copolymer architecture. Responsiveness of these brushes is determined by phase segregation of unlike polymer blocks; however, the structure of the brush layer depends on whether the AB block copolymer is tethered by the more (A) or the less (B) soluble block.27, 28 In poor solvents

Immersion in CH2Cl2

Gradual addition of cyclohexane

18.6 Morphology of the tethered poly(styrene–block– methylmethacrylate) brush. Reprinted with permission from Ref. 30. Copyright 2000 American Chemical Society.

© 2007, Woodhead Publishing Limited

478

Nanofibers and nanotechnology in textiles

the brushes grafted by the B blocks form dumbbell-like micelles with the Bpolymer segregated to the grafting surface and the A-polymer micelles on the top. The brushes grafted by the A blocks form B-polymer micelles shielded by A-polymers. If solvent selectivity is improved for the A-polymers, the dumbbell-like micelles of the B-grafted brushes are transformed to structures that are similar to the mixed brushes in selective solvents. The B-polymers form pinned micelles which are congregated at the grafting surface. The micelles are shielded by the brush of the A-polymers. The block–copolymer brushes have a complex phase diagram affected by the ratio between the molecular weights of the A and B blocks and the grafting density.29–32

18.3

Polymer–polymer hybrid layers

Mixed and block–copolymer brushes are polymer–polymer hybrid thin films tethered to solid substrates. The well-defined architecture of end-tethered chains in polymer brushes allows for a deep study of the mechanisms of the brushes’ response to their environment. Nevertheless, for practical applications a very similar mechanism can be approached using polymer layers when a polymer can be grafted by end groups, and also by side functional groups. The polymer chains randomly grafted by side functional groups will form loops and tails exposed to the solvent. Thus, two or more unlike polymers grafted by side chains will form polymer–polymer hybrid thin films with properties very similar to those for the mixed brushes. Practically, many combinations of polymer layer architectures could be used to design responsive hybrid layers (end tethers polymers A and grafted via side functional groups polymers B, two- and tri-block copolymers grafted via side functional groups of one of the blocks, etc.). Although many versions exist, in this chapter we will focus on the brush-like architecture which gives us a clear illustration of the responsiveness. There are two main approaches employed for synthesizing mixed polymer brushes: ‘grafting to’33 and ‘grafting from’.34 The fabrication of binary mixed brushes consists of two repeating steps of the grafting procedure; the grafting of the first polymer is followed by the grafting of the second polymer. The ‘grafting to’ approach to synthesize binary polymer brushes is based on subsequent grafting of end-functionalized polymers to a solid substrate. Before grafting, a surface treatment is usually applied to introduce appropriate functional groups onto the substrate surface. A plasma treatment was used to introduce amino or hydroxyl functional groups on polymeric substrates35, 36 or ω-functional silanes to introduce epoxy, amino and hydroxyl functional groups onto surfaces of inorganic oxides.33, 34 Polished Si-wafers are frequently used to study the grafting polymers using ellipsometry. As an example, we describe here the grafting of a P2VP and polystyrene (PS) mixed brush. The

© 2007, Woodhead Publishing Limited

Hybrid polymer nanolayers for surface modification of fibers

479

first step consists of chemisorption of 3-glycidoxypropyltrimethoxysilane (GPS) on cleaned Si-wafer (Si-wafer is protected by thin 1–2 nm native silica coating), which forms a layer with a very high concentration of epoxy groups on the surface. Its ellipsometric thickness around 0.7–1.1 nm corresponds to 1–1.5 theoretical monolayers of GPS. The second and the third steps consist of the grafting of carboxyl terminated poly(2-vinylpyridine) (P2VPCOOH) and polystyrene (PS-COOH), respectively. The polymers are grafted from melt. A thin 10–50 nm PS-COOH film is spin-coated on the substrate and the polymer is grafted by annealing the sample at a temperature higher than the glass transition temperature (Tg) of the polymer. The non-grafted polymer is removed by Soxhlet extraction with tetrahydrofuran (THF). Afterwards, a 50 nm thick film of the second polymer P2VP-COOH is spin-coated on top of the PS-COOH brush and grafted upon heating above Tg. The kinetics of grafting of PS-COOH and P2VP-COOH in terms of the ellipsometric thickness of the layer can be estimated by sampling and measuring the thickness of the grafted brush. Most of the polymer is grafted in 2 h; afterwards the grafting slows down, approaching a plateau value in 16–20 h. In this example, the plateau values for homopolymer PS and P2VP brushes are around 7 nm. The grafting kinetics is useful to regulate the mixed brush composition. If we stop grafting PS-COOH as soon as we approach 3.5 nm thickness of the homopolymer brush, then we graft the second polymer P2VP-COOH for 16 h to approach the 7 nm thick mixed brush. Finally, the composition of the mixed brush is 1:1. Typically, the second polymer can be successfully grafted only if the first polymer has much smaller affinity to the solid substrate than the second. In this case the strong affinity of the second polymer to the solid substrate acts as a driving force for chains of the second polymer to penetrate the brush layer formed by the first polymer. Binary brushes that are synthesized by the ‘grafting to’ method are macroscopically homogeneous but they exhibit phase segregation on a nanoscopic length scale. The composition of the brush (the amount of each polymer grafted) can be controlled by conditions of the grafting procedure: time, temperature and thickness of the spin-coated films. The drawback of the ‘grafting to’ procedure is that only a relatively small amount of polymer can be grafted onto the surface, although the number of grafted chains of polymers is much smaller than the number of functional reactive groups present on the surface. Diffusion of polymer chains to the surface is strongly limited by the already grafted polymer chains. The maximal thickness of binary brushes that could be achieved by the ‘grafting to’ method typically is around 10 nm, depending on molecular mass of the grafted polymers. Nevertheless, this method is very simple and robust. For many practical applications the grafting density achieved by the ‘grafting to’ method provides very pronounced responsive properties of the thin film.

© 2007, Woodhead Publishing Limited

480

Nanofibers and nanotechnology in textiles

Mixed polymer brushes of a very high grafting density can be prepared using the ‘grafting from’ approach by sequential one-after-another grafting of two incompatible polymers. Here we describe an example of the ‘grafting from’ procedure on the surface of polyamide (PA) plates and on the surface of a PA textile for the synthesis of PS and P2VP mixed brushes. Polyamides as semicrystalline thermoplastics have found numerous applications, particularly for the fabrication of excellent fibers due to their good thermal stability, flexibility and mechanical properties. Surface modification of PA is a widely used approach to regulate properties of fiber reinforced materials, textiles, etc. Responsive properties of a PA surface would extend the application of PA, particularly for cloth and biomedical materials. The aim of this example was to fabricate PA-based materials that change surface characteristics in response to environmental conditions. One of the possible routes to approach this goal comprises the grafting of mixed polymer brushes from a PA surface which can introduce adaptive and switching behavior in different surrounding media. Here we describe the ‘grafting from’ PA-6 surfaces via free radical polymerization initiated by the thermal decomposition of an azo initiator (the chloroanhydride derivative of 4,4′-azobis(4-cyanopentanic acid) – abbreviated as ACPC) covalently attached to the PA substrate. To attach the azo initiator to the PA substrate, the PA surface was treated with low-pressure ammonia plasma. NH3 plasma introduces N-containing functionalities such as amino (—NH2), imino (—CH==NH), cyano (—C≡≡N) and other functional groups, and in addition oxygen-containing groups such as amido (—CONH2) and hydroxyl groups due to post-discharge atmospheric oxidation. We used those functional groups to covalently bond the azo initiator to PA surfaces. Each step of the surface modification was controlled with ellipsometry, X-ray photoelectron spectroscopy (XPS), AFM, Fourier transform infrared spectroscopy in attenuated total reflection (FTIR-ATR) and contact angle measurements. The PA samples were treated with NH3 plasma for 60 s. The ratio between the number of atoms of oxygen and carbon ([O]:[C]) and nitrogen and carbon ([N]:[C]) in PA-6 calculated from the chemical composition equals 0.167. The surface composition determined from the XPS spectra gives the ratios [O]:[C] = 0.15 and [N]:[C] = 0.14, which is in good agreement with the chemical composition of the substrate. The plasma-treated samples of PA-6 show increased ratios [N]:[C] = 0.19, which prove the introduction of the Ncontaining functionalities. The azo initiator was covalently bound to the PA surface via the reaction of the amino and hydroxyl groups with ACPC. The reaction is well reproducible. The resulting layer of the initiator is about 2.1 ± 0.3 nm thick (as measured with ellipsometry in the reference experiment on the thin PA film deposited on the Si-wafer). This value corresponds to 5.7 × 10–6 mol/m2 surface concentration of the initiator and to 0.5 nm average distance between the grafted initiator molecules.

© 2007, Woodhead Publishing Limited

Hybrid polymer nanolayers for surface modification of fibers

481

Grafting of PS chains was performed by in situ radical polymerization initiated by the thermal decomposition of the azo initiator covalently attached to the PA surface. The amount of the grafted polymer and the residual initiator on the surface was regulated by polymerization time. The ungrafted polymer was washed out by a cold Soxhlet extraction. In the second polymerization step the residual amount of the azo initiator was used to carry out the graft polymerization of 2-vinylpyridine. The grafting of both polymers was proved with FTIR-ATR performed with the PA plates tightly pressed to the surface of the attenuated total reflection prism. The very well-pronounced differences in the spectra of individual polymers at 1400–1750 and 2750–3200 cm–1 allowed us to analyze the layer composition at least qualitatively. The characteristic bands of aromatic and aliphatic groups observed for the mixed brushes (obtained by the subtraction of the reference spectrum of the PA substrate from spectrum of the grafted brush) provide evidence for the grafting of PS and P2VP. In the spectra, we identified very pronounced PS bands at 1601 and 1493 cm–1 and the characteristic bands of P2VP at 1568 and 1590 cm–1. Responsive properties of the mixed brushes were investigated using contact angle experiments. The same samples of PA with the grafted PS/P2VP mixed brushes were exposed for 5 min to solvents of different thermodynamic quality for the polymers. After each treatment with a particular solvent the samples were dried in a flow of nitrogen and used for the contact angle investigations. The experiments were repeated several times with each sample to prove the reversibility of the switching of surface properties. In these experiments, we assume that the morphology of the dry film is directly correlated with the structure of the swollen film. The time of switching in a particular solvent is in the order of minutes (contact angle changes in 1–2 min and approaches to equilibrium in 5–10 min), which is much longer than the time to dry the film under nitrogen flux (several seconds). We may assume that we freeze the film morphology during solvent evaporation. At ambient conditions the polymers in the dry film are in a glassy state and the film morphology is stable for a long period of time. The switching of morphologies upon exposure to different solvents is affected by the phase segregation at the nanoscopic scale as discussed above. The contact angle measurements have shown very pronounced switching of the surface energetic state. For example, if we expose the sample to toluene, the top of the layer is preferentially occupied by PS. In this case the contact angle approaches the value of 90°, while in ethanol and water (pH 3.0) the surface is dominated by P2VP with contact angles of 60° and 35°, respectively. The data clearly show that a top layer of the binary brush switches from the hydrophobic to the hydrophilic energetic state and vice versa upon exposure to selective solvent for one of the polymers. The contact angles obtained on the mixed brush of 90° and 60° after toluene and ethanol, respectively, correspond to the same values of the contact

© 2007, Woodhead Publishing Limited

482

Nanofibers and nanotechnology in textiles

angles on the model PS and P2VP single homopolymer brushes, respectively. This fact is evidence that in a selective solvent the top layer of the mixed brush is formed due to the layered (perpendicular) segregation (Fig. 18.4b) and the top is occupied by the favorite polymer. In acidic water P2VP is protonated and charged. This layer is wetted by water much better than neutral P2VP. Wetting in this case is promoted by dissociation of the protonated P2VP below the water drop. In the cases of nonselective solvents after exposure to chloroform or THF, both polymers are present on the top of the film (contact angle 80°) (Fig. 18.4a). Using the Cassie equation, we calculated that this contact angle corresponded to the 65% PS fraction on the top of the brush when the surface of the brush is constructed from laterally segregated domains of PS and P2VP. Here we extend the study of the switching behavior on substrates with a complicated texture and present the results of contact angle measurements on the PA textile with the grafted PS/P2VP binary mixed brushes, comparing them with the wetting behavior on the PA plates with the same grafted brush. The fibers of about 200 µm in diameter forming the textile introduce an effect of a composite surface37 where the drop of water is in contact partially with the surface of PA fibers and partially with air. In this case we observed the much more pronounced switching effect amplified by the surface texture of the PA textile (Fig. 18.7). The exposure of the textile sample to toluene results in the highly hydrophobic properties of the material with an advancing

(a)

(b)

(d)

(c)

(e)

18.7 Video images of drops on substrates with grafted PS-P2VP brush from PA-6 textile: after exposure to toluene, Θ = 150° (a); ethanol, Θ = 50° (b); water, pH 3, wicking regime (c); from PA-6 plates after exposure to toluene, Θ = 90° (d) and water, pH 3, Θ = 20° (e). Reprinted with permission from Ref. 35. Copyright 2003 American Chemical Society.

© 2007, Woodhead Publishing Limited

Hybrid polymer nanolayers for surface modification of fibers

483

contact angle of 152°, while upon exposure to ethanol the advancing contact angle is 50°, and after treatment with acidic water the film is fully wetted and water soaks into the textile sample. These pictures demonstrate that, in contrast, the flat PA surface is less hydrophobic (90°) and less hydrophilic (20°) upon exposure to toluene and acidic water, respectively. Schematically this phenomenon is outlined in Fig. 18.8. The drop of water deposited on the mixed brush coated textile is in contact with the textile fibers and air. Each fiber is a twist of single fibers forming a yarn with a rough surface. Thus, the textile surface has a complicated hierarchical texture. Depending on the bare contact angle, on a flat surface the liquid might fill all the grooves of the rough substrate or might be in contact with the upper part of the relief and air can be trapped below a drop. The wicking criterion is determined as follows:38 cos Θ >

(1 – φs ) ( r – φs )

Θ0 < Θ

[18.1]

where Θ0 is the bare contact angle of water on a flat surface, φs is the solid surface fraction assigned with the upper part of the relief (in this case it is a fraction of the total area that is not in contact with the liquid), and r is the ratio between the increased contact area of the rough surface and the corresponding projected area. An example of this regime is shown in Fig. 18.7c. If the bare contact angle is smaller than 90° but the wicking criterion is not fulfilled, the liquid fills the grooves of the rough surface only below a

(a)

(b)

18.8 Schematic representation of a drop of water on PA textile: Θ >90°, Θ = 150°; air is trapped below the drop giving the Cassie regime if the brush was switched in a hydrophobic state (a). Θ <90°, water soaks into the textile sample if the brush was switched in a hydrophilic state (b). Reprinted with permission from Ref. 35. Copyright 2003 American Chemical Society.

© 2007, Woodhead Publishing Limited

484

Nanofibers and nanotechnology in textiles

droplet contacting with fibers and the contact angle corresponds to Wenzel’s regime:39 cos Θ =

r (γ vs – γ ls ) = r cos Θ 0 γ

[18.2]

where Θ is the contact angle of water on a rough surface. An example of this regime is shown in Fig. 18.7b. If the bare contact angle is larger than 90°, air can be trapped below a drop and the liquid is only in contact with the upper part of the relief of the rough surface of fibers, resulting in the Cassie wetting regime given by:40, 41 cos Θ = –1 + φs(1 + cos Θ0)

[18.3]

where φs is the fraction of the upper part of the relief (in this case, that is the fraction of the total area in contact with liquid), which itself depends on Θ0. An example of this regime is shown in Fig. 18.7a. Consequently, switching of the mixed brush on the surface of the textile material causes the transition between the Cassie regime (Fig. 18.8a, when the contact angle is larger than 90°) and the wicking regime (Fig. 18.8b, when wetting is characterized by a small contact angle and the wicking criterion is fulfilled). Therefore, the textile material with the grafted mixed brush demonstrated the behavior affected by the combination of two possible approaches to regulate surface wetting: chemical composition and roughness.

18.4

Polymer–particles hybrid layers

Direct grafting mixed or block–copolymer brushes on surfaces of various fibers and textile materials is a slow chemical process which cannot be easily introduced in existing technologies for treatment of fiber and textile surfaces. The situation was substantially improved with the following simple and elegant approach. The mixed or block–copolymer brushes can be grafted on the surface of colloidal particles, for example silica particles. Afterwards, the particles can be deposited on fibers and textiles from colloidal dispersions applying well-developed technologies for deposition of pigments on textile materials. In this approach the responsive particles (or smart particles) are used for the simple and robust modification of polymer surfaces.42 Here we describe an example of the fabrication and investigation of smart responsive nanoparticles by grafting block–copolymers. We grafted triblock copolymer of poly(styrene-b-2-vinylpyridine-b-ethyleneoxide) (P(S-b-2VPb-EO) to silica particles 200 nm in diameter (Fig. 18.9). The particles were modified by 11-bromoundodeciltrimethoxisilane (BUDTMS), then the block– copolymer was grafted by a quaternization reaction to the particle surface. The grafting of the block–copolymer to the silica nanoparticles was proved by FTIR using the diffuse reflection technique. Very well-pronounced

© 2007, Woodhead Publishing Limited

485

N

Hybrid polymer nanolayers for surface modification of fibers

O Si

N

Br O Si

Br

18.9 Tri-block copolymer grafted to the silica particles via a quaternization reaction between pyridine rings of the central P2VP blocks and 11-bromoundodeciltrimethoxisilane on the silica particle.

differences in the spectra of individual blocks of the block–copolymer allowed the analysis of the chemical composition of the smart silica particles. The characteristic bands of aromatic and aliphatic groups observed for the block– copolymer brushes were obtained by subtraction of the spectrum of the particles with chemosorbed BUDTMS from the spectrum of the particles with the grafted block–copolymer. In the spectra, we identified very pronounced PS block bands at 1601 and 1493 cm–1, the characteristic bands of P2VP blocks at 1568 and 1590 cm–1, and the characteristic bands of the PEO block at 1648 and 3378 cm–1, which provided evidence for the grafting of P(S-b2VP-b-EO). The quantitative analysis of FTIR data was used to estimate the grafted amount and the grafting density of the P(S-b-2VP-b-EO) on the silica surface as 6.7 mg/m2 and 0.066 nm–2, respectively. In the next step, the smart particles were deposited on the substrate surface. In a reference experiment we used Si-wafers modified with the BUDTMS (Fig. 18.10). In this case, the smart particles were covalently attached to the Si-wafers via quaternization reaction between P2VP blocks and BUDTMS. However, the particles just physisorbed on Si-wafers or on the surface of PA textiles have demonstrated the same responsive behavior as chemically grafted particles. The data of contact angle measurements on these surfaces demonstrate well-pronounced switching from hydrophobic to hydrophilic wetting behavior (Table 18.1) upon exposure to different solvents. This behavior is similar to the behavior of block–copolymer brushes on a flat surface. However, the range of switching on the particle coated substrate is much larger than on the flat surface owing to the higher roughness of the material.

18.5

Hierarchical assembly of nanostructured hybrid films

The successful approach to smart responsive particles was further transformed into the principle of hierarchical self-assembly in the system when the particle

© 2007, Woodhead Publishing Limited

Nanofibers and nanotechnology in textiles le artic Si p

486

Br N Br

Br

N N Br

Br

Si

Si

O

O

O

Si

Si

Si

O

Si

O

Si

(a)

Si-wafer (b)

18.10 Schematic representation of grafting silica nanoparticles with tri-block copolymer brushes to the Si-wafer surface (a); a layer of the nanoparticles on Si-wafer (b). Table 18.1 Wetting of block–copolymer brushes grafted to the nanoparticles on silica wafers Sample

P(S-b-2VP-b-EO)

Contact angles of water (°) Toluene

CH2Cl2

Ethanol

Water, pH 7

Water, pH 3

131

117

57

69

29

core was comparable in size or smaller than the thickness of the grafted mixed brushes. Recently, we reported on a unimolecular spherical mixed brush represented by the P2VP7-PS7 copolymer, consisting of seven poly(2vinylpyridine) and seven polystyrene arms emanating from the same core.43 The core of the unimer is formed by 106 cross-linked divinylbenzene molecules. The PS and P2VP arms (Mw(PS-arm) = 20 000 g/mol and Mw(P2VP-arm) = 57 000 g/mol, where Mw is the weight-average molecular weight) are randomly grafted to the core. The core is highly cross-linked and can be considered to consist of dense particles 3–6 nm in diameter embedded into the shell formed by PS and P2VP arms. The unimolecular spherical mixed brush demonstrates similar principles to that of environment-induced response. In an acidic environment at low concentrations, the copolymer forms unimolecular micelles with the PS segregated to the core and the protonated P2VP shell. Upon exposure to toluene, P2VP7-PS7 undergoes the inverse intramolecular

© 2007, Woodhead Publishing Limited

Hybrid polymer nanolayers for surface modification of fibers

487

segregation: the P2VP arms form a dense core surrounded by the swollen PS shell. We applied this mixed spherical brush for grafting onto flat substrates.44 The surface of silica substrate was chemically modified with Br-alkyl silane. Then the grafting was performed by a quaternization reaction of P2VP arms with surface Br-alkyl groups. The fabricated surfaces demonstrated an example of hierarchical transitions of the thin film morphology. This complex mechanism was approached using a very simple synthetic route to the responsive material via a one-step grafting procedure in contrast to much more complex procedures reported for the synthesis of mixed brushes. To investigate the responsive behavior of the star–copolymer (smart particles) grafted layers, the samples were exposed to selective (toluene, ethanol, pH 2 water) and less selective (chloroform) solvents for 10 min. After the exposure to each solvent, the samples were dried and investigated with AFM, XPS and the contact angle method. A comparison can be made between the conformational transitions of single particles in very dilute solutions (where the transitions occur as a single-molecule event without effects of any kind of intermolecular interactions and the solid substrate) with cooperative transition in the dense layer confined by two interfaces and other particles grafted in close proximity to each other. In all solvents, the single unimers possess a core–shell morphology caused by the phase separation of the PS and P2VP arms to avoid unfavorable interactions in the given solvent. The size of the shell increases as solvent selectivity increases. The largest shell was observed upon treatment with pH 2 water where the P2VP arms are strongly stretched because of the electrostatic repulsion. Similar behavior was observed for the grafted monolayers of the particles in selective solvents toluene (for PS) and ethanol or pH 2 water (for P2VP). The PS and P2VP arms in toluene and ethanol, respectively, form nanosize clusters segregated to the unimer cores. That was proven by the XPS and water contact angle data (Table 18.2). Upon exposure to toluene, the contact angle value corresponds to the wetting of PS (90°), which completely occupies the top layer (the lowest fraction of N atoms in the XPS spectra). Upon exposure

Table 18.2 Advancing contact angles and XPS data for the P2VP7-PS7 star–copolymer layer surface after exposure to different solvents. Reprinted with permission from Ref. 44. Copyright 2005 American Chemical Society Solvent

Contact angle(°)

Toluene Chloroform Ethanol H2O, pH 2

91 78 65 15

© 2007, Woodhead Publishing Limited

± ± ± ±

1 2 1 5

Nitrogen, at. % (XPS, 30° angle of incidence) 5.47 7.61 9.20 –

488

Nanofibers and nanotechnology in textiles

to ethanol or water, the contact angle value corresponds to the wetting of P2VP (63°) (in the case of acidic water, the contact angle value corresponds to the wetting of protonated P2VP), and the XPS spectra reveal the largest fraction of N atoms. Thus, in the monolayer each grafted single smart particle responds to outside changes via intramolecular phase segregation of PS, and P2VP arms form the first hierarchical level of response. The cooperative character of the conformational transitions in the densely packed monolayer of smart particles makes up the second level of response (Fig. 18.11). The combination of the smart particles in one layer introduces new kinds of morphology. The core–shell transitions of unimers are transformed into transitions between different phase-segregated morphologies in the grafted monolayer in selective solvents. Clusters of the unfavored polymer are embedded in the layer of extended chains (matrix) of the favored polymer. In other words, the core–shell transitions of unimers are transformed into the interplay between lateral and layered phase segregation. The increase in

(a)

(b)

(c)

18.11 Schematic representation of the switching behavior of the P2VP7-PS7 star–copolymer layer chemically grafted to the surface upon exposure to solvents: (a) selective for P2VP (dashed lines); (b) nonselective; and (c) selective for PS (solid lines). Reprinted with permission from Ref. 44. Copyright 2005 American Chemical Society.

© 2007, Woodhead Publishing Limited

Hybrid polymer nanolayers for surface modification of fibers

489

solvent selectivity results in an increase in upper layer thickness. However, the P2VP7-PS7 unimers possess new specific morphology in the grafted monolayer exposed to chloroform, where PS and P2VP arms segregate to different sides of the core. The result of the phase-segregation mechanism appears as a ripple morphology of the layer. Both the P2VP and PS arms are exposed to the top and form alternating stripes. This may be concluded from the AFM image, XPS and contact angle data. Chloroform is slightly selective for P2VP, and this promotes the formation of the ripple structures. The switching between different morphologies is a reversible process. We observed the transitions more than 10 times upon treatment of the sample with different solvents. It is noteworthy that the morphologies of the layer formed by the densely packed ‘spherical brushes’ are very similar to the morphologies of ‘flat’ mixed brushes.7 In both cases, the response is expressed in the form of morphological transitions between the dimple and ripple segregated phases.

18.6

Future trends

Application of thin hybrid polymer monolayers for coating of smart textiles is in its earliest stages. The long initial period involved development of synthetic approaches and study of basic mechanisms of the responsive behavior. Most of these studies were performed using idealized model systems for theory and experiments. Further development of the field will face the problem of switching into an applied stream of research targeting the important properties of textiles, realistic technology, durability, etc. To this end the smart particle approach seems to be very promising. Development of this approach will effect no substantial changes in the existing technologies for modification of fibers and textile materials. This approach will need intensive investigations for synthetic routes of the fabrication of smart particles. Not all approaches can be automatically transformed from plane substrates onto particle surfaces. A number of problems will increase as an inverse function of the particle size. Another set of problems lies in the appropriate selection of polymers for the synthesis of a responsive shell of the particles. This selection will be strongly connected to the practical applications of the materials. The response kinetics is also a very important issue. The kinetics can be regulated by molecular weight and glass transition temperature of the grafted polymers. Investigations of the smart particles in dynamic conditions will be needed for precise adjustment of the response kinetics. For small particles, if the particle diameter is comparable with the grafted chain dimension, the phase segregation mechanism can be modified by the spherical geometry. This mechanism has not been investigated yet and should be studied carefully. The structure of the particle core can be also involved in designing responsive properties of the smart particles.

© 2007, Woodhead Publishing Limited

490

Nanofibers and nanotechnology in textiles

Further investigations should study correlations between responsiveness of the thin films and properties of textiles and cloth. Although many useful properties of responsive textiles can be imagined, it is important to identify how practically they could be explored for textile materials of different applications. Many different scenarios could be suggested to employ the responsiveness. Some examples are as follows. For a direct switching effect (textile hydrophilic in water and hydrophobic in air) the cloth will adsorb excess perspiration and it can then be easily washed in water, but in a dry state it will be hydrophobic with less friction and less adhesion. For an inverse switching effect (textile hydrophilic in air and hydrophobic in water) the cloth will be permeable for vapors, but an excess of water will block permeability of the materials for aqueous solutions. Experiments should discover which properties will be most attractive and important to consumers.

18.7 • • • •

Sources of further information and advice

Selected chapters on responsive polymer systems1 Responsive thin polymer films2 Polymer brushes15, 16 Polyelectrolyte brushes18

18.8

Acknowledgment

Work described in this review was supported in part by the US ARO grant W911NF-05-1-0339, NTC award C04-CL06 and NSF award CTS 0456548.

18.9

References

1. Minko S., Responsive Polymer Materials: Design and applications, Ames, Blackwell Publishing, 2006. 2. Luzinov I., Minko S., and Tsukruk V. V., ‘Adaptive and responsive surfaces through controlled reorganization of interfacial polymer layers’, Prog Polym Sci, 2004 29(7) 635–698. 3. Roiter Y., and Minko S., ‘Single molecule experiments at the solid–liquid interface: in situ conformation of adsorbed flexible polyelectrolyte chains’, J Am Chem Soc, 2005 127(45) 15688–15689. 4. Holly F. J., and Refojo M. F., ‘Wettability of hydrogels. I. Poly(2-hydroxyethyl methacrylate)’, J Biomed Mater Res, 1975 9 315–326. 5. Ruckenstein E., and Lee S. H., ‘Estimation of the equilibrium surface free-energy components of restructuring solid surfaces’, J Colloid Interface Sci, 1987 120 153– 159. 6. Ruckenstein E., and Lee S. H., ‘Surface restructuring of polymers’, J Colloid Interface Sci, 1987 120 529–536. 7. Lee S. H., and Ruckenstein E., ‘Stability of polymeric surfaces subjected to ultravioletirradiation’, J Colloid Interface Sci, 1987 117 172–178.

© 2007, Woodhead Publishing Limited

Hybrid polymer nanolayers for surface modification of fibers

491

8. Mason R., Jalbert C. A., O’Rourke Muisener P. A. V., Koberstein J. T., Elman J. F., Long T. E., and Gunesin B. Z., ‘Surface energy and surface composition of endfluorinated polystyrene’, Adv Colloid Interface Sci, 2001 94 1–19. 9. Lukas J., Sodhi R. N. S., and Sefton M. V., ‘An XPS study of the surface reorientation of statistical methacrylate copolymers’, J Colloid Interface Sci, 1995 174(2) 421– 427. 10. Russell T. P., ‘Surface-responsive materials’, Science, 2002 297(5583) 964–967. 11. Zhang D., Ward R. S., Shen Y. R., and Somorjai G. A., ‘Environment induced surface structural changes of a polymer: an in situ IR + visible sum-frequency spectroscopic study’, J Phys Chem B, 1997 101 9060–9064. 12. Lewis K. B., and Ratner B. D., ‘Observation of surface rearrangement of polymers using ESCA’, J Colloid Interface Sci, 1993 159(1) 77–85. 13. Senshu K., Sh. Y., Mori H., Ito M., Hirao A., and Nakahama S., ‘Time-resolved surface rearrangements of poly(2-hydroxyethyl methacrylate-block-isoprene) in response to environmental changes’, Langmuir, 1999 15 1754–1762. 14. Milner S. T., ‘Polymer brushes’, Science, 1991 251(4996) 905–914. 15. Zhao B., and Brittain W. J., ‘Polymer brushes: surface-immobilized macromolecules’, Prog Polym Sci, 2000 25(5) 677–710. 16. Advincula R. C., Brittain W. J., Caster K. C., and Ruehe J., Polymer Brushes, Weinheim, Wiley-VCH, 2004. 17. Biesalski M., Johannsmann D., and Ruhe J., ‘Synthesis and swelling behavior of a weak polyacid brush’, J Chem Phys, 2002 117(10) 4988–4994. 18. Ruhe J., Ballauff M., Biesalski M., Dziezok P., Gröhn F., Johannsmann D., Houbenov N., Hugenberg N., Konradi R., Minko S., Motornov M., Netz R. R., Schmidt M., Seidel C., Stamm M., Stephan T., Usov D., and Zhang H., ‘Polyelectrolyte brushes’, in Schmidt, M., Polyelectrolytes with Defined Molecular Architecture I, New York, Adv. Polym. Sci.; Springer, 79–150, 2004. 19. Minko S., Luzinov I., Luchnikov V., Muller M., Patil S., and Stamm M., ‘Bidisperse mixed brushes: Synthesis and study of segregation in selective solvent’, Macromolecules, 2003 36(19) 7268–7279. 20. Minko S., Muller M., Usov D., Scholl A., Froeck C., and Stamm M., ‘Lateral versus perpendicular segregation in mixed polymer brushes’, Phys Rev Lett, 2002 88(3) 0355021–0355024. 21. Shusharina N. P., and Linse P., ‘Oppositely charged polyelectrolytes grafted onto planar surface: mean-field lattice theory’, Eur Phys J E, 2001 6 147–155. 22. Houbenov N., Minko S., and Stamm M., ‘Mixed polyelectrolyte brush from oppositely charged polymers for switching of surface charge and composition in aqueous environment’, Macromolecules, 2003 36(16) 5897–5901. 23. Biesheuvel P. M., and Stuart M. A. C., ‘Electrostatic free energy of weakly charged macromolecules in solution and intermacromolecular complexes consisting of oppositely charged polymers’, Langmuir, 2004 20(7) 2785–2791. 24. Zhulina E., and Balazs A. C., ‘Designing patterned surfaces by grafting Y-shaped copolymers’, Macromolecules, 1996 29(7) 2667–2673. 25. Julthongpiput D., Lin Y. H., Teng J., Zubarev E. R., and Tsukruk V. V., ‘Y-shaped polymer brushes: nanoscale switchable surfaces’, Langmuir, 2003 19(19) 7832– 7836. 26. Granville A. M., Boyes S. G., Akgun B., Foster M. D., and Brittain W. J., ‘Synthesis and characterization of stimuli-responsive semifluorinated polymer brushes prepared by atom transfer radical polymerization’, Macromolecules, 2004 37(8) 2790–2796.

© 2007, Woodhead Publishing Limited

492

Nanofibers and nanotechnology in textiles

27. Zhulina E. B., Singh C., and Balazs A. C., ‘Forming patterned films with tethered diblock copolymers’, Macromolecules, 1996 29(19) 6338–6348. 28. Zhulina E. B., Singh C., and Balazs A. C., ‘Self-assembly of tethered diblocks in selective solvents’, Macromolecules, 1996 29(25) 8254–8259. 29. Zhao B., Brittain W. J., Zhou W., and Cheng S. Z. D., ‘Nanopattern formation from tethered PS-b-PMMA brushes upon treatment with selective solvents’, J Am Chem Soc, 2000 122(10) 2407–2408. 30. Zhao B., Brittain W. J., Zhou W. S., and Cheng S. Z. D., ‘AFM study of tethered polystyrene-b-poly(methyl methacrylate) and polystyrene-b-poly(methyl acrylate) brushes on flat silicate substrates’, Macromolecules, 2000 33(23) 8821–8827. 31. Zhao B., and Brittain W. J., ‘Synthesis, characterization, and properties of tethered polystyrene-b-polyacrylate brushes on flat silicate substrates’, Macromolecules, 2000 33(23) 8813–8820. 32. Boyes S. G., Brittain W. J., Weng X., and Cheng S. Z. D., ‘Synthesis, characterization, and properties of ABA type triblock copolymer brushes of styrene and methyl acrylate prepared by atom transfer radical polymerization’, Macromolecules, 2002 35(13) 4960–4967. 33. Minko S., Patil S., Datsyuk V., Simon F., Eichhorn K. J., Motornov M., Usov D., Tokarev I., and Stamm M., ‘Synthesis of adaptive polymer brushes via “grafting to” approach from melt’, Langmuir, 2002 18(1) 289–296. 34. Sidorenko A., Minko S., Schenk-Meuser K., Duschner H., and Stamm M., ‘Switching of polymer brushes’, Langmuir, 1999 15(24) 8349–8355. 35. Motornov M., Minko S., Eichhorn K. J., Nitschke M., Simon F., and Stamm M., ‘Reversible tuning of wetting behavior of polymer surface with responsive polymer brushes’, Langmuir, 2003 19(19) 8077–8085. 36. Minko S., Muller M., Motornov M., Nitschke M., Grundke K., and Stamm M., ‘Two-level structured self-adaptive surfaces with reversibly tunable properties’, J Am Chem Soc, 2003 125(13) 3896–3900. 37. Öner D., and McCarthy T. J., ‘Ultrahydrophobic surfaces. Effects of topography length scales on wettability’, Langmuir, 2000 16(20) 7777–7782. 38. Bico J., Tordeux C., and Quere D., ‘Rough wetting’, Europhys Lett, 2001 55(2) 214– 220. 39. Wenzel R., ‘Resistance of solid surfaces to wetting by water’, Ind Eng Chem Res, 1936 28 988–994. 40. Cassie A. B., and Baxter S., ‘Wettability of porous surfaces’, Trans Faraday Soc, 1944 40 546–551. 41. Bico J., Marzolin C., and Quere D., ‘Pearl drops’, Europhys Lett, 1999 47(2) 220– 226. 42. Motomov M., Sheparovych, R., Lupitskyy, R., MacWilliams, E., Minko, S., Responsive colloidal systems: Reversible aggregation and fabrication of superhydrophobic surfaces. J Colloid Interface Sci 2007, 310(2), 481–488. 43. Gorodyska G., Kiriy A., Minko S., Tsitsilianis C., and Stamm M., ‘Reconformation and metallization of unimolecular micelles in controlled environment’, Nano Lett, 2003 3(3) 365–368. 44. Lupitskyy R., Roiter Y., Tsitsilianis C., and Minko S., ‘From smart polymer molecules to responsive nanostructured surfaces’, Langmuir, 2005 21(19) 8591–8593.

© 2007, Woodhead Publishing Limited

19 Structure–property relationships of polypropylene nanocomposite fibres C. Y. L E W, University of Oxford, UK and G. M. M C N A L L Y, Queen’s University Belfast, UK

19.1

Introduction

Montecatini (Roder, 1999) first commercialised polypropylene as a synthetic fibre-forming material in 1957, nearly two decades later than the production of polyamide fibre by E. I. du Pont de Nemours & Co. in 1939 (Carraher, 2003) and polyester fibre by ICI in 1941 (Whinfield, 1946). This was because the stereochemistry required to develop the spinnability properties for polypropylene was not achieved until 1954 with the independent discovery of stereospecific titanium halide-based coordination catalysts by Edwin Vandenberg (Vandenberg and Salamone, 1992) and Giulio Natta (Natta et al., 1955). The high molecular weight (Mw) and broad molecular weight distribution (MWD) arising from the heterogeneous Ziegler–Natta polymerization limited the earlier melt spinning of polypropylene to relatively low take-up speeds, usually less than 1000 m/min, due to high viscoelasticity and resultant susceptibility to spinline cohesive fracture (Ziabicki, 1976). Higher-speed spinning of isotactic polypropylene was later made possible with visbreaking technology developed by Exxon, facilitating the reduction of Mw and MWD by extensive chain-scission through twin-screw kneading in the presence of a peroxide species (Steinkamp and Grail, 1975). Recent progress in constrained geometry-catalysed technology using chiral metallocene catalysts has led to the polymerisation of highly stereospecific polypropylenes with even narrower Mw and MWD (Brintzinger et al., 1995), thus further improving spinnability. Studies of propylene random copolymers have recently gained importance over isotactic polypropylene in applications requiring high clarity, flexibility and low-temperature performance (Maier and Calafut, 1998). Traditionally, the enhancement in mechanical performance of melt-spun fibres relies primarily on the control of molecular chain orientation and crystalline structure development through take-up speed, drawing ratio and quenching conditions. Heterogeneous particulate reinforcement of polymers often leads to phase separation, increases the melt viscosity and creates hydrodynamic instabilities, 493

© 2007, Woodhead Publishing Limited

494

Nanofibers and nanotechnology in textiles

which affect the spinline stability and molecular orientation of the molten filament. Several reinforcement techniques have been introduced for the fabrication of composite fibres, such as: (i) the introduction of thermotropic liquid crystalline polymers (TLCP) to produce a matrix–fibril structure, (ii) use of multiphase polymer blends and hard/soft segmented thermoplastics, and (iii) bicomponent extrusion, where different polymers are brought in contact as separate streams just before the spinnerette to produce a sheath–core structure (Salem, 2000). However, the inapplicability of these techniques to highcommodity commercial polymers and other serious drawbacks has limited the appeal. For instance, fabrication of TLCP is very expensive and postprocessing may destroy its unique matrix–fibril structure. Incomplete microphase separation in some polymer blends often leads to a less desirable morphology in multiphase fibres and bicomponent spinning is sensitive to differences in viscosity between the polymers. The use of smectite layered-silicate for reinforcing commercial fibre matrix was first reported in 2000 by Giza et al. (2000a, b) and has recently attracted greater attention, in both academia and industry (Ergungor et al., 2002; Koo et al., 2002; Chang et al. 2003, 2004; McCord et al., 2004; Siochi et al., 2004; Yoon et al., 2004; Zhang et al., 2004). The fabrication of nanocomposite fibres involves exfoliating the layered silicate particles and integrally dispersing them in the host polymer matrix. Nanocomposite systems could potentially offer an overall property improvement without compromising the melt spinnability, flexibility and density of the host matrix. Furthermore, the nanocomposites would preserve a homogeneous polymer-filler phase compared with the heterogeneous morphology obtained in conventionally filled thermoplastic composites. The kinematics and dynamics of deformation that occur in the polymer melt leaving the spinneret during fibre spinning are affected by extensional flow characteristics and the associated elongational viscosity. The relationship between elongational viscosity and shear viscosity for Newtonian fluids defined by Trouton’s rule states that the extensional viscosity is over three times greater than the shear viscosity, while for non-Newtonian polymer melts it is considerably greater (Trouton, 1906). Pavlikova et al. (2003) reported on the enhanced layered-silicate exfoliation through solid-phase drawing of melt-spun polypropylene/hectorite fibres. More recently, Guan et al. (2005) observed that the intercalation of layered-silicate was further improved following the melt-spinning process, which can be related to the strong elongational stress. Lew (2004) reported that exfoliation of layeredsilicate can be promoted by the action of elongational deformation during the melt-drawing process of polypropylene, where the extent of exfoliation increased with drawing speed. McConnell et al. (2006) also observed improved exfoliation of layered-silicate in poly(ethylene-terephthalate) matrix under

© 2007, Woodhead Publishing Limited

Polypropylene nanocomposite fibres

495

the influence of elongational stress; however the degree of exfoliation decreased with increasing melt spinning speed. This chapter aims at exploring in greater detail than hitherto the effect of extensional flow deformation on the exfoliation behaviour of layered-silicate by investigating the structure and properties of a series of polypropylene nanocomposite fibres, obtained through melt-spinning processing, compared with their corresponding compounded feedstocks and injection-moulded specimens. A range of maleic anhydride (MA) oligomers were used to reduce physical phase separation between the hydrophobic matrix and the hydrophilic layered-silicate, and to study the effect of compatibiliser properties on the structure–property relationships of the resultant nanocomposites.

19.2

Materials, processing and characterisation techniques

The host polymer was a metallocene-catalysed propylene–ethylene random copolymer (PP), tradename Metocene X70293, supplied by Basell in pellet form, with a melt-flow index of 40 g/10 min (ASTM D1238). The layeredsilicate was synthetic tetrasilisic fluoromica, tradename SOMASIF MAE, with a cation-exchange capacity of 1.2 meq/g, manufactured by CO-OP Chemical Co. Ltd. The layered-silicate was derived from heat treating talc with alkali silicofluoride followed by isomorphous counterion-exchange with tri-(hexadecyl-octadecyl-octadecenyl)-methyl ammonium cations. A range of MA-grafted isotactic propylene oligomers were used as compatibilisers (Table 19.1). A range of PP/layered-silicate mixtures were prepared by cryo-grinding the PP pellets into powder form, turbo-blending with 3.5 pph fluoromica and thereafter, mixed with predetermined percentages of compatibilisers (Table 19.2). The mixtures were compounded using a Killion KN150 single-screw extruder, fitted with a Davis Standard DSBM-T Spiral Maddock screw. The mixtures were compounded twice, at 15 rpm and 45 rpm screw speed and the barrel temperature profile was ramped from 175 to 200 °C. The low-speed first pass was intended to improve distributive mixing and extend the melt

Table 19.1 Specifications of maleic anhydride-grafted propylene (MAPP) oligomers MAPP

Mw

Mw / M n

Tm (°C)

Acid number (mg KOH/g)

Functionality MAPP/MA (%) (molar ratio)

E43 G3003 G3015 P3150

9100 52 000 47 000 332 000

2.33 1.91 1.89 10.7

156.0 163.8 161.6 164.3

47 8 15 3

8.2 1.4 2.6 0.5

© 2007, Woodhead Publishing Limited

0.32 0.24 0.16 1.57

496

Nanofibers and nanotechnology in textiles Table 19.2 Composition of MAPP and MA in the samples Sample code

MAPP

MA (%)

Final MA (%)

PPEX PPEZ PPG5 PPG3 PPB3

E43 E43 G3015 G3003 P3150

4.17 12.5 12.5 12.5 12.5

0.342 1.025 0.325 0.175 0.063

intercalation time, while the second pass at higher rotation would promote delamination of the intercalated layered-silicate. The pelletised feedstock pellets were melt-spun at a constant shear rate of 300 s–1, at take-up speeds of 80 and 400 m/min, from a 16:1 L/D ratio, 1 mm diameter die, at 160 ºC into single filaments targeted at 150 and 50 µm diameter respectively. Fibres spun at 80 and 400 m/min are referred to as ‘lsv’ and ‘hsv’ respectively. Spinning was performed using a Rosand RH7 capillary rheometer equipped with a digitally controlled high-speed haul-off unit. For comparison, the feedstocks were also injection-moulded into tensile specimens (ASTM D638) using an Arburg 320S Allrounder injection-moulding machine. The injection-moulding temperature and pressure were 180–215 °C and 1200 bar with the mould maintained at 45 °C. The structure of layered-silicates in the nanocomposites was analysed using a Bruker AXS D8 Discover diffractometer (CuKα radiation), scanned at step size of 0.04° and step time of 0.8 s. X-ray diffraction (XRD) was performed on the surface of injection-moulded samples, the compounded feedstock and the melt-spun filament bundles, twisted to about 2 mm in diameter. The morphology of the nanocomposite fibres and the injectionmoulded samples was examined using a digital Philips FEI Technai F20 high-resolution transmission electron microscope (HRTEM), operated at 200 kV. The presented HRTEM images for each of the fibre samples were representative from at least four different ultra-microtomed specimens of the fibre, taken at random sections along the filament. The fibres’ surface textures, associated with spinline stability and cold-fractured surface morphologies related to phase homogeneity, were studied using a JOEL JSM 6400 scanning electron microscope (SEM). Rheological data were recorded using a Rosand RH7 capillary rheometer. Optical birefringence (refractive index) of the fibres was analysed using an Olympus BX50 microscope, equipped with a pair of mutually perpendicular cross-polarisers and a 30 order Berek compensator. The birefringence (double refractive index) ∆n was determined from the ratio of optical retardation Γ over the filament diameter d. The infrared deformation response of fibre molecular bonding was studied using a Perkin-Elmer Spectrum 1000 Fourier transform infrared (FTIR) spectrometer, scanned at resolution of 4 cm–1 for 25 scans on average. The

© 2007, Woodhead Publishing Limited

Polypropylene nanocomposite fibres

497

tensile properties of the fibres were determined using an Instron 4411 Universal extensometer.

19.3

Structure and morphology

Figure 19.1 shows the wide-angle XRD patterns of pristine layered-silicate and its structures in the injection-moulded specimens. The corresponding XRD patterns of the compounded feedstock and melt-spun fibres are shown in Fig. 19.2 and 19.3. The pristine layered-silicate exhibited six characteristic peaks, at 2.9, 4.4, 5.0, 5.8, 7.3 and 10.2°. These characteristic peaks were nearly absent in the injection-moulded, melt-spun and compounded PPEX, thus suggesting that the compatibiliser imparted favourable mixing thermodynamics which promoted exfoliation. PPEX was maleated with a highly functionalized/low Mw, MA oligomer (8.2% functionality, 9100 Mw) at a low concentration (4.2%). High functionality (and hence acid value) of 18 000

16 000

14 000

Relative intensity (a.u.)

12 000

10 000 PPEX 8000

PPEZ

6000

PPG3

4000

PPG5

PPB3

2000

Pristine fluoromica 0 1

3

5

7

9

11

2θ(°)

19.1 Wide-angle XRD patterns of injection-moulded nanocomposites.

© 2007, Woodhead Publishing Limited

498

Nanofibers and nanotechnology in textiles 9000

8000

Relative intensity (a.u.)

7000

6000 PPEX 5000

4000

PPEZ

3000

PPG3

2000 PPG5 1000 PPB3 0 1

3

5

7

9

11

2θ (°)

19.2 Wide-angle X-ray patterns of nanocomposites in feedstock compound form.

an MA oligomer improved the mixing enthalpy (Reichert et al., 2000), while low Mw enhanced its dispersibility and wetting efficiency, therefore creating a favourable condition for exfoliation to occur. In addition, the low concentration of the polar MA compatibiliser minimised physical phase separation with the host PP matrix. In contrast, PPEZ also incorporated the same oligomer, but did not exhibit good exfoliation at a concentration that was three times higher (12.5%). The higher loading of the MA oligomer in PPEZ led to phase separation as observed later in the SEM morphology (Lew, 2004) with the neat PP, or caused a reduction in the optimum shear viscosity required to delaminate the layered-silicate tactoids (Kawasumi et al., 1997; Kim, 2000). Furthermore, Manias et al. (2001) also reported that overloading of the MA oligomer could render a polymer system so robust that the layered-silicate will not mix well with the polymer given that the oligomer would agglomerate around the layered-silicate particles.

© 2007, Woodhead Publishing Limited

Polypropylene nanocomposite fibres

499

7000 lsv hsv 6000

PPEX

Relative intensity (a.u.)

5000 PPEZ 4000

PPG3 3000

PPG5

2000

1000 PPB3

0 1

3

5

7

9

11

2θ (°)

19.3 XRD patterns of melt-spun nanocomposite fibres.

The XRD spectra show the layered-silicate retains four interlayered peaks in the injection-moulded specimens and the compounded feedstock. The higher peak intensity recorded for the injection-moulded samples is attributed to the strong flow orientation effect (McNally et al., 2003). Although the peak reflection patterns for the injection-moulded and compounded samples are very similar, the 2θ peak angles for the feedstock compound were slightly lower than those of the corresponding injection-moulded specimens. In addition, the injection-moulded PPEX displayed three mildly adjoined peaks at 1.54, 1.78 and 2.22° which were not found in the feedstock. These two phenomena suggest shear-induced demixing (Madbouly, 1999), a not so rare phenomenon sometimes seen in a nanoparticle-filled polymer system when subjected to high shear flow (Dennis et al., 2001). In contrast, the XRD spectra for the nanocomposite fibres recorded only two peaks at 2θ position of ~2.2 and 4.4°. The disappearance of the third and fourth order peaks (at 6 to 9° region) further suggest exfoliation of layeredsilicate during the melt-spinning process. This may be associated with the elongational deformation of the the PP matrix during melt-spinning, which

© 2007, Woodhead Publishing Limited

500

Nanofibers and nanotechnology in textiles

can be greater than conventional shear-induced deformation by a factor of up to a hundred (Cogswell, 1975), and therefore draws on a more effective stress to delaminate the layered silicate. Considering the differences in surface characteristics (texture and dimension), the disappearance of the third and fourth order XRD peaks of the fibres could be attributed to the beam scattering effect on the filament due to its surface curvature compared with the flat surface of the injectionmoulded and feedstock compound. Comparisons between the first and second order retained peaks of the fibres suggested that the disappearance of the third and fourth order peaks was attributable in part, if not solely, to improved exfoliation of the layered-silicate, based on the following observations. The 2θ peaks of nanocomposite fibres spun at 400 m/min (hsv) appeared to be more depressed in shape (i.e. greater degree of exfoliation) than their corresponding lsv fibres spun at 80 m/min. Since elongational deformation during spinning of the hsv was greater than lsv because of the higher take-up speed, this would corroborate that the more depressed peaks (exfoliation) observed for hsv could be ascribed to the more effective elongational deformation derived from the higher melt drawing speed. The third and fourth order peaks of the injection-moulded PPB3 appeared slightly more depressed in shape than PPG3, PPG5 and PPEZ. This manifestation may be ascribed to the high MA oligomer concentration with very high Mw (332 000) and low functionality (0.5%). Fornes et al. (2001) extruded nanocomposites using polyamide 6 of different Mw and also observed that a higher matrix Mw, because of its greater interlayer penetration stress, was more effective in exfoliating the layered-silicate. Similarly, the high MA oligomer concentration together with its high Mw in PPB3 may be responsible for the shift of all the peaks to the lower 2θ angles (i.e. enhanced intercalation) observed for the PPB3 feedstock compound.

19.3.1 Morphology Figure 19.4 shows the HRTEM images of the hsv fibres and injection-moulded nanocomposites. Overall, the relative thickness and dispersion of the layeredsilicates in the PP matrices are consistent with the XRD trend. For example, the highly delaminated silicate platelets observed in the PPEX matrix are manifested as a level plateau in its XRD spectrum; an indication of full exfoliation. The observed smaller lateral dimension of layered-silicates in the hsv matrix relative to their corresponding lsv fibres would corroborate the more repressed XRD peaks seen for the hsv fibres. This further suggests the influence of extensional flow deformation on the exfoliation of layeredsilicate. The appearance of tactoid layered-silicate in the PPEZ matrix may be attributed to high-concentration (12.5%) incorporation of low Mw (9100)

© 2007, Woodhead Publishing Limited

Polypropylene nanocomposite fibres PPEX (F)

PPG3 (F)

0.1 µm

0.2 µm

PPEZ (F)

PPB3 (F)

0.2 µm

0.2 µm

PPEX (I)

PPEZ (I)

0.2 µm

0.2 µm

19.4 HRTEM images of hsv fibres (F) and injection-moulded (I) specimens.

© 2007, Woodhead Publishing Limited

501

502

Nanofibers and nanotechnology in textiles

MA oligomer, thus lowering the overall melt viscosity and hence the shear stress to break up the layered-silicate crystallites. Furthermore, the high MA functionality (6.0%) of the oligomer may cause conflux of layered-silicate particles by the compatibiliser species. Two possible consequences may arise as an effect. Envisage the formation of a strong hydrogen bond and van der Waals interaction between the carbonyl (C==O) group and the Si—O/Si— OH groups drawn from excess enthalpic interaction and thus developed into polar globules. These would energise more of the close MA oligomers to populate about the polar localities, therefore reducing the MA dispersibility and leading to a PP/oligomer phase separation. Furthermore, too much enthalpic interaction between the MA and the layered-silicate may reduce the translational entropy and miscibility of the compatibiliser with the interlayer aliphatic tails of the alkylammonium surfactant and, hence, overall energetic conditions for intercalation.

19.4

Phase homogeneity and spinline stability

The cold-fractured surface morphologies of the various fibres, associated with the matrix phase homogeneity are shown by the SEM photomicrographs in Fig. 19.5. The relatively smooth morphology exhibited by PPEX, close to that of the neat PP, is an indication of the presence of significantly exfoliated silicate layers that preserved a homogeneous PP phase. In contrast, PPEZ

PP

PPEX

PPEZ

PPG3

19.5 SEM fracture morphology of lsv fibres (×1000 magnification).

© 2007, Woodhead Publishing Limited

Polypropylene nanocomposite fibres

503

displayed an extremely jagged and micro-cracked surface with strong evidence of clay tactoids, which, as suggested by the XRD and HRTEM results, are attributed to poor exfoliation. PPG3, which as expected, exhibited an intermediate degree of exfoliation between PPEX and PPEZ, displayed a fracture morphology somewhere between that of the two samples. The surface textures of the various melt-spun hsv fibres were associated with their spinline stability. The PPEX fibre featured a smoother circumference than the neat PP fibre, which is usual for a filled polymer system. Given that the development of fibre surface texture is largely dependent on its spinline stability and crystallisation behaviour, this would imply a favourable alteration of the linear viscoelastic response of the PP that had led to better spinning properties and crystallisation attributes. On the contrary, the PPEZ fibre and PPB3 fibre (Lew, 2004) exhibited an uneven and jagged surface texture. Zhang et al. (2004) ascribed these flaws, which they observed in the polypropylene/montmorillonite nanocomposite fibres, to the presence of clay platelets or aggregates that dramatically altered the heterogeneous crystallisation process. This explanation is consistent with the aggregate of tactoids seen in the HRTEM image of PPEZ. In addition, the very high compatibiliser loading (12.5%), as well as low (9100), and high (332 000) Mw in PPEZ and PPB3 respectively would result in a significant broadening of the MWD and an increase in the viscoelasticity of PP melt, therefore becoming more sensitive to melt cohesive fracture. According to Ziabicki (1976), the broadening in MWD and hence elasticity can cause significant change in the heterogeneity of spinning melt, instability of flow within the spinneret channel, hydrodynamic instability of spinline and alteration of crystallisation behaviour, which may be responsible for defects and irregularities on the undrawn fibre surface. In addition, the presence of impurities (tactoids) in PPEZ is quite obviously responsible for the observed knots on its surface. PPG3 and PPG5 (Lew, 2004) which are not shown, displayed an intermediate surface texture between those of PPEX and PPEZ. This would agree with the observed trend since both PPG3 and PPG5 possess an Mw and functional level between those of PPEX, PPEZ and PPB3. It is not uncommon for the presence of bulk nanofiller particles in a conventional polymer system to lead to a reduction in melt spinnability or even provoke the formation of aggregates, owing to instabilities, such as localisation and phase segregation. However, in this work, because the layeredsilicate had already been intercalated and partially exfoliated via compounding prior to melt-spinning, the resultant particles would exhibit an improved aspect ratio and anisotropy. This, in effect, should lead to an enhanced mesoscopic reorientation ability of the silicate platelets in the shear flow direction, which, in turn, could promote the realignment of the polymer chain (Giannelis et al., 1999).

© 2007, Woodhead Publishing Limited

504

Nanofibers and nanotechnology in textiles 8 PPB3 PPG3 PPEZ PP PPG5 PPEX

7

Elongational viscosity (kPa s)

6

5

4

3

2

1

0 100

1000

Shear rate (s–1)

3700

4600

19.6 Plots of elongational viscosity vs. shear rate.

The rheological curves in Fig. 19.6 recorded a significant decrease in the elongational (extensional) viscosity for PPEX. Furthermore, it is intriguing that the PPEX melt would exhibit a Newtonian-like response in the range of shear rates studied, rather than the non-Newtonian behaviour that would be expected for a viscoelastic (polymer) system. From this viewpoint, PPEX, from its Newtonian-like behaviour as well as its adequate melt strength to resist viscoelastic fracture, could notionally be spun at extremely high speeds to achieve a very high tenacity and optical birefringence. The enhanced fibre surface texture of PPEX compared with the virgin PP fibre is another indication of its improved spinning property. This viscosity reduction for PPEX may be attributed to the enhanced plasticising effect of the PP matrix imparted by a nanodispersion effect of the highly exfoliated silicate platelets and alklyammonium species. In a highly exfoliated state, as indicated by its XRD and HRTEM results, contact between the PP matrix and the soft-waxy alkylammonium phase tethered to the silicate layers surface was increased significantly and therefore led to an enhanced plasticising effect. The use of alkylammonium as an alternative to conventional plasticiser to impart melt processing flexibility (i.e. plasticising effect) is well documented (Wittcoff et al., 2003). Xu et al. (2004) reported on the ability of layered-silicate to act as a plasticiser carrier. The increased

© 2007, Woodhead Publishing Limited

Polypropylene nanocomposite fibres

505

volume of silicate layers, and therefore plasticiser carriers attributed to extensive exfoliation in PPEX, could have contributed to the enhancement of the plasticising effect. Although the cause of the plasticising effect may be attributed to grafted alkylammonium species, several authors have reported on the unique effect of ungrafted pristine nanoparticles to plasticise a polymer matrix, leading to viscosity reduction. This is achieved either via a chain slippage effect or through formation of nanosize free volume in the polymer melt (Cho and Paul, 2001; Roberts et al., 2001; McNally et al., 2003; Mackay et al., 2003). However, the enhanced melt spinnability of PPEX is more likely to be associated with the Newtonian-like modification of its melt viscoelastic response and is thus a free melt fracture characteristic. However, the reasons for this radical alteration of viscoelastic behaviour in PPEX are not clear and will be the subject of further investigation. The considerable increase in elongational viscosity recorded for PPB3 is likely to be attributed to the excessive loading (12.5%) of a very high Mw (332 000) compatibiliser, therefore increasing the overall system viscosity. However, the MA oligomer (12.5% E43) used in PPEZ had an Mw (9100) that was significantly lower than that of the neat PP, but also recorded a substantial increase in viscosity. In this case, the increase in viscosity is associated with the presence of silicate tactoids, as observed previously in the HRTEM. The presence of tactoids would restrict the flow of PP melt such as in a conventionally filled polymer system. Kawasumi et al. (1997) reported that excessive loading of an MA oligomer, especially a highly polar one (such as in the PPEZ) in the PP matrix could lead to pronounced phase separation and result in an increase of the polymer melt viscosity, similar to that observed in a heterogeneous system.

19.5

Optical birefringence and infrared activation

The polarised monofilament images and double refractive index measured for lsv fibres melt-spun at lower take-up speed did not exhibit significant variation in birefringence. The observed dark spots along the core of lsv monofilaments would manifest the presence of isotropic type crystal formed through melt nucleating (Buchko et al., 1999). Nucleation of the isotropic crystal is attributed to lower cooling rate associated with lower spin speed and hence more favourable chain relaxation process. A similar crystal feature was also exhibited by the hsv PPEZ, PPG3, PPG5 and PPB3 which comprised a predominantly intercalated layered-silicate structure. However, this feature is not seen for the highly exfoliated hsv PPEX and neat PP fibres, which displayed an intense core brightening associated with enhanced molecular chain conformation. This is confirmed by their considerably higher refractive index plots shown in Fig. 19.7. The absence of isotropic crystal in the exfoliated

© 2007, Woodhead Publishing Limited

506

Nanofibers and nanotechnology in textiles 22

21

Refractive index, ∆n (× 103)

20

19 hsv fibres 400m/min)

18

17

16 lsv fibres (80m/min) 15

14 PP

PPEX

PPEZ

PPG5 PPG3

PPB3

19.7 Plots of optical birefringence.

PPEX compared with the intercalated nanocomposite fibres may be due to enhanced nanodispersion of the silicate platelets. In this respect, despite the fact that the exfoliated silicate platelets may increase point nucleating density, they reduce the volumetric space required for the nuclei to a spherulite growth. In addition, the nanodispersed silicate platelets would lead to a significant volume increase in the silicate layers in the polymer matrix, thus restricting the global self-diffusion ability of the polymer chain. The enhanced birefringence seen for the hsv PPEX fibre may also be because of the anisotropic ordering of silicate platelets. The presence of highly exfoliated silicate platelets would suppress the entropy associated with relaxation of the polymer chain. The higher birefringence of PPEX further corroborated enhanced melt spinnability of the hsv PPEX as indicated by the SEM and elongational viscosity plots. Because the enhancement in birefringence was only recorded for the PPEX fibre spun at higher take-up speed, this strongly suggests the influence of extensional stress deformation on exfoliation of layered-silicate. At an angle of 2θ between 13 and 29° in the fibre XRD spectra (not shown in Fig. 19.1; Lew, 2004), the equatorial reflection of the PP α-crystalline phase for PPEX fibre was greatly suppressed. This observation indicates that the improved PPEX birefringence may also be attributed to a decrease in the isotropic crystalline phase located towards the fibre circumference and an

© 2007, Woodhead Publishing Limited

Polypropylene nanocomposite fibres

507

increase in the growth of the anisotropic ‘shish kebab’ phase (Salem, 2000). It is also possible that the suppressed α-crystalline phase was due to the lateral migration of the highly exfoliated silicate platelets towards the fibre outer region during the spinning process and obstructed the diffraction of the X-ray beam from the PP crystal. Siochi et al. (2004), in their study of polymer carbon nanotubes fibres, reported an increased nanotube concentration from the centre towards outer fibre surface. This phenomenon may be related to the tendency of neutrally buoyant particles to migrate towards an equilibrium position between the wall and flow centre due to wall effects, velocity profile curvature and shear force (Segrè and Silberberg, 1961; Qi et al., 2002). Huang et al. (1997) simulated motion of a single anisotropic circular disc particle in a planar Poiseuille flow using the finite element technique and also demonstrated that, in a neutral case, the Segrè–Silberberg phenomenon was observed in which the equilibrium position for the particle would shift closer to the wall for a higher velocity flow.

19.5.1 Infrared activation Figure 19.8 shows the FTIR spectra of the lsv and hsv fibres. Wavebands of ~1763 and ~1710 cm–1 represent the asymmetric stretching of the hydrogenbonded carbonyl group (C=O) of the cyclic anhydride and the hydrolysed maleic acid, which are reversible on addition or removal of water (Barra et al., 1999; Bettini and Aganelli, 2000; Premphet and Chalearmthitipa, 2001). Two registered bands at 969 and 99 cm–1 are attributed to the vibrational stretching of the helical C—C backbone of the PP in the amorphous region (Verleye et al., 2001). The 1039 cm–1 band refers to both the symmetric stretching of Si—O and also deformation of PP, while the 1353 and 1371 cm–1 bands are assigned to the in-plane bending vibration of CH2 and CH3 bonds. The 1069 cm–1 band results from stretching of the Si—O bond in the tetrahedral coordinated silica plane (Gilman et al., 1999; Kang et al., 2000; Yeh et al., 2002; Kim et al., 2003). The spectra show a unique IR activation response for the PPEX fibre, which is very different from IR patterns exhibited by PPEZ, PPG3, PPG5 and PPB3. PPEX exhibited a waveband pattern that was consistent with those of the neat PP spectra as if the layered-silicate were absent in the matrix. This would suggest that PPEX, because of its highly exfoliated layered-silicate and uniform dispersion, had developed a homogeneous PP/layered-silicate phase, whereas the less exfoliated samples had yielded a more heterogeneous phase. Firstly, the 1069 cm–1 Si—O band and the 1003 cm–1 band are absent in the nanodispersed PPEX and neat PP, but are visible in other systems. Secondly, no alteration was registered for the 1039 cm–1 band of PPEX and PP, but it rose in intensity and merged with the 1003 cm–1 band in other, less homogeneous, systems that exhibited evidence of clay tactoids. Rzanek-

© 2007, Woodhead Publishing Limited

508

Nanofibers and nanotechnology in textiles

Relative absorbance (a.u.)

1710

1003

lsv (80 m/min)

PPB3

1096

PPEX

993 969

1763

PPG3

1039

1069

PPG5

PP 3100

2900

2700

2500

1750

1500

1250 Wavelength (cm–1)

1500

1250 Wavelength (cm–1)

1000

750

hsv (400 m/min)

Relative absorbance (a.u.)

PPB3 PPG5

1096

1039

PPEX

969

1710

PPEZ

993

1763

PPG3

PP

3100

2900

2700

2500

1750

1000

750

19.8 FTIR spectra of the lsv and hsv fibres.

Boroch et al. (2002) reported a similar observation of the 1039 cm–1 Si—O IR absorption band in their study of a range of organosilicon deposited film’s. When the film’s surface was further characterised using atomic force microscopy, they observed that the intensity of the Si—O IR band was weaker for films with improved phase homogeneity and was stronger for films with more heterogeneous surface deposition, exhibiting evidence of fine organosilicon aggregates. The manifestation of the 1003 cm–1 Si—O vibration band for predominantly phase-intercalated PPEZ, PPG3, PPG5 and PPB3 may be ascribed to a systematic distortion of the tetrahedral symmetry silicate plane structure by dipolar interaction between the Si—O or Si—OH and carbonyl group of the anhydride

© 2007, Woodhead Publishing Limited

Polypropylene nanocomposite fibres

509

species, or with the alkylammonium cations (NH δ4 + ) . Peak et al. (1999) previously demonstrated, in the absorption study of a goethite-based clay system, that increasing the absorption of protonated surfactants (e.g. cationic alkylammonium) would render a more severe distortion effect on the tetrahedral SiO4 symmetry due to an electrostatic effect, causing a split in their v3 (asymmetric stretch) absorption band and triggering activation of the usually IR-inert v1 bond vibration. The absence of this 1003 cm–1 band in PPEX therefore can be associated with its homogeneous exfoliated layered-silicate phase, where the effect of electrostatic distortion by onium cations was significantly reduced compared with other matrix/onium intercalated layered-silicate systems.

19.6

Crystallisation behaviour and mechanical performance

Figure 19.9 shows the endothermic differential scanning calorimetry (DSC) traces of the various fibres, with their corresponding peak melting temperature, the principal and secondary crystalline enthalpy of fusion attributed to melting of the host PP copolymer and MA oligomer respectively. The DSC traces of the PPG3, PPG5 and PPB3 fibres demonstrated two fusion peaks. In contrast, the second endotherm ascribed to the compatibiliser phase is absent in the PPEX fibre which exhibited DSC traces consistent with that of the virgin PP fibre consisting of a homogeneous phase. The absence of the second endotherm peak in PPEZ, on the other hand, may be ascribed to inhibition of nucleation activity or crystal growth by the highly MA-grafted low Mw oligomer incorporated where the pendant anhydride branching would interfere with, and restrict, the lamellae folding process. In contrast to most reports on the nucleating effect of layered silicates that often find an increase in the PP crystallinity of their nanocomposite systems (Ma, 2001; Xu et al., 2002; Li et al., 2003), the crystalline portion of the nanocomposite fibres, except for PPEX, was considerably lower than the neat PP and the magnitude of reduction was more significant for fibres exhibiting less exfoliated morphology. In general, this reduction in the crystalline phase may be associated with an anti-nucleation effect of the layered-silicate, as implied from its suppressed melt crystallisation temperature (Lew, 2004). In addition, the increase in viscosity measured for PPEZ, PPG3, PPG5 and PPB3 due to the presence of predominantly intercalated layeredsilicate phase may cause an increase in the diffusional activation energy of the PP chains, thus reducing their crystallinity. The nucleation rate of a polymer is the product of a mass transport and nucleation parameter (Avrami, 1939), where the transport factor is related to a chain’s self-diffusion; therefore the reduction in crystalline phase and crystallisation temperature could be explained by the coupling of functional carbonyl anhydride molecules with the reactive silanol sites on the silicate

© 2007, Woodhead Publishing Limited

510

Nanofibers and nanotechnology in textiles 16 hsv lsv 14 78.5 J/g 61.1 J/g 12

139.0 136.8

10 Heat flow (mJ/s)

137.7 °C 136.4 °C

75.2 J/g 56.6 J/g

PP

138.7 138.5

8

59.5 J/g 53.3 J/g

PPEX

47.7 J/g 35.6 J/g

PPEZ

6

53.4 J/g 39.2 J/g

PPG3 4

136.8 135.3 137.3 135.6

49.9 J/g 41.7 J/g

PPG5

137.0 135.5

7.5 J/g 6.8 J/g 3.5 J/g 4.3 J/g

4.0 J/g 5.3 J/g

2 PPB3 0 10

40

70

100 130 Temperature (°C)

160

190

220

19.9 DSC endothermic traces of nanocomposite fibres.

surface, thus inhibiting lamellae formation for the compatibiliser. In addition, nanodispersion of exfoliated silicate platelets, owing to their larger surface area, may constrain the chain mobility and hence global self-diffusion of the host PP matrix. Lew et al. (2004) and Gopakumar et al. (2002) observed a similar reduction in the crystallinity of polyethylene layered-silicate nanocomposites incorporating MA oligomer and ascribed it to restriction in polymer matrix mobility through association with exfoliated layered-silicate or pendant anhydride. The higher melting enthalpy measured for PPEX could be due to increased formation of the anisotropic ‘shish kebab’ phase as can be seen from the previous birefringence results. However, owing to the very large specific area of the exfoliated silicate platelets, the entropy of the melt matrix to initiate nucleation was greatly constrained and hence explained for the lower quiescent crystallisation temperature (Lew, 2004). In contrast to

© 2007, Woodhead Publishing Limited

Polypropylene nanocomposite fibres

511

the melt-spun nanocomposite fibres, the influence of the layered-silicate on the crystallinity of injection-moulded nanocomposites was not as significant; recorded melt enthalpy values were: PP (69.2 J/g), PPEX (66.8 J/g), PPEZ (69.5 J/g), PPG3 (70.1 J/g), PPG5 (67.6 J/g) and PPB3 (76.1 J/g). The melting enthalpy, associated with crystallinity for hsv fibres spun at greater take-up speed, despite registering lower XRD peak intensity (associated with lower fraction of crystalline phase), was significantly higher than for their corresponding lsv fibre. Given that the hsv fibre exhibited considerably higher birefringence, the higher melting enthalpy measured for hsv fibres may be ascribed to higher thermal energy required for the melting of the shish kebab phase. The slightly higher peak melting temperature registered for the hsv fibres relative to their corresponding lsv fibres may be attributed to the greater activation energy required to induce mobility to the better oriented core crystalline structure.

19.6.1 Mechanical performance Figure 19.10 shows the tensile properties graphs of the various hsv fibres, melt-spun at 400 m/min. The graphs for lsv fibres, melt spun at 80 m/min Tensile strength (MPa) Young’s modulus (× 10 MPa) Yield stress (MPa) Strain at break (%)

500

400

300

200

100

0 PP

PPEX

PPEZ

PPG5

19.10 Mechanical properties of hsv fibres.

© 2007, Woodhead Publishing Limited

PPG3

PPB3

512

Nanofibers and nanotechnology in textiles

(not shown here; Lew, 2004), exhibited lower values for all the corresponding properties compared with the hsv, except having significantly higher strain at break. Because the lsv fibres were spun at lower take-up speed, the degree of extensional flow stress deformation and hence chain orientation was lower. As a result, lsv will therefore contain a larger percentage of (stretchable) amorphous phase and isotropic crystal line phase compared with the more highly aligned ‘shish’ core crystalline phase in the hsv. Substantial correlation is observed between the previous birefringence study and the tensile properties of the fibre samples, especially the tensile strength, which appeared to be dependent on the birefringence. For instance, PPEX, which displayed the highest birefringence value, also recorded the highest tensile strength and the neat PP exhibited second highest tensile strength for a similar reason. All other nanocomposite fibres exhibited tensile strengths lower than the neat PP fibre. The very poor tensile strength recorded for PPEZ is attributed to the presence of layered-silicate tactoids which could accelerate the rate of void propagation. Despite that PPG3, PPG5 and PPB3 exhibited greater improvement than PPEX in some of the tensile performances, PPEX alone exhibited a combination of overall improvement in tensile strength, Young’s modulus, yield stress and stress at break. The substantial increase in stress at break for PPB3 on the other hand is attributed to the presence of the very high Mw (332 000) and broad MWD compatibiliser. Kim et al. (2001) proposed that the tensile deformation mechanism of injection-moulded nanocomposites consists of three main stages. In brief, the first two stages comprise microvoid formation, owing to debonding of the intercalated matrix, followed by a splitting and realignment of layeredsilicates. The criticality of the first and second deformation stages is less significant for PPEX fibre because the layered-silicates were already highly exfoliated as shown by the XRD and HRTEM results, and already oriented in the fibre axis (flow) direction. The third deformation stage could be adapted to explain the observed overall improvement in tensile properties for the PPEX fibre. During the third tensile deformation stage, silicate platelets bonded to the ‘shish’ fibrillar core and the overgrowth lamellae ‘kebab’ would interfere with the chain slip and fragmentation process. As the tensile deformation continued, the silicate platelets resist stress-induced recrystallisation of the disintegrated fibrils and the matrix/silicate interaction relaxes the stress imposed on the PP chain, while the soft alkylammonium tethered to the silicate platelets would impart a plasticising effect, thus reducing the slip friction between matrices and prolonging the fibre rupturing process. The greater break strain recorded for the PPG3, PPG5 and PPB3 fibres may be attributed to the second deformation stage, which allows for splitting and reorientation of the intercalated layered-silicate stacks as described. The higher amounts of microvoids created during this stage would lead to an improved toughening mechanism (McNally et al., 2003).

© 2007, Woodhead Publishing Limited

Polypropylene nanocomposite fibres

19.7

513

Exfoliation by extensional flow deformation

Fornes et al. (2001) proposed that the layered-silicate exfoliation during an extrusion process was initiated by the break-up of taller silicate stacks into smaller stacks, following a layer-by-layer peeling mechanism of the top and bottom silicate stack platelets by the extrusion shear stress. They also reported that this mechanism was proportional to the polymer Mw (viscosity), in which higher effective shear stress associated with higher Mw would lead to a more exfoliated morphology. Contrary to conventional shear deformation via an extrusion process, the elongational stress deformation associated to the spin speed in fibre extrusion can, in many instances, being inversely proportional to the Mw and MWD. This is because, although higher Mw polymers will produce higher spinline tension associated with the elongational viscosity, the molten filament is also more susceptible to spinline fracture and thus less spinnable. On the other hand, lower Mw polymers, providing they possess sufficient melt strength, are more spinnable at much higher take-up speeds and hence would experience greater elongational stress and chain alignment. A recent schema proposed to describe the exfoliation mechanism of layered-silicate under the influence of extensional flow stress deformation can be found in the literature (Lew, 2004). Correlations drawn from the experimental results have led to the postulation that melt-spinning and hence extensional flow stress per se will not evoke exfoliation of layered-silicate in a pristine state. However, interestingly, meltspinning is found to greatly promote exfoliation and nanodispersion of a precedently intercalated layered-silicate. This postulation is in part based on the XRD and HRTEM results and further corroborated by the following observations and contentions. Firstly, the two peaks attributed to the third and fourth orders of basal spacing of the layered-silicate seen in the XRD spectra for the injectionmoulded specimens and feedstock compounds were absent in the fibre spectra. Because the disappearance of the third and fourth order peaks could be an effect of exfoliation or because of low X-ray beam intensity, it is not possible to distinguish which of the two effects is more prevalent. However, the first and second order peaks of the hsv fibres manifested a more depressed peak shape and exhibited a lower peak intensity compared with the lsv fibres, but, in theory, the lsv fibre should exhibit a greater peak intensity because of the larger circumference and hence area exposed for beam diffraction. It is quite apparent, therefore, that the disappearance of the third and fourth order peaks, and the suppressed peak shape and intensity for the first and second order peaks, are mainly attributed to further exfoliation of the precedently intercalated layered-silicate through an extensional deformation effect during the melt-spinning process.

© 2007, Woodhead Publishing Limited

514

Nanofibers and nanotechnology in textiles

Secondly, because the XRD spectra of PPEX did not show any peak in the compounded feedstock form, its absence in the fibre spectra could not be taken as convincing evidence of layered-silicate exfoliation associated with the melt-spinning. However, given that the HRTEM images of PPEX fibre have manifested layered-silicates of thinner lateral dimension and more uniform dispersion than the injection-moulded PPEX, this would imply the occurrence of exfoliation during the melt-spinning process. Finally, the larger lateral dimension of layered-silicates observed in the feedstock compounds and the injection-moulded samples compared with the corresponding melt-spun fibres would confirm that the extensional flow stress associated with melt-spinning had indeed extended the exfoliation of the layered-silicate. The above analyses would lead to the following proposal of the exfoliation mechanism of layered-silicate during the melt-spinning process and its schematic diagram is given in the literature (Lew, 2004). • The coiled polymer chains intercalated in the layered-silicate disentangle, enabling the layered-silicates to begin to realign themselves in the flow direction. The polymer chains residing in the intercalated layered-silicate spacing will also be oriented in the flow direction. • Drawdown of the melt stretches the polymer chains and promotes further penetration of the co-entangled polymer chains into the layered-silicate spacing. • Under the very high elongational deformation rates, the silicate interlayers begin to shear apart and are subsequently delaminated by the polymer chains that reside in the silicate interlayers. • The exfoliated silicate platelets improve the slippage of the polymer chains during drawdown from the capillary, enhance the melt stability and result in highly regular chain orientation in the spinline direction. Depending on the extent of exfoliation, the exfoliated silicate platelets preserve the chain alignment by suppressing molecular relaxation in the fibre traverse axis, leading to an overall improvement in the birefringence and mechanical performance of the fibre, as shown by the PPEX fibre.

19.8

Conclusions

Propylene–ethylene copolymer nanocomposites, in fibre or injection-moulded form, were produced from precompounded PP/organoclay feedstocks, maleated with a range of isotactic propylene MA oligomers of different Mw, MWD and functionality. The XRD and HRTEM results revealed that a highly nanodispersed and exfoliated layered-silicate phase was obtained when the PP was loaded, with a very low Mw, low MWD and highly functionalised oligomer at low/optimum concentration (4.2%). The low Mw would induce a favourable wetting property and ease of dispersion in the PP melt while

© 2007, Woodhead Publishing Limited

Polypropylene nanocomposite fibres

515

high functionality improves enthalpic interaction with the layered-silicate. Low MWD reduces the tendency to viscoelastic fracture during the spinning process and low concentration is important to prevent phase separation with the PP matrix. Comparison of the compounded nanocomposite feedstocks and melt-spun fibres revealed the action of elongational stress on the PP matrix during the melt-spinning process would extend exfoliation of already intercalated layeredsilicate and enhances its nanodispersion, similar to observations by Guan et al. (2005). This is further corroborated by the SEM, birefringence and rheological analyses in which the PPEX nanocomposite exhibited more desirable fibre features than the virgin PP fibre, characterised by its enhanced melt spinnability, surface texture, birefringence and mechanical properties, while maintaining a homogeneous PP phase. PPEX melt had demonstrated a Newtonian-like modification which is unusual for a polymer system, given that the PP is viscoelastic and a viscoelastic system should develop a non-Newtonian flow behaviour. Therefore, this Newtonian-like modification of the PPEX melt is most likely to be attributed to a unique molecular level interaction between the PP matrix and the exfoliated silicate platelets. This could potentially lead to infinite increase of the meltspin speed and hence the fibre tenacity.

19.9

References

Avrami, M. (1939), ‘Kinetics of phase change. I general theory’, J Chem Phys, 7 (12), 1103–1112. Barra, G. M. O., Crespo, J. S., Bertolino, J. R., Soldi, V., Pires, A. T. N. (1999), ‘Maleic anhydride grafting on EPDM: Qualitative and quantitative determination’, J Braz Chem Soc, 10 (1), 31–34. Bettini, S. H. P., Agnelli, J. A. M. (2000), ‘Evaluation of methods used for analysing maleic anhydride grafted onto polypropylene by reactive processing’, Polym Testing, 19 (1), 3015. Brintzinger, H. H., Fisher, D., Mulhaupt, R., Rieger, B., Waymouth, R. M. (1995), ‘Stereospecific olefin polymerization with chiral metallocene catalysts’, Angew Chem Int Ed, 34 (11), 1143–1170. Buchko, C. J., Chen, L. C., Shen, Y., Martin, D. C. (1999), ‘Processing and microstructural characterization of porous biocompatible protein polymer thin films’, Polymer, 40 (26), 7393–7407. Carraher, C. E. (2003), Giant Molecules: Essential materials for everyday living and problem solving, 2nd edn, John Wiley & Sons Inc. Chang, J. H., An, Y. U., Kim, S. J., Im, S. (2003), ‘Poly(butylene terephthalate)/organoclay nanocomposites prepared by in situ interlayer polymerization and its fiber (II)’, Polymer, 44 (19), 5655–5661. Chang, J. H., Kim, S. J., Joo, Y. L., Im, S. (2004), ‘Poly(ethylene terephthalate) nanocomposites by in situ interlayer polymerization: the thermo-mechanical properties and morphology of the hybrid fibers’, Polymer, 45 (3), 919–926.

© 2007, Woodhead Publishing Limited

516

Nanofibers and nanotechnology in textiles

Cho, J. W., Paul, D. R. (2001), ‘Nylon 6 nanocomposites by melt compounding’, Polymer, 42 (3), 1083–1094. Cogswell, F. N. (1975), ‘Polymer melt rheology during elongational flow’, in White, J. L. Fiber and Yarn Processing, Appl Polym Symp No. 27, John Wiley & Sons Inc. Dennis, H. R., Hunter, D. L., Chang, D., Kim, S., White, J. L., Cho, J. W., Paul, D. R. (2001), ‘Effect of melt processing conditions on the extent of exfoliation in organoclaybased nanocomposites’, Polymer, 42 (23), 9513–9522. Ergungor, Z., Cakmak, M., Barut, C. (2002), ‘Effect of processing conditions on the development of morphology in clay nanoparticle filled nylon 6 fibers’, Macromol Symp, 185 (1), 259–276. Fornes, T. D., Yoon, P. J., Keskkula, H., Paul, D. R. (2001), ‘Nylon 6 nanocomposites: the effect of matrix molecular weight’, Polymer, 42 (25), 9929–9940. Giannelis, E. P., Krishnamoorti, R., Manias, E. (1999), ‘Polymer-silicate nanocomposites: model systems for confined polymers and polymer brushes’, Adv Polym Sci, 138, 107–147. Gilman, J. W., Kashiwagi, T., Nyden, M., Harris, R. H. Jr. (1999), New Flame Retardants Consortium Final Report, NISTIR 6357, NIST, pp. 1–26. Giza, E., Ito, H., Kikutani, T., Okui, N. (2000a), ‘Fiber structure formation in high-speed melt spinning of polyamide 6/clay hybrid nanocomposites’, J Macromol Sci – Phys, 39 B (4), 545–559. Giza, E., Ito, H., Kikutani, T., Okui, N. (2000b), ‘Structural control of polyamide 6/clay nanocomposite fibers by in-line drawing process’, J Polym Eng, 20 (6), 403–425. Gopakumar, T. G., Lee, J. A., Kontopoulou, M., Parent, J. S. (2002), ‘Influence of clay exfoliation on the physical properties of montmorillonite/polyethylene composites’, Polymer, 43 (20), 5483–5491. Guan, G. H., Li, C. C., Zhang, D. (2005), ‘Spinning and properties of poly(ethylene terephthalate)/organomontmorillonite nanocomposite fibers’, J Appl Polym Sci, 95 (6), 1443–1447. Huang, P. Y., Feng, J., Hu, H. H., Joseph, D. D. (1997), ‘Direct simulation of the motion of solid particles in Couette and Poiseuille flows of viscoelastic fluids’, J Fluid Mech, 343, 73–94. Kang, S., Kim, J., Park, M. Hong, S. I. (2000), ‘Effects of surface modification of silicas on the properties of organic/inorganic nanocomposites’, Polym Preprints, 41 (1), 596–597. Kawasumi, M., Hasegawa, N., Kato, M., Usuki, A., Okada, A. (1997), ‘Preparation and mechanical properties of polypropylene–clay hybrids’, Macromolecules, 30 (20), 6333– 6338. Kim, G. M., Lee, D. H., Hoffmann, B., Kressler, J., Stöppelmann, G. (2001), ‘Influence of nanofillers on the deformation process in layered silicate polyamide-12 nanocomposites’, Polymer, 42 (3), 1095–1100 Kim, K. N., Kim, H. S., Lee, J. W. (2000), ‘Mixing characteristics and mechanical properties of polypropylene–clay composites’, Proc ANTEC 2000, 3, 3872–3876. Kim, T. H., Lim, S. T., Lee, C. H., Choi, H. J., Jhon, M. S. (2003), ‘Preparation and rheological characterization of intercalated polystyrene/organophilic montmorillonite nanocomposite’, J Appl Polym Sci, 87 (13), 2106–2112. Koo, C. M., Ham, H. T., Kim, S. O., Wang, K. Y., Chang, I. J. (2002), ‘Morphology evolution and anisotropic phase formation of the maleated polyethylene-layered silicate nanocomposites’, Macromolecules, 35 (13), 5116–5122.

© 2007, Woodhead Publishing Limited

Polypropylene nanocomposite fibres

517

Lew, C. Y. (2004), Polymer-clay nanocomposites: preparation, processing and properties, Doctoral Thesis, Queen’s University Belfast. Lew, C. Y., Murphy, W. R., McNally, G. M. (2004), ‘Preparation and properties of polyolefin-clay nanocomposites’, Polym Eng Sci, 44 (6), 1027–1035. Li, J., Zhou, C., Gang, W. (2003), ‘Study on nonisothermal crystallization of maleic anhydride grafted polypropylene/montmorillonite nanocomposite’, Polym Testing, 22 (2), 217–223. Ma, J., Qi, Z., Hu, Y. (2001), ‘Synthesis and characterization of polypropylene/clay nanocomposites’, J Appl Polym Sci, 82 (14), 3611–3617. Mackay, M. E., Dao, T. T., Tuteja, A., Ho, D. L., van Horn, B., Kim, H. C., Hawker, C. J. (2003), ‘Nanoscale effects leading to non-Einstein-like decrease in viscosity’, Nature Mater, 2, 762–766. Madbouly, S., Ohmomo, M., Ougizawa, T., Inoue, T. (1999), ‘Effect of the shear flow on the phase behaviour of polystyrene/poly(vinyl methyl ether) blend’, Polymer, 40 (6), 1465–1472. Maier, C., Calafut, T. (1998), Polypropylene. The definitive user’s guide and databook, Plastics Design Library. Manias, E., Touny, A., Wu, L., Strawhecker, K., Lu, B., Chung, T. C. (2001), ‘Polypropylene/ montmorillonite nanocomposites. Review of the synthetic routes and materials properties’, Chem Mater, 13 (10), 3516–3523. McConnell, D., Hornsby, P. R., Lew, C. Y., Qua, E. H. (2006), ‘Structure-property of PET nanocomposite fibres’, ANTEC 2006, accepted. McCord, M. G., Matthews, S. R., Hudson, S. M. (2004), ‘Extrusion and analysis of nylon/montmorillonite nanocomposite filaments’, J Adv Mater, 36 (1), 44–56. McNally, T., Murphy, W. R., Lew, C. Y., McNally, G. M., Turner, R. J., Brennan, G. P. (2003), ‘Polyamide-12 layered silicate nanocomposites by melt blending’, Polymer, 44 (9), 2761–2772. Natta, G., Pino, P., Corradini, P., Danusso, F., Mantica, E., Mazzanti, G., Moranglio, G. (1955), ‘Crystalline high polymers of α-olefin’, J Am Chem Soc, 77 (6), 1708–1710. Pavlikova, S., Thomann, R., Reichert, P., Mülhaupt, R., Marcincin, A., Borsig, E. J. (2003), ‘Fiber spinning from poly(propylene)–organoclay nanocomposite’, J Appl Polym Sci, 89 (3), 604–611. Peak, D., Ford, R. G., Sparks, D. L. (1999), ‘An in situ ATR-FTIR investigation of sulfate bonding mechanisms on goethite’, J Coll Interface Sci, 218 (1), 289–299. Premphet, K., Chalearmthitipa, S. (2001), ‘Melt grafting of maleic anhydride onto elastomeric ethylene-octene copolymer by reactive extrusion’, Polym Eng Sci, 41 (11), 1978–1986. Qi, D. W., Luo, L. S., Aravamuthan, R., Strieder, W. (2002), ‘Lateral migration and orientation of elliptical particles in Poiseuille flows’, J Statistical Phys, 107 (1/2), 101–120. Reichert, R., Nitz, H., Klinke, S., Brandsch, R., Thomann, R., Mülhaupt, R. (2000), ‘Poly(propylene)/organoclay nanocomposite formation: Influence of compatibilizer functionality and organoclay modification’, Macromol Mater Eng, 275 (1), 8–17. Roberts, C., Cosgrove, T., Schmidt, R. G., Gordon, G. V. (2001), ‘Diffusion of poly(dimethylsiloxane) mixtures with silicate nanoparticles’, Macromolecules, 34 (3), 538–543. Roder, H. (1999), ‘Polypropylene – a material of the future’, Prog Polym Sci, 24 (8), 1205–1215. Rzanek-Boroch, Z., Schmidt-Szalowski, K., Janowska, J., Dudzin´ski, K., Szyman´ska, A.,

© 2007, Woodhead Publishing Limited

518

Nanofibers and nanotechnology in textiles

Misiak, M. (2002), ‘Thin films obtained by PE-CVD method from hexamethyldisiloxane under atmospheric pressure’, Proc International Synposium on High Pressure Low Temperature Plasma Chemistry, Pühajärve, Estonia, 415–419. Salem, D. R. (2000), Structure Formation in Polymeric Fibres, Hanser Gardner Publications Inc. Segrè, G., Silberberg, A. (1961), ‘Radial Poiseuille flows of suspensions’, Nature, 189, 209. Siochi, E. J., Working, D. C., Park, C., Lillehei, P. T., Rouse, J. H., Topping, C. C., Bhattacharyya, A. R. and Kumar, S. (2004), ‘Melt processing of SWCNT–polyamide nanocomposite fibers’, Composites Part B: Engineering, 35 (5), 439–446. Steinkamp, R. A., Grail, T. J. (1975), US Patent no. 3 863 265. Trouton, F. T. (1906), Proc Royal Soc, 77 (A), 426. Vandenberg, E. J., Salamone, J. C. (1992), Catalysis in Polymer Synthesis, Oxford University Press. Verleye, G. A. L., Roeges, N. P. G., Moor, M. O. D. (2001) Easy Identification of Plastics and Rubbers, Rapra Technology Ltd, Chapter 3, pp. 28–29. Whinfield, J. R. (1946), Nature, 158, 930. Wittcoff, H. A., Reuben, B. G., Plotkin, J. S. (2003), Industrial Organic Chemicals, 2nd edn, John Wiley & Sons Inc. Xu, W., Ge, M., He, P. (2002), ‘Nonisothermal crystallization kinetics of polypropylene/ montmorillonite nanocomposites’, J Polym Sci, Part B: Polym Phys, 40 (5), 408–414. Xu, W. B., Zhou, Z. F., Ge, M. L., Pan, W. P. (2004), ‘Polyvinyl chloride/ montmorillonite nanocomposites: Glass transition temperature and mechanical properties’, J Thermal Anal Calorimetry, 78 (1), 91–99. Yeh, J. M., Chen, C. L., Chen, Y. C., Ma, C. Y., Lee, K. R., Wei, Y., Li, S. (2002), ‘Enhancement of corrosion protection effect of poly(o-ethoxyaniline) via the formation of poly(o-thoxyaniline)–clay nanocomposite materials’, Polymer, 43 (9), 81–86. Yoon, K. H., Polk, M. B., Min, B. G., Schiraldi, D. A. (2004), ‘Structure and property study of nylon-6/clay nanocomposite fiber’, Polym Int, 53 (12), 2072–2078. Zhang, X., Yang, M., Zhao, Y., Zhang, S., Dong, X., Liu, X., Wang, D. and Xu, D. (2004), ‘Polypropylene/montmorillonite composites and their application in hybrid fiber preparation by melt-spinning’, J Appl Polym Sci, 92 (1), 552–558. Ziabicki, A. (1976), Fundamentals of Fibre Formation, John Wiley & Sons.

© 2007, Woodhead Publishing Limited